Datasets:

ccdv

/

govreport-summarization

like 34

IHS was established within the Public Health Service in 1955 to provide health services to members of AI/AN tribes primarily in rural areas on or near reservations. IHS provides these services directly through a network of hospitals, clinics, and health stations operated by IHS, and it also funds services provided at tribally operated IHS facilities. The federally operated system comprises 26 hospitals, 56 health centers, and 32 health stations in 33 states and received over 5 million outpatient visits and approximately 19,000 admissions in 2014. (See table 1.) Federally operated IHS hospitals range in size from 4 to 133 beds and are open 24 hours a day for emergency care needs. Health centers offer a range of care, including primary care services and at least some ancillary services, such as pharmacy, laboratory, and X-ray services, and they are open for at least 40 hours a week. Health stations offer only primary care services on a regularly scheduled basis and are open fewer than 40 hours a week. IHS oversees its health care facilities through a decentralized system of area offices, which are led by area directors and located in 12 geographic areas. (See fig. 1 for a U.S. map showing the IHS patient population by area). Nine of these 12 IHS areas have federally operated IHS facilities— Albuquerque, Bemidji, Billings, Great Plains, Nashville, Navajo, Oklahoma City, Phoenix, and Portland. According to IHS, the headquarters office is responsible for setting health care policy, ensuring the delivery of quality comprehensive health services, and advocating for the health needs and concerns of AI/AN people. The IHS area offices are responsible for distributing funds to the facilities in their areas, monitoring their operation, and providing guidance and technical assistance. (See fig. 2). According to IHS, its mission to raise the physical, mental, social, and spiritual health of AI/AN people to the highest level cascades through every organizational level and individual in the agency. This cascading method of accountability is often used by health care organizations with a decentralized management structure and is recommended by the Office of Personnel Management (OPM) for agencies with clear organizational goals and objectives. In addition, in 2009 IHS developed four agency- wide priorities that serve as a strategic framework for improvement within the agency. One of these four priorities is to improve the quality of and access to care. In 2016, IHS established the following revised agency-wide priorities: assess care, improve delivery of services, address behavioral health issues, strengthen management, bring health care quality expertise to IHS, and engage local resources. IHS’s oversight of the quality of care provided in its federally operated facilities has been limited and inconsistent. While some oversight functions are performed at the headquarters level, the agency has delegated primary responsibility for the oversight of the quality of care to the area offices. Area officials told us that the oversight they provide has generally included (1) holding periodic meetings with facility staff, such as governing board and other meetings; (2) reviewing available quality performance data; (3) reviewing data on adverse events occurring in their facilities; (4) monitoring compliance with facility certification and accreditation requirements; and (5) appraising employee performance. However, our review found that this oversight was limited and inconsistent across IHS areas and facilities, in part due to a lack of agency-wide quality performance standards and significant leadership turnover in some offices. Meeting with facility staff. Officials from all nine of the area offices that oversee federally operated IHS facilities told us that they monitor the quality of care provided by facilities through periodic meetings with facility staff—including governing board meetings and other meetings. However, according to area office officials, the frequency of these meetings varies widely by area. In general, these meetings are used to discuss a range of issues, such as quality of care, equipment problems, staff vacancies, and provider credentialing. For example, documentation of a governing board meeting with facility staff in the Phoenix area shows that board members and staff discussed a problem with the facility’s wireless internet connection, which was negatively affecting their bar code medication administration (BCMA) system. A board member noted that these connectivity issues caused a patient safety risk, but another board member noted that they had a short-term resolution in place and area office officials were working with the wireless carrier to resolve the problem over the long- term. Area offices vary, however, as to whether, or to what extent, these meetings focus on the quality of care. For example, officials from one area office stated that their governing board meetings include a standardized agenda that includes quality of care items, and that facilities are required to submit data reports that include information on quality issues such as rates of hospital acquired infections, patient complaints, and provider productivity. In contrast, officials from another area office told us that there are no standing agenda items for the discussion of quality of care, and that facility staff set the meeting agendas based on issues they want to discuss. Furthermore, the frequency of governing board meetings with facility staff varied widely among the area offices, ranging from quarterly to annually. Reviewing available quality performance data. According to IHS officials, clinical quality performance data are generally collected and reported consistently to IHS’s area and headquarters offices in response to requirements in the Government Performance and Results Act of 1993 (GPRA), but other data used to oversee the quality of care provided in facilities are not reported or reviewed consistently across IHS. Officials from all nine area offices in our review stated that they periodically review reports showing facility progress in meeting 24 annual GPRA clinical performance measure targets. These performance measures focus on health screening and prevention activities, such as cancer screening, immunization rates, and tobacco cessation activities, and do not include broader measures of quality, such as whether patients are receiving proper diagnoses and medications, and the extent to which facility staff properly perform infection control activities. Area officials reported that they review other quality performance data, such as the percentage of medication orders reviewed for therapeutic duplication, the number of mislabeled laboratory specimens, and patient satisfaction, but these data are not consistently obtained or reviewed by all area offices because IHS has not required that they be reviewed or reported. In addition, staff from the two facilities in our review told us that limitations of IHS’s electronic health record system—the Resource and Patient Management System (RPMS)—also contribute to variation in the quality performance data that are collected and reported to area offices. For example, staff told us that certain data elements, such as patient diagnoses, are difficult to extract from RPMS. Officials from facilities and area offices said that pulling these data may require special modifications to RPMS. Officials said that modifying RPMS requires knowledge of computer programming and can be costly, so some facility staff may manually enter and extract certain data. Officials from several areas also told us that some facilities have hired contractors or purchased software to assist them in monitoring their data. One such software package—QlikView— provides facility staff with multiple data reporting options. Staff from one facility said that, from an information technology standpoint, they face “massively complicated issues” when the lack of a standardized user interface leads to individual facilities across IHS customizing RPMS for their own needs. IHS officials told us that they are working on improving RPMS, in part, by developing software to stabilize the system; however, we have not assessed these efforts. Monitoring adverse events. Officials from all nine area offices told us that their oversight of the quality of care includes monitoring adverse events that occur at IHS facilities, such as medication errors or patient falls, and taking steps to prevent future adverse events. While all IHS facilities have the means to report and monitor adverse events through an IHS-wide web-based reporting tool—WebCident— officials told us reporting adverse events through WebCident has been inconsistent. Officials from one area office stated that adverse events are not always reported through WebCident, and therefore the appropriate staff are often not notified when adverse events occur, including those resulting in patient harm. These officials stated that this creates a lost opportunity to address the deficiency and improve, as well as to hold individuals accountable. Officials from IHS headquarters reported that they plan to enhance this reporting system to encourage consistent use by facility staff, or replace it with a new system after January 2017. Monitoring compliance with facility certification and accreditation requirements. Officials from all nine area offices in our review told us that they monitor the ongoing compliance with certification and accreditation requirements of the facilities in their area—such as through mock surveys and other interim monitoring—to help ensure that they maintain their certification by CMS to participate in the Medicare and Medicaid programs, as well as their accreditation by accrediting bodies such as The Joint Commission and the Accreditation Association for Ambulatory Health Care (AAAHC). For example, documentation of a mock survey of a facility in the Phoenix area states that the facility emergency department was improperly storing contaminated medical instruments. In addition, documentation of a mock survey of a facility in the Albuquerque area states that the surveyors found defective lead aprons, as well as a high-voltage power line sitting on the floor instead of behind a wall, and exposed electrical wiring blocking the door and wrapped around a door handle—which the surveyors concluded could be a serious hazard. These findings underscore the need for such surveys, but the frequency with which these mock surveys have been conducted varies by area. Officials from one area office told us that they have conducted mock accreditation surveys of facilities in their area annually for the past 15 years. Officials from other area offices stated that they have recently begun performing such mock surveys. IHS officials told us that in May 2016, IHS began a system-wide mock survey initiative at all 26 federally operated hospitals to assess compliance with the CMS Conditions of Participation and readiness for reaccreditation. Surveys conducted by CMS and accrediting bodies are relatively infrequent, however, and this infrequency highlights the importance of interim monitoring. For instance, area office staff told us that The Joint Commission conducts site visits every 3 years, and, while CMS attempts to conduct site visits every 3 to 4 years, staff of one facility we visited stated that CMS had not surveyed the facility in 10 years. Appraising employee performance. According to IHS, area directors are held accountable for achieving agency-wide goals and specific performance objectives through an appraisal process that also enables these goals and objectives to cascade down to chief executive officers (CEO) at individual facilities and to all agency employees. Area directors sign performance agreements documenting their accountability. The fiscal year 2016 performance requirements included a provision on ensuring that all IHS operated health care facilities achieve and maintain accreditation or certification by a national health care organization in fiscal year 2016. The performance requirements also included a provision on quality care that requires documentation of the implementation of “at least two activities to improve wait times and access to quality health care for patients that are based on enhanced implementation of current quality initiatives or new quality initiatives and that have measurable goals, measures and outcomes,” as well as improvements resulting from these efforts. However, area officials can choose activities to satisfy this requirement from a list of suggestions—such as improving customer service and expanding clinic hours—without directly addressing the quality of care in their facilities. These inconsistencies are exacerbated by significant turnover in area leadership. Officials from four of the nine area offices in our review reported that they had at least three area directors in the past 5 years, and officials from three area offices reported that they had at least three chief medical officers. (See fig. 3 and 4). Officials stated that inconsistent area office and facility leadership is detrimental to the oversight of facility operations and the supervision of personnel. For example, officials from multiple area offices told us that frequent leadership turnover can lead to instability in oversight initiatives if these initiatives are started but not completed. In addition, an area office’s review of a facility in the Navajo area documented that the majority of facility staff interviewed felt that there were too many people in acting leadership positions, that acting leaders were afraid to commit to decisions, and that the leaders needed additional supervisory training. In addition, the facility staff interviewed stated that those in acting leadership positions had their own work to contend with and were not always responsive to the responsibilities of the leadership position. According to IHS officials, the agency has not defined contingency or succession plans for the replacement of key personnel, including area directors. See appendix I for additional information on leadership turnover within IHS. IHS’s limited and inconsistent agency-wide oversight of the quality of care in its federally operated facilities, as well as its lack of contingency and succession plans for key personnel, is inconsistent with federal internal control standards. These standards suggest that agencies should establish and review performance standards and then monitor data to assess the quality of performance over time, and that agencies should define contingency and succession plans for key roles to help continue achieving objectives. As a result of IHS’s lack of consistent agency-wide quality performance standards, as well as the significant turnover in area leadership, IHS officials cannot ensure that facilities are providing quality health care to their patients, and therefore that the agency is making steps toward fulfilling its mission to raise the physical, mental, social, and spiritual health of AI/AN people to the highest level. Recognizing some of the challenges it faces with overseeing and providing quality health care in its facilities, IHS finalized the development of a quality framework in November 2016 that outlines, at a high level, IHS’s vision, goals, and priorities to develop, implement, and sustain an effective quality program that is intended to improve patient experience and ensure the delivery of reliably high quality health care for IHS direct service facilities. According to IHS, the priorities of the framework are to (1) strengthen organizational capacity to improve quality of care and systems, (2) meet and maintain accreditation for federally operated facilities, (3) align service delivery processes to improve patient experience, (4) ensure patient safety, and (5) improve transparency and communication regarding patient safety and quality to IHS stakeholders. While this framework is focused on initiatives related to improving the quality of care provided in its facilities—such as increasing staff training and technical assistance on achieving compliance with quality and safety standards, promoting a culture of patient safety, and developing a patient perception survey process—elements of the framework also describe IHS plans to improve oversight. For example, the framework directs IHS to establish a quality office that will be responsible for assessing area office and facility functions, staffing, and critical quality assurance activities. This quality office is to be developed as part of an overall realignment of offices in IHS, and according to the framework, the office will be responsible for identifying resource needs, structures, processes, and supports for an effective and sustainable quality assessment and performance improvement system. More specifically, the framework directs IHS to develop a process for monitoring performance measures, such as measures of clinical care, patient access, and financial performance, for periodic review by leadership. In addition, IHS’s quality framework states that the agency will implement annual mock accreditation surveys for all federally operated facilities and develop a standardized governing board structure to improve planning and oversight processes. The framework says that “transparency and accountability will be fostered through regular and frequent (i.e., monthly or quarterly) communications” between offices. The framework also explains that IHS will enhance its adverse event reporting system to encourage consistent use by facility staff, or replace it with a new system after January 2017. If effectively implemented, the quality framework could address the limited and inconsistent oversight of the quality of care provided in federally operated IHS facilities. However, as of November 2016, the quality office had not yet been formed, and officials told us the agency’s plan for realigning offices was out for tribal review and comment. In addition, IHS officials stated that the agency has not yet selected quality performance measures but has plans to do so. Furthermore, the quality framework states that IHS will support enhanced efforts to recruit and retain highly qualified executives. While IHS officials reported that they are implementing strategies to recruit and retain staff, the quality framework does not specifically mention contingency or succession plans for key personnel. American Indians and Alaska Natives die at higher rates than other Americans from many causes—such as lower respiratory infections and complications from diabetes—that can be mitigated through access to quality health care services, and concerns continue to be raised about the quality of care provided in federally operated IHS facilities, including misdiagnoses, incorrectly prescribed medications, and unsafe facility conditions. Despite IHS’s mission to raise the physical, mental, social, and spiritual health of American Indians and Alaska Natives to the highest level, IHS’s oversight of the quality of care in its federally operated facilities has been limited and inconsistent. In addition, several of the area offices in our review experienced frequent leadership turnover with no contingency or succession plans. While IHS has recognized the need for quality improvement and has drafted a quality framework to improve the oversight of the quality of care provided, it has not yet developed quality performance standards. Until IHS develops agency-wide standards for the quality of care provided in its federally operated facilities, systematically monitors facility performance in meeting these standards at all facilities, and develops contingency and succession plans for key personnel to address its significant leadership turnover, it cannot ensure that it is consistently providing quality medical care to the AI/AN population served in its facilities. To help ensure that quality care is provided to AI/AN people, the Secretary of HHS should direct the Director of IHS to take the following two actions: 1. As part of the implementation of its quality framework, ensure that agency-wide standards for the quality of care provided in its federally operated facilities are developed, that facility performance in meeting these standards is systematically monitored over time, and that enhancements are made to its adverse event reporting system. 2. Develop contingency and succession plans for the replacement of key personnel, including area directors. We provided a draft of this report to HHS for its review and comment. HHS provided written comments, which are reproduced in appendix II. HHS concurred with both of our recommendations. In its comments, HHS elaborated on steps that IHS has started taking to improve its oversight of the quality of care provided in its federally operated facilities, which we describe in our report. Specifically, HHS described the development of IHS’s quality framework and quality office, plans to develop agency-wide quality measures, the standardization of governing board by-laws, plans to enhance or replace its adverse event reporting system, and its annual mock survey initiative. In its comments, HHS also described IHS’s corrective action plan process related to the mock survey initiative, and we added this information to our report. HHS also provided information on steps that IHS is taking to improve the quality of care in its federally operated facilities, including steps taken toward the automation and systemization of provider credentialing. Regarding our second recommendation to develop contingency and succession plans for the replacement of key personnel, including area directors, HHS stated that IHS has already begun to address this recommendation. For instance, HHS reported that on December 2, 2016, IHS distributed succession planning instructions and descriptions of the competencies associated with each position in IHS headquarters, area offices, and facilities to all headquarters office directors and area directors. In addition, HHS reported that IHS has contingency plans in place to ensure continuity of operations in emergency situations. However, as explained in our report, standards for internal control in the federal government state that the agency should have contingency plans in place to respond to sudden personnel changes, which would include non-emergency situations as well. HHS also provided technical comments, which we incorporated where appropriate. We are sending a copy of this report to the Secretary of the Department of Health and Human Services. The report will be available at no charge on GAO’s website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or kingk@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. Officials from some area offices in our review reported significant turnover of staff in area office leadership positions. Officials from four of the nine area offices in our review reported that they had at least three area directors in the past 5 years, and three area offices reported that they had at least three chief medical officers. See table 2. In addition, area offices reported 19 federally operated facilities that had 4 or more chief executive officers (CEO) in the past 5 years. One area reported that one of its hospitals had 10 CEOs and 6 clinical directors in the past 5 years. See table 3. In addition to the contact named above, Kristi Peterson, Assistant Director; Kelly DeMots; Krister Friday; Keith Haddock; Lisa Rogers; Patricia Roy; Jennifer Whitworth; and Emily Wilson made key contributions to this report.

IHS is charged with providing health care to American Indian/Alaska Native (AI/AN) people who are members or descendants of 567 tribes. AI/AN people born today have a life expectancy that is 4.4 years lower than all races in the United States, and they continue to die at higher rates than other Americans from preventable causes. Concerns about the quality of care provided to AI/ANs in IHS facilities have been identified recently by federal officials and tribal members. GAO was asked to review how IHS oversees the quality of care provided in its facilities. This report examines IHS's oversight of the quality of care provided in its federally operated facilities. GAO reviewed policies and guidance related to quality of care in federally operated facilities and interviewed IHS officials at the headquarters level and all nine area offices with federally operated facilities. GAO also examined documents from governance meetings between area office and facility staff. The Indian Health Service's (IHS) oversight of the quality of care provided in its federally operated facilities has been limited and inconsistent. While some oversight functions are performed at the headquarters level, the agency has delegated primary responsibility for the oversight of care to nine area offices. Area officials stated that the oversight they provide has included, for example, holding periodic meetings with facility staff, reviewing available quality performance data and reviewing adverse events. However, GAO found that this oversight was limited and inconsistent across IHS facilities, due in part to a lack of agency-wide quality of care standards. Specifically, GAO found: variation in the frequency of governing board meetings and the extent to which quality was a standing agenda item at these meetings; limited and inconsistent reporting of quality data across IHS areas and facilities; and inconsistent reporting of adverse events at federally operated facilities. These inconsistencies are also exacerbated by significant turnover in area leadership. Officials from four of the nine area offices in our review reported that they each had at least three area directors in the past five years. According to IHS officials, the agency has not defined contingency or succession plans for the replacement of key personnel, including area directors. IHS's lack of agency-wide quality of care standards and lack of contingency and succession plans for key personnel are inconsistent with federal internal control standards. These standards suggest that agencies should establish and review performance standards and then monitor data to assess the quality of performance over time, and define contingency and succession plans for the replacement of key personnel to help IHS continue achieving its objectives. As a result, IHS officials cannot ensure that facilities are providing quality health care. Recognizing the challenges it faces with overseeing and providing quality health care in its facilities, IHS finalized the development of a quality framework in November 2016 that outlines, at a high level, IHS's plan to develop, implement, and sustain a quality program intended to improve patient experience and ensure the delivery of reliably high quality health care. For example, the framework directs IHS to develop a quality office that will be responsible for identifying resource needs, structures, processes, and supports for an effective and sustainable quality assessment and performance improvement system. More specifically, the framework directs IHS to develop a process for monitoring select performance measures, such as measures of clinical care, patient access, and financial performance, for periodic review by leadership. The framework also explains that IHS will enhance its current patient safety reporting systems to encourage consistent use by staff. If effectively implemented, the quality framework could address the limited and inconsistent oversight of the quality of care provided in federally operated IHS facilities. As of November 2016, IHS officials stated that the agency has not yet selected quality performance measures but has plans to do so. GAO recommends that the Secretary of the Department of Health and Human Services direct the Director of IHS to (1) as it implements its quality framework, ensure that agency-wide standards for the quality of care provided in its federally operated facilities are developed, that facility performance in meeting these standards is monitored over time, and that enhancements are made to its adverse event reporting system, and (2) develop contingency and succession plans for the replacement of key personnel. HHS concurred with GAO's recommendations.

VA provides medical services to various veteran populations—including an aging veteran population and a growing number of younger veterans returning from the military operations in Afghanistan and Iraq. VA operates approximately 170 VAMCs, 130 nursing homes, and 1,000 outpatient sites of care. In general, veterans must enroll in VA health care to receive VA’s medical benefits package—a set of services that includes a full range of hospital and outpatient services, prescription drugs, and long-term care services provided in veterans’ own homes and in other locations in the community. The majority of veterans enrolled in the VA health care system receive care in VAMCs and community-based outpatient clinics, but VA may authorize care through community providers to meet the needs of the veterans it serves. For example, VA may provide care through its Care in the Community (CIC) program, such as when a VA facility is unable to provide certain specialty care services, like cardiology or orthopedics. CIC services must generally be authorized by a VAMC provider prior to a veteran receiving care. In addition to the CIC program, VA may also provide care to veterans through the Veterans Choice Program, which was established through the Veterans Access, Choice, and Accountability Act of 2014 (Choice Act), enacted on August 7, 2014. Implemented in fiscal year 2015, the program generally provides veterans with access to care by non-VA providers when a VA facility cannot provide an appointment within 30 days or when veterans reside more than 40 miles from the nearest VA facility. The Veterans Choice Program is primarily administered using contractors, who, among other things, are responsible for establishing nationwide provider networks and scheduling appointments for veterans. The Choice Act created a separate account known as the Veterans Choice Fund, which cannot be used to pay for VA obligations incurred for any other program, such as CIC, without legislative action. The Choice Act appropriated $10 billion to be deposited in the Veterans Choice Fund. Amounts deposited in the Veterans Choice Fund are available until expended and are available for activities authorized under the Veterans Choice Program. However, the Veterans Choice Program activities are only authorized through fiscal year 2017 or until the funds in the Veterans Choice Fund are exhausted, whichever occurs first. As part of the President’s request for funding to provide medical services to veterans, VA develops an annual budget estimate detailing the amount of services it expects to provide as well as the estimated cost of providing those services. VA uses the Enrollee Health Care Projection Model (EHCPM) to develop most of the agency’s estimates of the budgetary needs to meet the expected demand for VA medical services. Like many other agencies, VA begins to develop these estimates approximately 18 months before the start of the fiscal year for which funds are provided. Different from many agencies, VA’s Veterans Health Administration receives advance appropriations for health care in addition to annual appropriations. VA’s EHCPM makes these projections 3 or 4 years into the future for budget purposes based on data from the most recent fiscal year. In 2012, for example, VA used actual fiscal year 2011 data to develop the budget estimate for fiscal year 2014 and the advance appropriation estimate for fiscal year 2015. Similarly, in 2013, VA used actual fiscal year 2012 data to update the budget estimate for fiscal year 2015 and develop the advance appropriation estimate for fiscal year 2016. Given this process, VA’s budget estimates are prepared in the context of uncertainties about the future—not only about program needs, but also about future economic conditions, presidential policies, and congressional actions that may affect the funding needs in the year for which the estimate is made—which is similar to budgeting practices of other federal agencies. Further, VA’s budget estimates are typically revised during the budget formulation process to incorporate legislative and department priorities as well as in response to successively higher level of reviews in VA and OMB. Each year, Congress provides funding for VA health care primarily through the following appropriation accounts: Medical Services, which funds, among other things, health care services provided to eligible veterans and beneficiaries in VA’s medical centers, outpatient clinic facilities, contract hospitals, state homes, and outpatient programs on a fee basis. The CIC program is funded through this appropriation account. Medical Support and Compliance, which funds, among other things, the administration of the medical, hospital, nursing home, domiciliary, construction, supply, and research activities authorized under VA’s health care system. Medical Facilities, which funds, among other things, the operation and maintenance of the Veterans Health Administration’s capital infrastructure, such as costs associated with nonrecurring maintenance, utilities, facility repair, laundry services, and groundskeeping. Our preliminary work suggests that the higher-than-expected obligations identified by VA in April 2015 for VA’s CIC program accounted for $2.34 billion (or 85 percent) of VA’s projected funding gap of $2.75 billion in fiscal year 2015. These higher-than-expected obligations for the CIC program were driven by an increase in utilization of VA medical services across VA, reflecting, in part, VA’s efforts to improve access to care after public disclosure of long wait times at VAMCs. VA officials expected that the Veterans Choice Program would absorb much of the increased demand from veterans for health care services delivered by non-VA providers. However, veterans’ utilization of Veterans Choice Program services was much lower than expected in fiscal year 2015. VA had estimated that obligations for the Veterans Choice Program in fiscal year 2015 would be $3.2 billion, but actual obligations totaled only $413 million. Instead, VA provided a greater amount of services through the CIC program, resulting in total obligations of $10.1 billion, which VA officials stated were much higher than expected for that program in fiscal year 2015. According to VA officials, the lower-than-expected utilization of the Veterans Choice Program in fiscal year 2015 was due, in part, to administrative weaknesses, such as provider networks that had not been fully established, that slowed enrollment in the program and that VAMC staff lacked guidance on when to refer veterans to the program. The unexpected increase in CIC obligations in fiscal year 2015 exposed weaknesses in VA’s ability to estimate costs for CIC services and track associated obligations. While VA officials first became concerned that CIC obligations might be significantly higher than projected in January 2015, they did not determine that VA faced a projected funding gap until April 2015—6 months into the fiscal year. They made this determination after they compared authorizations in the Fee Basis Claims System (FBCS)—VA’s system for recording CIC authorizations and estimating costs for this care—with obligations in the Financial Management System (FMS)—the centralized financial management system VA uses to track all of its obligations, including those for medical services. In its 2015 Agency Financial Report (AFR), VA’s independent public auditor identified the following issues as contributing to a material weakness in estimating costs for CIC services and tracking CIC obligations: VAMCs individually estimate costs for each CIC authorization and record these estimates in FBCS. This approach leads to inconsistencies, because each VAMC may use different methodologies to estimate the costs they record. Having more accurate cost estimates for CIC authorizations is important to help ensure that VA is aware of the amount of money it must obligate for CIC services. VAMCs do not consistently adjust estimated costs associated with authorizations for CIC services in a timely manner to ensure greater accuracy, and they do not perform a “look-back” analysis of historical obligations to validate the reasonableness of estimated costs. Furthermore, centralized, consolidated, and consistent monitoring of CIC authorizations is not performed. FBCS is not fully integrated with VA’s systems for recording and tracking the department’s obligations. Notably, the estimated costs of CIC authorizations recorded in FBCS are not automatically transmitted to VA’s Integrated Funds Distribution, Control Point Activity, Accounting, and Procurement (IFCAP) system, a procurement and accounting system used to send budgetary information, such as obligations, to FMS. According to VA officials, because FBCS and IFCAP are not integrated, at the beginning of each month, VAMC staff must record in IFCAP estimated obligations for outpatient CIC services, and they use historical obligations for this purpose. Depending on the VAMC, these estimated obligations may be entered as a single lump sum covering all outpatient care or as separate estimated obligations for each category of outpatient care, such as radiology. Regardless of how they are recorded, the estimated obligations recorded in IFCAP are often inconsistent with the estimated costs of CIC authorizations recorded in FBCS. In fiscal year 2015, the estimated obligations that VAMCs recorded in IFCAP were significantly lower than the estimated costs of outpatient CIC authorizations recorded in FBCS. VA officials told us that they did not determine a projected funding gap until April 2015, because they did not complete their analysis of comparing estimated obligations with estimated costs until then. In addition, the Chief Business Office (CBO) within the Veterans Health Administration, which is responsible for developing administrative processes, policy, regulations, and directives associated with the CIC program, had not developed and implemented standardized and comprehensive policies for VAMCs, regional networks, and the office itself to follow when estimating costs for CIC authorizations and for monitoring authorizations and associated obligations. This contributed to the material weaknesses the independent public auditor identified in the AFR. The AFR and VA officials we interviewed stated that because CIC was consolidated under CBO in fiscal year 2015 pursuant to the Choice Act, CBO did not have adequate time to implement efficient and effective procedures for monitoring CIC obligations. To address the fiscal year 2015 projected funding gap, on July 31, 2015, VA obtained temporary authority to use up to $3.3 billion in Veterans Choice Program funds for obligations incurred for medical services from non-VA providers, whether authorized under the Veterans Choice Program or CIC, starting May 1, 2015 and ending October 1, 2015. Based on our preliminary work, Table 1 shows the sequence of events that led to VA’s request for and approval of additional budget authority for fiscal year 2015. Our preliminary work also suggests that unexpected obligations for new hepatitis C drugs accounted for $0.41 billion of VA’s projected funding gap of $2.75 billion in fiscal year 2015. Although VA estimated that obligations in this category would be $0.7 billion that year, actual obligations totaled about $1.2 billion. VA officials told us that VA did not anticipate in its budget the obligations for new hepatitis C drugs —which help cure the disease—because the drugs were not approved by the Food and Drug Administration until fiscal year 2014, after VA had already developed its budget estimate for fiscal year 2015. The new drugs costs between $25,000 and $124,000 per treatment regimen, and according to VA officials demand for the treatment was high. Officials told us that about 30,000 veterans received these drugs in fiscal year 2015. In October 2014, VA reprogrammed $0.7 billion within its medical services appropriation account to cover projected obligations for the new hepatitis C drugs, after VA became aware of the drugs’ approval. However, in January 2015, VA officials recognized that obligations for the new hepatitis C drugs would be significantly higher by year end than they expected. VA officials told us that they assessed next steps and then limited access to the drugs to those veterans with the most severe cases of hepatitis C. In June 2015, VA requested statutory authority to transfer funds dedicated to the Veterans Choice Program to VA’s medical services appropriation account to cover the projected funding gap. Our preliminary work indicates that VA has developed new processes to prevent funding gaps for fiscal year 2016 and future years by improving its ability to track obligations for CIC services and hepatitis C drugs. In August 2015, VA issued a standard operating procedure to all VAMCs for recording estimated costs for inpatient and outpatient CIC in FBCS. The procedure, among other things, stipulates that VAMCs are to base estimated costs on historical cost data provided by VA. In addition, VA developed a software patch—released in December 2015 to all VAMCs—that automatically generates estimated costs for CIC authorizations, thereby eliminating the need for VAMC staff to individually estimate costs and record them in FBCS. According to VA officials, these changes should result in more accurate estimated costs for CIC authorizations. However, VA officials told us that accurately estimating the cost of CIC authorizations is challenging because of several unknown factors, such as the number of times a veteran may seek treatment for a recurring condition. In November 2015, VA allocated funds for CIC and hepatitis C drugs to each VAMC. In addition, VA officials told us that to identify VAMCs that may be at risk for exhausting their funds before the end of the fiscal year, VA began tracking VAMCs’ obligations for CIC and hepatitis C drugs through monthly reports. Officials from the Office of Finance within the Veterans Health Administration told us that once a VAMC had obligated its CIC and hepatitis C drug funds, it would have to request additional funds from VA. VA would, in turn, evaluate the validity of a VAMC’s request and determine whether additional funds may be made available. This practice could limit veterans’ access to CIC services or hepatitis C drugs in some locations. Officials told us that these steps are intended to reduce the risk of VAMCs obligating more funds than VA’s budgetary resources allow. In November 2015, VA also issued a policy requiring VAMCs to identify and report on potentially inaccurate estimated costs for CIC authorizations recorded in FBCS and any discrepancies between estimated costs for CIC authorizations recorded in FBCS and the amount of estimated obligations recorded in FMS. According to VA officials, these discrepancies may signal a risk of VA under obligating funds for CIC, leaving VA potentially unable to pay for authorized care. VA’s policy also requires VAMCs to address concerns identified by VAMCs in these reports—such as adjusting unreasonably low estimated costs for CIC authorizations and unreasonably low estimated obligations, to make the estimates more accurate. Under VA’s new policy, network directors are required to certify monthly that the reports have been reviewed and concerns addressed. VA officials told us that these new processes are necessary to help prevent future funding gaps because of the deficiencies in VA’s systems for tracking obligations, which we have described previously. Officials also told us that VA is exploring options for replacing IFCAP and FMS, which officials describe as antiquated systems based on outdated technology, and the department has developed a rough timeline and estimate of budgetary needs to make these changes. Officials told us that the timeline and cost estimate would be refined once concrete plans for replacing IFCAP and FMS are developed. Officials told us that replacing IFCAP and FMS is challenging due to the scope of the project and the requirement that the replacement system interface with various VA legacy systems, such as the Veterans Health Information Systems and Technology Architecture, VA’s system containing veterans’ electronic health records. However, as we have previously reported, VA has made previous attempts to update IFCAP and FMS that were unsuccessful. In October 2009, we attributed these failures to the lack of a reliable implementation schedule and cost estimates, among other factors, and made several recommendations aimed at improving program management. Our preliminary work indicates that VA updated its EHCPM to include data from the first 6 months of fiscal year 2015, reflecting increased health care utilization in that year, which VA officials told us will inform VA’s budget estimate for fiscal year 2017 and advance appropriations request for fiscal year 2018. Without this change, VA would have used actual data from fiscal year 2014 to make its budget estimate and inform the President’s budget request for fiscal years 2017 and 2018. However, as we have previously reported, while the EHCPM projection informs most of VA’s budget estimate, the amount of the estimate is determined by several factors, including the President’s priorities. Historically, the final budget estimate for VA has consistently been lower than the amount projected for modeled services. VA officials told us that they expect any difference between the fiscal year 2017 budget estimate and the amount projected by VA’s model to be made up by greater utilization of the Veterans Choice Program. However, VA’s authority to use Veterans Choice Program funds is only available through fiscal year 2017 or until the funds are exhausted, whichever occurs first. VA has also taken steps to help increase utilization of the Veterans Choice Program. VA issued policy memoranda to VAMCs in May and October 2015, requiring them to refer veterans to the program if timely care cannot be delivered by a VAMC, rather than authorizing care through the CIC program. With statutory authority, VA has also loosened restrictions on veterans’ use of the Veterans Choice Program, eliminating the requirement that veterans must be enrolled in the VA health care system by August 2014 in order to receive care through the program. While data from November 2015 indicate that utilization of care under the Veterans Choice Program has increased, VA officials expressed concerns that utilization would not reach the levels projected for fiscal year 2016 because of continuing weaknesses in implementing the program. For example, in November 2015, VA’s Office of Compliance and Business Integrity identified extensive noncompliance among VAMCs with VA’s policies for implementing the Veterans Choice Program and recommended training for VAMC staff responsible for implementing the program. The office also recommended that VA establish internal controls to ensure compliance with VA’s policies. As of January 2016, VA had not completed a plan for establishing these internal controls. Like other health care payers, VA faces uncertainties estimating the cost of emerging health care treatments—such as costly drugs to treat chronic diseases affecting veterans. VA, like other federal agencies, prepares its budget estimate 18 months in advance of the start of the fiscal year for which funds are provided. At the time VA develops its budget estimate, it may not have enough information to estimate the likely costs for health care services or these treatments with reasonable accuracy. However, by establishing appropriate internal controls, VA can help reduce the risks associated with the weaknesses in its budgetary projections and monitoring. Chairman Miller, Ranking Member Brown, and Members of the Committee, this concludes my statement for the record. If you or your staff members have any questions concerning this statement, please contact Randall B. Williamson, Director, Health Care, at 202-512-7114 or williamsonr@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this statement include Rashmi Agarwal, Assistant Director; Luke Baron; Krister Friday; Jacquelyn Hamilton; and Michael Zose. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

VA projected a funding gap in its fiscal year 2015 medical services appropriation account and obtained temporary authority to use up to $3.3 billion in Veterans Choice Program funding to close this gap. GAO was asked to examine VA's fiscal year 2015 projected funding gap and changes VA has made to help prevent potential funding gaps in future years. This statement is based on GAO's ongoing work and provides preliminary observations on (1) the activities or programs that accounted for VA's fiscal year 2015 projected funding gap in its medical services appropriation account and (2) changes VA has made to prevent potential funding gaps in future years. GAO reviewed data VA provided on its obligations and related documents to determine what activities accounted for the projected funding gap in its fiscal year 2015 medical services appropriation account, as well as the factors that contributed to the projected funding gap. GAO interviewed VA and Office of Management and Budget officials to identify the steps taken to address the projected funding gap. GAO also examined changes VA made to better track obligations and project future budgetary needs. GAO shared the information provided in this statement with VA and incorporated its comments as appropriate. GAO's ongoing work indicates that two areas accounted for the Department of Veterans Affairs' (VA) fiscal year 2015 projected funding gap of $2.75 billion. Specifically, Higher-than-expected obligations for VA's longstanding care in the community (CIC) program—which allows veterans to obtain care from providers outside of VA facilities—accounted for $2.34 billion or 85 percent of VA's projected funding gap. VA officials expected that the new Veterans Choice Program—which was implemented in fiscal year 2015 and also allows veterans to access care from non-VA providers under certain conditions—would absorb veterans' increased demand for care after public disclosure of long wait times. However, administrative weaknesses slowed enrollment into this new program. The unexpected increase in CIC obligations also exposed VA's weaknesses in estimating costs for CIC services and tracking associated obligations. VA officials did not determine that VA faced a projected funding gap until April 2015—6 months into the fiscal year, after they compared estimated authorizations with estimated obligations for CIC. Unanticipated obligations for hepatitis C drugs accounted for the remaining portion—$408 million—of VA's projected funding gap. VA did not anticipate in its budget the obligations for these costly, new drugs, which can help cure the disease, because the drugs did not gain approval from the Food and Drug Administration until fiscal year 2014—after VA had already developed its budget estimate for fiscal year 2015. VA officials told GAO that in fiscal year 2015 about 30,000 veterans received these drugs, which cost between $25,000 and $124,000 per treatment regimen. GAO's ongoing work indicates that VA has taken steps to better track obligations and project future healthcare utilization, but systems deficiencies and budgetary uncertainties remain. Specifically, GAO's preliminary results indicate that VA has taken the following steps: VA issued a standard operating procedure to help VA medical centers (VAMC) more accurately estimate the costs associated with authorizations for CIC. VA directed VAMCs to compare their estimated costs for CIC authorizations with estimated obligations for CIC on a monthly basis. VA allocated funds to each VAMC for CIC and hepatitis C drugs and began tracking VAMCs' obligations with monthly reports. Officials told GAO that once a VAMC has obligated its funds, it would have to request additional funds. VA would determine whether additional funds may be made available. These processes are necessary because continued deficiencies in VA's financial systems present challenges in tracking of obligations. VA updated the model it uses to inform most of its budget estimates for medical services. It now includes more recent data that reflect increased healthcare utilization among veterans in fiscal year 2015. However, VA officials noted uncertainties remain about the forecasted utilization of the Veterans Choice Program and emerging health care treatments, which could affect the accuracy of the health care budget estimates.

Congress has incrementally expanded the use and scope of “other transaction” authority since first authorizing its use more than a decade ago. In 1989, Congress gave DOD, acting through the Defense Advanced Research Projects Agency (DARPA), authority to temporarily use “other transactions” for basic, applied, and advanced research projects. In 1991, Congress made this authority permanent and extended it to the military services. In 1993, Congress enacted Section 845 of the National Defense Authorization Act for Fiscal Year 1994, which provided DARPA with authority to use, for a 3-year period, “other transactions” to carry out prototype projects directly relevant to weapons or weapon systems proposed to be acquired or developed by DOD. The National Defense Authorization Act for Fiscal Year 1997 temporarily extended DARPA’s Section 845 authority and provided similar authority to the military services and defense agencies. Congress subsequently extended this authority’s expiration date until September 30, 2004. In an era of a shrinking defense industrial base and new threats, DOD views “other transaction” prototype authority as a key to attracting nontraditional defense contractors. Section 803 of the Floyd D. Spence National Defense Authorization Act for Fiscal Year 2001 defined a nontraditional defense contractor as an entity that has not, for at least a period of one year prior to the date of entering into or performing an “other transaction,” entered into or performed (1) any contract subject to full coverage under the cost accounting standards or (2) any other contract in excess of $500,000 to carry out prototype projects or to perform basic, applied, or advanced research projects for federal agencies. DOD also views Section 845 authority as a way to test creative procurement strategies—such as the use of teaming and consortia—with traditional defense contractors and in industry areas not normally associated with government contracts. Under this authority, new business relationships, which could involve changes in traditional business processes or intellectual property rights agreements, are created to leverage commercial investments and to permit DOD to influence the design, development, and availability of commercial technologies to address national security needs. In fiscal year 2001, the most recent year for which complete data are available, DOD awarded 61 Section 845 agreements, totaling $392 million in federal government funds. Contractors contributed another $97 million in cost-sharing funds. Figure 1 shows these agreements by awarding organization. DOD is required to submit an annual report to Congress addressing both research and prototype “other transaction” agreements awarded in the preceding fiscal year. The report, which is prepared and signed by the Director, Defense Research and Engineering, includes input from the Director of Defense Procurement on Section 845 agreements. The report is to address (1) the technology areas in which the work was focused; (2) the extent of cost sharing among federal and nonfederal sources; and (3) how “other transactions” contributed to a broadening of the technology and industrial base and fostered new relationships and practices that support U.S. national security interests. In December 2000, the Under Secretary of Defense for Acquisition and Technology issued a revised guide that sets out the conditions and framework for using Section 845 agreements. The guide is effective for all solicitations issued after January 5, 2001, and provides a useful framework for tailoring the terms and conditions appropriate for each agreement. DOD agreements officers view the new guide as a significant improvement over the prior version. Several key improvements are as follows: The previous guide contained very limited information on the terms and conditions to be tailored when crafting a Section 845 agreement. The current guide provides additional details on the appropriate use of terms and conditions such as intellectual property, accounting systems, and cost sharing. It instructs agreements officers not to view previously issued agreements as a template or model, but to rely on their skill and experience and to consider Federal Acquisition Regulation clauses and commercial business practices, as well as prior “other transactions,” when formulating agreements. The current guide also requires an acquisition strategy that identifies and discusses the rationale for using a Section 845 agreement. The previous guide did not define “nontraditional” contractors. The current guide defines the term, based in part on the definition in Section 803 of the National Defense Authorization Act for Fiscal Year 2001. It also requires that information on these entities be collected. DOD considers nontraditional defense contractors to be “business units,” which can be any segment of an organization or an entire business organization that is not divided into segments. The previous guide listed eight examples of benefits to be considered under Section 845 agreements, including attracting business entities that normally do not do business with the government. However, it did not identify a specific metric that should be used on all Section 845 agreements. The current guide clearly states that DOD will track, as a metric, the participation of nontraditional defense contractors. DOD also included a draft audit policy in the revised guide. According to DOD officials, the impetus for including a draft audit policy came from two DOD Inspector General reports. The first, a 1997 report, questioned the adequacy of audit coverage on DARPA’s “other transactions” for research. Although “other transactions” agreements for research included an audit clause, the report noted that agency officials intended to require audits only if they suspected fraud. The Inspector General argued that without final cost audits, agency officials could not ensure compliance with the statutory requirement pertaining to cost-sharing provisions. In a 1999 follow-up study on cost-sharing, the Inspector General raised similar concerns about prototype projects and included recommendations regarding audit policy for “other transactions” for prototype projects. On August 27, 2002, DOD issued a final rule codifying the definition of a nontraditional defense contractor and setting forth the conditions for using Section 845 agreements consistent with Section 803 of the National Defense Authorization Act for Fiscal Year 2001. The notice accompanying the final rule stated that the audit policy is being discussed and will be addressed by a separate rule. After exploring a number of performance indicators for Section 845 agreements, DOD selected one quantitative performance metric—the extent of nontraditional contractor participation—which is tracked by the Office of Defense Procurement. Officials believe that this metric is key because involving firms that do not traditionally do business with DOD increases DOD’s opportunity to leverage commercial technology investments and to take advantage of commercial business processes, such as using an integrated team approach rather than a traditional prime- subcontractor structure. Congress also has encouraged the participation of commercial firms in the development of defense systems and has recognized the critical contributions of nontraditional participants in areas such as biotechnology and pharmaceuticals in today’s national security environment. DOD contracted with RAND, a nonprofit institution, for a study to assess the overall effectiveness of the Section 845 acquisition approach and to explore the possibility of using additional metrics. In addition to this effort, a DOD working group, composed of officials from across DOD, considered the types of metrics that could be used to assess the effectiveness of Section 845 agreements. These two efforts identified several difficulties, as follows: Traditional metrics—such as cost growth, schedule slips, and performance shortfalls—are inappropriate for Section 845 projects that are inherently risky. A “path not taken” cannot be measured; that is, when a Section 845 agreement is used rather than a procurement contract, a statistical comparison between the two acquisition approaches cannot be made. Too many variables and too few Section 845 agreements would limit the results of a quantitative analysis. Few Section 845 projects have been completed, limiting the results to date. RAND concluded that important new technological capabilities—a desirable benefit of “other transaction” agreements—mostly come from segments of major firms that formerly focused on commercial projects but are now willing to apply their skills to the development of military prototypes. RAND also pointed out that there are other benefits associated with the flexibility inherent in this authority. For example, the flexibility to change project plans based on mutual agreement between DOD and industry managers, with minimal documentation or administrative burden, provides more powerful opportunities to cope with the problems and opportunities that occur when developing new systems and components. However, RAND emphasized the difficulties in developing quantifiable metrics that would be accepted as credible. In its effort to focus on collecting information on nontraditional contractors, DOD uses the Report of Other Transactions for Prototype Projects (DD Form 2759), which is completed by the agreements officer. (App. II contains a sample form.) According to the DOD guide, when funding actions are taken, the agreements officer must record information on whether the prime or subcontractor awardees are traditional contractors, nontraditional defense contractors, or non-profit organizations. The agreements officer also must record the names and addresses of significant nontraditional defense contractors. The summary information is sent to DOD’s Office of Defense Procurement, where it is aggregated. According to the DD 2759 reports for Section 845 agreements awarded in fiscal year 2001, 16 nontraditional prime contractors and 29 significant nontraditional subcontractors participated in a total of 61 agreements. Nontraditional participants included commercial business units of U.S. traditional firms as well as foreign corporations. Congress requires DOD to report annually on all “other transaction” projects—for research as well as prototypes—awarded in the preceding fiscal year. While the Section 845 portion of the report addresses the issues set forth in the congressional reporting requirement, it does not present the number of nontraditional contractors in a clear, straightforward format, such as a summary table. Because information on nontraditional participants—DOD’s key performance metric—is not summarized, it is difficult for Congress to assess how successful DOD has been in achieving this metric. The annual report includes a 1- or 2-page summary of each project that discusses (1) government and contractor cost contributions, (2) the reason for using the Section 845 authority, and (3) how the agreement contributed to a broadening of the technology base or fostered relationships and practices that support U.S. national security interests. In the fiscal year 2001 report, these individual summaries totaled 152 pages. In a fiscal year 2000 supplemental report to Congress, DOD did present a narrative summary of the number of nontraditional contractors; however, this was the only occasion when the information was clearly imparted. DOD officials stated that they are reluctant to add another reporting element and that the current report format meets congressional requirements. They added that they view the number of nontraditional contractors as secondary to the agreement-level information presented in the report. DOD also is not regularly reporting on or assessing the benefits derived from completed Section 845 agreements. In 1996, the Under Secretary of Defense for Acquisition and Technology requested a comparison of the benefits and drawbacks of completed agreements with the expected benefits at the time of award. However, this attempt to compile “lessons learned” was abandoned because many DOD officials believed that the results were parochial and not useful across the department. A draft version of the current Section 845 guide included a requirement for an assessment of completed agreements, but the requirement was not incorporated in the final version because DOD officials believe that another reporting requirement was not likely to produce a meaningful assessment of Section 845 results. DOD officials commented that the law only requires them to report on projects awarded in the previous fiscal year. They acknowledged, however, that periodic assessments of the benefits derived from completed agreements could be useful. By updating the Section 845 guide and requiring the number of nontraditional contractors to be measured as a performance metric, DOD has implemented our April 2000 recommendations. However, the reporting on the benefits derived from this alternative acquisition approach could be improved. A summary table in the annual report to Congress, setting forth the number of nontraditional contractors brought in under Section 845 agreements during the preceding year, would provide a clear picture of the extent to which DOD’s performance metric is being achieved. The current report format, consisting of summaries of each agreement, requires the reader to review each summary sheet in order to determine how the Section 845 authority was used—including the number of nontraditional contractors participating in the agreement. Thus, its usefulness to Congress is limited. Further, in the absence of regular assessments of the benefits derived from completed projects, DOD and the Congress lack vital information on the results the government is deriving from this flexible procurement strategy. The experience that DOD has gained from the use of Section 845 authority can be useful to Congress as it makes decisions about subsequent extensions of this authority to DOD and in future congressional deliberations. We recommend that the Secretary of Defense incorporate in the annual report to Congress summary information on the extent of nontraditional contractor participation and periodically report to Congress the results of studies on the benefits derived from completed Section 845 projects, including how key private sector participants contributed to the results. In written comments on a draft of this report, DOD agreed to incorporate in the summary of the annual report to Congress information on the number of new agreements and to break out the reasons for using the authority. However, DOD did not agree with our recommendation to include the number of nontraditional contractors in the annual report, stating that a raw count does not necessarily indicate the extent of nontraditional contractor participation and that it is secondary information derived from a separate reporting system. We agree that a raw number alone can be misleading. However, we do not understand why DOD is reluctant to publish the total number of nontraditional contractors—along with the other information to be reported—when those numbers are being internally collected and when this is the key performance metric DOD has established. Including the number of nontraditional contractors, along with the other information DOD has agreed to provide, would give Congress a more complete basis on which to assess the achievements gained through the use of Section 845 authority. DOD concurred with our second recommendation but stated that it would oppose the establishment of a regular reporting requirement. We are not advocating a new reporting requirement; however, we believe that periodic assessments of completed Section 845 projects would provide Congress useful information on the benefits the department is deriving from use of this authority. To assess the comprehensiveness of DOD’s new Section 845 guide, we compared it to the November 1998 guide that was in effect during our prior review. To determine the adequacy and usefulness of the revised guide and the performance metrics used, we interviewed officials in the Office of Defense Procurement and in the Office of Acquisition Initiatives—Office of the Under Secretary of Defense for Acquisition, Technology and Logistics; Washington Headquarters Services’ Directorate for Information Operations and Reports; the headquarters offices of the Army, Navy, and Air Force; DARPA; and NIMA. We also reviewed reports prepared by DOD’s Office of the Inspector General, RAND, and GAO. In addition, we reviewed various directives, memorandums, publications, correspondence, and legislation concerning Section 845 authority. To determine the number and value of fiscal year 2001 Section 845 agreements and the number of agreements having nontraditional defense contractors, we analyzed data compiled by the Office of Defense Procurement. In addition, at each military service and DARPA, we reviewed the Reports of Other Transactions for Prototype Projects (DD Form 2759) for all agreements awarded in fiscal year 2001. We did not validate or verify the information reported on these forms, including whether the cited nontraditional defense contractors met the definition in Section 803 of the National Defense Authorization Act for Fiscal Year 2001. To determine whether Congress is receiving adequate information on the number of nontraditional defense contractors participating in Section 845 agreements and whether DOD is assessing the benefits derived from completed projects, we reviewed the Section 845 portion of the annual reports for fiscal years 1999 through 2001 and the supplemental reports provided to Congress in fiscal years 1999 and 2000. We also reviewed DOD’s guidance and memorandums and held discussions with officials from the Office of Defense Procurement. We conducted our review between April and August 2002 in accordance with generally accepted government auditing standards. We are sending copies of this report to interested congressional committees; the Secretaries of Defense, the Army, the Navy, and the Air Force; the Director, DARPA; the Director, NIMA; and the Director, Office of Management and Budget. We will also make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. Please contact me at (202) 512-4841 or Michele Mackin at (202) 512-4309 if you have any questions regarding this report. Other major contributors to this report were William M. McPhail, Rosa M. Johnson, and Kenneth E. Patton.

In April 2000, GAO reported on the Department of Defense's (DOD) use of Section 845 agreements, also referred to as "other transactions" for prototype projects. These are transactions other than contracts, grants, or cooperative agreements that generally are not subject to federal laws and regulations applicable to procurement contracts. In December 2000, DOD revised its Section 845 guide. The guide specifies when Section 845 agreements may be used and provides criteria for tailoring terms and conditions for each agreement. Officials from the military services and defense agencies have found the new guide useful and a significant improvement over the prior version. The Secretary of Defense has required a metric--the number of participating nontraditional defense contractors--which is measurable and directly related to each agreement. This metric is tracked and reported internally. DOD explored additional metrics, but concluded that the number of nontraditional contractors was the only one that was quantifiable and tied directly to Section 845 outcomes. DOD's annual report to Congress on Section 845 agreements consists of summaries on each agreement. However, the key metric--the number of nontraditional contractors--is not clearly presented in these reports, making it difficult to gauge DOD's progress in achieving success on this objective. Further, DOD is not regularly assessing reporting on the benefits derived from completed Section 845 projects. In the absence of such assessments, congressional and DOD decision makers lack a vital piece of information that would help them determine whether this flexible procurement authority is achieving expecting results.

The Navy has reported that more than $8.5 billion of Navy resources was applied in fiscal year 1996 to maintenance programs in support of fleet ships and aircraft. Each type of “platform,”—surface ships, submarines, aircraft carriers, and aircraft—has a separate maintenance infrastructure. Maintenance is done at three different levels—organizational, intermediate, and depot—depending on the nature and complexity of the work required. Organizational maintenance is done by military personnel on board ships or at aircraft squadrons. While at sea, intermediate maintenance on large ships such as aircraft carriers and tenders is done by military personnel; ashore, intermediate maintenance is done by military and civilian personnel at submarine refit facilities and aircraft and shore intermediate maintenance activities. Depot-level maintenance is done mostly by civilian personnel at aviation depots and shipyards. In 1996, the Navy had over 21,000 military and 42,000 civilians participating in maintenance activities at the intermediate and depot levels. In addition, the Navy has reported that up to 40 percent of depot-level maintenance is outsourced to private companies. In response to force structure reductions since the mid-1980s and subsequent defense planning guidance to reduce excess maintenance infrastructure, the Chief of Naval Operations (CNO), early in 1993, tasked the commanders of the Atlantic and Pacific Fleets to develop a strategy for streamlining and consolidating maintenance functions. This led to the Navy establishing the RM Program in March 1994. The Navy’s RM Program efforts have been focused on reducing excess maintenance infrastructure. However, the program has other objectives such as improving maintenance processes, integrating supply support and maintenance functions, and providing compatible data systems across the three maintenance levels. The program was to be implemented in three overlapping phases during fiscal years 1995-99. Since the RM Program began, the number of Navy ships and aircraft has continued to decline. For example, the Navy projects that by the end of fiscal year 1999, it will have 186 fewer aircraft and 22 fewer ships to maintain than in 1996. During the same period, the maintenance budget for ships and aircraft is also expected to be reduced to about $7.5 billion (in fiscal year 1996 dollars), a decrease of about $1 billion. The Navy has made substantial progress in implementing the infrastructure streamlining objective of the RM Program through such efforts as establishing a management structure, a phased execution plan, and a process for realigning and reducing its maintenance infrastructure. It also identified 102 initiatives aimed at regionalizing, consolidating, and streamlining the maintenance infrastructure and achieving savings. Implementation of the program has not been as rapid as predicted, however, and milestones may not be met. The Navy has established a management structure for planning and implementing the RM Program. The structure is linked at the CNO level and includes committees, systems commands, the fleets, and various quality boards and other groups. For example, the management structure within the CNO includes an Executive Steering Committee that provides overall program guidance and direction. This committee chartered the Fleet Support Quality Management Board to develop the transition strategy for moving to regional maintenance. Through this Board, regional maintenance was planned and developed using focused working groups. A Regional Maintenance Implementation Board (RMIB) was established to coordinate among the Pacific and Atlantic Fleets, the systems commands, and CNO-level units. The Board is co-chaired by the Fleet Maintenance Officers, who have key leadership responsibilities to implement regional maintenance. This Board meets on a regular basis to address regional maintenance and other issues. Each of the systems commands, the Atlantic Fleet, and the Pacific Fleet report separately to the CNO. Through this management structure, the Navy has developed concepts, guidance, fleet business plans, and milestones for the RM Program infrastructure streamlining objective. For example, the fleets developed program guidelines for establishing regional repair centers, adopted a business-case analysis approach for evaluating candidate activities for consolidation, and formulated cost templates for measuring the monetary impacts of consolidations. In March 1994, the CNO approved a three-phased execution plan that assigned the following primary tasks in each phase: Phase 1: Minimize intermediate-level redundant capacity through process improvements and resource sharing, and develop prototype centers of excellence, called Regional Repair Centers. Implement phase during fiscal years 1995-99. Phase 2: Integrate intermediate- and depot-level activities and establish Regional Maintenance Centers (RMC), consisting of a confederation of Regional Repair Centers. Implement phase during fiscal years 1996-99. Phase 3: Conduct fleet maintenance using a single maintenance process supported by common business and production practices. Implement phase during fiscal years 1997-99. As part of phase 2, the Mid-Atlantic and Northwest RMCs, and later six others, were established—a total of four in each fleet. The Navy has included one or more of the states around where the centers are located and where there are Navy maintenance activities into areas it refers to as RM regions (see fig. 1). Each RM region established an executive steering committee for maintenance, comprised of the commanders and maintenance managers of activities in the region and chaired by the region’s RMC commander. These committees have chartered process action teams to identify which activities in the region should be evaluated for consolidation. The two fleets have developed regional maintenance business plans, including initiatives and estimates of savings to be achieved in each of their respective regions, and the systems commands have added their own initiatives with estimates of savings. Although the Navy began implementation as planned, phase 2 has been redefined, and phase 3 has been delayed. As a result, implementation is taking longer than anticipated. Full implementation, initially projected for fiscal year 1999, is currently projected for fiscal year 2000 and could take longer. According to Navy officials, there have been delays in implementing the program because many of the issues involved are complex and require extensive studies and approvals. For example, the possible consolidation of calibration laboratories in the Northwest region was identified as an initiative in 1994, but it has taken 3 years and multiple studies to determine which activity would do the calibration work and whether it would be done using government or contract employees. Also, implementation among the regions has been uneven. While all regions have made some progress, the Mid-Atlantic region has led the way in establishing the program and piloting regional maintenance initiatives to achieve savings. For example, the Mid-Atlantic region has 18 (33 percent) of the 55 initiatives being implemented from 1994 to 1997, including consolidating three shore intermediate maintenance activities (SIMA) into one organization, consolidating calibration and material testing laboratories, and establishing other regional repair centers. It also initiated the fleet business plans and guidance for regional repair centers later used by other regions. By contrast, the Hawaii and Northwest regions had implemented eight initiatives each. The Northwest region established regional repair centers for pumps, periscopes, gas turbine engines, and eliminated a military construction project; and both the Northwest and Hawaii regions have consolidated nuclear regional maintenance work. Because of problems with the Navy’s financial information system and other coordination issues, in 1996 phase 2 of the execution plan was divided into a three-step process. During the first step, ship intermediate- and depot-level maintenance were to be consolidated; aircraft intermediate- and depot-level maintenance activities were to be collocated; and ship and aircraft maintenance consolidations were to take place where logical. During the second step, ship intermediate- and depot-level maintenance planning and engineering functions were to be consolidated into ship-planning and engineering centers, reducing the number of planning and other positions needed. As of July 1997, the third step had not been approved, and delays in phase 2 have postponed the approval of phase 3. In a regional maintenance briefing in May 1997, the Deputy Chief of Naval Operations (Logistics) briefed the CNO that execution of the RM Program would be completed in fiscal year 2000. Since then, in July 1997, the CNO noted that although phase 2 was on track, challenges remained, much still had to be done, and efforts must be accelerated. According to Navy officials, the tendency is to be optimistic in establishing milestones for such programs and organizational realignments are particularly difficult to accomplish. They also said that developing regional maintenance will not be completed in 1999 as planned, but will be a continuous process long after fiscal year 2000 as new initiatives and refinements to existing maintenance processes are identified. Through fiscal year 1996, 102 initiatives with projected savings of $944 million had been identified for the program. Of the 102, the Navy estimated that it would achieve net savings of $198 million through implementation of 55 initiatives during fiscal years 1994-97 and that these projects would continue to provide savings in fiscal years 1998-2001 amounting to $272 million, or a total of about $470 million (see table 1). It planned to implement 47 more between 1998 and 2001. Program savings are not being achieved at the levels the Navy originally estimated. According to fleet maintenance officials, initiatives that have been implemented are mostly the less controversial projects that are easier to implement and a few complex consolidations. They said they are proceeding slowly because they believe Navy managers should be encouraged, rather than forced, to accept regional maintenance. The Navy has also reduced its estimates of savings for a number of initiatives. For example, the savings estimate for an initiative to reduce the overhaul for certain diesel engines was reduced from $5.4 million annually to $1.2 million in fiscal year 1996 and $900,000 in fiscal years 1997-99. According to officials at the maintenance facility responsible for the repair of these engines, overhauls have not occurred at the anticipated rate per year because a maintenance process change reduced the requirements. In the Northwest region, a delay in a project to consolidate calibration functions has delayed the realization of potential savings. Also, an initiative to consolidate ship repair planning and engineering functions at the Naval Sea Systems Command is not occurring as expected, delaying planned reductions-in-force actions and affecting up to $92 million in RM Program savings projected to accrue between fiscal year 1998 and 2001. Although the Navy has incorporated its $944 million in estimated savings from the RM Program into its projected maintenance budget, actual RM costs and related savings are not systematically tracked to determine whether they have actually been accrued. The Navy’s accounting system, like all Department of Defense (DOD) accounting systems, tracks expenses and disbursements but not savings, and the Navy did not establish an independent system to track RM costs and related savings. CNO officials told us they recognized the need for such RM Program data but that efforts to collect it can involve many Navy activities and would be so labor intensive that there are no current plans to do so. The Navy Audit Service said it is in the process of evaluating RM savings through baseline studies of initiatives and follow-up studies 1 year after implementation. Only one of the studies has been completed. It showed that savings achieved through the consolidation of activities in an electric motor rewind shop in the Mid-Atlantic region was about $4.4 million a year, or about 44 percent of the $9.9 million baseline cost each year prior to consolidation. Baseline costs were being evaluated for some other selected projects so that post consolidation studies could be done. While the Naval Audit Service is looking at costs before and after consolidation, it is not tracking RM savings into budget and accounting records to determine if they have actually been accrued. Actual savings achieved through the RM Program have been questioned within OSD. An OSD Maintenance Policy, Programs, and Resources office study of the RM Program concluded that savings had been achieved from the restructuring of maintenance activities, but that some of the savings might have been the result of the four base realignment and closure rounds and other actions. The OSD Comptroller has gone further and concluded in 1996 that savings projected from the RM Program for fiscal years 1994 through 1997 have not materialized as anticipated and are not evident in actual Navy budgets submitted to Congress each year. In a program review for fiscal year 1995, the Navy decreased its planned fiscal year 1995-99 budgets for operations and maintenance by $1.28 billion, anticipating that savings from regionalizing maintenance would offset the impact of the reductions (see table 2). According to Navy officials in each fleet and in the CNO’s Supportability, Maintenance, and Modernization Division, RM savings did not materialize to cover the amounts taken from the programs, and the reductions had to be made up in other ways. While program budgets were reduced by $1.28 billion, the commands and fleets did not have records available showing how the reductions were finally absorbed. According to the OSD Comptroller, evidence indicates that other factors such as base closures, force structure reductions, directed civilian drawdowns and the general reduction in depot workloads resulting from force structure cuts during fiscal years 1992-97, have accounted for the actual reductions in costs. The OSD Comptroller stated in a November 1996 budget memorandum that the RM Program has not progressed enough to reap projected savings and that further review of regional maintenance might be in order to ensure savings occur and readiness is not degraded as a result of the reductions. In August 1997, OSD Comptroller officials said that savings anticipated from the RM Program have not materialized; in fiscal years 1995 and 1996 regional maintenance did not progress much past isolated, small and informal tests; and in fiscal years 1997 and 1998, savings were offset by the need to finance construction of new facilities in Navy SIMAs. The officials noted that the Navy has recently requested additional funding for depot maintenance and that more requests for additional funding were anticipated. They further noted that depot maintenance budgets in a number of areas have had to be increased over the Navy’s proposed budget levels. For example, in fiscal year 1995, rates were increased significantly over the Navy’s proposed budget levels to ensure full costs were recouped; and in fiscal year 1998, Navy air depot budgets were increased each year in the Future Years Defense Program, with over $200 million added in both fiscal years 1998 and 1999. According to Atlantic Fleet officials, they have thus far been able to absorb the reductions in planned budgets for ships with no impact on readiness. They said this is because they are focusing on fixing specific problems, which reduces the total amount of maintenance to be done, rather than performing entire scheduled depot-level maintenance overhauls. They also said that an initiative started in fiscal year 1995, to better balance expected naval shipyard workloads with the available workforce, has resulted in improved operating results for naval shipyards. CNO officials acknowledged that fixing specific maintenance problems rather than overhauling entire components would likely result in maintenance cost reductions. However, they were concerned that by using this approach the overall material condition of ships might be adversely affected over the long term, but noted that the Navy currently does not have adequate measures of material condition and its relationship to readiness. The Navy has many opportunities to build on its maintenance infrastructure streamlining progress. The Navy anticipates that the largest savings will accrue during fiscal years 1998-2001 (see fig. 2); that is, of the estimated $944 million its 102 initiatives are projected to save, $746 million would accrue during that period. The Atlantic and Pacific Fleets have identified additional opportunities for savings, and in 1997 added 34 more initiatives to their regional maintenance business plans. They have not estimated the amounts of savings from many of these initiatives, however. During our review, we identified additional opportunities for infrastructure reductions in two regions with potential savings of up to $48 million. These included potential annual savings of $26 million based on maintenance infrastructure consolidations in the Hawaii and Northwest regions, and $22 million in one-time savings by transferring work at the SIMA, Everett, Washington, to other existing shops and eliminating a military construction project and two barge overhauls. The Navy has identified other potential regional maintenance opportunities that need to be studied. In the regions we reviewed, for example, the Northwest region in June 1994 identified 41 areas of redundant capabilities, but still has not studied many of them to determine whether initiatives could be developed to reduce unnecessary infrastructure and achieve savings. In a February 1997 Regional Maintenance Implementation Board meeting, in an effort to spur progress, the fleets were tasked to identify regional maintenance consolidation initiatives that could be considered. In May 1997, the RM Program manager for the Hawaii region told us his region had followed the lead of the earlier Northwest region project and compiled a comprehensive inventory of regional maintenance capabilities to be used to help identify future initiatives. In addition, the Mid-Atlantic region, in its 1997 update to the fleet business plans, identified 29 new savings initiatives. The update did not determine when about half of them would be implemented or estimate the savings that could be achieved. These initiatives include establishing two regional repair centers—one for special tool design and manufacture and another for sheet metal component fabrication—and a regional training support center. Other regions identified a total of five additional initiatives. In the Hawaii and Northwest regions, we identified three examples of opportunities to consolidate intermediate- and depot-level maintenance activities that were not in current business plans. We observed common industrial facilities, called backshops, at six activities. The backshops consisted of electrical and electronic, machining and metal-forming shops and material testing laboratories. At most of these backshops, we were provided estimates showing unused infrastructure—facilities and equipment. We estimated that consolidating and reducing excess capacity in these shops could save up to $48 million—from about $2 million to $26 million annually and about $22 million in one-time savings. (See apps. I and II for details of our analysis.) These are not budget quality estimates, however, because complete and compatible data on the facilities were not available, alternative consolidation arrangements are possible, and there was no consensus on what workforce savings could be achieved. Both regions had considerably more maintenance capacity than workload, particularly at the shipyards. For example, we observed first shift operations in a total of 25 backshops at 6 activities in the 2 regions. Although usually the busiest shift, supervisors estimated that on average, this shift was operating at about 30-percent utilization, with a range of between 4 and 70 percent. Of the 25 shops, 16 had a second shift and only 7 had a third shift. Estimates of utilization during second shifts were markedly lower, an average of 12 percent and a range of from about 1 to 39 percent. Some shop supervisors noted their shops had supported several times the number of workers in the 1980s than were currently employed. Navy data indicated that excess facilities and equipment capacity were due to reductions in labor hours and numbers of employees at these shipyards. For example, Navy data on direct labor hours at the Pearl Harbor Naval Shipyard showed a reduction from about 6.1 million to 3.2 million (48 percent), and the Puget Sound Naval Shipyard showed a reduction from 12.0 million to 11.3 million (6 percent) between 1989 and 1996. (See fig. 3.) At the same time, employment was reduced from 6,044 to 2,879 ( 52 percent) at the Pearl Harbor Naval Shipyard, and from 12,240 to 9,424 (23 percent) at the Puget Sound Naval Shipyard. Our specific findings, suggestions for consolidations, and estimated savings for the Hawaii and Northwest regions are summarized in appendix I, with detailed data on labor, facilities, and costs provided in appendix II. The Navy has barriers to overcome before it can fully achieve expected infrastructure reduction savings and other RM Program objectives. According to the Navy, in May 1996, 54 percent of the $944 million in projected savings would come from projects considered high risk. Initiatives were considered high risk to achieving expected savings when a large number of organizations and funding accounts were involved and/or they required significant manpower reductions. For example, 1 high-risk initiative, to save $4 million a year by consolidating the calibration functions in the Mid-Atlantic region, involved 22 activities and 3 funding sources. The Navy recognizes that parochial and institutional resistance to the RM Program’s objectives and other issues will be difficult to resolve. The biggest barrier to overcome may be resistance to initiatives that eliminate organizations, reduce jobs and promotions, or reduce control over resources. Other barriers to integrating intermediate- with depot-level capabilities are (1) the lack of management visibility over all maintenance-related costs; (2) multiple, unconnected management information systems that do not provide adequate data for regional maintenance planning and decision-making; and (3) the large number of shore duty intermediate-level maintenance positions needed to support the Navy’s sea-to-shore rotation program compared to a lesser number needed to perform the work. The Navy has RM Program working groups and committees in place to address some of these issues. According to Navy officials, these issues are intertwined and some planned resolutions would be subject to legal and congressional review. The RM Program requires managers to forgo the traditional platform-oriented structure and substantially reduce or close some maintenance activities as work is eliminated or reassigned. Many commands involved in the RM Program have chains of command that are independent of each other, and visible commitment by the CNO is critical to program implementation. For example, reductions will result in fewer commands and promotion opportunities and a need to share resources, prioritize work, and reassign responsibilities. According to fleet officials, this organizational resistance may be the greatest inhibitor to RM progress. According to fleet maintenance officials, overcoming resistance to organizational changes is difficult. The fleets’ type commanders—shore-based commanders responsible for supporting the fleet, including providing maintenance for aircraft, surface ships, and submarines—did not fully support the proposed changes. These fleet officials told us that since the type commanders are responsible for the intermediate maintenance facilities for their respective platforms, they may view the regionalization of maintenance as a loss of control or responsibility, including a potential loss of their ability to assure readiness of their assigned units. These officials also noted that the Fleet Maintenance Officers’ influence over the type commanders is limited. More progress has been made in the Mid-Atlantic region, where (1) the fleet command, type commanders, and regional maintenance officials are collocated; (2) the program has had strong support from the fleet commander; and (3) the Fleet Maintenance Officer initially started regional maintenance. Also, according to CNO and fleet officials, RM initiatives that cut across major commands may prove difficult to achieve, particularly if they involve loss of control or responsibility. Initiatives to integrate and consolidate depot-level maintenance activities with intermediate-level maintenance activities require the cooperation and support of most of the Navy’s major commands and the CNO. Naval Air Systems commanders, Naval Sea Systems commanders, and fleet and their subordinate type commanders all have a stake in how RM initiatives are implemented and how the initiatives will affect their particular activities and staffing. Various representatives of the activities, regions, fleets, and headquarters offices expressed concern that CNO-level managers had not decisively endorsed regional maintenance and this had caused problems in participation, particularly outside the surface ship community. Fleet and headquarters officials also noted that the Naval Air Systems command activities have had limited involvement in the program primarily because they consider their maintenance systems different and airworthiness a critical criterion that surface ship maintenance activities are not used to. Similarly, submarine platform officials voiced their concerns to us about their strict maintenance requirements and safety standards. Fleet and headquarters officials further noted that many commands involved in the RM Program have chains of command that are independent of each other up to the CNO. Therefore, visible commitment by the CNO is critical to implementing the RM Program, as this involvement accelerates the provision of resources and the coordination needed for efficient and effective program implementation. For example, there was a significant increase in activity after the CNO directed the Hawaii region to implement a pilot project to study the consolidation of the Pearl Harbor shipyard with the naval intermediate maintenance facility and to complete the integration by September 30, 1998. In another streamlining effort, regionalizing base operations, the CNO has provided crucial support. For example, in September 1995, the CNO approved a major Navy-wide infrastructure reduction initiative to (1) reduce the number of activities that own and manage shore installations, (2) regionalize installation management functions where it makes sense, and (3) find excesses, duplications and redundancies among the numerous tenants on bases, using San Diego and Jacksonville as pilot locations. The San Diego project is to be completed as soon as possible, but no later than fiscal year 1999. According to these officials and the information provided, this effort has affected many activities, commands and the way business is conducted; therefore, the support of the CNO was crucial for accomplishing the components of the initiative. The Navy has identified the need to provide visibility over all maintenance-related costs as an issue in implementing the RM Program. The Navy has also identified a need for a flexible and responsive managerial accounting system because the Navy’s current financial system does not provide the data needed for informed decision-making. For example, the Navy has in some cases increased capacity in its shore intermediate activities’ backshops without regard to the fact that a nearby shipyard had excess capacity in similar backshops. Fleet maintenance officials said efforts to develop full cost visibility and the necessary financial system are underway. According to OSD Comptroller officials, a central issue is that Navy depot-level maintenance activities are funded under the Navy Working Capital Fund (formerly the Defense Business Operations Fund), and intermediate-level maintenance facilities are funded directly from the appropriations accounts. One of the basic tenets of the Working Capital Fund financial structure is to focus on total cost visibility and full cost recovery for depot-level maintenance activities. Operating under this tenet, managers of the fund’s activities are to be held accountable for the costs of all the resources that they manage, and military customers are to pay the full costs of the maintenance work performed. In contrast to the full costing visibility of the Navy’s depot-level maintenance, intermediate-level maintenance activities are not operated using the Working Capital Fund concept. Military customers at the intermediate activities are usually only charged the incremental costs of the work performed, such as the costs of materials. Most of their other costs are subsumed in the mission funded operating budget and have little to no visibility. The mission funded operating budget includes the costs of civilian personnel and all overhead type costs to include real property maintenance and utilities. In addition, these intermediate maintenance activities are manned with military personnel, and their personnel costs are directly borne by the Military Personnel Appropriation and are not costed as part of the repair work they perform. By excluding these costs, the full costs of products and services are concealed, and customers see the work done at the intermediate activities as significantly less expensive than the work done at the shipyards. As a result, there is an incentive for customers to use intermediate facilities to the maximum extent possible. For example, according to officials in the Northwest region, sailors from the aircraft intermediate-level maintenance activity at Whidbey Island Naval Air Station fabricated components at Whidbey Island and traveled to the shipyard, a distance of 35 miles, to install the components on a ship when the fabrication work could have been done at the underutilized shipyard sheet metal shop. According to OSD Comptroller officials, until the Navy can accumulate complete, comparable, and reliable data on the costs of its intermediate and depot-level maintenance facilities, decisions on how best to use and integrate these facilities will continue to be impaired. Fleet officials told us the Navy has recognized this problem and has pilot projects underway to obtain total cost visibility data at the job-order level in regional repair centers. They said that experience with regional repair centers that have been established under the RM Program has shown that such efforts are complicated, particularly by the problems associated with obtaining the required data from multiple systems. As a result, accumulating reliable cost data will be difficult, and require dual systems for some time, thereby reducing potential savings. The Navy does not have well-defined and consistent data on its maintenance shops’ capacity, capability, workforce, and current and projected workloads. Without such data, the Navy cannot systematically identify potential regional consolidations and related savings estimates. The Navy has recognized that it lacks compatible and interconnected maintenance information systems that could identify similar maintenance capabilities across activities. Although the Navy has made some attempts to address this issue, its systems do not yet collect the critical information needed to identify excess capacity. In the Northwest and Hawaii regions, incomplete and unreliable data has hindered the Navy’s ability to identify excess maintenance capacity. For example, data is not available or compatible within and among activities in such areas as shops’ capacity, productivity, labor efficiency, workloads, and equipment utilization rates. According to the fleet business plans, having separate maintenance infrastructures for ships, submarines, aircraft carriers, and aircraft has fostered the development of unique maintenance management information systems for the different platforms and levels of maintenance. A Northwest region process action team has studied the issue and found a wide disparity in the information available among its regional activities. It developed a strategic implementation plan to establish, first, an interconnection among information systems; second, an ability for these systems to exchange data; and third, the ability to manage, control, and use the data. Although phases 1 and 2 were to have been implemented by fiscal years 1995 and 1996, respectively, as of August 1997, the team was still in phase 1. According to CNO and fleet officials involved in the establishment of an automated information system, it is critical to have a system that allows for the exchange of technical and management data among various maintenance activities. In one instance where several databases were evaluated, none provided sufficient common data to determine capabilities across activities. Although the Navy contracted for the development of a concept model that would recognize capabilities among activities in two regions, it concluded that the model developed required intensive data collection and was not cost-effective to implement. Navy fleet officials said that some progress has been made in providing access and linkages of data among platforms and RM regions, but efforts have been delayed because activities have not made it a priority or do not have the computer equipment needed. The Navy’s need to support requirements other than workload at shore intermediate-level maintenance facilities can hinder RM regions’ efforts to reduce military positions. For example, these facilities need positions to support maintenance workload, the Battle Force Intermediate Maintenance Activity (BFIMA) program and sea-to-shore rotation requirements. These facilities’ positions are also used for personnel identified as excess to the requirements or on limited duty. According to the Navy, however, the number of shore intermediate-level positions should not be less than BFIMA program requirements and should not exceed sea-to-shore rotation requirements. As of March 1996, the Navy had 12,668 shore intermediate-level positions. The Navy needed only 11,704 of these positions to support the maintenance workload. Thus, it had an excess of 964 positions. Also, the workforce is unevenly distributed across the regions. For example, three regions had 1,409 positions that exceeded their maintenance workload requirements, while four regions had 445 positions less than their projected maintenance workload requirements. The Navy identified a need for 4,649 positions to support the BFIMA program; thus, the 12,668 existing positions far exceed BFIMA requirements. On the other hand, the number of intermediate-level maintenance positions desired to support the sea-to-shore rotation program far exceeds the number needed to support the maintenance workload. In March 1996, the Navy reported to the CNO that 19,819 shore intermediate-level positions were desired to support sea-to-shore rotation. Thus, it had a shortfall of 7,151 positions. This shortfall acts as a disincentive for the Navy to reduce the number of shore intermediate-level positions. The Navy also uses positions at intermediate-level maintenance facilities for personnel awaiting reassignment or on limited duty. This practice further hinders efforts to reduce excess maintenance capacity. For example, the Navy indicated that of the Southwest region’s 753 excess positions, 335 were positions for sailors displaced by the decommissioning of a Navy tender. Sailors affected by this decommissioning are typically waiting for funding for permanent changes of station or reassignment. Also, in June 1997, the intermediate-level maintenance facility at Everett, Washington, and its detachment at Bremerton, Washington, reported a workforce of 521, of which 84 (over 16 percent) were on limited duty. The facility had recommended a reduction of its detachment workforce of 91 positions—from 197 to 106. An efficiency review to determine the appropriate number of staff has been done but was not finalized during our review. Although the Navy has made substantial progress in establishing a structured RM Program to achieve its infrastructure streamlining objective, it has reported only limited progress in accruing savings from the program. Thus far, the reported savings have not materialized as anticipated because projects have been changed and delayed. Further, the accuracy of claimed savings is questionable because they are not tracked and verified. Consequently, the Navy’s actual savings may be far less than the $944 million it originally projected. They also may be achieved much later than expected. These conditions could negatively affect maintenance programs, the overall material readiness of ships and aircraft, or future fleet readiness, since reductions have already been made to spending plans in anticipation of savings. Nonetheless, the Navy can still achieve significant savings by studying and, where appropriate, implementing other initiatives that can yield savings without impacting readiness. To implement such initiatives, it must also resolve difficult organizational, financial, management information system, and sea-to-shore rotation issues that have slowed the RM Program’s progress. Further, overcoming resistance to change, perhaps the greatest inhibitor to RM Program implementation, will require continued high-level commitment, cooperation, and coordination from the CNO, the fleet, and type and systems commanders, to ensure that regional initiatives reach fruition and achieve the savings projected. The Navy’s RM Program is extremely important to improving the effectiveness and efficiency of its maintenance activities and we encourage DOD to move forward as quickly as possible. If successful, the program can result in a more streamlined, regionalized maintenance program. As we stated in our high-risk report on the defense infrastructure, breaking down cultural resistance to change, overcoming parochialism, and setting forth a clear framework for a reduced infrastructure are key to effectively achieving savings. We recommend that the Secretary of Defense direct the Secretary of the Navy to annually report on the RM Program initiatives identified, savings achieved that have been verified in Navy budget and accounting records, and the progress made to overcome the barriers to achieving infrastructure reductions and savings. We also recommend that program implementation plans be established and tied to milestones, with regular reporting to the CNO. DOD’s written comments on the draft of this report are presented in appendix IV. DOD stated that the Navy has many actions underway to address the issues contained in this report. Specifically, DOD noted that the RM Program was started to help the Navy perform maintenance more efficiently, not to offset specific budget reductions. We agree that the program was designed to generate greater efficiencies; however, as noted in our report, it was also expected to generate significant cost savings. We revised our report to clarify that our work focused on the infrastructure streamlining objective, which has been the program’s principal focus thus far and to which savings projections are linked. DOD also stated that the Navy varied from the original plans for achieving efficiencies, because it wanted to ensure that its operational commitments would continue to be met while efforts to reduce its infrastructure were being implemented. We agree that achieving savings through regional maintenance should not be done at the expense of meeting operational commitments. However, our work indicates the greatest impediments to progress are nonoperational issues, such as resistance to initiatives that eliminate organizations, reduce jobs and promotions, and reduce control over resources. DOD concurred in principle with our recommendation that the Secretary of Defense direct the Secretary of the Navy to annually report on the RM Program initiatives identified, savings achieved that have been verified in Navy budget and accounting records, and the progress made to overcome the barriers to achieving infrastructure reductions and savings. DOD stated that the Navy, through the staffs of the CNO, Naval Sea Systems Command, and the Atlantic and Pacific Fleet Maintenance Officers are already in regular communication with the OSD staff on all matters relating to the Navy’s RM Program. We agree that there is communication between the OSD staff and the Navy on various program matters. However, we believe that the communication needs to be more formal and comprehensive and cover such items as savings achieved and verified and progress made to overcome barriers to program implementation. DOD also agreed with our recommendation that program implementation plans be established and tied to milestones, with regular reporting to the CNO. DOD commented that the Navy has a management structure in place that provides unfettered information to the CNO on relative merits of potential initiatives as well as the success or failure of ongoing initiatives. While we agree that the CNO does get program information, the program lacks a strategic plan that identifies the Navy’s ultimate goal for the program and provides a baseline and a roadmap, with milestones, for achieving the goal. Such a plan is needed to show the Navy has made a high-level commitment to the program and to increase the likelihood of successful program implementation. DOD had several suggested technical and editorial changes; we considered them and made changes as appropriate. We are sending copies of this report to the Ranking Minority Member, Subcommittee on Military Readiness, House Committee on National Security; the Chairmen and Ranking Minority Members of the Subcommittee on Defense, Senate Committee on Appropriations; the Senate Committee on Armed Services; and the Subcommitee on National Security, House Committee on Appropriations. We are also sending copies of the report to the Secretaries of Defense and the Navy; the CNO; and to the Director, Office of Management and Budget. We will make copies available to others upon request. If you or your staff have any questions concerning the report, please contact me on (202) 512-8412 or my Assistant Director, George A. Jahnigen, on (202) 512-8434. Major contributors to this report are listed in appendix V. We identified examples of opportunities for consolidation of activities with potential annual savings of up to $26 million based on infrastructure reductions in the Pacific Fleet’s Hawaii and Northwest regions. Depending on the extent to which operations are consolidated in Hawaii, we estimate the range of annual savings to be from about $1 million to about $14 million. In the Northwest region, again depending on the extent of the consolidation, we estimate the range of annual savings to be from about $1 million to about $12 million. In addition, work at the Shore Intermediate Maintenance Activity (SIMA) at Everett, Washington, might be transferable to other existing shops, eliminating the need for a military construction project estimated to cost about $17 million and two barge overhauls planned at an estimated cost of about $5 million. In the Hawaii region, the Pearl Harbor Naval Shipyard and the Naval Intermediate Maintenance Facility are adjacent to each other, and the Public Works Center is about 1 mile away. At all three locations, the backshops have excess facilities and equipment. At the shipyard, for example, one electrical shop was not in use during the first shift at the time we observed operations. Also, a separate machine shop for tool-making supported the machine shop that did the repair work. Maintenance managers said this separate tool-making shop was unnecessary. (See fig. I.1 for pictures of machine shop capabilities in this region.) For purposes of this review, we estimated a range of potential savings. If selected backshop industrial work was combined and done by the shipyard, (1) facility savings alone might be about $1 million annually and (2) facility and personnel savings could be about $14 million annually if the work could be done just at the shipyard by a workforce the size of the current shipyard workforce. When we first reviewed operations in the Hawaii region in December 1996, we observed that a consolidation of intermediate-level maintenance activities with the shipyard appeared practicable; the Pacific Fleet Maintenance Officer agreed. On our return, in May 1997, fleet maintenance officials said that the Navy had begun to study issues surrounding the consolidation of the Intermediate Maintenance Facility and the Pearl Harbor Naval Shipyard, with a target date for complete integration by September 30, 1998. In the Northwest region, the Trident Refit Facility and the Naval Undersea Warfare Center, Keyport, are located within 4 miles of each other and about 14 miles from the Puget Sound Naval Shipyard. As in the Hawaii region, there were indications of excess facilities and equipment. The shipyard has a greatly reduced workload in 1996 compared to 1992, and the Keyport facility was subject to downsizing based on base realignment and closure action. Also, officials at these facilities told us that the shipyard had the facilities and equipment to do all of the region’s backshop industrial work. As other indicators, the shipyard had four machine shops scattered throughout the facility, and the sheet-metal shop was noticeably underused, employing about 65 workers on three shifts versus about 100 when it operated at full capacity, according to the shop supervisor. Similar to the Hawaii region, we estimated a range of savings. For example, if the industrial backshops at the Trident Refit Facility and the Warfare Center were declared excess and if all the workers needed to do that work were moved to the shipyard and used just the shipyard’s facilities and equipment, then there might be annual savings of about $1 million. If this industrial backshop work could be done just at the shipyard by a workforce the size of the current shipyard workforce, then facility and personnel savings could be about $12 million annually. Chief of Naval Operations (CNO), Regional Maintenance (RM) Program, fleet, and Northwest region officials agreed that there are significant amounts of excess industrial backshop facilities and equipment and that consolidation is possible and necessary. They said that consolidating industrial backshop work of all types into one industrial complex is key and that the goal should be to have one regional backshop for each type of capability. Figure I.2 shows electric motor industrial backshops at the Puget Sound Naval Shipyard and Trident Refit Facility in the Northwest region that have similar facilities and equipment for rewinding electric motors. A SIMA, Everett, Washington, detachment located at the shipyard also had its own facilities to rewind smaller electric motors. The Naval Undersea Warfare Center does not repair electric motors. In addition to the consolidation suggested above, the work of the SIMA at Everett, Washington, might be transferable to existing shops at Whidbey Island Naval Air Station, the Puget Sound Naval Shipyard, and the Trident Refit Facility. This transfer might eliminate the need for a military construction project at the SIMA. This military construction project is estimated to cost about $17 million, according to information provided by Everett SIMA officials. Northwest regional maintenance officials told us this military construction project is currently in the budget for fiscal year 2000. Also, other facilities at the SIMA could be converted to support waterfront maintenance activity requirements and eliminate the need for two barges and planned docking and repairs that Everett SIMA officials estimated could cost about $5 million. Tables II.1 and II.2 show the potential range of annual savings for the Hawaii region from consolidating at the Pearl Harbor Naval Shipyard selected industrial backshop work of the shipyard, the Intermediate Maintenance Facility, and the Public Works Center. Savings of about $1 million annually (table II.1) would be realized from lower facility operations (maintenance, utilities, and janitorial) costs if the entire workforce from all three activities is retained, but located at the shipyard. However, additional savings of about $13 million annually (table II.2) could be realized if the work were to be consolidated into the shipyard and could be absorbed by a smaller workforce the size of the one at the shipyard. Tables II.3 and II.4 show a similar range of annual savings for the Northwest region from consolidating at the Puget Sound Naval Shipyard selected industrial backshop work from the shipyard, the Trident Refit Facility, and the Naval Undersea Warfare Center, Keyport. The savings would be about $1 million annually (table II.3) from lower facility operations costs if the entire labor force is retained after consolidation. It shows additional savings of about $11 million annually (table II.4) if the work is absorbed by a smaller workforce the size of the one at the shipyard. Total annual savings from consolidations in both regions would be about $2 million if just facilities were consolidated and current staffing levels relocated to the shipyards, or about $26 million if the facilities were consolidated and all the work was done at the shipyards using a reduced labor force the size of the two shipyards. Data in the tables on labor-years and square footage of facilities were obtained from the six activities identified. Estimated total square footage costs and estimated costs of retaining only shipyard square footage were developed by multiplying the number of square feet identified by a cost factor used for RM studies. (See tables II.1 and II.3, footnote a.) Estimated savings is the difference between total square footage costs and the costs of retaining just the shipyard square footage. The lower range of projected total annual savings is derived by adding estimated facilities savings for both the Hawaii and Northwest regions, about $1 million each, or a total of about $2 million. Estimated total workforce costs and estimated costs of retaining just the shipyard level workforce were developed by multiplying the number of labor years identified by a cost factor also used the Navy uses for its RM studies. (See tables II.2 and II.4, footnote a.) Estimated labor savings is the difference between estimated total costs and estimated costs retaining just the shipyard force. The upper range of projected total annual savings is derived by adding the facilities and labor labor savings for both regions, about $1 million and $13 million, respectively, for the Hawaii region and about $1 million and $11 million, respectively, for the Northwest region, for a total of about $26 million. Table II.1: Potential Annual Facilities Savings From Consolidating Activities in the Hawaii Region Intermediate Maintenance Facility (square footage) Public Works Center (square footage) Total (square footage) Table II.2: Potential Annual Labor Savings From Consolidating Activities in the Hawaii Region Pearl Harbor Naval Shipyard (labor years) Intermediate Maintenance Facility (labor years) Public Works Center (labor years) Total costs (labor years) Table II.3: Potential Annual Facilities Savings From Consolidating Activities in the Northwest Region Trident Refit Facility (square footage) Naval Undersea Warfare Center (square footage) Total (square footage) Table II.4: Potential Annual Labor Savings From Consolidating Activities in the Northwest Region Trident Refit Facility (labor years) Naval Undersea Warfare Center (labor years) Total (labor years) To identify the Navy’s progress made in implementing the RM Program, we interviewed officials from the Office of the Secretary of Defense (OSD), Office of the Comptroller, the Deputy Under Secretary of Defense (Logistics), the Deputy Chiefs of Naval Operations for Manpower and Personnel and for Logistics, the Naval Sea and Air Systems Commands, and the Assistant Secretary of the Navy for Financial Management and the Comptroller and reviewed studies, briefings, and other documents on the RM Program. At the Atlantic Fleet headquarters, we interviewed the Fleet Maintenance Officer, reviewed documents, and obtained briefings from the Mid-Atlantic region—one of the four regions under the Atlantic Fleet. For the Pacific Fleet, we met with the Fleet Maintenance Officer and his staff, reviewed documents, and obtained briefings and other information from the Hawaii and the Northwest region—two of the four regions under the Pacific Fleet. Further, we talked to the officials of the Naval Audit Service about regional maintenance progress and its management consulting work for the RM Program. The Navy has identified seven objectives for the RM Program: (1) process improvement to maintain customer responsiveness and fleet readiness, (2) elimination of excess maintenance infrastructure, (3) integrated supply support, (4) maintenance cost visibility, (5) compatible maintenance management automated data processing, (6) positive control of technical elements, and (7) support the Department of Defense’s (DOD) industrial core policy. However, the program’s principal efforts thus far have been on the elimination of excess maintenance infrastructure; therefore, we focused our work on that program objective. To obtain cost and related-savings information for the RM Program, we interviewed officials with the Navy Financial Management and Comptroller offices, the CNO’s Naval Operations Supportability, Maintenance, and Modernization Division, and financial managers with the Naval Sea Systems Command, Naval Air Systems Command, and the Atlantic and Pacific Fleets. We also reviewed documents generated during the budget program review, the fiscal year 1998 program objective memorandum review, the net savings summary, and various memoranda discussing the budget reductions and projected savings. To identify opportunities for additional excess maintenance infrastructure reductions and cost savings in the Hawaii and the Pacific Northwest regions, we reviewed Atlantic and Pacific Fleet business plans, regions’ lists and studies of redundant capabilities. From the lists, we selected for further analysis industrial backshops for electric motor repair, electronics equipment repair, machining, and metal-forming shops and material testing laboratories. We obtained data, observed work, and discussed issues with maintenance officials and shop supervisors at six activities in two regions—the Pearl Harbor Naval Shipyard, the Intermediate Maintenance Facility, and the Public Works Center in the Hawaii region; and the Puget Sound Naval Shipyard, the Trident Refit Facility, and the Naval Undersea Warfare Center, Keyport, in the Northwest region. To calculate costs for these shops, we obtained information on square footage of facilities, and direct labor years and RM Program cost estimate factors ($13 per square foot for costs to operate and maintain facilities; $45,000 per year for each staff; 40 percent of direct labor for production support costs for all shops and 30 percent for material testing laboratories; and 15 percent of direct labor for administrative and general expense costs). We used this data to calculate estimated total costs, estimated total costs to retain the total workforce in just the shipyard facility, and estimated total costs of doing the work just at the shipyard with just a shipyard-level workforce. We compared the difference in these estimated total costs to identify estimated savings from retaining the total workforce at the shipyard, and estimated total savings with just a shipyard-level workforce at the shipyard. We conducted our work between December 1996 and September 1997 in accordance with generally accepted government auditing standards. Lionel C. Cooper, Jr. Dennis A. DeHart Samuel S. VanWagner Gary W. Kunkle Jean M. Orland The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed the Navy Regional Maintenance (RM) Program, focusing on the: (1) progress made in implementing the program; (2) savings that have been achieved; (3) opportunities for additional savings; and (4) barriers that inhibit full implementation of the program and achievement of projected savings. GAO noted that: (1) the Navy has made progress in achieving its infrastructure streamlining objective, but it has not been as great as anticipated and challenges remain for accomplishing future plans; (2) to implement the infrastructure streamlining objective, the Navy established steering committees, initiated a phased execution plan, identified a regional structure, and developed business plans; (3) through fiscal year (FY) 1996, the Navy identified 102 initiatives, 55 of which had been started by the end of FY 1997, and 47 of which are to be implemented between FY 1998 and 2001; (4) the Navy projected that its 102 initiatives would save about $944 million, of which $198 million was expected to accrue during FY 1994 to 1997 and $746 million was expected to accrue during FY 1998 to 2001; however, some of the initiatives are not progressing as projected; (5) the Navy cannot identify actual savings achieved because its accounting system does not track RM Program costs and related savings, and the Navy did not establish an independent system to track these costs and related savings; (6) the Navy has opportunities to build on its progress by working to achieve the $746 million in expected savings during FY 1998 to 2001, moving more quickly to implement savings initiatives, and pursuing other opportunities with high potential for significant savings; (7) the Navy identified many of its savings initiatives as high risk because of barriers to implementation; (8) the Navy faces parochial and institutional resistance to the RM Program's objectives and has other complex issues to resolve; (9) the biggest hurdle to overcome may be resistance to initiatives that eliminate organizations, reduce jobs and promotions, or reduce a command's or organization's control over resources; (10) other barriers are: (a) the lack of management visibility over all maintenance related costs; (b) multiple, unconnected management information systems that do not provide adequate data for regional maintenance planning and decisionmaking; and (c) the large number of shore positions desired to support the sea-to-shore rotation program compared to the smaller number needed to perform the intermediate maintenance workload; and (11) visible commitment by the Chief of Naval Operations (CNO) is critical to overcome resistance, accelerate decisionmaking, and provide the necessary resources and coordination needed for efficient and effective program implementation.

DOD’s primary medical mission is to maintain the health of its 1.6 million active duty service personnel and to provide health care for them during military operations. DOD also offers health care to 6.7 million nonactive duty beneficiaries, including dependents of active duty personnel, military retirees, and dependents of retirees. Under TRICARE, DOD provides health care to its eligible beneficiaries through military-operated hospitals and clinics worldwide and supplements this care with civilian providers. DOD contracts with MCS contractors to administer its TRICARE program on a regional basis. The MCS contractors’ responsibilities include claims processing, customer service, and developing and maintaining an adequate network of civilian providers. Since 1994, DOD has awarded seven MCS contracts covering the 11 TRICARE regions. These contracts were awarded for a base period and 5 option years. (See table 1.) Four of the MCS contracts have used all of the option years and three of these have been extended for an additional 2 years. A fourth contract’s extension is under way. TMA anticipates that all of its MCS contracts will eventually be extended. TMA, within DOD’s Office of the Assistant Secretary of Defense (Health Affairs), is responsible for administering the MCS contracts. TMA’s contracting officers have the ultimate responsibility for contract administration, including the issuance of change orders. Contracting officers are assigned to each MCS contract and are supported by other TMA staff as well as by the Lead Agents. Change orders may result from new laws or regulations, or from DOD initiatives. The most recent data available, which were in our 1997 report, showed that one-third of all TRICARE change orders resulted from new laws or regulations, while the remaining two-thirds were self-initiated. TMA officials told us that they were unable to provide updated statistics for our current review because the data are maintained by several different departments and would require significant effort to compile. Most changes are incorporated into the MCS contracts by issuing the change order as an amendment to the applicable TRICARE program manual: the Policy Manual, the Automated Data Processing (ADP) Manual, or the Operations Manual. Policy changes include the authorization of new benefits or changes in the administration or payment of current benefits. ADP changes involve modifications to how data are created, maintained, or reported, as well as changes to systems requirements. Operations changes include those involving the administration of the TRICARE program, such as revisions to home health care billing procedures. Changes can also be classified as “multiple,” meaning they involve modifying two or more of the manuals. Changes classified as “other” fall outside these categories. Examples of such changes are the authorization of travel costs incurred by the MCS contractors for government training and orders directing MCS contractors to report information about ongoing provider fraud investigations. Similar to our 1997 report, we found that as of June 30, 2000, the Operations Manual was modified most frequently, followed by the Policy Manual, as shown in figure 1. Most change orders affect all of the MCS contracts. For example, if TMA issues a change to the TRICARE program, such as a new benefit, all of the MCS contracts must be changed. Each change order, even for the same change, has to be negotiated separately with each MCS contractor because the cost of the change can vary by geographic region and the number of beneficiaries a contract covers. The Federal Acquisition Regulation, the Defense Federal Acquisition Regulation Supplement, and other internal DOD guidance set forth the requirements governing the administration of change orders. The requirements include time frames within which contractors should submit cost proposals to the government as well as guidance as to the time frames within which change orders should be settled. All change orders have been issued by TMA as unilateral changes, which means that they were implemented before their costs were negotiated and settled. The process of issuing and settling change orders involved many steps. When a change was identified for the TRICARE program, TMA defined its requirements, solicited comments about the change from the MCS contractors, and obtained an independent government cost estimate (IGCE), which was used to obligate, or set aside, funds for the change order. Then, the change orders were issued to the MCS contractors for implementation. MCS contractors were asked to submit a cost proposal within 60 days. When it received the proposal, TMA performed technical reviews and cost analyses of the information and then negotiated with the MCS contractor to determine the final price. Change order settlements can result in payments to the MCS contractors, savings to TMA, or no cost to either party. TMA’s goal for settling change orders is 180 days from issuance. TMA is also responsible for formulating the DHP budget request, which encompasses costs for MCS contracts, including change orders. TMA prepares both a current-year budget and a budget for the Future Years Defense Plan, which represents the estimated appropriation needs for the budget years for which funds are being requested and at least 4 years after. TMA also maintains the accounting system used to obligate and disburse funds for change orders. In our July 1997 report we cited a series of actions TMA had under way to address change order problems and reduce the backlog, such as hiring a management consulting firm to recommend improvements. In that report, we recommended that TMA continue providing high-level management attention to implement needed improvements to the process. However, TMA’s actions failed to make a measurable improvement to the process, and by June 30, 2000, the number of unsettled change orders had peaked at 562—more than double what it had been at the time of our previous report. Realizing that they had to “clear the books” in order to prepare for sweeping program changes that would result from the Floyd D. Spence National Defense Authorization Act for Fiscal Year 2001 (P.L. 106-398), TMA officials began an ambitious effort on July 1, 2000, to eliminate the backlog by the end of the calendar year. As of February 2001, TMA’s change order backlog totaled 121—a reduction of 78 percent in 7 months. To reach this goal within the short time frame, TMA modified its normal change order process, negotiating with its MCS contractors about $900 million in global settlements for the current and prior fiscal years that included change orders as well as other contract adjustments. Because the total cost of these settlements had not been included in the current-year defense budget, these costs contributed to about $500 million of the overall DHP funding shortfall for fiscal year 2001, which TMA estimates at $1.4 billion. According to our 1997 report, the initiatives TMA was implementing to address change order problems included engaging a consulting firm to prepare IGCEs for individual change orders, establishing a new requirement that all proposed change orders be reviewed and approved by Health Affairs before issuance, and hiring a management consulting firm to review and recommend improvements to the process. In that report, we stated that while it was too soon to determine the effectiveness of these efforts, they could bring needed discipline to the system by helping to ensure the need for, cost of, and timely settlement of change orders. However, for a variety of reasons that included a major staff reorganization and reduction, TMA management did not remain consistently focused on change order improvements, and these initiatives either were not fully implemented or had limited success. Furthermore, TMA did not always have data available to measure the impact of these specific initiatives on the change order process. (See app. II for a list of the initiatives and their outcomes.) These initiatives notwithstanding, between July 1997 and June 30, 2000, the backlog more than doubled to 562 (see table 2) although the percent of unsettled change orders decreased from 62 percent to 52 percent. On July 1, 2000, TMA initiated an ambitious, short-term effort, referred to as Mobilization, to settle its 562 open change orders as well as other contract adjustments by the end of the calendar year. TMA officials stated that clearing the backlog was necessary in order to prepare for changes mandated in the Floyd D. Spence National Defense Authorization Act for Fiscal Year 2001 (P.L. 106-398) that would result in significant program changes, including expanded health care and pharmacy benefits for military retirees who are age 65 and older. In addition, TMA officials recognized that they needed to reimburse MCS contractors for implementing past change orders that had not been negotiated and paid. To help achieve this goal, TMA increased the technical support for its Contract Management staff by using 14 staff on contract from the Center for Naval Analysis (CNA), who were originally slated to analyze and evaluate the proposals submitted in response to TMA’s solicitations for the next round of TRICARE contracts, which had been postponed. These additional staff were used primarily to help with change order proposal reviews by providing technical and pricing expertise. TMA also hired an information systems consulting firm to help review proposals and increased the numbers of other contracted staff who support Contract Management. TMA worked with its MCS contractors to make this initiative a priority by setting deadlines for the cost proposal submissions, which are needed to negotiate and settle the change orders. TMA officials soon realized, however, that reviewing and negotiating individual change orders would be too time-consuming to meet its 6-month goal. To expedite negotiations, TMA sent a team of contract staff to each MCS contractor to negotiate global settlements, which included all change orders that had not yet been settled under this effort, as well as REAs, claims, and other outstanding contract adjustments. As of February 8, 2001, TMA had completed payments to four of its five MCS contractors and had made partial payments to the fifth MCS contractor. Through this effort, TMA settled all but 71 change orders. According to TMA, some change orders were not settled under this effort because it lacked information or disagreed with the MCS contractors about settlement terms. The 71 change orders, combined with the 50 change orders issued after July 1, 2000, resulted in a backlog of 121 by February 2001. As of February 2001, TMA estimated a DHP funding shortfall of about $1.4 billion for fiscal year 2001. This shortfall amount includes about $500 million of the negotiated settlement amounts from TMA’s Mobilization initiative as well as other DHP requirements, such as the direct care system of military treatment facilities and the National Mail Order Pharmacy. TMA officials indicated that this shortfall would have to be satisfied through either an emergency supplemental budget request for fiscal year 2001 or a reprogramming of DOD funds. Negotiated settlements from the Mobilization initiative totaled about $900 million for current and prior fiscal years. However, it is not possible to identify the amounts related specifically to change orders for each of the MCS contracts because the change orders and other contract adjustments were jointly settled. Furthermore, the total cost of this effort is difficult to determine because the settlements also affected future-year costs. TMA officials stated that future-year cost estimates will be included in the President’s Budget for Fiscal Year 2002. In 1997 we reported that TMA was not budgeting separately for change orders, and we cautioned that settling TRICARE’s backlog of change orders could be very costly. We noted that the MCS contractors’ estimates to settle only a portion of the open change orders at the time totaled $423 million, yet TMA estimated that settling all open change orders would cost $38 million. At the time of our 1997 report, TMA officials told us that they were developing a methodology to include change order costs in their budget. That methodology, developed in 1998 and first employed in fiscal year 1999, was to use 3 percent of TMA’s annual adjusted MCS contract costs as an estimated budget for change orders. Using this methodology, the amount in the fiscal year 2001 budget for change orders is approximately $90 million. However, since TMA has negotiated settlement amounts for most of its change order backlog and has a clearer picture of change order costs, TMA officials plan to review the adequacy of the 3 percent budget estimate. The high volume of change orders issued and the consistently slow pace of settlements allowed the backlog to grow. Although TMA’s goal is to settle change orders within 180 days after issuance, it had met this goal less than 20 percent of the time as of June 30, 2000. In July 1997 we reported an average settlement time of 340 days and an average age per unsettled change order of 273 days. By June 30, 2000, both averages had increased: settlement time had risen to 499 days and the age of unsettled change orders reached 547 days—1½ years. Settlement delays occurred because of the slow submissions of cost proposals by the MCS contractors, the subsequent slow reviews of the proposals by TMA, and TMA’s periodic problems with obtaining funds for payment. TMA contract staffing shortages and high turnover further impeded the process. As of May 1, 1997, TMA had issued 357 change orders to its TRICARE contracts, with an average of 71 per contract; as of June 30, 2000, change orders totaled 1,091 and averaged 156 per contract. Since our 1997 report, TMA awarded two additional MCS contracts, whose change orders resulted in a 25 percent increase to the overall number issued. Table 3 compares the number of change orders issued for each TRICARE contract and region for these time periods. Figure 2 shows that the rate of issuance for change orders since May 1, 1997, has varied substantially by month. The number of change orders issued each month from May 1, 1997, through June 30, 2000, varied from a low of 3 during July 1998 to a high of 80 during September 1998. An average of 19 change orders were issued each month during this time period. The change order process has consisted of sequential steps, such as the submission of cost proposals by the MCS contractors, the subsequent proposal reviews by TMA, and ultimately, payment by TMA, if needed, for settlement. Delays in completing these steps slowed settlements. Furthermore, long-standing contract staffing problems, including both limited numbers and high turnover, diminished the contract staff’s overall knowledge base and impeded its ability to handle the heavy workload. To begin the change order negotiation and settlement process, TMA must receive an adequate cost proposal from the MCS contractor. TMA asks MCS contractors to submit a cost proposal within 60 days from receipt of the change order, but TMA does not often receive them within this time frame. As of June 30, 2000, TMA had not received proposals for 381 of the 562 unsettled change orders (68 percent)—90 percent of which were older than 60 days. In addition, we found that it took an average of about 9 months from change order issuance to proposal submission. One of the most common reasons MCS contractors cited for slow proposal submission was that they believe TMA does not always provide sufficient specifications for them to appropriately price out the costs of the changes. We cited this same problem in our 1997 report. As a result, MCS contractors said that during the implementation of the change they must spend time clarifying the details and scope of the change. One MCS contractor told us that TMA’s insufficient specifications leaves many items open to contractor interpretation, and therefore the MCS contractor’s efforts and corresponding costs sometimes vary significantly from TMA’s expectations. Another reason some contractors gave us for slow proposal submissions is the time it takes to collect relevant data for determining the total cost of the change. This process involves obtaining information from various internal departments, which can include systems and actuarial personnel, as well as from subcontractors, such as those used in processing health care claims. Also, two of the MCS contractors stated that if the benefit is new and unique, they prefer to gather actual costs because it is hard to predict the extent to which beneficiaries will use the benefit. Another factor that delays proposal preparation is TMA’s issuance of additional changes to previous unsettled change orders. This complicates proposal preparation because it is difficult to determine where to assign costs—to the original or to the subsequent change. In order to meet TMA’s goal to settle change orders in 180 days, it has 120 days after proposals are submitted to review, negotiate, and settle them. However, as of June 30, 2000, the average time between proposal submission and settlement was 295 days. A number of factors contributed to this delay. One was that when a proposal is received, the TMA contracting officer may need to obtain multiple cost evaluations, including health care, information systems (IS), operations, and possibly a Defense Contract Audit Agency audit. These evaluations usually are not performed by TMA’s Contract Management staff but by consultants or other TMA staff. The logistics of obtaining timely evaluations from these different sources can be time-consuming, and negotiations cannot proceed until the appropriate evaluations are complete. An audit or review that results in the revision of a proposal also adds to the settlement time. Furthermore, as time passes, data in the proposal may need to be updated. There is no standard length of time for which a proposal is considered current, and either TMA or the MCS contractor may initiate these updates. MCS contractors also may choose to revise their proposals if additional information becomes available that affects the cost of the change order. If TMA determines that a proposal is inadequate, the MCS contractor must make the necessary changes and resubmit it for review. According to TMA, an adequate proposal must meet the requirements of the Federal Acquisition Regulation. TMA told us that some of the more common reasons a proposal must be revised are that it does not provide sufficient detail on costs and that it does not include adequate supporting documentation. TMA officials said they rarely issue formal notices of inadequacy and prefer to resolve problems informally in a collaborative approach. As a result, even though TMA officials told us that proposals are frequently inadequate, they have not maintained statistics to demonstrate this. TMA officials acknowledged that contract staff exercise a fair amount of judgment in determining the adequacy of a proposal. For example, one contracting officer may find a proposal adequate, while another, who prefers more detailed cost data, may not. This can be a problem for MCS contractors if the TMA contract official they work with changes. One MCS contractor expressed frustration when proposals it submitted were initially considered adequate, then inadequate, and then adequate again, as their TMA contracting officer changed. Another MCS contractor told us it generally experienced an influx of inadequate proposals when new and inexperienced TMA staff were assigned to its contract. Once these staff became familiar with the MCS contractor’s proposal format, fewer proposals were returned as inadequate. All of the MCS contractors told us that once they and TMA agreed upon a payment amount, there was sometimes a lengthy delay before they received the official settlement paperwork allowing them to bill TMA. As a result, the change order backlog can include change orders that have been negotiated but not paid. Although TMA does not maintain statistics showing the length of time between negotiation and payment, MCS contractors told us that in some instances they waited many months after negotiation to receive the final settlement paperwork. For example, one MCS contractor waited 9 months for settlement paperwork, and another said it waited 6 months. TMA officials acknowledged that one of the primary reasons for delays has been insufficient funding. After negotiation, TMA officials prepare settlement paperwork and coordinate with Resource Management staff, who perform budgeting and accounting functions, to determine whether adequate funding is available. If it is not, TMA must wait until additional funding is obtained before sending paperwork to the MCS contractor. Delays also can occur if there is a problem with any of the numerous steps that must be completed in sequence after negotiation before the MCS contractor can be paid. Once TMA determines it has adequate funding, it sends the settlement paperwork to the MCS contractor for signature. The MCS contractor signs and returns the paperwork to TMA, where it is signed and returned to the MCS contractor, who may then bill TMA. TMA officials told us that they are required to pay the bill within 30 days of receipt. In 1997, TMA officials stated that the backlog was caused in part by a shortage of staff as well as TMA’s decision to allocate existing resources to the higher priority work of awarding the TRICARE contracts. Until recently, TMA officials continued to cite staff shortages as a problem contributing to the growth of the change order backlog. These officials stated that the continuation of this shortage was partially due to the Defense Reform Initiative that began in late 1997 and resulted in both a reorganization and an overall staff reduction of about 20 percent. TMA also had difficulty obtaining staff with certain areas of expertise needed to review MCS contractors’ proposals. For example, TMA did not have adequately trained staff who were dedicated to perform information systems technical evaluations of proposals for approximately 1½ years from January 1999 to June 2000. Without these evaluations, proposals with IS data could not be satisfactorily negotiated and settled. To avoid future staffing difficulties, TMA officials told us that they intend to retain some of the contracted staff who assisted them with Mobilization. In addition, TMA is currently conducting an internal assessment of current and future workload requirements to better align its resources, which could affect staffing levels. Therefore, TMA officials told us that they are not sure how many contracted staff will be used in the future. Another problem was the turnover among TMA’s contract staff—with staff both leaving and shifting among contracts. TMA officials stated that staff turnover is primarily due to staff shortages, which results in more work being distributed among fewer staff, leading to burnout. TMA officials stated that their experienced contract staff are highly marketable and can readily obtain other jobs for higher pay. Turnover of contract staff has been a long-standing problem for TMA. A 1998 Defense Logistics Agency (DLA) procurement management review of contracting activities at TMA reported that since January 1995, the turnover of contracting staff was high—about 33 percent over 3 years—and morale was low. The DLA report stated that these problems were due, in part, to staff burnout resulting from the change to managed care contracts, the sometimes hectic work pace, and the lack of program managers to make decisions and focus priorities. TMA concurred with the findings in that report. The DLA report also said that contract staff need a minimum of 18 months of TRICARE experience to properly learn about MCS contracts. Therefore, high turnover of contract staff compromises the overall knowledge level about contracts as well as the specific business operations of the MCS contractors, which can delay change order settlements. For example, a few of the MCS contractors said that by the time they reached final negotiations, the TMA contract staff they initially worked with were no longer there, and they had to spend time educating the new staff about past actions on the change order. In May 1999, TMA began revising the change order process with the goals of reducing the backlog and facilitating the effective management of future change orders. The resulting streamlined Change Management Process is intended to address the fundamental problems with change orders, such as controlling the number issued by ensuring that the changes are necessary and by making certain that money is available to fund them. The new process includes the creation of the Change Management Board, an executive-level body charged with reviewing, approving, and prioritizing new changes. Although the Change Management Board began meeting in January 2000, it is premature to evaluate the effectiveness of the new process because no change orders have yet been issued under it. Under the revised process, TMA’s Program Executive Officer assigns each potential change order to a Program Manager, who is responsible for all activities associated with it. The Program Manager’s initial task is to determine whether the potential change is “operational” or “significant.” Operational changes are those that cost less than $500,000 per MCS contract and include administrative modifications, such as annual updates of provider reimbursement rates, ADP system updates, and routine modifications such as the clarification of current requirements. Operational changes are not reviewed by the Change Management Board for approval and proceed directly to negotiation and implementation. Significant changes include benefits that represent a major program shift or changes with a financial impact over $500,000. For such changes, the Program Manager establishes an Integrated Program Team (IPT) consisting of a cross-section of personnel including senior TMA, military services, and Lead Agent officials and others as needed. The IPT is responsible for determining specifically how a new change will be implemented and for estimating the cost of implementation to the MCS contracts, if applicable. Although the amount of time needed to determine these details may vary, TMA officials estimate that the IPT’s portion of the process could take 6 to 12 months. Within this time period, the IPTs have specific milestones to meet, such as receiving approval of an implementation plan from the Change Management Board. The Change Management Board is composed of senior officials of both TMA and the Armed Services. The Board reviews proposed changes and determines which ones will become change orders. Approved changes are prioritized for negotiation and implementation on the basis of their importance and the availability of funds. Some approved changes will be delayed if funding is not available and thus may be included in the next year’s budget. For example, the Board may approve 20 change orders with total estimated costs of $34 million. However, if only $20 million is available to fund these changes, only changes with the highest priority will be implemented immediately. TMA’s new Change Management Process allows for change orders to be issued either unilaterally, as they were previously, or bilaterally, which means that the changes are negotiated before implementation. TMA’s goal is to issue all of its contract changes bilaterally. As with the previous process, contract changes issued bilaterally must have an adequate proposal with the requisite reviews before negotiations begin. Because the entire bilateral process may take over a year, any delays with proposal submissions and reviews would delay not only settlement but also implementation. To avoid such delays, in the new Change Management Process, TMA and MCS contractor officials will work together to develop the proposal, eliminating the more time-consuming consecutive steps of proposal submission and review. When possible, TMA plans to have negotiation teams that can be deployed to the MCS contractors’ locations to jointly develop proposals and negotiate costs, with the goal of leaving with signed bilateral agreements. Changes now going through the new process include one for the National Enrollment Database, which is to be issued bilaterally; TMA expects it to be issued in spring 2001. Twenty-three other changes are pending approval by the Board and will not be considered until funding is available. These include the expansion of the mammography benefit and elimination of the preauthorization requirement for beneficiaries with other health insurance. In addition, the Floyd D. Spence National Defense Authorization Act for Fiscal Year 2001 (P.L. 106-398) contained mandates that resulted in 22 changes not yet in the process, including the elimination of copayments for family members under TRICARE Prime as well as the extension of TRICARE Prime Remote for family members. Other mandates in this act include the recently added benefits for military retirees who are Medicare-eligible—TRICARE Senior Pharmacy and TRICARE for Life, both of which will be issued bilaterally. These 46 changes, when applied to all seven contracts, could create a total of 322 change orders. According to TMA officials, potential advantages of the new process include better control over the volume of change orders issued by prioritizing approved changes for issuance. In addition, TMA should be able to better manage its financial resources by issuing changes bilaterally because it will know the costs of changes before implementation and can ensure that funds will be available to pay for them. Potential benefits to contractors include better-defined specifications and assurance of timely payments. Although the bilateral process appears to be a step in the right direction, it may not work for some changes because sufficient lead-time may not be available. For example, congressionally mandated changes may have implementation dates that this process cannot meet because the amount of lead-time will not be sufficient. According to TMA, the unilateral change order process can still be used under such time constraints. As of June 30, 2000, the number of change orders issued had almost tripled, while the number of unsettled change orders had more than doubled since our last report. Although we recommended in that report that DOD devote high-level attention to managing improvements to the change order process, this was not consistently done. As a result, until recently, none of TMA’s numerous initiatives effected much improvement to the process or reduced the backlog. The current small backlog is the result of a recent concerted effort, not better management over time. TMA’s new Change Management Process appears to address many of TMA’s problems with change orders by controlling the volume of issuance, using a more collaborative negotiation process, and settling costs before implementation. However, past initiatives that appeared promising ultimately delivered little in terms of preventing or reducing the backlog. The high volume of change orders soon to enter the new Change Management Process makes it imperative that TMA management closely monitor the process to prevent future backlogs. We recommend that the Secretary of Defense direct the Assistant Secretary of Defense (Health Affairs) to monitor the new Change Management Process on a continuous basis and take immediate corrective action if problems, such as a growing backlog, are identified. We requested comments from DOD, but none were provided. We are sending copies of this report to the Honorable Donald H. Rumsfeld, Secretary of Defense, and relevant congressional committees. Please contact me on (202) 512-7101 if you or your staff have any questions concerning this report. Another GAO contact and staff acknowledgments are listed in appendix III. Our objectives were to update the 1997 report and provide information on (1) the status of the change order backlog and whether DOD reduced it, (2) factors that contributed to the growth of the backlog, and (3) DOD’s new change order process. To provide information on the status of the change order backlog and whether DOD reduced it, we analyzed a copy of TMA’s Change Order Tracking System (COTS) dated January 8, 2001. This file contains data about all change orders that have been issued by TMA to the five MCS contractors. With this database, we identified change orders that had been issued and settled on August 1, 1996, May 1, 1999, and June 30, 2000. We also used this database to identify the number of change orders issued since May 1, 1997, the number that had an independent government cost estimate (IGCE), and the average number of days between change order issuance and proposal submission. We compared these data with the corresponding data from our 1997 report to assess the difference. We also interviewed and obtained documentation from TMA about its prior initiatives to address change order problems, its Mobilization effort, the funding shortfall, and the number of unsettled change orders as of February 2001. We also assessed how TMA estimates, budgets, and accounts for change orders by obtaining studies and supporting documentation used to estimate costs, tracing estimated costs to the budget, and reviewing accounting data from fiscal year 1997 through fiscal year 2000. In addition, we interviewed and obtained documentation from each of the five MCS contractors about prior initiatives to address change order problems, and the Mobilization effort. We also interviewed and obtained supporting documentation from the TMA consultant who prepares budget estimates and IGCEs. To provide information on factors that contributed to the previous growth of the backlog, we used the January 8, 2001, COTS database to calculate the average amount of time needed to finalize all change orders settled by June 30, 2000; average age of unsettled change orders as of June 30, 2000; number of change orders issued each month between May 1, 1997, and June 30, 2000; number of proposals that had been submitted as of June 30, 2000; average amount of time from change order issuance to proposal submission; average number of days between proposal submission and change order settlement as of June 30, 2000; and number of change orders issued since May 1997 that had an IGCE. We compared these data with the corresponding data from our 1997 report to assess the difference. We also interviewed and obtained documentation from TMA about the number of change orders issued, the pace of settlements, the process TMA uses to settle change orders, factors affecting the pace of proposal review, and reasons why payments to MCS contractors are delayed. In addition, we interviewed the five MCS contractors about how they process change orders, factors affecting the timeliness of proposal submission, and delayed payments from TMA. We met with officials of the Defense Contract Audit Agency and Defense Contract Management Agency to determine their roles in the change order process. We were not able to determine whether change orders resulted from new laws or regulations or whether they were self-initiated because TMA does not regularly maintain these data. TMA officials stated that they were unable to provide this information in time for this report because the data are filed in several different departments and would require significant effort to compile. To provide information on DOD’s new change order process, we interviewed officials from TMA and obtained documentation that described this new process. We also discussed the new process with each of the MCS contractors to obtain their views. We did our work from July 2000 through March 2001 in accordance with generally accepted government auditing standards. At the time we issued our 1997 report, TMA had begun numerous efforts to improve the change order process and expedite settlements. These initiatives and their outcomes include the following: TMA assembled a team of contract specialists in August 1996 to expedite the settlement of change orders for all MCS contracts. TMA told us that the team was focused on reducing the change order backlog to a manageable and consistent level—a goal of 100 to 150 open change orders—by early 1999. However, by May 1999, when the team was disbanded, the backlog had grown from 197 to 408 change orders. TMA officials explained that this approach did not work for several reasons, including staff shortages. In November 1996, TMA engaged a consulting firm to prepare independent government cost estimates (IGCE) for new change orders instead of making “guesstimates” or basing cost estimates on contractors’ informal estimates. While this was a needed improvement, it was not expected to have an impact on the timeliness of the process since the IGCE is obtained before the change order is issued. Since our 1997 report, approximately 90 percent of the change orders issued had an IGCE. IGCEs are an important financial management tool because they serve as the basis for determining the amount of funds to be obligated for the change orders. Inaccurate estimates could result in either the underobligation of funds, which would result in the need for additional funding, or overobligation, which would unnecessarily obligate funds that could be used for another DHP program activity. According to TMA’s consultant who prepared the estimates, differences between IGCEs and proposal amounts occur for many reasons. For example, the MCS contractors may use different actuarial assumptions than TMA’s consultant. Differences can also result from the time delay between the preparation of the IGCE, which is developed shortly before the change order is issued, and the contractor’s proposal, which may be developed many months later and include actual costs. As of June 30, 2000, we found that it took an average of about 9 months from issuance to proposal submission. TMA’s analysis found that IGCEs were generally lower than both the MCS contractors’ proposed costs and settlement amounts. In March 1997, TMA established a new requirement that all proposed change orders be reviewed and approved by Health Affairs before issuance. The review was to evaluate each order’s effects on the health care system, its costs, and the availability of funds. In short, the Deputy Assistant Secretary, Health Services Financing, was to evaluate the need for each change order and decide whether to implement it. However, under the reorganization prompted by the Defense Reform Initiative, this responsibility was transferred to the Director of Military Health Systems Operations, who created a Change Management Board consisting of senior Service and TMA officials. Although TMA officials said that the Board reviewed and approved change orders before issuance, they could not provide us data to assess the impact of these reviews on the numbers of change orders issued. In March 1997, TMA hired a management consulting firm to review and recommend improvements to TMA’s change order process, specifically, ways to help reduce the current backlog and to prevent future backlogs. However, after reviewing a draft of the firm’s report, TMA officials discontinued the study because they were already aware of the consultant’s principal finding—that the change order problems were caused by high volume and a lack of discipline within the process. Nevertheless, some of TMA’s subsequent initiatives mirrored the report’s recommendations, such as the use of predetermined milestones and the establishment of a centralized review board to evaluate changes before they are issued. TMA notified contractors to begin submitting overdue (beyond the 60-day post-issuance requirement) proposals for low-cost or no-cost change orders. Contractors were told that proposals not received within 30 days could be unilaterally settled by TMA, meaning that TMA would pay the price it deemed appropriate. TMA officials told us that they periodically prompted MCS contractors to submit cost proposals for particular change orders and that the MCS contractors responded, obviating the need for TMA to settle any change orders unilaterally. However, these officials did not have data to illustrate the outcome of this and similar initiatives. When we issued our July 1997 report, TMA was developing provisional payment procedures, which were implemented in January 1998. These procedures allow MCS contractors to bill TMA on a monthly basis for costs incurred to implement changes, which is contingent upon the receipt of an adequate cost proposal. TMA makes provisional payments at the lesser of 100 percent of incurred contractor costs or up to 75 percent of the amount TMA has obligated for the change. If incurred costs are greater than obligated amounts, TMA requests a Defense Contract Audit Agency audit to validate the additional costs. TMA will then pay up to 75 percent of the validated costs. Despite TMA’s effort to pay contractors in a more timely manner, MCS contractors told us that the provisional payment process is cumbersome and slow, especially when their incurred costs are greater than TMA’s funded amounts. TMA officials said that the biggest obstacle to MCS contractors’ receipt of provisional payments is the submission of an adequate proposal. Bonnie Anderson, Mario Artesiano, Rathi Bose, Cynthia Forbes, Linda Garrison, Elizabeth T. Morrison, and Dayna K. Shah made key contributions to this report.

Under TRICARE, the Department of Defense's (DOD) managed care program, military-operated hospitals and clinics are supplemented by contracted civilian services. Since the inception of TRICARE, DOD has made many changes to these contracts via contract change orders. Since July 1997, when GAO reported that DOD was trying to improve its change order process, the backlog of change orders has continued to grow. This report evaluates (1) the status of the change order backlog and how DOD addressed it, (2) factors that contributed to the growth of the backlog, and (3) DOD's new initiative to improve the change order process. GAO found that as of June, 2000, the number of change orders issued had almost tripled, while the number of unsettled change orders had more than doubled since July 1997. Despite recommendations to devote high-level attention to managing improvements to the change order process, this was not done. Until recently, none of TRICARE Management Activity's (TMA) many initiatives significantly improved the process or reduced the backlog. The current small backlog is the result of recent concerted effort, not better management over time. TMA's new Change Management Process (CMP) appears to address many of TMA's problems with change orders. However, past initiatives have made similar promises and delivered little when they were abandoned or eclipsed by higher priorities. The high volume of change orders now or soon to enter the new CMP makes it imperative that TMA management closely monitor the process to prevent future backlogs.

As the nation’s largest cash-assistance program for workers with disabilities, DI provides benefits to eligible individuals under Title II of the Social Security Act. An individual is eligible to receive DI benefits if he or she has a medically determinable physical or mental impairment that (1) has lasted (or is expected to last) at least 1 year or is expected to result in death and (2) prevents the individual from engaging in SGA. SGA is defined as work activity that involves significant physical or mental activities performed for pay or profit. For individuals whose impairment is anything other than blindness, earnings averaging over $1,000 a month for calendar year 2010 generally demonstrate SGA. For blind individuals, earnings averaging over $1,640 a month for the year 2010 generally demonstrate SGA for DI. The amount of earnings that generally demonstrates SGA can vary from year to year. For example, the SGA amount for individuals with disabilities, other than blindness, was $980 in 2009. Individuals with disabilities must also have a specified number of recent work credits under the Social Security program at the onset of medical impairment. An individual may qualify on the basis of the work record of a deceased spouse or the work record of a parent who is deceased, retired, or considered eligible for disability benefits, meaning one disability beneficiary can generate multiple monthly disability payments. DI benefits are financed by payroll taxes paid into the Federal Disability Insurance Trust Fund by covered workers and their employers, on the basis of each worker’s earnings history. Cash benefits are payable monthly, as long as the worker remains eligible for benefits, until the worker reaches full retirement age or dies. In fiscal year 2010, more than 10 million beneficiaries received DI benefits totaling $121.6 billion, and the program’s average monthly benefit was about $922. As directed by federal law, SSA must reduce DI benefits for individuals receiving certain other government disability benefits, such as worker’s compensation. However, SSA may not reduce DI benefits for individuals receiving UI or for individuals earning less than SGA. As mentioned, the Social Security Board of Trustees projects that the DI trust fund will be exhausted in 2016 and noted that changes designed to improve the financial status of the DI program are needed soon. Established by the Social Security Act of 1935, the federal-state UI program temporarily and partially replaces the lost earnings of those who become unemployed through no fault of their own. To be eligible for UI benefits, unemployed workers must meet eligibility requirements established by state laws that conform to federal law, including that they have worked recently, be involuntarily unemployed, and be able and available for work. Whereas federal statutes and regulations provide broad guidelines on UI eligibility, the specifics of UI eligibility are determined by each state. According to DOL, all states require that a claimant must have earned a specified amount of wages, worked a certain number of weeks in covered employment, or must have met some combination of the wage and employment requirements within his/her base period. To be eligible for benefits, claimants must also be free from disqualification for acts such as voluntary leaving without good cause, discharge for misconduct connected with the work, and refusal of suitable work. In addition to these eligibility requirements, all states require that a claimant must be able and available for work. However, “able and available for work” requirements vary among the states, according to DOL. For example, a few states specify that a worker must be physically able, or mentally and physically able, to work. Likewise, while some states require that a worker must be available for work, other states require that a worker must be available for suitable work; still other states require that a worker be available for work in the worker’s usual occupation or for work in which the worker is reasonably fitted by training and experience. According to DOL, in addition to being able and available for work, all states require by law or by practice that a worker be actively seeking work or making a reasonable effort to obtain work. Finally, some state laws expressly prohibit denying UI eligibility on the basis of illness or disability under certain circumstances. UI benefits and administrative costs are financed primarily by taxes levied on employers. These taxes are deposited in the appropriate accounts within the Unemployment Trust Fund, which consists of 53 state accounts and other federal accounts dedicated to special purposes. The severity and length of the recent recession, and the slow pace of recovery, have placed a heavy demand on state UI trust funds, resulting in very large numbers of workers receiving benefits for very long periods of time. Since mid-2008, Congress and the states have temporarily extended the period of time that displaced workers can receive UI benefits to up to 99 weeks, though the maximum number of weeks of available benefits varies among the states. In April 2010, GAO reported that state UI trust funds were at historically weak levels, with most requiring federal loans to pay benefits. During fiscal year 2010, state agencies paid 11.3 million beneficiaries $156.4 billion in federal and state unemployment benefits. UI benefits vary substantially during a business cycle. As shown in figure 1, UI benefits varied substantially from 2005 to 2011, while DI benefits steadily increased during those years. In fiscal year 2010, 117,000 individuals received concurrent cash benefit payments of more than $850 million. As shown in figure 2, these individuals represented less than 1 percent of the total beneficiaries of both programs. However, estimated overlapping cash benefits paid to these individuals totaled over $281 million from the DI program and more than $575 million from the UI program. For individuals receiving overlapping benefits in fiscal year 2010, we estimate the average quarterly amount of overlapping cash benefit payments to be $1,093 in DI and $2,231 in UI, for a quarterly average of $3,324 in overlapping benefits. Differences in program rules and definitions allow individuals in certain circumstances to receive overlapping DI and UI benefits without violating eligibility requirements. As mentioned, SSA’s definition of a disability involves work that does not rise to the level of SGA. For 2010, a non-blind person who is earning more than a $1,000 a month is ordinarily considered to be engaging in SGA. In contrast, states’ determination of “able and available for work” criteria for UI benefits may include performing work that does not rise to the level of SGA. As a result, some individuals may have a disability under federal law but still be able and available for work under state law, thus eligible to receive DI and UI concurrently. SSA officials stated that UI is considered unearned income and therefore does not affect DI benefits. DOL officials acknowledged that certain individuals may be eligible for both DI and UI, depending on the applicable state laws regarding UI eligibility. Because these overlapping payments may be allowed under both programs’ eligibility requirements, and no federal law authorizes an automatic reduction or elimination of benefits if a recipient receives both payments, neither SSA nor DOL have any processes to identify these overlapping payments. As such, the costs associated with establishing mechanisms to reduce or eliminate these overlapping payments are not readily available. While the DI and UI programs generally serve separate populations and provide separate services—thus not meeting our definition for overlapping programs—the concurrent cash benefit payments made to individuals eligible for both programs are an overlapping service for the replacement of their lost earnings. We define overlaps as programs that have similar goals, devise similar strategies and activities to achieve those goals, or target similar users. Our prior work on overlapping government programs has found that, in some instances, overlapping programs or activities have led to inefficiencies, and we have determined that greater efficiencies or effectiveness might be achievable. However, in other instances, it may be appropriate for multiple agencies or entities to be involved in the same programmatic or policy area due to the nature or magnitude of the federal effort. Although current program rules allow overlapping benefits under certain circumstances, concurrent receipt of DI and UI benefits can also be an indicator of improper payments. For example, some individuals who have a disability as determined by SSA may be receiving improper UI payments because they are not “able and available” for work. Similarly, some individuals receiving UI benefits may be receiving improper DI payments because they no longer have a disability as defined by SSA. Specifically, being “able and available” for work may indicate that an individual’s medical condition no longer prevents him or her from performing work that rises to the level of SGA. As mentioned, neither SSA nor DOL have any processes to identify overlapping DI and UI payments. As a result, neither SSA nor DOL currently evaluates whether overlapping payments made to these individuals may be proper or improper. For our review, we obtained DI and UI information for these eight individuals beyond fiscal year 2010. Of the eight individuals we selected for further investigation, SSA determined that one individual had a DI benefit overpayment. Additionally, three individuals had UI benefit overpayments, as determined by the appropriate state UI office. Because we selected a small number of individuals for our review, the results cannot be projected to the population of individuals receiving overlapping DI and UI benefits. earnings may be related to work that makes this individual ineligible for UI benefits. The Massachusetts UI benefits were exhausted after 99 weeks as of June, 2011. As of April, 2012, the individual remains in current pay status in the DI program, with a monthly DI benefit amount of $2,377. Six of the individuals we selected for further investigation received overlapping DI and UI benefits for 18 months or more. For example, one individual began receiving DI benefits in 2004 originally due to disorders of the back, and received overlapping DI and UI payments, which totaled over $107,000, in 36 different months from 2008 to 2011. During that period, this individual worked for construction companies and received UI benefit payments from New Mexico in 2008, Wisconsin in 2009, Kansas in 2010, and Montana in 2011. Montana officials stated that they also received wage data from North Dakota for use in adjudicating the UI claim in their state. As of April, 2012, this individual was no longer receiving UI benefits from these states, but continued to receive cash benefits from the DI program. SSA officials told us that this individual is currently under a continuing disability review to determine if the beneficiary is ineligible for DI due to work at or above the SGA level. DI and UI provide important safety nets for American workers who have lost their income. However, both trust funds face serious fiscal sustainability challenges, prompting the need to examine opportunities for potential cost savings. While the programs target different populations and generally provide separate services, existing rules and definitions result in a limited number of individuals being eligible for overlapping DI and UI payments. However, the concurrent receipt of these benefits can also provide an indicator of improper payments related to DI or UI. Because these overlapping payments may be allowed under both programs’ eligibility requirements, and no federal law authorizes an automatic reduction or elimination of benefits if a recipient receives both payments, neither SSA nor Labor has a process to identify these overlapping benefit payments. As a result, for individuals receiving both DI and UI benefits, the government is replacing a portion of their lost earnings not once, but twice. Reducing or eliminating this overlap and potential improper payments could offer substantial savings, though actual savings are difficult to estimate because the potential costs of establishing mechanisms to do so are not readily available. We recommend that the Secretary of Labor work with the Commissioner of SSA to (1) evaluate the circumstances under which individuals are receiving overlapping DI and UI payments, taking appropriate action, as necessary, for any payments determined to be improper, and (2) assess whether cost savings or other benefits might be achieved by reducing or eliminating overlapping DI and UI cash benefit payments being made within the existing laws and regulations, seeking congressional authority to do so as appropriate. We provided a draft of this report to SSA and DOL for comment. DOL and SSA provided written comments to the draft which can be found in appendices I and II. DOL and SSA agreed with our recommendation that DOL work with SSA to evaluate overlapping DI and UI benefits, taking appropriate action for any payments determined to be improper, and assessing whether cost savings or other benefits might be achieved by reducing or eliminating overlapping DI and UI cash benefit payments. DOL and SSA also both recognized that the states play an important role in the UI program, and DOL recommended that we encourage states to participate in addressing the report’s recommendations. In this regard, we agree that states’ programmatic knowledge would significantly contribute to the evaluation of overlapping DI and UI benefits and encourage state participation as appropriate. We also believe that it will be important for DOL to reach out to the states in carrying out our recommendations to evaluate these overlapping benefits. DOL and SSA also provided technical comments, which we incorporated as appropriate. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Commissioner of the Social Security Administration, the Secretary of Labor, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have questions about this report, please contact me at (202) 512-6722 or hillmanr@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report.

The DI and UI trust funds face serious fiscal sustainability challenges. In addition to other services, both programs provide cash benefits to their targeted populations to replace lost earnings. DI is available to workers who are unable to engage in SGA because of physical or mental impairments expected to last at least 12 months or result in death. SGA is defined as work activity that involves significant physical or mental activities performed for pay or profit. UI provides temporary cash benefits to eligible workers who are able to work but remain involuntarily unemployed. GAO was asked to determine the extent to which individuals received DI and UI benefits concurrently. To do so, GAO matched unemployment files with SSA disability files for fiscal year 2010. GAO also reviewed DI and UI case files for a nongeneralizable selection of 8 individuals – 4 from the top 50 recipients of concurrent DI and UI benefits in fiscal year 2010, and 4 who received UI benefits based on wages from multiple states. These examples cannot be generalized beyond those presented. In fiscal year 2010, 117,000 individuals received concurrent cash benefit payments from the Disability Insurance (DI) and Unemployment Insurance (UI) programs of more than $850 million, which is allowable in certain circumstances under current program authority. While these individuals represented less than 1 percent of the total beneficiaries of both programs, the cash benefits they received totaled over $281 million from DI and more than $575 million from UI. One individual GAO selected for further investigation received over $62,000 in overlapping benefits in a year. Based on GAO inquiries, state UI officials are reviewing the person’s UI eligibility because of earnings that may be related to work that makes the person ineligible for UI benefits. Under certain circumstances, individuals may be eligible for concurrent cash benefit payments due to differences in DI and UI eligibility requirements. Specifically, the Social Security Administration’s (SSA) definition of a disability involves work that does not rise to the level of substantial gainful activity (SGA). In 2010, a monthly income of $1,000 or more for a non-blind beneficiary generally demonstrated SGA. In contrast, the Department of Labor allows states’ determination of “able and available for work” eligibility criteria for UI benefits to include work that does not rise to the level of SGA. Therefore, some individuals may have a disability under federal law but still be eligible for UI under state law because they are able and available for work that does not rise to the level of SGA. Although DI and UI generally provide separate services to separate populations—and thus are not overlapping programs—the concurrent cash benefit payments for individuals eligible for both programs are an overlapping benefit when both replace lost earnings. While SSA must reduce DI benefits for individuals receiving certain other government disability benefits, such as worker’s compensation, no federal law authorizes an automatic reduction or elimination of overlapping DI and UI benefits. As a result, neither SSA nor DOL has any processes to identify these overlapping payments. Reducing or eliminating overlapping or improper payments could offer substantial savings, though actual savings are difficult to estimate because the potential costs of establishing mechanisms to do so are not readily available. DOL should work with SSA to (1) evaluate overlapping DI and UI cash benefit payments, taking appropriate action for any improper payments, and (2) assess whether cost savings or other benefits might be achieved by reducing or eliminating overlapping DI and UI cash benefit payments being made within the existing laws and regulations, seeking congressional authority to do so as appropriate. DOL and SSA agreed with the recommendations.

As part of its constitutional authority to regulate commerce with foreign nations, Congress has long delegated to the President authority to proclaim reciprocal tariff reductions with U.S. trading partners and has encouraged the President to enter into certain trade agreements that meet congressionally mandated objectives. Congress established the trade advisory committee system in Section 135 of the Trade Act of 1974 as a way to institutionalize domestic input into such U.S. trade negotiations from interested parties outside the federal government. This system was considered necessary because of complaints from some in the business community about their limited and ad hoc role in previous negotiations. The 1974 law created a system of committees through which such advice, along with advice from labor and consumer groups, was to be sought. In practice, USTR has primary responsibility within the executive branch for developing U.S. trade policy, and the President has delegated to USTR the role of leading the trade advisory committee process. Additional players in developing U.S. trade policy include other executive branch agencies, particularly the departments of Commerce and Agriculture; the private sector, including business and nonbusiness groups; and state and local governments. USTR also maintains close consultation with Congress. The advisory committee system is one of several ways that USTR obtains input from the private sector (see fig. 1). In fact, Section 135 of the Trade Act also requires USTR to provide an opportunity to private organizations or groups outside the advisory committee system to present their views on trade issues. The system, created in 1974, was originally intended to provide private sector input to global trade negotiations occurring at that time (the Tokyo Round). Since then, the original legislation has been amended to expand the scope of topics on which the President is required to seek information and advice from “negotiating objectives and bargaining positions before entering into a trade agreement” to the “operation of any trade agreement, once entered into,” and on other matters regarding administration of U.S. trade policy. The legislation has also been amended to include additional interests within the advisory committee structure, such as those represented by the services sector and state and local governments. Finally, the amended legislation requires the executive branch to inform the committees of “significant departures” from their advice. The trade advisory committees are subject to the requirements of the Federal Advisory Committee Act (FACA), with limited exceptions pertaining to holding public meetings and public availability of documents. One of FACA’s requirements is that advisory committees be fairly balanced in terms of points of view represented and the functions the committees perform. FACA covers most federal advisory committees and includes a number of administrative requirements, such as requiring rechartering of committees if they are to continue for more than 2 years. The structure of the trade advisory committee system consists of three tiers, with the top tier directed by law to provide “overall policy advice,” the second tier to provide “general policy advice,” and the third tier to provide “technical advice and information.” However, Section 135 of the Trade Act does not establish any formal relationship among these tiers, nor does it authorize the first tier to exercise any control over the other two. USTR, working jointly with the other relevant executive departments, has the discretion to create, change, and terminate committees in the second and third tiers. The system comprises about 735 advisers spread across 34 committees, with the bulk of the advisers and committees in the third, technical tier. This tier consists of 17 industry sector advisory committees (ISACs), 4 industry functional advisory committees (IFACs), a committee of ISAC and IFAC chairpersons, and 5 agricultural technical advisory committees (ATACs). The second tier currently consists of five policy advisory committees. The first tier consists of just one committee, the Advisory Committee for Trade Policy and Negotiations (ACTPN), whose members are appointed by the President. Figure 2 illustrates the committee structure. The advisory committees are administered by USTR, which assumes a leadership role, along with the departments of Agriculture, Commerce, and Labor. USTR is responsible for administering ACTPN and three of the tier-2 Policy Advisory Committees, and shares responsibility with the other agencies for administering the rest of the committees. The Department of Commerce co-administers the majority of these committees—the ISACs, IFACs, and the Committee of Chairs. The Department of Agriculture (Agriculture) coadministers six others—the ATACs and the tier-2 Agricultural Policy Advisory Committee. The Department of Labor (Labor) is responsible for coadministering a tier-2 Policy Advisory Committee. The Environmental Protection Agency (EPA) plays a supportive role in the activities of the tier-2 Trade and Environment Policy Committee (TEPAC) but does not administer it directly. The advisory committee system’s unique features give it an important role in U.S. trade policy. Many negotiators use the system and report that the committees have made important contributions to successful U.S. trade agreements. Our analysis of documents indicates that committees have been given numerous opportunities to provide formal advice at committee meetings and through informal channels. The advisory committee system is unique in U.S. trade policy because it provides a forum in which business and other interested groups can consult confidentially with and provide advice to the executive branch on trade negotiations, U.S. trade policy, and implementation of trade agreements. The formal nature of advisory meetings helps ensure that representatives of the private sector and other groups have regular access to officials engaged in U.S. trade policy. Further, the system provides government officials with a body of private sector experts with whom they can develop an ongoing dialogue. Since USTR’s administrative procedures for the advisory committees require advisers to obtain security clearances before participating, the committees offer an environment conducive to discussing sensitive negotiating information. Many participants said the advisory committee system serves an important role in U.S. trade policy. A former USTR official and current committee member termed the advisory committee system “one of the great strengths of U.S. trade policy.” Among the comments made by negotiators whom we interviewed and members responding to our survey were that the formal advisory committee system is often preferable to more ad hoc means of obtaining input because it is institutionalized and seeks to be representative. Moreover, they said, the system provides assurance to Congress that domestic interests with a stake in trade matters have a voice when trade policy is formulated and will support the final agreements. It thus helps make the executive branch accountable to Congress and, ultimately, to the American public. According to multiple responses, the system strengthens the U.S. bargaining position by bringing to bear on-the- ground perspective and information from the private sector that the U.S. government lacks; establishing a clear set of U.S. priorities and fuller appreciation of various American interests; and enabling the United States to present a unified front when it faces foreign nations at the negotiating table. Without the system, some participants commented, U.S. negotiators would be operating in a vacuum and businesses would be unable to effectively resolve with foreign governments issues that only the U.S. government can pursue. The bottom line, negotiators and members agree, is that when it works properly, the system results in better trade agreements. Not only does it help the United States achieve commercial benefits, it can help keep the trading system vital and responsive to actual needs. Agency officials also cited the system’s value and contributions to U.S. trade policy. According to USTR, the advisory committee process was extremely successful during negotiations (1) on China’s accession to the WTO; (2) multilateral agreements on information technology, financial services, and basic telecommunications; (3) the Uruguay Round of negotiations that led to establishment of the WTO; (4) as well as regional initiatives such as the North American Free Trade Agreement, the Summit of the Americas, and the Asia-Pacific Economic Cooperation forum. Of the 27 USTR trade negotiators whom we interviewed, 18 indicated that they had obtained useful advice from the system, as did most of the 12 Commerce officials we interviewed. They cited numerous specific situations where advisory committee input had been helpful to negotiations. For example, an Assistant U.S. Trade Representative indicated that the advisory committees are playing a vital role in identifying market-opening priorities for the more than 140 nations currently involved in WTO negotiations. A Department of Commerce official described a committee as instrumental in helping monitor implementation of China’s accession commitments to the WTO. A USDA negotiator said a committee was helpful in setting the tone regarding the language on tariff reductions in the comprehensive U.S. agriculture proposal to the WTO. A USTR negotiator reported that a committee helped develop a position on defining “international standards” in the WTO’s Technical Barriers to Trade Agreement and helped gain the international community’s support for the U.S. proposal, expediting acceptance of U.S. goods in foreign markets. Committee members also value the advisory committee system and devote considerable resources to participating in it on a voluntary basis. Just over one-half of committee members live outside of Washington, D.C., and pay their own travel expenses to attend committee meetings. Further, when the Department of Commerce renewed the charter for the ISACs and IFACs in March 2002, more than 80 percent of those members continued. Members whom we surveyed highlighted numerous benefits of committee membership, including access to USTR and other agency officials, insights into other members’ views, and face-to-face dialogue with all members. Parties outside the system, such as U.S. subsidiaries of foreign-owned businesses and nongovernmental organizations (NGO), have sought representation on the committees, arguing that they should not be excluded from such an influential system. Our analysis of committee documents found ample evidence that USTR and other executive branch agencies are consulting committees on a wide range of trade initiatives at formal committee meetings. For example, agendas for the committee meetings during the 3 years leading up to the 4th WTO Ministerial, held in Doha, Qatar, in November 2001, listed the ministerial 60 times. Twenty of the advisory committees discussed U.S. preparations for the ministerial. Also, during fiscal years 1999 through 2001, different elements of the Free Trade Area of the Americas agreement were listed as items on more than 190 meeting agendas, at meetings of almost every committee. Such scheduled advisory committee meetings, usually held in Washington, D.C., are the formal channels for the executive branch to consult with the private sector advisory committees. In fiscal year 2001 there were approximately 110 formal meetings across the committee system. The number of meetings varied considerably by committee. The meetings generally lasted 3 to 5 hours. According to our analysis, about 80 percent of the meetings for fiscal years 1999 to 2001 were closed to the public. Negotiators and other trade officials attend portions of the meetings, each in turn briefing, discussing, and consulting with the committee. The private sector committee chair and the managing agency’s designated federal official (DFO) generally schedule meetings and select the agenda topics, although occasionally negotiators seek out specific committees to consult on a particular topic. Consultation during meetings is oral, but some committees send their positions in writing to USTR and the corresponding secretary or head of the agency. Many committee chairmen said their committees seek to provide consensus advice, which may include dissenting opinions. For the first- and second-tier committees, only classified transcripts were kept until recently. For the third-tier committees, DFOs prepare classified minutes of closed meetings for internal committee use only, as well as unclassified public summaries. In addition to formal meetings, USTR, Commerce, USDA, and others informally request advice from committee members through faxes, E-mails, ad hoc meetings, and teleconferences when they need a rapid response. However, committee members consulted at ad hoc meetings or through other means may provide advice only as personal opinions because, in keeping with FACA rules, formal committee advice generally can only be provided through formal committee meetings. In some cases, this informal advice is solicited by a request from a negotiator to the coordinating offices at USTR and other agencies, which then transmit the request to all advisers. In other cases, direct contact between negotiators and selected committee members occurs. Regardless of how contact is initiated, members typically provide advice directly to the relevant official and no central record is kept. Nevertheless, our review of existing agency records indicates that such informal consultation is active. In fiscal year 2001, USTR scheduled at least nine ad hoc meetings, mostly teleconferences or in-person meetings, to which trade advisers were invited an average of 2 to 3 days in advance. During this same period, USTR and Commerce faxed or E-mailed approximately 63 requests for advice, usually addressed to the entire advisory system membership or all of the industry sector and functional committees; according to our analysis of available data, the advisers had an average of 7.5 days to respond. Figure 3 shows the different processes for obtaining formal and informal advice from the committees. Most advisory committee members are satisfied with key aspects of the advisory process. However, some would prefer to be included more fully in the deliberations before actual trade policies are made, and many cited several problems with the consultation process that have hindered the system’s effectiveness. In addition, accountability for the use or consideration of advice could be improved. More than 60 percent of the members who responded to our survey reported that they were very satisfied or generally satisfied with 11 of the 16 areas of committee composition, operations, and effectiveness listed in table 1. In addition, about half of the committee members responding to our survey indicated that the system is fulfilling its statutory mandate to a “very great extent” or “a great extent.” The areas with the greatest levels of satisfaction (very or generally satisfied) were the knowledge of government speakers (85 percent), the committee’s opportunity to ask questions of government officials (84 percent), and the opportunity for members with dissenting views to provide input at meetings (79 percent). The areas with the lowest levels of satisfaction (very or generally satisfied) were the executive branch’s response to committee advice (39 percent), the use of technology to facilitate meetings (39 percent), and the time it takes to appoint new committee members (16 percent), which had by far the lowest level of satisfaction. Despite reporting general satisfaction with many aspects of the system, more than a quarter of survey respondents felt that the system has not realized its potential contribution to U.S. trade policy (see app. IV, question 23). Many members responding to our survey supported actions to improve committee operations. Through our survey, member interviews, and document analysis, we identified several problems with the (1) timeliness, (2) quality, and (3) accountability of consultations between the executive branch and the committees. These problems have, at times, limited member input into and influence over trade policy. Some of these problems were particularly acute for specific issues or committees. The timeliness of consultations was a concern to many advisers, who stated that consultations sometimes occur too late to affect policy. Overall, 30 percent of respondents felt that the executive branch scheduled its consultations so that the committees’ advice could be used in trade negotiations to “some or little extent” or “no extent,” while only 25 percent of the respondents believed the consultations were scheduled appropriately to a “very great extent” or “great extent.” Members whom we surveyed as well as interviewees reported that advice is often sought after the executive’s policy direction is already set. Several members reported that, in the past 5 years, the tendency has been for negotiators to come to committee meetings and say, “Here is the agreement, what do you think?” In one case that we documented, the administration did not consult with the President’s overall policy advisory committee—ACTPN—before submitting its proposed international trade agenda to Congress, including the principles to be included in Trade Promotion Authority legislation. The staff liaisons for ACTPN and TEPAC were only briefed on the matter the day after the agenda was submitted. Furthermore, the advisory committees were not consulted before the Clinton administration announced its decision to pursue a Free Trade Agreement with Singapore. The announcement provoked considerable concern across the private sector for a variety of reasons, not least because the original proposed time frame of completing the negotiations within 6 weeks would have allowed little time for advisers to provide input. Another problem with timeliness cited by members is that certain committees meet infrequently. USTR, Commerce, and USDA procedures generally indicate that agency officials are responsible for calling meetings. The problem of meeting frequency is particularly acute for the first- and second-tier committees, which averaged 1.7 and 2.5 meetings each year, compared with the third-tier committees, which met an average of 3.7 times each year. The ACTPN, which consists of CEO-level advisers and is designed to provide overall policy advice, met twice in fiscal years 2000 and 2001. It did not meet for more than 16 months between March 2000 and July 2001. During that period, the Jordan Free Trade Agreement—which broke new ground by including labor and environmental provisions in the text of a U.S. trade agreement for the first time—was finalized without formal executive branch consultation with the ACTPN. Members and negotiators reported that the lack of regular meetings was a barrier to the effective functioning of the committees. Although Section 135 of the Trade Act requires the executive branch to consult with the committees “on a continuing and timely basis” and “to the maximum extent feasible . . . before the commencement of negotiations,” the agencies involved have not adopted guidelines to implement these directives. For example, Commerce’s and USTR’s procedures and rules for managing these committees do not address the principle of timeliness or consulting to the maximum extent feasible. USDA’s procedures also do not refer to these issues, but in practice, the agency has developed a calendar of key negotiation events to use in scheduling advisory committee meetings in an effort to ensure that consultations are timely. Committee members, agency officials, and negotiators reported several problems that sometimes affect the quality and meaningfulness of consultations. These problems included too little time for discussion at meetings, limited access to background documents, insufficient consultation on certain issues, and poor participation by some negotiators. In survey responses and interviews, many committee members said that they did not have enough time to discuss issues or provide advice at committee meetings. While respondents to our survey were broadly satisfied with the amount of time spent on presentations by USTR and the committees’ principal agencies, 42 percent of the respondents reported that not enough time was devoted to providing advice, 43 percent reported that not enough time was spent on members’ discussing trade issues, and 39 percent said that not enough time was devoted to presentations by other executive branch agencies (see table 2). Commerce officials confirmed that they were aware that the amount of time available for discussion is an issue, but explained in agency comments that, because of the costs and travel time associated with ISAC and IFAC meetings and the number of issues to be discussed, meeting agendas are often packed. Negotiators stated that this imposes practical constraints on the time devoted to each agenda item. According to a USDA official, the agency is carefully reviewing the number of items on committee agendas and scheduling full- day meetings for its ATACs to ensure that there is sufficient time for member discussion. Limited access to certain background documents also affects the quality of consultations. USTR often provides national security classified and trade sensitive documents, such as proposed negotiating objectives or text of draft agreements, to committee members for comment. However, access to these documents—which are kept in Washington, D.C., in secured reading rooms—is often not feasible for advisers who live outside of the Washington, D.C., area, and not always convenient for advisers who work in Washington. Numerous survey respondents complained that current arrangements for reviewing such documents are inadequate. Officials and members said that being able to access documents electronically, such as through an encoded Internet site, would improve the quality of committee advice. In agency comments, USTR, Commerce, and USDA indicated they are exploring options for electronic access, but stressed that safeguarding sensitive or classified negotiating material must remain paramount. Taking a detailed look at these documents is important to members because it can materially affect negotiating outcomes. A Commerce official related an example pertaining to the Chile FTA, when ISAC members felt they had not had an opportunity to look at the negotiating text because it was put into the reading room at the last minute before a holiday. A subcommittee of the ISAC (8 to 10 people, mostly lawyers) reviewed the text line by line at the Commerce reading room and provided numerous changes to USTR, which were presented to the Chileans. The Commerce official noted that what happens with the Chile negotiations is considered extremely important because it will set a precedent for future trade agreements. Consultation on certain issues appears to be particularly problematic. First, although Section 135 of the Trade Act requires consultation regarding “the development, implementation, and administration” of U.S. trade policy, 30 percent of respondents reported that they were dissatisfied (“very dissatisfied” or “generally dissatisfied”) with the extent of consultation on implementation of trade agreements, 26 percent were dissatisfied with consultation on bilateral trade negotiations, 25 percent were dissatisfied with consultation on general trade policy issues, and 23 percent were dissatisfied with consultation on multilateral trade negotiations (see table 3). In one case, the administration prepared and issued, without first consulting the top-tier committees, a comprehensive report reviewing the WTO’s operation during its first 5 years, advocating that the United States should continue participation in the WTO. Advisers were only briefed on the 126-page report’s contents 5 days after the date it was signed by the USTR. Second, consultations with the tier-1 and tier-2 policy committees have not been satisfactory, some respondents said.Table 4 illustrates differences in tiers’ satisfaction rates for selected aspects of committee operations for those who responded to our survey. At the tier 1 ACTPN, as discussed above, lack of meetings and timely consultation were additional concerns. Although the tier-2 Labor Advisory Committee’s (LAC) formal steering committee met six times in 1999, six times in 2000, and three times in 2001, the full committee did not meet at all between 1994 and 2002. Steering Committee Members whom we interviewed felt that USTR consulted them more out of obligation rather than to obtain advice. One member added that USTR has treated the committee like a “dissent group” and did not provide the same level of briefings as it did to other advisory committees. Although members singled out a few USTR negotiators for their willingness to listen to the committee’s views, USTR acknowledges that it did not even have an official liaison to the LAC between 1993 and 2001. None of the labor respondents to our survey reported that they were satisfied with the degree of attention the executive branch paid to their committee’s issues or the executive branch’s response to the committee’s advice. Further, members of the tier- 2 Trade and Environment Policy Advisory Committee, comprising business and environmental interests, generally agreed that the committee had not been a successful vehicle for addressing environmental aspects of trade policy. While members and negotiators said that the committee made significant contributions to the development and implementation of President Clinton’s executive order regarding the review of environmental implications of trade agreements, many members cited frustrations over the committee’s inability to provide advice on other environmental policy issues. Few recognized environmental organizations still participate in the committee, and some members reported that the diverse interests represented in the group meant they had difficulty reaching agreement and providing clear advice. Finally, despite USTR’s efforts to convene meetings of the tier-2 committee designed to address the trade issues of concern to state and local governments, the Intergovernmental Policy Advisory Committee met only once in fiscal year 2000 and once in fiscal year 2001, both via telephone conference calls. Both members and negotiators reported that the lack of regular meetings was a barrier to the optimal functioning of the committees. Third, members and negotiators believe the system’s capacity for cross- fertilization among committees should be strengthened. Although functional committees on issues such as customs and standards have been established and are supposed to include representatives from industry committees, participation by such representatives is reported to be limited. Many issues—such as antidumping, biotechnology, and transparency in trade regulation—cut across several committees, but the system’s capacity to handle them is limited. A negotiator for antidumping, for example, generally consults only with the ferrous ores and metals committee (ISAC 7), which includes steel, on such matters. Although FACA and Section 135 of the Trade Act do not preclude agencies from consulting with a cross- section of members on such issues, USTR and Commerce have not generally taken advantage of this opportunity. Consulting with members from different committees on an ad hoc basis also would not produce the formal committee advice negotiators prefer. Some mechanisms for cross- fertilization already exist. A Committee of Chairs of the ISACs and IFACs is empowered to provide collective advice and provide a cross-section of views. Joint meetings of committees have also been convened. For example, members of the labor committee reacted favorably to an initiative by the USTR services negotiator to conduct a joint meeting of the services and labor committees to discuss the issue of temporary entry of foreign workers into the United States. Survey respondents supported additional steps to better address cross-cutting issues, such as sharing meeting agendas and recommendations—an idea that USDA is exploring. Some agency officials and committee members believe that the quality of consultations suffers because USTR is not as engaged as it should be with the advisory process. While 88 percent of the members responding to our survey supported (“strongly support” or “generally support”) actions to ensure that USTR officials attend committee meetings on a regular basis (see app. IV, question 18), we found that some committees have had little or no contact with USTR. For example, although the head of the USTR office that manages the advisory committee system said the office works actively to ensure that USTR’s negotiators consult with the advisory committees, several DFOs told us that arranging for USTR negotiators to meet with their committees is one of their most difficult tasks. In one example, two DFOs said they had never met their USTR liaison, nor had they been able to arrange for the liaison to attend their committees’ meetings. One of these DFOs added that, despite attempts, they have not been able to identify anyone at USTR who covers their issues, and, consequently, no one from USTR has attended the last five committee meetings. Even obtaining negotiating calendars is difficult, another official reported, making it hard to ensure that committee meetings are scheduled to support trade policy demands. Our analysis of documents provided by USTR revealed that USTR negotiators have not been actively working with some committees. Attendance records do not indicate who was present for what portions of committee meetings. As a proxy for this data, we reviewed scheduled speakers at tier-3 committee meetings in fiscal years 1999 to 2001. We found that, on average, USTR negotiators were scheduled to brief committees on 42 percent of the topics raised at the meetings. However, for 10 of the tier-3 committees, the USTR negotiators were scheduled to brief on 32 percent or fewer of the topics discussed. USTR argues that at tier-3 committee meetings the most knowledgeable speaker is often an employee of another agency, not a USTR official, because that agency works most closely with the technical information that is important to the committees. However, a perceived lack of attention by USTR was a source of concern to some members, who believe that the committee system was intended as a mechanism for negotiators to obtain advice on trade policy and agreements. For example, a committee chairman told us that a USTR negotiator had not been meeting with its committee and put forward a tariff proposal in ongoing FTA negotiations that placed the committee’s product in the longest phase-out category. However, because many U.S. producers now import, the committee actually favors lowering tariffs more rapidly. Seven of the 27 USTR negotiators with whom we met stated that they prefer to obtain advice outside the system because advisory committees cannot provide the type or quality of advice that they need. For example, three of the seven negotiators handle bilateral issues with key trading partners and shared this view. One negotiator said that the committees could not provide guidance on cross-cutting regulatory issues, so he speaks to associations or key companies that can provide the necessary advice. The negotiator also said that the committees generally do not provide timely, targeted responses orally or on paper. The second negotiator does not work with the ISACs or IFACs at all and, instead, uses informal contacts to obtain industry input. She explained that the ISACs are too broad to assist with the detailed issues she handles in bilateral trade negotiations. The third negotiator agreed that the ISACs were most useful when dealing with major, comprehensive negotiations like the Uruguay trade round. She had not used the system very much in the past 4 years and stated that, in her opinion, the system was not designed to handle the specific disputes, often involving litigation, that dominate bilateral trade relations. One of the two USTR agriculture negotiators we interviewed said the committees these negotiators work with often fail to advance policy because they do not narrow differences among members’ competing interests. A senior industry negotiator, meanwhile, indicated that the wealth of information she obtains through informal channels is more helpful than advisory committee input. The committee system provides limited accountability to ensure that committee positions on trade negotiating objectives are considered, as called for in Section 135 of the Trade Act. Prior to January 15, 1994, trade advisory committees affected by certain bilateral or multilateral trade negotiations were required to report to the President, Congress, and USTR at the conclusion of negotiations. This requirement was linked to legislation that gave the President the authority to negotiate certain trade agreements and submit them for congressional approval under expedited legislative procedures. The reporting requirement lapsed when the negotiating authority expired in 1994 and was not renewed until the recent passage of the Trade Act of 2002, which granted the President Trade Promotion Authority. According to a former USTR official, this lapsed requirement was an essential element in the trade advisory committee process because it assured Congress that the executive branch had sought and considered private sector advice. Without this reporting requirement, there was limited accountability in the advisory committee system. Moreover, mechanisms for tracking and distributing committee advice to senior agency officials are not routine or reliable. Instead, agency officials report that advice is transmitted through diffuse channels that range from formal to informal. At the formal end of the spectrum, there is no requirement that advisory committee input be sought before USTR officials submit documents on U.S. trade policy for interagency clearance by the Trade Policy Staff Committee. Although such documents sometimes include a section on private sector views, our examination of selected documents drafted by USTR in 2000 to 2002 revealed that many did not acknowledge solicitation or use of advisory committee input. The problems of tracking and distributing committee advice are aggravated bathe predominance of oral, nonconsensus committee advice offered during discussions at meetings. While oral advice from a range of perspectives can be valuable, it does not provide as clear guidance as written, consensus advice, which is easier to track and respond to. Commerce’s training manual for DFOs notes that “advisors should be encouraged to provide advice in writing, as advice imparted at meetings is often not captured for follow-up and is difficult to document. Members often incorrectly assume that resolutions made at meetings are passed to action officials by the DFO or that minutes are widely circulated in a timely manner.” Questions have been raised about how responsive agencies are to written committee advice. A number of chairmen felt such advice received more serious consideration, but several chairmen expressed frustration to us about nonsubstantive or untimely replies to their committees’ letters. One Commerce DFO stated that, at Commerce, committee letters are not always sent to officials responsible for the issues involved and instead go up the administrative chain and end up in a bureaucratic “black hole”; another DFO reported that the letters from the committee on which he serves have not been answered. Commerce denies that this is typical, indicating that committee letters are considered controlled correspondence that involves distribution of the incoming letter and review of the draft reply by responsible officials. Although officials at Commerce told us that it is common to send pro forma, rather than substantive, responses to committee letters, they noted in agency comments that this is generally because final U.S. policy has not been decided. USDA recently initiated a practice of summarizing resolutions made and sensitive issues raised at advisory committee meetings for senior USDA and USTR officials in an effort to improve agency awareness of and accountability for committee advice. Finally, Section 135(i) of the Trade Act requires the executive branch to inform committees of “significant departures” from committee advice. However, 41 percent of survey respondents reported that agency officials informed committees less than half of the time when their agencies pursued strategies that differed from committee input; only 22 percent reported that they were always or almost always informed of significant departures from committee advice (see app. IV, questions 8 and 9). About 86 percent of the respondents reported that they would support obtaining more feedback from USTR (see app. IV, question 18). Mismatches between the advisory committee system and the U.S. economy and trade policy issues suggest that the system is not positioned to provide the executive branch with all the advice it needs or to assure Congress that negotiated agreements are fully in U.S. interests. While most U.S. agricultural and industry sectors are represented in the committee structure, the composition of the system does not proportionally match each sector’s economic significance. Also, some specific industry committees have gaps in coverage. The structure of the system has not evolved fully to address new trade policy issues and stakeholders, and incorporating nonbusiness groups has been difficult. In the 28 years since the advisory system’s creation, the U.S. economy and trade have shifted toward services and high-technology industries (see fig. 4). However, membership composition and the number of committees that comprise the system’s structure are still heavily weighted toward the agriculture and manufacturing sectors (see fig. 5). In 1974, the committee structure was largely designed to enable the private sector to provide input on tariff negotiations, the principal issue in multilateral trade negotiations at that time. To determine whether the advisory committee system’s structure and composition reflect the current U.S. economy, we examined calendar year 2000 U.S. industry sector trade and output data and compared these with the corresponding membership data from the tier-3 industry sector advisory committees and agricultural technical advisory committees. We found the following: The services sector accounts for the largest share of U.S. output (more than 50 percent) and a sizable share of U.S. exports (almost 30 percent); these shares both increased sharply since 1974. Yet the committee system’s structure has only two services sector committees (the same number it had 20 years ago), and its composition includes fewer than 50 members from services. The number of committees in the system’s structure is heavily weighted toward manufacturing, which has 15 of the 33 committees. This appears to be consistent with manufacturing’s large share of U.S. goods exports. However, within manufacturing, some sectors such as textiles and apparel, nonferrous ores, and lumber and wood appear to be overrepresented in the committee system’s membership compared with their shares of U.S. trade, while large, exporting sectors such as electronics (18.3 percent of U.S. exports) are underrepresented (see fig. 6). Committee member composition is heavily focused on agriculture, even though agriculture accounts for less than 1.5 percent of U.S. output and 2.7 percent of exports. In 2001, USDA boosted the number of agricultural technical advisory committee members from 111 to 180. As a result, 222 of the 745 members in the entire system during fiscal year 2001 represented agricultural interests. This is not to suggest that there should be an automatic and linear relationship between trade levels and committee membership. In a few cases, other factors, such as policy considerations, might justify the imbalances between economic importance and committee representation. For example, sizable agriculture committees may be appropriate, since exports represent 40 percent of agricultural output and trade barriers are high. In services, the main services committee has been meeting monthly to keep up with comprehensive negotiations to improve the WTO General Agreement on Trade in Services. The government services negotiators we spoke with believed that the committee represents the sector well, and 70 percent of services committee respondents to our survey reported satisfaction with their sector’s representation in the system. Certain manufacturing sectors such as textiles are recognized as import sensitive. Nevertheless, such reasons may not apply to the remaining committees. Indeed, according to Commerce and USDA officials, in most cases, current committee membership levels are functions of private sector interest in participating, rather than a deliberate effort by the agencies to determine appropriate levels of representation. Membership in the system is also not fully aligned with the economy because of gaps in industry representation that occur for at least two reasons. First, there are gaps based on whether companies choose to join the system or not, resulting in a lack of balance needed by negotiators to cover all the industry-specific issues they must address in trade negotiations. For example, according to one USTR negotiator, a major telecommunications services provider opted not to participate because it had access to USTR through other venues. The electronics committee does not yet have a representative from the software industry, and the intellectual property rights committee does not have a representative from the generic drug or noncontent producing copyright industry. Second, major foreign companies, such as DaimlerChrysler, cannot participate because foreign-owned companies are generally prohibited from membership on committees under USTR and Commerce procedures and rules. In commenting on a draft of this report, Commerce stated that the rationale for this long-standing policy is the sensitivity of the subject matter considered by the committees and possible conflicts that might be experienced by U.S. firms that have foreign owners. U.S. subsidiaries of foreign-owned firms accounted for more than 5 percent of U.S. employment and more than 20 percent of U.S. goods exports in 1999; such foreign ownership has grown with globalization and is particularly high in certain manufacturing sectors, such as transportation equipment and chemicals. These gaps in industry representation have encouraged negotiators to seek advice outside the advisory committee system, including from foreign-owned firms or trade associations that include such firms. Committee membership is significantly below the levels authorized in committee charters, averaging 49 percent of authorized capacity (see fig. 7) in 2001 and 48 percent of authorized capacity as of August 2002. The low membership rates can at times severely limit the availability of advice for negotiators from certain committees, particularly since just over half of the members attend meetings, on average, according to the attendance records made available to us. One negotiator with overall responsibility for a major bilateral FTA said he had hoped to rely on the committee system exclusively for advice, but had concerns that certain committees were not sufficiently filled to provide a meaningful cross-section of industry views. In addition, some meeting records we reviewed indicated that more government officials were in attendance than committee members. On the other hand, the fact that some committees are far below authorized membership levels means agencies have opportunities to fill gaps in industry representation. Officials at USTR, Commerce, and USDA acknowledge the need for increased outreach to fill gaps in membership and have recently taken steps toward this end. In the past, agencies primarily relied on recruitment through Federal Register notices to attract new members, rather than targeting specific needs or groups. Furthermore, negotiators were not always actively involved in identifying candidates to fill gaps in composition or representation. However, Commerce has stepped up its outreach by encouraging current members to recruit applicants, directly soliciting applicants at trade shows, holding meetings outside of Washington, and speaking before trade associations and outside groups. In addition, USDA solicited applications through different means, including widely disseminating notices to state departments of agriculture and other farm groups. Some USTR negotiators reported urging industry contacts and experts to become involved, but USTR reports that a key obstacle to filling vacancies is the difficulty in identifying qualified individuals in the private sector who are willing to join the advisory committees, due to the significant amount of time and resources required to serve. With little restructuring to mirror emerging trade policy issues and new stakeholders, the committee system is unable to provide some negotiators with all of the advice necessary to support trade policy development. New trade issues and stakeholders have emerged since 1974, as trade negotiations expanded beyond tariffs to include nontariff barriers to trade and other complex trade-related issues, such as intellectual property rights and health and safety. Moreover, the WTO negotiations launched in November 2001 cover topics such as the relationship between WTO rules and multilateral environmental agreements, and negotiations on a free trade agreement with Chile cover investment and competition (antitrust) policy. These issues require functional expertise and expand the number of U.S. interests concerned with and affected by trade agreements. Trade negotiators with whom we spoke stated that there are gaps in the committee system structure regarding functional issues such as investment and government procurement and in representation of stakeholders in such areas as public health. There have been few changes in the committee system’s structure to address these new issues and avoid gaps in coverage. Section 135 of the Trade Act gives USTR flexibility to restructure the committee system to reflect changes in U.S. international trade interests. However, in the past decade, only 3 committees—the Trade and Environmental Policy Advisory Committee, the Trade Advisory Committee for Africa, and the Industry Functional Advisory Committee on Electronic Commerce—have been added to the 34-committee structure. Most of the remaining committees have existed for more than 20 years (see fig. 8). USTR officials acknowledge the need to update the committee system to reflect the current economy and new trade issues, but add that the agency would need external guidance to support any sensitive decisions affecting existing committees. According to a former USTR official, a USTR effort to review the committee system in 2000 did not even address the question of how well the system was meeting USTR needs, because the agency did not have the time to ask negotiators what they wanted from industry advisers. In some cases, agencies forego addressing some recognized needs for advice on new issues because of the time and effort required to create and amend committees. For example, two USTR and Commerce negotiators, who are in charge of their offices and oversee other negotiators, told us that current ad hoc methods for obtaining advice on investment policy are inadequate and that they believe a separate committee on investment would be desirable. However, they expressed reservations about undertaking the considerable effort involved to form one. (Commerce spent more than a year establishing the E-commerce advisory committee.) Although many new trade issues impinge upon domestic regulatory areas that are of concern to nonbusiness groups, USTR and the other managing agencies have had difficulty incorporating nonbusiness stakeholders into the committees. Some nonbusiness interests from the labor, environment, and consumer communities participate in the committee system but stated that they feel marginalized within it. Most nonbusiness members currently participating in the system are placed on a few committees in the second tier, where committees are less active and productive than in the third tier, as shown in figures 9 and 10. New stakeholders in the trade process, such as public health, development, and gender advocates, have limited or no participation in the formal committee system, even though topics such as intellectual property are of interest to them. Some negotiators on this topic and on services believe that nonbusiness stakeholders’ perspective is useful and necessary in formulating U.S. trade policy. However, the extent of participation by nonbusiness members on tier-3 committees is still an unresolved legal issue. Nonbusiness participants with whom we spoke also feel marginalized because they have difficulty ensuring their views get serious consideration. For example, the ACTPN is meant to provide overall policy advice and is required to be broadly representative of key sectors and groups affected by trade. Six of the 33 current members represent nonbusiness interests. In 2000, the three labor representatives temporarily resigned from this presidential committee because the chair said the committee would only meet once and its sole focus would be the granting of Permanent Normal Trade Relations to China. The labor representatives felt that their issues were not being addressed and in their resignation letter said that the advisory process “relegates minority views to a marginalized dissent.” Most of the 22 committees administered by Commerce do not routinely allow representation from nonbusiness interests. As a result of legal challenges to the business-only composition of several of the Commerce committees, two of these committees (lumber and paper products) now have environmental representatives and are reported to be functioning productively. A third committee, chemicals, represents the second-leading manufacturing export sector but still lacks a permanent environmental representative as called for by a settlement order, resulting in the committee’s operations being interrupted for the second time in 2 years. Outside of these three committees, the extent to which nonbusiness interests, including environmental interests, can be represented on tier-2 and tier-3 committees has not been completely resolved. Neither Section 135 of the Trade Act nor its legislative history is clear about how that statute relates to the Federal Advisory Committee Act’s “fair balance” requirement or about how to apply “fair balance” in the context of a trade advisory committee system largely composed of discrete interests (see app. II). Recently, the Commerce Department published a notice indicating that except for environmental representation in the three committees where representation has been successfully challenged, “non-government organizations and academic institutions do not qualify for representation on a committee.” Nevertheless, without further clarification by U.S. appellate courts or amendments to the current legislation about what fair balance means for trade advisory committees, some ambiguity about this issue will remain. Negotiators, agency officials, and committee members have suggested the need for Congress to clarify its intent for representation in the committee system. The current legal uncertainty also raises several practical issues. First, the lack of clarity in what fair balance means for these committees makes the system more vulnerable to court challenges. Second, our interviews with agency officials suggest that these uncertainties may make it difficult for agencies to consider revisions in the committee structure to better address functional issues such as investment and public health, which are of interest to business and nonbusiness groups. The appointment of environmental representatives to tier-3 committees has generated concerns among some current committee members. For example, the committee of ISAC/IFAC chairmen wrote to USTR expressing concern that having nontraditional members on their industry sector and functional committees would make the committees less productive in performing their primary mission of ensuring that U.S. negotiators were aware of industry interests and positions. One committee chair said that business members would be less forthcoming about discussing trade issues because of concern that nonbusiness representatives might release sensitive information to the public, thus undermining candor and confidence. More than 60 percent of respondents to our survey opposed adding more nonbusiness interests to their committees (see app. IV, question 18). These concerns were echoed many times in our interviews with members who feared they would lose their voice in trade policy or said that they would be unwilling to participate if the committees become unproductive “debating clubs.” As an alternative, the ISAC/IFAC chairmen recommended “the establishment of a functional committee or committees to serve as parallel and equal fora for involvement by non-traditional groups.” Our interviews and review of agency documents found that USTR has been making efforts to provide information to, and obtain input from, various nongovernmental organizations outside the formal advisory committee process. We contacted several such organizations that had demonstrated an interest in U.S. trade policy by submitting formal comments in response to USTR Federal Register notices or attending USTR public briefings. While these groups welcomed increased outreach by USTR, most felt that having a role in the formal advisory committee system was still desirable, saying it would enhance accountability and add balance to U.S. trade policy. However, several feared that creating NGO-only committees would “ghettoize” them within the system and fail to ensure equal access to information and decision makers. These NGOs favor a broader overhaul of the system but acknowledge that NGOs often do not have the requisite resources or desire to participate. Despite relatively high rates of member satisfaction with the support by USTR and other agencies such as Commerce and USDA, we found in our review that lack of policy direction and weak system administration at executive branch agencies are limiting the advisory committee system’s capacity to accomplish its statutory mission and contribute to U.S. trade policy. USTR, as the lead agency, has not provided clear policy direction. Execution of administrative tasks needed to keep advisory committees operating and relevant has been slow. The limited resources USTR and the other key agencies devote to managing the advisory committee system have not been sufficient to position them to maximize input from the committees. Several experts and committee members stressed the importance of organizational leadership from the top in creating an environment for vital and effective advisory committee input into U.S. trade policy. However, USTR has taken a decentralized and delegated approach to obtaining private sector advice and has not demonstrated a commitment to assume a leadership role in the advisory committee system. Through interviews with USTR negotiators and other officials, we learned that the agency’s overall policy of consulting with the private sector generally has not ensured that the formal statutory advisory committees are systematically consulted. Agency officials explained that negotiators are encouraged to consult with the private sector but that they exercise individual discretion over whether to consult with the advisory committees. As noted earlier, some negotiators whom we interviewed reported using the committee system to obtain advice, while others consult the committees only on a pro forma basis or do not consult them at all. This unevenness has economic consequences: In one example, USTR did not inform a committee that a general effort to reduce discriminatory tariffs against U.S. goods in central and eastern Europe was under way, and as a result, its industry sector was not included in the final package of agreed tariff cuts. Our examination of USTR and Commerce procedures found that they do not provide broad guidance to USTR officials and other negotiators on their obligation to consult with advisory committees or on when, how, and with whom to consult. Instead, they are largely aimed at committee members and agency administrators and focus on committee operations. A USTR negotiator and committee members have suggested that clearer expectations for the consultation process need to be developed for both negotiators and advisers. Without them, misunderstandings do occur. For example, one committee chairman, who is generally satisfied with USTR’s use of advisory committees, was outraged when USTR neglected to consult his committee on an issue of long-standing interest and, instead, sent a position paper in a broadcast E-mail to all advisers in the system with a 2- day deadline and then presented the proposal to other nations. Although, at the time, a USTR administering official said the broadcast E-mail was typical, the negotiator responsible later acknowledged that USTR mishandled the process for seek advice in this instance. Slow and cumbersome administrative and security processes have also hindered committee operations. Under FACA, Section 135 of the Trade Act and implementing guidance and procedures, USTR and other federal agencies are responsible for placing new members on committees, rechartering committees, and creating new committees. These are important functions that keep the advisory committee system operating and relevant. However, our work at three key administering agencies— USTR, Commerce, and USDA—suggests that present methods for accomplishing these responsibilities do not ensure that the system functions reliably. Turnover of membership occurs regularly given the pace of global business, industry consolidation, and distress in certain segments of the U.S. economy. Yet, applications for prospective members spend months in the approval pipeline before the members can participate (see app. III). For example, USTR submitted a list of candidates for appointment to the White House for a presidential appointment to the Advisory Committee for Trade Policy and Negotiations in mid-February 2002, which, as of early September 2002 had not yet been cleared by the White House. Agency officials acknowledged that these delays are frustrating for potential members and can be a disincentive to joining the system. Indeed, 40 percent of our survey respondents were dissatisfied with the time it takes to appoint new members to committees and 35 percent said the time to appoint new members has deteriorated (see app. IV, questions 17 and 22). Our analysis of agency documents indicates that the full appointment process, which includes the time for members to complete the application materials and the time for a required security clearance, regularly takes 6 months or longer. Some time-consuming elements of the clearance process are beyond the trade agencies' control. However, all three agencies now pay for the expedited security investigation offered by the Office of Personnel Management. Some agencies have taken other streamlining steps, such as providing interim security clearances. According to DFOs and other agency officials, applying these reforms more widely could alleviate this major irritant. FACA’s requirement that committees continued beyond 2 years must be rechartered has been disruptive for the trade advisory committee system, posing a particular burden for new administrations until their key policymaking vacancies are filled. In several cases during our audit period, committees ceased to meet and thus could not provide advice, because the agencies had not adopted new charters and appointed members. For example, the agriculture advisory committees did not meet between April and October 2001 while USDA went through the process of appointing members for its six committees. The committee charters and rosters expired before the United States was able to vet its market access proposal for the Chile FTA negotiations, and as a result, a lead USDA negotiator reported that he was not able to use the committees to obtain input on the proposal. The Labor Advisory Committee’s charter expired in July 2001 and was not renewed until February 2002. As a result, the LAC Steering Committee could no longer meet or provide formal committee advice as efforts to launch new WTO negotiations at the Doha Ministerial in Qatar were under way. (Only one of the LAC respondents to our survey reported that the system was fulfilling its statutory role in U.S. trade policy.) Commerce successfully avoided disruptions in its most recent rechartering by starting the process for the 22 committees it manages well before their charters expired. However, the effort to appoint new members and obtain security clearances required the full-time attention of two of the three Commerce employees responsible for managing the committees and took 7 months to complete. Commerce, USDA, and USTR officials said the tasks associated with the rechartering process—such as preparing new charters, analyzing the attendance records of members up for reappointment, and reviewing member application information—places a significant burden on their ability to manage the committee system and detracts from their ability to support committee operations. The resources USTR and the other agencies devote to managing the advisory committee system do not match the tasks that must be accomplished to keep the system running reliably and well. (We recently testified on human capital shortages at trade agencies, including USTR and Commerce.) According to annual reports that the agencies prepare for the General Services Administration, federal staff time allocated to managing all the committees totaled 15.60 full-time equivalent (FTE) positions in fiscal year 2001 and averaged 0.47 FTEs per committee. USTR officials said the current staffing levels in the office responsible—three positions with multiple responsibilities besides the committee system—do not allow them time to proactively manage committee operations. The recent head of the office said that simply restarting all the lapsed committees and keeping the rest of the system operating were occupying much of the time she could devote to the system. Commerce and USDA manage more committees and face similar challenges. Commerce officials responsible for managing the Commerce committees reported that they must focus their limited staff on the rechartering and appointment processes, which has not allowed them to meet their responsibilities to attend all the committee meetings. However, some improvement may be forthcoming: In its official comments on our draft report, Commerce stated that it will shortly hire two new full-time staff to support administration of the committee system. Meanwhile, a USDA official in the office responsible for managing Agriculture’s committees—which has one professional position devoted to advisory committee work and two other positions with multiple responsibilities in addition to managing the committees—said the reappointment process in 2001 took more than 85 percent of her time and prevented her from fulfilling other key job responsibilities, such as legislative liaison. Resource limitations also affect the use of technology. Although committee members supported the use of technology to improve committee operations (79 percent strongly or generally supported increased technology to inform members and 60 percent supported the use of videoconferencing technology to enable greater participation in meetings), the cost of new technology is a significant determining factor in its adoption and use. Commerce is examining options to expand the use of its Web site for committee members, but the cost of options at the high end of estimates ($200,000)—which include the security safeguards needed for improved member access to sensitive documents—far exceeds available funding. Finally, a USTR official reported that the agency’s live Web casts from the WTO Doha Ministerial were very costly (estimated at $50,000) and cannot be done on a routine basis. However, USTR plans to examine less expensive technological options, such as taped presentations through its Web site. Despite several weaknesses we identified, negotiators, agency officials, and members told us that the advisory committee system Congress created 28 years ago still provides value to U.S. trade policy. Many negotiators report that input from the system has helped the United States achieve more beneficial trade agreements. Members devote time and contribute much to the process and report generally high satisfaction with many aspects of committee operations and effectiveness. Nevertheless, our work suggests that the committee system is not being used to full advantage and has lost some of its vitality in providing useful advice on trade policy matters. Consultations are not always timely or meaningful, and when advice is provided, there is little assurance that executive branch officials are held accountable for considering it. Furthermore, the committee structure has not evolved fully to reflect today’s economy. Some key trade interests that have recently surfaced—industries, issues, and stakeholders—are missing or poorly represented in the system. Conflicts over interpreting how FACA’s fair balance requirement applies to the advisory committees have complicated the task of incorporating nonbusiness stakeholders. Low membership rosters for most existing committees further reduce the opportunity for negotiators to obtain a full range of private sector views. Finally, USTR’s decentralized management of the committees has left the system without sufficient direction or support. With limited resources devoted to the system’s functioning, agencies are struggling with administrative tasks such as security clearances associated with appointments and 2-year rechartering requirements. To perform the unique role in U.S. trade policy Congress has given it, the advisory committee system’s capacity to provide frank and representative advice needs strengthening. Because important multilateral, regional, and bilateral negotiations are currently under way for which ongoing advisory committee input is expected and desirable, improvements should be made to the existing system, particularly with regard to the timeliness and quality of consultations, gaps in representation, and committee administration. However, given the issues we identify, improving the system’s readiness to play its envisaged role in U.S. trade policy will also require more fundamental reform. As Congress seeks to provide new direction to the President on U.S. trade policy, we recommend that the U.S. Trade Representative, as the lead agency for the committee system, work with the Secretaries of Agriculture, Commerce, and Labor and the EPA Administrator to make the existing system’s consultation process more meaningful and reliable. 1. Specifically, we recommend that the agencies adopt or amend guidelines and procedures to ensure that advisory committee input is sought on a continual and timely basis, consultations are meaningful, committee advice is considered and committees receive substantive feedback on how agencies respond to their advice; 2. continue to increase outreach efforts to fill gaps in committee composition and revitalize membership; 3. streamline the nomination and appointment process for committee members and prevent disruptions in committee activity due to lapses in charters; and 4. provide sufficient technological and human resources to support meaningful consultations and ensure effective functioning of the system. In addition, we recommend that the U.S. Trade Representative work with the Secretaries of Agriculture, Commerce, and Labor and the EPA Administrator to conduct an assessment of the entire system and update it to make it more relevant to the current U.S. economy and trade policy needs. In conducting this assessment and updating the system, USTR, in conjunction with the other agencies, should seek to 1. more closely align the system’s structure and composition with the 2. better incorporate new trade issues and interests, 3. more reliably meet negotiator needs, and 4. better match agency resources to the tasks associated with managing the system. To assist the U.S. Trade Representative and the other agencies in updating the system and improving advisory committee operations, Congress may wish to consider 1. clarifying its intent regarding how to apply the FACA fair balance requirement to the trade policy advisory committee system, and 2. providing an exception to FACA administrative requirements by extending the charter period for the trade policy advisory committees beyond 2 years. We provided draft copies of this report to the following agencies for review: the Office of the U.S. Trade Representative, the Department of Agriculture, the Department of Commerce, the Department of Defense, the Department of Labor, and the Environmental Protection Agency. We received formal comments from USTR, Agriculture, and Commerce (see apps. VI through VIII). The three agencies, as well as Labor and EPA, also provided technical comments, which we incorporated in the report as appropriate. The Department of Defense reviewed the report but did not provide formal comments. USTR and USDA agreed with our overall findings and reported on initial steps they are taking to implement our recommendations. Commerce characterized the report as thorough and fair, but urged us to make a number of modifications. In general, Commerce believes that we underplay member satisfaction with the system. Commerce also took issue with our conclusions on apparent mismatches between the committee structure and the current U.S. economy and agencies’ administrative capacity. Some of Commerce’s comments contain new information or useful clarifications that we have added to the reportfor example, language about the agency’s concerns over security breaches and additional details about outreach efforts. However, as explained in appendix VIII, we do not agree with Commerce’s changes related to members’ concerns about the timeliness and quality of consultations, accountability for seeking and responding to committee advice, and the need to update the system’s structure. We believe that the recent passage of Trade Promotion Authority and the ambitious negotiating plans that have since been announced only heighten the urgency of taking steps to ensure that U.S. negotiators have timely, meaningful, and representative input from the private sector on U.S. trade policy. As agreed with your office, we plan no further distribution of this report until 30 days from its issue date. At that time, we will send copies to appropriate congressional committees and to the U.S. Trade Representative, the Secretary of Agriculture, the Secretary of Commerce, the Secretary of Defense, the Secretary of Labor, and the Administrator of the Environmental Protection Agency. Copies will also be made available to others upon request. In addition, this report is also available on GAO’s Web site for no charge at http://www.gao.gov. If you or your staff has any questions about this report, please contact me on (202) 512-4128. Other GAO contacts and staff acknowledgments are listed in appendix IX. This appendix discusses the scope and methodology for our work. We have included a separate segment at the end of appendix IV providing technical information on our methodology for the survey of committee members. The scope of our review included analysis of 34 private sector advisory committees on all aspects of committee activities, as well as the 4 agencies that currently administer them: the Departments of Commerce, Labor, and Agriculture and the Office of the United States Trade Representative (USTR). The time period covered by our review was generally fiscal years 1999 to 2001. At the time we initiated our review, three policy advisory committees in the second tier were in uncertain stages of activity. The charters for the Labor Advisory Committee and the Defense Policy Advisory Committee on Trade had expired in 2001, but Labor and Defense officials indicated that their agencies were rechartering the committees. A third policy advisory committee dealing with trade with Africa has a charter and seven members, but the committee had not met during our review period. Our survey of members, which focused on committee operations, was under way when these uncertainties existed. Defense and Labor committee members were included in our survey, but the Africa committee’s members were not. Ultimately, the Department of Labor rechartered its committee in February 2002. However, in January 2002, Defense officials informed us that the Department of Defense did not intend to reconstitute its committee, which had 10 members on its latest roster, as part of a departmental effort to reduce the number of advisory committees. Regarding the Africa committee, USTR informed us in March 2002 that it had rechartered the committee and was seeking to appoint more members to it. As a result, the Labor and Africa committees are included in our discussion of the committee’s current structure and count of committees and members, but the Defense committee is not. For our first and second objectives--determining the advisory system’s value to U.S. trade policy and which aspects of the consultation process participants indicate are and are not satisfied--we used three methods of inquiry: interviews, a survey, and document analysis. Regarding interviews, in initial meetings with agency officials and other trade experts involved in the committee process, we were told that the best way to obtain information on how well the trade advisory committee system functions is to interview the key participants. These officials and experts stated that the available documentation on committee activities would not provide as comprehensive a picture as interviews. We therefore first conducted 168 interviews with every type of participant in the process, including 25 executive branch negotiators, 40 other agency officials, 30 committee chairmen, 50 committee members, and 15 trade experts. To gain the perspectives of organizations that do not currently participate in the advisory committee system, we interviewed selected representatives of nonbusiness non-governmental organizations (NGOs) having a demonstrated interest in trade policy. Interest was evidenced by submitting formal comments in response to USTR Federal Register notices or attendance at USTR public briefings. In addition to interviews, we conducted a Web-based survey of 720 committee members and staff liaisons between January and March 2002 to obtain views on matters such as overall satisfaction with committee operations and effectiveness. We surveyed all the members and staff liaisons whose names appeared on lists obtained from the Departments of Agriculture, Commerce, Defense, Labor, and USTR. We developed our questionnaire in November and December, 2001. We put the instrument on a special Web site on the GAO server, activated it on January 17, 2002, and kept it open until March 15, 2002. In all, we received a total of 515 usable responses to our survey, for an overall adjusted response rate of 72 percent. The response rate varied considerably by committee and by tier. For example, seventy-eight (78) percent of tier-3 members responded to the questionnaire, compared with 55 percent of tier-2 members and 57 percent of tier-1 members. Consequently, while we present the aggregated responses for all committee members who responded, we are not generalizing to the universe of all committee members. The survey also allowed for some open-ended responses. Members provided considerable commentary, which is reflected in the body of the report but is not summarized statistically. The survey results and a technical description of the survey methodology are in appendix V. Third, we collected and analyzed documents from four agencies that currently administer committees. Documentation generally covered fiscal years 1999 through 2001. Specifically, we collected and reviewed applicable laws, legislative history, and implementing rules; committee charters and rosters; agency operating procedures and other guidance; meeting notices, agendas, summaries, minutes, and transcripts; interagency decision memos; formal committee reports; and agency correspondence with advisory committees. We also reviewed written responses to an April 2000 Federal Register notice requesting suggestions to improve the advisory committee system. To investigate whether the system matches the current U.S. economy and supports U.S. trade policy needs, we obtained and analyzed U.S. trade data and committee membership rosters, as well as information obtained during our interviews and Web-based survey. Specifically, we examined data on annual industry sector contributions to U.S. gross domestic product from 1974 to June 2000 and determined commodity shares of U.S. imports and exports using data collected by Commerce, the U.S. International Trade Commission, and the Department of Treasury. We defined commodity groups using Commerce’s 4-digit Standard Industrial Classification level codes, and we used Commerce’s determination of how best to match each advisory committee to a commodity group. We determined which industries are over- and under-represented in the committee system with respect to their U.S. import and export contributions by comparing this data with lists showing annual numbers of members on each committee in fiscal years 1999 through 2001. We also discussed the adequacy of coverage of industry sectors with agency and industry officials. We identified trade issues and associated stakeholders that have emerged since 1974 by reviewing academic and agency literature. We discussed the system’s coverage of these issues and stakeholders in interviews with agency officials and selected business and nonbusiness organizations. To examine how well USTR and the other agencies are managing the advisory committee system, we collected and examined available data from USTR, USDA, and Commerce about the time involved in the appointment process for new members. We also interviewed agency officials, negotiators, and committee members about agency practices and other factors that affect the extent of consultation with the advisory committees and the capability of the managing agencies to maintain full and active committees. Finally, for information about agency resources devoted to the committee system we obtained and reviewed the annual reports for each advisory committee for 1999 to 2001 from the General Services Administration (GSA) and conducted interviews with agency officials. We conducted our work from August 2001 through May 2002 in accordance with generally accepted government auditing standards. The Federal Advisory Committee Act (FACA) includes a fair balance requirement that applies to each advisory committee covered by the act. In this regard, the legislative history of FACA shows that the focus of committee membership should be on the groups directly affected by the work of a committee, rather than whether these groups represent business or nonbusiness interests. The broad language of section 135 of the Trade Act of 1974 making FACA generally applicable to the trade advisory committees indicates that the fair balance requirement applies to them. Nevertheless, there is still some legal ambiguity about what this means within the context of the trade advisory committee structure. Aside from lack of clarity in the legislation, at this point, there appear to be too few decided court cases to show any trend in fair balance challenges by nonbusiness groups to the composition of trade advisory committees. FACA, passed in 1972, sets forth certain requirements for Congress to follow in creating federal advisory committees. One such requirement states that any legislation establishing an advisory committee shall require that the membership of the committee be fairly balanced in terms of points of view represented and the functions the committee performs. GSA guidelines implementing FACA indicate that to attain a fair balance of membership on an advisory committee, agencies should ensure that they consider a cross-section of those directly affected, interested, and qualified, as appropriate to the nature and functions of the committee. The legislative history of FACA shows that the fair balance requirement was intended to ensure that persons or groups directly affected by the work of a particular advisory committee would have some representation on the committee. In this regard, the House Report on FACA criticized the composition of an advisory council for only having industry representatives. The report suggested that representatives of conservation, environment, clean water, consumer or other public interest groups should have been present at meetings with government officials to consider a proposed questionnaire regarding national industrial wastes inventory. The Trade Act of 1974, which mandated creation of advisory committees on trade policy, was enacted 2 years after FACA was passed. Section 135(f) of the Trade Act states that the provisions of FACA do apply to the trade advisory committees, with limited exceptions relating to open meetings and public availability of documents. As the fair balance requirement is not one of the excepted FACA provisions, the requirement and the implementing GSA guidance would apply to the trade advisory committees established under section 135 of the Trade Act. This was one of the findings made by one of the two United States courts that have considered application of the FACA fair balance requirement to section 135. Although the language of FACA indicates that the fair balance requirement applies to each advisory committee, there is some ambiguity about what this means within the context of the trade advisory committee structure. Section 135 of the Trade Act called for formation of three different kinds of trade advisory committees for the purpose of creating an institutional framework to ensure that representative elements from the private sector have the opportunity to present their views to U.S. negotiators. The three- tier structure established by section 135, as amended,(1) requires establishment of an Advisory Committee for Trade Policy and Negotiations (ACTPN) whose function is to provide overall trade policy advice (tier 1); (2) authorizes establishment of general policy advisory committees whose function is to provide general policy advice (tier 2); and (3) requires establishment of industry sector and functional advisory committees, as may be appropriate, whose functions are to provide technical advice and information about negotiations over particular products and other factors relevant to positions of the United States in trade negotiations (tier 3). The language of section 135(b), as amended, does show that ACTPN, the tier-1 committee, is to include both business and nonbusiness interests. Specifically, ACTPN is to be broadly representative of the key sectors and groups of the economy affected by trade and “shall include representatives of non-federal governments, labor, industry, agriculture, service industries, retailers, non-governmental environmental and conservation organizations, and consumer interests.” However, section 135 of the 1974 Trade Act and its legislative history do not specifically discuss how the fair balance requirement of FACA was intended to apply to the tier-2 and tier-3 committees. With regard to the general policy advisory committees of tier 2, section 135 authorizes, but does not require, the President to establish such committees for industry, labor, agriculture, services, investment, defense, and other interests, as appropriate. Section 135 states that these committees, to the extent practicable, are to be representative of all industry, labor, agricultural, service, investment, defense, and other interests, including small business interests. Regarding the industry sector and functional advisory committees of tier 3, the President is directed to establish them as appropriate, and similar to the tier-2 committees, to the extent practicable each tier-3 committee is to be representative of all industry, labor, agriculture, or service interests, including small business interests in the sector or functional areas concerned. The language of section 135 suggests that each of the tier-2 and tier-3 committees is to be composed of members involved in the particular sector, and does not indicate any intention to expand these committees to include other interests. The legislative history of the 1974 act, which shows that Congress was concerned that in prior trade negotiations there had not been adequate input from U.S. producers, would appear to support this view. In this regard, the Senate report stated that the purpose of the procedures in section 135 were to “strengthen the hand of U.S. negotiators by improving their knowledge and familiarity with the problems domestic producers face in obtaining access to foreign markets.” Similarly, the House report stated that in past trade negotiations “there has not been adequate input from U.S. producers who are in the best position to assess the effects of removing U.S. and foreign trade barriers on their particular products.” Nevertheless, the legislative history of the 1979 amendments to section 135 indicates congressional interest in broadening representation of the tier-2 and tier-3 committees to include other interests. In this regard, the Senate report states that in establishing the membership of the policy, sector, or functional advisory committees, it was expected that each of these committees “will fully represent the interests of the Government, small business, retailers, wholesalers, distributors, consumers and the general public, as well as labor, industry, agriculture and services, as the case may be.” The House report has similar language and also stated that “ll major recognized organizations, regardless of their point of view, should be invited to participate in appropriate advisory groups.” These statements are consistent with the legislative history of FACA, which shows that the focus of committee membership was intended to be on the groups directly affected by the work of a committee, rather than whether those groups represent business or nonbusiness interests. An additional problem in applying the FACA fair balance requirement to the trade advisory committees concerns the relatively small number of court decisions that have considered the issue. Although several U.S. Courts of Appeal had rejected challenges under FACA to the composition of other federal advisory committees, until 1999 no case had involved a civil- society, fair-balance challenge to membership on a trade advisory committee. Since then, two rulings have been issued, and a settlement agreement has been reached in another case. These dispositions have affected three tier- 3 advisory committees. In November 1999, several environmental organizations brought an action in the Federal District Court for the Western District of Washington, Northwest Ecosystem Alliance v. USTR, challenging the composition of two tier-3 industry advisory committees that deal with forest products. The district court found that fair balance meant balanced representation within each trade advisory committee, not among all advisory committees, and ruled that the two committees should include environmental representatives. Two of the factors the court relied on in making its ruling were that (1) the forest product committees routinely advised the government on trade issues that affected the environment, both nationally and internationally, and (2) the positions supported by the committees were directly contrary to those supported by the environmental organizations challenging their fair balance. Importantly, the court also rejected USTR’s position that fair balance is fulfilled if the membership of an industry sector advisory committee is broadly representative of the industry sector for which the committee was established. The court found that this position contradicted one of the primary purposes of FACA, which was to end industry domination of advisory bodies. To implement its holding, the court ordered USTR to make a good faith effort to expedite the appointment of at least one properly qualified environmental representative to each of the two committees. USTR and Commerce appealed the case to the U.S. Court of Appeals for the Ninth Circuit, and the United States filed a brief in support of the appeal. Nevertheless, the United States later dropped the appeal, and environmental representatives were appointed to the two forest product advisory committees. After the decision in Northwest Ecosystem Alliance, various public interest groups filed a lawsuit in the same federal district court, Washington Toxics Coalition v. USTR, asking the court to require USTR and Commerce to appoint one or more environmental representatives to the chemical and allied products industry sector advisory committee. In March 2001, the parties entered into a settlement agreement in which USTR and Commerce agreed to make a good faith effort to expedite the appointment of one or more qualified environmental representatives to this committee. In response to the Washington Toxics Coalition case, in early 2001, several members of the chemical and allied products advisory committee brought an action before the U.S. District Court for the District of Columbia, Gamble v. Zoellick,asking the court to preclude environmental representatives from becoming members of their committee. The court rejected this position and held that the committee members lacked standing to challenge the appointment of an environmental representative to their committee. In support of its ruling, the court also found that there was nothing in the Trade Act of 1974 that prohibited USTR and Commerce from appointing an environmental representative. The court noted that the appointment of other members was not precluded by the mandatory language of section 135 requiring that the industry sector advisory committees be representative of all industry, labor, agricultural, or service interests in the sector concerned. In this regard, the court endorsed the U.S. government’s position that the language of section 135 gave the government considerable discretion in making appointments to the chemical and applied products committee beyond those required. To date, there have been no further court challenges by environmental or other civil society groups to the composition of trade advisory committees. Without further clarification by U.S. appellate courts or the Congress about how to apply the FACA fair balance requirement to the trade advisory committee system, some ambiguity about this issue will remain. Current executive branch policy is that tier-3 committees are generally not open to nonbusiness groups. A March 20, 2002, Federal Register notice issued by the U.S. Department of Commerce states that with the exception of the 3 committees affected by fair balance challenges—ISAC 3 (chemicals), ISAC 10 (lumber and wood products) and ISAC 12 (paper and paper products)—“non-government organizations do not qualify for representation on a committee.” Regarding to the Washington Toxics Coalition case, the settlement agreement provided that until an environmental representative was appointed, USTR and Commerce could call meetings of the chemical and applied products advisory committee but had to make a good-faith effort to include an interim qualified environmental representative at any such meetings. An interim environmental representative has attended all but one of the nine committee meetings held since the settlement, but he declined to continue to serve beyond the renewal of the committee's charter in March, 2002. The committee--which represents the second-leading manufacturing export sector--has not met since March 13, 2002. One potential environmental representative has applied to serve as environmental representative on the committee, and the application is being considered. To date no appointment has been made. Commerce, USDA, and USTR follow slightly different procedures in screening applicants for advisory committees and in obtaining security clearances. Generally, the vetting process for new members includes an internal agency review and a security clearance investigation performed by the Office of Personnel Management (OPM). This appendix provides information on the nomination and security process, based on data provided by the three agencies. Agencies begin a review process after they receive a nomination or a letter of interest from a prospective member. Figure 11 illustrates the screening process by each agency and the approximate time for applicants to move through different stages of the process toward committee membership. For example, the initial review process averages 70 days at the Department of Agriculture, while the Department of Commerce conducts an initial, 5-day review and then saves time by continuing the review concurrently with the security clearance process. Figure 11 does not include the time spent by committee members in completing the application materials and assembling the documents required for the security clearance because agency data on this part of the process is not systematic or complete. Based on our review of available agency data and interviews with agency officials about their typical experience, we found that the appointment process can regularly take 6 months or longer to complete, if additional time for completing application materials is added. The security clearance process can take about 3 months, according to agency officials and data. Although we found an average waiting time for clearances at USTR of 227 days for the period fiscal years 1999 to 2001, the average wait time for a clearance fell to 84 days when USTR began using the OPM to perform its security clearances in 2000. Department of Agriculture officials said the process of obtaining a clearance takes about 3 months once the completed paperwork is submitted. Security clearance data provided by the Department of Commerce show that the process takes an average of approximately 105 days. All members receive a secret-level national security clearance, following a background investigation from the OPM. The clearance is valid for 10 years. Section I: Committee Activities and Advice Q2) In general, do you feel that too much, too little, or about the right amount of time was devoted to the following activities at your committee meetings? n=480 (number of responses to question) Q3) How satisfied or dissatisfied are you with the extent to which the Executive Branch sought your advice on the following matters during the last 3 years? Members of the ACTPN, TEPAC, and IGPAC are permitted to appoint one or more staff liaisons to help them prepare for and participate in committee deliberations. These liaisons have clearances and meet on their own; some also participate in member meetings. Q4) How satisfied or dissatisfied are you with the opportunities your committee had to provide advice for the following trade agreements/ negotiations? Section III: Committee Composition and Membership Q13) In your opinion, do the current members of your committee generally have similar, different or mixed views on trade policy? ❒ 9HU\VLPLODU 6RPHZKDWVLPLODU ❒ 0L[HGYLHZV ❒ 6RPHZKDWGLIIHUHQW ❒ ❒ Q14) In your opinion, did the dissimilarity of views on your committee make it easier or more difficult for your committee to provide advice to the executive branch? ❒ ❒ ❒ ❒ ❒ Q15) During the past 3 years, how often has your committee provided written or oral advice to the executive branch that reflected a consensus position? Q17) Overall, how satisfied or dissatisfied are you with the following aspects of your committee? Q19) In your opinion, what effect have the following recent actions had on the effectiveness of your committee operations? Section VI: Benefits of Committee Membership Q20) In your opinion, to what extent do you obtain the following benefits as a result of being a committee member? To examine how well the committee structure reflects the current economy, we identified the range of goods or services represented by individual third tier committees and the export and import shares of those goods and services in total U.S. exports and imports. We then compared this data to membership data for each committee obtained from the GSA, which maintains annual reports covering each fiscal year covered by our review. Based on these calculations, table 5 shows the export and import shares as well as the relative percentage of membership for each committee in 2000. The following is GAO’s comment on the Department of Agriculture letter dated August 5, 2002. 1. Regarding our findings that the nomination and appointment process is slow and cumbersome, USDA indicated that it took steps to streamline the process during the most recent rechartering period. We appreciate that the rechartering was completed within 4 months of being started. However, we note that it did not begin until May 2001---more than a month after the APAC and ATAC charters had expired in March 2001. As a result, as our report indicates, none of the six agricultural advisory committees met during the April to October 2001 period. Moreover, we note that USDA indicated in a July 2002 meeting with us that many advisers appointed to the current charter term have yet to receive final security clearances. The following are GAO’s comments on the Department of Commerce’s letter dated August 5, 2002. 1. The Department of Commerce stated that our draft report understated member satisfaction with the timeliness of consultations, arguing that 62 percent of the respondents to our survey reported that consultations were held on a timely basis to a moderate or great extent. We do not agree with this characterization of our survey data. In our survey, we asked to what extent the executive branch timed requests so that committee input could be used in trade negotiations. Respondents answered this question according to a five-point extent scale that ranged from “No extent” through “Very great extent.” Only 25 percent of respondents checked the top two categories, “Great extent” and “Very great extent.” As the Department of Commerce notes, another 37 percent of respondents checked the middle category on the scale, which was “Moderate extent.” If all three of these categories are added together, they total 62 percent of respondents. However, we do not agree that all three categories should be added together. Our report includes the full range of responses to the question, adding together only the top two (very great and great extent) and bottom two (some or little and no extent) categories, and reporting those who checked “to a moderate extent” separately. As the report already notes, 37 percent of respondents reported that they were satisfied to a moderate extent, the third point on a five-point scale. Furthermore, 30 percent checked the final two points on the scale, “Some or little extent” and “No extent.” Consequently, we believe our finding that “onsultations were not always timely enough to have an impact on U.S. policy . . .” is justified. Finally, we are accurately reporting member statements in both the survey and interviews that there were instances when advice was sought after the fact or not sought at all. 2. We agree that the frequency of meetings varies considerably across committees, and we have added language to this report to that effect. With respect to scheduling meetings on a timely basis, we recognize that there is a tension between scheduling meetings far enough in advance and scheduling additional meetings as needed. However, it is clearly important to have timely consultations. 3. We recognize that Commerce, USTR, and USDA have made extensive use of electronic transmissions to provide information to and seek input from committee members. To capture the extent of such communication, we reported on our analysis of such transmissions during fiscal year 2001. Specifically, we calculated the number of times during fiscal year 2001 that USTR officials used electronic means to request comment from advisers, including when USTR sent requests for comment to Commerce’s Industry Consultations Program (ICP) office, which relayed them to advisers electronically. Our analysis yielded a result of 63 requests for comments, rather than the 84 suggested by Commerce in its agency comments. Commerce may have additional information not made available to GAO about the content of each communication that could account for the discrepancy between our counts of agency requests for comment in fiscal year 2001. However, because GAO and Commerce are analyzing the same data for the same period, the number of requests for comment is certainly not 63 plus 84, as Commerce’s comments imply. We welcome the fact that use of electronic means to communicate with advisers is continuing in fiscal year 2002, a period that was outside the scope of our document review. 4. Commerce reports that USTR has been responsive to ICP requests to extend deadlines for ISAC and IFAC members to provide comments on fast-moving issues. However, we note that in earlier interviews ICP officials told us that the reason they have requested extensions from USTR was because members complained that the given deadlines were too short to provide meaningful input. 5. We believe that, if implemented, the technological improvements Commerce and USTR are pursuing to allow sensitive documents to be viewed on a secure interactive Web site could help remedy member concerns over access to key documents required for meaningful and timely advisory committee input. 6. Regarding agency procedures, Commerce does not disagree with our statement that Commerce’s and USTR’s procedures and rules “do not address the principle of timeliness or consulting to the maximum extent feasible.” However, it requests a clarification in the report to the effect that these rules and procedures only apply to the ISACs and IFACs operating at the third, technical tier of the advisory committee system. But in reaching the conclusion that Commerce’s and USTR’s procedures and rules do not address the principle of timeliness and consulting to the maximum extent feasible, we examined procedures that apply to all three tiers of the advisory process, including the first- and second-tier committees having the most severe scheduling problems. Specifically, we examined the procedures for the USTR-only, Commerce-USTR, and USDA-USTR committees. The procedures for USTR-only committees pertain to the first- and second-tier committees having the most acute meeting scheduling problems. The Commerce- USTR procedures pertain to 22 of the third-tier committees. Neither the USTR-only nor the Commerce-USTR procedures address the principle of timeliness or consulting to the maximum extent feasible. The USDA procedures—which apply to six committees at the second and third tier also do not address these issues. Although we recognize that the procedures are based on the legal framework created by Section 135 of the Trade Act as well as other laws and orders, Section 135 (i) states that it shall be the responsibility of the United States Trade Representative, in conjunction with the Secretary of Commerce and other executive departments, “to adopt procedures for consultation with and obtaining information and advice from the advisory committees” on “a continuing and timely basis.” 7. Commerce recognizes that committee meetings frequently include a very full agenda, but stresses that this reflects efforts to balance a variety of factors, including cost, members’ time, and the number of issues to be addressed. We have added language to the report to this effect, but we note that many survey respondents expressed a desire for more time for committees to discuss issues and formulate advice. Survey respondents and interviewees also indicated that the format of meetings is sometimes not conducive to the two-way dialogue that would characterize quality consultations. Formulation of advice is the fundamental purpose of the advisory committees, and we urge Commerce to consider time available for committee deliberations as it seeks to structure meetings to make best use of members’ time. 8. We have added language to the report noting Commerce and USTR’s concerns over safeguarding classified information. 9. We have added language to the report noting that Commerce and USTR already have some mechanisms to bring to the table crosscutting issues including the Committee of the Chairs of the ISACs and IFACs. We note that according to documents we obtained from Commerce, that committee met three times in fiscal year 1999, twice in fiscal year 2000, and twice in fiscal year 2001. 10. Commerce notes that the statute places limits on sharing of advice and information across advisory committees that could inhibit the trade advisory committee system’s capacity for cross-fertilization. Although we agree that Section 135 places some limitations on the disclosure of trade secrets and confidential information, it does not appear to preclude provision of confidential information to designated advisory committee members who possess the requisite security clearances. 11. Commerce asserted that significant departures from committee advice are rare, and that in those infrequent instances committee members are appropriately informed. This point of view is supported by the GAO survey, in Commerce’s opinion. In our survey, we asked committee members how often the executive branch had pursued negotiating strategies that significantly differed from the committee advice. One hundred twenty committee members responding to our survey reported that the executive branch significantly departed from their committees’ advice about half of the time, or more frequently. These 120 members constitute 25 percent of all respondents to our survey, and about one-third of those who provided an answer to this question. While they, by no means, constitute a majority of respondents, they do represent a sizable minority. In any case, significant departures from committee advice do not seem to be a rare event, as Commerce suggests. Our survey then asked a follow-on question for respondents who indicated that there had been significant departures from committee advice. Thirty percent of those answering this question indicated that they had rarely or never been informed of these significant departures. Another 21 percent of those who answered this question indicated that they had been informed of significant departures about or less than half of the time. As a result, we do not agree with Commerce’s statement that committee members are appropriately informed when there are significant departures from advice. Section 135(i) clearly states that USTR “shall inform the advisory committees of significant departures” from committee advice or recommendations. 12. We have updated this report with the information Commerce provided about its practices for handling formal letters from advisory committees. We note that chairmen and members with whom we spoke expressed some frustration about lack of feedback from the government as to how it intends to use or respond to committee advice—a sentiment not inconsistent with Commerce’s practice of providing pro forma responses to committee advice unless it has already made a final decision on policy. Moreover, 21.9 percent of committee chairmen responding to our survey reported that their committees written advice was not acknowledged most of the time (see Q25). In general, the members told us they want to have an opportunity to influence policy before it is finalized and expressed dissatisfaction when feedback on committee input was not substantive or timely in nature. 13. Regarding changes to sectoral and functional committee, our report already notes that only three committees have been created in the past decade to respond to emerging needs. We believe that continued efforts by Commerce and USTR to reevaluate the sectoral and functional advisory committee alignments with the economy and trade policy needs are warranted. 14. Commerce’s position on the services committee is consistent with the statements in this report that certain services negotiators and 70 percent of ISAC 13’s members said that the services sector is well represented in the system. However, we note that some negotiators with whom GAO spoke made a point of saying that the services sector is a large share of U.S. output and trade and that it is only represented in 2 of the 17 industry sector advisory committees; in the scheme of the whole committee system, therefore, they stated that services is underrepresented relative to manufacturing. This report has been updated to note that Commerce has efforts under way to fill the gap in representation of the software industry. 15. Commerce asks us to “note and explain” USTR’s long-standing policy against including foreign-owned or –controlled firms among committee membership. Commerce indicates that this policy is based on the sensitivity of the matters considered by the committees and the possible conflicts that would be experienced by U.S. firms that have foreign owners, and we have added language to this report to that effect. However, we note that first, the U.S. government does not have a uniform policy against inclusion of foreign-owned firms on the trade advisory committees. USDA stated in its technical comments on our draft report that it does not preclude foreign-owned firms from participating in its trade advisory committees. Indeed, USDA indicates that at least one foreign-owned or -controlled firm already participates. USDA officials indicate that although foreign ownership can be considered in the nomination review process, in practice, it was not actually considered during the 2001 rechartering of the six USDA trade advisory committees. Second, as to the rationale for the USTR/Commerce exclusion, we note that there does not appear to be any bar in Section 135, FACA, and GSA implementing regulations specifically precluding participation by foreign-owned or -controlled firms from having representatives on trade advisory committees. The legislative history of Section 135 does not deal directly with this issue, and in their comments, neither USTR nor Commerce bases its long- standing policy on a legal prohibition. Third, while we recognize Commerce’s and USTR’s concerns about the sensitivity of the subject matters considered by the committees, we note that neither Commerce nor USTR has provided us with requested explanations of why the requirements that advisory committee members obtain security clearances and sign a legally binding nondisclosure agreement to protect classified information, along with giving members procedural guidance on safeguarding trade sensitive information, are not sufficient to address these concerns. Fourth, we acknowledge that a majority of our survey respondents expressed reservations about inclusion of foreign-owned firms in the system. However, several members and negotiators still suggested that the long-standing policy barring foreign- owned firms from membership altogether should be revisited, in part because of the contribution to U.S. employment and production that some of these firms provide. Indeed, several U.S. negotiators reported to GAO that they already actively work with foreign-owned firms on an informal basis during trade negotiations, many of which are already members of key trade associations. 16. Regarding participation levels and outreach, Commerce took issue with our position that the number of members specified in each committee’s charter represents a proper level of membership. We note Commerce’s assertion that the “authorized capacity” numbers specified in each committee’s charter are “somewhat arbitrary,” but we hold that they do provide useful guidance regarding committee size. Each committee charter specifies that it “consists of approximately X members,” and each committee’s charter specifies a distinct membership number, ranging from 30 to 50 members. For example, the charter for the Small and Minority Businesses Committee states that it “consists of approximately 35 members,” while the charter for the Chemicals and Allied Products Committee specifies approximately 50 members. Further, while for some trade advisory committees managed by other agencies the charter states that these numbers represent a maximum, this is not the case for the committees that Commerce administers. Even if the numbers specified in the charters do not represent an absolute ideal, our conclusion that the trade advisory committees were at 49 percent of their authorized capacity in fiscal year 2001 highlights the ample room available on the committees that could be used to fill gaps in representation. 17. We appreciate that Commerce provided us with current membership numbers as of August 2002, although the scope of our document review was through fiscal year 2001 (September 30, 2001). We note that according to these current membership numbers, at 48.3 percent of charter levels, the committees administered by Commerce remain just below half of their authorized capacity, and well below the 55 percent of capacity they had reached in fiscal 2000. 18. We recognize Commerce’s efforts to recruit new members and have updated the report to reflect them more fully. These efforts may alleviate the difficulties of maintaining robust and representative membership, concerns that both Commerce and USTR officials expressed during our review. 19. Commerce believes that the draft report’s treatment of the issue of nonbusiness participation may be somewhat misleading and states that the report should contain a more detailed and specific discussion of the congressional delegation in Section 135 of the Trade Act of 1974, particularly in distinguishing the functions and makeup of each of the three “tiers” of committees. We believe this report’s treatment of the nonbusiness issue is fair and accurate and note that appendix II of our draft report contains a detailed discussion of the functions and committee structure for each tier. 20. We recognize that many members expressed satisfaction with the support provided to committees by USTR and other managing agencies, including Commerce and USDA, and we have added language to this report to that effect. However, certain members also expressed concerns about overall leadership of the system and stated that delays or disruptions associated with agency execution of administrative tasks such as rechartering and new appointments were hindering the system’s ability to fulfill its statutory purpose. Our report already notes that unlike USTR, USDA, and Labor Commerce’s ICP successfully avoided disruptions in committee operations typically associated with rechartering. This report has been updated to note that Commerce is taking steps to fill administrative support needs by hiring additional staff. With the renewal of trade promotion authority on August 6, 2002, the U.S. negotiating agenda and resulting demands on the committee system are likely to increase. In addition to the person named above, Dennis Richards, Venecia Rojas Kenah, Kay Halpern, Jon Rose, Janet Lewis, Sharla Draemel, Martin De Alteriis, Richard Seldin, and Janey Cohen made key contributions to this report. The General Accounting Office, the investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to daily E-mail alert for newly released products” under the GAO Reports heading.

In 1974, Congress mandated creation of a private sector advisory system to ensure that representatives from private business and other groups with a stake in trade policy could provide input as negotiations unfolded. The hope was that such involvement would result in trade agreements that Congress could approve with confidence. The law established a three-tier structure of committees to advise the President on overall U.S. trade policy, general policy area, and technical aspects of trade agreements. Four agencies, led by the Office of the U.S. Trade Representative (USTR), currently administer the committee system. According to many negotiators, agency officials, and committee members, the trade policy advisory committee system plays an important role in U.S. trade policy and has made valuable contributions to U.S. trade agreements. Although GAO's survey of committee members found high levels of satisfaction with many aspects of committee operations and effectiveness, more than a quarter of respondents indicated that the system has not realized its potential to contribute to U.S. trade policy. GAO found that consultations could be more timely and meaningful and that the consultation process needs greater accountability. The structure and composition of the committee system have not been fully updated to reflect changes in the U.S. economy and U.S. trade policy. In general, the system's committee structure is largely the same as it was in 1980, even though the focus of U.S. trade policy has shifted from border taxes toward other complex trade issues, such as protection of intellectual property rights and food safety requirements. Leadership direction and administrative support by USTR and the other managing agencies have not been sufficient to ensure that the advisory committee system works reliably. GAO found that negotiators have used inconsistent approaches to solicit committee member views, with some negotiators not consulting with committees at all.

This section discusses key aspects of relevant laws and history related to the implementation of Superfund and the reporting and cleanup of hazardous substances and hazardous waste at DOD installations. In 1976, Congress passed the Resource Conservation and Recovery Act (RCRA), establishing requirements, as well as giving EPA regulatory authority, for the generation, transportation, treatment, storage, and disposal of hazardous waste. Section 7003 authorizes EPA to issue administrative cleanup orders where an imminent and substantial endangerment to health and the environment may exist; if a nonfederal recipient fails to comply, EPA can enforce the order, including fines, by requesting that the Department of Justice (DOJ) file suit in federal court. RCRA also authorizes citizen and state suits, including those to enforce an administrative cleanup order. The passage of CERCLA in 1980 gave the federal government the authority to respond to actual and threatened releases of hazardous substances, pollutants, and contaminants that may endanger public health and the environment. EPA’s program implementing CERCLA is better known as “Superfund” because Congress established a trust fund that is used to pay for, among other things, remedial actions at nonfederal installations on the NPL. Federal agencies cannot use the Superfund trust fund to finance their cleanups and must, instead, use their own or other appropriations. CERCLA does not establish regulatory standards for the cleanup of specific substances, but requires that remedial actions—which are long- term cleanups—comply with “applicable or relevant and appropriate requirements.” These requirements may include a host of federal and state standards that generally regulate exposure to contaminants. CERCLA also establishes authorities for removals, including expeditious response actions by EPA and DOD to reduce dangers to human health, welfare, or the environment such as an emergency response required within hours or days to address acute situations involving actual or potential threat to human health, the environment, or real or personal property due to the release or threatened release of a hazardous substance. Generally, removals are quicker, short-term responses to reduce risks, while remedial actions are the culmination of the full CERCLA process to provide long- term protection of human health and the environment. The National Oil and Hazardous Substances Pollution Contingency Plan (NCP) outlines procedures and standards for implementing the Superfund program. The NCP designates DOD as the lead agency for cleanup at defense installations. CERCLA requires DOD to comply with the law and the NCP to the same extent as a nonfederal entity; thus, the same process and standards for cleanup apply. Where there has been a release of a hazardous substance where DOD is the lead agency, CERCLA section 103 requires DOD to report such releases above reportable quantities as soon as it has knowledge of such release to the National Response Center, and section 111(g) requires DOD to notify potentially injured parties of such release, and promulgate regulations pertaining to notification. In addition, DOD must carry out its responsibilities consistent with EPA’s oversight role under section 120 of CERCLA, including EPA’s final authority to select a remedial action at NPL installations if it disagrees with DOD’s proposed remedy. CERCLA section 120 establishes specific requirements governing IAGs between EPA and federal agencies. The contents of the IAGs must include at least the following three items: (1) a review of the alternative remedies considered and the selection of a remedial action by the agency head and EPA (or, if unable to reach agreement, selection by EPA); (2) the schedule for completing the remedial action; and (3) arrangements for long-term operations and maintenance at the installation. Federal agencies and EPA are required to enter into an IAG within 180 days of the completion of EPA’s review of the remedial investigation and feasibility study (RI/FS) at an installation. An RI/FS is performed at the site, typically after a site is listed on the NPL. The RI serves as the mechanism for collecting data to characterize site conditions; determine the nature of the waste; assess risk to human health and the environment; and conduct treatability testing to evaluate the potential performance and cost of the treatment technologies that are being considered. The FS is the mechanism for the development, screening, and detailed evaluation of alternative remedial actions. Because such study culminates in a record of decision (ROD), EPA has interpreted this requirement as triggered by the first ROD at an NPL site with multiple cleanup activities. EPA and federal agencies often enter IAGs earlier so the agreement may guide the study process as well. IAGs between EPA and DOD include a site management plan, which is an annually amended document providing schedules and prioritization for cleanup of the installation, addressing all response activities and associated documentation, as well as milestones. IAGs also specify requirements for documents throughout the cleanup process, addressing DOD’s submission, EPA’s review, and DOD’s response to EPA’s comments. For “primary” documents, such as the site management plan, RI/FS work plans and reports, RODs, final remedial action designs, and remedial action work plans, the IAG establishes a review and comment process intended to result in no further comment––essentially agency agreement on the document; if either agency disagrees, it can submit the issue to dispute resolution procedures. Hence, for purposes of this report we consider that formal EPA approval is effectively required for these key steps. IAGs do not subject removals to formal EPA approval, although submission of certain documents is required (unless shown impracticable) before an action is taken to allow EPA to comment. Removals are intended to prevent, minimize, or mitigate a release or threat of release, and are not subject to required cleanup goals, whereas a remedial action is intended to implement remedies that eliminate, reduce, or control risks to human health and the environment and generally involve establishing numerical cleanup goals. Removals do not relieve DOD of completing additional steps––such as RI/FS completion––or the full cleanup process for the site with formal EPA approval, if required to ensure long-term protection of human health and the environment. In some cases, however, a removal action does fully address the threat posed by the release, and additional cleanup is not necessary. In 1986, the Superfund Amendments and Reauthorization Act (SARA) added provisions to CERCLA—including section 120—specifically governing the cleanup of federal facilities. Under section 120 of CERCLA, as amended, a preliminary site assessment is to be completed by the responsible agency for each property where the agency has reported generation, storage, treatment, or disposal of hazardous waste. This preliminary assessment is reviewed by EPA, together with additional information, to determine whether the site poses a threat to human health and the environment or requires further investigation or assessment for potential proposal to the NPL. SARA’s legislative history explains that, while the law already established that federal agencies are subject to and must comply with CERCLA, the addition of section 120 provides the public, states, and EPA increased involvement and a greater role in assuring the problems of hazardous substance releases at federal facilities are dealt with by expeditious and appropriate response actions. The relevant congressional conference committee report establishes that IAGs provide a mechanism for (1) EPA to independently evaluate the other federal agency’s selected cleanup remedy, and (2) states and citizens to enforce federal agency cleanup obligations, memorialized in IAGs, in court. Specifically, the report states that while EPA and the other federal agency share remedy selection responsibilities, EPA has the additional responsibility to make an independent determination that the selected remedial action is consistent with the NCP and is the most appropriate remedial action for the affected facility. The report also observes that IAGs are enforceable documents just as administrative cleanup orders are under RCRA and, as such, are subject to SARA’s citizen suit and penalties provisions. Thus, IAGs can provide for the assessment of penalties against federal agencies for violating terms of the agreements. However, at installations without IAGs, EPA effectively has only a limited number of enforcement tools to use in compelling a recalcitrant agency to comply with CERCLA; similarly, states and citizens also lack a key mechanism to enforce CERCLA. Section 211 of SARA established DOD’s Defense Environmental Restoration Program (DERP), providing legal authority and responsibility to DOD for cleanup activities at DOD installations and properties, including former defense sites. The statute requires DOD to carry out the program subject to and consistent with CERCLA section 120. Among other things, the DERP provisions require the Secretary of Defense to take necessary actions to ensure that EPA and state authorities receive prompt notice of the discovery of a release or threatened release, the associated extent of the threat to public health and the environment, proposals to respond to such release, and initiation of any response. Executive Order (E.O.) 12580, Superfund Implementation, was issued in 1987 to respond to SARA. E.O. 12580 delegates to EPA certain regulatory authorities that the statute assigns to the President, while delegating to DOD authority for removal and remedial actions at its facilities, subject to section 120 and other provisions of CERCLA. The E.O. also constrains EPA’s authorities under CERCLA section 106(a) to issue cleanup orders and under section 104(e)(5)(A) to issue compliance orders for access, entry, and inspections by the requirement that the Attorney General, DOJ concur in such actions. In practice, EPA told us it has requested DOJ concurrence approximately 15 times on unilateral section 106 orders to federal agencies and, to date, DOJ has concurred only once, when the recipient federal agency did not object. CERCLA authorizes the filing of civil actions to assess and collect penalties for certain violations—such as failing to provide notice of a release—and section 120 makes each federal department subject to the full procedures and substance of CERCLA. RCRA similarly authorizes the filing of civil actions to enforce— including by assessing fines—orders issued under its imminent and substantial endangerment provision. Nonetheless, as a practical matter, court action is not an available enforcement tool to EPA against another federal agency. Federal law generally reserves the conduct of litigation in which the United States is a party exclusively to DOJ. EPA officials told us the agency has not sought DOJ assistance for such actions because it is DOJ’s policy that one department of the executive branch will not sue another in court. The Federal Facility Compliance Act of 1992, which amended RCRA, authorizes EPA to initiate RCRA administrative enforcement actions against a federal agency for the cleanup of contaminated properties, among other things, as well as subjects federal agencies to RCRA’s existing fines and penalties provisions. The act directs EPA to initiate administrative enforcement actions against federal agencies as it would against a private party. In March 2009, we issued a report that suggested Congress may wish to consider expanding EPA’s enforcement authority to give the agency more leverage to better satisfy statutory responsibilities with agencies that are unwilling to enter into IAGs where required under CERCLA. The report was issued following DOD’s February 2009 agreement with EPA that appeared to resolve a long dispute by determining that the 11 IAGs outstanding at the time would be completed using an IAG between the Army and EPA for Fort Eustis, Virginia, as a template. In addition, EPA agreed to rescind each administrative cleanup order upon the effective date of an installation’s IAG. Soon after this approach for resolving outstanding disputes was agreed to by EPA and DOD leadership, some progress was made in signing IAGs. For example, the Army signed an IAG for Fort Meade Army Base in June 2009. Likewise, the Air Force signed IAGs for McGuire AFB, Brandywine Defense Reutilization and Marketing Office Salvage Yard, Langley AFB, and Hanscom Field AFB, by November 2009. In the absence of the required IAGs, DOD, at some installations, took a few actions toward cleanup and, at others, proceeded with some cleanup activities—including investigations, removals, and remedial actions— without EPA approvals, according to EPA officials. To address continued challenges, EPA issued administrative cleanup orders at four DOD installations, either under EPA’s RCRA authority, or under EPA’s Safe Drinking Water Act authority. According to EPA officials, the agency took the unusual step of issuing the orders because it needed them to fulfill EPA’s cleanup oversight responsibilities at the sites in the absence of IAGs. These administrative cleanup orders were issued as final in 2007 and 2008. In response, DOD challenged the validity of the administrative cleanup orders and asked DOJ to resolve certain questions in dispute between DOD and EPA over the terms of the IAGs and the circumstances under which EPA may issue administrative cleanup orders at such NPL installations. In December 2008, DOJ issued a letter upholding EPA’s authority to issue administrative cleanup orders at DOD NPL installations in general, without discussing whether the facts supported these specific orders. DOJ’s letter also supported including provisions in IAGs, such as the types of provisions that EPA regularly includes in its cleanup agreements with private parties, in addition to those specifically in CERCLA, while stating the opinion that DOD does not necessarily have to agree to all extra-statutory terms. After DOJ’s letter, the Fort Meade Army Base recognized EPA’s 2007 administrative cleanup order under RCRA and gave formal notice to EPA that the Army would comply with the order. However, at about the same time, the state of Maryland filed a lawsuit in December of 2008 against the Army “to force the Army to investigate fully and remediate soil and groundwater contamination resulting from years of mismanagement of hazardous substances, solid waste, and hazardous waste,” and to enforce EPA’s 2007 administrative cleanup order. In November 2009, the state voluntarily withdrew the suit after the Army, EPA, and two other federal agencies signed an IAG for Fort Meade. By the terms of the IAG, EPA withdrew the administrative cleanup order in October 2009. In contrast, Air Force officials at Tyndall AFB and McGuire AFB did not give formal notice of intent to comply with EPA’s administrative cleanup orders and never complied with the terms of the orders. For example, the Air Force stated in a May 2008 letter to EPA regarding the Tyndall order, “the Air Force continues to challenge this Order as lacking legal and factual basis…I have directed my staff and Tyndall AFB to continue to conduct cleanup actions under using our lead agency functions, authorities and responsibilities delegated to DOD.” The Air Force continues to assert that the IAG proposed by EPA does not match the agreed-to template, whereas EPA asserts the IAG does follow the template; both EPA and DOD officials told us the dispute over the IAG relates to the appendices listing the areas to be investigated and, if required, cleaned up. McGuire AFB’s IAG was since signed and became effective December 2009, and EPA’s 2008 administrative cleanup order was withdrawn. While Tyndall remains without an IAG and its administrative cleanup order is still in effect, the Air Force counsel has asserted they are continuing “substantive compliance” with the administrative cleanup order using the CERCLA process—although EPA’s order specifically requires Tyndall to use the RCRA process. EPA officials stated that the agency cannot on its own impose penalties or otherwise compel compliance with the administrative cleanup order at Tyndall; to do so would require concurrence from DOJ to proceed with court action against another federal agency, which is contrary to fed policy. A summary of the current status of IAGs is provided in Table 1. In summary, seven IAGs have been signed and have become effective. There are also four installations that do not yet have signed IAGs as of June 2010 These installations have continued to lack the IAGs required by CERCLA for an extended time frame, and include three Air Force installations and . Because EPA and DOD use different terminology and metrics to report investigative and remedial work at defense installations, determining the status of cleanup at Fort Meade, McGuire AFB, and Tyndall AFB is challenging. EPA’s data suggest that DOD’s progress at these installations was limited primarily to the early study or investigative phase, whereas DOD’s data suggest that some work in the later remedial action or cleanup phase has taken place at these installations. As DOD did not obtain EPA’s concurrence with some of the cleanup actions it took at these installations, it may need to conduct additional work even on reported completed actions as a result of EPA requirements. Twenty or more years after contamination was first reported at Fort Meade, McGuire AFB, and Tyndall AFB, EPA reports that environmental cleanup generally remains in the early, investigative phases of the CERCLA process, with little progress in achieving long-term remediation of contaminated sites at these installations. While DOD’s data suggest that some remedial action work has taken place, EPA and DOD have differing interpretations of the level of cleanup achieved at these installations, in part because the agencies use different terminology and performance metrics to assess cleanup. EPA’s terminology and metrics are based on the Superfund program, including some that are unique to federal facilities, while DOD’s terminology and metrics are based on the DERP program, which DOD is directed to conduct in accordance with CERCLA. Specifically: EPA divides installations into numbered “operable units” (OU), which may represent the type of action to be taken, such as the removal of drums and tanks from the surface of an installation; the geographic boundaries of the contamination; or the medium that is contaminated, such as groundwater. DOD divides installations into smaller geographic areas of contamination called “sites.” These sites are typically scoped narrowly to allow for targeting work on actions that can be accomplished efficiently—for example, a building or waste disposal area where a potential or actual release of hazardous substances, pollutants, or contaminants may have occurred may be considered a “site,” while adjacent buildings with similar operations are considered as separate sites. DOD’s sites are sometimes smaller than EPA’s OUs; therefore there may be multiple DOD sites in one EPA OU. The differing nomenclature can make it difficult to interpret and compare the information DOD reports annually to Congress with what EPA lists in its Comprehensive Environmental Response, Compensation, and Liability Information System (CERCLIS) database on the status of environmental cleanup at NPL sites. For example, as seen in figure 1, EPA reports the progress of cleanup at McGuire AFB by tracking advancements achieved at the installation’s 8 EPA OUs, while DOD reports progress according to advancements achieved at 36 DOD sites. EPA, as the regulator under CERCLA, must track progress made under the statute, and EPA officials said that units in program regulations must have precedence over DOD’s internal system of measuring progress. According to EPA data, most of the OUs at Fort Meade, McGuire AFB, and Tyndall AFB are in the RI/FS phase of environmental cleanup, which as seen in figure 2 occurs early in the CERCLA cleanup process. While the RI/FS phase historically has an average duration of 5.2 years for EPA OUs at federal facility sites on the NPL, many EPA OUs at these three bases have already been in the RI/FS phase for twice that long and are not yet complete. In fact only 3 of a total 37 OUs at these three installations have completed the RI/FS phase of the CERCLA process; those 3 EPA OUs are located at Fort Meade, and none of the OUs at McGuire AFB or Tyndall AFB have completed the RI/FS phase according to EPA. DOD, on the other hand, reports that cleanup is further along at all three of these installations. For example, officials at Fort Meade said that environmental cleanup at their installation is at a very mature stage. In a 2008 report to Congress, DOD reported that Fort Meade had achieved response complete at 61 percent of its 54 sites. The achievement of “response complete,” a DOD term, occurs either late in the CERCLA process after the remedy selected in the RI/FS phase is implemented, or at any time when DOD deems cleanup goals have been met and no further action is required at the site. As we previously reported, we are concerned about the lack of clarity in DOD’s use of this term to describe sites that have been administratively closed, with no physical cleanup. In addition, EPA and DOD report dissimilar pictures of cleanup progress because each agency reports cleanup progress in a different way. For example, DOD reports on removals, which CERCLA defines as short-term and emergency actions to reduce risk, and for which EPA’s formal approval is not required unless specified in an enforceable agreement. These actions are not necessarily designed to provide long-term protectiveness of human health and the environment, and sites where a removal has been conducted are still subject to the full CERCLA process, until no further action is appropriate. EPA tracks removals through its CERCLIS database, which also shows the remaining steps in the full CERCLA process; a removal may be the first response action taken, although one can occur at any time during the process. Furthermore, EPA tracks approved cleanup actions under CERCLA that have been completed or are under way for an entire EPA OU, and records these cleanup actions by EPA OU in the CERCLIS database, where key information is made available to the public on EPA’s Web site. Also, EPA’s current reporting system does not show cleanup progress unless the action has been achieved at all DOD sites within that OU. In contrast, DOD tracks cleanup by site through various cleanup phases as defined in the DERP, which generally aligns with CERCLA but includes additional milestones, and then reports the number of sites in each cleanup phase in its annual report to Congress. For example, Tyndall AFB includes 12 EPA OUs with 12 DOD sites, with an additional 39 other sites that are not contained within any EPA OU. These additional sites are still in stages of preliminary investigation under CERCLA, according to EPA officials; DOD officials said that a number of these are regulated completely as petroleum sites under a separate program that is administered by the state of Florida, but EPA officials said they want to evaluate all of them under CERCLA, to ensure that any non-petroleum contamination that may exist is accounted for and cleaned up under CERCLA. According to EPA officials, Tyndall AFB has achieved no completed cleanup actions at the base, and it recognizes only one RI/FS action as ongoing. In contrast, DOD reported in fiscal year 2008 that Tyndall staff had completed 36 of 51 study actions for sites at Tyndall AFB, amounting to 71 percent of the study phase complete at the base. The fact that DOD measures progress in smaller increments can lead to differing interpretations of cleanup. As we said earlier, DOD counts as progress the completion of each contaminated DOD site located within an EPA OU, although EPA does not count progress until action is taken at all DOD sites in that OU. In June 2009, EPA and DOD formed a working group to review and harmonize both agencies’ environmental cleanup goals and metrics, with the goal of better communication between the agencies regarding cleanup progress at DOD installations on the NPL. DOD officials said they hope that the working group will minimize the inconsistencies between DOD’s and EPA’s goals and metrics. EPA officials said they believe the recommendations of the working group will ultimately result in fewer misunderstandings and surprises between parties that can stall cleanup actions in the future. The proposed timeline for the working group suggests the drafting of proposed recommendations in June 2011. EPA and DOD also report very different cleanup progress at defense installations because some of DOD’s reported claims of completed cleanup phases were never approved by EPA, and therefore EPA does not recognize them. In addition, where DOD has already taken actions, EPA has in some cases found that DOD’s supporting documentation in the record is insufficient for EPA to approve the cleanup actions that DOD has already taken. Specifically at Tyndall, after a change in personnel at EPA, the new project manager reviewed the files and found the documentation was insufficient to support many of the previous decisions made at the base. EPA officials told us that once IAGs are in place at these installations, any unilateral cleanup actions previously taken are likely to be revisited and EPA may require work to be redone. According to EPA officials, DOD and EPA have long agreed that, because EPA has ultimate authority under CERCLA for remedies at DOD NPL installations, EPA approval of key steps toward remedy selection is required. In practice, according to EPA officials, it is difficult for a federal facility to obtain EPA concurrence on its cleanup decisions in the absence of a signed IAG for several reasons. First, from a project management perspective, EPA lacks assurance that it has had adequate involvement in key steps in the process. Second, from a compliance standpoint, EPA told us it must incorporate, among other things, an enforceable schedule and arrangements for long-term management of a remedy into a ROD, in order to approve the selected remedy at a federal facility without an IAG. At least one installation has gained EPA’s concurrence with cleanup actions without an IAG through effective interagency cooperation. However, two of the three DOD installations we examined for this report—Tyndall AFB and Fort Meade—moved forward with cleanup actions, including remedies, without a signed IAG or ROD. For example, EPA’s records for Tyndall AFB show that DOD made decisions at a number of sites without the required concurrence of EPA. Despite the lack of IAGs, DOD submitted a variety of documents for EPA review at each of the three selected installations. However, without an IAG, there are no agreed-upon time frames for review and comment and no overall work plan to provide predictable schedules for DOD or EPA. With an IAG, EPA’s typical primary document review times would be 60 days; however, DOD officials told us that EPA reviews sometimes take longer, with or without an IAG. As a result, DOD officials said that, in some cases, DOD moved forward without EPA concurrence, while in other cases DOD may have delayed planned actions. For example, EPA provided comments on a preliminary assessment of munitions sites at Tyndall AFB that included concerns about how the munitions at these sites could affect other nearby hazardous substances sites. However, EPA took approximately 4 months after receiving the assessment from DOD to submit the comments. As a result, DOD officials told us they finalized the preliminary assessment before receiving EPA’s comments because they wanted to close out the contract. On the other hand, without the predictable schedules provided by an IAG, EPA officials told us they could not predict the flow of documents from DOD they would have to review. EPA officials told us that DOD at times submitted few documents for review, while at other times, an overwhelming number of documents, making it difficult for EPA to allocate resources for review and comment. We could not verify long-term trends in the volume of document submission and in document review times because neither EPA nor DOD maintains a consistent, verifiable, and long-term management system for tracking documents submitted or reviewed. For example, DOD said that the three installations have only maintained document tracking systems for the last 2 to 4 years. DOD officials told us they received EPA approval of some cleanup actions in informal meetings—referred to as partnering meetings—but could not provide documentation. EPA officials noted that these meetings were never intended to replace the formal process mandated by CERCLA and that such decisions were not formally documented, as needed for EPA to approve the proposed remedy selection and as required for the administrative record. CERCLA requires the lead agency, in this case DOD, to establish an administrative record upon which DOD bases the selection of a response action. This record (1) serves as the basis for judicial review of the adequacy of the response action and (2) acts as a vehicle for public participation, since it must be made available for public inspection and comment during appropriate comment periods. Several obstacles have delayed cleanup at the three selected DOD installations in our review. First, the lack of IAGs has made managing installation cleanup and addressing routine matters challenging for both EPA and DOD. DOD contract management issues at some installations have affected how the work at these installations has been scoped and conducted and placed effective and efficient use of the public’s resources at risk, further undermining cleanup progress. In addition, at Fort Meade Army Base, a lack of coordination with EPA and incomplete record reviews resulted in DOD personnel occupying housing at risk of contamination until they were evacuated. Further, the Air Force has failed to disclose some contamination risks at Tyndall AFB promptly, resulting in delays in taking cleanup action. We also found particular problems at Tyndall AFB, where long-standing noncompliance regarding environmental cleanup and notification has contributed to the lack of cleanup progress. Finally, EPA’s ability to address noncompliance by federal facilities is limited by provisions in law, executive order, and executive branch policy. The lack of IAGs has contributed to delays in cleanup progress at the three installations in our review. Without an IAG, EPA lacks the mechanisms to ensure that cleanup by an installation proceeds expeditiously, is properly done, and has public input, as required by CERCLA. For example, DOD officials said that EPA reviewed the proposed remedial action and provided written agreement for the Army’s decision to use monitored natural attenuation—relying on natural processes to reduce the contamination in soil or groundwater without human intervention—as the remedy for groundwater contamination at the Ordnance Demolition Area at Fort Meade, which had been historically used for the demolition of unexploded munitions. However, Fort Meade did not have EPA’s signature on the ROD and did not seek formal public comment. EPA officials said that additional documentation was needed to support the use of that remedy and advised Fort Meade that it was exceeding its authority. The IAG for Fort Meade provided that Fort Meade withdraw this decision document and submit a new one for EPA’s review, which could result in the Army being required to carry out additional cleanup actions for that site. Whereas an IAG would provide for negotiated deadlines designed to reflect the specific complexities at an installation, DOD’s national cleanup goals may drive installations to take actions without EPA approval to meet deadlines. In particular, DOD recently set a cleanup goal for reducing risk or achieving remedy in place or response complete by 2014 for sites under DOD’s Installation Restoration Program at active installations, including those at NPL-listed installations. The Air Force set an even more stringent deadline of 2012 for its sites, which Air Force officials have said is a “stretch goal” imposed to ensure that the 2014 goal is met. These deadlines were not based on evaluations of field conditions, and therefore do not necessarily reflect remaining required cleanup actions. However, DOD’s use of these deadlines has acted as an incentive for DOD to proceed with actions that have not been fully vetted with EPA and the public, according to EPA officials. For example, EPA officials said that, under the pressure of the 2012 deadline, McGuire AFB has proposed monitored natural attenuation, which EPA has not approved, as a remedy for contaminated groundwater at the installation despite not having performed required analyses. EPA typically only approves monitored natural attenuation as a remedy when certain conditions exist, such as a low potential for contaminant migration and a time frame comparable to other methods of remediation. EPA said DOD did not provide evidence of these conditions to EPA, which is necessary for EPA to concur in the remedy selection, as required by CERCLA. One consequence of this gap is that the public lacks assurance that human health and the environment are adequately protected by DOD’s remedy. At installations with IAGs, the Site Management Plans include detailed schedules and become part of the IAG, thus providing a legal basis for when DOD must complete the work. Moreover, with IAGs to provide an enforceable cleanup schedule, DOD must move forward with cleanup or there will be consequences, such as penalties, for violating the terms of the agreements. These legal obligations are a key factor in DOD’s sequencing of cleanup activities for funding. DOD officials told us that, in the early 1990s, the installations that had IAGs were moved to the top of the list for funding, while other installations were considered a lower priority. Also, DOD headquarters makes its funding decisions from budget requests submitted by installations; therefore, if an installation does not have an IAG and does not submit a request for funding for a particular contaminated area, DOD does not consider it in its national funding decisions. DOD contracting management issues have affected how the cleanup work at the selected installations was scoped and conducted, placing effective and efficient use of the public’s resources at risk, and further undermining cleanup progress. Specifically, two of the installations, Tyndall and Fort Meade, have relied extensively on performance-based contracts (PBC) to clean up installations. The third, McGuire, in 2008 awarded a PBC for 21 sites. However, PBCs can create pressure on contractors to operate within price caps and meet deadlines, which may conflict with regulatory review times and encourage DOD to take shortcuts. Both EPA and DOD officials told us that PBCs may frequently be inappropriate for some Superfund cleanup work—particularly in the investigative stages—since there can be a great deal of uncertainty in these phases. For example, initial sampling during a site investigation may lead to the need for extensive follow-up sampling that was not anticipated and therefore not provided for in the contract incentives. While the federal government has advocated the use of PBCs in recent years for procurement of most services, federal acquisition regulations generally requiring the use of PBCs specifically exclude engineering services from this requirement. DOD policy directs the services to use PBCs whenever possible—establishing the goal that PBCs be used for 50 percent of service acquisitions—but acknowledges that not all acquisitions for services can be conducted using PBCs. According to federal guidelines, PBCs are not generally appropriate for work that involves a great deal of uncertainty concerning the parameters of the work to be performed. For example, Air Force guidance establishes the first step in using PBCs is to screen the particular project for suitability, noting that in general, a PBC may not be the right approach when the site is poorly characterized or the project would pose inordinately high risk to contractors, among other characteristics. PBCs are generally better suited to work that has highly prescribed goals, such as the provision of food service or janitorial services. The general intent of PBCs is to allow contractors to determine the best way to achieve specific goals within a certain time frame for a fixed cost. When used in appropriate circumstances, PBCs can reduce costs by allowing contractors flexibility in how they provide the services. EPA officials cited a number of problems resulting from the use of PBCs for cleanup at these three installations. One problem cited is that, when PBCs are used, the contractor typically may not explore the full range of alternatives during the remedial investigation and feasibility study due to the pressure of PBC price caps to reduce the costs involved in developing these alternatives. In addition, EPA officials said, the remedies or proposals put forward by the PBCs tend to be those that do not require construction, such as monitored natural attenuation for groundwater contamination, in order to save money on the contract. For example, EPA officials said that the sole PBC contractor for 21 DOD-designated sites at McGuire AFB proposed in its contract a remedy of “no further action” for soil, sediment, and groundwater for nearly all 21 sites, along with monitored natural attenuation for groundwater at many of the sites; these approaches to address contamination at the sites were proposed prior to completing the remedial investigation, which would include a human and ecological risk assessment, feasibility study, proposed plan, public meeting, and ROD. In addition, EPA has specific guidelines on the selection of monitored natural attenuation as a remedy. Other problems that EPA cited with using PBCs for environmental cleanup work include contractor’s inability to carry out cleanup-related work required by EPA or other stakeholders that was not contained in the original PBC contract, such as installing monitoring wells, without contract amendment; unrealistic time frames for cleanup work that have not been agreed to by EPA or other stakeholders and that create an incentive for rushed work, resulting in possible rework later on; poor quality of documents submitted to EPA, including lack of legal review and routine failure of the installation to perform quality reviews of contractors’ work, which EPA officials said were due to pressure to meet the fixed price aspect of these contracts, and which result in significant redrafting by EPA’s legal staff; and PBC contractors—rather than DOD officials—acting as project managers to the point of decision making, rather than supporting DOD, when critical cleanup decisions require interaction between EPA and DOD officials. In responding to a draft of this report, DOD noted that the department believes it has successfully used PBCs for some environmental remediation and munitions response activities. According to DOD, the PBCs include identifiable and measurable costs, schedules, and outcomes, such as acceptance by DOD and the regulatory agencies. DOD stated PBCs can benefit DOD by providing flexibility of scope, rather than prescriptive methods; allowing DOD to benefit from the expertise and emerging technologies of the private sector in solving problems during various phases of the cleanup process; ensuring cost control with known outcomes at the completion of the encouraging contractors to look for ways to reduce time and cost. Nonetheless, Tyndall AFB officials told us that after shifting toward PBCs for cleanup work in 2004, they are now migrating away from them because there is too much uncertainty in the cleanup work needed at the base. Conversely, the Army told us that in its view, PBCs are better suited for complex work because they foster innovation from the private sector. At Fort Meade Army Base, a lack of coordination with EPA and incomplete record reviews led to the necessity to evacuate military personnel from housing that was at risk of methane contamination due to its construction near a dump. A contractor for Fort Meade building military housing on the base—as part of the Army’s new national privatized housing construction effort—in 2003 discovered an old dump site in the area of the new housing and near an existing elementary school. Prior to construction, the Army Corps of Engineers prepared an environmental baseline survey, but it was later determined that the Corps apparently did not review key historical maps in the possession of Fort Meade indicating a former dump and incinerator in the area. The Corps, in conducting the survey, also apparently failed to use a relevant EPA report, which provided an interpretation of historical aerial photographs to identify potential hazards. According to Fort Meade documentation, once the dump was discovered, the housing contractor attempted to determine the limits of the dump and continued with construction, avoiding building directly on top of the dump site. However, according to EPA officials, Fort Meade did not involve EPA in these assessments prior to construction after the dump was discovered. Nonetheless, EPA, which had an on-site manager at the Fort Meade installation, was aware of the discovery of the dump and did not assert a role in decisions about where to locate housing. For example, EPA did not provide any written advice concerning the matter to Fort Meade. After construction was completed and the housing was occupied, methane fumes were found in 2004 below the ground in soils adjacent to the 20 houses that were built nearest the dump site and elementary school. The Army installed and operated a methane abatement system but in 2005 determined that methane was reaching the homes, and families were evacuated. These houses remain empty, and DOD is monitoring both the houses and the school for methane gas intrusion into indoor air. Thus far methane gas has not been found at an unacceptable level in the school. In addition to methane, Fort Meade has documented other contamination at the dump site, including volatile organic compounds (VOC) in the groundwater, and heavy metals, polychlorinated biphenyls (PCB), and VOCs in soil. Fort Meade has since prepared a preliminary assessment and site inspection(PA/SI) and a draft RI, which EPA has reviewed. While the Army has a policy requiring that the environmental conditions of properties be assessed, it is unclear whether local Fort Meade officials were adequately involved in the preconstruction assessment, which was performed by a contractor to the Corps under the Army’s national housing privatization initiative. While the Army has acknowledged that the preconstruction assessment apparently missed evidence pointing to the incinerator and dump, the Army has not explained the source of the omission—for example whether lack of adherence to policy or shortcomings in coordination and review were contributing factors. As such, it is unclear how the Army could prevent a recurrence of this situation in which review of key documents available to the Army may have averted construction of housing near a waste site. Of the three installations we selected to review, only Tyndall AFB remains without an IAG. Furthermore, Tyndall has delayed cleanup progress by generally demonstrating a pattern of not complying with federal laws and regulations concerning environmental cleanup. In addition, Tyndall has on multiple occasions delayed disclosures about newly found contaminants or associated risks for months or failed to disclose them entirely, furthering delay of cleanup. After 13 years on the NPL, Tyndall AFB stands out as the only one of the three installations that received EPA administrative cleanup orders for sitewide cleanup and has not signed an IAG even though IAGs are required under CERCLA. Following DOD’s issuance in February 2009 of a letter to EPA indicating its willingness to sign IAGs for the 11 installations that did not have them, most of the other installations have resolved differences with EPA and signed IAGs or are close to signing them. As previously noted, in the absence of a signed IAG, Tyndall has delayed cleanup progress by generally demonstrating a pattern of not complying with federal laws and regulations concerning environmental cleanup under CERCLA. For example, Tyndall proceeded with remedies with which EPA had not concurred, did not seek required public input, failed to disclose contamination risks in a timely fashion, and refused to comply with the terms of the EPA-issued administrative cleanup order. EPA officials told us DOD proceeded with cleanup remedies without EPA’s written concurrence—such as signed RODs or other form of documented agreement—to protect human health and the environment, despite knowing that the work may need to be redone. Whereas the CERCLA process requires regulator oversight at federal NPL properties during cleanup activities to provide assurance of such protection, DOD officials said they relied on quarterly partnering meetings with EPA in lieu of written approvals. Tyndall has also issued contracts for work for which EPA hasn’t formally concurred, potentially resulting in rework and jeopardizing public resources. For example, Tyndall authorized a PBC in June 2006 that included selecting and putting a remedy in place at a DDT- contaminated bayou within 5 years without having obtained EPA concurrence on how to proceed with the work. At an informal meeting in April 2003 that included officials from Tyndall, the Army Corps of Engineers, Fish and Wildlife Service, and the National Oceanic and Atmospheric Administration, but at which EPA officials were not present, Tyndall reportedly reached the initial decision to leave the DDT- contaminated sediment in place, with the rationale that having the DDT trapped in the sediment would be preferable to a release that could result from removing the sediment. In January 2009, Tyndall officials put forth the option to EPA officials of dredging the DDT-contaminated sediments from the bayou with the highest concentrations of contamination, proposing to carry out this ecologically sensitive and potentially risky action as a removal action for which Tyndall would not need concurrence from EPA. EPA said that a human and ecological risk assessment—which would estimate how threatening a hazardous waste site is to human health and the environment—would be needed for EPA to evaluate the proposed Air Force removal action and to determine whether it would protect the local population who catch and eat fish from the bayou. Without this information, the adequacy and protectiveness of the response action is in question. Tyndall AFB delayed disclosures about newly found contaminants or associated risks for months or failed to disclose them entirely. The DERP provisions of SARA require defense installations to promptly notify EPA and state regulatory agencies of the discovery of releases or threatened releases of hazardous substances, as well as the extent of the associated threat to public health and the environment. However, we found that Tyndall failed to make such reports. Tyndall was also required to immediately report releases of hazardous substances to EPA according to the RCRA administrative cleanup orders, but did not do so. It also did not provide potentially affected individuals with information on such releases in a timely manner, despite the requirement in CERCLA. Because Tyndall AFB failed to notify EPA of newly discovered releases, cleanup was delayed or conducted without regulatory agency oversight in recent incidents, potentially putting human health and the environment at risk. An example of Tyndall’s failure to notify EPA concerns the presence of lead—a hazardous substance under CERCLA—at the Tyndall Elementary School. Tyndall’s actions have included failing to promptly report to regulators key information about the lead and its threat to public health; failing to take action to prevent children’s exposure to lead shot; and potentially representing inaccurately its actions related to a cleanup, as detailed below: In 1992, children discovered lead shot in their playground at Tyndall Elementary School. Despite the discovery and the SARA requirement, Tyndall AFB officials did not notify EPA. Instead, Tyndall officials worked with county health officials to collect soil samples and Tyndall officials assured the public that the area was safe for children. From 1997 to 2000, ATSDR conducted a health assessment, which was triggered by Tyndall’s listing on the NPL. According to ATSDR officials, ATSDR examined Tyndall records that said the lead shot was removed and clean sand was deposited. As such, ATSDR based its assessment solely on the soil sampling results from 1992, found the contamination below levels of concern, and concurred with Tyndall taking no further action. Tyndall did not conduct any follow-up surveying or sampling of the school area. In 2007, Tyndall issued a base-wide report—the Comprehensive Site Evaluation Phase I—that, based on a records search and visual site survey, identified inactive areas of the base where munitions, munitions constituents, and unexploded munitions may have been released. The report noted that Tyndall Elementary School is located on a portion of a former target range. In 2008, Tyndall initiated the next phase of work, commencing with a site walk. Tyndall representatives observed lead shot and clay target debris on the ground surface of the playground, but Tyndall did not notify EPA of this information and did not take any other action to ensure protection of the health of the children attending the school. In March 2009, officials from the Air Force Center for Engineering and the Environment (AFCEE) visited the base and became aware of the situation and pressed Tyndall to expedite sampling that would assess potential risks. As a result, sampling of the school yard was included in the next phase of work. Once these samples were taken in May 2009, they showed elevated lead in the soils exceeding state standards. The base did not notify EPA until 22 days later—in contrast to the DERP statute’s requirement of prompt notification of a threat, as well as the RCRA order’s requirement, which states that the EPA must be notified immediately of any release of a hazardous substance. Once notified, EPA officials said they called for Tyndall to take appropriate action, including an emergency removal to reduce risk and notifying students’ parents. Tyndall officials told us they initiated funding for a removal action before notifying EPA of sampling results and discussing the action with EPA. In 2009, ATSDR also became involved at the site again, and is currently conducting a health consultation. According to ATSDR officials, EPA requested the consultation in June 2009. Following the request, ATSDR notified its Air Force liaison, who then initiated the formal request on July 7, 2009. When asked about these events, Tyndall officials stated they had always known lead shot could be there, and said they believed EPA also knew of this potential. Tyndall officials told us they did not conduct a cleanup following the 1992 discovery, although they agree that lead shot was found in the playground in 1992 and Tyndall officials subsequently assured parents that the area was safe. Furthermore, Tyndall representatives disagreed with ATSDR’s account that the lead shot had been removed and clean sand placed in the area – information upon which ATSDR relied in focusing its 2000 review on lead in soil exclusively and concluding the site did not pose a health hazard. In 1992, CERCLA and the DERP statute were in effect and well-established, and since lead is a CERCLA hazardous substance, DOD was legally required to conduct any response in accordance with CERCLA and its standards. Thus, Tyndall officials either left the lead shot in place with essentially no response other than to reassure parents of the schoolchildren, or conducted a response outside of CERCLA. While Tyndall officials now state that the lack of response with respect to the lead shot itself was based on its belief that ATSDR found the lead shot not to pose a health hazard, the ATSDR report was not issued until 2000 while Tyndall decided not to conduct a response action years earlier, in 1992. Regarding Tyndall’s lack of action on the discovery of lead shot, Tyndall officials did not take steps until 2009 to protect children from potential exposure, despite their statements that they knew from 1992 forward that lead shot could be present at the school, because they did not believe there were any health risks. Tyndall officials further stated that they believed the ATSDR health assessment found no health risk from the lead shot. However, because ATSDR understood the lead shot had been removed, the ATSDR assessment was based solely on the soil lead levels reported by the Air Force to have been found in 1992 and did not address any subsequent risks from the presence of lead shot after 1992 (e.g., from direct contact and the possibility of increased soil levels from leaching). Moreover, the ATSDR assessment had a narrow objective—to evaluate the potential human health effects associated with exposure to certain environmental conditions at several areas on the base—and was not intended as a substitute for the CERCLA process, which provides for investigations to determine whether a remedial action is required based on both human health and the environment. For example, as ATSDR focused on the likely exposure of children, it discounted certain soil samples with lead levels above its screening values because the agency determined children were unlikely to play in those areas; however, those samples are relevant for CERCLA purposes. Finally, while Tyndall officials have not denied knowledge of the presence of lead shot in the playground prior to June 2009 (when Tyndall reported high lead levels in the soil), they were unable to identify or document when base officials or contractors became aware of the lead shot and clay target debris on the ground surface of the playground. Because Tyndall failed to promptly notify EPA of the release observed prior to March 2009, as required by the administrative cleanup order as well as the DERP provisions of SARA, EPA did not have the information needed to ensure Tyndall’s actions were protective of the health of the schoolchildren. Only at the urging of the Air Force Center for Engineering and the Environment did the base conduct sampling, and only when the results showed high levels of lead in soils did the base inform EPA of the lead shot. In summary, the base failed to take appropriate action to prevent lead exposure until June 2009—months after discovering the debris at the surface during the school year, when children were potentially exposed to lead in this material. Figure 3 shows how visible the lead shot was on the school playground. Tyndall’s failure to disclose the lead at the schoolyard is not an isolated failure to disclose contamination risks. In late 2007, Tyndall discovered the Mississippi Road Landfill but delayed a year before reporting the discovery to EPA in October 2008. Tyndall discovered discarded smoke signal flares, which are hazardous waste under RCRA, in late October 2009 and delayed reporting this to EPA for about a month until November 2009. EPA’s ability to pursue enforcement actions against federal agencies is limited by provisions of law, executive order, and executive branch policy. Specifically, EPA may issue CERCLA orders seeking information, entry, inspection, samples, or response actions from federal agencies only with DOJ’s concurrence. In practice, EPA told us it has requested DOJ concurrence approximately 15 times on unilateral section 106 orders to federal agencies and, to date, DOJ has concurred only once, when the recipient federal agency did not object. Moreover, under federal law, DOJ—and not EPA—is the sole representative authorized to conduct litigation on behalf of the federal government in judicial proceedings, including those arising under CERCLA. This provision, in conjunction with a long-standing DOJ policy against one federal agency suing another in court, has effectively precluded EPA judicial actions against sister federal agencies. However, EPA retains whatever enforcement provisions are contained within an IAG, such as stipulated penalties that may be established within a penalty provision in the agreement. For those installations without an IAG, EPA effectively has no enforcement tools available, without DOJ concurrence, to compel agency compliance with CERCLA. Cleaning up the most seriously contaminated DOD installations is a daunting task, especially when these properties are in ongoing use by DOD components. We recognize that DOD’s primary mission is ensuring the nation’s defense, and that DOD is currently focused on ensuring its components’ readiness for wars in Iraq and Afghanistan. Nonetheless, the environmental problems at the three installations addressed in this report have persisted for more than 20 years since laws requiring their cleanup were enacted. DOD and its components have environmental responsibilities to EPA as well as responsibilities to the public and the military personnel stationed at its installations. Despite some progress in the early investigative stages made by the installations we reviewed, we believe that DOD, the Air Force, and the Army are not fully upholding these responsibilities at the three installations. DOD has expressed its commitment to full and sustained compliance with federal, state, and local environmental laws and regulations that protect human health and preserve natural resources. However, until the current challenges—including the lack of uniform measures for DOD and EPA to report cleanup progress, the absence of IAGs at some installations, the failure to disclose newly discovered contamination at some installations as required by provisions in SARA, and the continued disagreement over proposals for the use of monitored natural attenuation and other nonconstruction remedies, and over DOD’s use of PBCs—are addressed, delays in cleaning up these three installations will likely persist. Section 120 of CERCLA was enacted in 1986 amidst concerns that federal facilities on the NPL were taking too long to get cleaned up and contained key provisions aimed at eliminating stalemates, such as those that were occurring over IAGs. Yet, the IAGs required by law are still outstanding at several NPL installations after more than a decade of effort. While EPA is charged with regulating cleanup of federal NPL sites, without IAGs and lacking independent authority to enforce CERCLA, EPA has little leverage to facilitate compliance at such sites. While EPA ultimately issued administrative cleanup orders at these three installations under other environmental laws, the agency is nonetheless limited in its ability to enforce these orders because DOJ policy generally precludes bringing suit on behalf of one federal agency against another. In the absence of the IAGs, EPA attempted to work with the services over the past decade by offering technical support and in many cases participating in informal meetings with DOD officials, while the services provided numerous documents to EPA. However, we believe that these interactions, while well intentioned, contributed to a less rigorous approach that interfered with the collection of documents such as formal approvals for the administrative record, and led to insufficient communication between the agencies on significant issues such as risk and approvals. Further, without the more predictable time frames as would be provided with an IAG, EPA and DOD resorted to less formal document review processes—including a lack of clarity on document review times and on whether agreements had been reached on key decisions—leading DOD to sometimes move forward in the cleanup process without EPA’s concurrence. Together, these informal approaches contributed to disagreements between the agencies, further delayed cleanup, and resulted in a lack of transparency and accountability to Congress and the public. We are making six recommendations, as follows: To provide greater assurance that cleanup progress is being measured accurately and consistently, and to build off of the existing DOD and EPA working group’s initial efforts, we recommend that the Secretary of Defense and Administrator of EPA develop a plan with schedules and milestones to identify and implement a uniform method for reporting cleanup progress at the installations and allow for transparency to Congress and the public. To ensure that outstanding CERCLA section 120 IAGs are negotiated expeditiously, should the agencies continue to be unable to execute a signed IAG within 60 days of this report, we recommend the Administrator of EPA pursue amendments to E.O. 12580 to (1) delegate to EPA unconditionally the independent authority to issue unilateral administrative orders under section 106(a) to executive agencies, and (2) cause the existing delegation of CERCLA remedial action authorities at NPL-listed sites to DOD to be conditional on, for example, the existence of a signed IAG or on DOD’s submission of detailed monthly reports to CEQ and Congress concerning the status of IAG negotiations at such sites. To ensure that DOD promptly reports new hazardous releases to EPA and other stakeholders (including potentially injured parties, the National Response Center, and the states), we recommend that the Secretary of Defense develop guidance for components concerning the proper notification when a new release is discovered or significant new information about a previously known release is obtained. The guidance should at a minimum address timing and contents of such notice, as well as meet the requirements of CERCLA § 103(a) and 111(g) and 10 U.S.C. § 2705(a). To improve project management at DOD NPL sites regarding the use of contractors, we recommend that the Secretary of Defense ensure that the services make a determination of appropriateness, using Office of Management and Budget criteria and service guidance, before using PBCs for Superfund cleanup. To ensure that DOD NPL sites utilize monitored natural attenuation as the sole remedy at contaminated sites only when it is documented to meet remediation objectives that are protective of human health and the environment, we recommend that the Secretary of Defense direct the services to document compliance with relevant EPA guidance when selecting monitored natural attenuation. To ensure that the document review process is used effectively and to facilitate oversight and transparency between DOD and EPA, even where there are no IAGs in effect, we recommend that the Administrator of EPA establish a record-keeping system for DOD NPL sites, consistent across all regions, to accurately track documents submitted for review, including the status of approvals. While EPA is charged with regulating cleanup of federal NPL sites, it has little leverage to facilitate compliance at such sites. Specifically, when a federal agency refuses to enter an IAG at an NPL site or to comply with an administrative cleanup order issued pursuant to RCRA’s imminent hazard provision, EPA cannot take steps to enforce the law, such as initiating a court action to assess fines, as it would do in the case of a private party. As we suggested in 2009, Congress may want to consider amending section 120 of CERCLA to authorize EPA—after an appropriate notification period—to administratively impose penalties to enforce cleanup requirements at federal facilities. This review provides further reason to emphasize such authorities to facilitate more timely and efficient compliance at federal facilities. We provided a draft of this report to the EPA Administrator and the Secretary of Defense for their review and comment. In written comments, EPA’s Assistant Administrator for the Office of Solid Waste and Emergency Response and Assistant Administrator for the Office of Enforcement and Compliance Assurance indicated agreement with the three recommendations directed at EPA and discussed actions that EPA is taking to address one of them. EPA indicated general agreement with our findings and conclusions, noting in particular that our observations are consistent with its experience at Tyndall AFB. EPA also provided technical comments, which we addressed, as appropriate. EPA’s written comments are included in appendix VI. In written comments, the Deputy Under Secretary of Defense agreed with all recommendations directed to the Secretary of Defense, noting that our report raises several good points, some of which DOD has already implemented. The Deputy Under Secretary also commented on our recommendations directed at EPA, disagreeing with one of them as well as with our Matter for Congressional Consideration. In its disagreement with our recommendation that EPA pursue amendments to Executive Order 12580 if outstanding CERCLA section 120 IAGs are not negotiated expeditiously, DOD suggested that we incorrectly characterized the entire IAG process as flawed due to five outstanding site agreements that represent more complex cleanup issues than most sites. However, we disagree because while we acknowledge that IAGs have successfully been entered into at most DOD NPL sites, DOD’s refusal to enter into IAGs — required by CERCLA section 120—for more than a decade at four of the five sites nonetheless suggests, in our view, that there is a significant problem requiring additional attention by the Administration. DOD disagreed with our Matter for Congressional Consideration that Congress consider amending section 120 of CERCLA to provide additional enforcement authority to EPA because it believes EPA has adequate existing means—including informal tools such as interagency dispute mechanisms, and statutory authorities—to enforce cleanup requirements at federal facilities without a negotiated IAG. However, we disagree with DOD’s position for a number of reasons. For example, despite the informal tools pursued by EPA, a decade passed without negotiated IAGs at the three installations. EPA then resorted to more formal means to attempt to compel cleanup at these installations. Nonetheless, even when EPA attempted to use its RCRA authority, DOD initially refused to comply with RCRA cleanup orders issued by EPA at the three installations and is still in noncompliance at one installation. Moreover, while EPA has remedy selection authority under CERCLA, it has no enforceable schedule to ensure DOD installations make progress on the technical steps leading up to a remedy decision. We therefore believe it is critically important that Congress consider additional EPA enforcement authority to ensure that cleanup is being pursued properly at federal facility NPL sites. In addition, DOD provided technical comments, which we addressed, as appropriate. DOD’s written comments and our responses are included in appendix VII. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the appropriate congressional committees, the Secretary of Defense, the Administrator of EPA, and other interested parties. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov If you or your staffs have any questions about this report, please contact me at (202) 512-3841 or stephensonj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix VIII. We were asked to determine (1) the status of Department of Defense (DOD) cleanup of hazardous substances at selected DOD installations subject to administrative orders and (2) obstacles, if any, to progress in cleanup at these selected sites and the causes of such obstacles. To select installations for more detailed study from the 11 installations that were out of compliance with the Comprehensive Environmental Response, Compensation, and Liability Information System (CERCLA) in February 2009 because they did not have interagency agreements (IAG), we reviewed the 4 that were issued additional Environmental Protection Agency (EPA) cleanup orders under the Resource Conservation and Recovery Act (RCRA) or under the Safe Drinking Water Act (SDWA). These 4 installations are Air Force Plant 44 in Arizona, Fort Meade Army Base in Maryland, McGuire Air Force Base (AFB) in New Jersey, and Tyndall AFB in Florida. EPA and DOD agreed that one of these—Air Force Plant 44, the only 1 of the 4 installations that was issued the SDWA order—was near cleanup completion and we therefore eliminated it from our selection of installations. To determine the status of DOD cleanup of hazardous substances at the three selected installations, we toured the three installations; interviewed officials from DOD, EPA, DOD contractors, and the Public Employees for Environmental Responsibility, a public interest group; and attended an installation’s Restoration Advisory Board meeting. We reviewed numerous laws, guidance, and technical documents, including CERCLA, RCRA, DOD Defense Environmental Restoration Program (DERP) guidance and annual reports to Congress, decision documents, and correspondence between EPA and DOD. We reviewed and analyzed information on cleanup progress from EPA’s Comprehensive Environmental Response, Compensation, and Liability Information System (CERCLIS) information system, the three EPA regions that monitor cleanup at the installations, and from the individual DOD installations. To identify any obstacles to progress in cleanup at the selected installations and the causes of such obstacles, we interviewed officials from DOD, EPA, the Agency for Toxic Substances Disease Registry (ATSDR), the Fish and Wildlife Service, and the Architect of the Capitol, as well as state officials from Florida, Maryland, and New Jersey, and the Public Employees for Environmental Responsibility. We reviewed numerous laws, guidance, orders, and technical documents, including EPA guidance on the appropriate selection of cleanup remedies; decision documents; correspondence between EPA and DOD; internal EPA and DOD documents; ATSDR reports; federal contracting guidelines; and GAO reports on government contracting and project management. We conducted this performance audit from January 2009 to July 2010 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In February 2009 DOD sent EPA an e-mail indicating its renewed willingness to accept the Fort Eustis Federal Facility Agreement as the model for DOD’s remaining site agreements under CERCLA. At that time EPA reported there were 12 DOD installations on the National Priorities List (NPL) without agreed-upon IAGs, as required under CERCLA. (Since then, DOE and EPA acknowledge there are only 11 installations without IAGs for which DOD is responsible. They exclude the Middlesex Sampling Plant, which is the responsibility of the Army Corps of Engineers.) For a detailed list of the 11 DOD installations, see table 2. EPA told us that since February 2009, progress has been made and IAGs were signed and made effective for Fort Meade in Maryland, Naval Computer and Telecommunications Area Master Station in Hawaii, and Whiting Field in Florida. In addition, as of June 2010 the remaining four installations that lack signed IAGs include Andrews AFB in Maryland, Tyndall AFB in Florida, Redstone Arsenal in Alabama, and Air Force Plant 44 in Arizona. The Fort Meade Army Installation is located approximately halfway between Baltimore, Maryland, and Washington, D.C., near Odenton, Maryland, and has been a permanent United States Army Installation since 1917. Fort Meade once occupied approximately 13,500 acres of land, but currently occupies approximately 5,142 acres after parcels of land were transferred to the U.S. Department of the Interior, the U.S. Architect of the Capitol, and Anne Arundel County, Maryland. Fort Meade’s mission is to provide base operations support for activities of over 80 partner organizations from all four Department of Defense (DOD) military services and several federal agencies. Some of the major tenant agencies include the National Security Agency, the Defense Information School, the U.S. Army Intelligence and Security Command, the Naval Security Group Activity, the 70th Intelligence Wing (Air Force), the 902nd Military Intelligence Group (Army), and the U.S. Environmental Protection Agency (EPA). The EPA placed Fort Meade on the National Priority List (NPL) on July 28, 1998, after an evaluation of contamination due to past storage and disposal of hazardous substances at the Defense Reutilization and Marketing Office, Closed Sanitary Landfill, Clean Fill Dump, and Post Laundry Facility. Contamination at these sites included solvents, pesticides, polychlorinated biphenyls (PCB), heavy metals, waste fuels, and waste oils. Moreover, elevated levels of volatile organic compounds (VOC), pesticides, and explosives compounds have been detected in underlying aquifers and low levels of VOCs, including tetrachloroethylene (PCE) and trichloroethylene (TCE), and pesticides have been detected in residential wells located off-base in Odenton, Maryland. On August 27, 2007, EPA issued a unilateral Administrative Order under the Resource Conservation and Recovery Act (RCRA) section 7003 for Fort Meade under its authority to address solid and hazardous wastes that may present an imminent and substantial endangerment to health or the environment. The RCRA Order requires the Army to assess the nature and extent of contamination, determine appropriate corrective measures, and implement those measures. The Order was motivated by the absence of a signed interagency agreement (IAG) between EPA and DOD, as required by section 120 of CERCLA, and which would establish a framework for EPA’s involvement. EPA and the Army could not come to an agreement on the IAG due to several issues. For many years, the Army maintained the position that since EPA took only four sites into consideration for listing Fort Meade on the NPL, it would negotiate an IAG for only those four sites. EPA’s position on the other hand has been that the 14 Areas of Concern on the Fort Meade property and 3 Areas of Concern on the adjacent transferred property should be included in the language of the IAG. Another major disagreement centers on groundwater contamination issues at the base, a common problem on DOD installations. The RCRA Order consequently required the Army to move forward with cleanup of all these hazardous waste sites. Fort Meade officials accepted the order in December 2008. While as of March 2009, Fort Meade was out of compliance with the RCRA Order, in June of 2009, DOD and EPA reached an agreement and an IAG for Fort Meade was signed by all parties. The IAG became effective in October of 2009, after the required public comment period. Per the terms of the IAG, the EPA has rescinded the RCRA Order at Fort Meade. McGuire Air Force Base (AFB) is located in south-central New Jersey near the town of Wrightstown, which is approximately 20 miles southeast of Trenton, and occupies about 3,536 acres within the boundaries of the Pinelands National Reserve. McGuire AFB began operations in 1937 functioning under the control of the U.S. Army until 1948 when the facility’s jurisdiction was transferred to the Air Force. McGuire AFB is home to five units of command, including the 87th Air Base Wing (the host wing), 108th Air Refueling Wing, 305th Air Mobility Wing, 514th Air Mobility Wing, and 621st Contingency Response Wing. McGuire AFB’s mission is to provide joint installation support for McGuire AFB, Fort Dix (Army), and the Naval Air Engineering Station Lakehurst. McGuire AFB is the Department of Defense’s (DOD) first and only joint base to consolidate Air Force, Army, and Navy installations. The base provides airlift capabilities to place military forces into combat situations. The Environmental Protection Agency (EPA) placed McGuire AFB on the National Priorities List (NPL) on October 22, 1999. The initial sites responsible for McGuire AFB’s inclusion on the NPL include: (1) Zone 1 Landfills (comprised of Landfill Nos. 4, 5, and 6; (2) Landfill No. 2; (3) Landfill No. 3; and (4) the Defense Reutilization and Marketing Office. Examples of contaminants found on McGuire AFB sites include volatile organic compounds; polychlorinated biphenyls; trichloroethylene; semivolatile organic compounds; polycyclic aromatic hydrocarbons; total petroleum hydrocarbons; pesticides; and metals, such as nickel and mercury. There are 42 contamination sites in total at McGuire AFB, where 36 sites are located on the base and 6 sites, which are not included in McGuire AFB’s NPL listing, are located at the Boeing Michigan Aeronautical Research Center Missile Facility. According to McGuire AFB officials, the sites that have the greatest priority for cleanup include the landfill sites, which were responsible for McGuire AFB’s listing on the NPL, the Bulk Fuel Storage Area, the Triangle area, the Defense Reutilization and Marketing Office site, the C-17 Hangar site, the Fuel Hydrant Area, and the Pesticide Shop Area. On July 13, 2007, EPA issued a RCRA Administrative Order under section 7003 for McGuire AFB, which became effective on November 26, 2007. EPA issued the order under its RCRA authority to address solid and hazardous wastes that may present an imminent and substantial endangerment to health or the environment. The RCRA Order requires McGuire AFB to assess the nature and extent of contamination, determine appropriate corrective measures, and implement those measures. The Order was motivated by the absence of an IAG between EPA and DOD at McGuire AFB, according to EPA officials. On December 7, 2007, the Air Force notified EPA by letter that it considered the RCRA Order for McGuire AFB to be invalid. The Air Force officials said that the contamination sites listed in the Order, which were also included in a draft IAG for the base, had not been updated since 2001. According to EPA, the RCRA Order was based on site information from McGuire AFB’s outdated documents, since those were the only sources of the information available to EPA at the time. In addition, the officials at McGuire AFB believed that EPA’s issuance of the RCRA Order was politically motivated and that it slowed cleanup progress at the base. For example, they believed that EPA did not approve McGuire AFB’s site management plan (SMP)— related to cleanups under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA)—because the RCRA Order was in place. However, prior to the issuance of the RCRA Order, McGuire AFB had not submitted an SMP and only provided EPA with individual fact sheets for contamination sites on the base. McGuire AFB submitted a revised draft SMP in July 2009. Officials from the Air Force said that the Air Force would continue to exercise its CERCLA responsibilities at McGuire AFB to accomplish the substantive cleanup work that EPA sought to impose in the RCRA Order. However, this did not stop EPA’s involvement with the cleanup activities at McGuire AFB, as EPA continued to work with Air Force officials on the RCRA Facility Investigation phase at McGuire AFB. According to EPA officials, McGuire AFB was not in compliance with the RCRA Order as it had not complied with deadlines set forth in the Order and refused to follow the outlined cleanup process. It is EPA’s opinion that only after EPA’s issuance of the RCRA Order did McGuire AFB begin submitting the required documentation. However, McGuire AFB overwhelmed EPA’s document review process by submitting the required documents all at once. Following DOJ’s letter upholding EPA authority to issue the RCRA Order, as a matter of law, DOD asserted that fulfilling CERCLA requirements fulfilled the Order’s RCRA requirements. Nonetheless, progress was made on the IAG negotiations at McGuire AFB. In October 2009, an IAG was signed by all the appropriate parties for McGuire AFB and it became effective on December 1, 2009, following a public comment period. Tyndall Air Force Base (AFB) occupies approximately 29,000 acres on a peninsula near Panama City, Florida. The base was initially activated in 1941 as a gunnery school for the Army Air Corps, then as an air tactical training school in 1946, and finally designated as an Air Force base in 1947. Currently, Tyndall AFB contains the 325th Fighter Wing, which has a mission of pilot and maintenance training for the F-15 Eagle and F-22 Raptor fighter jet squadrons, weapons system controllers training, and the 601st Air Operations Center activities. Tyndall AFB is also part of the Air Education and Training Center. The Environmental Protection Agency (EPA) placed Tyndall AFB on the National Priorities List (NPL) on April 1, 1997, primarily due to DDT contamination in the sediment of Shoal Point Bayou. Shoal Point Bayou is a tidal creek used as a waterway for barges and small ships to deliver petroleum, oil, lubricant products, and building supplies to the base. In October 1985, the U.S. Fish and Wildlife Service conducted sediment sampling throughout St. Andrew Bay, including Shoal Point Bayou, and found the presence of DDT and DDT metabolites. Then in 1990, the same contaminants were detected in fish, soil, and sediment throughout the Bayou. After multiple investigations, a remedial investigation (RI) was completed for this site in 2002 by the Department of Defense (DOD); however, EPA later deemed the investigation insufficient. Additional investigations have been completed, which found higher concentrations of DDT and metabolites than previously determined. However EPA officials report that the new information on the contamination at Shoal Point Bayou was never integrated into the previous RI findings. Other areas of contamination at Tyndall AFB include the flight line and aircraft maintenance areas, oil/water separators, landfills, fire training pits, petroleum release sites, and munitions testing, disposal, and burial areas. The other contaminants of concern in soil, sediment, surface water, and groundwater at Tyndall AFB include petroleum, DDT, chlordane, TCE, vinyl chloride, pesticides, lead, benzene, arsenic, chromium, barium, and munitions constituents. DOD officials claim that Tyndall AFB currently has 16 active contamination sites after beginning its Installation Restoration Program with 39 sites. Tyndall AFB has many cleanup challenges due to its geography and topography, which cover approximately 110 miles of coastal shoreline with a maximum elevation of less than 30 feet above mean sea level. In addition, approximately 40 percent of the land on Tyndall AFB is wetlands and there are three underlying groundwater aquifers on the base. Tyndall AFB is proceeding at many of the sites by employing a cleanup remedy of natural attenuation. One challenge is that the groundwater at the installation is highly susceptible to contamination and is used as a drinking water source on base. Another challenge is protecting Tyndall AFB’s extensive wetlands and bayous, which includes protecting over 40 species of threatened and endangered plant and animal species. Finally, it is a challenge to control civilian, military, visitor, and trespasser access to areas of contamination on the base. For example, Tyndall AFB has over 110 miles of uncontrolled shoreline where recreational boaters and trespassers may gain access and be exposed to contamination. Furthermore, military and civilian workers may access areas of contamination throughout Tyndall AFB because the installation does not have a land use controls program or physical barriers, such as fences, to prevent unacceptable exposures. Tyndall AFB cleanup and remedial investigation activities have continued in the absence of a signed IAG and without EPA concurrence. On November 21, 2007, EPA issued an Administrative Order under RCRA section 7003 for Tyndall AFB to provide EPA with an instrument to enforce cleanup and which EPA hoped would lead to a signed IAG. EPA issued the Order, which was finalized in May 2008, under its Resource Conservation and Recovery Act (RCRA) authority to address solid and hazardous wastes that may present an imminent and substantial endangerment to health or the environment. The RCRA Order requires Tyndall AFB to assess the nature and extent of contamination, determine appropriate corrective measures, and implement those measures. Tyndall AFB has maintained progress schedules for individual sites, but EPA officials say that Tyndall AFB has not submitted an integrated site cleanup schedule as part of a larger site management plan (SMP) for the entire base. EPA officials stated that outside of their goal to reach an agreed-upon IAG, one of their other priorities is to get Tyndall AFB to submit a draft SMP in the near future. Tyndall AFB submitted one in the past, but according to EPA officials it was deficient, lacked integrated schedules, and only addressed approximately 30 contaminants on the base. However, according to EPA, Tyndall AFB is currently out of compliance with the deadlines and scope of work requirements as defined in the RCRA Order. In addition EPA officials said the Air Force has denied the Order’s legitimacy by calling it a “potential Order.” As of June 2010, Tyndall AFB still does not have a signed IAG. The following are GAO’s comments on the Department of Defense’s letter, dated July 5, 2010. 1. For this recommendation DOD agreed that it is vitally important to track cleanup progress at its installations and to make that information available to Congress and the public in a manner that is transparent and easily understandable. DOD also discussed working actively with EPA through a federal working group. However, DOD indicated that if the working group decides a common metric is essential, DOD would require that the metric meet DOD criteria, such as continuing use of DOD’s site level measure as compared to EPA’s operating unit level of measure, suggesting the agencies are unlikely to implement a uniform method for reporting cleanup progress at the installations. We continue to believe that such uniformity is essential to provide greater assurance that cleanup progress is being measured accurately and consistently across all Superfund sites, and to provide for transparency to Congress and the public. An agency may need more detailed information for management purposes, but information comparable to other Superfund sites is essential to providing adequate transparency. 2. DOD disagreed with our recommendation that EPA pursue amendments to Executive Order 12580 to condition delegation of CERCLA authorities to DOD on the existence of a signed IAG. DOD stated that because all but 5 of the 141 IAGs remain in negotiation, DOD should maintain lead agency CERCLA authority so it can continue executing cleanup actions pending resolution of any IAG issues and indicated its intention to sign the remaining 5 IAGs using as a template an IAG between the Army and EPA for Fort Eustis, Virginia, as has been agreed upon by the agencies. However, given that 4 remaining agreements have been pending for over a decade, we continue to believe that outstanding CERCLA Section 120 IAGs need to be negotiated expeditiously and that amendments to Executive Order 12580 could facilitate such action. 3. For this recommendation, the Deputy Under Secretary agreed that proper notification of new releases that exceed statutory thresholds and significant new information about previously known releases is necessary. DOD noted that DOD guidance on this issue is already in existence; however, GAO reviewed these documents during the engagement and found them to lack adequate specificity for use by installation personnel, particularly in the area of new information about previous releases. Although the Deputy Under Secretary notes that when DOD personnel obtain new information about a previously known release they are already required to review and evaluate any potential impacts to the cleanup process in consultation with relevant stakeholders, to include regulators, we found several instances where DOD personnel did not share such information with regulators in a timely fashion. When we asked why, installation personnel stated they were not required to provide regulators with such information. For example, our report highlights an example of Tyndall’s failure to notify EPA about the presence of lead—a hazardous substance under CERCLA—at the Tyndall elementary school, and failure to take action to prevent children’s exposure to lead shot, among other issues. 4. For this recommendation, DOD agreed and referenced its policy. However, our review found inconsistencies in how this policy was interpreted. While federal guidelines indicate that performance-based contracts (PBC) are not generally appropriate for work that involves a great deal of uncertainty, officials from the Army told us that in their view, PBCs are better suited for complex work because they foster innovation from the private sector. DOD policy directs the services to use PBCs whenever possible—establishing the goal that PBCs be used for 50 percent of service acquisitions. Nonetheless, Tyndall AFB officials told us that after shifting toward PBCs for cleanup work in 2004, they are no longer using them for new contracts because of the uncertainty in the cleanup work needed at the base. 5. For this recommendation, DOD agreed and referenced its DERP guidance, which outlines the process for developing and proposing remedies. The guidance, however, does not provide specific requirements regarding monitored natural attenuation. As DOD notes, when DOD selects monitored natural attenuation as its remedy, DOD is to present the basis for its selection in a ROD or proposed plan. However, DOD and its contractors are not uniformly demonstrating that EPA’s specific criteria for selection of monitored natural attenuation are met before selecting such a remedial alternative, according to EPA. These criteria require that certain conditions exist such as a low potential for contaminant migration and a time frame comparable to other methods of remediation. 6. The Deputy Under Secretary of Defense disagreed with our Matter for Congressional Consideration, in which we suggested that Congress should consider amending section 120 of CERCLA to authorize EPA to administratively impose penalties to enforce cleanup requirements at federal facilities without a negotiated CERCLA interagency agreement. DOD presented several reasons for its position, including its belief that EPA has existing statutory enforcement tools under the Resources Conservation Recovery Act (RCRA) and the Safe Drinking Water Act (SDWA). However, there is little evidence that these other mechanisms have been effective. For example, in 2007 EPA issued administrative cleanup orders under RCRA at all three installations that the services disagreed with and they all initially refused to comply while DOD sought DOJ review of the orders’ validity. The orders stated that an imminent and substantial endangerment from contamination may be present on the sites and required DOD to notify EPA of its intent to comply and clean up. The Air Force and Army did not notify EPA of their intent to comply with the order within the time frame required and stated they would continue to clean up the sites under their CERCLA removal and lead agency authority. After DOJ issued a letter stating its opinion that EPA had the authority to issue the orders, as a matter of law, the Army informed EPA of its intent to comply and initiated work under RCRA at Fort Meade, while the Air Force did not take similar actions for its installations. Subsequent negotiations between DOD and EPA resulted in IAGs at Fort Meade and McGuire AFB. However, at Tyndall AFB, where there is still no signed IAG, DOD continues to refuse to comply with the RCRA order. In regards to SDWA, we recognize there can be installations with contamination that do not threaten a public water supply, and therefore SDWA would not apply. DOD also commented that EPA has authority to negotiate administrative penalties in IAGs under CERCLA and that existing IAGS include stipulated penalties. However, as we stated previously, several of the most challenging sites do not yet have IAGs, including Tyndall AFB. For more than a decade DOD has failed to enter into IAGs required by CERCLA section 120 to clean up DOD National Priorities List (NPL) sites. As we note in our report, without an IAG EPA lacks the mechanisms to ensure that cleanup by an installation proceeds expeditiously, is properly done, and has public input, as required by CERCLA. We disagree that providing EPA with the authority to issue CERCLA penalties at facilities without an IAG will be a disincentive to EPA’s negotiating interagency agreements. EPA has stated on numerous occasions its commitment to complete negotiations for such agreements. Finally, DOD noted that EPA has remedy selection authority at NPL installations regardless of whether the installation has a signed IAG. Despite having authority for choosing a final cleanup remedy, EPA has not been able to force progress toward remedy selection because it has no enforceable schedule to ensure DOD installations make progress on the technical steps leading up to the ROD, which documents the remedy selected for cleanup. Hence, as at the three installations reviewed in this report, installations may not complete cleanup for a decade or more without an IAG. We believe our report demonstrates that EPA has experienced considerable difficulty employing its existing enforcement authorities and that DOD has resisted EPA’s use of such authority to compel DOD to enter into IAGs at NPL sites. Hence, we continue to assert that an expansion in EPA’s enforcement authority is warranted. In addition to the contact named above, Diane B. Raynes, Assistant Director; Elizabeth Beardsley; Pamela Davidson; Michele Fejfar; Justin Mausel; Alison D. O’Neill; Ilga Semeiks; and Amy Ward-Meier made major contributions to this report. Vasiliki Theodoropoulos also made key contributions. This glossary is provided for reader convenience. It is not intended as a definitive, comprehensive glossary of all aspects of the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) process for the cleanup of environmental contamination at Superfund sites. When a federal agency identifies an actual or suspected release or threatened release to the environment on a federal site, it notifies EPA, which then lists the site on its Federal Agency Hazardous Waste Compliance Docket. The docket is a listing of all federal facilities that have reported hazardous waste activities under RCRA or CERCLA. RCRA and CERCLA require federal agencies to submit to EPA information on their facilities that generate, transport, store, or dispose of hazardous waste or that has had some type of hazardous substance release or spill. EPA updates the docket periodically. The lead agency (DOD, in this case) conducts a preliminary assessment of the site by reviewing existing information, such as facility records, to determine whether hazardous substance contamination is present and poses a potential threat to public health or the environment. EPA regions review these preliminary assessments to determine whether the information is sufficient to the likelihood of a hazardous substance release, a contamination pathway, and potential receptors. EPA regions are encouraged to complete their review of preliminary assessments of federal facility sites listed in EPA’s CERCLA database within 18 months of the date the site was listed on the federal docket. EPA may determine the site does not pose a significant threat to human health or the environment and no further action is required. If the preliminary assessment indicates that a long-term response may be needed, EPA may request that DOD perform a site inspection to gather more detailed information. The lead agency (DOD, in this case) samples soil, groundwater, surface water, and sediment, as appropriate, and analyzes the results to prepare a report that describes the contaminants at the site, past waste handling practices, migration pathways for contaminants, and receptors at or near the site. EPA reviews the site inspection report and, if it determines the release poses no significant threat, EPA may eliminate it from further consideration. If EPA determines that hazardous substances, pollutants, or contaminants have been released at the site, EPA will use the information collected during the preliminary assessment and site inspection to calculate a preliminary HRS score. If EPA determines that a significant hazardous substance release has occurred, the EPA region prepares an HRS scoring package. EPA’s HRS assesses the potential of a release to threaten human health or the environment by assigning a value to factors such as (1) the likelihood that a hazardous release has occurred; (2) the characteristics of the waste, such as toxicity and the amount; and (3) people or sensitive environments affected by the release. If the release scores an HRS score of 28.50 or higher, EPA determines whether to propose the site for placement on the NPL. CERCLA requires EPA to update the NPL at least once a year. Within 6 months after EPA places a site on the NPL, the lead agency (DOD, in this case) is required to begin a remedial investigation and feasibility study to assess the nature and extent of the contamination. The remedial investigation and feasibility study process includes the collection of data on site conditions, waste characteristics, and risks to human health and the environment; the development of remedial alternatives; and testing and analysis of alternative cleanup methods to evaluate their potential effectiveness and relative cost. EPA, and frequently the state, provides oversight during the remedial investigation and feasibility study and the development of a proposed plan, which outlines a preferred cleanup alternative. After a public comment period on the proposed plan, EPA and the federal facility sign a record of decision (ROD) that documents the selected remedial action cleanup objectives, the technologies to be used during cleanup, and the analysis supporting the remedy selection. Within 6 months of EPA’s review of DOD’s remedial investigation and feasibility study, CERCLA, as amended, requires that DOD enter into an IAG with EPA for the expeditious completion of all remedial action at the facility. (EPA’s policy however, is for federal facilities to enter into an IAG after EPA places the site on the NPL.) The IAG is an enforceable document that must contain, at a minimum, three provisions: (1) a review of remedial alternatives and the selection of the remedy by DOD and EPA, or remedy selection by EPA if agreement is not reached; (2) schedules for completion of each remedy; and (3) arrangements for the long-term operation and maintenance of the facility. During the remedial design and remedial action process, the lead agency (DOD, in this case) develops and implements a permanent remedy on the site as outlined in the record of decision and IAG. Long-term monitoring occurs at every site following construction of the remedial action. This includes the collection and analysis of data related to chemical, physical, and biological characteristics at the site to determine whether the selected remedy meets CERCLA objectives to protect human health and the environment. For NPL or non-NPL sites where hazardous substances, pollutants, or contaminants were left in place above levels that do not allow for unlimited use and unrestricted exposure, every 5 years following the initiation of the remedy, the lead agency (DOD, in this case) must review its sites. The purpose of a 5-year review, similar to long-term monitoring, is to assure that the remedy continues to meet the requirements contained in the record of decision and is protective of human health and the environment.

Before the passage of federal environmental legislation in the 1970s and 1980s, Department of Defense (DOD) activities contaminated millions of acres of soil and water on and near DOD sites. The Environmental Protection Agency (EPA) has certain oversight authorities for cleaning up contaminants on federal property, and has placed 1,620 of the most contaminated sites--including 141 DOD installations--on its National Priorities List (NPL). As of February 2009, after 10 or more years on the NPL, 11 DOD installations had not signed the required interagency agreements (IAG) to guide cleanup with EPA. GAO was asked to examine (1) the status of DOD cleanup of hazardous substances at selected installations that lacked IAGs, and (2) obstacles, if any, to cleanup at these installations. GAO selected and visited three installations, reviewed relevant statutes and agency documents, and interviewed agency officials. EPA and DOD use different terms and metrics to report cleanup progress; therefore, the status of cleanup at Fort Meade Army Base, McGuire Air Force Base (AFB), and Tyndall AFB is unclear. EPA reports that cleanup at all three installations is in the early investigative phases, while DOD's data suggest that cleanup is further along and, in some cases, in mature stages. EPA and DOD have differing interpretations of cleanup progress because they describe and assess cleanup differently. In particular, while both agencies divide installations into smaller cleanup projects, DOD divides them into units generally smaller than EPA's; therefore, DOD measures its progress in smaller increments. Further, because DOD did not obtain EPA's approval for key cleanup decisions, EPA does not recognize them. Unless key cleanup decisions are justified, documented, and available to the public for review and comment, they are not sufficient under the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA), and once an IAG is in place, some DOD cleanup work may have to be redone. When an agency refuses to enter into an IAG and cleanup progress lags, because of statutory and other limitations, EPA cannot take steps--such as issuing and enforcing orders--to compel CERCLA cleanup as it would for a private party. A variety of obstacles have delayed cleanup progress at these installations. First, DOD's persistent failure to enter IAGs, despite reaching agreement with EPA on the basic terms, has made managing site cleanup and addressing routine matters challenging at these installations. For example, in the absence of IAGs, DOD may fund work at other sites ahead of these NPL sites. Second, DOD failed to disclose some contamination to EPA and the public in a timely fashion, including lead shot on a playground, delaying cleanup and putting human health at risk. Third, the extensive use of performance-based contracts at these installations has created pressure to operate within price caps and fixed deadlines. In some cases, these pressures may have contributed to installations not exploring the full range of cleanup remedies, or relying on nonconstruction remedies, such as allowing contaminated groundwater to attenuate over time rather than being cleaned up. In particular, Tyndall AFB's long-standing lack of full compliance with environmental cleanup requirements, such as notification of hazardous releases and EPA's 2007 administrative order, has been an obstacle to verifiable cleanup of that installation. GAO is recommending, among other things, that EPA and DOD identify options that would provide a uniform method for reporting cleanup progress at the installations and allow for transparency to Congress and the public. EPA and DOD agreed with the recommendations directed at them. GAO is also suggesting that Congress may want to consider giving EPA certain tools to enforce CERCLA at federal facilities without IAGs. DOD disagreed with this suggestion. GAO believes EPA needs additional authority to ensure timely and proper cleanup at such sites.

Electricity-related R&D encompasses both basic and applied research and includes all aspects of electricity generation, including nuclear, fossil, and renewable energy technologies; transmission and distribution technologies; energy storage technologies; and environmental studies of electricity-related issues, according to DOE’s Deputy Assistant Secretary for Utility Technologies. Electricity-related R&D is funded from several sources. For the last 4 years, DOE has provided about $4.6 billion in funding at its national laboratories, at universities, and in co-funded collaborative research with utilities and manufacturers. Over the same period, electric utilities, primarily private and investor-owned, have spent about $2.3 billion; over the last 3 years, state programs have spent about $200 million. Manufacturers of electric utility equipment have also funded electricity-related R&D; however, current estimates of such funding are unavailable. As the electric power industry moves toward deregulation and increased competition, utilities face significant changes. Historically, utilities have operated as monopolies in protected geographic areas. Many of these utilities were regulated by state public utility commissions that approved the inclusion of electricity R&D expenditures in the rate base. By including these expenditures in the rate base, the utilities have been allowed to earn a fixed rate of return on these expenditures. Driven by a combination of factors, the move toward deregulation gained impetus with the Energy Policy Act of 1992, which promotes increased competition in the wholesale power market. Other factors spurring the move toward competition include large differences in electricity rates among utilities; new low-cost electricity generation technologies; and recent experiences in reduced regulation in other industries, such as telecommunications and natural gas. In April 1996, as a result of the Energy Policy Act of 1992, the Federal Energy Regulatory Commission issued a final rule that now requires electric utilities to make their transmission lines accessible to other utilities or power producers for the transmission of wholesale power. It requires that this open access be made available at the same cost that these public utilities incur to transmit their own power. Regulatory commissions in 44 states and the District of Columbia had adopted or were evaluating deregulation alternatives as of June 30, 1996. Electricity-related R&D funding was generally reduced in 1996 by the federal government, the electric utility industry, and most states that we reviewed. Since fiscal year 1993, DOE’s electricity-related R&D budget has increased, except for fiscal year 1996 when it was reduced to near its 1993 level. Meanwhile, the electric utilities began making reductions 3 years ago. Most state programs we reviewed are also experiencing reductions. The primary reasons for the funding declines are overall reductions in federal and state funding and the increased competition expected from the deregulation of the utilities. Current data on the manufacturers’ R&D spending were unavailable. Figure 1 shows the funding for the two largest sources of R&D for which we have data for the last 4 years. DOE’s 1993 and 1996 budget amounts are similar, while the 1994 and 1995 budgets experienced increases. Meanwhile, the utilities’ investments have decreased each year. After approving increases in previous years, the Congress reduced DOE’s electricity-related R&D budget by about 20 percent in fiscal year 1996 compared to 1995. The reductions occurred in electricity-related R&D activities under both of DOE’s appropriations—Energy and Water Development and Interior and Related Agencies. DOE’s 1997 budget request for electricity-related R&D is about 14 percent higher than the 1996 appropriation. Table 1 presents the major R&D program budgets over 5 years. According to the House and Senate Appropriations Committees’ reports on DOE’s fiscal year 1996 appropriations, the primary reason for the decline was to meet overall budget constraints. For example, the House and Senate Appropriations Committees’ reports on Energy and Water Appropriations made repeated references to budget constraints and budget realities in their reports on DOE’s fiscal year 1996 budget for energy supply R&D activities. Compared to DOE’s request, the House Committee recommended a 24-percent decrease, and the Senate Committee recommended an 18-percent decrease. In separate reports, the House and Senate Appropriations Committees responsible for the Interior and Related Agencies Appropriations recommended reducing DOE’s fiscal year 1996 appropriation for fossil energy R&D programs by about 10 percent below the fiscal year 1995 level and stated their intent to continue reducing this program by a similar percentage each year for the next several years. According to these reports, the reductions will permit the agency to gradually phase down to a funding level more in line with the recommendations of the legislative committee of jurisdiction in the House. DOE’s fiscal year 1997 overall budget request for electricity-related R&D is greater than the 1996 appropriations; however, the budget request for some technologies decreased. For example, the request for the renewable energy and energy-efficiency programs is $146.4 million (or 35 percent) greater, whereas the request for the fossil energy programs is $31.8 million (or 10 percent) less. DOE’s budget attributes the reductions in the fossil energy programs to congressional guidance to reduce these programs by 10 percent per year. R&D spending by the nation’s investor-owned utilities has declined by nearly one-third in 3 years (from 1993 to 1996) after being level in real dollars for the previous 10 years. We gathered data from 80 companies representing the 112 largest operating utilities from a total of 3,000 utilities. These 112 investor-owned utilities, which are privately owned, account for over 93 percent of all nonfederal utility R&D spending and are responsible for about three-quarters of all electricity sales. They reduced their spending for R&D from about $708 million in 1993 to about $476 million in 1996. In 1992, the National Association of Regulatory Utility Commissioners recommended that utilities devote 1 percent of their revenues to R&D. In 1993, 6 of the 112 investor-owned utilities met that target, but since then all 6 have substantially cut back their R&D spending. In 1994, utilities on average devoted about 0.3 percent of their revenues to R&D. Utility R&D managers told us that this average will most likely continue to decline. Of the 80 companies we contacted, the R&D managers of 38 companies predicted cutbacks in R&D spending after 1996, while the managers of only 2 companies predicted increases. The managers from the remaining 40 companies were either unsure, thought their expenditures would remain about the same, or did not provide the information. According to utility R&D managers who were asked why their budgets were being reduced, the main reason was that their companies are preparing for deregulation and competition by cutting costs wherever they can. In the past, utilities were allowed to earn a fixed rate of return on all R&D projects that the public utility commission allowed in the rate base. In a more competitive marketplace, utilities will be forced to price electricity to compete with other utilities and independent power producers. As a result, R&D managers evaluate potential R&D projects on the basis of their likelihood of providing a near-term return to the utility that will allow them to reduce electricity rates. Increased competition was cited as the primary reason for the biggest cutbacks to date by utilities in California, New York, and Florida. The 13 investor-owned utilities in these states have been among the leaders in R&D investments, accounting for 39 percent of the R&D funded by investor-owned utilities in 1993. But they have reduced their R&D spending by 52 percent since 1993. According to utility R&D managers in New York and California, they currently charge customers considerably more than the average price for electricity, and they are under pressure to cut costs in order to be able to compete in a deregulated market. Florida’s major utilities have eliminated nearly all of their R&D funding in order to be cost-competitive with each other and with other electricity suppliers in the region. Other reasons given by 10 companies’ R&D managers for reductions in their R&D were that no new DOE co-funded projects were being initiated and ongoing projects were either reaching completion or being cut back. These projects included coal technology development, renewable energy, and other projects for advanced electricity generation technology. The electricity-related R&D programs that we reviewed at the state level are also experiencing reductions. Of the 11 large programs in the nine states that we reviewed, 7 have been reduced in the past 3 years. Overall the programs have seen a 30-percent reduction in funding, from $83 million to $58 million, since 1993. Most of these programs involved energy-efficiency R&D, and some involved generation technologies of particular interest to that state, such as coal power and renewable energy. The state program officials attributed the declines in these programs to the decreases in major funding sources: (1) utilities’ contributions; (2) oil overcharge revenues; (3) co-funding available for R&D projects from DOE, the Electric Power Research Institute (EPRI), utilities, and industry; and (4) state appropriations. For example, the budget of the Empire State Electric Energy Research Corporation, funded by voluntary contributions from New York utilities, was cut nearly in half, from $19 million to $10 million, between 1993 and 1996. The California Institute for Energy Efficiency, which has funded energy-efficiency R&D at Lawrence Berkeley National Laboratory and various California universities, has also been affected by cost-cutting. The Institute’s primary source of funding was about $4 million per year from California utilities. By late 1994, the utilities no longer provided funding. The Institute is maintaining a skeleton operation using carryover funds but will be unable to continue if another source of funding is not found by the end of 1996. Information on spending on electricity-related R&D by the manufacturers of electric utility equipment is unavailable. Data from manufacturers are considered proprietary and therefore difficult for organizations that collect and analyze R&D financial information to obtain. The organizations, which include the National Science Foundation and DOE’s Energy Information Administration, said that they had data on the manufacturers’ energy R&D but could not isolate the electricity-related R&D spending. The most recent such information available was an EPRI study that estimated a 1988 total for all U.S. manufacturers of $200 million. In a restructured industry in which other companies are reducing R&D spending, manufacturers may take on the development of new products. In the absence of current data, the degree to which this is occurring, if at all, is uncertain. Utility, EPRI, and DOE officials told us that on the one hand, the manufacturers are increasingly being relied on to meet technology needs, especially by independent power producers, which are producing a growing portion of the nation’s electricity but are generally not investing in R&D themselves because they operate on a thin profit margin. On the other hand, officials from these organizations told us that electricity-related manufacturers may not invest in new technology R&D for the following reasons: (1) the cutbacks in the availability of co-funding to help support projects; (2) the restructuring of the electricity industry, which has created uncertainties in the domestic market; and (3) the difficulty of competing in international markets where foreign competitors have the strong backing of their governments. Concurrent with the declines in funding, a shift in the types of R&D being funded has also occurred, primarily resulting in a decrease in collaborative and longer-term projects. Many utilities are shifting their R&D from such projects to proprietary R&D and to projects with a short-term payback. In addition, as a result of these changes and last year’s reductions in DOE’s funding, advanced technology projects in the six technology areas we reviewed were often delayed, scaled back, or canceled. Given the inherent difficulties in measuring the benefits of R&D, the economic consequences of these program changes are unclear. According to many utility R&D managers, their companies have been shifting the focus of their R&D from collaborative projects benefiting all utilities, to proprietary R&D, giving their individual companies a competitive edge. R&D managers at more than half of the 80 utilities we contacted reported reducing funding for collaborative R&D. Some R&D managers said they believe that continued investment in R&D that could benefit all companies would put their company at a competitive disadvantage in comparison with other utilities and with independent power producers that are not making such investments. This shift is reflected in the declining support for EPRI, which is the utilities’ main vehicle for collaborative R&D. According to a National Rural Electric Cooperative Association official, many of his members belong to EPRI and look to it for larger, industrywide innovations. Traditionally amounting to more that half of the utilities’ R&D dollars, the utilities’ contributions to EPRI over the last few years have declined faster than the utilities’ R&D spending overall. Between 1994 and 1996, membership contributions to EPRI declined by nearly 30 percent, from $424 million to $300 million, and EPRI officials expect a further decline in 1997 (see fig. 3). Of the 80 utility companies we contacted, 12 dropped out of EPRI between 1994 and 1996, but most remained members and simply decreased their contributions. To address the changes that are occurring, EPRI has tried to encourage membership for independent power producers—which an EPRI official estimates will account for more than 35 percent of future generating capacity—but such efforts have been unsuccessful. In addition, EPRI plans to establish a taxable subsidiary that can participate in proprietary R&D. According to many utility R&D managers, their companies are also shifting the focus of their R&D away from long-term, advanced-technology R&D, like the advanced gas turbine and new fuel cells, to short-term projects that will be profitable and provide a competitive edge in the near term. The R&D managers at about half of the 80 utility companies we contacted reported such a change. In fact, the R&D managers at the nation’s two largest utilities, Pacific Gas & Electric and Southern California Edison, said that their advanced-technology R&D programs have been eliminated. R&D managers from 52 of the 80 utility companies we contacted expressed concern that if the trend in funding decreases continued, it would result in slowing technology development, sacrificing future prosperity to meet short-term goals, and failing to meet national energy goals. In addition, DOE officials said that the reductions in the renewable and fossil energy programs will delay penetration of technologies into the market and change the way that some projects are being carried out. With the move toward deregulation, some R&D managers said that they are more concerned with whether their companies will continue to exist in the face of widespread restructuring and mergers than with the potential long-term benefits from advanced technology that may take 8 or more years to develop and market. They also said that they view the shift to short-term R&D as part of the recasting of utility companies as businesses rather than regulated public-service providers. A 1996 DOE study found that private industry in general is shifting its R&D priorities away from the longer-term benefits of basic and applied research to an emphasis on product development and process enhancements supporting shorter-term market strategies and “bottom lines.” The R&D managers at some utilities told us that their companies are shifting from R&D activities related to long-term, advanced-technology power generation R&D because, under restructuring, they will become transmission and distribution companies and will no longer be involved in power generation. Thus, some utilities see themselves purchasing new power rather than adding generating facilities. Additional reasons for the shift mentioned by some utilities’ R&D officials were that there is no immediate need for additional electricity supplies, the available gas-turbine technology is adequate as long as natural gas is plentiful and relatively inexpensive, and market uncertainties are associated with deregulation. EPRI’s R&D programs for advanced power generation have also been affected by cutbacks in the utilities’ contributions. For example, the budgets for fuel cells, coal gasification, advanced gas turbines, and wind and solar power have declined by a total of 66 percent, from $40.7 million to $13.9 million, within the past 3 years. EPRI program managers said that they no longer have funds to initiate new projects; instead, their role is increasingly one of information transfer rather than R&D funding. The projects that we reviewed in six technologies in which DOE participated were beginning to be delayed, scaled down, or canceled as a result of funding reductions, according to DOE, state, and industry officials that we contacted. We chose to review these areas because, according to the 1995 DOE Task Force study, the projects in these areas have a high or medium long-term potential for meeting the national energy goals and because they were significantly reduced in the utilities’ budgets. The technologies reviewed were fuel cells, coal gasification, advanced gas turbines, wind power, photovoltaics, and electricity storage (see app. II for details). The reductions in DOE’s funding are delaying the development of several technologies, according to DOE officials. For example, the unavailability of DOE and EPRI co-funding is delaying the development of a fuel-cell system—whose goal is the highly efficient, environmentally benign conversion of fossil fuel to electricity. Funding reductions are also delaying the development of one of DOE’s fuel-cell vehicle programs and a demonstration of superconducting magnetic energy storage, whose goal is a highly efficient new technology for storing electricity. The funding reductions by DOE, utilities, and EPRI are resulting in the scaling down of collaborative projects with industry for the development of cost-efficient photovoltaic systems, which convert sunlight directly to electricity. Several projects are being scaled down, such as (1) a program to reduce the cost of photovoltaic manufacturing and (2) a center that aids in designing new photovoltaic applications. The funding reductions are also resulting in the cancellation of two programs to encourage wind-power development. A collaborative program involving DOE, utilities, and EPRI to test new wind turbines in utility settings will be terminated following the completion of the three projects currently under way. A program to support utilities’ wind turbine purchases to reduce the utilities’ perceived risk of introducing a new and unfamiliar technology has also been eliminated. Utility R&D managers and industry, DOE, and state government officials who expressed concerns about the funding of electricity-related R&D suggested alternative funding sources. They are (1) a state-administered surcharge on all retail sales of electricity within the state and (2) a nationwide non-bypassable wires charge that could provide an alternative funding source for EPRI. Several states that are considering deregulating their utilities have proposed surcharges to fund public-benefit R&D; the states include California, New York, Massachusetts, and Rhode Island. For example, in January 1996 the California Public Utilities Commission published its deregulation proposal. It recognized that California utilities’ R&D budgets have decreased significantly in the transition to a competitive market. The Commission’s proposal calls for a non-bypassable surcharge to be instituted no later than January 1, 1998, on all retail electricity sales to fund public-benefit R&D and energy-efficiency activities. The surcharge would fund R&D that served a broad public interest which might otherwise be lost in the transition to a more competitive market place. The proposal calls for establishing a consortium or public authority to administer the funds but does not specify a funding level. However, some utility R&D managers and state and EPRI officials pointed out weaknesses in the state-by-state administration of surcharges. These officials believe that although surcharges may be suitable for programs that focus on locally available natural resources, local conditions, and partnerships with local industries, the states’ administration of more broadly based programs would likely be inefficient, uncoordinated, and duplicative and not achieve the critical mass necessary for projects of nationwide scope. Also, they are concerned that if some states implemented a surcharge and others did not, the problem of “free riders” would continue, where some would receive the benefits without helping to pay for the R&D, putting states that did pay at a competitive disadvantage. Some utility R&D managers and state and EPRI officials suggested that a non-bypassable national wires charge could provide an alternative funding mechanism for EPRI and longer-term collaborative R&D. It would ensure that those who do not fund R&D do not achieve a competitive advantage over those who do. Under this proposal, a small charge would be assessed on all electricity entering the transmission grid, whether it be interstate or intrastate. Furthermore, the National Association of Regulatory Commissioners in November 1994 adopted a resolution recognizing the need for a system of support for public benefits, which include electricity-related R&D in the restructured electricity industry. Subsequently, an Association official told us that in commenting on the Federal Energy Regulatory Commission’s open-access rulemaking, the Association recognized that a nationwide wires charge was one possible technique to fund public benefits. The Gas Research Institute, the R&D counterpart to EPRI for the natural gas industry, is funded by a somewhat similar charge on gas flowing through interstate pipelines. This pipeline charge has enabled funding for the Institute to be maintained despite reduced regulation. Recently, the Institute has encountered problems with this funding mechanism because individual pipeline companies are allowed to reduce their payments to the Institute if their rates are discounted due to competition from other pipeline companies. As a result, the Institute experienced a 21-percent shortfall in its 1996 R&D budget. In an order issued on May 3, 1996, the Federal Energy Regulatory Commission approved an amended R&D program for the Institute. Many utility R&D managers with whom we spoke, although generally reluctant to support any additional charges for electricity, said that a non-bypassable wires charge would be a more equitable way to provide funding than the current system, to which some utilities and independent power producers were not contributing. The managers also said that if there were a wires charge, they would like to have considerable say over how the money was spent. We transmitted a draft of this report to the Secretary of Energy for review and comment. We received written comments from DOE’s Assistant Secretary, Energy Efficiency and Renewable Energy, who stated that the agency had only minor editorial comments on the draft. We incorporated these suggestions where appropriate. We conducted our work from October 1995 through July 1996 in accordance with generally accepted government auditing standards. Appendix I describes the objectives, scope, and methodology of our review in detail. Appendix III lists the major contributors to this report. As arranged with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after the date of this letter. At that time, we will send copies to the appropriate congressional committees, federal agencies, and other interested parties. We will also make copies available to others on request. If you have any questions about this report, call me at (202) 512-3841. Our objectives were to determine (1) what changes have occurred in the amount of electricity-related research and development (R&D) funding and the primary reasons for these changes and (2) what has been the impact of these changes on the types of R&D being funded. For the impact of changes to the Department of Energy’s (DOE) R&D, we agreed to provide information on six technologies. You also wanted to know, given these changes, what alternate funding approaches R&D managers and others have proposed. To obtain information on the federal electricity-related R&D programs, we contacted DOE officials and program managers and extracted electricity-related R&D data from DOE’s budget documents. Because DOE does not separately identify electricity-related R&D, we relied heavily on an analysis of utility-related activities for fiscal year 1993 performed by DOE’s Office of Energy Efficiency and Renewable Energy. We used this analysis together with other budget data to estimate DOE’s budget for electricity-related R&D activities for fiscal years 1993 through 1997. We did not gather data on possible electricity-related R&D funding by other federal agencies, such as the Department of Defense, National Aeronautics and Space Administration, and Department of Commerce. To determine changes in electric utilities’ R&D spending, we analyzed data on R&D expenditures collected by the Federal Energy Regulatory Commission from investor-owned utilities through 1994. These utilities accounted for about 93 percent of nonfederal utilities’ R&D spending. For information for 1994 through 1996, future trends, and other responses, we interviewed and obtained data from utility R&D managers or other corporate officials at 80 companies representing 112 investor-owned electric utilities, which accounted for over 99 percent of investor-owned utilities’ R&D. We also interviewed and collected information from corporate officials, program managers, and officials of the Electric Power Research Institute (EPRI), and officials of trade associations representing municipal utilities and rural electric cooperatives. To gather information on state-sponsored electricity-related R&D programs, we interviewed and obtained data from officials involved with 11 state programs of significant size from 9 states—California, Florida, Illinois, Iowa, New York, North Carolina, North Dakota, Ohio, and Wisconsin. The 11 programs include the Empire State Electric Energy Research Corporation, New York State Energy Research and Development Authority, Florida Solar Energy Center, Ohio Coal Development Office, California Energy Commission’s Energy Technology Advancement Program, California Institute for Energy Efficiency, Energy Center of Wisconsin, North Carolina Alternative Energy Corporation, North Dakota Lignite Energy Council, Illinois Clean Coal Institute, and Iowa Energy Center. To obtain information on industry spending on electricity-related R&D, we contacted manufacturers’ trade associations and private nonprofit research organizations, including EPRI. We also interviewed officials from several companies involved with specific technologies we selected to examine. In addition, we reviewed studies published by and contacted officials at other organizations, including DOE, the National Science Foundation, the Industrial Research Institute, and the Energy Information Administration. To determine the effects of the changes on the types of R&D being funded, we interviewed and obtained information from EPRI and DOE program managers, utility R&D managers, and industry and state officials. To determine the impact of changes to DOE’s R&D, we selected six technologies—fuel cells, coal gasification, advanced gas turbines, wind power, photovoltaic, and electricity storage. We selected these technologies because the Secretary’s 1995 Task Force study designated them as having high and medium long-term potential for meeting national energy goals and they have experienced funding reductions by utilities and EPRI. These technologies are advanced electricity generation technologies, except for electricity storage, which EPRI officials predict will be of increasing importance in an era of deregulation. To obtain information on alternative funding approaches, we relied on interviews and documents from utility R&D managers and state, DOE, EPRI, and industry officials, as well as the National Association of Regulatory Utility Commissioners. We also discussed the Gas Research Institute’s fuel line charge with Institute and Federal Energy Regulatory Commission officials. We did not determine whether the changes and trends in funding levels are appropriate and therefore whether the alternative funding proposals are necessary. We conducted our work from October 1995 through July 1996 in accordance with generally accepted government auditing standards. We reviewed the following six technologies—fuel cells, coal gasification, advanced gas turbines, wind power, photovoltaic, and electricity storage. Projects in these six technologies are beginning to be delayed, scaled down, or canceled as a result of funding reductions, according to DOE, state, and industry officials that we contacted. Fuel cells is a new generating technology that converts the energy of chemical reactions directly into electricity. It is intended to be the most efficient, environmentally benign of the fossil-fueled technologies. As fuel cell applications are being tested, developers are seeking ways to bring down the cost so that the systems can compete with other technologies. The various fuel cell technologies—phosphoric acid, molten carbonate, solid oxide, and proton exchange membrane fuel cells—are at different stages of development. Phosphoric acid fuel cell technology is on the market as relatively small power plants in hospitals, research laboratories, and remote sites. Neither EPRI nor DOE is any longer involved with this R&D. However, DOE is cooperating with the Department of Defense on a buy-down program aimed at decreasing the unit cost by increasing production and sales. Defense is providing $15 million to reduce the purchase price, with preference given to Defense sites. For molten carbonate fuel cells, two manufacturers are currently building and testing demonstration plants. The demonstrations involve scaling up the technology into commercial-size powerplant systems. The projects were co-funded by DOE, EPRI, the Gas Research Institute, and several utility companies. Both manufacturers have experienced significant problems in scaling up their systems to demonstration plant size. Because extensive system modifications are being made, another generation of demonstration plants will likely be needed before the systems are market ready. The availability of funding for a second round of demonstration plants is questionable. Solid oxide fuel cell systems are currently being developed by several companies. DOE is co-funding the development by Westinghouse Electric Corporation of a tubular system, while several smaller companies are working on developing planar systems. While Westinghouse’s tubular system is more fully developed, planar technology may prove simpler and cheaper. The developers, along with EPRI, are seeking funding to scale up the technology into larger systems. Proton exchange membrane fuel cells operate at low temperatures and thus can be turned on and off readily, making them suitable for transportation vehicles. Several auto makers, with DOE co-funding, are investigating whether this technology can meet the cost and performance standards under the Partnership for a New Generation of Vehicles program, an effort to spur the development of more efficient and lower emission vehicles. The stationary fuel cell budget for 1995 was $48.2 million, which according to DOE’s program manager was more than two-thirds of total U.S. fuel cell R&D expenditures. In fiscal year 1996, the program operated at a reduced level under a continuing resolution for most of the year. The program manager expects that the Congress will reduce the budget by about $5 million per year for the next 4 years, which would reduce DOE’s ability to co-fund demonstration projects. Even without these cuts, DOE lacks the funds to support the development of planar solid oxide systems unless it were to drop funding for other on-going projects, according to the program manager. EPRI is trying to get a consortium of utilities to invest in the development of a planar solid oxide system but is finding it difficult. The program manager also said that the lack of DOE or EPRI/utility support will delay efforts by the developer, a small company, to scale up its technology into a marketable system. DOE’s transportation fuel cell budget for fiscal year 1995 was $25 million. The program manager expects the fiscal year 1996 budget to be reduced by 14 percent and expects additional reductions in 1997. For 1996, DOE eliminated programs for the transportation applications of phosphoric acid fuel cells, such as in buses and locomotives. For proton exchange membrane fuel cells, DOE reduced by 50 percent its support for General Motors’ fuel cell vehicle program, which according to DOE’s program manager will delay the program. DOE is maintaining its support for the Ford and Chrysler fuel cell vehicle programs, which are not as far along. The program manager said that further budget cuts in the transportation fuel cell program will jeopardize advanced concept research, such as work at Los Alamos National Laboratory to develop direct methanol oxidation, which could potentially eliminate the need for a heavy on-vehicle fuel reformer. He is also concerned that delays due to budget cutbacks could keep fuel cell vehicle development from meeting the time frames for selection under the Partnership for a New Generation of Vehicles program. EPRI’s fuel cell budget has dropped 67 percent, from $9 million to $3 million, in the past 2 years and is likely to decrease further in the future. EPRI officials are concerned that they will no longer be able to support fuel cell demonstration projects and that fewer funds will be available for exploratory research on advanced fuel cell concepts. In addition, the largest California utilities, which have supported fuel cell development, have discontinued funding for the advanced generation technologies. Coal gasification is an advanced electricity generation technology that converts coal into gaseous fuel and cleans the fuel of pollutants in certain powerplants. Three powerplants to demonstrate the currently available technology are under construction as a part of another DOE program. The program was to consist of the Gasification Product Improvement Facility in West Virginia and the Power System Development Facility in Alabama. The West Virginia facility was designed to do R&D on advanced concepts to increase efficiency and lower costs. The Alabama facility, for which DOE provides 80 percent of the funding, was designed to test high-temperature particulate filters, but it has the potential to do some of the advanced concept R&D work planned for the West Virginia facility. DOE’s budget for coal gasification-related R&D was reduced by 23 percent from $26.7 million in fiscal year 1995 to about $20.6 million in fiscal year 1996. To achieve this cutback, DOE has decided to eliminate the Gasification Product Improvement Facility. DOE, however, is preserving funding for the Power System Development Facility. For fiscal year 1996, DOE’s funding for the project is $12 million. Also contributing funding are EPRI and the Southern Company, the host utility. DOE expects some cost sharing from filter manufacturers and developers and more participation from industry once the facility is in operation. DOE is also funding related research projects at the Morgantown Energy Technology Center and other locations. EPRI’s budget for advanced coal technology has dropped 71 percent, from $8.7 million in 1993 to $2.5 million in 1996, of which $1.6 million is for the Alabama facility. The EPRI program manager said that if funding keeps shrinking, EPRI may not be able to continue funding the facility. Although EPRI has supported the three demonstration projects, EPRI’s continued support is jeopardized because one of the host utilities, Tampa Electric, has dropped out of EPRI and another, Cinergy, has stopped funding EPRI’s advanced fossil business unit. The outlook for industry’s potential contribution, according to EPRI and DOE officials, is mixed. Many of the companies are small and not capable of doing much R&D on their own. However, some of the large oil and gas companies, which have more resources, may become more involved, especially in international markets. The potential benefits of advanced gas turbines, another electricity generating technology, is greater energy efficiency and economy and reduced emissions. The turbines can be fueled by natural gas, oil fuels, coal-derived gas, or biomass gas. Gas turbines for utility applications are typically combined with steam turbines to form a combined-cycle system. The waste heat from the gas turbine is used to generate steam, which is converted into additional electricity. DOE is cost-sharing with industry, developing both large and small advanced gas turbines. DOE is to fund no more than 65 percent of the $700 million, 8-year program to develop advanced turbines by the year 2000; industrial participants will contribute at least 35 percent. DOE’s Fossil Energy Office is responsible for developing the large-scale turbine, while DOE’s Office of Energy Efficiency and Renewable Energy is responsible for the small-scale turbine for distributed, industrial, and co-generation applications. On a four-phase schedule, both programs are now in phase III, which involves full-scale component development and a steep increase in expenditures, according to DOE officials. Two manufacturers are independently developing turbine systems under each program—General Electric and Westinghouse are the developers of the large turbine systems and Allison Engines and Solar Turbines are the developers of the small turbine systems. Meanwhile a collaborative initiative, including several utilities and EPRI, is seeking to develop a mid-sized (about 100 megawatt) advanced turbine using aeroderivative technology developed for jet aircraft, such as the wide-bodied Boeing 777. Organized in 1991 by Pacific Gas & Electric, other California utilities, and the state of California, the collaborative has been managed by EPRI and supported by the Gas Research Institute since 1994 when Pacific Gas & Electric dropped out. Now, more than 50 percent of its funding comes from overseas members, including Canadian, British, French, Danish, Dutch, and Italian power companies. Although DOE was an early participant in the collaborative, DOE officials believe that more of a market, albeit overseas, exists for the big plants and may provide benefits to the United States in terms of exports and job creation. The collaborative is seeking to encourage a manufacturer to develop a mid-sized system by getting potential customers to step forward and ensure a sufficient initial market. The collaborative has requested federal seed money for this marketing activity, which, if successful, would result in a largely private-sector system development effort. The collaborative has received no response from DOE or the White House Office of Science and Technology Policy. If such federal funds are not forthcoming, a program official said the collaborative will try to raise the funds from other sources, such as from utilities or independent power producers. DOE’s budget requests for big and small turbines increased from fiscal year 1995, but approval for smaller increases is expected. DOE’s fiscal year 1995 budget for the big turbine project was $36.6 million. DOE requested $44 million for fiscal year 1996, but $36.7 million was approved. While the projects may be slowed somewhat as a result, they were delayed about 9 months in the solicitation process; therefore, the budget impact will likely occur in fiscal year 1997. According to the DOE program manager, General Electric is cost-sharing at 65 percent and Westinghouse at 40 percent, both above the minimum 35 percent for phase III called for in the program plan. The program manager also said that if future federal funds are not available for the program, the manufacturers will likely forgo the development of machines for the domestic market and participate with international partners in developing machines for overseas markets. DOE’s fiscal year 1995 budget for the small turbine project was $18.8 million. DOE requested $27.5 million for fiscal year 1996, but $22.1 million was received. According to the DOE program manager, Allison Engines and its partners and subcontractors, which include EPRI and Indianapolis Power & Light, are cost-sharing 40 percent, and Solar Turbines and its partners and subcontractors, which include the California Energy Commission and the Gas Research Institute, are cost-sharing 60 percent. The DOE manager does not believe either group can afford a bigger share and that budget reductions are likely to delay by 2 years the completion of the projects. In the past 3 years, EPRI’s budget for advanced gas turbines has been cut 67 percent, from $15 million to $5 million. As a result, EPRI has started no new innovations and is forgoing several areas of research. For example, EPRI is no longer doing any control or balance of plant R&D. According to the program manager, EPRI’s ability to monitor the performance of new technology in utility settings and identify problems has been reduced, and no one is picking up the slack from the cutbacks in EPRI’s program. The Department of Defense and the National Aeronautics and Space Administration are continuing to be involved in turbine R&D. However, according to DOE, EPRI, and industry officials, cutbacks in these R&D programs and the shifting of funds from turbine R&D to other activities will mean that less turbine technology will flow from these programs to the U.S. turbine industry than in the past. DOE’s wind program seeks to assist the wind industry in designing, developing, and testing technologically advanced wind turbines that can compete with conventional electricity generation. Wind energy is a renewable resource that does not use fossil energy supplies; has no air pollutant emissions; and is compatible with other land uses, such as farming and recreation. While good wind resources exist in many areas of the country, over 90 percent of the usable wind resource is in the Great Plains, stretching from Montana, North Dakota, and Minnesota south to Texas. Over 1,700 megawatts of wind power capacity are currently installed in the United States, mostly in California. DOE is providing funds for several wind power projects. Since 1992, DOE has provided funds under its near-term product development and prototype testing program to three companies to develop turbines capable of generating electricity at a cost of 5 cents per kilowatt-hour (kwh). Also, in 1994 DOE began co-funding R&D projects by five companies on subsystems that could be incorporated into advanced turbine systems capable of generating electricity at a cost of 4 cents per kwh from 13-mile-per-hour winds. In addition, DOE is negotiating contracts with two companies to develop advanced turbine systems over the next 3 to 5 years. Scheduled to commence in September 1996, the contracts are expected to total $33.7 million; DOE’s share will be approximately $19.7 million. One developer is covering 50 percent of the cost and the other is covering 30 percent. In 1993, DOE and EPRI began the Utility Wind Turbine Performance Verification Program to promote utilities’ participation in wind power projects and evaluate the latest commercial prototype wind turbines in typical utility operating environments. The program also provides a limited market for newly designed wind turbines prior to their achieving fully commercial status and documents and communicates the project’s experiences and lessons learned to interested U.S. utilities and turbine manufacturers. DOE has provided $2.75 million of the total program cost of $22.4 million. The two utility companies involved in the program (Central & SouthWest in Texas and Green Mountain Power in Vermont) are covering 50 percent and 65 percent of the cost, respectively, and EPRI is contributing the balance. The Texas project has been built and plant performance evaluation is under way. Construction of the Vermont project is scheduled for summer 1996. To further encourage the utilities’ involvement, in 1994 DOE initiated the Wind Energy Deployment Project under which DOE would contribute up to 20 percent of the cost of constructing 25-megawatt wind powerplants. In fiscal year 1995, DOE selected one project in Wyoming and two projects in Iowa under this program. DOE’s wind program budget was reduced 34 percent, from $45.4 million in fiscal year 1995 to $30 million in fiscal year 1996. The previous year’s budget was increased from the fiscal year 1994 level of $28.6 million, primarily to fund the Wind Energy Deployment Project. As a result of the reduction in 1996, DOE canceled further funding of the project. In addition, because of DOE’s, EPRI’s, and utilities’ budget reductions, no further funding in fiscal year 1996 was provided for the Turbine Verification Program, beyond the two projects already under way. DOE, however, is sustaining funding for the advanced turbine contracts, which DOE officials believe are of increased importance because the industry needs outside support to ride out the current domestic utility market stagnation and continue developing new technology. Because the projects require multiyear funding, however, the budget reductions in fiscal year 1997 would require cutbacks, potentially not allowing completion of the turbine development program. According to EPRI’s program manager for wind and solar projects, the wind budget has declined from $2.3 million to $2 million since 1993. The 1996 budget will be used primarily to complete the two turbine verification projects under way. According to DOE and EPRI officials, the domestic market for wind turbines is currently depressed because of utilities’ uncertainty about electric power market restructuring. Additionally, even though the cost of wind power is coming down, the target price for power generation has declined further due to the availability of cheap natural gas and turbines. A further setback occurred in 1995 when the Federal Energy Regulatory Commission nullified a California set-aside plan under which California utilities would have purchased over 1,000 megawatts of additional wind power. According to DOE, the U.S. wind industry is badly lagging in sales and behind in technology development compared to European competitors, who have expanded R&D funding for wind energy since 1985, much faster than any other renewable technology. These countries are currently spending over $150 million annually, according to DOE. Photovoltaic technology uses various devices to convert sunlight directly into electricity without any moving parts. Photovoltaic systems are aimed at providing an alternative to fossil fuel-based electricity generation and its residual environmental impacts. Hundreds of photovoltaic applications are currently cost-effective for off-grid electric power needs, and research is directed at making more applications cost-effective. DOE has several programs to assist the photovoltaic industry. DOE’s Photovoltaic Manufacturing Technology Project, a DOE-industry partnership, is aimed at reducing manufacturing costs and increasing production capacity. DOE also provides U.S. manufacturers with some international marketing assistance. Another research program that DOE funds is in advanced materials and devices, the major focus of which is developing more efficient and durable thin-film photovoltaic technology, which is cheaper to manufacture. Under this program, DOE funds research at national laboratories and universities and co-funds selected industry research projects. DOE also has several programs to encourage utilities to use photovoltaics. DOE provides funds to operate Photovoltaics for Utility Scale Applications, which tests the performance of new systems in a utility setting. Through another program, DOE helps to buy down the cost for utilities purchasing photovoltaic systems. DOE also funds the Design Assistance Center at Sandia National Laboratory, which provides information and technical assistance to utilities and other entities to design photovoltaic projects. DOE’s budget for photovoltaics was reduced by 29 percent, from $84.6 million in fiscal year 1995 to $60.1 million in fiscal year 1996. Specific reductions include (1) 57 percent, from $14 million to $6 million, to buy down the cost for utilities purchasing photovoltaic systems; (2) 57 percent, from $10 million to $4.3 million, to provide information and technical assistance to federal agencies, utilities, and other entities to design photovoltaic projects; (3) 49 percent, from $5 million to $2.6 million, to support the development and testing of new equipment designs and applications; (4) 33 percent, from $3 million to $2 million, to support international marketing; and (5) 23 percent, from $11 million to $8.5 million, to fund the Photovoltaic Manufacturing Technology Project. According to DOE officials, the effect of these reductions on market expansion programs will be magnified because in most cases the cost-shared contributions from utilities and industry will also be reduced. Furthermore, according to the DOE program manager, the result of this reduction is that efforts to improve cost-effectiveness in manufacturing are slowing down and the goals are being extended. EPRI’s solar budget has declined 75 percent, from $5.7 million in 1993 to $1.4 million in 1996. EPRI is continuing to fund some thin-film research and is assisting a manufacturer in the marketing of its photovoltaic concentrator technology that was developed with past EPRI assistance. DOE officials do not believe that industry will pick up the slack from DOE’s reductions, since only a few of the 19 U.S. photovoltaic manufacturers are profitable or close to making a profit. Furthermore, according to an industry spokesperson, DOE’s 23- percent reduction in the Photovoltaic Manufacturing Technology Project will affect the manufacturers’ initiatives. Some of the manufacturers have reinvested their revenues and have been able to attract venture capital to develop new automated processes and equipment for manufacturing photovoltaic modules on the expectation that the government would follow through in assuming some of the technology development risk. Electricity storage technologies store electrical energy for stationary or transportation applications and can absorb or release energy upon demand. Advanced batteries for electric vehicles and superconducting magnetic energy storage are examples of such technologies that could provide economic and environmental benefits. EPRI officials believe that electricity storage will become increasingly important as utilities are deregulated and more entities are involved in electric power. The U.S. Advanced Battery Consortium, which includes DOE, automakers, battery manufacturers, EPRI, and several utility companies, is spearheading efforts in this country to develop advanced batteries whose performance, weight, durability, and cost will enable electric vehicles to compete in the marketplace. Begun in 1991, the consortium has funded the development of several batteries expected to be in production in 2000 and beyond. For stationary applications, DOE and EPRI have agreed to each sponsor the development of a transportable battery storage system that could be moved to utility sites to demonstrate and quantify the extent of reliability improvements, network stability enhancements, and other system benefits. Several companies have bid on the development contract. Planned completion of the project is scheduled for mid-1997. According to DOE, completion of the project is questionable if funds are cut further. Superconducting magnetic energy storage is the only storage technology that stores electricity as electricity. It is about 90 percent efficient, compared to other storage systems that are about 70 percent efficient. It uses a large coil of conductor maintained at a superconducting low temperature. DOE and EPRI funded early research in the 1980s. DOE-funded Los Alamos National Laboratory was involved in the research. Several U.S. companies have also pursued the technology. EPRI has proposed the construction of a pilot plant for a system that would enable greater utilization of existing transmission capacity and forgo the need for the construction of additional transmission lines. The estimated cost of the project is $80 million, but funding for this project has not been found. EPRI’s only active project is a cooperative program with the Navy under which the Navy is seeking to identify potential applications in the military and EPRI is doing the same for utilities and private industry. The Department of Defense has provided some funds for a smaller project designed to show how superconducting magnetic energy storage can meet the specific needs of a utility. The project would provide stored power to Anchorage Municipal Power & Light for a short period if a turbine plant went down until the utility got backup power going from its reserve. Full funding for this project has not yet been secured, and continued Defense funding is unlikely unless a concrete military application is identified. A limited amount of research is ongoing at several national laboratories and universities to develop flywheels and ultracapacitors. The Department of Defense and DOE are funding these efforts. Flywheels, which involve storing energy in a heavy wheel spun very fast, potentially have both stationary and transportation applications. Ultracapacitors also enable the rapid storage and release of large amounts of energy. Ultracapacitors are electrochemical double-layer energy storage devices that use electrodes with a very high surface area per unit volume; they have potential transportation applications. DOE’s budgets for transportation and stationary applications have declined 17 percent, from $34.5 million to $28.6 million. Specifically, DOE’s budget for transportation applications declined 7 percent, from $28.7 million in fiscal year 1995 to $26.6 million in fiscal year 1996. However, some major shifts were made within the program, specifically shifting funding from the advanced battery program to high-power energy storage for vehicles. For example, DOE’s funding for high-power energy storage for hybrid vehicles was increased from $470,000 to $9.6 million. Hybrid vehicles use piston engines, gas turbines, or fuel cells to produce electricity that is stored and used to run the vehicle. Meanwhile, DOE’s funding for the advanced battery consortium was reduced 43 percent, from $26.4 million to $15.1 million. Several utility consortium members have also dropped out or are considering dropping out because of cutbacks in their R&D programs. To accommodate these cuts, the program is narrowing to funding only one mid-term and one advanced battery technology, deemed the most promising for meeting the consortium’s goals. According to the DOE program manager, steady funding at this level will be needed for the next 3 to 4 years to continue developing these technologies. Finally, the program continues to fund nearly $1.9 million in exploratory and applied research, most of which is at the national laboratories and universities. According to the head of the exploratory research program, the program has been cut back significantly because the cooperative R&D agreements have been eliminated. DOE’s stationary energy storage program budget has decreased 66 percent, from $5.8 million in fiscal year 1995 to $2 million in fiscal year 1996, and EPRI’s budget has decreased 67 percent in the past 2 years, from $6 million to $2 million. As a result, neither EPRI nor DOE is funding pilot projects to demonstrate newly developed superconducting magnetic energy storage technology. EPRI and DOE have funded separate stationary battery storage projects, but neither has funds to explore the demonstration and testing of new advanced batteries in stationary utility applications. DOE focuses on benefits to the nation’s utility networks that can result from storage systems employing currently available batteries. Electric Vehicles: Efforts to Complete Advanced Battery Development Will Require More Time and Funding (GAO/RCED-95-234, Aug. 17, 1995). Electricity Supply: Consideration of Environmental Costs in Selecting Fuel Sources (GAO/RCED-95-187, May 19, 1995). Electric Vehicles: Likely Consequences of U.S. and Other Nations’ Programs and Policies (GAO/PEMD-95-7, Dec. 30, 1994). Fossil Fuels: Lessons Learned in DOE’s Clean Coal Technology Program (GAO/RCED-94-174, May 26, 1994). Electricity Supply: Efforts Under Way to Develop Solar and Wind Energy (GAO/RCED-93-118, Apr. 16, 1993). Energy R&D: DOE’s Prioritization and Budgeting Process for Renewable Energy Research (GAO/RCED-92-155, Apr. 29, 1992). The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO provided information on: (1) changes in the amount of funding for electricity-related research and development (R&D); and (2) the impact of these changes on the types of R&D being funded. GAO found that: (1) Department of Energy (DOE) funding for electricity-related R&D increased between fiscal years (FY) 1993 and 1995, but began to decrease by FY 1996 due to budget constraints and congressional appropriation committees' recommendations; (2) the electric utility industry began to reduce R&D funding in 1993, and R&D managers expect the decreases to continue as the industry prepares for deregulation and increased competition; (3) some state programs are also experiencing funding reductions, primarily due to decreases in contributions from utilities, oil overcharge revenues, DOE funding, and state appropriations; (4) proprietary data and industry restructuring make it difficult to assess the current level or project future levels of industry spending on electricity-related R&D; (5) many utilities are shifting their R&D focus from collaborative, long-term projects to proprietary, short-term projects to gain a competitive edge; (6) some R&D projects are being delayed, scaled down, or cancelled as a result of the funding reductions; and (7) DOE, industry, and state government officials who expressed concerns about electricity-related R&D funding suggested such alternative funding sources as a state-administered surcharge on all retail sales of electricity and a national wires charge on all electricity entering the transmission grid.

As we reported in October 2009, insufficient site-specific data, such as local projections of expected changes, make it hard for federal, state, and local officials to predict the impacts of climate change, and thus hard for these officials to justify the current costs of adaptation efforts for potentially less certain future benefits. Based on the responses by a diverse array of federal, state, and local officials knowledgeable about adaptation to a web-based questionnaire designed for that report, related challenges generally fit into two main categories: (1) translating climate data—such as projected temperature and precipitation changes—into information that officials need to make decisions and (2) difficulty in justifying the current costs of adaptation with limited information about future benefits. The process of providing useful information to officials making decisions about adaptation can be summarized by the following: First, data from global-scale models must be “downscaled” to provide climate information at a geographic scale relevant to decision makers. About 74 percent (133 of 179) of the officials who responded to our questionnaire rated “availability of climate information at relevant scale (i.e., downscaled regional and local information)” as very or extremely challenging. Second, the downscaled climate information must be translated into impacts at the local level, such as increased stream flow. Some respondents and officials interviewed for our October 2009 report said that it is challenging to link predicted temperature and precipitation changes to specific impacts. For example, one federal official said that “we often lack fundamental information on how ecological systems/species respond to non-climate change related anthropogenic stresses, let alone how they will respond to climate change.” Third, local impacts must be translated into costs and benefits, since this information is required for many decision making processes. Almost 70 percent (126 of 180) of the respondents to our questionnaire rated “understanding the costs and benefits of adaptation efforts” as very or extremely challenging. As noted by one local government respondent, it is important to understand the costs and benefits of adaptation efforts so they can be evaluated relative to other priorities. Fourth, decision makers need baseline monitoring data to evaluate adaptation actions over time. Nearly 62 percent (113 of 181) of the respondents to our questionnaire rated the “lack of baseline monitoring data to enable evaluation of adaptation actions (i.e., inability to detect change)” as very or extremely challenging. These challenges make it difficult for officials to justify the current costs of adaptation efforts for potentially less certain future benefits. A 2009 report by the National Research Council (NRC) discusses how officials are struggling to make decisions based on future climate scenarios instead of past climate conditions. According to the report, requested by the Environmental Protection Agency and NOAA, usual practices and decision rules (e.g. for building bridges, implementing zoning rules, using private motor vehicles) assume a stationary climate—a continuation of past climate conditions, including similar patterns of variation and the same probabilities of extreme events. According to the NRC report, that assumption, which is fundamental to the ways people and organizations make their choices, is no longer valid. Federal actions to provide and interpret site-specific information would help address challenges associated with adaptation efforts, based on our analysis of responses to the web-based questionnaire and other materials analyzed for our October 2009 report. The report discussed several potential federal actions that federal, state, and local officials identified as useful to inform adaptation decision making. These included state and local climate change impact and vulnerability assessments and the development of processes and tools to access, interpret, and apply climate information. In that report, we also obtained information regarding the creation of a climate service—a federal service to consolidate and deliver climate information to decision makers to inform adaptation efforts. About 61 percent (107 of 176) of the federal, state, and local officials who responded to the web-based questionnaire developed for our October 2009 adaptation report rated the “creation of a federal service to consolidate and deliver climate information to decision makers to inform adaptation efforts” as very or extremely useful. Respondents offered a range of potential strengths and weaknesses for such a service. Several said that a climate service would help consolidate information and provide a single-information resource for local officials, and others said that it would be an improvement over the current ad hoc system. A climate service would avoid duplication and establish an agreed set of climate information with uniform methodologies, benchmarks, and metrics for decision making, according to some officials. According to one federal official, consolidating scientific, modeling, and analytical expertise and capacity could increase efficiency. Similarly, some officials noted that with such consolidation of information, individual agencies, states, and local governments would not have to spend money obtaining climate data for their adaptation efforts. Others said that it would be advantageous to work from one source of information instead of different sources of varying quality. Some officials said that a climate service would demonstrate a federal commitment to adaptation and provide a credible voice and guidance to decision makers. In an announcement on February 8, 2010, the Department of Commerce proposed establishing a NOAA climate service. Though not yet established, information is available on the NOAA climate service website, including draft vision and strategic framework documents. According to NOAA documents, such a climate service would provide a single, reliable, and authoritative source for climate data, information, and decision support services to help individuals, businesses, communities, and governments make smart choices in anticipation of a climate changed future. A September 2010 report by the National Academy of Public Administration discusses the factors needed for a NOAA climate service to succeed—such as the designation of a lead federal agency to be the day-to-day integrator of the overall federal effort regarding climate science and services—and makes recommendations on how to achieve those factors. Other respondents to our questionnaire, however, were less enthusiastic about the creation of a climate service. Some voiced skepticism about whether it was feasible to consolidate climate information, and others said that such a system would be too rigid and may get bogged down in lengthy review processes. Furthermore, certain officials stated that building such capacity may not be the most effective place to focus federal efforts because the information needs of decision makers vary so much by jurisdiction. Several officials noted that climate change is an issue that requires a multidisciplinary response, and a single federal service may not be able to supply all of the necessary expertise. For example, one federal official stated that the information needs of Bureau of Reclamation water managers are quite different from the needs of Bureau of Land Management rangeland managers, which are different from the needs of all other resource management agencies and programs. The official stated that it seems highly unlikely that a single federal service could effectively identify and address the diverse needs of multiple agencies. Several respondents also said that having one preeminent source for climate change information and modeling could stifle contrary ideas and alternative viewpoints. Moreover, several officials who responded to our questionnaire were concerned that a climate service could divert attention and resources from current adaptation efforts by reinventing duplicative processes without making use of existing structures. The 2009 NRC report on informing decisions in a changing climate recommends that the federal government’s adaptation efforts should be undertaken through a new integrated interagency initiative with both service and research elements but that such an initiative should not be centralized in a single agency. Doing so, according to this report, would disrupt existing relationships between agencies and their constituencies and formalize a separation between the emerging science of climate response and fundamental research on climate and the associated biological, social, and economic phenomena. Furthermore, the report states that a climate service located in a single agency and modeled on the weather service would by itself be less than fully effective for meeting the national needs for climate-related decision support. The NRC report also notes that such a climate service would not be user-driven and so would likely fall short in providing needed information, identifying and meeting critical decision support research needs, and adapting adequately to changing information needs. We have not made recommendations regarding the creation of a climate service within NOAA or any other agency or interagency body, although the provision of climate data and services will be an important consideration in future governmentwide strategic planning efforts, particularly in an era of declining budgets. Federal strategic planning efforts could be improved for many aspects of the climate change enterprise. Our October 2009 report on climate change adaptation concluded that, to be effective, related federal efforts must be coordinated and directed toward a common goal. This report recommended the development of a strategic plan to guide the nation’s efforts to adapt to a changing climate, including the identification of mechanisms to increase the capacity of federal, state, and local agencies to incorporate information about current and potential climate change impacts into government decision making. Some actions have subsequently been taken to improve federal adaptation efforts, but our May 2011 report on climate change funding found that federal officials do not have a shared understanding of strategic governmentwide priorities. This report recommended, among other things, the clear establishment of federal strategic climate change priorities, including the roles and responsibilities of the key federal entities, taking into consideration the full range of activities within the federal climate change enterprise. In other reports, we also noted the need for improved coordination of climate- related activities. For example, our April 2010 report on environmental satellites concluded that gaps in satellite coverage, which could occur as soon as 2015, are expected to affect the continuity of important climate and space weather measurements. In that report, we stated that, despite repeated calls for interagency strategies for the long-term provision of environmental data from satellites (both for climate and space weather purposes), our nation still lacks such plans. Of particular importance in adaptation are planning decisions involving physical infrastructure projects, which require large capital investments and which, by virtue of their anticipated lifespan, will have to be resilient to changes in climate for many decades. The long lead time and long life of large infrastructure investments require such decisions to be made well before climate change effects are discernable. Our ongoing work for the Senate Committee on Environment and Public Works Subcommittee on Oversight and Subcommittee on Transportation and Infrastructure will explore this issue by reviewing the extent to which federal, state, and local authorities consider the potential effects of climate change when making infrastructure investment decisions. Chairman Begich, Ranking Member Snowe, and Members of the Subcommittee, this concludes my prepared statement. I would be happy to respond to any questions that you or other Members of the Subcommittee may have. For further information about this testimony, please contact David Trimble at (202) 512-3841 or trimbled@gao.gov. Contact points for our Congressional Relations and Public Affairs offices may be found on the last page of this statement. Barb Patterson, Anne Hobson, Richard Johnson, Ben Shouse, Jeanette Soares, Kiki Theodoropoulos, and Joseph Dean “Joey” Thompson also made key contributions to this statement. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Climate change is a complex, crosscutting issue that poses risks to many existing environmental and economic systems, including agriculture, infrastructure, ecosystems, and human health. A 2009 assessment by the United States Global Change Research Program (USGCRP) found that climate-related changes--such as rising temperature and sea level--will combine with pollution, population growth, urbanization, and other social, economic, and environmental stresses to create larger impacts than from any of these factors alone. According to the National Academies, USGCRP, and others, greenhouse gases already in the atmosphere will continue altering the climate system into the future, regardless of emissions control efforts. Therefore, adaptation--defined as adjustments to natural or human systems in response to actual or expected climate change--is an important part of the response to climate change. This testimony addresses (1) the data challenges that federal, state, and local officials face in their efforts to adapt to a changing climate, (2) the actions federal agencies could take to help address these challenges, and (3) federal climate change strategic planning efforts. The information in this testimony is based on prior work, largely on GAO's recent reports on climate change adaptation (GAO-10-113) and federal climate change funding (GAO-11-317). These reports are based on, among other things, analysis of studies, site visits to areas pursuing adaptation efforts, and responses to a web-based questionnaire sent to federal, state, and local officials. As GAO reported in October 2009, challenges from insufficient site-specific data--such as local projections--make it hard for federal, state, and local officials to predict the impacts of climate change, and thus hard to justify the current costs of adaptation efforts for potentially less certain future benefits. Based on responses from a diverse array of federal, state, and local officials knowledgeable about adaptation, related challenges generally fit into two main categories: (1) translating climate data--such as projected temperature and precipitation changes--into information that officials need to make decisions and (2) the difficulty in justifying the current costs of adaptation with limited information about future benefits. Federal actions to provide and interpret site-specific information would help address data challenges associated with adaptation efforts, based on responses to GAO's web-based questionnaire sent to federal, state, and local officials and other materials analyzed for its October 2009 report. In addition to several potential federal actions identified as useful by respondents to GAO's questionnaire, including the development of state and local climate change vulnerability assessments, GAO's 2009 report also contained information about the creation of a federal climate service. Specifically, about 61 percent (107 of 176) of respondents rated the "creation of a federal service to consolidate and deliver climate information to decision makers to inform adaptation efforts" as very or extremely useful. Respondents offered a range of potential strengths and weaknesses for such a service. For example, several respondents stated that a climate service would help consolidate information and provide a single information resource for local officials. However, some respondents to GAO's questionnaire voiced skepticism about whether it was feasible to consolidate climate information, and others stated that such a service would be too rigid and may get bogged down in lengthy review processes. GAO has not made recommendations regarding the creation of a climate service within the National Oceanic and Atmospheric Administration or any other agency or interagency body. Federal strategic planning efforts could be improved for many aspects of the climate change enterprise. For example, GAO's October 2009 report on climate change adaptation concluded that, to be effective, related federal efforts must be coordinated and directed toward a common goal. This report recommended the development of a strategic plan to guide the nation's efforts to adapt to a changing climate, including the identification of mechanisms to increase the capacity of federal, state, and local agencies to incorporate information about current and potential climate change impacts into government decision making. Some actions have subsequently been taken to improve federal adaptation efforts, but GAO's May 2011 report on climate change funding found that federal officials do not have a shared understanding of strategic governmentwide priorities.

GPRA is intended to shift the focus of government decision making, management, and accountability from activities and processes to the results and outcomes achieved by federal programs. New and valuable information on the plans, goals, and strategies of federal agencies has been provided since federal agencies began implementing GPRA. Under GPRA, annual performance plans are to clearly inform the Congress and the public of (1) the annual performance goals for agencies’ major programs and activities, (2) the measures that will be used to gauge performance, (3) the strategies and resources required to achieve the performance goals, and (4) the procedures that will be used to verify and validate performance information. These annual plans, issued soon after transmittal of the president’s budget, provide a direct linkage between an agency’s longer-term goals and mission and day-to-day activities. Annual performance reports are to subsequently report on the degree to which performance goals were met. The issuance of the agencies’ performance reports, due by March 31, represents a new and potentially more substantive phase in the implementation of GPRA—the opportunity to assess federal agencies’ actual performance for the prior fiscal year and to consider what steps are needed to improve performance, and reduce costs in the future. Treasury is responsible for a broad scope of activities that touch the lives of all Americans, including collecting taxes, managing the government’s finances, securing U.S. borders, controlling firearms-related crime, and managing seized assets. This section discusses our analysis of Treasury’s performance in achieving its selected key outcomes and the strategies the agency has in place, particularly human capital and information technology, for accomplishing these outcomes. In discussing these outcomes, we have also provided information drawn from our prior work on the extent to which the agency provided assurance that the performance information it is reporting is credible. On the basis of information in Treasury’s 2000 performance report, we could not assess its progress in effectively and fairly administering the tax laws because the report lacked information on strategic measures directly related to this outcome. However, the results of our work and other reported information below the strategic level on the performance of Treasury’s agency responsible for relevant programs—the Internal Revenue Service (IRS)—indicated that IRS improved its performance on one key indicator while losing ground on others. As was the case last year, neither Treasury nor IRS had any measures that were specifically linked to the outcome of effective and fair administration of tax laws. IRS is modernizing all aspects of the agency’s operations, such as its organizational structure, business processes, technology, and performance management. As part of its modernization, IRS developed three agencywide strategic goals, which we used to assess this outcome. The goals are top quality service to each taxpayer in every interaction, top quality service to all taxpayers through fair and uniform application of the law, and productivity through a quality work environment. In fiscal year 2000, Treasury reported that the IRS made progress in conceptualizing and identifying the measures it needs for its goals.Currently the IRS has only an employee satisfaction measure, which relates to the third goal. While IRS-wide strategic measures to assess this outcome are not available, our work on measures dealing with collecting revenues, providing taxpayer service, and enforcing tax laws indicated mixed results. On the plus side, during fiscal year 2000, IRS issued refunds without significant problems and taxpayers had an easier time getting through to telephone assistors. On the down side, the quality of service for taxpayers who visited taxpayer assistance centers and trends in enforcement functions continue to be troubling. For example, Treasury’s fiscal year 2000 performance report listed five IRS-wide measures that Treasury designated as key performance indicators related to the outcome. The agency did not meet its target for four of the five indicators including all three measures of quality. Treasury’s explanations for not meeting its targets included reasons such as the decline “…was caused by a failure to meet any one, several, or all of the standards measured in this category, thus resulting in a lower overall composite score.” Figure 1 shows IRS’ performance over time on the five key indicators. In its fiscal year 2002 plan for IRS, Treasury lacked specific strategies to demonstrate how it would achieve its strategic goals and objectives and, thereby, be ensured of effectively and fairly administering the tax laws. Instead, IRS’ plan included individual strategies for meeting each of the 73 measures in the plan. Of those measures, 18 are outcome-oriented and focused on quality, timeliness, customer satisfaction, or employee satisfaction. While 5 of these measures will establish baseline data, the strategies to implement the remaining 13 measures are, in general, continuations of those used in the previous year. For example, the strategy for improving field collection quality in fiscal year 2001 included using data analysis to target areas for improvement, planning training around identified needs, and delivering a redesigned, web-based manual. For fiscal year 2002, the strategy to improve quality in that area included continued data analysis and training and added a strategy to develop a system for embedding responsibilities for quality at the organizational level closest to customers. Treasury did not explain why it was likely to be more successful in fiscal year 2002 by continuing similar strategies for the five measures where the targets were not met. In February 2001, we again recommended that IRS more clearly link its measures to its goals and objectives. In response, the Commissioner of Internal Revenue stated that IRS would make such refinements as it gains more experience with modernization. We are unable to assess progress toward achieving less waste, fraud, and error relating to the Earned Income Tax Credit (EITC) because Treasury did not report on measures for any aspect of IRS’ administration of the program, and IRS also lacked performance measures for the program. We first identified this program as a high-risk area in 1995 and, more recently, noted in our December 2000 report that the overall impact of the compliance initiative remained unclear despite its success in identifying hundreds of millions of dollars in erroneous EITC claims in fiscal year 2000. Yet, no performance measures specific to EITC activities are planned at the agency level. Although IRS already collects some data for an internal quarterly tracking report, Treasury does not plan to use that data to assess and report program performance. Treasury plans to continue emphasizing increased customer service for and compliance activities affecting both taxpayers and preparers in its current strategy for reducing problems in the EITC program. However, Treasury’s performance plan for IRS does not include measures for those areas or others related to this outcome. The 5-year strategy for EITC, instituted in 1998, includes expanding customer service and taxpayer education, reviewing preparers’ compliance, and improving return selection methods for audits, among others. However, the strategy does not address training needs, performance management initiatives, or measurement of the strategy’s effectiveness. Without performance measures and an evaluation strategy, Treasury will not be able to assess progress. Limitations in the performance measures reported by Treasury make it difficult to gauge the progress of IRS in collecting tax debt and the Financial Management Service (FMS) in collecting non-tax debt. However, other available information showed continuing declines in most of IRS’ collection actions to collect delinquent tax debt and roughly stable collections by FMS of non-tax debts. Also, Treasury’s plans for the two agencies provided little information on how their strategies for improvement will increase debt collections. On the basis of information in Treasury’s 2000 performance report, we could not assess Treasury’s progress in improving its collection of delinquent taxes because none of the performance measures were linked to this outcome. However, based on other information about collection programs at the IRS—the Treasury agency responsible for the relevant programs—we are troubled by the performance with respect to this outcome. Treasury’s performance report measures output for this effort in terms of volume of collection cases closed and timeliness of certain types of collection actions. Treasury did not report performance measures that would provide perspective on whether IRS is collecting the correct amount of taxes under proper collection procedures. For example, the performance report did not contain measures for amounts collected as a percentage of the total value of the collection cases that were closed, or at the IRS-wide level, a measure for enforcement revenue collected as a percent of unpaid taxes. Such measures, over time, would give a clearer indication than case closure data as to whether Treasury is making any headway in improving delinquent tax collections. As figure 2 illustrates, enforcement revenue collected has not kept pace with the growth in the levels of unpaid taxes. Furthermore, our recent work showed declines in important collection actions including seizures, liens, and levies. Although the Commissioner of Internal Revenue predicted last year that the downward trends for these actions would be reversed, by and large, they were not. In addition, the reliability and accuracy of these output measures in Treasury’s performance report is questionable. Our audit of IRS’ fiscal year 2000 financial statements found that IRS was unable to provide documentation that it had performed validation and verification procedures on its key performance indicators. Treasury’s strategy in its fiscal year 2002 IRS plan for increasing collections of tax debts focused on reducing the diversion to noncompliance duties of staff who are experienced in compliance efforts and enhancing the efficiency of the delinquent tax account management. This included (1) adopting a risk-based approach for identifying better yielding accounts for collection and examination and (2) making more effective use of technology and specialization for processing unpaid tax transactions through IRS systems. However, although these strategies potentially could improve tax debt collections, the performance plans did not contain performance measures for assessing its progress. In addition, we previously reported that IRS does not have adequate records on its unpaid assessments to properly manage its accounts receivable inventory. Without such information, IRS may not be able to successfully implement these strategies and, therefore, not achieve the desired outcome. While Treasury reported that it had many significant accomplishments in improving non-tax delinquent debt collection in fiscal year 2000, the non- tax debt collections in fiscal year 2000 were about the same as in fiscal year 1999—about $2.6 billion, primarily from tax refund offsets. The performance target for fiscal year 2000 was to collect $2.08 billion. These collections came from primary collection programs administered by FMS—Treasury Offset Program and Cross-servicing. Of the total collected, about $41 million was collected through the cross-servicing program. Treasury also measured non-tax debt collection progress in terms of the amount of debt referred to FMS for collection. For example, as of September 30, 2000, agencies had referred 83 percent of delinquent debts over 180 days old reported as eligible for the Treasury Offset and Cross-servicing programs. Treasury’s fiscal year 2000 target was to refer 75 percent of the amount eligible for referral. While the performance measure for total collections is a good indicator of the overall progress being made in collecting non-tax debt, it does not adequately capture important distinctions between the offset and cross- servicing programs. We suggested in our June 2000 report that combining the performance achievements of the two programs can mask potential performance issues with cross-servicing. By breaking out—in the performance report—the total collections amounts by the results of the offset program and the cross-servicing program would give decision- makers a better indication of the effectiveness of the two programs relative to the resources being applied to each program. Our prior report also suggested that breaking out total non-tax collections by amounts collected as a result of federal delinquent non-tax debt referrals and amounts collected for debts associated with state child support would give decision-makers more information on the types of debt that is being collected. As we previously reported, collections for child support represent a significant percentage of total collections and are forwarded to the states. Reporting such collections separately from amounts related to the collection of federal delinquent non-tax debts would provide a more accurate indication of FMS’ performance. In addition, the amount of delinquent non-tax debt that is referred to Treasury for collection as compared with the amount of delinquent non-tax debt that is eligible for referral is not fully indicative of FMS’ performance. Specifically, since the measure is an indicator of the efforts of other agencies to participate in the program, it might be unduly influenced by factors outside FMS’ full control. Therefore, it could be difficult to attribute changes in the measure to the effectiveness of FMS’ debt collection efforts. Treasury’s fiscal year 2002 strategies in its FMS plan for increasing non-tax debt collections revolved around efforts to improve the efficiency of the cross-servicing program. For example, FMS plans to analyze the types of cross-servicing debts collected, review cross-servicing costs and fee structure, and develop a methodology to periodically evaluate the process of distributing debts to private collection agencies. FMS is also preparing audit guidance on procedures to monitor agency debt referrals. Regarding the offset program, FMS plans to expand the program by including federal salary and other payment types. However, while FMS has strategies to increase non-tax debt collections, it did not discuss a timeframe for incorporating these payments into the program. FMS’ target for total non- tax debt collections is $2.3 billion for fiscal year 2001 and $2.4 billion for fiscal year 2002. Both of these targets are less than the $2.6 billion FMS collected in fiscal years 1999 and 2000. As such, it does not appear that FMS expects its planned actions for these programs to result in increased collections. It was difficult to fully gauge Treasury’s progress in reducing the availability and/or use of illegal drugs because some of its performance measures did not directly measure Treasury's progress toward achieving this outcome. Treasury acknowledges that some of its measures used to track performance may not provide the best performance information and indicated that it is working toward improving performance measures. Given the measures that Treasury did use, it made some progress in reducing the availability and/or use of illegal drugs. For example, the U.S. Customs Service (Customs)—the Treasury agency primarily responsible for programs related to this outcome—exceeded its targets for three of its nine measures of illegal drugs seized. For the targets that were not met, Treasury attributed this shortfall to external factors such as “an expanding cocaine market in Europe where prices and profit margins are higher than in the United States.” However, Treasury’s report did not identify actions to evaluate the effects of external factors on Treasury’s seizure targets and programs although it acknowledged that it would work with various federal, state, local, and international law enforcement agencies on this crosscutting outcome. Measures of illegal drugs seized provided only a partial assessment of Customs’ success in reducing the availability and/or use of illegal drugs. Without an underlying measure of the amount of drugs moving into the country in total, interpretations of measures of drugs seized is problematic. This overall measure, given the clandestine and diffused nature of illegal drug traffic, is illusive even with rigorous measurement efforts. In its performance report, Treasury noted that the Office of National Drug Control Policy is developing models that will better estimate the amount of cocaine, heroin, and marijuana being smuggled into the U.S. In the meantime, Customs plans to rely on its targeting efficiency measures to quantitatively assess the effectiveness of the criteria to target potential violators—a measure applied to both air passengers and vehicles. Until better measures of Customs’ performance are developed, Treasury may want to explore the relative effectiveness of several Customs anti-smuggling programs. Customs relies on intelligence, surveillance, investigations, random inspections of incoming passengers and cargo, technology, and arrangements with exporters, importers, and carriers to increase the likelihood that it will detect drugs being smuggled into the U.S. For example, Customs could compare the difference in drug detection at different border crossings where one site had a new scanning technology and another site did not have the technology. As we noted in our September 2000 report, agencies could use program evaluation for several purposes such as exploring the benefits of programs, measuring program performance, and explaining performance results. While program evaluations will also be hampered by the lack of underlying data about the flow of drugs, they might provide some indications of the comparative effectiveness of different interdiction programs. In addition, we noted in a March 2000 report that Treasury could improve processes related to its performance measures for target efficiency. In response to our recommendations, Customs stated it would collect more complete and accurate data on persons subjected to personal searches as well as closely monitor data on personal searches. While Treasury’s report notes that one of its actions to increase the targeting efficiency and effectiveness is to improve training, Treasury’s performance report for these measures could have included a discussion on either progress to date for collecting this data or results of its data evaluations of passengers targeted for searches in relation to its efforts to seize drugs. Treasury’s fiscal year 2002 plan for Customs provided specific strategies and programs for fiscal year 2002 designed to help Customs reduce the availability and use of illegal drugs. For example, Customs plans to reduce the availability and/or use of illegal drugs by using air and sea interdiction units designed to protect our borders from the continually shifting narcotics and contraband smuggling threat. However, some measures for the strategies and programs that are designed to achieve this outcome provide an incomplete measure of performance. For example, the agency plans to measure the number of landings made by suspect aircraft that occur shortly before the aircraft crosses the border into the U.S., called short landings. The agency could refine its measure by using information about planes that are found after landing in the U.S. to have carried illegal drugs. In addition, Customs did not identify actions that are to be taken to mitigate the effects on its activities of external influences such as changes in drug smuggling routes in response to law enforcement pressures. The fiscal year 2002 performance plan does not discuss any human capital initiatives as strategies to support this outcome nor does it include information on technology initiatives for this outcome. The plan also described coordination efforts underway with another agency and annotated the performance goal to show where such crosscutting coordination occurred. As with drug flow discussed above, the clandestine nature of the underlying activities renders performance measurement in the area problematic. Thus, it is unclear whether Treasury made progress in achieving this outcome because none of the measures for this outcome directly targeted whether criminal access to firearms was reduced. For example, two measures tracked aspects of firearms-tracing activities related to crime guns—the number of firearms trace requests submitted and the average trace response time. According to the Bureau of Alcohol, Tobacco, and Firearms (ATF)—the Treasury agency responsible for programs related to this outcome—the tracing process assists law enforcement agencies in identifying possessors of recovered crime guns, and it enables ATF to develop investigative leads to identify illegal suppliers of firearms. However, neither of these measures provided an assessment of the extent to which tracing activities by ATF helped deny criminals access to firearms or reduced crime. In addition, the measures do not address one issue stated in the performance goal—community exposure to firearms-related crime. Although Treasury reported statistical data that indicated reductions in crimes committed with firearms, these data were not reported as performance measures. As discussed in an earlier section of this report, our September 2000 report noted that agencies could also use program evaluation to identify program benefits, among other uses. Treasury’s fiscal year 2002 plan for ATF provided specific strategies and programs designed to help it reduce criminal access to firearms and related crime. For example, for the Integrated Violence Reduction Strategy (IVRS), the ATF strategy aims to remove violent firearms offenders from the community, deny criminals access to firearms, and prevent violence and firearms crimes through community outreach. However, this strategy is not currently addressed by the ATF performance measures, as none of these measures determine how successful the IVRS will be at denying criminals access to firearms and reducing related crimes. For example, one ATF performance measure tracks the number of firearm trace requests submitted during the fiscal year, a measure that does not provide sufficient information regarding IVRS’ success at helping deny access to firearms. In addition, the fiscal year 2002 performance plan strategies to achieve this outcome do not discuss human capital or information technology initiatives. For the selected key outcomes, this section describes major improvements or remaining weaknesses in Treasury’s (1) fiscal year 2000 performance report in comparison with its fiscal year 1999 report, and (2) fiscal year 2002 performance plan in comparison with its fiscal year 2001 plan. It also discusses the degree to which the agency’s fiscal year 2000 report and fiscal year 2002 plan addresses concerns and recommendations by the Congress, GAO, the Inspectors General and others. Treasury’s fiscal year 2000 performance report contained the same weaknesses that we identified in its fiscal year 1999 report and one additional limitation. However, Treasury made a few improvements. The strategic goals and objectives of IRS, FMS, Customs, and ATF were not always directly reflected in the broader departmental goals, limiting the reports’ usefulness in determining whether these agencies are making progress in meeting their strategic goals in general and these outcomes in particular. When measures were dropped from use, Treasury usually did not explain the reason for the changes. When agencies did not meet their targets, Treasury provided only brief explanations for the shortfalls. In general, for fiscal year 2000, a full understanding of the reasons for the shortfalls in performance was either not provided or was speculative. For example, in explaining why IRS did not meet its target for overall quality of field examination cases, Treasury stated the obvious—that it was unable to meet new quality standards. The reports generally did not discuss crosscutting issues or the impact of external factors on Treasury’s abilities to meet its targets. For example, Customs did not identify actions to mitigate the effects on its activities of external influences, such as changes in drug smuggling routes in response to law enforcement pressures. The reports provided minimal assurance that the performance information and data reported was credible by inserting an overall data accuracy statement at the beginning of the report. Data accuracy is one of several important elements to consider when examining the quality of agency performance data. However, decision-makers may need more detailed explanations about such things as data validity, completeness, consistency, timeliness, and/or access. As we noted earlier in this report, our audit of IRS’ fiscal year 2000 financial statements raised questions about the reliability and accuracy of some of IRS’ performance indicator data. Unlike the fiscal year 1999 report, the fiscal year 2000 report did not fully discuss the findings of any program evaluations performed. Instead, the report provides brief summaries of four evaluations as examples. Treasury made two changes that strengthened its presentation of program performance data in general and also increased consistency with agency strategic goals. First, Treasury elevated its objective of Improve Customer Satisfaction to a strategic goal. Second, it added a strategic goal of Improve Employee Satisfaction. Treasury’s fiscal year 2002 performance plans for its agencies contained the same weaknesses that we identified in its fiscal year 2001 report. First, the performance goals and measures of Treasury’s agencies will still not provide much results-oriented information related to broader departmental goals. Second, the performance plan sections on IRS and FMS provided minimal information on each measure’s data source, accuracy, and limitations, among other things. The sections on Customs and ATF presented more data-related information than was presented in the IRS section. However, while Customs and ATF discussed the source and accuracy of the data, they did not present complete assessments. For example, the Treasury performance plan did not discuss data collection and storage, data validation and verification, or data limitations. Under the Reports Consolidation Act of 2000 (P.L. 106-531) agencies are to assess the completeness and reliability of the performance data included in their reports. The assessments are to describe any material inadequacies in the completeness and reliability of the performance data, and the actions the agency can take and plans to take to resolve such inadequacies. Third, the sections generally did not describe program evaluations currently underway and how they would be used to assess agency performance. Only the outcome on reducing the availability and use of illegal drugs was affected by changes in Treasury’s performance plans for its agencies. These measures attempt to better link performance with this outcome but they may not provide complete information. For example, one of these measures tracks the number of suspect aircraft that land short of the U.S. border, an indicator that Customs’ interdiction effort is successful. While the number of short landings is an indicator of successful reduction in the availability of illegal drugs, Customs should also consider measuring the number of suspect aircraft that actually make it past the border to deliver their illegal drug cargo. GAO has identified two governmentwide high-risk areas: human capital and information security. Regarding human capital, we found that Treasury’s performance plan had one measure related to human capital which was to measure the extent to which Treasury has implemented a new human resources system. With respect to information security, we found that Treasury’s performance plan had one related goal and measure that captured the percent of all Treasury information technology systems that are certified and accredited to operate. In addition, Treasury addressed this management challenge at the individual agency level, as discussed in app. I. In addition, GAO has identified five major management challenges facing the Department of the Treasury. We found that Treasury’s performance report discussed the agency’s progress in resolving its challenges. Of the agency’s seven major management challenges, identified by GAO, its performance plan had (1) goals and measures that were directly related to one of the challenges, (2) had goals and measures that were indirectly applicable to five of the challenges (3) had no goals and measures related to one of the challenges, but discussed strategies to address it. As agreed, our evaluation was generally based on the requirements of GPRA, the Reports Consolidation Act of 2000, guidance to agencies from the Office of Management and Budget (OMB) for developing performance plans and reports (OMB Circular A-11, Part 2), previous reports and evaluations by us and others, our knowledge of Treasury’s operations and programs, GAO identification of best practices concerning performance planning and reporting, and our observations on Treasury’s other GPRA- related efforts. We also discussed our review with agency officials in IRS, FMS, Customs, ATF, and the Treasury’s Office of Inspector General. The agency outcomes that were used as the basis for our review were identified by the Ranking Minority Member of the Senate Governmental Affairs Committee as important mission areas for the agency and do not reflect the outcomes for all of Treasury’s programs or activities. We also used information from our January 2001 report on major management challenges and program risks for Treasury, and similar reports by Treasury’s Inspector General and the Treasury Inspector General for Tax Administration to identify challenges related to the outcomes we reviewed. We did not independently verify the information contained in the performance report and plan, although we did draw from other GAO work in assessing the validity, reliability, and timeliness of Treasury’s performance data. We conducted our review from April 2001 through June 2001 in accordance with generally accepted government auditing standards. We discussed this report with Treasury officials on June 5, 2001, and received written comments on it. The full text of Treasury's written comments is in appendix II. Treasury noted that it appreciated our reviews and insight on how it can make its GPRA products more useful. Treasury divided its comments into two categories, general and specific. Regarding the general comments, Treasury expressed agreement that it needs to use program evaluation to determine the impact of its programs on outcomes. It also expects to improve the link between agency measures and Treasury goals and objectives by using information from a review of its measures that it plans to conduct as part of the fiscal year 2003 budget. While we are not recommending specific measures to Treasury at this time, we are available to work with Treasury on performance measurement issues. In addition, Treasury plans to take steps to ensure data validity for each of its performance measures by reviewing its control processes. We agree that such a step is useful. As part of that review, Treasury may also want to identify and implement changes to ensure that its performance measurement data are relevant, timely, and accurate. Treasury also noted that it continues to face conflicting pressures to keep its GPRA products streamlined and yet to include more detailed information; Treasury characterized our position as one desiring considerably more detail in their GPRA products. We agree that it is difficult to strike a balance in order to provide information that is useful and easily understood yet is sufficiently inclusive. However, we believe that it is only after Treasury establishes the information basic to a GPRA orientation such as its (1) planned outcomes, (2) actions to accomplish the outcomes, and (3) measures that are meaningful and reliable indicators of progress that presentation issues should be addressed. Producing documents that are responsive to GPRA requirements is an additional consideration. Treasury made a number of specific comments that provide additional perspective on issues that we discuss in the report. We note below those instances where Treasury disagreed with a point we made and where we incorporated technical clarifications as appropriate. Administration of Tax Law. Treasury agreed that it needs to further develop measures for this outcome. We recognize that development of such measures is difficult. EITC. Treasury agreed that it lacks performance measures for the EITC program and describes other measures that it uses in managing the program. However, as we note in the report, without performance measures and an evaluation strategy, Treasury will not be able to assess progress in achieving less waste, fraud, and error in the program. Delinquent Tax Collection. Treasury commented that IRS' reorganization, new mission, and strategic goals caused IRS to find methods of measuring success without considering dollars collected. Treasury went on to note that for various reasons, IRS' traditional debt collection activities have declined and that with additional staffing it has been authorized, IRS believes it will realize significant improvement in critical debt collection areas by the end of fiscal year 2002. While IRS' current performance measures provide some perspective on its collections success, we believe, as explained in our report, that some measures that consider whether IRS is collecting the correct amount of taxes under proper collection procedures would provide a more balanced performance perspective. Non-tax Debt Collection. Treasury disagreed with our assessment that one of its measures--involving a comparison of delinquent non-tax debts referred and eligible for referral--for this outcome is not a good gauge of success. While we agree that FMS has spent considerable resources working with other agencies so that the agencies will refer their debt, it is also true that the measure is unduly influenced by factors outside FMS' full control. We modified the wording in our report to better reflect our concern about that measure. Customs Drug Interdiction Program. Treasury disagreed with our statement that some of its performance measures were not linked directly to its goals. In response, we revised our wording to better articulate our assessment of the progress Treasury has made in achieving its outcomes. Personal Search Data. Treasury provided additional perspective on its efforts in response to our previous recommendation related to its performance measures for target efficiency. Short Landings. Treasury disagreed with our assessment that some of its drug interdiction performance measures do not support the outcome of reduced drug availability. It also took issue with our use of an example regarding the performance measure on the number of suspect aircraft that land shortly before crossing the border into the U.S. The intent of the example was to illustrate the need for better outcome measures, a need that Treasury acknowledged in its comments. We clarified our wording in the report. Automated Commercial Environment (ACE). We deleted reference to the development of the ACE initiative, as suggested by Treasury. Data Accuracy, Incomplete Assessment. At Treasury's request, we revised the report to include an example of components of an assessment of data issues. FMS' Computer Security. Treasury provided information in its comments about the performance measures it established for its Information Technology Security Program in the Self- Assessment Framework. While that action is commendable, it does not substitute for performance measures in Treasury's performance report or plan. We clarified this point in our report. FMS Non-compliance with the Federal Financial Management Improvement Act (FFMIA). Treasury acknowledged that its financial management systems did not comply with FFMIA and that it is addressing deficiencies through a remediation plan and corrective actions. While those actions are commendable, they do not substitute for performance measures in Treasury's performance report or plan. As arranged with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after the date of this letter. At that time, we will send copies to appropriate congressional committees; the Honorable Paul H. O’Neill; and the Honorable Mitchell E. Daniels, Jr., Director, Office of Management and Budget. Copies will also be made available at www.gao.gov. If you or your staff have any questions, please call me at (202) 512-9110. Key contributors to this report were Ralph T. Block, Kerry Gail Dunn, and Elwood D. White. Additional staff acknowledgments are listed in app. III. The following table identifies the major management challenges confronting the Department of the Treasury, which includes the government-wide high-risk areas of human capital and information security. The first column of the table lists the management challenges that we and/or the Department of the Treasury’s Office of Inspector General (OIG) or Treasury Inspector General for Tax Administration (TIGTA) have identified. Treasury has two offices of Inspector General— TIGTA, which covers the Internal Revenue Service and the OIG, which covers all other Treasury bureaus. The second column discusses what progress, as discussed in its fiscal year 2000 performance report, Treasury made in resolving its challenges. The third column discusses the extent to which Treasury’s fiscal year 2002 performance plan includes performance goals and measures to address the challenges that we and the two Treasury IGs identified. We found that Treasury’s performance report discussed the agency’s progress in resolving many of its challenges, but it did not discuss the agency’s progress in resolving the following challenge: Strategic Human Capital Management. Of the agency’s 16 major management challenges, its performance plan had goals and measures that were directly related to five of the challenges, goals and measures that were indirectly applicable to five of the challenges and no goals and measures related to six of the challenges, but discussed strategies to address them. In addition to those named in the report, Mark T. Bird, Charles R. Fox, Meafelia P. Gusukuma, Gary N. Mountjoy, Paula M. Rascona, Gregory C. Wilshusen, and Ellen T. Wolfe made contributions to this report.

This report reviews the Department of the Treasury's fiscal year 2000 performance report and fiscal year 2002 performance report plan required by the Government Performance and Results Act. Specifically, GAO discusses Treasury's progress in addressing several key outcomes that are important to Treasury's mission. In general, GAO could not adequately determine Treasury's progress on five key outcomes because the fiscal year 2000 performance report lacked at least some measures needed to directly assess each of the outcomes. However, other information that GAO reviewed and GAO's past work suggest that Treasury may be at risk of not achieving these outcomes. In assessing Treasury's strategies, GAO identified shortcomings in its plans for each of the outcomes it reviewed. Chief among the limitations common to both the Treasury's fiscal year 1999 and 2000 performance reports was that the performance goals and measures of Treasury's agencies were not always directly reflected in the broader departmental goals, limiting the reports' usefulness in determining whether these agencies are making progress in meeting their strategic goals in general and the outcomes GAO reviewed in particular. Treasury improved the fiscal year 2000 report by elevating its objective of Improve Customer Satisfaction to a strategic goal and adding a strategic goal of Improve Employee Satisfaction. Treasury's performance report discussed the progress made in resolving many of its major management challenges, but it did not specifically discuss the agency's progress in resolving challenges related to strategic human capital management.

FTC is an independent agency headed by five Commissioners appointed by the President and confirmed by the Senate, each serving staggered 7-year terms. FTC’s mission is, in part, to prevent business practices that are anticompetitive, deceptive, or unfair to consumers. It acts to prevent business practices that restrain competition and attempts to ensure that the marketplace continues to provide a full range of product and service options among which consumers can choose. The Bureau of Competition is FTC’s antitrust arm. FTC’s Bureau of Competition and Bureau of Economics have the responsibility for merger review. The Bureau of Economics helps to ensure that FTC considers the economic impact of its actions. To achieve this, the Bureau of Economics provides economic analysis and support to the Bureau of Competition and the Commission in carrying out FTC’s antitrust responsibilities. FTC generally shares responsibility for enforcing the federal antitrust laws with the Department of Justice’s (DOJ) Antitrust Division. The federal antitrust statutes are the Sherman Act, as amended (15 U.S.C. 1-7); the Clayton Act, as amended (15 U.S.C. 12-27), which includes the Robinson- Patman Act; and the FTC Act, as amended (15 U.S.C. 41 et seq.). The acts’ objectives are to prevent anticompetitive behavior and preserve and promote competition in the marketplace. FTC is solely responsible for enforcing the FTC Act, while DOJ is solely responsible for enforcing the Sherman Act. Both FTC and the Antitrust Division are responsible for enforcing Section 7 of the Clayton Act, which prohibits mergers and acquisitions that may substantially lessen competition or tend to create a monopoly in any market. FTC and the Antitrust Division have clearance procedures to determine which agency will investigate a potential antitrust violation. The agencies decide which one will conduct a particular investigation primarily by examining current agency expertise in the industries at issue. The HSR Act, which added Section 7A to the Clayton Act, requires certain parties to provide premerger notification of proposed acquisitions and mergers prior to consummation to assist FTC and the Antitrust Division in investigating whether a proposed acquisition would violate Section 7 of the Clayton Act, in that the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly. The premerger notification provisions of the HSR Act require companies exceeding certain thresholds of company size and value of the transaction to notify FTC and the Antitrust Division of the proposed merger transaction, submit documents and other information to the agencies concerning the transaction, and refrain from closing the transaction until a specific waiting period has expired, or their request for early termination of the waiting period has been granted. FTC and the Antitrust Division then have up to 30 days (15 days for cash tender offers and bankruptcy sales) from the time of the filing of the proposed merger to review the filing and determine whether to send the parties a request for additional information (a second request). A second request extends the waiting period to enable further review. If FTC or the Antitrust Division sends a second request to the parties, the parties have to wait 30 days (10 days for cash tender offers and bankruptcy sales) from the date the parties substantially complied with the request before consummating the acquisition to allow the agencies time to complete the investigation and to determine whether to take law enforcement action. If the reviewing agency does not send a second request, or the parties have substantially complied with a second request, and the waiting period has expired, or if the parties’ request for early termination of the waiting period has been granted, the parties can consummate the merger or acquisition. For merger investigations conducted by FTC, prior to the expiration of the waiting period, FTC staff are to seek to complete the analysis of likely competitive effects of the transaction and prepare recommendations to the FTC Commissioners on whether enforcement action is warranted. If FTC staff determine that a merger is likely to be anticompetitive, FTC has wide discretion in choosing an effective remedy to provide relief that addresses the competitive problems it identified and maintains or restores competition, without unnecessarily limiting the parties’ lawful objectives, such as achieving efficiencies. To meet this challenge, the agency can choose any number of possible actions, including blocking the entire transaction; requiring full or partial divestiture, either broadly or in specific geographic markets; requiring contractual arrangements; or requiring some form of behavioral relief, such as establishing firewalls in vertical transactions to prevent the sharing of competitive information. In cases where FTC staff and the merging parties have negotiated a remedy and agreed upon a proposed settlement, staff recommend that the Commission accept the proposed order and place it on the public record to enable the public to comment. The public is given the opportunity to provide comments for the record about the proposed consent order. FTC staff are to analyze the public comments received and may recommend appropriate changes to the final complaint and order issued by FTC. FTC also may terminate the HSR waiting period and allow the merging parties to begin to consummate the transaction, including any divestitures, if the buyer is identified in the proposed order. However, for consent orders in which FTC has required a divestiture(s), and the buyer(s) of the divested asset(s) is not identified in the proposed order, the merging parties must subsequently submit an application to FTC requesting approval to divest the assets to a proposed buyer(s) and await FTC approval before consummating the divestiture. These applications also must be placed on the record for public comment generally for 30 days. The staff analyze the comments received and make recommendations to the Commission on whether it should approve the divestiture application. (Appendix IV provides additional information on the merger review process.) The number of mergers reported annually to FTC and the Antitrust Division pursuant to the HSR Act more than doubled from 2,262 transactions in fiscal year 1990 to 4,926 transactions in fiscal year 2000. During this same period, FTC announced 192 consent orders involving mergers, of which 153 (about 80 percent) were divestiture orders. (Appendix V provides information on the 153 divestiture orders FTC announced for public comment during fiscal years 1990 through 2000). According to FTC staff, FTC decisions to use particular divestiture approaches are (1) based on the unique facts of each case and do not readily translate into written guidelines or systematic aggregation and (2) tied to proprietary company information that FTC is statutorily prohibited from disclosing to the public. Accordingly, FTC does not systematically compile and make publicly available data that show under what circumstances clean sweep divestitures, single buyers, and up-front buyers should be or have been used. However, available information from FTC staff, speeches, and other public documents reveal that during the mid-1990s, based on lessons learned from past divestitures, FTC began to develop preferences for divestiture approaches designed to restore competition more quickly and reduce the likelihood that assets would deteriorate while awaiting final action on the proposed merger. These preferences may include clean sweeps, single buyers and up-front buyers. However, according to FTC staff, it depends on the industry whether they are appropriate. The history of FTC's clean sweep, single buyer, and up-front buyer divestiture approaches is difficult to chronicle because FTC does not have readily available public data that show under what circumstances they should be or have been used. Our review of divestiture orders and related public documents, such as FTC’s analysis to aid public comment, revealed that FTC provides limited information on the rationale for the use of particular divestiture approaches. FTC staff told us that the staff who worked on the order can readily provide information on the basis for using the approaches based on their knowledge of the case and their personal case files. FTC staff acknowledged that public documents typically do not provide detailed, meaningful information on why a particular approach was used. However, they said that staff document information on the rationale for the approaches and provisions used in a divestiture order in internal, confidential documents, such as staff memorandums to the Bureau Directors and Commissioners. They also told us that because FTC’s decisions are largely tied to companies’ trade secret information, which FTC is statutorily prevented from disclosing to the public, FTC can provide to the public only limited information on the basis for its decisions. “Our approach to remedies evolves, as does our approach to merger enforcement generally. We learn from each case what works and what doesn't work. Our past actions provide guidance, but there are no absolute rules. We evaluate remedies based on the facts in each individual case. We also go back and evaluate our remedy process…to see if expectations are borne out and the remedies are effective.” “One aspect of our merger relief that might bear scrutiny is the FTC’s insistence in most of its divestiture orders that the merging firm locate an up-front buyer for the divested assets. I believe that this requirement has been warranted in a number of the orders that I’ve reviewed in more than three years at FTC, but I would be concerned if the agency became too rigid and unflinching in its insistence on this element of an order.” Although it is not always clear under what circumstances FTC has used one or a combination of divestiture approaches, recent speeches and articles and our discussions with FTC staff do provide some insight into the evolution of these approaches during the mid 1990s. Specifically, it appears that in 1996, based on lessons learned from past divestitures, FTC developed preferences for how to structure divestitures and began to modify its divestiture approaches. According to an April 1997 speech by the then Senior Deputy Director of the Bureau of Competition, FTC had not been satisfied with the effectiveness of past divestiture orders. He said that FTC began a retrospective study of nine divestiture orders selected to assess the effectiveness of particular types of divestiture orders, which showed the need for changes to the way FTC approached merger remedies. He further said that FTC staff had found that the divestiture process was less effective than they would have hoped and “came to appreciate better the difficulties in creating a viable divestiture package of assets that had not previously been a stand-alone business”—a business that contains all the assets needed to enable a buyer to be operational the day after purchasing the assets, selling to all the same customers. By 1996, according to an October 1996 speech by the then Director of the Bureau of Competition, FTC had begun to take steps to shorten the time that it took to complete a divestiture because of FTC’s desire to fashion remedies that (1) restored competition more quickly and (2) reduced the likelihood that assets would deteriorate while awaiting final action on the proposed merger. These steps included, among others, the identification of up-front buyers and the requirement for the divestiture of "broader asset packages" where assets are grouped into one divestiture package to ensure continued marketability, viability, and competitiveness following the merger. In an April 1997 speech, the then Senior Deputy Director of the Bureau of Competition noted that because of these and other steps, the average time between the date a divestiture order is provisionally approved and the ordered divestiture is approved dropped from 15 months in fiscal year 1995 to 7 months in fiscal year 1996. FTC staff acknowledged that they have developed a preference for certain practices in negotiating some divestiture orders, particularly among retail industries, like grocery stores and gas stations. Regarding clean sweeps, single buyers, and up-front buyers, FTC staff said that all three approaches tend to provide them greater assurance that the divestitures will help restore or maintain competition at levels existing before the merger. “Although the Division has not adopted this as a policy, we do consider clean sweep as an option when we look at a divestiture package. The obvious advantage to requiring a clean sweep is that the sale of an ongoing business, as opposed to various stand-alone assets pieced together, may provide greater assurance that the assets will be viable in the hands of a suitable purchaser. Such a policy also prevents the parties from choosing the least attractive assets from each company for divestiture…The potential disadvantage for requiring a clean sweep is that it prevents the parties from realizing possible efficiencies by integrating the different assets of both companies.” Regarding up-front buyers, which FTC also refers to as buyers up-front, FTC staff said that they have made greater use of up-front buyers because the approach reduces the time for the divestiture(s) to take place and gives them the opportunity to evaluate the marketability of the assets to be divested with more concrete evidence. According to FTC staff, the up-front buyer approach also enables staff to better determine whether, among other things, there is a viable buyer(s) for the proposed divestiture assets. Additionally, as pointed out in various speeches by FTC staff, the up-front buyer approach reduces the amount of time needed for the assets to be divested because a buyer can be identified before the merger transaction occurs—a factor in grocery store divestitures and those of other retail operations, where assets may quickly deteriorate during the search for a buyer. FTC staff also said that whether an up-front buyer will be required is dependent upon the circumstances of each individual case and generally not on the industry in which the merging parties operate. During fiscal years 1990 through 2000, clean sweep divestitures, single buyers, and/or up-front buyers were used in the 31 divestiture orders in the four industries we reviewed—grocery stores, drug stores, funeral services, and gas stations—although up-front buyers were not used at all prior to fiscal year 1996. The three approaches were used most frequently in the 16 grocery store divestiture orders, particularly in 1996 or later. During the 11-year period, there was a wide variation in the use of these approaches in the 15 divestiture orders covering the other three industries. According to FTC staff, the remedies approved in each of the 31 divestitures were based on the unique facts of the case, with the goal of remedying an otherwise unlawful acquisition, and only after FTC determined the remedy would achieve that objective. FTC used clean sweeps, single buyers, and/or up-front buyers to remedy anticompetitive concerns for 16 grocery store mergers throughout fiscal years 1990 through 2000, but used these approaches with increased frequency during the latter part of the period. We analyzed the use of clean sweeps, single buyers, and up-front buyers in grocery store divestitures over the 11-year period. Because FTC speeches about its merger remedies indicated that FTC began to make changes in its divestiture approaches during fiscal year 1996, we focused on two periods—the period from fiscal years 1990 through 1995, before FTC started to make these changes, and fiscal years 1996 through 2000, the period during which FTC staff announced that FTC had begun to make these changes. Clean sweep divestitures were used about three-fourths of the time in the latter period, compared with one-third of the time in the earlier period. We examined each of the 16 grocery store divestiture orders to determine how many geographic markets were covered under the order and determined that, in total, FTC had delineated 131 geographic and product markets. Because FTC's divestiture orders and related documents, such as the analysis to aid public comment, typically do not indicate whether a divestiture is a clean sweep, we asked FTC to designate which of the 131 markets involved clean sweep divestitures. FTC provided data on 128 markets that showed that the number of clean sweep divestitures involving grocery store mergers had become much more prevalent during fiscal years 1996 through 2000, the period in which FTC staff said that they had initiated reforms. Whereas about 33 percent (3 of 9) of the markets defined in divestiture orders announced during fiscal years 1990 through 1995 involved clean sweep divestitures, about 75 percent (89 of 119) of the markets involved clean sweep divestitures during fiscal years 1996 through 2000. Table 1 shows the extent to which clean sweeps were used for the 6 divestiture orders announced during fiscal years 1990 through 1995, the 10 divestiture orders announced during fiscal years 1996 through 2000, and for the entire period. FTC staff said that the increased use of clean sweeps occurred because FTC’s past experience with grocery store mergers and divestitures taught staff that individual stores might not be as competitive as chains and packages containing all of the assets of one of the merging parties are easier to sell. Additionally, they said that this type of divestiture gives FTC staff greater confidence that they are preserving competition in the affected geographic market. For the 16 grocery store divestiture orders, the use of single buyers more than tripled between the periods. As shown in table 2, single buyers were used in 20 percent (1 of 5) of the geographic markets with multiple divested assets during fiscal years 1990 through 1995, but in 72 percent (36 of 50) of the geographic markets in fiscal years 1996 through 2000. According to FTC staff, their preference for single buyers started when the nature of grocery store mergers began to change. They said that recent mergers include mergers of chains that are direct competitors in the same geographic market, whereas in past years, merging parties purchased certain stores from one another, not the entire chain. They added that because of chain-wide distribution efficiencies of divested stores, a single buyer can operate divested assets more efficiently and easily. FTC staff told us that if the stores of the merging parties were competing directly throughout the geographic market, a single buyer who can operate all the assets will give FTC greater confidence that competition will be replicated in the market. Up-front buyers were not included in any of the six grocery store divestiture orders issued during fiscal years 1990 through 1995. However, about 76 percent of the buyers for divestiture orders announced during fiscal years 1996 through 2000 were up-front buyers. Table 3 shows the extent to which up-front buyers were used in grocery store divestiture orders announced during fiscal years 1990 through 1995, fiscal years 1996 through 2000, and for the entire period. FTC staff told us that as FTC became more familiar with grocery store mergers and the potential effect of long divestiture periods on the viability of the grocery store assets to be divested, it became apparent that up-front buyers were a material factor in making divestitures successful. They said that, in the early to mid-1990s, merging parties were given up to 12 months, and sometimes longer, to find a buyer(s) for the divested assets after a divestiture order became final, regardless of the industry in which the merger was taking place. Additionally, in a March 2002 FTC document titled, “Frequently Asked Questions About Merger Consent Order Provisions,” FTC staff noted that “...supermarkets and other retail operations (e.g., retail pharmacies) are particularly vulnerable to having their assets deteriorate during the search for a post order buyer; this affects the ability of the assets to be operated in a manner that maintains or restores competition in the relevant market.” FTC staff told us that once FTC determined with respect to grocery store mergers that, among other things, the longer the merging parties have control of the assets, the more likely it is that the assets will deteriorate, FTC established a preference for up-front buyers, particularly in the retail sector. In our discussions with FTC staff about their use of clean sweeps, single buyers, and up-front buyers, they said that grocery store divestitures that were accepted by FTC 10 years ago had long divestiture periods and no up-front buyers, but these divestitures would not likely be accepted today. FTC staff told us that, in recent years, generally FTC’s starting point for settlement discussions with the merging parties in a grocery store merger is a clean sweep divestiture with a single, up-front buyer. According to FTC staff, such a remedy ensures restoration of the status quo before the merger. Additionally, FTC staff told us that it would take additional time and considerable amount of evidence to convince FTC that competition could be restored through selling a combination of the merging parties’ assets to multiple buyers in a single geographic market. In such cases, FTC staff said they would have to analyze information on each individual store and its role in the merging parties’ overall operations. FTC staff also noted that the grocery store mergers in more recent years tended to be much larger and involve more extensive geographic overlaps than previous mergers. While FTC staff have said that FTC typically prefers up-front buyers in divestiture orders involving grocery stores and sometimes in other retail operations, they told us that they are willing to diverge from their preference for clean sweep divestitures and single buyers when the proposed divestiture will restore or maintain competition. A key example is the Albertson’s, Inc., and American Stores Company divestiture order, which at the time it was announced for public comment in fiscal year 1999 was the largest retail divestiture ever required by FTC. The order differed from many recent divestiture orders in the grocery store industry in that many of the markets involved the divestiture of a combination of selected assets from both of the merging parties—a practice sometimes referred to as “mix and match”—versus clean sweeps, and there were several multiple buyer markets, that is markets in which several buyers were purchasing the assets FTC ordered to be divested. According to FTC staff, mix-and-match divestitures require a more careful analysis of each retail location than does the divestiture of only one of the merging parties stores because staff must examine more closely whether the mixed assets can compete effectively. FTC must determine, for example, whether the mixed assets will be capable of producing efficiencies and economies of scale and scope comparable to those existing in the market before the merger. FTC staff estimated that the Albertson’s and American divestiture order took at least 6 additional months to negotiate with the merging parties because of the need to assess the mix and match approach and the geographic markets where assets were divested to multiple buyers as well as to the buyers purchasing assets in multiple geographic markets. There was a wide variation in the use of clean sweeps, single buyers, and/or up-front buyers in the 15 divestiture orders FTC announced in the drug store, funeral services, and gas station industries during fiscal years 1990 through 2000. For the five drug store divestiture orders FTC announced during fiscal years 1990 through 2000, all of the divestitures were to single buyers, but the use of clean sweep divestitures and up-front buyers varied. Three of these orders involved 11 geographic markets between fiscal years 1990 through 1995 and two involved 7 geographic markets between fiscal years 1996 through 2000. In terms of clean sweep divestitures, single buyers, and up-front buyers: All of the 11 geographic markets in the three drug store divestiture orders announced during fiscal years 1990 through 1995 involved clean sweeps of drug store assets within those markets. By contrast, 4 of the 7 (57 percent) geographic markets in the two orders announced between fiscal years 1996 through 2000 involved clean sweeps of drug store assets. All five drug store divestiture orders involved the divestiture of a total of 294 assets from fiscal years 1990 through 2000, and all of the divestitures were to single buyers. For the two divestiture orders announced during fiscal years 1996 through 2000, FTC explicitly required single buyers in 4 of the 7 geographic markets to ensure that buyers were large enough and had the coverage to serve as an alternative anchor pharmacy chain for a pharmacy benefit management firm’s retail pharmacy network. While there were no up-front buyers designated among the 10 buyers of drug store assets in the three divestiture orders announced during the earlier period, 2 of the 3 buyers of drug store assets for the two divestiture orders announced from fiscal years 1996 through 2000 were up-front buyers. Our analysis of funeral services divestitures showed that the use of clean sweeps, single buyers, and up-front buyers remained relatively unchanged for the seven funeral services divestiture orders FTC announced during fiscal years 1990 through 2000. FTC announced four divestiture orders requiring the divestiture of funeral services assets in 8 geographic markets between fiscal years 1990 through 1995, and three requiring the divestiture of funeral services assets in 19 geographic markets between fiscal years 1996 through 2000. Specifically: The divestitures in 6 of the 8 (75 percent) geographic markets designated in the four funeral services divestiture orders announced during fiscal years 1990 through 1995 were clean sweep divestitures and 14 of the 19 (about 74 percent) geographic markets for the three funeral services divestiture orders announced during fiscal years 1996 through 2000 were clean sweep divestitures. In all cases, a single buyer purchased the assets in each of the multiple asset markets. Uniquely, in six of these seven divestiture orders, a single buyer (although not the same buyer in all six cases) purchased all of the divested assets in all of the geographic markets. For example, in the Service Corporation International and Equity divestiture order announced in fiscal year 1999, there were 14 geographic markets in which FTC ordered divestitures; a single buyer purchased all of the assets in all of the geographic markets. Over the 11-year period, only one of the eight buyers of funeral services assets was an up-front buyer—this occurred in a divestiture order announced in fiscal year 1999. For gas station divestiture orders, we could not analyze differences in the use of clean sweeps, single buyers, and up-front buyers between the two periods—fiscal years 1990 through 1995 and fiscal years 1996 through 2000—because there were no divestiture orders requiring the divestiture of gas stations announced in the first period. However, FTC announced three divestiture orders requiring the divestiture of gas stations during the latter period. In terms of the use of clean sweeps, single buyers, and up-front buyers in fiscal years 1996 through 2000, we found that: The three divestiture orders involved 14 geographic markets, and 13 of the 14 (about 93 percent) geographic markets had clean sweep divestitures. Two of the divestiture orders involved clean sweep divestitures in all of the markets in which gas stations were divested. All three orders required single buyers to purchase the assets. (In total, approximately 980 assets were divested across the 14 geographic markets.) According to FTC, the divestiture of large packages of retail gasoline assets should allow the buyer to efficiently advertise a brand, develop credit card and other marketing programs, persuade distributors to market the buyer’s brand and otherwise compete in the sale of branded gasoline. Only one of seven buyers that purchased divested assets in the three divestiture orders was an up-front buyer. FTC staff told us that the remedy approved in each of the divestitures covered in our review was based on the unique facts of that case—not on any formula. Instead, FTC staff said that they examined each market in order to determine how most effectively to remedy the anticompetitive effects of the particular merger. They said that in one case, FTC may have ordered divestiture of all of the assets of one of the merging parties relating to a particular product market in every affected geographic market because FTC determined that any buyer would need a minimum scale and/or scope of operations. In another case, FTC may have ordered divestiture of all of the assets of one of the merging parties relating to a particular product market in most, but not all, of the geographic markets because the buyer had demonstrated that the remaining assets of that merging party were not profitable and, therefore, were undesirable. They added that FTC may have ordered divestiture of carefully selected assets in each geographic market without regard for which of the merging parties owned the assets, but only after an extensive and time-consuming financial analysis of each asset. According to FTC staff, in all cases, FTC’s objective has been to remedy the anticompetitive effects that FTC concluded would result from an otherwise unlawful acquisition, and the remedy was accepted only after a demonstration that it would achieve that objective. Although we did not have enough observations to analyze statistically the differences in the level of smaller business participation in purchasing divested drug store, funeral services, and gas station assets from fiscal years 1990 through 2000, our analysis of the buyers of divested grocery store assets showed that smaller buyers, including those that met SBA’s definition of a small business, were significantly less likely to directly purchase divested grocery store assets after fiscal year 1996. However, when we account for divested assets that were initially purchased by grocery wholesalers then, per the divestiture order, sold to other businesses—which reflects the level of indirect participation—the decline in the level of smaller business participation between the two time periods is not as great. Our discussions with FTC staff and antitrust practitioners indicated that the decline in the level of smaller business participation in purchasing divested assets occurred for a variety of reasons, including FTC divestiture practices and consolidation in the grocery store industry, specifically a significant difference in the type and size of mergers during the 1990s. Because there were so few buyers of divested assets in the drug store, funeral services, and gas station industries during the period, we were unable to determine if there was a statistically significant difference in the level of smaller business participation in purchasing these assets. While there were sufficient numbers of buyers of divested assets in the grocery store industry to enable us to analyze changes in the profile of the buyers over the 11-year period, there appeared to be no generally accepted definition of a small grocery store business. FTC staff, grocery industry officials, and representatives of relevant small business associations, told us that there was not a generally accepted standard by which one could measure the relative size of businesses using revenues. They acknowledged that SBA’s definition of a small business, which ranged from $13.5 million to $20 million in receipts from 1990 to 2000 for the grocery store industry, might be one indicator. However, industry members told us that SBA's threshold for a small grocery store business is too low given the relative size of grocery store businesses. Because there appeared to be no generally accepted definition of a small grocery store business, we used two approaches to determine the level of smaller business participation in purchasing divested grocery store assets in fiscal years 1990 through 1996 and fiscal years 1997 through 2000. First, we analyzed the median of the average annual revenues of the buyers. Specifically, we calculated the average annual revenues of each of the buyers of divested grocery store assets using the buyers’ 3 years revenues prior to their purchase of the divested assets, when available, in constant 2000 dollars. The average annual revenues ranged from about $5 million to about $35 billion, with 9 of 41 direct buyers having average annual revenues greater than $15 billion. We then computed the median of the average annual revenues for all the buyers for the two periods and for fiscal years 1990 through 2000. We used the overall median of the buyers’ average annual revenues for fiscal years 1990 through 2000 as a benchmark and determined the number of buyers in each of the periods that were below the overall median. We considered buyers below the overall median to be “smaller.” Second, we used SBA’s size standard for a small grocery store business. However, we used the buyers’ average annual revenues in constant 2000 dollars before they purchased the divested assets as a proxy for receipts, because we were unable to obtain data on the buyers’ receipts. We also adjusted SBA’s size standards to constant 2000 dollars. (Appendix II provides information on the adjusted SBA size standards.) We then determined the number of buyers in each of the periods that were below the adjusted SBA size standard. Additionally, to examine changes in the size of the buyers, we compared the median of the average annual revenues of the buyers for the two periods. Regardless of the approach we used, the results were similar—there were fewer smaller buyers purchasing divested grocery store assets after fiscal year 1996. Our analysis showed that significantly fewer smaller businesses directly purchased divested assets after fiscal year 1996, the first full fiscal year after which FTC began showing preferences for certain divestiture approaches, such as up-front buyers, in the grocery store industry. (Appendix II shows the results of our statistical test for the grocery store industry.) We calculated that the median of the average annual revenue of 41 direct buyers of divested grocery store assets, using constant 2000 dollars, was about $1.8 billion. Figure 1 shows that using less than $1.8 billion as a benchmark for a smaller business, 80 percent (12 of 15) of the buyers of the divested grocery store assets during fiscal years 1990 through 1996 were smaller, compared with about 31 percent (8 of 26) in the latter period. We also looked at the level of smaller business participation in directly purchasing divested grocery store assets from another perspective—small businesses as defined by SBA—and found that, when comparing the two periods, there was a significant decline in the number of smaller businesses that purchased divested grocery store assets. We used buyers' revenues before they purchased the divested assets as a proxy for receipts and determined the extent to which buyers in both periods did or did not meet SBA's thresholds (in 2000 constant dollars). Our analysis showed that, using SBA’s definition of a small grocery store business, smaller business participation in purchasing divested grocery store assets declined significantly from about 27 percent (4 of 15) in the period from fiscal years 1990 through 1996 to about 4 percent (1 of 26) in the latter period. Not only did the percentage of smaller businesses purchasing divested grocery store assets decline, but the size of businesses that purchased divested grocery store assets grew significantly over time as well as, according to FTC staff, the size of the parties to the acquisition and the acquisition value. We calculated the median of the average annual revenues of the buyers in each period and found that the typical buyers in the latter period were significantly larger than buyers in the earlier period. Specifically, the median of the average annual revenues of the buyers in the earlier period was about $89 million. By contrast, the median of the average annual revenues of the buyers in the latter period was about $3.3 billion. In recent years, grocery wholesalers have been a key factor in enabling smaller businesses to purchase divested assets in the grocery store industry—75 percent of the smaller buyers after 1996 purchased the divested assets from wholesalers. Under this practice, some grocery wholesalers, per the divestiture agreement reached with FTC, have purchased divested grocery store assets directly from the merging parties and under the order provision, in turn, sold them to grocery store operators. Thus, a direct buyer (the wholesaler) sells the asset(s) to indirect buyers (the grocery store operators). According to grocery wholesalers we interviewed, this practice has occurred in part because wholesalers were losing market share due to increasing consolidation in the grocery store industry. To determine if reselling the assets had an effect on the extent to which smaller businesses participated in purchasing divested grocery store assets, we performed a separate analysis. Whereas in the analysis of direct buyers, we considered the wholesalers as the buyers of divested assets, in our analyses that included indirect buyers, we replaced the revenues of the wholesalers with those of the buyers to which they resold the assets. When we accounted for the effect of indirect buyers, there was still a significant decline in smaller businesses participation in purchasing divested assets after fiscal year 1996; however, the extent of the decline was not as great. Specifically, when the buyers who purchased the divested assets indirectly from wholesalers were included in our analysis, the overall median of the average annual revenues of 54 buyers of divested grocery store assets for the fiscal years 1990 to 2000 time period was about $208 million—about 88 percent less than the overall median (about $1.8 billion) for direct buyers. Figure 2 shows that using less than $208 million as a benchmark for a smaller business, about 73 percent (11 of 15) of the buyers of the divested grocery store assets during fiscal years 1990 through 1996 were smaller buyers, compared with 41 percent (16 of 39) during fiscal years 1997 through 2000. We also examined the level of smaller business participation in indirectly purchasing divested grocery store assets using SBA’s definition of a small grocery store business. Again, we found that even though there continued to be a decline in the extent to which smaller businesses purchased these assets, the decline was not as great when we included indirect buyers in the analysis. Specifically, using SBA's definition, as in our earlier analysis, we found that about 27 percent (4 of 15) of the businesses that purchased divested grocery store assets during fiscal years 1990 through 1996 (including indirect buyers) were smaller businesses. By contrast, about 15 percent (6 of 39) of the buyers of divested assets during fiscal years 1997 through 2000 were smaller businesses. As in the case of the direct buyers, we found that in addition to the percentage of smaller businesses purchasing divested grocery store assets declining between the two periods, the size of the businesses that purchased the assets grew significantly over time. However, the median size of the buyers in the latter period was substantially less than that of the direct buyers. The median of the average annual revenues of the buyers in the earlier period was about $78 million, compared with about $288 million in the latter period—about 91 percent below the approximately $3.3 billion median for the direct buyers in the latter period. Through interviews with FTC staff, antitrust practitioners, and smaller buyers of divested grocery stores, we identified factors that may have contributed to the decline in the level of smaller business participation in purchasing divested assets in the grocery store industry. First, FTC’s clean sweep, single buyer, and up-front buyer divestiture practices and the merging parties’ desire to close the deal quickly may have impacted the ability of smaller businesses to purchase divested assets. However, FTC staff told us that FTC’s role is to protect competition and consumers, not particular competitors or businesses, whether large or small. Second, growing consolidation in the grocery store industry has resulted in fewer smaller businesses because many have either been acquired by larger companies or have gone out of business. According to FTC staff and antitrust practitioners, FTC’s divestiture practices and the desire of the merging parties to consummate the merger quickly may impact the ability of smaller buyers to purchase divested assets. Regarding FTC’s divestiture practices, FTC staff told us that FTC’s preference in retail industries for clean sweep divestitures combined with its preference for single buyers could impact the ability of smaller businesses to purchase divested assets, particularly in larger markets where there may be a greater number of assets to be divested, especially where economies of scale exist. FTC staff also said that it is easier and quicker for the merging parties to convince FTC that a clean sweep divestiture to a single buyer will be successful than a clean sweep divestiture to multiple buyers within a geographic market. They acknowledged that a larger package of divestiture assets may make it difficult for smaller buyers to purchase the divested assets because they might not have the financial strength to purchase all of the assets that have to be divested in a geographic market. FTC staff and antitrust practitioners also told us that negotiating a divestiture order is generally easier and quicker with a single, well- established chain as the buyer because FTC has to analyze the financial, managerial, and operational strength of only one buyer, and the merging parties have to negotiate with only one buyer. Similarly, antitrust practitioners said that FTC’s up-front buyer preference may create a bias against smaller businesses because the merging parties cannot consummate the merger until a viable buyer(s) has been identified, reviewed, and provisionally approved by FTC. Antitrust practitioners told us that because the merging parties want to consummate the merger quickly and it may take longer to (1) convince FTC that a smaller business is a viable buyer and (2) negotiate with a smaller business, the merging parties generally would prefer larger, well-established buyers. Additionally, the practitioners said that for these reasons they are likely to advise their clients to select a strong, well-established buyer that would clearly be acceptable to FTC. They added that the cost of delaying a merger while the merging parties are waiting for FTC to approve a buyer can be very expensive to the merging parties because of the delay in achieving the potential efficiencies they sought through the merger. FTC staff and antitrust practitioners also told us that factors other than clean sweeps, single buyers, and up-front buyers may impact the ability of smaller buyers to purchase divested assets. For example, antitrust practitioners said that FTC’s preference for buyers that do not already have a presence in the geographic market as well as FTC’s definition of the product and geographic markets may also impact the ability of smaller businesses to purchase divested assets. They told us that smaller businesses interested in purchasing divested assets may be more likely to be located in the geographic market where the assets are being divested. According to FTC staff, if any business, large or small, interested in purchasing the divested assets already has a significant presence in the geographic market where the assets are being divested, it reduces the chances of FTC approval because divesting to a market incumbent does not replace the acquired firm and thus reduces the competitive effectiveness of the divestiture. Thus, according to FTC staff, FTC may not approve proposed buyers that already have a significant presence in the geographic market because this raises concerns about the anticompetitive effects of the divestiture transaction, including an increase in concentration and failure to maintain the number of market participants. For example, in fiscal year 1996, FTC did not approve a proposed buyer for a grocery store in the Stop and Shop and Purity Supreme merger, because the proposed buyer already had two stores relatively close to the store being divested. However, FTC staff said that in certain instances smaller businesses operating within the same geographic and product markets have been found to be acceptable buyers. Additionally, FTC staff and antitrust practitioners told us that the size of the divestiture package and FTC staff’s definition of the geographic and product markets may affect the opportunity for smaller businesses to purchase divested assets. They said that the larger the geographic market, as defined by FTC, the less opportunity there may be for smaller businesses to purchase the divested assets. The Food Marketing Institute (FMI) and the National Grocers Association (NGA)—the two largest associations that represent grocers—have submitted comments to FTC stating that FTC’s clean sweep, single buyer, and up-front buyer divestiture practices have hindered the ability of small businesses to purchase divested assets. Additionally, they said that FTC’s strong preference for buyers located outside the geographic market in which the assets are to be divested have disadvantaged small businesses, which are struggling to expand and keep pace with large corporations. Several smaller buyers of divested grocery store assets also told us that other factors, such as the merging parties’ bidding process, create additional challenges for smaller businesses in purchasing and maintaining the viability of divested assets. (Appendix VII provides information on the public comments that FTC received regarding all 31 divestiture orders included in our review as well as the results of our discussions with several smaller buyers of divested grocery store assets, selected associations that represent smaller businesses, and SBA officials.) FTC staff told us that the antitrust statutes are designed to protect competition and consumers, not particular competitors or businesses, whether large or small. They also said that in several public statements, FTC staff have noted that FTC does not have a preference in favor of large- chain buyers, or a preference against small chains, independents, or wholesalers that will eventually spin off stores to other buyers. For example, in a March 2001 speech before the American Bar Association, the then Acting Director of FTC’s Bureau of Competition said that FTC neither favors nor disfavors any particular category of purchaser. According to the speech, “...the Commission’s approach to supermarket divestitures has not precluded smaller or local grocery stores from participating as buyers because it believes that effective competitors come in all shapes and sizes.” Additionally, the then Acting Director said that FTC is sensitive to the fact that small supermarket chains often offer greater product variety and choice than other supermarkets. However, FTC staff told us that the size of a possible buyer is a factor in determining the acceptability of that buyer, but only to the extent that size affects the buyer’s financial viability and ability to operate the divested assets competitively. Some antitrust practitioners we interviewed also told us that they do not perceive FTC as having a role in protecting small businesses. Like FTC, they said FTC’s mandate is to protect and preserve competition and consumers, not to protect or promote small businesses. They told us that protecting competition and promoting small businesses are not always completely consistent goals. Growing consolidation in the grocery store industry has resulted in fewer smaller businesses because many have either been acquired by larger companies or have gone out of business. An article by the U.S. Department of Agriculture’s (USDA) Economic Research Service states that in recent years, the U.S. food retailing industry has undergone unprecedented consolidation and structural change through mergers, acquisitions, divestitures, internal growth, and new competition. Widespread consolidation in the grocery store industry, driven by expected efficiency gains from economies of size, has had a significant effect on the share of total grocery stores sales accounted for by the largest food retailers. The concentration levels of the industry have also increased. According to a August 1999 American Antitrust Institute (AAI) article, in 1992, the top five supermarket chains had 19 percent of the national market. In 1999, that share had increased to at least 33 percent. According to FMI data, the number of chain supermarkets increased from 17,460 in 1990 to 20,825 in 2000. Over the same period, the number of independent supermarkets declined from 13,290 to 11,005. The Progressive Grocer 2001 annual report, reports that market share continues to be consolidated among a handful of traditional players, including Kroger Company; Safeway, Inc.; Albertson’s, Inc.; and Royal Ahold. According to the report, larger chains will acquire smaller and mid-size independents to solidify their market share and increase their buyer power. However, the report also notes that the closings of grocery stores resulting from larger chains that are rejecting leases on older, smaller stores when they come up for renewal will afford independents and smaller chains the chance to expand by acquiring those stores. FTC has not systematically measured the success or failure of the divestitures it has approved since it developed preferences for approaches like clean sweep and up-front buyers. In 1999, FTC reported the results of the Bureau of Competition staff’s study of divestiture orders made final during fiscal years 1990 through 1994 that, according to FTC staff, confirmed the need to make the changes that FTC had made starting in the mid-1990s. The report was drafted by FTC’s Bureau of Competition. Although the study had some methodological limitations, it appears to have been instrumental in helping FTC staff better understand the divestiture process. Nonetheless, antitrust practitioners and representatives of the grocery store industry have questioned key aspects of FTC's study, including whether the approach used supported the conclusions drawn and whether the study went far enough in measuring the impact of its divestiture practices on buyers of divested assets and the markets in which they operate. To perform its study, FTC staff examined 35 divestiture orders made final during fiscal years 1990 through 1994 to identify problems with the divestiture process and determine whether buyers of divested assets were able to begin operating in the relevant market relatively quickly and maintain operations. The study found that the acquirers of divested assets generally were viable competitors in the markets of concern and found, among other things, that across all orders studied divestitures of ongoing businesses succeeded at a higher rate than divestitures of selected assets; parties to a potential merger that were required to divest assets sometimes looked for buyers who were not the strongest competitors and sometimes engaged in strategic conduct to impede the success of the buyer; many buyers of divested assets did not have sufficient information to prevent mistakes in the course of their negotiations and subsequent acquisitions, particularly where the buyers had never operated in the industry or the to-be-divested business; and smaller buyers succeeded at least at the same rate as larger buyers and, therefore, should not be presumed to be less competitive buyers than larger firms. The study confirmed much of what FTC staff had suspected and discussed in earlier speeches and recommended a number of consent provisions and approaches designed to mitigate some of the problems FTC staff identified. According to FTC staff, the recommendations were designed to correct the informational and bargaining imbalance that had occurred in previous divestitures. We were unable to fully assess the divestiture study because, according to FTC, the nonpublic version of the study contained proprietary financial information about the merging parties and the buyers of the divested assets. However, our analysis of the public version of the study showed that it had some key limitations. For example, The study discussed the advantages of up-front buyers, and in speeches, FTC staff used the study as a basis for articulating FTC’s preference for up-front buyers. However, the study did not provide information on the number of up-front buyers it examined, nor does it discuss what factors made those buyers successful. A finding in the report is that smaller firms across all orders studied succeeded at least at the same rate as larger firms and, therefore, should not be presumed to be less competitive buyers than larger firms. The study does not define “smaller” nor does it provide information on the industries in which the smaller businesses operated. FTC staff told us that the divestiture study did not use quantitative criteria for categorizing firms as smaller or larger. Instead, FTC staff looked at (1) the size of buyers on a continuum of smaller revenues to larger revenues for the sample of businesses surveyed and/or (2) whether the firm was a single product firm, often a newly formed company, operating in a fairly localized area, as opposed to an established, multiproduct, highly capitalized, multinational company. FTC staff told us that it was clear to them which buyers were smaller and which were larger. The study covered divestitures to 50 buyers that were approved by FTC, but FTC staff were only able to interview 37 of the buyers, and the study does not provide information on the distribution of the 13 buyers (26 percent) that did not participate in the study. If these buyers are mostly smaller or larger, or fall into the same industry, the results could potentially be skewed. The methodology for the study published in the Federal Register stated that FTC would interview 147 third parties, such as customers, suppliers, and trustees. However, the final report only reports on interviews with two third parties, both of which were trustees. According to FTC staff, due to resource constraints, they were limited in the number of interviews they could complete. They told us that the study focused on the buyers and the merging parties, and although it would have been useful to obtain information from third parties, it would have been costly, and would have diverted Commission resources from pressing merger enforcement activities. Nonetheless, FTC staff told us that in reviewing a proposed merger, they rely heavily on information provided by third parties to determine whether a merger is likely to result in price increases or reduction in quality or output. FTC staff told us that they recognize that the divestiture study had limitations. Nevertheless, they said that the study provided them sufficient information to adjust their approaches to working on divestiture orders. For example, FTC staff said that they learned that buyers of divested assets often did not know all of the assets they needed to effectively operate a business. As a result, FTC staff stopped relying as heavily on buyers of divested assets to inform them of the assets that should be included in the divestiture packages. FTC staff said that while the divestiture study is not the ultimate competitive analysis, it is an important step in determining whether the divestiture orders are satisfying FTC’s mission. Antitrust practitioners have praised FTC for taking the initiative to do the divestiture study, but have also questioned key aspects of the study, including its design and methodology. For example, some antitrust practitioners we spoke with said that FTC should be commended for undertaking the study, but questioned whether the methodology employed was sufficient for drawing conclusions. They suggested that the study should have been designed in a way to enable FTC to reach conclusions about how the divestiture(s) impacted the availability, quality, and price of products because these are the market factors that ultimately affect the consumer. “…the Divestiture Study has taken on a significance, in terms of justification of Commission policies, that was not foreseen at the time it was issued, especially since it was never formally adopted by the Commission. In addition, it is not clear from the description of the Study whether the authors, in evaluating the apparent failure of a remedy, considered whether this was consistent with the lack of a need for a remedy in the first place.” "Studies such as this would be more useful contributions to the dialogue between the agencies and the private bar if their methodology was fully discussed and disclosed prior to the study, and the study included the participation (with appropriate confidentiality constraints) of outside academics and practitioners.” Given questions that have been raised about FTC’s divestiture study, we spoke with FTC staff about the benefits and cost of an updated, more comprehensive study. They acknowledged that another study might be beneficial because FTC has not studied the effect of its recent approaches on the viability of buyers of divested assets or on competition in the marketplace. Although they could not pinpoint the actual cost of doing another study, they said that it could be labor-intensive and burdensome on businesses. FTC staff told us that the 1999 study was extremely labor intensive and another study, especially one attempting to assess market competition in each market, would require extensive resource allocations. FTC staff said that they do not currently have resources to commit to such a study at this time. For example, in doing the 1999 study, FTC staff had to first manually review all FTC actions, during the fiscal years 1990 through 1994 time period, to identify consent orders involving mergers and then, determine which orders required divestitures. They said that it might be easier today because recent divestiture orders are more readily available through FTC’s Web site. However, our review of divestiture orders announced for public comment during fiscal years 1990 through 2000 revealed that the Web site does not provide a way to readily identify all of the divestiture orders FTC announced or made final each year. Additionally, FTC staff told us that they would be concerned that such an effort could be burdensome on businesses. For example, they said that they would have to rely on buyers of divested assets to obtain information post-divestiture, and currently, buyers are not parties to divestiture orders. Thus, currently buyers are not required to provide information to FTC post- divestiture. In performing the 1999 study, FTC staff relied on the cooperation of buyers to collect information for its divestiture study and, as noted previously, 13 of the 50 buyers (26 percent) did not participate in the study. However, FTC is currently examining whether it should take steps to overcome this problem. In June 2002, FTC began holding a series of public workshops to obtain insights into its merger remedy process. The workshops are designed to address, among other things, whether FTC should require the buyers of divested assets to report on their operations of the divested assets to permit FTC to better determine whether its remedy provisions have achieved the objective of maintaining or restoring competition in the relevant market. The workshops also will consider what any such reporting requirements should entail, how long they should be in place, and the impact of any such obligations on potential buyers’ interest in acquiring divestiture assets. FTC officials have frequently noted in speeches that if a merger does not result in restored competition, the remedy is not a success. For example, in a March 2001 speech, the then Acting Director of the Bureau of Competition said that consumers should not bear the risk of an inadequate or ineffective remedy. “A merger is forever… and, therefore, so is the harm caused by an incomplete remedy.” These views are consistent with the then Chairman’s statement in a February 2000 speech: “The law is clear that divestiture and other restructuring remedies should not be adopted unless they are likely to restore fully the competition lost as a result of the merger.” He further said, “Enforcement agencies should not be expected and would not be justified in making the same mistake over again. If restructuring in a particular industry and in similar circumstances has been unsuccessful, enforcement officials have a responsibility to determine why.” Moreover, FTC in explaining antitrust laws has said that a merger that lessens competition can lead to higher prices, reduced availability of goods or services, lower quality of products, and less innovation. Despite this recognition, FTC has not fully evaluated whether the divestitures it has approved have achieved its goal of maintaining or restoring competition in the marketplace. FTC’s mission is, in part, to prevent business practices that are anticompetitive, deceptive, or unfair to consumers. FTC attempts to achieve its mission through preventing anticompetitive mergers from taking place. FTC’s approaches to merger remedies have evolved over time. In fiscal year 1996, based on past experiences and preliminary findings of its divestiture study, FTC developed preferences for how to structure divestitures and began to modify its divestiture approaches. FTC has not studied the effect of its recent approaches on the viability of buyers of divested assets or on competition in the marketplace. For example, in the grocery store industry, clean sweep divestitures, single buyers, and up- front buyers have been used with greater frequency in the late 1990s—a period not covered in FTC’s 1999 divestiture study. Because FTC has not systematically reviewed divestiture orders made final after fiscal year 1994, FTC does not know how these practices have impacted the viability of the buyers of divested assets or prices and/or innovation in the marketplace— key factors FTC uses to measure competition. Consequently, FTC cannot state that recent divestiture orders have, among other things, maintained or restored competition in the affected markets, or that smaller buyers continue to be as competitive as their larger counterparts in operating the divested assets. Although the cost of doing a study of its recent divestiture orders could be considerable, such a study would give FTC the opportunity to develop and design a methodology to (1) overcome some of the limitations encountered during the earlier study; (2) provide greater insights into the impact of its divestiture preferences on potential buyers, including smaller businesses; and (3) better examine the short and long-term effects of various divestiture preferences on the markets in which divestitures take place. FTC staff acknowledged the benefit of conducting a more rigorous and comprehensive study to determine whether recent merger remedies are achieving their intent—to adequately maintain or restore competition and protect consumers so that consumers have the benefit of low prices and good product variety. Similarly, antitrust practitioners have said that collecting relevant economic data and including the data within the scope of future studies would place FTC in a better position to reach firm conclusions regarding the success or failure of its divestiture practices and to shape future divestiture orders. We recognize the difficulties and costs inherent in conducting an evaluation that assesses the impact of FTC’s divestiture practices on the marketplace. However, the more that economic data are brought to bear on the questions of how FTC’s divestiture practices impact the marketplace, including smaller businesses, the more confident FTC can be that divestitures are having the intended effect of maintaining or restoring competition. We also recognize that the need for and benefits of conducting the analysis must be balanced against the costs of collecting the data as well as balanced against any logistical and legal implications. FTC could not provide us with an estimate of the costs to perform a more comprehensive divestiture study, nor could we estimate the costs. While we are aware that the costs could be significant, we believe that the lack of a more rigorous and comprehensive study places FTC at risk of not being in the position to fully understand the effects of its divestiture practices, including possible negative effects, on competition in the marketplace—a risk that FTC officials have said that FTC cannot afford to take. To ensure that FTC has complete and up-to-date information on the effectiveness of divestitures in industries in the retail sector, we recommend that the Chairman of FTC direct the Bureaus of Competition and Economics to undertake a study of the impact of divestiture orders made final since fiscal year 1994 that require divestitures in the retail sector on (1) the viability of buyers of divested assets and (2) competition in the marketplace. If the findings show that FTC's intended results have not been achieved, we further recommend that FTC explore expanding the study to include divestiture orders for other sectors of the economy that have been impacted by changes to FTC divestiture practices during the mid-1990s. In commenting on our report, FTC said that the report adds important information about the Commission’s role in enforcing antitrust laws related to mergers and should lead to further improvement in the Commission’s merger enforcement efforts. In its comments, which are included as appendix VIII, FTC said that our recommendation is consistent with the Commission’s own objectives and its most recent Government Performance and Results Act report, which states that the Commission plans to “study and evaluate the remedies used in past antitrust cases, particularly divestiture orders used to resolve merger cases.” FTC noted that it currently has two related studies underway concerning hospital and petroleum mergers. FTC also provided additional information on the history of the Commission’s divestiture and remedy practices. This information describes the impetus for Congress’ passage of the HSR Act and reiterated that FTC is taking certain actions to seek public input on merger remedy issues, as discussed in our report. Additionally, FTC previously provided technical comments, which have been incorporated in this report where appropriate. We are sending copies of this report to the Chairman of the Federal Trade Commission and interested congressional committees. We will also provide copies to others on request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions about this report, please contact John F. Mortin at 202-512-5727 or me at 202-512-4636. Our objectives were to describe (1) the history of Federal Trade Commission’s (FTC) clean sweep divestiture, single buyer, and up-front buyer practices within the context of FTC’s overall merger remedies, and the circumstances under which these practices have been used; (2) the extent of FTC’s use of these practices in the grocery store, drug store, funeral services, and gas station industries; (3) the level of small business participation in purchasing divested assets in the four industries and the factors that may explain the level of small business participation; and (4) FTC’s efforts to gauge the success or failure of these divestiture practices and the impact of these practices on the marketplace, especially small businesses. We also obtained information on the nature of comments FTC received pertinent to these practices in the four industries as well as comments on how they impacted the ability of small businesses to purchase divested assets. As agreed with your staff, we focused on divestitures required by consent orders that FTC announced for public comment between fiscal years 1990 through 2000 for four industries— grocery store, drug store, funeral services, and gas station industries—that sell products and/or services directly to consumers. We focused on proposed consent orders that were announced for public comment during fiscal years 1990 through 2000, rather than consent orders made final during this period, because in most cases, at the point that FTC accepts a proposed consent order for public comment, the parties are allowed to consummate their merger. Our work was limited to publicly available information because, according to FTC, under the Hart-Scott-Rodino Act (HSR Act) of 1976, FTC is prohibited from disclosing information provided by parties to merger transactions. To address these objectives, we performed our work at FTC headquarters in Washington, D.C. We also contacted officials from various associations that represent or deal with small and/or independent businesses—the National Grocers Association (NGA), the Food Marketing Institute (FMI), the National Community Pharmacists Association (NCPA), the National Funeral Directors Association (NFDA), the Petroleum Marketers Association of America (PMAA), the National Federation of Independent Business (NFIB), the National Business Association (NBA), the National Small Business United (NSBU), the SCORE Association (SCORE), and the U.S. Chamber of Commerce. In addition, we obtained and reviewed relevant documents prepared by and discussed FTC’s divestiture practices with antitrust practitioners—attorneys and economists in private practice who specialize in antitrust issues—academicians and officials with the Small Business Administration (SBA). To address the first objective concerning the history of FTC’s clean sweep divestiture, single buyer, and up-front buyer practices within the context of FTC’s overall merger remedies, and the circumstances under which these practices have been used, we met with staff from FTC’s Bureaus of Competition and Economics and its Office of General Counsel and interviewed a former FTC Commissioner and former staff. We also reviewed relevant FTC public speeches, articles, and documents prepared by current and former FTC Commissioners and staff between the years 1995 and 2002 pertaining to FTC’s divestiture practices. Furthermore, we obtained and reviewed relevant documents prepared by and talked with antitrust practitioners, academicians, associations that represent small and/or independent businesses, and officials from SBA to obtain their views of FTC’s divestiture practices. In addition, we met with an official of the Department of Justice’s (DOJ) Antitrust Division to obtain general information on DOJ’s merger remedies. To address the second objective concerning the extent of FTC’s use of the clean sweep divestiture, single buyer, and up-front buyer practices in the grocery store, drug store, funeral services, and gas station industries, we analyzed FTC consent orders that required a divestiture(s), also called divestiture orders, that were announced for public comment during fiscal years 1990 through 2000. Because FTC did not have a publicly available listing of the divestiture orders announced for public comment during fiscal years 1990 through 2000, we reviewed several sources of information. To identify divestiture orders FTC announced for public comment during fiscal years 1990 through 1995, we reviewed FTC decision books for calendar years 1989 through 1995 to identify all divestiture orders issued by FTC during that period. We then reviewed FTC press statements announcing each of the proposed divestiture orders to determine the fiscal year in which FTC announced the order. To identify divestiture orders FTC announced for public comment during fiscal years 1996 through 2000, we reviewed a summary of FTC’s Bureau of Competition activity for fiscal year 1996 through March 31, 2000. We then reviewed FTC monthly actions announced for public comment and posted on FTC’s Web site for the period April 1, 2000, through September 30, 2000, to identify the divestiture orders FTC announced for public comment during April 2000 through September 2000. Finally, we reviewed FTC’s HSR Act annual reports to Congress for fiscal years 1990 through 2000. After we identified the 153 divestiture orders that were announced for public comment during fiscal years 1990 through 2000, we reviewed them to determine which ones required divestitures in the grocery store, drug store, funeral services, and gas station industries. We determined that during the 11-year period, FTC announced for public comment 33 divestiture orders that required divestitures in the four industries. However, we eliminated two funeral services divestiture orders from our review because, according to FTC staff, as of June 30, 2002, the divestitures had not taken place. After we identified the divestiture orders for the four industries, we used a structured data collection instrument to gather information from each of the 31 divestiture orders and related documents, such as press releases announcing a proposed divestiture order, complaints, analyses to aid public comment, final divestiture orders, modifications to the orders, divestiture applications, and FTC letters approving divestiture applications. Specifically, we collected information on, among other things, (1) the date FTC announced the divestiture order for public comment; (2) the geographic market in which FTC required a divestiture; (3) the buyer(s) of the divested assets in each geographic market, both the “direct” buyers— buyers that purchased the divested assets directly from the merging parties—and the indirect buyers—third parties that purchased the assets from direct buyers of divested assets, generally a wholesaler; (4) whether the buyers were identified in the divestiture order at the time FTC announced it for public comment; and (5) whether the divestiture order required a single buyer and/or an up-front buyer. Because FTC's divestiture orders and related documents typically do not indicate whether a divestiture is a clean sweep, we relied on clean sweep data provided by FTC. Because FTC speeches about its merger remedies indicated that FTC began to make changes in its divestiture practices during fiscal year 1996, we focused our analysis of the use of clean sweeps, single buyers, and up-front buyers on two periods—the period from fiscal years 1990 through 1995, before FTC started to make these changes, and fiscal years 1996 through 2000, the period during which FTC staff announced that FTC had begun to make changes. We determined the extent to which these practices were used in each of the two periods for the 31 divestiture orders included in our review. Additionally, because according to FTC staff each industry operates differently, we performed separate analyses for each of the four industries. To address the third objective concerning the level of small business participation in purchasing divested assets in the four industries and the factors that may explain the level of small business participation, we obtained revenue data for the buyers of the divested assets, both the direct and indirect buyers. Because according to FTC staff, due to confidentiality restrictions, they could not provide us with the revenue data for the buyers of the divested assets, we obtained the revenue data from several sources. For publicly held buyers, we obtained the data from public sources, such as the Securities and Exchange Commission filings and Valueline Magazine. For privately held buyers for which we could not locate the data through publicly available sources, we contacted the buyers directly to obtain the data. When possible, we obtained the revenue data for the 3 most recent completed fiscal years prior to the buyers being identified in the proposed divestiture orders or in the divestiture applications. If revenue data were not available for the 3 most recent completed fiscal years, we obtained the data for the prior 2 years or year. For the direct buyer analyses, we were not able to obtain revenue data for 3 of the buyers of divested grocery store assets, and for the analyses that included direct and indirect buyers, we were not able to obtain revenue data for 5 of the buyers. Because there were so few buyers of divested assets in the drug store, funeral services, and gas station industries during the period, we limited our analysis to the grocery store industry. According to FTC staff, SBA officials, grocery industry officials, and representatives of relevant small business associations, there is no generally accepted standard by which one could measure the relative size of businesses using revenues. They acknowledged that SBA's definition of a small business, which ranged from $13.5 million to $20 million in receipts from 1990 through 2000 for the grocery store industry, might be one indicator. However, they told us that SBA's threshold for a small grocery store business is too low given the relative size of grocery store businesses. Because there appeared to be no generally accepted definition of a small grocery store business, we used two approaches to determine the level of smaller business participation in purchasing divested grocery store assets in fiscal years 1990 through 1996 and fiscal years 1997 through 2000. We used these two time periods for the size of buyers analyses, instead of the fiscal years 1990 through 1995 and fiscal years 1996 through 2000 time periods that we used in our analyses of FTC’s divestiture practices, primarily because fiscal year 1997 is the first full fiscal year after FTC began altering its approach to divestitures. Additionally, there were too few observations in the fiscal years 1990 through 1995 period. Our analyses of the size of the buyers is based on the dates of the merging parties’ divestiture applications or, in the case of up-front buyers, the dates of the proposed divestiture orders—generally the point at which FTC receives the revenue data for the proposed buyers to include in its review of the viability of the proposed buyers. Because FTC did not start to use up-front buyers in grocery store divestitures until fiscal year 1996, divestitures in the earlier period took much longer to occur. Additionally, fiscal year 1997 is the first full fiscal year after FTC began altering its approach to divestitures. In regard to the two approaches we used to determine the level of smaller business participation in purchasing divested assets, we first analyzed the median of the average annual revenues of the buyers. Specifically, we calculated the average annual revenues of each of the buyers of divested grocery store assets using the buyers’ 3 years revenues, when available, prior to their purchase of the divested assets in constant 2000 dollars. We then computed the median of the average annual revenues for all the buyers for the two periods and for fiscal years 1990 through 2000. We used the overall median of the buyers’ average annual revenues for fiscal years 1990 through 2000 as a benchmark and determined the number of buyers in each of the periods that were below the overall median. We considered buyers below the overall median to be “smaller.” Second, we used SBA’s size standard for a small grocery store business at the time the buyers of divested assets were either identified in the divestiture order or in the divestiture application to determine the extent to which buyers in both periods did or did not meet SBA's thresholds. However, we used the buyers’ average annual revenues in constant 2000 dollars before they purchased the divested assets as a proxy for receipts, because we were unable to obtain data on the buyers’ receipts. We also adjusted SBA’s size standards to constant 2000 dollars. We then determined if there was a statistically significant change in the number of buyers that met the adjusted SBA thresholds for the two periods. Additionally, to examine changes in the size of the buyers, we compared the median of the average annual revenues of the buyers for the two periods and determined whether differences in the medians were statistically significant. (Appendix II provides more detailed information on our approaches for determining the level of smaller business participation in purchasing divested assets in the grocery store industry.) As part of our work addressing the factors that may explain the level of small business participation in purchasing divested assets in the four retail industries included in our review, we interviewed staff from FTC’s Bureaus of Competition and Economics and its Office of General Counsel. We also interviewed antitrust practitioners and academicians to obtain their views on factors that may explain a decline in the level of small business participation in purchasing divested assets. In addition, we reviewed relevant articles to determine factors that may explain the level of small business participation. Unless provided in a public comment, we could not obtain information on small businesses that may have attempted to purchase divested assets but were not selected or approved as a buyer. To address the fourth objective concerning FTC’s efforts to gauge the success or failure of its divestiture approaches and the impact of its approaches on the marketplace, especially small businesses, we reviewed FTC’s 1999 Divestiture Study and interviewed and reviewed relevant speeches and documents prepared by current and former FTC Commissioners and staff and antitrust practitioners. Specifically, we reviewed the study to determine whether it was methodologically sound and contextually sophisticated. We also interviewed and reviewed documents prepared by economists who consult merging parties and academic experts to obtain their views on FTC’s 1999 Divestiture Study and how FTC should measure the success of its divestiture approaches. To obtain information on the nature of comments FTC received pertinent to its clean sweep divestiture, single buyer, and up-front buyer practices and how these practices impact the ability of small businesses to purchase divested assets, we reviewed FTC public case files for the 31 divestiture orders included in our review to identify the public comments FTC received concerning the divestiture orders and divestiture applications. FTC staff told us that in some cases, the commenter requests that their comments be kept confidential. FTC keeps the confidential comments in nonpublic files. We did not review the nonpublic files nor did we verify that the public files contained all of the nonconfidential public comments received by FTC. We relied on FTC to provide us with the information. We reviewed each of the 1,902 public comments contained in the files and, when available, FTC’s responses to the comments to determine if they related to the ability of small businesses to purchase divested assets due to FTC’s divestiture practices. In addition, we interviewed FTC staff to obtain information on their process for obtaining public comments and to determine whether FTC had received any complaints, outside of the public comment process, concerning its clean sweep divestiture, single buyer, and up-front buyer practices and how these practices impact the ability of small businesses to purchase divested assets. To determine if associations that represent small and independent businesses had concerns related to the impact of FTC’s divestiture practices on the ability of small businesses to purchase divested assets, we also interviewed representatives from 10 associations whose membership in general, consists of the following: NGA is the national trade association representing retail and wholesale grocers that comprise the independent sector of the food distribution industry. FMI conducts programs in research, education, industry relations, and public affairs on behalf of 2,300 food retailers and wholesalers in the United States and around the world. Its retail membership comprises of independent supermarkets, large multistore chains, and regional firms. NCPA represents the pharmacist owners, managers, and employees of nearly 25,000 independent community pharmacies across the United States. The independent community pharmacists are small business entrepreneurs and multifaceted health care providers. NFDA provides advocacy, education, information, products, programs, and services to help funeral directors enhance the quality of service to families. Ninety-three percent of its members are small business owners/operators with an average of four full-time employees. PMAA is a federation of 42 state and regional trade associations representing approximately 7,850 independent petroleum marketers nationwide. It seeks to further the common business interest of the petroleum marketing industry. NFIB is an advocacy organization representing 600,000 small and independent businesses. Its membership includes independent professional, retailers, service providers, manufacturers, farmers, and wholesalers. NBA is a nonprofit organization that supports and educates the small business community and the self-employed. NSBU is an advocacy organization representing over 65,000 U.S. members. It informs small business owners about legislative and regulatory issues that affect them. The U.S. Chamber of Commerce is a nonprofit business federation representing businesses of all sizes. More than 96 percent of its members are small businesses with 100 or fewer employees. The SCORE Association is a national, nonprofit association with 11,500 volunteer members and 389 chapters throughout the United States and its territories. It provides general business advice to small businesses and to persons interested in starting a business. We also reviewed comments provided to FTC for its June 2002 workshop on merger remedies. In addition, we met with officials from SBA’s Office of Advocacy to determine if the agency has received any complaints from small businesses about their ability to purchase divested assets due to FTC’s divestiture practices. Finally, while obtaining revenue data from buyers of divested assets, representatives of 11 grocery store businesses with average annual revenues of $200 million or less provided anecdotal comments about the challenges that smaller businesses face in purchasing divested assets and maintaining their viability. We did our work between April 2001 and August 2002 in accordance with generally accepted government auditing standards. As discussed in appendix I, to address our objective to determine the level of smaller business participation in purchasing divested assets in the grocery store, drug store, funeral services, and gas station industries, we collected publicly available data on the revenue of buyers of divested assets, both the direct and indirect buyers, for the four industries and, when public data were not available, from the buyers of divested assets. We did not have enough buyers in the drug store, funeral services, and gas station industries to analyze statistically the differences in the level of small business participation in purchasing divested assets in these industries from fiscal years 1990 through 2000. Therefore, we limited our analysis to the grocery store industry. We used two approaches to determine the level of smaller business participation in purchasing divested grocery store assets in fiscal years 1990 through 1996 and fiscal years 1997 through 2000. The analysis was done using these two time periods primarily because fiscal year 1997 was the first full fiscal year after FTC began altering its approach to divestitures and began increasingly using clean sweep, single buyer, and up-front buyer divestiture practices in the grocery store industry—practices that may impact the level of smaller business participation in purchasing divested assets. We did our analysis for the direct and indirect buyers of divested grocery store assets. First, we analyzed the median of the average annual revenues of the buyers. Specifically, we calculated the average annual revenues of each of the buyers of divested grocery store assets using the 3 years revenues prior to the purchase of the divested assets, when available, in constant 2000 dollars. We then computed the overall median of the average annual revenues for all the buyers for fiscal years 1990 through 2000. Using the overall median of the buyers’ average annual revenues for fiscal years 1990 through 2000 as a benchmark, we determined the number of buyers in each of the two periods—fiscal years 1990 through 1996 and fiscal years 1997 through 2000—that were below the overall median. We considered buyers with average annual revenues below the overall median to be “smaller.” To determine if the changes in the level of smaller business participation in purchasing the divested assets in the two time periods were statistically significant, we used the Fisher Exact test (in MegaStat, by J.B. Orris, Butler University), which compares the percentage of smaller buyers in each time period. The results, presented in table 4, indicate that the percentage of smaller buyers FTC approved to purchase divested assets in the grocery industry during fiscal years 1997 through 2000 was significantly less than the percentage of smaller buyer approved during fiscal years 1990 through 1996, when smaller businesses were both direct buyers and indirect buyers of divested grocery store assets. Second, we used the Small Business Administration’s (SBA) size standard for a small grocery store business to identify smaller buyers. We used the buyers’ average annual revenues in constant 2000 dollars before they purchased the divested assets as a proxy for receipts, because we were unable to obtain data on the buyers’ receipts. We also adjusted SBA’s size standard to constant 2000 dollars. Table 5 shows SBA’s size standard in current and in constant 2000 dollars for each year for the fiscal years 1990 through 2000 time period. We then determined the number of direct and indirect buyers of divested grocery store assets in each of the two periods that were below the adjusted SBA size standard for the grocery store industry. The results, presented in table 6, indicate that the percentage of smaller buyers FTC approved to purchase divested assets in the grocery industry during fiscal years 1997 through 2000 was significantly less than the percentage of smaller buyer approved during fiscal years 1990 through 1996 when smaller businesses were both direct and indirect buyers of divested grocery store assets. Additionally, to examine changes in the size of the buyers, for both the direct and the indirect buyers, we calculated the overall medians of the average annual revenues for fiscal years 1990 through 1996 and fiscal years 1997 through 2000. We then compared the overall medians of the annual revenues of the buyers for the two time periods. To determine if the changes in the sizes of the buyers of the divested assets in the two time periods were statistically significant, we used the Wilcoxon-Mann/Whitney test for two independent samples (in MegaStat, by J.B. Orris, Butler University), which compares the locations (medians) of two independent samples. The results, presented in table 7, indicate that the sizes of the buyers have increased significantly from the fiscal years 1990 through 1996 to fiscal years 1997 through 2000, for both the direct and indirect buyers. Product market(s) Retail sale of food and grocery items in supermarkets. The Vons Companies, Inc. & Williams Bros. Markets, Inc. Retail sale and distribution of food and grocery items in supermarkets. Red Apple Companies, Inc. + John A. Catsimatidis + Supermarket Acquisition Corp. + Designcraft Industries, Inc. & Sloan’s Supermarkets, Inc. Retail sale of food and grocery products in supermarkets. Retail sale of food and grocery products in supermarkets. Schnuck Markets, Inc. & National Holdings, Inc. Retail sale of food and grocery products in supermarkets. Schwegmann Giant Super Markets, Inc. & National Holdings, Inc. Retail sale of food and grocery products in supermarkets. The Stop and Shop Companies, Inc. & Purity Supreme, Inc. Retail sale of food and grocery products in supermarkets. Koninklijke Ahold nv + Ahold USA, Inc. & The Stop & Shop Companies, Inc. Retail sale of food and grocery products in supermarkets. Jitney-Jungle Stores of America, Inc. + Delta Acquisition Corporation & Delchamps, Inc. Retail sale of food and grocery products in supermarkets. Albertson’s, Inc. + Locomotive Acquisition Corporation & Buttrey Food and Drug Store Company, Inc. Retail sale of food and grocery products in supermarkets. Koninklijke Ahold nv & Giant Food Inc. Retail sale of food and grocery products in supermarkets. The Kroger Co. & Fred Meyer, Inc. Retail sale of food and grocery products in supermarkets. Retail sale of food and grocery products in supermarkets. Product market(s) Shaw’s Supermarkets, Inc. + J Sainsbury plc & Star Markets Holdings, Inc. Retail sale of food and grocery products in supermarkets. The Kroger Co. & The John C. Groub Company, Inc. Retail sale of food and grocery products in supermarkets. Delhaize America, Inc. (Food Lion) + Etablissements Delhaize Freres et Cie "Le Lion" S.A. & Hannaford Bros. Co. Retail sale of food and grocery products in supermarkets. Revco D.S., Inc. & Hook-SupeRx, Inc. Sale of prescription drugs in retail stores. Sale of prescription drugs in retail stores. Rite Aid Corporation & LaVerdiere’s Enterprises, Inc. Sale of prescription drugs in retail stores. Retail sale of pharmacy services to third-party payors. CVS Corporation & Revco D.S., Inc. Retail sale of pharmacy services to third-party payors. Sentinel Group, Inc. Provision of funeral services. Service Corporation International & Sentinel Group, Inc. Provision of funeral services. Provision of funerals. Provision of funerals and provision of perpetual care cemetery services. Provision of funerals, the provision of perpetual care cemetery services, and the provision of crematory services. Product market(s) The Loewen Group Inc. + Loewen Group International, Inc. & Garza Memorial Funeral Home, Inc. + Thomae-Garza Funeral Directors, Inc. (Texas) Provision of funerals. Funeral services; cemetery services. Shell Oil Co. & Texaco Inc. Refining, transportation, terminaling, wholesale sales, and retail sales of conventional unleaded gasoline, CARB-II gasoline; diesel fuel, kerosene jet fuel, and asphalt; and the transportation of undiluted heavy crude oil to the San Francisco, Cal. area. Terminaling of gasoline and other light petroleum products; wholesale sale of gasoline. Marketing of motor gasoline; refining and marketing of CARB gasoline; bidding for and refining of jet fuel for the U.S. Navy; terminaling of gasoline and other light petroleum products; pipeline transportation of light petroleum products; pipeline transportation of crude oil; refining and marketing of paraffinic base oil; production and sale of jet turbine oil. Excludes two divestiture orders in the funeral services industry. The Service Corporation International and LaGrone divestiture order (C-3959; 981-0108) announced for public comment on May 18, 2000, required a divestiture of a funeral home only if the corporation acquired a particular funeral home. As of June 30, 2002, the corporation had not acquired the funeral home; therefore, it had no obligation to divest the funeral home. The Loewen Group Inc., and Loewen Group International, Inc., and Heritage Family Funeral Services, Inc., divestiture order (C-3678; 931-0084) announced for public comment on May 14, 1996, required a divestiture of a funeral home. As of June 30, 2002, the merger had not taken place; therefore, it had no obligation to divest the funeral home. The Hart-Scott-Rodino Act (HSR Act) of 1976 reviews usually begin with the parties filing for a proposed merger or acquisition. The premerger notification provisions of the HSR Act require companies exceeding certain thresholds of company size and value of the transaction to notify the Federal Trade Commission (FTC) and the Department of Justice’s (DOJ) Antitrust Division of the proposed merger transaction, submit documents and other information to the agencies concerning the transaction, and refrain from consummating the transaction until a specified waiting period has expired or their request for early termination of the waiting period is granted by the appropriate antitrust enforcement agency. Table 8 shows that the number of transactions reported has more than doubled from 2,262 in fiscal year 1990 to 4,926 in fiscal year 2000, and the percent of early terminations granted has ranged from 65.8 percent to 81.3 percent over the same time period. In fiscal 2001, however, after the filing thresholds and filing fees were raised, there were only 2,376 filings. The HSR Act was amended in late 2000 to significantly increase the filing thresholds. FTC and the Antitrust Division also may become aware of mergers that are not subject to HSR Act requirements but that are potentially anticompetitive, using techniques such as (1) monitoring the trade press and Internet resources; (2) responding to and following up on case leads from congressional offices, other executive branch agencies, and state and local governments; and (3) encouraging consumers, businesses, and the bar to notify the FTC and the Antitrust Division of possibly anticompetitive mergers. FTC’s Performance Plan for fiscal years 2002 through 2003 states that after the reporting thresholds were increased on February 1, 2001, the FTC began to devote more effort to identifying mergers that may harm (or have harmed) competition but are not subject to HSR Act requirements. There are three tests, all of which must be met, in order for a transaction to be reportable. The first test is the commerce test, which requires that either the acquiring party or the acquired party must be engaged in commerce or in any activity affecting interstate commerce, as defined by Section 1 of the Clayton Act. The second test is the size-of-person test, which is based on the annual sales or assets of the merging parties. For the period we reviewed, one party to the transaction had to have annual sales or assets of at least $100 million and the other party of at least $10 million. The third test is the size-of-transaction test. Under this test, for the period we reviewed, as a result of such acquisition, the acquiring party had to hold (1) voting securities or assets worth in the aggregate more than $15 million or (2) voting securities that confer control (50 or more percent) of an issuer with annual sales of $25 million or more. FTC has 30 days (15 days for cash tender offers and bankruptcy filings) from the date the filing for the proposed acquisition is accepted to review the filing and to determine whether to seek additional information and documents from the merging parties and thereby extend the waiting period to enable further review. If FTC’s initial review does not indicate a need for further investigation, the merging parties can consummate the merger at the end of the waiting period, or when their request for early termination has been granted. If FTC’s initial review indicates a need for further review, a second request may be issued to the parties. A second request extends the waiting period for an additional 30 days (20 days prior to February 1, 2001, and 10 days for cash tender offers) after the parties have substantially complied with the request. The Assistant Directors in the Bureau of Competition offices have the authority to request clearance to investigate a merger under the FTC-DOJ liaison agreement and to open an initial phase investigation, even before an HSR Act filing is actually received. Obtaining clearance to investigate and open an initial phase investigation allows FTC staff to contact third parties as well as the parties to the merger. The Bureau Director has the authority to authorize a full phase investigation and to approve or reject staff’s recommendations to the FTC to authorize the use of compulsory process. The Commissioner assigned to the matter must approve the issuance of particular subpoenas to the merging parties or to others. While the full Commission acts on recommendations to authorize compulsory process, the Chairman approves, rejects, or modifies staff’s second request recommendations as approved by the Bureau Director. After FTC issues its second request to the merging parties, the parties begin assembling their responses. In the meanwhile, FTC staff are to employ other appropriate investigative techniques to obtain additional information relevant to determining the legality of the transaction and to evaluate the possible effects of the proposed transaction, such as concerns of third parties that may be affected by the merger. The third parties may include, among others, competitors, customers, and suppliers. In addition, when appropriate, staff may consult with state attorneys general, other U.S. governmental officials, and foreign antiturst authorities. When the parties comply with the second requests by submitting their responses to FTC, a second statutory waiting period is triggered. During that time, FTC staff are to review the material submitted by the parties, continue assembling third-party information, and complete staff’s evaluation and recommendations to FTC on whether enforcement action is warranted. According to FTC staff, parties often agree to extend the statutory time limits for FTC action to permit settlement negotiations. Antitrust practitioners and FTC staff told us that the merger investigation work generally is done with both the parties and FTC preparing for potential litigation. Table 9 shows the number of adjusted HSR Act transactions in which a second request could have been issued and the number of second requests issued by the FTC and the DOJ for fiscal years 1990 through 2000. According to FTC staff, generally at least two conditions are necessary for a merger to have a likely anticompetitive effect. The market must be substantially concentrated after the merger, with the merger substantially increasing concentration, and it must be difficult for new firms to enter the market in the near-term to provide effective competition. The majority of mergers that raise antitrust concerns are horizontal mergers. According to FTC staff and Antitrust Division officials, the joint Horizontal Merger Guidelines accurately outline how the agencies generally conduct their analysis of proposed mergers. The guidelines were originally developed by DOJ in 1968 and were updated in 1982 when FTC and DOJ issued separate statements and again in 1984 by the DOJ. The DOJ and the FTC issued joint Horizontal Merger Guidelines in 1992. In 1997, the agencies amended and expanded the efficiencies section of the guidelines. The unifying theme of the guidelines is that mergers should not be permitted to create or enhance market power or facilitate its exercise. The guidelines define a seller’s market power as the ability to profitably maintain selling prices above competitive levels for a significant period of time. Similarly, a buyer’s market power is defined as the ability to profitably maintain buying prices below competitive levels for a significant period of time. The Horizontal Merger Guidelines outline the five-step analytical process the FTC and the Antitrust Division use to determine whether a merger is likely to substantially lessen competition and, ultimately, whether to challenge a merger. FTC staff are to: (1) define the relevant product market and geographic market, identify the market participants, assign market shares, and assess whether increased market concentration from the proposed merger raises concern about potential adverse competitive effects, that is, increase in prices or a decrease in quality or output; (2) assess the potential competitive effects of the proposed merger, and the factors in addition to market concentration relevant to each; (3) assess whether entry into the market would be timely, likely, and sufficient either to deter or to counteract the competitive effects of concern; (4) assess any efficiency gains that cannot be reasonably achieved by the parties absent the proposed merger; and (5) if the parties have raised the “failing firm” as an affirmative defense, determine whether, but for the merger, either party to the transaction would be likely to fail and result in its assets exiting the market. According to FTC staff, while the analytical framework for merger analysis is generally the same regardless of the industry in which the merger is taking place, the particular facts of each merger are unique and outcome determinative. FTC staff told us that investigations are extraordinarily fact- specific. For example, supermarkets constitute a differentiated product market. No two supermarkets or chains are the same. Each store sells similar or the same products, but each store competes differently on price, level of service, and has different direct competitors. Similarly, FTC staff told us that the size of a geographic market also varies based on the facts of the case and the location of the merging parties’ and competitors’ assets. Once it is determined that a merger is likely to be anticompetitive, FTC may take action designed to prevent an anticompetitive result from a proposed merger by: (1) conducting successful litigation to block the merger; (2) negotiating a settlement to resolve anticompetitive aspects of the merger while allowing the underlying transaction to go forward; or (3) identifying antitrust concerns sufficient to cause the parties to abandon the transaction without court action. As FTC staff are reviewing the transaction, they are to invite the merging parties to discuss the competitive problem(s) and potential remedies to address the problem(s). FTC may decide that no remedy, short of blocking the transaction, will fully and effectively resolve the competitive concerns. FTC may also decide that to resolve the competitive concerns one of the merging parties must divest all of its assets in a single geographic market, or that a partial divestiture would be acceptable. Additionally, the anticompetitive effects of a merger may be remedied through contractual arrangements, such as the licensing of intellectual property or a short-term supply agreement. Alternatively, FTC may decide to use some form of behavioral relief, such as establishing firewalls in vertical mergers, to prevent the sharing of competitive information. Finally, the anticompetitive effects of some mergers may be addressed with a combination of these remedies. According to FTC staff, the remedy is based on the particular facts of the case. In response to FTC’s concerns, the merging parties might attempt to solve the competitive problem(s), abandon the transaction, or pursue the case through litigation. If an effective consent order cannot be negotiated and the merger has not yet been consummated, FTC staff are to recommend, in appropriate cases, that FTC authorize the filing of an action in federal district court for a preliminary injunction to stop the merger. If authorized, staff litigate preliminary injunction actions and appellate review proceedings. According to FTC staff, the team responsible for a merger investigation and consent negotiations typically includes the Bureau of Competition staff investigating the merger; staff from the Bureau of Competition’s Compliance Division, who are experienced in order drafting and enforcement; Bureau of Economics staff; and staff from FTC’s Office of the General Counsel. The team is to work together to develop and negotiate a consent agreement containing a proposed order with the merging parties. According to FTC staff, the parties may not agree with how the agency has defined the competitive harm, but they are aware of FTC’s concerns. FTC staff told us that generally the merging parties come forward with the first set of proposed remedies. FTC staff advised us that the merging parties often have extensive information about their industry and the effect of the proposed remedy. They also told us that FTC staff’s job is to make clear what their concerns are about the proposed transaction, to be available for constructive dialogue on how the problem can be adequately addressed, and to evaluate the adequacy of proposed remedies. According to a March 2000 speech by the then FTC General Counsel, FTC generally considers several factors in evaluating a proposed restructuring remedy. Specifically, she said that FTC staff generally consider whether the remedy is likely to protect, promote, or restore competition in the affected market. For example, in the case of a remedy involving divestiture of assets, FTC determines whether the buyer is obtaining sufficient assets, and the right kind of assets, to be able to create and operate a successful, competitive business, and whether that buyer has sufficient expertise, experience, incentives, and resources to accomplish its goal of maintaining or restoring competition in the relevant markets; whether efficiencies justify restructuring, rather than condemning, a the complexity of a proposed remedy; and whether its acceptance of a restructuring remedy might have implications for future matters. For example, according to the March 2000 speech, when FTC accepts a restructuring remedy, parties to subsequent proposed mergers tend to insist on a similar resolution to the competitive problems raised by their transactions, although FTC may view the matters as distinguishable on factual or economic grounds. FTC staff negotiate a proposed consent agreement with the merging parties. Once the merging parties sign the proposed consent agreement and the Director of the Bureau of Competition approves the execution of the draft consent agreement, it is reviewed by the full Commission, which determines if the agency will accept the agreement for public comment. For an agreement containing a proposed order to be approved subject to public comment, a majority of the Commissioners must vote to accept the agreement. In general, if the Commission votes to accept the agreement subject to public comment, the HSR Act waiting period is terminated, and the parties can consummate the merger. In consent agreements that require up-front buyers, the divestitures also may be consummated, but the divestiture must include a clause requiring its rescission if FTC does not give final approval to the proposed consent order. The merging parties generally can consummate the transaction as soon as they are notified of FTC action. This notification occurs before the public announcement, which generally occurs within a few days of the Commission’s vote to accept the agreement. Although not statutorily required, FTC places the consent agreement along with an analysis to aid public comment on the public record, generally for a 30-day comment period, in which the public is invited to comment on the merger and the proposed relief. At end of the comment period, FTC staff are to analyze the comments received and forward a second set of recommendations to the Commission on whether the agency should issue the final consent order. According to FTC staff, typically the Commission accepts the consent agreement and issues a complaint and decision and order. They told us that it is highly unusual for FTC to receive new antitrust-related information during the public comment period that staff had not already considered as part of their investigation. In addition, they said that after considering public comments, FTC has rarely modified a consent order because in most instances, the complaint and order, as proposed, provided the most appropriate relief. FTC staff have said that every order has the same goal: to preserve fully the existing competition in the relevant market or markets. According to FTC staff, while developing appropriate merger remedies is a very fact-specific process, many provisions have been developed over the years that appear in almost every order. These required provisions are readily apparent from even a cursory review of merger consent orders posted on FTC’s Web site. In March 2002, FTC’s Bureau of Competition posted on its Web site a document that provides responses to frequently asked questions about merger consent order provisions. This document as well as a review of FTC orders, other public documents prepared by FTC staff, and our discussions with FTC staff provide some insight into FTC’s consent order provisions. Most orders relating to a merger will require a divestiture—the selling of a business or assets by one or both of the merging parties—in order to maintain or restore the level of competition that existed before the merger. During fiscal years 1990 through 2000, FTC announced 192 proposed consent orders involving mergers for public comment, of which 153 required divestitures. FTC has the authority to decide the extent of the divestiture. A divestiture can range from a partial divestiture to a divestiture of an on-going business. Table 10 shows that most (about 80 percent) of the consent orders involving mergers that FTC announced for public comment during fiscal years 1990 through 2000 required divestiture of assets. (Appendix V provides information on the 153 consent orders requiring divestitures that FTC announced for public comment during fiscal years 1990 through 2000.) According to FTC staff, FTC generally requires divestitures in the geographic and product markets where the merging parties have competitive overlaps and FTC has competitive concerns. However, it is within FTC’s power to require divestiture of a greater set of assets than those which participate in the overlap markets in order to effectively replace competition. According to FTC staff, sometimes the buyers of the divested assets will need other ancillary assets in order to effectively restore competition. Without these ancillary assets the buyer will not be able to replicate the economies of scale of the firm that has been acquired. In other cases, these additional assets will be necessary to give the buyer both the incentive and ability to fully restore competition. FTC staff told us that the merging parties generally are the first to respond to staff’s concerns by proposing appropriate assets to be divested. According to FTC staff, an acceptable divestiture package is one that maintains or restores competition in the relevant market. FTC staff have said that the divestiture of an entire business of either the acquired or acquiring firm relating to the markets in which there is concern about anticompetitive effects, is most likely to maintain or restore competition in the relevant market, and thus will usually be an acceptable divestiture package. FTC has issued orders that require divestitures of less than the entire business operating in, or producing for, the relevant market. In those cases, FTC concluded that the assets to be divested were sufficient to allow the buyer of the assets to begin to compete in the market immediately and to remedy the likely or actual anticompetitive effects of the challenged acquisition. According to FTC staff, the burden is on the merging parties to provide concrete and convincing evidence that the asset package is sufficient to allow the proposed buyer, whether large or small, to operate in a manner that maintains or restores competition in the relevant market. FTC also generally requires that a divestiture be absolute. This means that the merging parties are to have no continuing ties to the divested business or assets, no continuing relationship with the buyer, and no financial stake in the buyer’s success. According to FTC staff, divestiture proposals in which the buyer intends to rely on the merging parties to finance the divestiture, or where the proposal includes performance payments by the buyer generally have been rejected. For consent orders in which FTC has required a divestiture(s), and the buyer(s) of the divested asset(s) is not identified in the proposed order, the merging parties must submit an application to FTC requesting approval to divest the asset(s) to a proposed buyer(s) and await FTC approval before consummating the divestiture. According to FTC staff, the more information the divestiture application contains about the transaction and the proposed buyer, the more likely the approval can be obtained quickly. The application must show how the proposed divestiture will remedy the competitive problem identified in FTC’s complaint and restore competition. Additionally, there must be a final signed contract for a divestiture application to be sufficient. The contract should conform to the requirements of the consent order or, if it does not, explain how the respondent will satisfy the order’s requirements. The divestiture proposal should demonstrate that the proposed buyer will be an effective competitor after the divestiture. The proposed buyers of the divested assets are to provide directly to the FTC information on their (1) financial capability, including the financing in place both to pay for the acquisition and to fund working capital and other needs associated with an on-going business; (2) technical and management skills; and (3) business plans and other evidence of the proposed buyer’s intention and ability to compete. In evaluating whether a proposed buyer has the financial resources to remain a vigorous competitor in the market, FTC staff are to examine the proposed buyer’s commitment to remain in the market by analyzing its past operations and business plans as well as its future business plans for the divested assets. The staff also will likely talk with industry members familiar with the proposed buyer, such as competitors, suppliers, and customers. The staff also are to evaluate the proposed buyer’s experience and expertise to operate effectively in the market. They also may examine information on the proposed buyer’s debt structure to determine whether the transaction is very risky for the company. However, information on debt structure is not collected for each buyer. FTC staff told us that the information provided by each buyer varies on a case-by-case basis. FTC does not use a standard form or checklist to collect information from buyers because the facts of each case and, thus, the requirements for information from the buyers differ. FTC frequently requires that the merging parties find an acceptable buyer(s) for the assets to be divested and that it execute an acceptable purchase agreement and all the necessary ancillary agreements with the buyer(s) before the FTC accepts the proposed consent order for public comment. FTC staff refers to such a buyer as an up-front buyer or buyer up-front. According to FTC staff, an up-front buyer may be the best way to ensure a successful divestiture in the supermarket industry because the approach enables staff to evaluate the marketability of the divestiture assets with more concrete evidence and better determine whether, among other things, there is a viable buyer(s) for the proposed divestiture assets. Additionally, FTC staff told us that an up-front buyer reduces the amount of time needed for the assets to be divested because a buyer can be identified before the merger transaction occurs—a factor in supermarket mergers and those of other retail operations, where assets may quickly deteriorate during the search for a buyer; and generally increases the likelihood that a buyer will restore the competition that otherwise would be lost through the merger. To preserve FTC’s ability to reject an up-front buyer following the public comment period, consent orders accepted subject to public comment require the merging parties to include a rescission provision in any divestiture contracts in which closing on the divestiture will occur before final FTC approval of the consent order. According to FTC staff, as of June 30, 2002, FTC had not ordered rescission of an up-front divestiture. FTC staff told us that there have been instances in which the buyers will purchase the assets only after the final order has been approved. Additionally, they said that one FTC order, not in the retail industry, specified that the assets could not be divested until the order was final. Amount of Time to Divest Assets Divestiture orders specify the timeframe in which the assets must be divested. The merging parties must find a buyer(s), negotiate a contract(s), submit the contract(s) to the FTC for its approval, and complete the divestiture(s) within that time. If there is an up-front buyer, the divestiture is required almost immediately upon consummation of the subject merger and FTC approves the up-front buyer when it issues the final consent order. In other cases, however, the merging parties must submit their divestiture application(s) to FTC early enough to allow for the 30-day public comment period required by FTC rules. According to FTC staff, to satisfy its obligation to divest by the date required in the order, the merging parties have to actually consummate the sale by that date. Executing an agreement or filing an application for the FTC’s approval by that date does not satisfy the obligation to divest by that date. FTC may appoint trustees or independent auditors in three situations. First, most orders authorize FTC to appoint a trustee to divest the assets if the merging parties fail to divest them within the time frame required by the consent order. In addition, where there is a hold separate order, FTC appoints an independent auditor to ensure the independence of the assets to be divested when they must be operated separately from the other assets of the merging parties until they are sold. Finally, where there is to be a short-term continuing relationship after the divestiture, FTC may appoint a trustee to ensure that the merging parties fully perform their responsibilities by the order. Some orders authorize a divestiture trustee to divest a different or larger package of assets, referred to as a crown jewel, if the assets are not divested on time. According to FTC staff, a crown jewel is used where there is risk that, if the respondent fails to divest the original divestiture package on time or if the original divestiture does not take place for any reason, a divestiture trustee may be needed to divest an expanded or alternative package of assets to accomplish the divestiture remedy. FTC staff have said that a crown jewel provision may be particularly valuable when there are some uncertainties about the marketability or viability of the initial divestiture package. FTC staff also have said that a crown jewel provision increases the incentive for the merging parties to accomplish the divestiture within the time required by FTC’s divestiture order, and it provides a bigger, and presumably more attractive, package for the trustee in the event the merging parties are unsuccessful in divesting the assets. Asset Maintenance and Hold Separate Agreements A consent order may include an asset maintenance agreement that requires the merging parties to maintain the viability, marketability, and competitiveness of the assets. FTC also has included hold-separate agreements to protect all the assets to be divested to prevent interim competitive harm and to preserve the viability and competitiveness of the assets pending divestiture. According to FTC staff, a hold separate agreement is designed to keep the divestiture assets from being intermingled with the divesting party’s other assets pending divestiture by requiring the divestiture assets to be operated separately from and independently of the remaining business. Additionally, a “hold separate” agreement prevents the transfer of competitively sensitive information and, by taking the assets out of the hands of the divesting party, better protects the assets from intentional or unintentional physical or intangible deterioration that would affect their ability to be operated in a manner that maintains or restores competition. According to FTC staff, hold-separate agreements are generally required if there is no up front-buyer. A provision that required the merging parties to obtain FTC’s prior approval for future transactions in the same product and geographic market(s), usually for a period of 10 years was among the commonly included provisions. However, on June 21, 1995, FTC issued a new policy regarding the use of prior approval requirements in FTC orders. Previously, whenever FTC found reason to believe that the respondent had attempted (or completed) a merger transaction that was anticompetitive, a prior approval provision was included in the order. In some cases, according to FTC staff, FTC also required prior notice of transactions that would not be reportable under the HSR Act. According to FTC staff, under the new policy that is no longer the case. The general rule now is that FTC will not use prior approval or prior notice requirements except in special cases. In cases in which the merging parties have a post-order divestiture obligation, the merging parties generally are required to keep FTC informed of their divestiture efforts every 30 or 60 days by submitting verified written compliance reports. To the extent there are obligations in the order beyond the divestiture obligation, the merging parties are usually required to submit verified written annual reports to FTC on their continued compliance with those obligations. In the case of a post-order divestiture requirement, when the merging parties file an application seeking approval of a particular divestiture, that application is also placed on the public record for a 30-day comment period. Again, staff must analyze the comments received and make recommendations to FTC. In some cases, buyers are identified before FTC accepts the consent agreement for public comment. If the buyer is identified early in the consent process and is identified in the proposed consent order, there is no separate comment period because the public is commenting on both the proposed order and divestiture at the same time. Consequently, according to FTC staff, the public comment serves the twin purposes of giving the public a chance to comment on the substance of relief obtained as well as the opportunity to comment on the suitability of the proposed buyers. According to FTC staff, FTC has negotiated with merging parties for divestitures of additional assets based on comments received during the public comment period. They also told us that after considering public comments, FTC has rarely modified a consent order accepted subject to public comment or disapproved an up-front divestiture. FTC can reopen and modify a final consent order when a request to reopen identifies significant changes in circumstances and shows that the changes eliminate the need for the order or make continued application of it inequitable or harmful to competition. FTC also may modify an order when, although changed circumstances would not require reopening, it determines that the public interest so requires. According to standard language in FTC orders reopening and modifying an order, FTC will balance the reasons favoring the requested modification against any reasons not to make the modification. FTC also considers whether the particular modification sought is appropriate to remedy the identified harm. Product market(s) Retail sale of food and grocery items in supermarkets. The Vons Companies, Inc. & Williams Bros. Markets, Inc. Retail sale and distribution of food and grocery items in supermarkets. Red Apple Companies, Inc. + John A. Catsimatidis + Supermarket Acquisition Corp. + Designcraft Industries, Inc. & Sloan’s Supermarkets, Inc. Retail sale of food and grocery products in supermarkets. Retail sale of food and grocery products in supermarkets. Schnuck Markets, Inc. & National Holdings, Inc. Retail sale of food and grocery products in supermarkets. Schwegmann Giant Super Markets, Inc. & National Holdings, Inc. Retail sale of food and grocery products in supermarkets. The Stop and Shop Companies, Inc. & Purity Supreme, Inc. Retail sale of food and grocery products in supermarkets. Koninklijke Ahold nv + Ahold USA, Inc. & The Stop & Shop Companies, Inc. Retail sale of food and grocery products in supermarkets. Jitney-Jungle Stores of America, Inc. + Delta Acquisition Corporation & Delchamps, Inc. Retail sale of food and grocery products in supermarkets. Albertson’s, Inc. + Locomotive Acquisition Corporation & Buttrey Food and Drug Store Company, Inc. Retail sale of food and grocery products in supermarkets. Koninklijke Ahold nv & Giant Food, Inc. Retail sale of food and grocery products in supermarkets. Product market(s) The Kroger Co. & Fred Meyer, Inc. Retail sale of food and grocery products in supermarkets. Retail sale of food and grocery products in supermarkets. Shaw’s Supermarkets, Inc. + J Sainsbury plc & Star Markets Holdings, Inc. Retail sale of food and grocery products in supermarkets. The Kroger Co. & The John C. Groub Company, Inc. Retail sale of food and grocery products in supermarkets. Delhaize America, Inc. (Food Lion) + Etablissements Delhaize Freres et Cie "Le Lion" S.A. & Hannaford Bros. Co. Retail sale of food and grocery products in supermarkets. Manufacture and sale of propylene oxide; manufacture and sale of urethane polyether polyol; manufacture and sale of propylene glycol . Licensing of polypropylene technology; polypropylene technology; licensing, production, and sales of high-yield/high-specificity polypropylene catalysts and catalyst technology; production and sales of polypropylene resin; production and sales of polypropylene impact copolymer resin. NGC Corporation & Chevron U.S.A., Inc. Fractionation of natural gas liquids. Product market(s) SoftSearch Holdings, Inc.+ Dwight’s Energydata, Inc. & Petroleum Information Corporation + GeoQuest International Holdings, Inc. Sale or licensing of well data and production data. Natural gas gathering services. Shell Oil Co. & Texaco, Inc. Refining, transportation, terminaling, wholesale sales, and retail sales of conventional unleaded gasoline, CARB-II gasoline (specially formulated gasoline required in California), diesel fuel, kerosene jet fuel, and asphalt; and the transportation of undiluted heavy crude oil to the San Francisco, Cal. area. Mining, production, and sale of coal; wholesale electricity sales. The Williams Companies, Inc. & MAPCO, Inc. Transportation by pipeline and terminaling of propane; transportation by pipeline of raw mix. Development, manufacture, marketing, and sale of viscosity index improver or viscosity modifiers for motor oil for automobiles and trucks. Natural gas gathering services. Product market(s) Terminaling of gasoline and other light petroleum products; wholesale sale of gasoline. El Paso Energy Corporation & Sonat, Inc. Transportation of natural gas out of producing fields; transportation of natural gas into gas consuming areas. Generation of electric power and the distribution of natural gas. Marketing of motor gasoline; refining and marketing of “CARB” gasoline (specially formulated gasoline required in California); bidding for and refining of jet fuel for the U.S. Navy; terminaling of gasoline and other light petroleum products; pipeline transportation of light petroleum products; pipeline transportation of crude oil; refining and marketing of paraffinic base oil; production and sale of jet turbine oil. Duke Energy Corporation & Phillips Petroleum Company & Duke Energy Field Services L.L.C. Natural gas gathering. Product market(s) Production, sale, and delivery of Alaska North Slope crude oil; production, sale, and delivery of crude oil used by targeted West Coast refiners; production, sale, and delivery of all crude oil used by refiners on the West Coast; purchase of exploration rights; pipeline transportation of Alaska North Slope crude oil; development for commercial sale of natural gas; oil pipeline and storage services into and in Cushing. Product market(s) Amersham International plc & Medi-Physics, Inc. Formulating, manufacturing, marketing, and selling radiopharmaceutical brain perfusion imaging agents for use with Single Positron Emission Tomography equipment. E-Z-EM, Inc. & Lafayette Pharmacal, Inc. Formulating, manufacturing, marketing, and selling barium diagnostic products and related accessories. Institut Merieux S.A. & Connaught BioSciences, Inc. Rabies-vaccines. Roche Holdings, Inc. + Hoffman- La Roche Inc. + Roche Holdings Ltd. & Genentech, Ltd. Research and development and production and manufacture of: vitamin C; therapeutics for treatment of human growth hormone deficiency or other short stature deficiency; and CD4- based therapeutics for the treatment of AIDS and HIV infection. Columbia Hospital Corporation & Galen Health Care, Inc. Production and sale of acute care inpatient hospital services. Sale, rental, or lease of oxygen systems. Sale, rental, or lease of oxygen systems. Sale, rental or lease of oxygen systems. Production and sale of acute care inpatient hospital services. Sale of prescription drugs in retail stores. Product market(s) Production and sale of acute care inpatient hospital services. Revco D.S., Inc. & Hook-SupeRx, Inc. Sale of prescription drugs in retail stores. Manufacture and sale of drugs of abuse reagent products. Rite Aid Corporation & LaVerdiere’s Enterprises, Inc. Sale of prescription drugs in retail stores. Columbia/HCA Healthcare Corporation & Medical Care America, Inc. Production and sale of outpatient surgery services. Manufacture and sale of combined tetanus and diphtheria vaccine (adult Td); manufacture and sale of combined diphtheria and tetanus vaccine (pediatric DT); manufacture and sale of tetanus vaccine (tetanus toxoid); research and development of vaccine against Rotavirus infection in humans; research, development, production, and sale of cytokines for white blood cell and platelet restoration. Wright Medical Technology, Inc. + Kidd, Kamm Equity Partners, L.P. + Kidd, Kamm Investments, L.P. + Kidd, Kamm Investments, Inc. & Orthomet, Inc. Manufacture and sale of orthopedic implants used or intended for use in the human hand approved by the Federal Drug Administration; research and development of orthopedic implants used or intended for use in the human hand. Product market(s) Healthsouth Rehabilitation Corporation & ReLife, Inc. Production and sale by rehabilitation hospital facilities of comprehensive, acute inpatient medical rehabilitation services. Boston Scientific Corporation & Cardiovascular Imaging Systems, Inc. & SCIMED Life Systems, Inc. Research, development, manufacture, and sale of intravascular ultrasound catheters. Research and development of noninjectable 5HT-ID agonists. Production and sale of acute care in-patient hospital services. Production and sale of psychiatric hospital services. Hoechst AG & Marion Merrell Dow, Inc. Research, development, manufacture, and sale of: once a day diltiazem; oral dosage forms of mesalamine; rifampin; drugs approved by FDA for treatment of intermittent claudication. Research, development, manufacture, and sale of Topoisomerase I inhibitors for the treatment of colorectal cancer. Manufacture and sale of neurological shunts. Fresenius AG + Fresenius USA, Inc. & National Medical Care, Inc. Hemodialysis concentrate. Retail sale of pharmacy services to third-party payors. Product market(s) Gene therapy technology and research and development of gene therapies; research, development, manufacture, and sale of flea control products; research, development, manufacture, and sale of corn herbicide. Research, development, manufacture, and sale of Factor VIII Inhibitor Treatments; research and development, manufacture, and sale of Fibrin Sealant. Production and sale of acute care inpatient hospital services. American Home Products Corporation & Solvay, S.A. Research, development, manufacture, and sale of canine lyme, canine corona virus, and feline leukemia vaccines. CVS Corporation & Revco D.S., Inc. Retail sale of pharmacy services to third-party payors. Research, development, manufacture, and sale of Cardiac Thrombolytic agents; research, development, manufacture, and sale of drug abuse testing reagents used in workplace testing. Medtronic, Inc. & Avecor Cardiovascular, Inc. Research, development, manufacture, and sale of non-occlusive arterial pumps. Manufacture and sale of Long-Acting Local Anesthetics. Product market(s) Research, development, manufacture, and sale of heart-lung machines. Hoechst AG + RhÔne-Poulenc S.A. & Aventis S.A. Research, development, manufacture, and sale of direct thrombin inhibitors; manufacture, marketing, and sale of cellulose acetate. Research, development, manufacture, and sale of over the counter pediculicides; research, development, manufacture, and sale of selective serotonin reuptake inhibitor/selective norepinephrinr reuptake inhibitors drugs for treatment of depression; research, development, manufacture, and sale of drugs for the treatment of Alzheimer’s disease; research, development, manufacture, and sale of EGFr-tk inhibitors for the treatment of cancer. Litton Industries, Inc. & PRC, Inc. Research, development, manufacture, and sale of Aegis destroyers; Systems Engineering and Technical Assistance Services. Product market(s) Research, development, manufacture, and sale of air traffic control systems; provision of Systems Engineering and Technical Assistance services; research, development, manufacture, and sale of commercial low earth orbit satellites; research, development, manufacture, and sale of commercial geosynchronous earth orbit satellites; research, development, manufacture, and sale of military aircraft; research, development, manufacture, and sale of NITE Hawk systems (same as nonpublic military aircraft information); research, development, manufacture, and sale of simulation and training systems; research, development, manufacture, and sale of electronic countermeasures; research, development, manufacture, and sale of mission computers; research, development, manufacture, and sale of unmanned aerial vehicles; research, development, manufacture, and sale of integrated communications systems. Product market(s) Research, development, manufacture, and sale of high altitude endurance unmanned air vehicles, research, development, manufacture, and sale of space launch vehicles, research, development, manufacture, and sale of space launch vehicle propulsion systems. TRW Inc. & BDM International, Inc. Research, development, manufacture, and sale of a ballistic missile defense system; Systems Engineering and Technical Assistance Services. Tele-communications, Inc., and Liberty Media Corporation & QVC Network, Inc. Subscription television program distribution to consumers and/or in cable premium movie channels. Development and sale of professional illustration software for use on Apple Macintosh and Power Macintosh computers. Multichannel video programming. Sale of cable television program services to Multichannel Video Programming Distributors; sale of cable television programming services to households. Product market(s) Automatic Data Processing, Inc., and AutoInfo, Inc. Integrated group of information products and services that form the complete salvage yard information systems network, consisting of an interchange integrated with yard management systems and electronic communications systems; development and sale of automotive parts and assemblies interchanges; development and sale of yard management systems integrated with interchange; development and sale of electronic communications systems used by salvage yards to locate parts through searches of a central database of parts; collection and provision of salvage yard inventory data to customers who provide such data as a part of estimating products sold to insurance companies. Cablevision Systems Corporation & Tele-Communications, Inc. Distribution of multichannel video programming by cable television. Manufacture and sale of high-performance, general purpose microprocessors capable of running Windows NT; manufacture and sale of general-purpose microprocessors; design and development of high- performance, general- purpose microprocessors. Product market(s) Emerson Electric Company + Emerson Power Transmission Co. & McGill Manufacturing Company, Inc. Production and distribution of mounted ball bearings. Rug cleaning products business. T&N plc & J. P. Industries, Inc. Manufacture and sale of thinwall engine bearings; design, manufacture, and sale of tri-metal heavywall engine bearings. Manufacture and sale of portland cement. Mannesmann, AG + Mannesmann Capital Corporation & Rapistan Corp. + Lear Siegler Holdings Corp. Manufacture and sale of high speed, light-to-medium duty unit handling roller and belt conveyor systems for distribution end users. Acrylic emulsion polymers for exterior architectural coatings. Premium silver alloy business. Cooper Industries, Inc.+ Cooper (U.K.) Limited & The Fusegear Group of BTR p.l .c. Low voltage industrial fuse market. Consol, Inc. + E.I. du Pont de Nemours + RWE Aktiengesellshaft & Island Creek Coal, Inc. Coal export terminal services. Manufacture and sale of acrylic plastic. Manufacture and sale of coating resins and other markets contained therein. Product market(s) Textron, Inc. & Avdel PLC and Banner Industries, Inc. Design, manufacture, and sale of aerospace blind rivets; design, manufacture, and sale of nonaerospace structural blind rivets. Alvey Holdings, Inc. + Alvey, Inc. & White Storage & Retrieval Systems, Inc. Manufacture and sale of horizontal carousels. The Dow Chemical Company + Marion Merrell Dow Inc. & Rugby- Darby Group Companies, Inc. Dicyclomine hydrochloride capsules and tablets. Manufacture and sale of aluminum polyester powder. Manufacture, distribution, and sale of turbomolecular pumps; manufacture, distribution, and sale of compact disc metallizers. Reckitt & Colman plc & L&F Products, Inc. Development, manufacture, marketing, and sale for resale of carpet deodorizer products. Mustad Connecticut Inc. + Mustad International Group NV & Cooper Horseshoe Nail Co., Ltd. Manufacture and sale of rolled horseshoe nails. Praxair, Inc. & CBI Industries, Inc. Manufacture and sale of merchant argon, merchant oxygen, merchant nitrogen. Industrial power sources, industrial engine drives, battery chargers, and aircraft ground power units. Fused cast refractories; hot surface igniters; silicon carbide refractory bricks. Product market(s) Electronic automotive parts catalogs; management information systems integrated with an electronic catalog. Mahle GmbH + Mahle, Inc. & Metal Leve, S.A. Research, development, design, production, and sale of articulate pistons; research, development, design, production, and sale of large bore two-piece pistons. Large Welded Aluminum Tubes; small Welded Aluminum Tubes. Research, development, manufacture, and sale of chelants. S.C. Johnson & Son, Inc. & Dow Brands Inc. + Dow Brands L.P. + Dow Brands Canada Inc. Research, development, manufacture, and sale of soil and stain removers; research, development, manufacture, and sale of glass cleaner product. Federal-Mogul Corporation & T&N p.l.c. Development, manufacture, and sale of thinwall bearings; development, manufacture, and sale of light duty engine bearings; development, manufacture, and sale of heavy duty engine bearings; m- manufacture and sale of aftermarket bearings. Global Industrial Technologies, Inc. & AP Green Industries, Inc. Glass-furnace silica refractories. Nortek, Inc. + NTK Sub, Inc. & NuTone, Inc. Manufacture, production, and sale of hard-wired residential intercoms. Product market(s) Research, development, manufacture, and sale of water-based polymers. Smelting and refining of lead; recycling of junkers. Development, manufacture, and sale of Titanium Aerospace Investment Cast Components; development, manufacture, and sale of Large Stainless Steel Aerospace Investment Cast Components; development, manufacture, and sale of Large Nickel-based Superalloy Aerospace Investment Cast Components. Reckitt + Colman plc & Benckiser N.V. Research, development, formulation, manufacture, marketing, and sale of hard surface bathroom cleaners; research, development, formulation, manufacture, marketing, and sale of fine fabric wash products. MacDermid, Inc. & Polyfibron Techonologies, Inc. Research, development, manufacture, and sale of liquid photopolymers; research, development, and sale of solid sheet photopolymers. Product market(s) RHI AG & Global Industrial Technologies, Inc. Research, development, manufacture, and sale of magnesia-carbon refractory bricks for basic oxygen furnaces; research, development, manufacture, and sale of magnesia- carbon bricks for electric arc furnaces; research, development, manufacture, and sale of magnesia- carbon refractory bricks for basic oxygen furnaces steel ladles; research, development, manufacture, and sale of magnesia chrome refractory bricks for steel degassers; research, development, manufacture, and sale of high-alumina refractory bricks for basic oxygen furnaces steel ladles; research, development, manufacture, and sale of high-alumina refractory bricks for torpedo cars. Manufacture, marketing, and sale of pure phosphoric acid. Production, distribution, and wholesale sale of nitrogen- based fertilizer urea; production, distribution, and wholesale sale of Urea Ammonia Nitrogen 32 solution; production, distribution, and wholesale sale of ammonia. Product market(s) Branded carbonated soft drinks. Production and sale of bulk bakery wheat flour. Rhone-Poulenc & Marschall Dairy Products + Miles Inc. Manufacture and sale of dairy cultures. Branded carbonated soft drinks; all carbonated soft drinks. American Stair-Glide Corporation + Access Industries, Inc. & The Cheney Company, Inc. Manufacture and sale of curved stairway lifts; manufacture and sale of straight stairway lifts; manufacture and sale of vertical wheelchair lifts. Production and sale of title plant information; production and sale back plant information. High-purity alcohol process alumina. Sentinel Group, Inc. Service Corporation International & Sentinel Group, Inc. Provision of funerals. Continuous action air freshener products business and the instant air freshener products business; furniture care products business. Product market(s) Residential nonselective herbicide market. McCormick & Company, Inc. & Haas Foods, Inc. + John I. Hass, Inc. Sara Lee Corporation + Kiwi Brands Inc. & Knomark, Inc. + Reckitt & Colman p.l.c. Sale of chemical shoe care products used in the maintenance, cleaning, and protection of shoes. First Data Corporation and Western Union Financial Services, Inc. Consumer money transfer services. B.A.T. Industries p.l.c. & Brown and Williamson Tobacco Corporation & American Tobacco Company & American Brands, Inc. Manufacture and sale of cigarettes for U.S. consumption. Provision of funerals and provision of perpetual care cemetery services. Scotts Company & Stern's Miracle-Gro Products, Inc. Water soluble fertilizer for consumer use. Sale of consumer money wire transfer services. Provision of funerals, the provision of perpetual care cemetery services, and the provision of crematory services. Manufacture and distribute collagen sausage casings. The Loewen Group Inc. + Loewen Group International, Inc. & Heritage Family Funeral Services, Inc. (Virginia) Provision of funerals. Product market(s) The Loewen Group Inc. + Loewen Group International, Inc. & Garza Memorial Funeral Home, Inc. + Thomae-Garza Funeral Directors, Inc. (Texas) Provision of funerals. Wesley-Jessen Corporation & Pilkington Barnes Hind International, Inc. Manufacture and sale of opaque contact lenses. General Mills, Inc. & Ralcorp Holdings, Inc. Sale of branded and private label ready-to-eat cereals. Premium scotch; premium gin. Sale of timeshare exchange services. Production and sale of title plant services. ABB AB + ABB AG & Elsag Bailey Process Automation N.V. Manufacture and sale of Process Gas Chromatograph; manufacture and sale of Process Mass Spectrometer. Funeral services; cemetery services. Provision of fleet card services to over the road trucking companies; development, manufacture, and sale of truck stop fuel desk automation services. VNU N.V. & Nielsen Media Research, Inc. Advertising expenditure measurement service. Provision of title information services. Product market(s) FMC Corporation & Solutia, Inc. Manufacture, marketing, and sale of pure phosphoric acid; manufacture, marketing, and sale of phosphorus pentasulfide. Funeral services. In recent years, antitrust practitioners have raised concerns regarding certain FTC’s divestiture practices, particularly its preference for up-front buyers. Some practitioners we spoke with also raised concerns that the basis for FTC’s merger remedy practices is not always clear and that the extent of information provided to them on the rationale for FTC’s preferred remedies varies by staff. While interviewees had differing opinions on whether merger remedy guidelines are needed, the commenting practitioners generally agreed that there is a need for greater transparency in the merger remedy process. “First, it can delay consummation of the transaction while the parties find, and obtain approval for, a buyer. Assuming the transaction has procompetitive components, this delay may have costs in the market. Second, an upfront buyer requirement can lead to strategic behavior by the potential purchasers who are given greater leverage in the negotiations. This may skew the bidding process in a way that is inconsistent with the agencies’ [FTC and DOJ] goal of preserving competition in the marketplace.” The American Bar Association (ABA) also has questioned FTC’s up-front buyer approach. In 2001, ABA’s Task Force on Federal Antitrust Agencies of the Section of Antitrust Law issued a report on the state of federal enforcement of the antitrust laws. According to the report, there is some uncertainty in the legal and business communities concerning the circumstances under which FTC policy requires an up-front buyer. The report also stated that “The policy needs to be clarified. In addition, the circumstances when up- front buyers should be required should be carefully examined to balance the need to preserve effective competition with the imposition of unnecessary costs on the merging parties.” The lack of transparency in FTC’s merger remedy process was another area of concern. Antitrust practitioners we spoke with told us that while FTC speeches, public workshops, and the 1999 divestiture study provide some information on FTC’s divestiture practices, the basis for the practices is not always clear, and the extent of information provided to the merging parties on the rationale for FTC’s preferred remedy varies by staff. In the 2001 ABA Task Force report, ABA stated that “…press releases issued by FTC explaining its enforcement actions merely highlight the remedies achieved and provide conclusionary reviews of the competitive concerns, but generally do not meaningfully explain the market context, specific competitive concerns, and the mode of analyzing competitive effects.” Similarly, in a recent article entitled Toward Guidelines for Merger Remedies, the President of American Antitrust Institute (AAI) said that FTC's analysis to aid public comment documents do not offer insight into trade-offs accepted during negotiations or reasons for accepting settlements that may differ in important respects from settlements in other apparently similar cases. FTC staff acknowledged that FTC does not always provide a detailed rationale for its divestiture approaches. They told us that because FTC’s decisions are largely tied to companies’ trade secret information, which FTC is statutorily prevented from disclosing to the public, FTC can provide to the public only limited information on the basis for its decisions. ABA and AAI have concurred that FTC’s lack of explanation of the basis for the remedies included in a divestiture order is due in part to the proprietary nature of the information. Nonetheless, in the AAI article on merger remedies, the President of AAI said that FTC has overused confidentiality as an excuse for not providing more transparency in the merger remedy process. In fact, both ABA and AAI have said that there needs to be a more transparent approach to the remedy phase of the merger review process. Specifically, ABA has urged FTC to make greater efforts to more meaningfully explain the factors that give rise to competitive concerns, the type of evidence viewed as relevant, the econometric analysis used (if any), and other key considerations that led to the decision to bring a complaint or enter into a consent order. AAI also has recommended that FTC (1) study its enforcement patterns, (2) derive best practices, (3) formalize rules, (4) provide transparency so that the public can understand and evaluate decisions being made, and (5) conduct regular post hoc evaluations to determine how well a program is working. FTC staff told us that FTC does in fact value transparency and has taken steps to provide the business community with information on FTC’s merger review and merger remedy processes. In a June 10, 2002, speech commemorating the twentieth anniversary of the merger guidelines, which apply to the merger investigation phase of the merger review process, FTC’s Chairman said that “… the Guidelines demonstrated the value of transparency—having public antitrust authorities clearly state their enforcement attentions , even at the risk of relinquishing the capacity to employ enforcement approaches that well-specified guidelines might disavow or disfavor. Experience with the 1982 Merger Guidelines has shown how the quality of policy improve when public officials specify clearly the bases on which they exercise their authority.” There has been some discussion among antitrust practitioners about whether merger-remedy guidelines are needed. Some practitioners have said that merger-remedy guidelines would provide a more structured, coherent, and transparent approach to the remedy phase of the merger review process. In fact, AAI prepared a proposal for merger-remedy guidelines and presented it to FTC in March 2002 for review and discussion at FTC’s merger-remedies workshops. Other practitioners have called for “practice guides” that could be used to educate antitrust practitioners and FTC staff about when and why to use certain divestiture practices, but would allow for greater flexibility in structuring a remedy. Regardless of whether they believed that guidelines are needed, they all agreed that there is a need for greater transparency in the merger-remedy process. Conversely, FTC staff told us that merger-remedy guidelines are not needed for the following reasons: Because each case is unique and fact-based they draw on their past experiences and advice from experienced senior staff, rather than developing written policies and procedures to guide staff in fashioning merger remedies. The merging parties have very sophisticated antitrust counsel who are aware of how to structure remedies in order to obtain FTC approval. FTC speeches, workshops, consent orders, and public documents provide antitrust practitioners and the business community with information on FTC’s merger remedy preferences. According to FTC staff, in an effort to continue to build on its relationship with antitrust practitioners and other interest groups (such as consumer groups and corporate personnel) and to increase the transparency of its merger review and remedy processes, FTC planned two sets of public workshops, one focusing on merger investigations and the other focusing on remedies. A March 15, 2002, FTC press release announcing the workshops states that the remedies workshops will consider whether the agency’s remedy provisions are necessary or sufficient and if the process through which they are negotiated can be improved. The first remedy workshop was held in Washington, D.C., in June 2002. Associations that represent grocery store businesses have complained to FTC that its clean sweep, single buyer, and up-front buyer divestiture practices have impaired the ability of small and independent businesses to purchase divested assets. Additionally, several smaller buyers of divested grocery store assets told us that other factors, such as the merging parties’ bidding process, create additional challenges for smaller businesses in purchasing and maintaining the viability of divested assets. However, our review of public comments and discussions with associations that represent small and independent drug stores, funeral services, or gas station businesses (all the businesses we studied except groceries), revealed that few concerns have been raised concerning the impact of FTC’s clean sweep, single buyer, and up-front buyer divestiture practices on the ability of small buyers in these industries to purchase divested assets. The National Grocers Association (NGA) and the Food Marketing Institute (FMI) submitted comments to FTC saying that FTC’s clean sweep, single buyer, and up-front buyer divestiture practices have hindered the ability of small and independent grocery store businesses to purchase divested assets. Both associations told FTC that these practices, described by FTC as preferences, have been interpreted in the business community as inflexible rules or policies. “These policies ignore the dynamics of the industry and the ever changing affects that these policies can have on the competitive environment. The FTC’s criteria of having a single buyer replace things as they are is inherently flawed and plays directly into the hands of power buyers. The FTC criteria do not recognize the inherent difference between the top five multi-billion dollar chains and privately owned regional and community retailers.” FMI submitted a paper to FTC on June 18, 2002, as part of FTC’s workshops on merger investigations and remedies. According to FMI, since approximately 1996, FTC has applied increasingly rigorous policies to supermarket divestitures. FMI noted that while FTC staff describe these policies as preferences, they often have been interpreted as inflexible rules. Similar to NGA’s comments, FMI said that the most significant policies that make it difficult for small and independent businesses to purchase divested assets are clean sweeps, single buyers, up-front buyers, and no change in market concentration policies. Specifically, FMI said that: FTC’s preference for clean sweep divestitures may prevent small retailers from assembling divestiture packages that best suit their needs. According to FMI, to the extent that FTC’s clean sweep policy is based on concerns about upsetting existing store networks or customer relationships, the concerns are overstated because (1) supermarket chains with a number of stores in a geographic market will inevitably contain some strong stores and some weaker stores and (2) there is no basis for assuming that a package of assets comprised of assets from both merging parties will contain a greater proportion of weaker stores. Additionally, FMI said that FTC’s claim that the risk that divestiture of a mixed package of assets will cause a loss of distribution efficiencies also is exaggerated because there is no basis for FTC’s assumption that an unmixed group of stores can always be supplied more efficiently than a mixed group of stores. FMI noted that “whatever logistical advantage lies in having all of the stores come from one prior owner rather than two generally is modest.” Like NGA, FMI indicated that FTC’s use of clean sweeps reflects the staffs’ desire to avoid the burden of having to evaluate individual stores to examine the viability of each store and determine whether the merging parties may be attempting to divest their least profitable assets. FTC’s preference for a single buyer limits opportunities for independents and small chains to purchase divested assets. FMI said that FTC staff rarely, if ever, allow stores to be sold to more than one buyer per market. Independents and small chains often are interested in buying, or may only have the ability to buy, a portion of the stores that are being divested in a market, but a single-buyer policy prevents this. According to FMI, a firm does not need to have complete market coverage or engage in marketwide advertising in order to be an effective competitor because no particular scale or degree of market coverage is necessary to compete effectively in grocery retailing. Additionally, FMI said that if more than one buyer is allowed to purchase the divested assets, the buyers of divested assets could seed a market with several growing chains. NGA and FMI also raised several other concerns about the impact of FTC’s divestiture practices on the ability of small businesses to purchase divested assets. Both NGA and FMI said that FTC’s strong preference for out-of-market buyers (buyers that at the time of the proposed divestiture, are not already operating within the same product and geographic markets as the assets to be divested) has disadvantaged small businesses by preventing them from expanding to provide increased competition for the merged firm. According to FMI, as long as there is any possibility of a buyer that does not currently operate in the geographic market, FTC staffs’ typical approach has been to warn counsel that securing approval for a buyer within the geographic market would be difficult and time- consuming. FMI said that in response to FTC’s warnings, most counsel have found out-of-market buyers. NGA indicated that FTC’s definition of the relevant geographic market also impacts the ability of small businesses to purchase divested assets because FTC does not allow buyers already operating in the relevant geographic market to purchase the assets to be divested if their market share would increase above that permitted by the Hortizontal Merger Guidelines. According to NGA’s public comment on the Food Lion and Hannaford merger, FTC should have defined one of the relevant geographic markets—Richmond, Virginia—in which to analyze the competitive concerns differently because there was strong evidence that inner city Richmond was a separate geographic market. NGA claimed that if FTC had defined inner city Richmond as a separate market, the market share of an independent Richmond grocer interested in purchasing the divested assets would have increased only to a level acceptable under the Horizontal Merger Guidelines. NGA noted that the merging parties’ bidding process for the assets to be divested may disadvantage small businesses. NGA suggested that FTC ensure a fair and open divestiture process by requiring the merging parties to provide FTC with information on the bidding process, such as copies of all expressions of interest in purchasing the assets to be divested and the merging parties’ evaluations of each bid. According to NGA, in numerous meetings with FTC staff and Commissioners, NGA has consistently been told that FTC leaves the divestiture process in the hands of the merging parties, and its role is merely to approve or disapprove the proposed buyer(s). NGA said that FTC should play a greater role in the bidding process for the assets to be divested. According to FMI, independents and small firms believe they face closer scrutiny of their financial viability, experience, supply arrangements, and business plans than large chains. FMI reported that the additional scrutiny of independents and small firms takes time and brings added uncertainty to the process, usually at a late stage in the investigation, when the parties are becoming increasingly worried that the merger may fall apart. Additionally, FMI noted that given the perception that independents and small chains are subjected to more scrutiny than large chains to be approved as buyers of divested assets, merging parties are likely to seek a buyer that is a “sure thing”—a well-established, out-of- market chain, rather than proposing divestiture to one or more independents. FMI concluded that in supermarket mergers, as perhaps no other industry, FTC in recent years has imposed a series of increasingly inflexible divestiture policies that place small chains and independent businesses at a disadvantage in the divestiture process, to their detriment and the detriment of consumers. Additionally, FMI said that FTC could go a long way toward dispelling the perception of bias against small chains and independent businesses in part by not insisting on clean sweeps and single buyers and by being open to in-market buyers. These changes, according to FMI, would give small chains and independent businesses realistic opportunities to become stronger and larger competitors. While obtaining revenue data from buyers of divested assets, representatives of 11 grocery store businesses with average annual revenues of $200 million or less provided anecdotal comments about the challenges that smaller businesses face in purchasing divested assets and maintaining their viability. Regarding their ability to purchase divested grocery store assets, the buyers provided the following comments. One representative noted that currently there is limited opportunity for smaller grocery store businesses to purchase divested assets because the merging parties sell the assets as a package to a single buyer. Smaller businesses may have the necessary financing to purchase one or two of the assets, but usually do not have the financial strength to purchase all of the assets. Four representatives told us that the merging parties usually offer smaller businesses the stores that the larger chains have declined to purchase. They said that they generally have access to some of the least desirable grocery stores being divested in terms of the condition of the stores, and have had to spend a considerable amount of money to refurbish them. Two representatives indicated that smaller businesses usually do not have an opportunity to express an interest in purchasing divested assets because, in practice, the bidding process is often closed to small businesses. They indicated that typically the merging parties have already identified buyers for the divested assets by the time that small businesses become aware that the merging parties must divest assets. One of the representatives said that FTC may need to play a greater role in the bidding process. The anecdotal accounts suggested that the current environment, particularly the practices of merger parties, makes it difficult to maintain the viability of divested grocery store assets: Two representatives told us that they received inaccurate or incomplete information from the merging parties or wholesaler. They said that after purchasing the stores, they discovered that the stores generated considerably less sales than reported to them by the merging parties. For example, a buyer of one divested grocery store said that he had to close the store 5 months after purchasing the store because it generated only a fraction of the sales quoted by the merging parties. Two representatives said that buyers of divested assets, particularly smaller buyers, have faced challenges remaining in business when the merging parties build state of the art supermarkets in the same geographic markets in which they were required to divest assets. They said their sales declined and several supermarkets went out of business because they were unable to compete with the larger, state of the art supermarket. Two representatives told us that the biggest problem smaller grocers face in maintaining the viability of the divested stores they purchase is competing with grocery chains that have supercenters. They told us that it is becoming extremely difficult for small businesses to compete with supercenters because the supercenters offer one-stop shopping— the convenience of being able to purchase groceries at the same time as other household items, such as clothes, appliances, and other household products. They also noted that while supercenters and large grocery chains are able to purchase and/or sell some products below costs, small businesses cannot afford to sell products below cost and remain profitable. Only one of the 1,902 public comments contained in FTC’s public files for the 31 divestiture orders included in our review raised a specific concern about how FTC’s clean sweep, single buyer, and/or up-front buyer divestiture practices have impacted the ability of small businesses to purchase divested assets. However, we were unable to fully examine the public comments for the 31 divestiture orders included in our review because some of them were confidential and others appeared to be missing from the files. Of the 1,902 public comments, 1,441 related to grocery store divestiture orders, 455 related to gas stations, four to drug stores, and two to funeral services. Our review of the public comments for each of the industries showed that: About 80 percent (1,146 of 1,441) of the comments relating to the grocery store industry were petitions or form letters from individuals expressing objections to the divestiture of a supermarket to a proposed buyer for the Ahold and Stop and Shop divestiture order announced for public comment in fiscal year 1996. Specifically, the customers of an existing local grocery store were concerned that the proposed buyer of the divested supermarket would harm the market share of the existing local grocery store business. Most of the remaining comments concerning grocery store divestitures were from customers who did not want the store where they shopped to close or change ownership. Other comments were from employees of stores being divested that did not want to work for a proposed buyer(s) when the buyer(s) did not have a good history of managing its employees or they feared they would lose their jobs if the stores were divested. Two of the four drug store industry comments opposed the merger, one requested that a particular drug store be included in the divestiture order and one requested that a specific drug store be sold to an independent drug buyer rather than another drug store chain. One of the two comments concerning the mergers in the funeral services industry expressed objection to the merging parties’ selection of the funeral homes to be divested. The other comment questioned whether an investigation was made to determine if the required divestiture would restore competition. About 60 percent (275 of 455) of the comments relating to the gas station industry were petitions from retirees of one of the merging parties who were concerned about their medical coverage after the merger or citizens who wanted to continue the gasoline brands of the 2 merging parties. Most of the remaining comments were from independent gas station dealers or their representatives expressing concerns that the divestiture orders would remove dealers from their business place with no right of first refusal on the purchase of their gas stations. FTC received only one public comment relating to small businesses that concerned the impact of FTC’s clean sweep divestiture, single buyer, and up-front buyer practices on the ability of small businesses to purchase divested assets. Specifically, in response to the Food Lion and Hannaford divestiture order announced for public comment in fiscal year 2000, NGA, as discussed previously, commented that FTC’s preference for clean sweeps, single buyers, and buyers that do not currently have a presence in the geographic market in which the assets are being divested have tremendously disadvantaged small businesses. Other public comments that related to small business ranged from whether a smaller business could effectively compete in a market dominated by larger companies to how a divestiture will impact supply contacts that small businesses had in place with one of the merging parties prior to the merger. FTC staff told us that the formal public comment period is not the only opportunity for concerned parties to express their views on an FTC matter. According to FTC staff, it is customary for affected third parties, including smaller businesses, to communicate with Commissioners and staff throughout the merger investigation. As a result, they said that the formal public record of comments on a proposed divestiture order and FTC’s responses do not represent the totality of FTC’s responsiveness to small businesses who have concerns about a pending merger. They cited two instances in which FTC has received comments outside of the public comment process regarding small and independent businesses being unable to purchase divested assets. First, they told us that some time ago, an attorney representing a grocery wholesaler presented arguments for why FTC should accept wholesaler-suppliers as acquirers of supermarkets. They said that FTC has considered those arguments in all matters, as appropriate. Additionally, in regard to the divestitures of gasoline retailing assets in the Exxon and Mobil matter, FTC received communications from a small regional “jobber” who complained that the requirement that all assets be divested to a single acquirer foreclosed the jobber from bidding on the divested assets. In that matter, FTC determined that an effective remedy precluded dividing the assets into smaller packages. We also talked with associations that represent small and independent businesses as well as SBA officials who generally told us that they were unaware of concerns directly related to the impact of FTC’s clean sweep, single buyer, and up-front buyer divestiture practices on the ability of small businesses to purchase divested assets. Specifically, The Petroleum Marketers Association of America expressed some concern about FTC’s divestiture practices in general. The official we spoke with noted that FTC has the ability to influence the merging parties’ choice of a buyer of divested assets even though FTC does not ultimately choose the buyer. A National Community Pharmacists Association (NCPA) official told us that NCPA did not have any concerns directly related to FTC’s divestiture practices. However, the official said that he was not aware that in certain situations FTC explicitly required merging parties to divest drug store assets to a single buyer. The official also said that it was his view that this requirement would definitely make it difficult for small businesses to purchase divested drug store assets. Officials of 4 of the 10 associations we contacted—the National Funeral Directors Association, the National Business Association, the National Small Business United, and the U.S. Chamber of Commerce—said that to their knowledge their members did not have any concerns directly related to the impact of FTC’s divestiture practices on the ability of small businesses to purchase divested assets. Officials of the National Federation of Independent Business and the SCORE Association told us that their associations do not have an official position on the issue. We also contacted staff in the Small Business Administration’s (SBA) Office of Advocacy to determine whether any small businesses or associations representing small businesses had raised this issue with SBA. According to SBA staff, SBA had not been made aware of concerns about the impact of FTC’s divestiture practices on the ability of small businesses to purchase divested assets. They noted that the issue that has been raised concerning mergers in the retail sector is whether small businesses will have access to products or services after the merger takes place. For example, they told us that in response to the Exxon and Mobil as well as the British Petroleum and Amoco Atlantic Richfield Company mergers, small businesses raised concerns regarding whether the merged companies would continue to provide them with access to crude oil. In addition to the persons named above, the following persons made key contributions to this report: Brodi L. Fontenot, John A. Karikari, Jan B. Montgomery, Katrina R. Moss, Cheryl M. Peterson, Jerry Sandau, Maria D. Strudwick, Carrie Wilks, and Gregory H. Wilmoth. The General Accounting Office, the investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to daily E-mail alert for newly released products” under the GAO Reports heading.

The Federal Trade Commission (FTC) seeks to prevent business practices that are anticompetitive, deceptive, or unfair to consumers. If FTC determines that a merger may harm competition in the marketplace, the agency may decide to block the merger or select a remedy that addresses the anticompetitive problems it has identified. FTC's preferred remedy is divestiture--the selling of a business or assets by one or both of the merging parties to maintain or restore competition where it might be harmed by the merger. When divestiture is chosen as a remedy, FTC usually drafts a proposed agreement with the merging parties that contains an order requiring the divestiture needed to remedy the anticompetitive problems. If all parties agree, FTC issues a proposed order which is made available to the public for comment for 30 days and, in most cases, authorizes the parties to consummate the merger. According to FTC staff, FTC decisions to use particular divestiture approaches are (1) based on the unique facts of each case and do not readily translate into written guidelines or systematic aggregation and (2) tied to proprietary company information that FTC is statutorily prohibited from disclosing to the public. From fiscal years 1990 through 2000, FTC used clean sweep divestitures, single buyers, or up-front buyers in the 31 divestiture orders announced for public comment in the grocery store, drug store, funeral services, and gas station industries (up-front buyers were not used at all in these industries prior to fiscal year 1996). Although there were too few buyers to analyze the level of smaller business participation in purchasing divested drug store, funeral services, and gas station assets for the period reviewed, GAO's analysis found that, after 1996, the smaller buyers were significantly less likely to purchase divested assets. FTC has not systematically measured the success or failure of the divestitures it has approved since it developed preferences for approaches like clean sweep and up-front buyers. In 1999, FTC reported the results of a study on divestiture orders made final during fiscal years 1990 through 1994 that found (1) FTC's divestiture orders had created viable competitors in the relevant markets and (2) smaller buyers succeeded at least at the same rate as larger buyers. However, without current information on the economic impact of its divestiture practices on the marketplace, especially given the changes it has made since the period covered by its 1999 study, FTC cannot state that divestiture orders have, among other things, restored or maintained competition in the affected markets.

The United States is the world’s largest net importer of oil. In 2006, the United States had net imports of 12.2 million gallons of oil per day, more than twice as much as Japan and over three times as much as China, the world’s next largest importers. The transport of oil into the United States occurs primarily by sea with ports throughout the United States receiving over 40,000 shipments of oil in 2005. In addition, vessels not transporting oil, such as cargo and freight vessels, fishing vessels, and passenger ships, often carry tens of thousands of gallons of fuel oil to power their engines. With over 100,000 commercial vessels navigating U.S. waters, oil spills are inevitable. Fortunately, however, they are relatively infrequent and are decreasing. While oil transport and maritime traffic have continued to increase, the total number of reported spills has generally declined each year since 1990. OPA forms the foundation of U.S. maritime policy as it pertains to oil pollution. OPA was passed in 1990, following the 1989 Exxon Valdez spill in Alaska, which highlighted the need for greater federal oversight of maritime oil transport. OPA places the primary burden of liability and the costs of oil spills on the vessel owner and operator who was responsible for the spill. This “polluter pays” system provides a deterrent for vessel owners and operators who spill oil by requiring that they assume the burden of spill response, natural resource restoration, and compensation to those damaged by the spill, up to a specified limit of liability—which is the amount above which responsible parties are no longer financially liable under certain conditions. For example, if a vessel’s limit of liability is $10 million and a spill resulted in $12 million in costs, the responsible party only has to pay up to $10 million—the Fund will pay for the remaining $2 million. Current limits of liability, which vary by type of vessel and are determined by a vessel’s gross tonnage, were set by the Congress in 2006. The Coast Guard is responsible for adjusting limits for significant increases in inflation and for making recommendations to the Congress on whether adjustments are necessary to help protect the Fund. OPA also requires that vessel owners and operators must demonstrate their ability to pay for oil spill response up to their limit of liability. Specifically, by regulation, with few exceptions, owners and operators of vessels over 300 gross tons and any vessels that transship or transfer oil in the Exclusive Economic Zone are required to have a certificate of financial responsibility that demonstrates their ability to pay for oil spill response up to their limit of liability. OPA consolidated the liability and compensation provisions of four prior federal oil pollution initiatives and their respective trust funds into the Oil Spill Liability Trust Fund and authorized the collection of revenue and the use of the money, with certain limitations, with regards to expenditures. The Fund has two major components—the Principal Fund and the Emergency Fund. The Emergency Fund consists of $50 million apportioned each year to fund spill response and the initiation of natural resource damage assessments, which provide the basis for determining the natural resource restoration needs that address the public’s loss and use of natural resources as a result of a spill. The Principal Fund provides the funds for third-party and natural resource damage claims, limit of liability claims, reimbursement of government agencies’ removal costs, and provides for oil spill related appropriations. A number of agencies— including the Coast Guard, EPA, and DOI—receive an annual appropriation from the Fund to cover administrative, operational, personnel, and enforcement costs. From 1990 to 2006, these appropriations amounted to the Fund’s largest expense (see fig. 2). The Fund’s balance has generally declined from 1995 through 2006, and since fiscal year 2003, its balance has been less than the authorized limit on federal expenditures for the response to a single spill, which is currently set at $1 billion (see fig. 3). The balance has declined, in part, because the Fund’s main source of revenue—a $0.05 per barrel tax on U.S. produced and imported oil—was not collected for most of the time between 1993 and 2006. As a result, the Fund balance was $604.4 million at the end of fiscal year 2006. The Energy Policy Act of 2005 reinstated the barrel tax beginning in April 2006. With the barrel tax once again in place, NPFC anticipates that the Fund will be able to cover its projected noncatastrophic liabilities. OPA also defines the costs for which responsible parties are liable and for the costs for which the Fund is made available for compensation in the event that the responsible party does not pay or is not identified. These costs, or “OPA compensable” costs, are of two main types: Removal costs: Removal costs are incurred by the federal government or any other entity taking approved action to respond to, contain, and clean up the spill. For example, removal costs include the equipment used in the response—skimmers to pull oil from the water, booms to contain the oil, planes for aerial observation—as well as salaries and travel and lodging costs for responders. Damages caused by the oil spill: OPA-compensable damages cover a wide range of both actual and potential adverse impacts from an oil spill, for which a claim may be made to either the responsible party or the Fund. (Table 1 provides a brief definition of OPA-compensable removal costs and damages.) Claims include natural resource damage claims filed by trustees, claims for uncompensated removal costs and third-party damage claims for lost or damaged property and lost profits, among other things. The Fund also covers costs when responsible parties cannot be located or do not pay their liabilities. NPFC encounters cases where the source of the spill, and therefore the responsible party is unknown, or where the responsible party does not have the ability to pay. In other cases, since the cost recovery can take a period of years, the responsible party may be bankrupt or dissolved. Based on our analysis of NPFC records, excluding spills with limit of liability claims, the recovery rate for costs from the 51 major oil spills since 1990 is 65 percent, which means that responsible parties have paid 65 percent of costs. The 35 percent of nonreimbursed costs to the Fund for these major spills have amounted to $53.9 million. Response to large oil spills is typically a cooperative effort between the public and private sector, and there are numerous players who participate in responding to and paying for oil spills. To manage the response effort, the responsible party, the Coast Guard, EPA, and the pertinent state and local agencies form the unified command, which implements and manages the spill response. Beyond the response operations, there are other stakeholders, such as accountants who are involved in documenting and accounting for costs, and receiving and processing claims. In addition, insurers and underwriters provide financial backing to the responsible party. The players involved in responding to and/or paying for major spill response are as follows: Government agencies: The lead federal authority, or Federal On-Scene Coordinator, in conducting a spill response is usually the nearest Coast Guard Sector and is headed by the Coast Guard Captain of the Port. The Federal On-Scene Coordinator directs response efforts and coordinates all other efforts at the scene of an oil spill. Additionally, the on-scene coordinator issues pollution removal funding authorizations—guarantees that the agency will receive reimbursement for performing response activities—to obtain services and assistance from other government agencies. Other federal agencies may also be involved. NOAA provides scientific support, monitoring and predicting the movement of oil, and conducting environmental assessments of the impacted area. The federal, state, and tribal trustees join together to perform a natural resource damage assessment, if necessary. Within the Coast Guard, the NPFC is responsible for disbursing funds to the Federal On-Scene Coordinator for oil spill removal activities and seeking reimbursement from responsible parties for federal costs. Additionally, regional governmental entities that are affected by the spill—both state and local—as well as tribal government officials or representatives may participate in the unified command and contribute to the response effort, which is paid for by the responsible party or are reimbursed by the responsible party or the Fund. Responsible parties: OPA stipulates that both the vessel owner and operator are ultimately liable for the costs of the spill and the cleanup effort. The Coast Guard has final determination on what actions must be taken in a spill response, and the responsible party may form part of the unified command—along with the Federal On-Scene Coordinator and pertinent state and local agencies—to manage the spill response. The responsible parties rely on other entities to evaluate the spill effects and the resulting compensation. Responsible parties hire environmental and scientific support staff, specialized claims adjustors to adjudicate third- party claims, public relations firms, and legal representation to file and defend limit of liability claims on the Fund, as well as serve as counsel throughout the spill response. Qualified individuals: Federal regulations require that vessels carrying oil as cargo have an incident response plan and, as part of the plan, they appoint a qualified individual who acts with full authority to obligate funds required to carry out response activities. The qualified individual acts as a liaison with the Federal On-Scene Coordinator and is responsible for activating the incident response plan. Oil spill response organizations: These organizations are private companies that perform oil spill cleanup, such as skimming and disposal of oil. Many of the companies have contractual agreements with responsible parties and the Coast Guard. The agreements, called basic ordering agreements, provide for prearranged pricing, response personnel, and equipment in the event of an oil spill. Insurers: Responsible parties often have multiple layers of primary and excess insurance coverage, which pays oil spill costs and claims. Pollution liability coverage for large vessels is often underwritten by not-for-profit mutual insurance organizations. The organizations act as a collective of ship owners, who insure themselves, at-cost. The primary insurers of commercial vessels in U.S. waters are the Water Quality Insurance Syndicate, an organization providing pollution liability insurance to over 40,000 vessels, and the International Group of P & I Clubs, 13 protection and indemnity organizations that provide insurance primarily to foreign- flagged large vessels. On the basis of information we were able to assemble about responsible parties’ expenditures and payments from the Fund, we estimate that 51 oil spills involving removal costs and damage claims totaling $1 million or more have occurred since 1990. In all, the Fund spent $240 million on these spills, and the responsible parties themselves spent about $620 million to $840 million, for a total of $860 million to $1.1 billion. The number of spills and their costs varied from year to year and showed no discernable trends in either frequency or cost. Less than 2 percent of oil spills from vessels, since 1990, had removal costs and damage claims of $1 million or greater. Each year, there are thousands of incident reports called into the National Response Center that claim oil or oil-like substances have been spilled from vessels sailing in coastal or inland waters in the United States—-but only a small percentage of these reported incidents are oil spills from vessels that received federal reimbursement for response efforts. Specifically, there have been 3,389 oil spills from vessels that sought reimbursement from the Fund for response efforts. Of these spills, we estimate that 51 were major oil spills. As figure 4 shows, there are no discernable trends in the number of major oil spills that occur each year. The highest number of spills was seven in 1996; the lowest number was zero in 2006. These 51 spills occurred in a variety of locations. As figure 5 shows, the spills occurred on the Atlantic, Gulf, and Pacific coasts and include spills both in open coastal waters and more confined waterways. The total cost of the 51 spills cannot be precisely determined, for several reasons: Private-sector expenditures are not tracked: The NPFC tracks federal removal costs expended by the Fund for Coast Guard and other federal agencies’ spill response efforts, but it does not oversee costs incurred by the private sector. There is also no legal requirement in place that requires responsible parties to disclose costs incurred for responding to a spill. The various parties involved in covering these costs do not categorize them uniformly: For example, one vessel insurer we spoke with separates total spill costs by removal costs (for immediate spill cleanup) and loss adjustment expenses, which contain all other expenses, including legal fees. In contrast, the NPFC tracks removal costs and damage claims in terms of the statutory definitions delineated in OPA. Spill costs are somewhat fluid and accrue over time: In particular, the natural resource and third-party damage claims adjudication processes can take many years to complete. Moreover, it can take many months or years to determine the full effect of a spill to natural resources and to determine the costs and extent of the natural resource injury and the appropriate restoration needed to repair the damage. For example, natural resource damage claims were recently paid for a spill that occurred near Puerto Rico in 1991, over 16 years ago. Because spill cost data are somewhat imprecise and the data we collected vary somewhat by source, the results described below will be reported in ranges, in which various data sources are combined together. The lower and higher bounds of the range represent the low and high end of cost information we obtained. Our analysis of these 51 spills shows their total cost was approximately $1 billion—ranging from $860 million to $1.1 billion. This amount breaks down by source as follows: Amount paid out of the Fund: Because the NPFC tracks and reports all Fund expenditures, the amount paid from the Fund can be reported as an actual amount, not an estimate. For these 51 spills, the Fund paid a total of $239.5 million. Amount paid by responsible parties: Because of the lack of precise information about amounts paid by responsible parties and the differences in how they categorize their costs, this portion of the expenditures must be presented as an estimate. Based on the data we were able to obtain and analyze, responsible parties spent between $620 million and $840 million. Even at the low end of the range, this amount is nearly triple the expenditure from the Fund. Costs of these 51 spills varied widely by spill, and therefore, by year (see fig. 6). For example, 1994 and 2004 both had four spills during the year, but the average cost per spill in 1994 was about $30 million, while the average cost per spill in 2004 was between $71 million and $96 million. Just as there was no discernible trend in the frequency of these major spills, there is no discernible trend in their cost. Although the substantial increase in 2004 may look like an upward trend, 2004 may be an anomaly that reflects the unique character of two of the four spills that occurred that year. These two spills accounted for 98 percent of the year’s costs. Location, time of year, and type of oil are key factors affecting oil spill costs, according to industry experts, agency officials, and our analysis of spills. Data on the 51 major spills show that spills occurred on all U.S. coasts, across all seasons, and for all oil types. In ways that are unique to each spill, however, each of these factors can affect the breadth and difficulty of the response effort or the extent of damage that requires mitigation. For example, spills that occur in remote areas can make response difficult in terms of mobilizing responders and equipment, as well as complicating the logistics of removing oil—all of which can increase the costs. Officials also identified two other factors that may influence oil spill costs to a lesser extent—the effectiveness of the spill response and the level of public interest in a spill. The location of a spill can have a large bearing on spill costs because it will determine the extent of response needed, as well as the degree of damage to the environment and local economies. According to state officials with whom we spoke and industry experts, there are three primary characteristics of location that affect costs: Remoteness: For spills that occur in remote areas, spill response can be particularly difficult in terms of mobilizing responders and equipment, and they can complicate the logistics of removing oil from the water—all of which can increase the costs of a spill. For example, a 2001 spill in Alaska’s Prince William Sound—which occurred approximately 40 miles from Valdez, AK—resulted in considerable removal costs after a fishing vessel hit a rock and sank to a depth of approximately 1,000 feet. Response took many days and several million dollars to contain the oil that was still in the vessel, but the effort was eventually abandoned because it was too difficult from that depth. Proximity to shore: There are also significant costs associated with spills that occur close to shore. Contamination of shoreline areas has a considerable bearing on the costs of spills as such spills can require manual labor to remove oil from the shoreline and sensitive habitats. The extent of damage is also affected by the specific shoreline location. For example, spills that occur in marshes and swamps with little water movement are likely to incur more severe impacts than flowing water. A September 2002 spill from a cargo vessel in the Cooper River near the harbor in Charleston, SC, spread oil across 30 miles of a variety of shoreline types. The spill resulted in the oiling of a number of shorebirds and a temporary disruption to recreational shrimp-baiting in area waters, among other things. As of July 2007, a settlement for natural resource damages associated with the spill was still pending. Proximity to economic centers: Spills that occur in the proximity of economic centers can also result in increased costs when local services are disrupted. A spill near a port can interrupt the flow of goods, necessitating an expeditious response in order to resume business activities, which could increase removal costs. Additionally, spills that disrupt economic activities can result in expensive third-party damage claims. For example, after approximately 250,000 gallons of oil spilled from a tanker in the Delaware River in 2004, a large nuclear plant in the vicinity was forced to suspend activity for more than a week. The plant is seeking reimbursement for $57 million in lost profits. Overall, for the 51 major oil spills, location had the greatest effect on costs for spills that occurred in the waters of the Caribbean, followed by the East Coast, Alaska, and the Gulf states. (See fig. 7). The range of average per spill costs for the spills that occurred in the East Coast locations ranged from about $27 million to over $37 million, higher than the average costs in any other region besides the two spills in Caribbean. The high spill costs in the East Coast locations were caused by several spills in that geographic area that had considerably higher costs. Specifically, four of the eight most expensive spills occurred on the waters off the East Coast. The time of year in which a spill occurs can also affect spill costs—in particular, impacting local economies and response efforts. According to several state and private-sector officials with whom we spoke, spills that disrupt seasonal events that are critical for local economies can result in considerable expenses. For example, spills in the spring months in areas of the country that rely on revenue from tourism may incur additional removal costs in order to expedite spill cleanup, or because there are stricter standards for cleanup, which increase the costs. This situation occurred in March of 1996 when a tank barge spilled approximately 176,000 gallons of fuel oil along the coast of Texas. Because the spill occurred during the annual spring break tourist season, the time frames for cleaning up the spill were truncated, and the standards of cleanliness were elevated. Both of these factors contributed to higher removal costs, according to state officials we interviewed. The time of year in which a spill occurs also affects response efforts because of possible inclement weather conditions. For example, spills that occur during the winter months in areas of the country that experience harsh winter conditions can result in higher removal costs because of the increased difficulty in mobilizing equipment and personnel to respond to a spill in inclement weather. According to a state official knowledgeable about a January 1996 spill along the coast of Rhode Island, extremely cold and stormy weather made response efforts very difficult. Although the 51 spills occurred during all seasons of the year, they were most prevalent in the fall and winter months, with 20 spills occurring in the fall and 13 spills during the winter, compared with 9 spills in the spring and 9 in the summer months. On a per-spill basis, the cost range for the 51 spills was highest in the fall (see fig. 8). The type of oil spilled affects the degree to which oil can be cleaned up and removed, as well as the nature of the natural resource damage caused by the spill—both of which can significantly impact the costs associated with an oil spill. The different types of oil can be grouped into four categories, each with its own set of impacts on spill response and the environment (see table 2). For example, lighter oils such as jet fuels, gasoline, and diesel dissipate quickly, but they are highly toxic, whereas heavier oils such as crude oils and other heavy petroleum products do not dissipate much and, while less toxic, can have severe environmental impacts. Very light and light oils naturally dissipate and evaporate quickly, and as such, often require minimal cleanup. However, light oils that are highly toxic can result in severe impacts to the environment, particularly if conditions for evaporation are unfavorable. For instance, in 1996, a tank barge that was carrying home-heating oil grounded in the middle of a storm near Point Judith, Rhode Island, spilling approximately 828,000 gallons of heating oil (light oil). Although this oil might dissipate quickly under normal circumstances, heavy wave conditions caused an estimated 80 percent of the release to mix with water, with only about 12 percent evaporating and 10 percent staying on the surface of the water. The natural resource damages alone were estimated at $18 million, due to the death of approximately 9 million lobsters, 27 million clams and crabs, and over 4 million fish. Medium and heavy oils do not evaporate much, even during favorable weather conditions, and thus, can result in significant contamination of shoreline areas. Medium and heavy oils have a high density and can blanket structures they come in contact with—boats and fishing gear, for example—as well as the shoreline, creating severe environmental impacts to these areas, and harming waterfowl and fur-bearing mammals through coating and ingestion. Additionally, heavy oils can sink, creating prolonged contamination of the sea bed and tar balls that sink to the ocean floor and scatter along beaches. These spills can require intensive shoreline and structural cleanup, which is time consuming and expensive. For example, in 1995, a tanker spilled approximately 38,000 gallons of heavy fuel oil into the Gulf of Mexico when it collided with another tanker as it prepared to lighter its oil to another ship. Less than 1 percent (210 gallons) of the oil was recovered from the sea, and as a result, recovery efforts on the beaches of Matagorda and South Padre Islands were labor intensive, as hundreds of workers had to manually pick up tar balls with shovels. The total removal costs for the spill were estimated at $7 million. Spills involving heavy oil were the most prevalent among the 51 spills; 21 of the 51 major oil spills were from heavy oils. On a per-spill basis, costs among the 51 spills, varied by type of oil, but the cost ranges for medium and heavy oils were higher than light and very light oils (see fig. 9). Although available evidence points to location, time of year, and type of oil spilled as key factors affecting spill costs, some industry experts reported that the effectiveness of the spill response and the level of the public interest can also impact the costs incurred during a spill. Effectiveness of spill response: Some private-sector officials stated that the effectiveness of spill response can impact the cost of cleanup. The longer it takes to assemble and conduct the spill response, the more likely it is that the oil will move with changing tides and currents and affect a greater area, which can increase costs. Some officials also stated that the level of experience of those involved in the incident command is critical to the effectiveness of spill response, and they can greatly affect spill costs. For example, poor decision making during a spill response could lead to the deployment of unnecessary response equipment, or worse, not enough equipment to respond to a spill. In particular, several private-sector officials with whom we spoke expressed concern that Coast Guard officials are increasingly inexperienced in handling spill response, in part because the Coast Guard’s mission has been increased to include homeland security initiatives. Additionally, another noted that response companies, in general, have less experience in dealing with spill response and less familiarity with the local geography of the area affected by the spill, which can be critical to determining which spill response techniques are most effective in a given area. They attributed the limited experience to the overall decline in the number of spills in recent years. Further, one private-sector official noted that response companies can no longer afford to specialize in cleaning up spills alone, given the relatively low number of spills, and thus, the quality, effectiveness, and level of expertise and experience diminish over time. Public interest: Several officials with whom we spoke stated that level of public attention placed on a spill creates pressure on parties to take action and can increase costs. They also noted that the level of public interest can increase the standards of cleanliness expected, which may increase removal costs. For example, several officials noted that a spill along the Texas coast in February 1995 resulted in increased public attention because it occurred close to peak tourist season. In addition to raising the standards of cleanliness at the beaches to a much higher level than normal because of tourist season, certain response activities were completed for primarily aesthetic reasons, both of which increased the removal costs, according to state officials. The Fund has been able to cover costs from major spills that responsible parties have not paid, but risks remain. Although liability limits were increased in 2006, the liability limits for certain vessel types, notably tank barges, may be disproportionately low relative to costs associated with such spills. There is also no assurance that vessel owners and operators are able to financially cover these new limits, because the Coast Guard has not yet issued regulations for satisfying financial responsibility requirements. In addition, although OPA calls for periodic increases in liability limits to account for significant increases in inflation, such increases have never been made. We estimate that not making such adjustments in the past potentially cost the Fund $39 million between 1990 and 2006. Besides issues related to limits of liability, the Fund faces other potential drains on its resources, including ongoing claims from existing spills, claims related to already-sunken vessels that may begin to leak oil, and the threat of a catastrophic spill such as occurred with the Exxon Valdez in 1989. Major oil spills that exceed the vessel’s limit of liability are infrequent, but their impact on the Fund could be significant. Limits of liability are the amount, under certain circumstances, above which responsible parties are no longer financially liable for spill removal costs and damage claims. If the responsible party’s costs exceed the limit of liability, they can make a claim against the Fund for the amount above the limit. Of the 51 major oil spills that occurred since 1990, 10 spills resulted in limit of liability claims on the Fund. The limit of liability claims of these 10 spills ranged from less than $1 million to over $100 million, and totaled over $252 million in claims on the Fund. Limit of liability claims will continue to have a pronounced effect on the Fund. NPFC estimates that 74 percent of claims under adjudication that were outstanding as of January 2007 were for spills in which the limit of liability had been exceeded. The amount of these claims under adjudication was $217 million. We identified three areas in which further attention to these liability limits appears warranted: the appropriateness of some current liability limits, the need to adjust limits periodically in the future to account for significant increases in inflation, and the need for updated regulations for ensuring vessel owners and operators are able to financially cover their new limits. The Coast Guard and Maritime Transportation Act of 2006 significantly increased the limits of liability from the limits set by OPA in 1990. Both laws base the liability on a specified amount per gross ton of vessel volume, with different amounts for vessels that transport oil commodities (tankers and tank barges) than for vessels that carry oil as a fuel (such as cargo vessels, fishing vessels, and passenger ships). The 2006 act raised both the per-ton and the required minimum amounts, differentiating between vessels with a double hull, which helps prevent oil spills resulting from collision or grounding, and vessels without a double hull (see table 3 for a comparison of amounts by vessel category). For example, the liability limit for single-hull vessels larger than 3,000 gross tons was increased from the greater of $1,200 per gross ton or $10 million to the greater of $3,000 per gross ton or $22 million. Our analysis of the 51 spills showed that the average spill cost for some types of vessels, particularly tank barges, was higher than the limit of liability, including the new limits established in 2006. We separated the vessels involved in the 51 spills into four types (tankers, tank barges, cargo and freight ships, and other vessels such as fishing boats); determined the average spill costs for each type of vessel; and compared the costs with the average limit of liability for these same vessels under both the 1990 and 2006 limits. As figure 10 shows, the 15 tank barge spills and the 12 fishing/other vessel spills had average costs greater than both the 1990 and 2006 limits of liability. For example, for tank barges, the average cost of $23 million was higher than the average limit of liability of $4.1 million under the 1990 limits and $10.3 million under the new 2006 limits. The nine spills involving tankers, by comparison, had average spill costs of $34 million, which was considerably lower than the average limit of liability of $77 million under the 1990 limits and $187 million under the new 2006 limits. In a January 2007 report examining spills in which the limits of liability had been exceeded, the Coast Guard had similar findings on the adequacy of some of the new limits. Based on an analysis of 40 spills in which costs had exceeded the responsible party’s liability limit since 1991, the Coast Guard found that the Fund’s responsibility would be greatest for spills involving tank barges, where the Fund would be responsible for paying 69 percent of costs. The Coast Guard concluded that increasing liability limits for tank barges and nontank vessels—cargo, freight, and fishing vessels— over 300 gross tons would positively impact the Fund balance. With regard to making specific adjustments, the Coast Guard said dividing costs equally between the responsible parties and the Fund was a reasonable standard to apply in determining the adequacy of liability limits. However, the Coast Guard did not recommend explicit changes to achieve either that 50/50 standard or some other division of responsibility. Although OPA requires adjusting liability limits to account for significant increases in inflation, no adjustments to the limits were made between 1990 and 2006, when the Congress raised the limits in the Coast Guard and Maritime Transportation Act. During those years, the Consumer Price Index rose approximately 54 percent. OPA requires the President, who has delegated responsibility to the Coast Guard, through the Secretary of Homeland Security, to issue regulations not less often than every 3 years to adjust the limits of liability to reflect significant increases in the Consumer Price Index. We asked Coast Guard officials why no adjustments were made between 1990 and 2006. Coast Guard officials stated that they could not speculate on behalf of other agencies as to why no adjustments had been made prior to 2005 when the delegation to the Coast Guard was made. The decision to leave limits unchanged had financial implications for the Fund. Raising the liability limits to account for inflation would have the effect of reducing payments from the Fund, because responsible parties would be responsible for paying costs up to the higher liability limit. Not making adjustments during this 16-year period thus had the effect of increasing the Fund’s financial liability. Our analysis showed that if the 1990 liability limits had been adjusted for inflation during the 16-year period, claims against the Fund for the 51 major oil spills would have been reduced 16 percent, from $252 million to $213 million. This would have meant a savings of $39 million for the Fund. Certificates of Financial Responsibility have not been adjusted to reflect the new liability limits. The Coast Guard requires Certificates of Financial Responsibility, with few exceptions, for vessels over 300 gross tons or any vessels that are lightering or transshipping oil in the Exclusive Economic Zone as a legal certification that vessel owners and operators have the financial resources to fund spill response up to the vessel’s limit of liability. Currently, Certificate of Financial Responsibility requirements are consistent with the 1990 limits of liability and, therefore, there is no assurance that responsible parties have the financial resources to cover their increased liability. The Coast Guard is currently making Certificates of Financial Responsibility consistent with current limits of liability. The Coast Guard plans to initiate a rule making to issue new Certificate of Financial Responsibility requirements. Coast Guard officials indicated their goal is to publish a Notice of Proposed Rulemaking by the end of 2007, but the officials said they could not be certain they would meet this goal. The Fund also faces several other potential challenges that could affect its financial condition: Additional claims could be made on spills that have already been cleaned up: Natural resource damage claims can be made on the Fund for years after a spill has been cleaned up. The official natural resource damage assessment conducted by trustees can take years to complete, and once it is completed, claims can be submitted to the NPFC for up to 3 years thereafter. For example, the NPFC recently received and paid a natural resource damage claim for a spill in U.S. waters in the Caribbean that occurred in 1991. Costs and claims may occur on spills from previously sunken vessels that discharge oil in the future: Previously sunken vessels that are submerged and in threat of discharging oil represent an ongoing liability to the Fund. There are over 1,000 sunken vessels that pose a threat of oil discharge. These potential spills are particularly problematic because, in many cases, there is no viable responsible party that would be liable for removal costs. Therefore, the full cost burden of oil spilled from these vessels would likely be paid by the Fund. Spills may occur without an identifiable source and therefore, no responsible party: Mystery spills also have a sustained impact on the Fund, because costs for spills without an identifiable source—and therefore no responsible party—may be paid out of the Fund. Although mystery spills are a concern, the total cost to the Fund from mystery spills was lower than the costs of known vessel spills in 2001 through 2004. Additionally, none of the 51 major oil spills was the result of a discharge from an unknown source. A catastrophic spill could strain the Fund’s resources: Since the 1989 Exxon Valdez spill, which was the impetus for authorizing the Fund’s usage, no oil spill has come close to matching its costs. Cleanup costs for the Exxon Valdez alone totaled about $2.2 billion, according to the vessel’s owner. By comparison, the 51 major oil spills since 1990 cost, in total, between $860 million and $1.1 billion. The Fund is currently authorized to pay out a maximum of $1 billion on a single spill. Although the Fund has been successful thus far in covering costs that responsible parties did not pay, it may not be sufficient to pay such costs for a spill that has catastrophic consequences. The “polluter pays” system established under OPA has been generally effective in ensuring that responsible parties pay the costs of responding to spills and compensating those affected. Given that responsible parties’ liability is not unlimited, the Fund remains an important source of funding for both response and damage compensation, and its viability is important. The Fund has been able to meet all of its obligations, helped in part by the absence of any spills of catastrophic size. This favorable result, however, is no guarantee of similar success in the future. Even moderate spills can be very expensive, especially if they occur in sensitive locations or at certain times of the year. Increases in some liability limits appear warranted to help ensure that the “polluter pays” principle is carried out in practice. For certain vessel types, such as tank barges, current liability limits appear disproportionately low relative to their historic spill costs. The Coast Guard has reached a similar conclusion but so far has stopped short of making explicit recommendations to the Congress about what the limits should be. Absent such recommendations, the Fund may continue to pay tens of millions of dollars for spills that exceed the responsible parties’ limits of liability. As the agency responsible for the Fund, it is important that the Coast Guard regularly assess whether and how the limits of liability for all vessel types should be adjusted—and recommends a course of action to the Congress on the adjustments that are warranted. Further, to date, liability limits have not been adjusted for significant changes in inflation. Consequently, the Fund was exposed to about $39 million in liability claims for the 51 major spills between 1990 and 2006 that could have been saved if the limits had been adjusted for inflation. Authority to make such adjustments was specifically designated to the Coast Guard in 2005, and with this clear authority, it is important for the Coast Guard to periodically adjust the limits of liability for inflation, as well. Without such actions, oil spills with costs exceeding the responsible parties’ limits of liability will continue to place the Fund at risk. To improve and sustain the balance of Oil Spill Liability Trust Fund, we recommend that the Commandant of the Coast Guard take the following two actions: Determine whether and how liability limits should be changed, by vessel type, and make specific recommendations about these changes to the Congress Adjust the limits of liability for vessels every 3 years to reflect significant changes in inflation, as appropriate. We provided a draft of this report to the Department of Homeland Security (DHS), including the Coast Guard and NPFC, for review and comment. DHS provided written comments, which are reprinted in appendix II. In its letter, DHS agreed with both recommendations. Regarding our recommendation that the Coast Guard review limits of liability by vessel type and make recommendations to the Congress, DHS stated that it has met the intent of the recommendation by issuing the first of its annual reports, in January 2007, on limits of liability. As stated in our report, however, our concern is that the current annual report made no specific recommendations to the Congress regarding liability limit adjustments. Therefore, we continue to recommend that in its next annual report to the Congress on limits of liability, the Coast Guard make explicit recommendations, by vessel type, on how such limits should be adjusted. Regarding our recommendation that the Coast Guard adjust the limits of liability for vessels every 3 years to reflect significant changes in inflation, DHS stated that the Coast Guard will make adjustments to limits as appropriate. In response to other concerns that DHS expressed, we modified the report to clarify the Coast Guard’s responsibility for adjusting liability limits in response to Consumer Price Index increases, and to deal with the Coast Guard’s concern that the report not imply that responsible parties’ liability is unlimited. In addition, we provided a draft report to several other agencies—the Departments of Commerce, Transportation, DOI and EPA—for review and comment, because some of the information in the report was obtained from these agencies and related to their responsibilities. The agencies provided technical clarifications, which we have incorporated in this report, as appropriate. We are sending copies of this report to the Departments of Homeland Security, including the Coast Guard; Transportation, Commerce, DOI, and EPA; and appropriate congressional committees. We will also make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you have any questions about this report, please contact me at flemings@gao.gov or (202) 512-4431. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix III. To address our objectives, we analyzed oil spill removal cost and claims data from the National Pollution Funds Center (NPFC); the National Oceanic and Atmospheric Administration’s (NOAA) Damage Assessment, Remediation, and Restoration Program; and the Department of the Interior’s (DOI) Natural Resource Damage Assessment and Restoration Program; and the U.S. Fish and Wildlife Service (FWS). We also analyzed data obtained from vessel insurers, and in contract with Environmental Research Consulting. We interviewed NPFC and NOAA officials and state officials responsible for oil spill response, as well as industry experts and representatives from key industry associations and a vessel operator. In addition, we selected five oil spills that represented a variety of factors such as geography, oil type, and spill volume for an in-depth review. During this review, we interviewed NPFC officials involved in spill response for all five spills, as well as representatives of private-sector companies involved in the spill and spill response; we also conducted a file review of NPFC records of the federal response activities and costs associated with spill cleanup. We also reviewed documentation from the NPFC regarding the Fund balance and vessels’ limits of liability. Based on reviews of data documentation, interviews with relevant officials, and tests for reasonableness, we determined that the data were sufficiently reliable for the purposes of our study. This report focuses on oil spills that have occurred since the enactment of OPA—August 18, 1990—for which removal costs and damage claims exceeded $1 million, and we refer to such spills as major oil spills. We conducted our review from July 2006 through August 2007 in accordance with generally accepted government auditing standards. For the purposes of this review, we included removal (or response) costs and damage claims that are considered OPA compensable; that is, the OPA-stipulated reimbursable costs that are incurred for oil pollution removal activities when oil is discharged into the navigable waters, adjoining shorelines, and the Exclusive Economic Zone of the United States, as well as costs incurred to prevent or mitigate the substantial threat of such an oil discharge. OPA compensable removal costs include containment and removal oil from water and shorelines; prevention or minimization of a substantial threat of discharge; contract services (e.g., cleanup contractors, incident management support, and wildlife rehabilitation); equipment used in removals; chemical testing required to identify the type and source of oil; proper disposal of recovered oil and oily debris; costs for government personnel and temporary government employees hired for the duration of the spill response, including costs for monitoring the activities of responsible parties; completion of documentation; and identification of responsible parties. OPA compensable damage claims include uncompensated removal costs, damages to natural resources, damages to real or personal property, loss of subsistence use of natural resources, loss of profits or earning capacity, loss of government revenues, and increased cost of public services. In order to present the best available data on spill costs, we gathered cost information from a number of sources, including federal agencies, vessel insurance companies and other private-sector companies involved in oil spill response, and Environmental Research Consulting—a private consultant. Federal agencies: We gathered federal data on OPA compensable oil spill removal costs from the NPFC. Additionally, we gathered federal data on OPA compensable third-party damage claims from the NPFC, and natural resource damage claims from NOAA’s Damage Assessment, Remediation, and Restoration Program, DOI’s Natural Resource Damage Assessment and Restoration Program, and FWS. Insurers and other private-sector companies: We collected the best available data for OPA-compensable removal costs and damage claims from private-sector sources, including vessel insurers such as the Water Quality Insurance Syndicate and the International Group of Protection and Indemnity Clubs; oil spill response organizations, including the Alaska Chadux Corporation and Moran Environmental Recovery; and a vessel operator. We made many attempts to contact and interview the responsible parties involved in the five spills we reviewed in-depth. One was willing to speak to GAO directly. Environmental Research Consulting: Environmental Research Consulting is a consulting firm that specializes in data analysis, environmental risk assessment, cost analyses, and the development of comprehensive databases on oil/chemical spills and spill costs. Environmental Research Consulting supplied cost estimates based on reviews of court documents, published reports, interviews with responsible parties, and other parties involved with major oil spills. In addition, Environmental Research Consulting verified its data collection by relying exclusively on known documented costs, as opposed to estimated costs. Environmental Research Consulting, therefore, did not include general estimates of spill costs, which can be inaccurate. A complete and accurate accounting of total oil spill costs for all oil spills is unknown, primarily because there is no uniform mechanism to track responsible party spill costs, and there are no requirements that private sector keep or maintain cost records. The NPFC tracks federal costs to the Coast Guard and other federal agencies, which are later reimbursed by the Fund, but does not oversee costs incurred by the private sector. There is also no legal requirement in place that requires responsible parties to disclose costs incurred for responding to a spill. We cannot be certain that all private-sector cost information we gathered included only OPA- compensable costs. However, we explicitly outline which costs are included in our review. Furthermore, private-sector data were obtained primarily from insurance companies, and one official told us that insurance coverage for pollution liability usually defines compensable losses in the same manner as OPA. For instance, while responsible parties incur costs ancillary to the spill response, such as public relations and legal fees, these costs are not generally paid by oil spill insurance policies. In addition, spill costs are somewhat fluid and accrue over time, making it sometimes difficult to account for the entire cost of a spill at a given time. In particular, the natural resource and third-party damage claims adjudication processes can take many years to complete. Based on consultation with committee staff, we agreed to present the best available data for major oil spills between 1990 and 2006, and we determined that the data gathered were sufficiently reliable for the purposes of our study. Because of the imprecise nature of oil spill cost data, and the use of multiple sources of data, the data described in this report were combined and grouped into cost ranges. Using ranges of costs to provide upper and lower estimates of total costs and damage claims allows us to report data on major oil spills from all reliable sources. To establish the universe of vessel spills that have exceeded $1 million in total removal costs and damage claims since 1990, we used—in consultation with oil spill experts—a combination of readily available data and reasoned estimation. Since federal government cost data are available, we first established an estimate of the probable share of spill costs between the federal government and the private sector to determine what amount of federal costs might roughly indicate the total costs were over $1 million. We interviewed Environmental Research Consulting, as well as agency officials from the NPFC and NOAA, to determine a reasonable estimated share of costs between the private and public sectors. The officials with whom we spoke estimated that in general, at least 90 percent of all spill costs are typically paid by the private sector. Based on that estimation, any spill with at least $100,000 in federal oil spill removal costs and damage claims probably cost at least $1 million in total—-that is, 90 percent of the total costs being paid by the private sector, and the remaining 10 percent paid by the public sector. Therefore, we initially examined all spills with at least $100,000 in federal oil spill removal costs and damage claims. We obtained these data on federal oil spill removal costs and damage claim payments from the NPFC. Of 3,389 federally managed spills since 1990, there were approximately 184 spills where the federal costs exceeded $100,000. From this group of spills, we limited our review to spills that occurred after the enactment of OPA on August 18, 1990. Additionally, we omitted (1) spill events in which costs were incurred by the federal government for measures to prevent a spill although no oil was actually spilled and (2) spills of fewer than 100 gallons, where, according to the NPFC, the likelihood of costs exceeding $1 million was minimal. Lastly, in consultation with Environmental Research Consulting, we used estimated spill costs and additional research to determine spills that were unlikely to have had total costs and claims above $1 million. Through this process, we concluded that since the enactment of OPA, 51 spills have had costs and claims that have exceeded $1 million. To assess the costs of oil spills based on various factors, we collected data from federal government, private sector, and a consultant, and combined the data into ranges. In addition to collecting data on removal costs and damage claims, we collected additional information on major oil spills. We categorized and grouped spill costs based on the vessel type, time of year, location, and oil type to look for discernable trends in costs based on these characteristics. We collected information on the limits of liability of the vessels at the time of the spill and the limits of liability for vessels after changes in liability limits in the Coast Guard and Maritime Transportation Act of 2006. In addition, to analyze the effects of inflation on the Fund and liability limits, using the Consumer Price Index, we calculated what the limits of liability would have been at the time of each spill if the OPA- stipulated limits had been adjusted for inflation. We used the Consumer Price Index as the basis for inflationary measures because OPA states that limits should be adjusted for “significant increases in the Consumer Price Index.” In reporting spill cost data by year and by certain categories, we use ranges, including the best available data. For certain statistics, such as the public-sector/private-sector cost share, where costs are aggregated for all spills, we calculated percentages based on the mid-point of the cost ranges. To test the reliability of using the mid-point of the ranges, we performed a sensitivity test, analyzing the effects of using mid-point versus the top and bottom of the cost range. We determined that presenting the certain figures based on the mid-point of the ranges is reliable and provides the clearest representation of the data. To supplement our data analysis and in order to determine the factors that affect the costs of major oil spills, we interviewed officials from the NPFC, NOAA, and EPA regarding the factors that affect major oil spill costs. We also interviewed state officials responsible for oil spill response from Alaska, California, New York, Rhode Island, Texas, and Washington to determine the types of costs incurred by states when responding to oil spills and the factors that affect major oil spills costs. Additionally, we interviewed industry experts and a vessel insurer about the factors that affect major oil spill costs. To determine the implications of major oil spills on the Fund, we interviewed agency officials from the NPFC and the Coast Guard as well as vessel insurers and industry experts to get the private sector’s perspective on the major oil spills’ impact on the Fund. In addition, we reviewed recent Coast Guard reports to Congress on the status of the Fund and limits of liability. Lastly, we conducted in-depth reviews of five oil spills. The spills were selected to represent a variety of factors that potentially affect the costs of spills—geography, oil type, and spill volume. During this review, we interviewed the NPFC case officers who were involved with each spill, state agency officials; insurance companies; and private-sector companies, such as oil spill response organizations that were involved in the spill and the spill response. To the best of our ability, we attempted to interview the responsible parties involved in each spill. We were able to speak with one vessel operator. Our interviews were designed to gain perspectives on the response effort for each spill, the factors that contributed to the cost of the spill, and what actual costs were incurred by the responsible party. Finally, we also conducted a file review of NPFC records of federal response activities, removal costs, and damage claims made to the Fund for each of the five spills we reviewed in-depth. We conducted our review from July 2006 through August 2007 in accordance with generally accepted government auditing standards, including standards for data reliability. In addition to the contact named above, Nikki Clowers, Assistant Director; Michele Fejfar; Simon Galed; H. Brandon Haller; David Hooper; Anne Stevens; Stan Stenersen; and Susan Zimmerman made key contributions to this report.

When oil spills occur in U.S. waters, federal law places primary liability on the vessel owner or operator--that is, the responsible party--up to a statutory limit. As a supplement to this "polluter pays" approach, a federal Oil Spill Liability Trust Fund administered by the Coast Guard pays for costs when a responsible party does not or cannot pay. The Coast Guard and Maritime Transportation Act of 2006 directed GAO to examine spills that cost the responsible party and the Fund at least $1 million. This report answers three questions: (1) How many major spills (i.e., $1 million or more) have occurred since 1990, and what is their total cost? (2) What factors affect the cost of spills? and (3) What are the implications of major oil spills for the Oil Spill Liability Trust Fund? GAO's work to address these objectives included analyzing oil spill costs data, interviewing federal, state, and private-sector officials, and reviewing Coast Guard files from selected spills. On the basis of cost information collected from a variety of sources, GAO estimates that 51 spills with costs above $1 million have occurred since 1990 and that responsible parties and the federal Oil Spill Liability Trust Fund (Fund) have spent between about $860 million and $1.1 billion for oil spill removal costs and compensation for damages (e.g., lost profits and natural resource damages). Responsible parties paid between about 72 percent and 78 percent of these costs; the Fund has paid the remainder. Since removal costs and damage claims may stretch out over many years, the costs of the spills could rise. The 51 spills, which constitute about 2 percent of all vessel spills since 1990, varied greatly from year to year in number and cost. Three main factors affect the cost of spills: a spill's location, the time of year, and the type of oil spilled. Spills that occur in remote areas, for example, can increase costs involved in mobilizing responders and equipment. Similarly, a spill occurring during tourist or fishing season might produce substantial compensation claims, while a spill occurring during another time of year may not be as costly. The type of oil affects costs in various ways: fuels like gasoline or diesel fuel may dissipate quickly but are extremely toxic to fish and plants, while crude oil is less toxic but harder to clean up. Each spill's cost reflects a unique mix of these factors. To date, the Fund has been able to cover costs from major spills that responsible parties have not paid, but risks remain. Specifically, the Coast Guard and Maritime Transportation Act of 2006 increased liability limits, but GAO's analysis shows the new limit for tank barges remains low relative to the average cost of such spills. Since 1990, the Oil Pollution Act required that liability limits be adjusted above the limits set forth in statute for significant increases in inflation, but such changes have never been made. Not making such adjustments between 1990 and 2006 potentially shifted an estimated $39 million in costs from responsible parties to the Fund.

The commercial space launch industry continues to develop and evolve, with changes in technology and facilities. Historically, commercial space launches carried payloads, generally satellites, into orbit using expendable launch vehicles that did not return to earth. Figure 1 shows examples of expendable launch vehicles. However, launch companies are testing reusable elements of expendable launch vehicles. For example, after launch, SpaceX has recovered four Falcon 9 first stages— three on a barge located at sea and one on land, according to FAA. United Launch Alliance is also developing capabilities to reuse the first stage of its Vulcan launch vehicle. Since the Space Shuttle fleet was retired in 2011, NASA has procured commercial cargo transportation services to the International Space Station from commercial providers such as SpaceX and Orbital ATK on these types of vehicles. In addition, the commercial space launch industry is further changing technology with the emergence of suborbital reusable launch vehicles that are capable of being launched into space more than once and could be used for space tourism. Several companies such as Virgin Galactic, Blue Origin, and XCOR are in the process of developing and testing manned, reusable launch vehicles for commercial space tourism. For example, according to Blue Origin it has launched, recovered, and re- flown the same booster four times. Companies like Virgin Galactic and Stratolaunch Systems are also developing vehicles that will have the capability to launch small satellites into orbit. See figure 2. Further, private companies and states are developing commercial spaceports—sites used for commercial space launches to support the expected growth in the launch industry. See figure 3. FAA’s primary means of authorizing space launch activities is through its licensing process which includes: licensing launch and reentry vehicle operations, reviewing applications for experimental permits, reviewing safety approvals, and conducting safety inspections and oversight of licensed and permitted activities, among other activities. For fiscal year 2016 for the Office of Commercial Space Transportation, FAA’s budget request was $18.1 million and 92 full-time equivalent positions. Congress provided $17.8 million for commercial space activities for fiscal year 2016. The federal government is authorized to provide catastrophic loss protection in the event of a launch accident for all FAA-licensed commercial launches through the Commercial Space Launch Amendments Act as amended. Thus, subject to congressional appropriations, the U.S. government may pay third-party liability claims for injury, damage, or loss that result from a commercial launch-related accident in excess of the required “maximum probable loss,” an amount which is calculated by FAA and is capped at $500 million per launch. The federal government, subject to the availability of appropriations, is then liable for claims over the maximum probable loss up to $1.5 billion which when adjusted for post-1988 inflation is about $3.06 billion in 2015 dollars. Launch companies are responsible for third-party liability claims up to the maximum probable loss and over $3.06 billion. Figure 4 illustrates this regime. FAA’s risk methodology to calculate the maximum probable loss uses an “overlay” method that entails reviewing the specific circumstances of the launch including the planned launch vehicle, launch site, payload, flight path, and the potential casualties and fatalities that could result from varying types of launch failures at different points along that path. There has not been a commercial launch-related accident that has invoked indemnification and thus the federal government has not paid any third-party liability claims to date. As we reported in our 2015 report, during the last decade, U.S. companies conducted fewer orbital commercial launches in total than companies in Russia or Europe, which are among the main foreign competitors. However, in recent years such as 2014 and 2015, U.S. companies have conducted an increasing number of orbital commercial launches. As shown in figure 4, the number of orbital launches conducted by U.S. companies varied over the last 11 years. For example, recently the number of launches increased from zero in 2011 to eight in 2015. In 2015, U.S. companies conducted more orbital launches than companies in Russia, which conducted five, or Europe, which conducted six. In 2015 we found that a number of factors are responsible for the recent expansion of the U.S. commercial space launch industry. First, increase in demand through federal government contracts, such as NASA’s commercial cargo program, have supported the industry and have resulted in an increase in the number of U.S. commercial launches. For example, in 2015, SpaceX conducted three cargo resupply missions for NASA. NASA also procured eight launches from Orbital ATK in 2008 that were scheduled to occur between 2014 and 2016 with one launch taking place in 2015, one launch taking place in 2016, and another scheduled for July 2016. In addition, in January 2016, NASA announced its selections for companies to conduct Commercial Resupply Services (CRS2) to the ISS. SpaceX and Orbital ATK were selected again, and Sierra Nevada Corporation was added as a new participant. According to NASA, these awards require a minimum of six missions to the ISS from each participant between 2019 and 2024. In addition to fulfilling government contracts, these companies also conduct launches for other customers, including international customers. Second, according to representatives from two commercial space launch companies, including SpaceX, and an advisory group and an expert whom we interviewed for our 2015 report, the growth in the U.S. commercial space launch industry is largely due to SpaceX because it is more price competitive compared with foreign launch providers. The Chairman of the Commercial Space Transportation Advisory Committee said that SpaceX’s prices are significantly lower than foreign providers. Some companies are seeking ways to further reduce costs. For example, Blue Origin is developing new main engine elements for United Launch Alliance’s expendable launch vehicle. Representatives from one company and an industry association and an expert told us that reusable stages may further lower launch prices. In previous work, we reported that— according to industry stakeholders—launch prices, along with launch vehicle reliability, were the major factors that customers focus on when selecting launch providers. Third, the emerging space tourism industry and small satellite industry in the United States also may help the U.S. commercial space launch industry expand. As noted earlier, some U.S. companies are developing launch vehicles to carry spaceflight participants on suborbital flights and to place small satellites into orbit. In our 2015 report, we asked FAA officials, representatives from nine commercial space launch companies, and three experts to identify the challenges that FAA faces—and is likely to face in the near future—to address significant developments in the commercial space launch industry over the last decade. The challenges for FAA that they identified included: (1) determining whether and when to regulate the safety of crew and spaceflight participants and (2) handling an increased workload relating to licensing and permitting launches and launch sites. In addition, in our 2015 report, we noted that changes in the number and types of commercial space launches could affect the government’s overall exposure and indemnification for launches. Determining whether and when to regulate the safety of crew and spaceflight participants: In 2014, FAA released a set of recommended practices on human spaceflight occupants’ safety that the agency indicated could be a starting point for the industry to develop standards, or if needed, for FAA to develop regulations. In 2015, we reported that FAA officials said that the agency did not have plans to issue regulations regarding the safety of crew and spaceflight participants but was looking to industry to develop industry consensus standards detailing validation and verification criteria that are needed to implement the agency’s recommended practices. As part of the U.S. Commercial Space Launch Competitiveness Act, Congress required FAA in consultation with an industry advisory group—the Commercial Space Transportation Advisory Committee—to submit two reports to Congress on this topic. The first report is on metrics that could indicate FAA’s and the industry’s readiness to transition to a safety framework that may include regulating crew, government astronaut and spaceflight participant safety and is due by August 2016. The second report is on the industry’s progress in developing voluntary industry consensus standards and is required to be submitted by December 31, 2016 and periodically afterwards until December 31, 2021. Increased workload relating to licensing and permitting launches and launch sites: Licensing more launches: In fiscal year 2015, FAA licensed and permitted 14 launches and re-entries, up from seven in fiscal year 2006 and compared with an average of about 11 launches and re- entries during each fiscal year from 2006 to 2015. We found a large part of this increase was due to launches for NASA’s commercial cargo program. In the future, FAA also will need to license launches for NASA’s commercial crew program and potentially launches of companies placing small satellites in orbit. Conducting more inspections: In fiscal year 2015, FAA conducted 216 commercial launch inspections, up from 27 in fiscal year 2006 and compared with an average of 90 inspections during each fiscal year from 2006 to 2015. Officials said that FAA has conducted more safety inspections, especially those associated with pre-launch and reentry activities, to allow the agency to identify safety issues early for correction and to avoid launch companies’ noncompliance with regulations and the conditions set forth in the launch license. FAA conducts different types of inspections such as launch and reentry operations and launch site operations, and FAA inspectors are present at launches. Licensing new types of vehicles and technologies: Companies are developing a variety of new vehicles and technologies. For example, the space tourism industry is developing hybrid launch systems such as SpaceShipTwo, which have elements of both aircraft and rocket-powered components. Some companies are also testing autonomous flight safety systems, which would allow a launch vehicle that is off course to be terminated without humans taking action. Most licensed launches as of August 2015 have involved flight termination systems that were human- operated. Licensing more and complex launch sites: Although launch sites traditionally have been located in coastal areas at federal launch facilities, in 2014 FAA licensed an inland launch site that is co- located with a commercial airport in Midland, Texas. In addition, FAA is licensing more nonfederal launch sites. As of June 2015, there were 10 FAA-licensed commercial launch sites, compared with six in 2006. In addition, as of May 2015, FAA had received partial applications for four additional launch sites. Also, in our 2015 report we noted that changes in the number and types of commercial space launches could affect the government’s overall exposure and indemnification for launches for several reasons. First, the number of launches and reentries covered by federal indemnification is forecasted to increase and the federal government’s potential exposure to third-party liability claims would increase with the added volume. In general, by increasing the volume of launches and reentries, the probability of a catastrophic accident occurring is also increased. A catastrophic accident could result in third-party losses over the maximum probable loss, which would invoke federal indemnification. Second, forecasted types of launches and reentries include newly developed launch vehicles that have a shorter launch history than “legacy” launch vehicles. For example, Virgin Galactic’s SpaceShipTwo, XCOR’s Aerospace’s Lynx, and Blue Origin’s New Shepard are new vehicles. However, increased flights of a launch vehicle could also make a vehicle more reliable. We have previously reported that although some industry changes may alter the government’s exposure, an accurate maximum probable loss calculation will mitigate the effects to some extent. If the maximum probable loss calculation is accurate, the estimated losses will adjust for the risk profile of each license, in such a way that the likelihood the government would indemnify a third-party remains the same regardless of the industry change. However, in July 2012, we reported that FAA’s risk methodology—which was first established in the 1980s— could be updated given advances in catastrophe modeling. We recommended that FAA review its maximum probable loss methodology. Congress mandated that FAA review the methodology and report back to the Congress by May 2016. FAA officials told us that in June 2016 that they have drafted a report which is currently under agency review. In 2015 we found that FAA’s budget requests for its commercial space launch activities generally were based on the number of projected launches, but that in recent years the actual number of launches was much lower than FAA’s projections. For example, during 6 of the 10 years from fiscal years 2005 to 2014, FAA generally based its budget submissions on the number of launches that it was projecting for the following year; none of those projections was realized in the actual number of licensed and permitted launches. FAA officials said at that time that although other metrics existed besides the number of projected launches, they were not consistently used in the agency’s budget submissions. In addition, other activities, such as time spent on pre- application license consultations, were not included in the metrics used in preparing the budget requests. According to FAA officials, more detailed information was not provided in their budget submissions because the agency lacked certain workload metrics regarding its commercial space launch oversight activities. We also found that the Office of Commercial Space Transportation did not track the amount of time spent on the office’s various activities. However, the officials indicated that they were continuing to develop a labor analysis methodology that began in fiscal year 2014 and that the office was considering implementing a new time recordkeeping system in 2016 to supplement the development of additional workload metrics. To provide Congress with more information about the resources requested to address developments in the commercial space launch industry, we recommended that FAA provide more detailed information in its budget submissions about its workload. FAA agreed with the recommendation, but DOT also had some concerns about how issues were presented. FAA has taken steps to implement our recommendation. In the 2017 budget submission, FAA provided workload indices based on the number of authorizations which the agency uses to authorize companies to conduct one or more launches, the number of licenses and permits, the number of on-site inspections as part of licensing launch sites, and staffing levels since fiscal year 2006. We will continue to monitor FAA’s progress toward implementing this recommendation. Chairman LoBiondo, Ranking Member Larsen and Members of the Subcommittee, this concludes my prepared statement. I would be pleased to answer any questions at this time. For further information on this testimony, please contact Gerald. L. Dillingham, Ph.D., at (202) 512-2834 or dillinghamg@gao.gov. In addition, contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Individuals making key contributions to this testimony include: Catherine Colwell, Bob Homan, Dave Hooper, Maureen Luna-Long, Stephanie Purcell, Namita Bhatia Sabharwal, and Travis Schwartz. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

The U.S. commercial space launch industry has changed considerably since the enactment of the Commercial Space Launch Amendments Act of 2004. FAA is required to license or permit commercial space launches; however, to allow space tourism to develop, the act prohibited FAA from regulating crew and spaceflight participant safety before 2012—a moratorium that was extended to 2023. The U.S. Commercial Space Launch Competitiveness Act, enacted in November 2015, addressed other aspects of the commercial space launch industry. This testimony summarizes and updates findings from GAO's 2015 report, specifically industry developments and FAA challenges, including FAA's launch licensing workload and budget. For its 2015 report, GAO reviewed FAA's guidance on its launch permit, licensing, and safety oversight activities; interviewed FAA officials, industry stakeholders, and experts who were selected on the basis of their knowledge of FAA's oversight of the commercial space launch industry; and visited spaceports where two 2014 launch mishaps occurred. To update this information GAO reviewed FAA information on the industry and FAA's budget request. In 2015, GAO reported that during the last decade, U.S. commercial space launch companies conducted fewer orbital launches in total than companies in Russia or Europe, which are among their main foreign competitors. However, the U.S. commercial space launch industry has expanded recently. In 2015, U.S. companies conducted eight orbital launches, compared with none in 2011. In addition, in 2015, U.S. companies conducted more orbital launches than companies in Russia, which conducted five, or Europe, which conducted six. In 2015, GAO reported that the Federal Aviation Administration (FAA)—which is responsible for protecting the public with respect to commercial space launches, including licensing and permitting launches—faces challenges. According to FAA officials and industry stakeholders, FAA faces an increasing workload licensing and permitting launches for transporting cargo, and in the future, crew for NASA's commercial space programs, space tourism, and potentially launching small satellites. FAA also faces the challenges of whether and when to regulate the safety of crew and spaceflight participants—in 2015 Congress extended the moratorium to 2023—and overseeing new types of vehicles and technologies. (See figure for commercial space launch vehicles.) Challenges also include updating FAA's method to calculate maximum probable loss—the amount above which the federal government indemnifies the industry for catastrophic loss. Virgin Galactic's SpaceShipTwo and SpaceX's Falcon 9 GAO reported in 2015 that FAA's budget requests for its commercial space launch activities generally were based on the number of projected launches, but that in recent years the actual number of launches was much lower than FAA's projections. GAO also reported that, according to FAA officials, more detailed information was not provided in FAA's budget submissions because the agency lacked information on its workload overseeing commercial space launch activities. In addition, GAO reported that the Office of Commercial Space Transportation did not track the amount of time spent on various activities. FAA has taken steps to implement GAO's recommendation that it provide more detailed information in its budget submissions regarding commercial space transportation activities. In its 2017 budget submission, FAA provided workload indices regarding authorizations under which companies conduct one or more launches; on-site inspections; licensing of spaceports; and staffing levels since 2006.Why GAO Did This Study In 2015, GAO recommended that FAA, in its budget submissions, provide more detailed information about the Office of Commercial Space Transportation's workload. FAA agreed with the recommendation. GAO is not making new recommendations in this testimony.

The 1987 Act requires that an individual or entity be actively engaged in farming in order to receive farm program payments. To be considered actively engaged in farming, the act requires an individual or entity to provide a significant contribution of capital, land, or equipment, as well as a significant contribution of personal labor or active personal management to the farming operation. Hired labor or hired management may not be used to meet the latter requirement. The act’s definition of a “person” eligible to receive farm program payments includes an individual, as well as certain kinds of corporations, partnerships, trusts, or similar entities. Recipients must also demonstrate that their contributions to the farming operation are in proportion to their share of the operation’s profits and losses and that these contributions are at risk. The 1987 Act also limits the number of entities through which a person can receive program payments. Under the act, a person can receive payments as an individual and through no more than two entities, or through three entities and not as an individual. The statutory provision imposing this limit is commonly known as the three-entity rule. Under the Farm Security and Rural Investment Act of 2002, “persons”—individuals or entities—are generally limited to a total of $180,000 annually in farm program payments, or $360,000 if they are members of up to three entities. Some farming operations may reorganize to overcome payment limits to maximize their farm program benefits. Larger farming operations and farming operations producing crops with high payment rates, such as rice and cotton, may establish several related entities that are eligible to receive payments. However, each entity must be separate and distinct and must demonstrate that it is actively engaged in farming by providing a significant contribution of capital, land or equipment, as well as a significant contribution of personal labor or active personal management to the farming operation. Within USDA, the Farm Service Agency (FSA) is responsible for enforcing the actively engaged in farming and payment limitation rules. FSA field offices review a sample of farming plans at the end of the year to help monitor whether farming operations were conducted in accordance with approved plans, including whether payment recipients met the requirement for active engagement in farming and whether the farming operations have the documents to demonstrate that the entities receiving payments are in fact separate and distinct legal entities. FSA selects its sample of farming operations based on, among other criteria, (1) whether the operation has undergone an organizational change in the past year by, for example, adding another entity or partner to the operation and (2) whether the operation receives payments above a certain threshold. These criteria have principally resulted in sampling farming operations in areas that produce cotton and rice—Arkansas, California, Louisiana, Mississippi, and Texas. Many recipients meet one of the farm program payments’ eligibility requirements by asserting that they have made a significant contribution of active personal management. Because FSA regulations do not provide a measurable, quantifiable standard for what constitutes a significant management contribution, people who appear to have little involvement are receiving farm program payments, according to our survey of FSA field offices and our review of 86 case files. Indeed, most large farming operations meet the requirement for personal labor or active personal management by asserting a significant contribution of management. Survey respondents provided information on 347 partnerships and joint ventures for which FSA completed compliance reviews in 2001; these entities comprised 992 recipients, such as individuals and corporations that were members of these farming operations. Of these 992 recipients, 46 percent, or 455, asserted that they contributed active personal management; 1 percent, or 7, asserted that they contributed personal labor; and the remaining 53 percent (530) asserted they provided a combination of active personal management and personal labor to meet the actively engaged in farming requirement. While FSA’s regulations define active personal management more specifically to include such things as arranging financing for the operation, supervising the planting and harvesting of crops, and marketing the crops, the regulations lack measurable criteria for what constitutes a significant contribution of active personal management. FSA regulations define a “significant contribution” of active personal management as “activities that are critical to the profitability of the farming operation, taking into consideration the individual’s or entity’s commensurate share in the farming operation.” In contrast, FSA provides quantitative standards for what constitutes a significant contribution of active personal labor, capital, land, and equipment. For example, FSA’s regulations define a significant contribution of active personal labor as the lesser of 1,000 hours of work annually, or 50 percent of the total hours necessary to conduct a farming operation that is comparable in size to such individual’s or entity’s commensurate share in the farming operation. By not specifying quantifiable standards for what constitutes a significant contribution of active personal management, FSA allows recipients who may have had limited involvement in the farming operation to qualify for payments. Some recipients appeared to have little involvement with the farming operation for 26 of the 86 FSA compliance review files we examined in which the recipients asserted they made a significant contribution of active personal management to the farming operation. For example, in 2001, 11 partners in a general partnership operated a farm of 11,900 acres. These partners asserted they met the actively engaged in farming requirement by making a significant contribution of equipment and active personal management. FSA’s compliance review found that all partners of the farming operation were actively engaged in farming and met all requirements for the approximately $1 million the partnership collected in farm program payments in 2001. However, our review found that the partnership held five management meetings during the year, three in a state other than the state where the farm was located, and two on-site meetings at the farm. Some of the partners attended the meetings in person while others joined the meetings by telephone conference. Although all 11 partners claimed an equal contribution of management, minutes of the management meetings indicated seven partners participated in all five meetings, two participated in four meetings, and two participated in three meetings. All partners resided in states other than the state where the farm was located, and only one partner attended all five meetings in person. Based on our review of minutes documenting the meetings, it is unclear whether some of the partners contributed significant active personal management. If FSA had found that some of the partners had not contributed active personal management, the partnership’s total farm program payments would have been reduced by about 9 percent, or $90,000, for each partner that FSA determined was ineligible. State FSA officials agreed that the evidence to support the management contribution for some partners was questionable and that FSA reviewers could have taken additional steps to confirm the contributions for these partners. According to our survey of 535 FSA field offices, FSA could make key improvements to strengthen the management contribution standard. These offices reported that the management standard can be strengthened by clarifying the standard, including providing quantifiable criteria, certifying actual contributions, and requiring management to be on-site. More than 60 percent of those surveyed, for example, indicated that clarifying the standard would be an improvement. In addition, in 2003, a USDA commission established to look at the impact of changes to payment limitations concluded that determining what constitutes a significant contribution of active management is difficult and lack of clear criteria likely makes it easier for farming operations to add recipients in order to avoid payment limitations. We also found that some individuals or entities have engaged in transactions that might constitute schemes or devices to evade payment limitations, but neither FSA’s regulations nor its guidance address whether such transactions could constitute schemes or devices. Under the 1987 Act as amended, if the Secretary of Agriculture determines that any person has adopted a “scheme or device” to evade, or that has the purpose of evading, the act’s provisions—in other words, the payment limitations—then that person is not eligible to receive farm program payments for the year the scheme or device was adopted and the following crop year. According to FSA’s regulations, this statutory provision includes (1) persons who adopt or participate in adopting a scheme or device and (2) schemes or devices that are designed to evade or have “the effect of evading” payment limitation rules. The regulations state that a scheme or device shall include concealing information that affects a farm program payment application, submitting false or erroneous information, or creating fictitious entities for the purpose of concealing the interest of a person in a farming operation. We found several large farming operations that were structured as one or more partnerships, each consisting of multiple corporations that increased farm program payments in a questionable manner. The following two examples illustrate how farming operations, depending on how the FSA regulations are interpreted, might be considered to evade, or have the effect of evading, payment limitations. In one case, we found that a family had set up the legal structures for its farming operation and also owned the affiliated nonfarming entities. This operation included two farming partnerships comprising eight limited liability companies. The two partnerships operated about 6,000 acres and collected more than $800,000 in farm program payments in 2001. The limited liability companies included family and non-family members, although power of attorney for all of the companies was granted to one family member to act on behalf of the companies, and ultimately the farming partnerships. The operation also included nonfarming entities—nine partnerships, a joint venture, and a corporation—that were owned by family members. The affiliated nonfarming entities provided the farming entities with goods and services, such as capital, land, equipment, and administrative services. The operation also included a crop processing entity to purchase and process the farming operation’s crop. According to our review of accounting records for the farming operation, both farming partnerships incurred a small net loss in 2001, even though they had received more than $800,000 in farm program payments. In contrast, average net income for similar- sized farming operations in 2001 was $298,000, according to USDA’s Economic Research Service. The records we reviewed showed that the loss occurred, in part, because the farming operations paid above-market prices for goods and services and received a net return from the sale of the crop to the nonfarming entities that appeared to be lower than market prices because of apparent excessive charges. The structure of this operation allowed the farming operation to maximize farm program payments, but because the farm operated at a loss these payments were not distributed to the members of the operation. In effect, these payments were channeled to the family-held nonfarming entities. Figure 1 shows the organizational structure of this operation and the typical flow of transactions between farming and nonfarming entities. Similarly, we found another general partnership that farmed more than 50,000 acres in 2001 and that conducted business with nonfarming entities, including a land leasing company, an equipment dealership, a petroleum distributorship, and crop processing companies, with close ties to the farming partnership. The partnership, which comprised more than 30 corporations, collected more than $5 million in farm program payments in 2001. The shareholders who contributed the active personal management for these corporations were officers of the corporations. Each officer provided the active personal management for three corporations. Some of these officers were also officers of the nonfarming entities—the entities that provided the farming partnership goods and services such as the capital, land, equipment, and fuel. The nonfarming entities also included a gin as well as grain elevators to purchase and process the farming partnership’s crops. Our review of accounting records showed that even though the farming partnership received more than $5 million in farm payments, it incurred a net loss in 2001, which was distributed among the corporations that comprised the partnership. As in the first example, factors contributing to the loss included the above- market prices for goods and services charged by the nonfarming entities and the net return from the sale of crops to nonfarming entities that appeared to be lower than market prices because of apparent excessive charges for storage and processing. For example, one loan made by the nonfarming financial services entity to the farming partnership for $6 million had an interest rate of 10 percent while the prevailing interest rate for similar loans at the time was 8 percent. Similarly, the net receipts from the sale of the harvested crop, which were sold almost exclusively to the nonfarming entities, were below market price. For example, in one transaction the gross receipt was about $1 million but after the grain elevators deducted fees for the quality of the grain and such actions as drying and storing the grain, the net proceeds to the farming entity were only about $500,000. In this particular operation, all of the nonfarming entities had common ownership linked to one individual. This individual had also set up the legal structure for the farming entities but had no direct ownership interest in the farming entities. It is unclear whether either of these operations falls within the statutory definition of a scheme or device or whether either otherwise circumvents the payment limitation rules. State FSA officials in Arkansas, Louisiana, Mississippi, and Texas, where many of the large farming operations are located, believed that some large operations with relationships between the farming and nonfarming entities were organized primarily to circumvent payment limitations. In this manner, these farming operations may be reflective of the organizational structures that some Members of Congress indicated were problematic when enacting the 1987 Act and the scheme or device provision. The House Report for the 1987 Act states: “A small percentage of producers of program crops have developed methods to legally circumvent these limitations to maximize their receipt of benefits for which they are eligible. In addition to such reorganizations, other schemes have been developed that allow passive investors to qualify for benefits intended for legitimate farming operations.” In our discussions with FSA headquarters officials in February 2004 on the issue of farming operations that circumvent the payment limitation rules, they noted that while an operation may be legally organized, it may be misrepresenting who in effect receives the farm program payments. FSA has no data on how many of the types of operations that we identified exist. However, FSA is reluctant to question these operations because it does not believe current regulations provide a sufficient basis to take action. Other FSA officials said that USDA could review such an operation under the 1987 Act’s scheme or device provision if it becomes aware that the operation is using a scheme or device for the purpose of evading the payment limitation rules. However, these FSA officials stated it is difficult to prove fraudulent intent—which they believe is a key element in proving scheme or device—and requires significant resources to pursue such cases. In addition, they stated that even if FSA finds a recipient ineligible to receive payments, its decision might be overturned on appeal within USDA. The FSA officials noted that when FSA loses these types of cases, the loss tends to discourage other field offices from aggressively pursuing these types of cases. It is not clear whether either the statutory provision or FSA’s regulations require a demonstration of fraudulent intent in order to find that someone has adopted a scheme or device. As discussed above, the statute limits payments if the Secretary of Agriculture determines that any person has adopted a scheme or device “to evade, or that has the purpose of evading,” the farm payment limitation provisions. The regulations state that payments may be withheld if a person “adopts or participates in adopting a scheme or device designed to evade or that has the effect of evading” the farm payment limitations. The regulations note that schemes or devices shall include, for example, creating fictitious entities for the purpose of concealing the interest of a person in a farming operation. Some have interpreted this provision as appearing to require intentionally fraudulent or deceitful conduct. On the other hand, FSA regulations only provide this as one example of what FSA considers to be a scheme or device. The regulations do not specify that all covered schemes or devices must involve fraudulent intent. As previously stated, covered schemes or devices under FSA regulations include those that have “the effect of evading” payment limitation rules. Finally, guidance contained in FSA Handbook Payment Limitations, 1-PL (Revision 1), Amendment 40, does not clarify the matter because it does not provide any additional examples for FSA officials of the types of arrangements that might be considered schemes or devices. This lack of clarity over whether fraudulent intent must be shown in order for FSA to deny payments under the scheme or device provision of the law may be inhibiting FSA from finding that some questionable operations are schemes or devices. In addition to the weaknesses described above, FSA does not effectively oversee farm program payments in five key areas, according to our analysis of FSA compliance reviews and our survey of FSA field offices. First, FSA does not review a valid sample of recipients to be reasonably assured of compliance with the payment limitations. In 2001, FSA selected 1,573 farming operations from its file of 247,831 entities to review producers’ compliance with actively engaged in farming requirements. FSA’s sample selection focuses on entities that have undergone an organizational change during the year or received large farm program payments. Field staff responsible for these reviews seek waivers for farming operations reviewed within the last 3 to 5 years—the time frame varies by state. As a result, according to FSA officials, of the farming operations selected for review each year, more than half are waived and therefore not actually reviewed. Many of the waived cases show up year after year because FSA’s sampling methodology does not take into consideration when an operation was last reviewed. In 2001, the latest year for which data are available, only 523 of 1,573 sampled entities were to be reviewed. Field offices sought and received waivers for 966 entities primarily because the entities were previously reviewed or the farming operation involved only a husband and wife. According to FSA headquarters officials, the sampling process was developed in the mid- 1990s and it can be improved and better targeted. Second, field offices do not always conduct compliance reviews in a timely manner. Only 9 of 38 FSA state offices responsible for conducting compliance reviews for 2001 completed the reviews and reported the results to FSA headquarters within 12 months, as FSA policy requires. FSA headquarters selected the 2001 sample on March 27, 2002, and forwarded the selections to its state offices on April 4, 2002. FSA headquarters required the state offices to conduct the compliance reviews and report the results by March 31, 2003. Six of the 26 FSA state offices that failed to report the results to headquarters had not yet begun these reviews for 470 farming operations as of summer 2003: Arkansas, California, Colorado, Louisiana, Ohio, and South Carolina. Until we brought this matter to their attention in July 2003, FSA headquarters staff were unaware that these six states had not conducted compliance reviews for 2001. Similarly, they did not know the status of the remaining 20 states. Because of this long delay, FSA cannot reasonably assess the level of recipients’ compliance with the act and may be missing opportunities to recapture payments that were made to ineligible recipients if a farming operation reorganizes or ceases operations. Third, FSA staff do not use all available tools to assess compliance. For one-half of the case files we reviewed for 2001, field offices did not use all available tools to determine whether persons are actively engaged in farming. FSA compliance review policy requires field staff to interview persons asserting that they are actively engaged in farming before making a final eligibility decision, unless the reason for not interviewing the person is obvious and adequately justified in writing. Indeed, 83 percent of the field offices responding to our survey indicated that interviews are helpful in conducting compliance reviews. However, in 27 of the 86 case files we reviewed in six states, field staff did not interview these persons and did not adequately document why they had not done so. In one of the states we visited, field staff had not conducted any interviews. We also found that some field offices do not obtain and review certain key financial information regarding the farming operation before making final eligibility decisions. For example, our review of case files indicated that for one-half of the farming operations, field staff did not use financial records, such as bank statements, cancelled checks, or accounting records, to substantiate that capital was contributed directly to the farming operation from a fund or account separate and distinct from that of any other individual or entity with an interest in the farming operation, as required by FSA’s policy. Instead, FSA staff often rely on their personal knowledge of the individuals associated with the farming operation to determine whether these individuals meet the requirement for active engagement in farming. Fourth, FSA does not consistently collect and analyze monitoring data. FSA has not established a methodology for collecting and summarizing compliance review data so that it can (1) reliably compare farming operations’ compliance with the actively engaged in farming requirements from year to year and (2) assess its field offices’ conduct of compliance reviews. Under Office of Management and Budget Circular A-123, agencies must develop and implement management controls to reasonably ensure that they obtain, maintain, report, and use reliable and timely information for decision-making. Because FSA has not instituted these controls, it cannot determine whether its staff are consistently applying the payment eligibility requirements across states and over time. Finally, these problems are exacerbated by a lack of periodic training for FSA staff on the payment limitations and eligibility rules. Training has generally not been available since the mid-1990s. In conclusion, the Farm Program Payments Integrity Act of 1987, while enacted to limit payments to individuals and entities actively engaged in farming, allows farming operations to maximize the receipt of federal farm payments as long as all recipients meet eligibility requirements. However, we found cases where payment recipients may have developed methods to circumvent established payment limitations. This seems contrary to the goals of the 1987 Act and was caused by weaknesses in USDA’s regulation and oversight. The regulations need to better define what constitutes a significant contribution of active personal management and clarify whether fraudulent intent is necessary to find that someone has adopted a scheme or device. Without specifying measurable standards for what constitutes a significant contribution of active personal management, FSA allows individuals who may have had limited involvement in the farming operation to qualify for payments. Moreover, FSA is not providing adequate oversight of farm program payments under its current regulations and policies. In our report to you, we made eight recommendations to the Secretary of Agriculture for improving FSA’s oversight of compliance with the 1987 Act, including: developing measurable requirements defining a significant contribution of active personal management; clarifying regulations and guidance as to what constitutes a scheme or device; improving its sampling method for selecting farming operations for review; and developing controls to ensure all available tools are used to assess compliance with the act. USDA agreed to act on most of our recommendations. However, USDA stated that its current regulations are sufficient for determining active engagement in farming and assessing whether operations are schemes or devices to evade payment limitations. Mr. Chairman, this concludes my prepared statement. We would be happy to respond to any questions that you or other Members of the Committee may have. For further information about this testimony, please contact Lawrence J. Dyckman, Director, Natural Resources and Environment, (202) 512-3841, or by email at dyckmanl@gao.gov. Ron Maxon, Thomas Cook, Cleofas Zapata, Carol Herrnstadt Shulman, and Amy Webbink made key contributions to this statement. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Farmers receive about $15 billion annually in federal payments to help produce major crops, such as corn, cotton, rice, and wheat. The Farm Program Payments Integrity Act of 1987 (1987 Act) limits payments to individuals and entities--such as corporations and partnerships--that are "actively engaged in farming." This testimony is based on GAO's report, Farm Program Payments: USDA Needs to Strengthen Regulations and Oversight to Better Ensure Recipients Do Not Circumvent Payment Limitations ( GAO-04-407 , April 30, 2004). Specifically, GAO (1) determined how well USDA's regulations limit payments and (2) assessed USDA's oversight of the 1987 Act. GAO's survey of USDA's field offices showed that for the compliance reviews the offices conducted, about 99 percent of payment recipients asserted they met eligibility requirements through active personal management. However, USDA's regulations to ensure recipients are actively engaged in farming do not provide a measurable standard for what constitutes a significant contribution of active personal management. By not specifying such a measurable standard, USDA allows individuals who may have limited involvement with the farming operation to qualify for payments. Moreover, USDA's regulations lack clarity as to whether certain transactions and farming operation structures that GAO found could be considered schemes or devices to evade, or that have the purpose of evading, payment limitations. Under the 1987 Act, if a person has adopted such a scheme or device, then that person is not eligible to receive payments for the year in which the scheme or device was adopted or the following year. Because it is not clear whether fraudulent intent must be shown to find that a person has adopted a scheme or device, USDA may be reluctant to pursue the question of whether certain farming operations, such as the ones GAO found, are schemes or devices. According to GAO's survey and review of case files, USDA is not effectively overseeing farm payment limitation requirements. That is, USDA does not review a valid sample of farm operation plans to determine compliance and thus does not ensure that only eligible recipients receive payments, and compliance reviews are often completed late. As a result, USDA may be missing opportunities to recoup ineligible payments. For about one-half of the farming operations GAO reviewed for 2001, field offices did not use available tools to determine whether persons were actively engaged in farming.

Competitive sourcing is a process under which federal agencies subject the performance of their commercial activities to competition among public and private sector sources. It is intended to contribute to cost savings, improved performance, and a better alignment of the agency’s workforce to its mission. OMB’s Circular A-76, Performance of Commercial Activities, establishes federal policy and prescribes the procedures to be used in determining whether commercial activities should be performed by a federal agency or by the private sector. Circular A-76 contains uniform procedures to be used by agencies for calculating costs so that cost comparisons between private sector proposals and government estimates are fair. The Circular mandates use of a standard and consistent process designed to ensure that evaluated costs reflect the full cost of performance by public and private sector sources. This is consistent with the Commercial Activities Panel’s final report, which recommended that these competitions should be conducted on as nearly equal terms as possible, using clear, consistent, and transparent processes for all offerors. As part of this process, the Circular is intended to help ensure that the estimated cost of government performance fairly reflects all of the personnel and non-pay costs of an agency source performing the work. When preparing estimates of government performance, agencies are required to use standard cost factors that are in effect as of the solicitation closing date and make adjustments to reflect changes projected to occur during the performance period. To estimate personnel costs for example, agencies add to basic pay (for full-time and part-time permanent civilian positions) a standard overall costing factor of 32.85 percent to account for fringe benefits. This overall factor is comprised of several components, including a standard cost factor of 5.7 percent to account for life insurance and health benefits as shown in table 1. According to OMB officials, the 5.7 percent factor consists of 0.2 percent for life insurance and 5.5 percent for health benefits. To conduct public-private competitions under Circular A-76, agencies may use either a standard or a streamlined competition process, depending on the number of positions involved. Agencies must use a standard competition process for activities with more than 65 full-time equivalent (FTE) positions. As part of the standard process agencies issue solicitations with a performance work statement describing the work to be performed, appoint an agency tender official to prepare a response to the solicitation based on a “most efficient organization” (MEO), and evaluate that response along with the proposals submitted by private offerors. Also, under the standard competition process, unless contractor performance would save the government $10 million or 10 percent of agency personnel-related costs (whichever is less), the work will be retained within the agency. This ensures that an agency does not convert to contract performance in cases where only marginal savings are anticipated. For activities with 65 or fewer FTEs, agencies may use a streamlined competition process. Streamlined competitions are based only on a comparison of public and private sector costs. Private sector costs are obtained either from documented market research or soliciting cost proposals in accordance with the Federal Acquisition Regulation. Use of the streamlined process enables agencies to complete the comparison more quickly. The Circular was revised in May 2003 based largely on the Commercial Activities Panel’s sourcing principles and recommendations for improving the government’s competitive sourcing processes. Among other things, the panel recommended that the government’s sourcing decisions be based on a clear, transparent, and consistently applied competitive sourcing process. This principle is key to ensuring the integrity of the process, as well as to creating trust in the process on the part of those it most affects: federal managers, users of the services, federal employees, the private sector, and the taxpayers. The revised Circular A-76 states that agencies should centralize oversight responsibility to foster fairness in their public- private competitions, and effectively apply a consistent process based on lessons learned and best practices. The Department of Defense has a long-established competitive sourcing program and is the leader among federal agencies in terms of the number of public-private competitions conducted and positions competed. In fiscal year 2004, DOD reported that it made sourcing decisions in 58 public- private competitions, with projected net savings of approximately $740 million. DOD has a centralized management structure to oversee its competitive sourcing program and those of the DOD components. The Deputy Under Secretary of Defense (Installations and Environment) in the Office of the Secretary of Defense has responsibility for establishing and overseeing DOD-wide policies, procedures, and guidance. DOD components—such as the Army, Navy, Marine Corps, and Air Force—as well as the defense agencies and DOD field activities have their own centralized management structures to operate their competitive sourcing programs based on DOD’s policies, procedures, and guidance. Under legislation applicable only to DOD for activities with more than 10 FTEs, unless contractor performance would save the government $10 million or 10 percent of agency personnel-related costs (whichever is less), the work will not be converted to contractor performance. Federal employees and the employees of the government’s service contractors may receive health insurance benefits based on different statutory requirements. The Federal Employees Health Benefits Act of 1959 established the framework for government civilian employees’ health insurance benefits through the Federal Employees Health Benefits Program (FEHBP). Participation in FEHBP is voluntary for civilian employees and their dependents and retirees. This statute sets the government’s share of each participant’s health insurance premium cost at an amount equal to 72 percent of the weighted average of the premiums of all FEHBP plans, but caps the government’s share at 75 percent of any individual plan’s premium. This formula is applied to the self-alone and self-and-family plans separately. For example, in fiscal year 2005, the government’s annual share of FEHBP premiums for two major FEHBP plans ranged between $2,600 to $3,400 for self coverage and $5,900 to $7,800 for self-and-family coverage. The McNamara-O’Hara Service Contract Act (SCA) of 1965 requires minimum wages and fringe benefits for employees working on government service contracts that exceed $2,500. The Department of Labor administers the SCA and determines the prevailing wages in geographic localities for various job categories. In June 2005, the department increased the standard SCA health and welfare minimum benefit rate to $2.87 per hour from $2.59 per hour. Government contractors have flexibility in the types of health and welfare benefits they provide, as long as they meet or exceed the $2.87 minimum health and welfare requirement. For example, contractors can meet their SCA benefits obligations by providing health insurance benefits, by allowing their employees to place some or all of the SCA benefits in a retirement plan, or by providing cash payments. Most DOD components implemented the health benefit cost provision by ensuring that private sector proposals included an amount for health insurance benefits at least equal to the amount that Circular A-76 requires to be added to agency cost estimates to account for health benefit costs. Under Circular A-76, this amount is 5.5 percent of direct labor costs. The Defense Logistics Agency (DLA), however, used a process based on the monthly premium contributions DOD is required to make towards civilian employees’ health insurance plans under the FEHBP. Either of the processes used by DOD or DLA provides a reasonable approach for ensuring that private offerors do not receive a competitive advantage for less costly health benefits. Use of two different processes, however, results in health benefit costs being treated inconsistently within DOD and could even result in different competitive sourcing outcomes. The health benefit cost requirement established for DOD’s public-private competitions in section 8014(a)(3) of the Department of Defense Appropriations Act, 2005 requires that a private offeror not receive a competitive advantage by not offering health insurance for its employees, or by paying less for employee health benefits than the government contributes for civilian employee health benefits. (See app. III for the text of Section 8014.) To implement this legislation, DOD’s competitive sourcing officials told us they consulted with officials from OMB’s Office of Federal Procurement Policy, DOD’s Office of General Counsel, and DOD components’ competitive sourcing offices to develop their interpretation of the legislation and a process for implementation. Officials from OMB advised DOD that the Circular A-76 standard insurance and health benefits cost factor of 5.7 percent consisted of 0.2 percent to account for the cost of federal employees’ life insurance benefits and 5.5 percent to cover health benefit costs. DOD decided to use the Circular A-76 standard health benefits cost factor as the benchmark for ensuring that the costs of health benefits provided by private offerors are sufficient to comply with the legislation. In November 2004, DOD communicated this approach as the preferred process throughout the department, but gave discretion to the competitive sourcing program offices of the DOD components to use alternate processes, as long as they consulted with DOD’s competitive sourcing office. Under DOD’s process, contracting officials are to follow a multistep approach to implement the health benefit cost comparability provision. First, for competitions conducted subsequent to the issuance of DOD’s guidance, contracting officials should obtain data from the private offeror regarding the company’s costs for contributions to employee health insurance (i.e., benefits) as well as its proposed direct labor costs for the performance of the competed commercial activity. Second, the contracting officials calculate the private offeror’s costs of employee health benefits as a percentage of direct labor costs. Finally, contracting officials make any necessary adjustments to their calculation of the private offeror’s proposed costs using the following criteria: If the health benefit cost percentage is lower than 5.5 percent, then the private offeror’s proposed cost is adjusted upward by the amount necessary to make the contribution equal 5.5 percent. If the percentage contribution is equal to or greater than 5.5 percent, no adjustment is necessary. Hypothetical examples of this process are shown in table 2. According to DOD officials, any health benefit cost adjustment made to the private offeror’s proposed costs is for evaluation and cost comparison purposes only. For work currently performed within the agency, if contracting officials determine that the private offeror has a higher priced proposal than the agency’s cost estimate, either before or after any adjustments for health benefit contributions, DOD will retain the work within the agency. If a private offeror selected for award under the solicitation’s evaluation criteria has a lower cost proposal after any adjustment for health benefit contributions—and contractor performance would save DOD at least $10 million or 10 percent of the agency team’s personnel-related costs—the offeror will be awarded a contract at its original proposed amount. Section 8014 does not compel DOD officials to reject a private offeror’s proposal based solely on the cost or extent of the company’s health benefit coverage. Nor does it require the private offeror to match the DOD’s health benefit costs, since according to DOD officials this would in effect have to be subsidized by DOD through higher awarded costs. DOD officials told us that using the 5.5 percent Circular A-76 cost factor to implement the health benefit cost legislation accomplishes several objectives. First, the approach is consistent with the requirement of the statute that private sector offerors not receive a competitive advantage by offering to pay less for health benefits than what the government pays. Second, DOD officials believe the approach is fair because it ensures that proposals from both the public and private sectors have an equal health benefit cost component of at least 5.5 percent. Third, the approach is consistent with the standard adjustment agency sources already make to account for health benefit costs when preparing agency cost estimates. Fourth, DOD officials said that the process reduces the chances of human error and miscalculations inherent in alternative approaches that might attempt to compare the quality of public and private health benefits. DOD officials commented that it would be difficult to do a true “apples-to- apples” comparison of federal and private sector health benefit plan costs because of the wide variation among federal civilian and private sector plan benefits and employee participation. Finally, according to DOD officials, this process avoids the problem of comparing aggregate employer contribution costs for health benefits, and better accounts for differences in proposed staffing across offers without penalizing a smaller company that may pay less for health benefits overall than the agency source. Except for DLA, which implemented its own process, the DOD components we reviewed adopted DOD’s preferred process to implement the health benefit cost comparability provision. Competitive sourcing program officials in the Air Force, Navy, Marine Corps, and Army Corps of Engineers told us that they have taken actions to implement the DOD process in their fiscal year 2005 competitive sourcing programs. According to these officials, implementation actions ranged from offering instructions to contracting staff about incorporating the DOD preferred process in ongoing competitions to more formal actions such as incorporating the preferred process in competitive sourcing manuals. For example, the Marine Corps’ competitive sourcing program officials added a section with guidance for implementing the health benefit cost provision in its draft competitive sourcing program manual, which contracting officers will use to run Marine Corps public-private competitions. Early in fiscal year 2005, some components took steps to implement the health benefit cost provision in advance of communication from DOD about its preferred process because these components had immediate needs to comply with the requirement in several pending public-private competitions. These early implementation efforts were generally consistent with the preferred process that DOD later communicated in November 2004. For example, Navy and Marine Corps contracting officials told us they issued amendments to ongoing solicitations in which they requested information from private offerors to implement the health benefit provision. This information included whether the offeror would provide an employer-sponsored health insurance plan, the total cost of the employer’s contribution to the plan on behalf of employees, and their direct labor costs to perform the commercial activity being competed. Navy and Marine Corps contracting officials told us that they collected this information in order to compare this offeror information against the standard Circular A-76 insurance and health benefit cost factor. Because of a pending competitive sourcing decision early in fiscal year 2005, DLA also moved ahead and implemented the health benefit cost comparability provision before DOD communicated its preferred process. DLA’s process, which it continues to use, differs from DOD’s and is based on using the monthly premium contributions DOD is required to make under the FEHBP towards civilian employees’ health insurance as the benchmark for comparing private offerors’ health benefit coverage and costs. DLA’s process requires detailed data collection and the use of a complex benefit and cost comparison method. Specifically, for a private offeror to demonstrate that it meets DLA’s health benefit cost comparability benchmark, the company first must provide data showing that its health insurance plan allows employees to enroll either self-alone or self-and-family, and the amount the company contributes towards the plan’s premium cost is at least the lower of the following two benchmarks: (1) the monthly maximum amount of DOD’s premium contribution for self-alone and self- and-family coverage under FEHBP—$298.23 and $646.17, respectively, or (2) 75 percent of the cost of the company plan’s monthly premium, which is the same cap set for any government contributions under the FEHBP. Next, DLA’s process requires that the contracting officer calculate the offeror’s health benefit costs for self-alone and self-and-family coverage, and compare those costs with the agency’s health benefit cost benchmarks under FEHBP, and make any cost adjustments based on the following criteria: If the private offeror’s health benefit plan cost equals or exceeds the lesser of DLA’s two premium contribution benchmarks, no upward adjustment is made to its cost proposal. If the private offeror’s health plan cost does not meet one of DLA’s two premium contribution benchmarks, a “health benefit cost factor” is added to the private offeror’s proposal cost to make up the shortfall. As with DOD’s process, such adjustments, if necessary, are made by DLA only for the purpose of determining compliance with the health benefit cost provision. If the private offeror still has the lower costs after such adjustment and completion of the cost comparison—and meets the minimum $10 million or 10 percent savings margin required for contractor conversion—the private offeror may be awarded a contract at its original proposal amount. In explaining the rationale for this process, DLA officials told us that their interpretation of section 8014 focused on determining that private offerors not receive a competitive advantage when they contribute less towards the premium share than the amount that is paid by DOD for civilian employees’ health benefits under FEHBP. DLA consulted in advance with DOD’s competitive sourcing office, which concurred with DLA’s proposed process for implementation. DOD’s competitive sourcing officials told us that they consider DLA’s process to be more complicated to administer than the preferred process of using the 5.5 percent health benefit cost benchmark. Nevertheless, they told us that DLA’s process is consistent with DOD’s current guidance which allows the use of an alternative process to implement the requirement for a health benefit cost comparison, as long as components consult in advance with DOD. Either of the processes used by DOD or DLA provides a reasonable approach for ensuring that private offerors do not receive a competitive advantage for less costly health benefits. Use of two different processes, however, results in health benefit costs being treated inconsistently within DOD and could even result in different competitive sourcing outcomes. For example, in one of the competitions we reviewed, the company’s contribution for health benefits totaled about 15 percent of its total direct labor costs. Under DOD’s preferred process for determining health benefit cost comparability, the company’s cost proposal would have required no adjustment since the offeror contributes substantially more than the 5.5 percent benchmark. Under DLA’s process, however, the contracting officer found that the private offeror’s share of the health insurance premium fell short of DLA’s benchmark for self-and-family coverage. As a result, the contracting officer added about $280,000 to the private offeror’s cost proposal to make up for the shortfall. Ultimately, because the agency cost estimate was lower, regardless of the health care addition, this adjustment did not change the competitive sourcing decision. Had the cost competition between the public and private sources been closer, however, the use of a different cost comparison approach could have resulted in a different outcome. The health benefit cost comparability provision has had minimal impact on DOD’s fiscal year 2005 competitive sourcing program and the offerors that participated. Of the 54 public-private competitions we reviewed, the health benefit provision was applicable in only 12 sourcing decisions. In 7 of these 12 competitions, DOD collected health benefit cost information from private sector offerors and found that most of their health benefit costs exceeded 5.5 percent of direct labor costs. This is mostly due to the requirements of the Service Contract Act—which mandates minimum wages and fringe benefits (which could include health insurance) for employees on government service contracts. DOD contracting officials and the private sector offerors told us that complying with the health care cost provision was not unduly burdensome. Implementation of the health care provision did not alter the outcome of any of the competitions. We reviewed the 54 public-private competitions that were either in progress or completed between October 1, 2004, and June 30, 2005. As shown in figure 1, only 12 public-private competitions that reached a sourcing decision involved some consideration of the requirements of the health benefit cost provision. Also as shown in figure 1, in the remaining 42 of the 54 competitions, the health benefit cost provision was not a factor for various reasons. For example, DOD contracting officers did not need to implement the health benefit cost comparability provision in 13 competitions that involved 10 or fewer FTEs since the requirement applies only to competitions involving more than 10 FTEs. (See app. IV for more information on the remaining 42 competitions where the health benefit cost provision was not yet applied or not a factor in sourcing decisions.) The DOD component conducting the public-private competition determined that there was no need to collect data on health benefit costs in 5 of the 12 competitions for which the legislative provision was applicable. As shown in table 3, in one of those competitions, the work was retained for agency performance. In that case, the cost estimate for agency performance was about 45 percent lower than the private offer. In the remaining four cases, a private offeror submitted a lower cost proposal than the agency’s cost estimate, and the difference was so great (ranging between 8.1 and 14.8 percent less) that even adding the full health cost factor of 5.5 percent would not have made a difference. In the remaining 7 of the 12 competitions, DOD components collected and assessed health benefit data from private offerors. As shown in table 4, most private offerors’ proposed costs for health benefits far exceeded DOD’s 5.5 percent benchmark. According to DOD component contracting officers and our review of competitive sourcing documents, the administrative steps taken to collect health benefit data were not unduly burdensome and generally did not significantly delay competition schedules. For the seven competitions in which cost data were obtained from the offerors, the components usually obtained the data through solicitation amendments. This step was necessary because the solicitations had been issued prior to the health benefit cost provision becoming effective. Component contracting officers generally told us that collecting the health benefit data imposed neither unusual burden nor unacceptable delays. In one case, instead of a solicitation amendment, the contracting officer simply contacted the offeror and asked the company to submit the health care cost data. Contracting officers told us that they plan to include the health benefit cost provision in the future solicitations. Our discussions with the offerors in the public-private competitions also indicated the process created little difficulty for them, and required minimal efforts. The health benefit data needed were readily available and generally maintained in the company accounting systems. This was the case for both small and larger companies. According to the offerors we interviewed (including one firm that submitted to DLA detailed data about health benefits), submitting the health benefit data was not considered unusually burdensome. The private offeror involved in DLA’s process we contacted raised no concern with us about any burden. Our review of DLA’s competition documents, however, indicated that much more documentation about health benefits and costs is expected to be submitted by a private offeror participating in a DLA public-private competition than what is expected of private offerors participating in other competitions following DOD’s preferred process. In all seven competitions we reviewed where DOD obtained health benefits data, private offerors were subject to the Service Contract Act (SCA). At the time DOD reviewed their health benefit costs, the SCA required that these offerors pay at least $2.59 per hour for employees’ fringe benefits. We contacted 6 of the 7 private offerors who submitted health benefit data for these competitions. Four of these offerors allowed their employees to use all of the SCA minimum benefit rate towards the cost of the health insurance, and they easily met the 5.5 percent health benefit cost benchmark. The fifth offeror allowed its employees to use a portion of the SCA benefits towards health insurance cost and receive the remainder as an increased hourly wage. As a result, this company’s offer fell short of the 5.5 percent benchmark for health benefit costs. No adjustment was made, however, because the agency cost estimate was lower. The sixth private offeror’s proposal included the cost of the required SCA fringe benefits, but the company notified DOD and also told us that it does not offer to pay for employee health insurance. Company officials told us that because most of their employees are former military or civilian employees with military or federal retiree health benefits, the company’s business decision under the SCA fringe benefit requirement is not to contribute towards employee health benefits. Instead, company officials told us they contribute towards a retirement benefit. Even after adjusting the offeror’s cost proposal by adding the 5.5 percent health benefit cost factor, the offeror had the lowest cost proposal and won the contract. DOD competitive sourcing and legal officials told us that they did not consider the availability or cost of SCA minimum requirements for health and other fringe benefits when they developed their approach for implementing the health benefit provision. DOD competitive sourcing officials acknowledged most private offerors will be able to match or exceed the 5.5 percent health benefit cost benchmark simply by meeting existing SCA fringe benefit requirements. Our analysis of established SCA rates for wages and benefits indicates that the ratio of benefit costs to labor costs is usually much greater than the 5.5 percent health benefit cost comparability benchmark under DOD’s process. For example, a general maintenance worker paid $17.28 an hour and receiving the current SCA benefit of $2.87 an hour for employer-paid health insurance would result in that employer paying roughly 16.6 percent of its direct labor costs for health benefits. The Department of Defense is currently using two different processes to implement the legislative health benefit cost provision. Although both are reasonable approaches for ensuring that private offerors do not gain a competitive advantage from lower health benefit costs, and neither one has yet affected the outcome of any public-private competition, the use of two different processes is problematic. The lack of a consistent DOD-wide process may—in future competitions where agency and private offerors’ proposal costs are close—result in different competitive sourcing outcomes depending on which approach is used. Such a result would be inconsistent with the purpose of Circular A-76, which is to provide for greater consistency in the competitive sourcing process and with the sourcing principles adopted by the Commercial Activities Panel. DOD currently lacks a uniform process for implementing the health benefit cost comparability provision that is in keeping with the sourcing principle that public-private competitions be guided by clear, transparent, and consistently applied processes. With legislation pending to extend this health benefit cost comparability provision through fiscal year 2006, DOD should not continue to permit this inconsistency to persist. To align DOD’s competitive sourcing program more fully with governmentwide policy contained in Circular A-76 and the sourcing principles of the Commercial Activities Panel, we recommend that if the health benefit cost provision is extended, the Secretary of Defense should direct the Deputy Under Secretary of Defense for Installations and Environment to require use of a uniform and consistent process for the DOD components in evaluating the health benefits costs of private sector offerors in public-private competitions. In comments on a draft of this report, DOD concurred with the recommendation. Both DOD and OMB said they remain concerned that the health care cost provision may harm small business participation in DOD’s competitive sourcing program. As such, both agencies said they will continue to seek elimination or amendment of the provision. Written comments from DOD and OMB are reprinted in appendices V and VI, respectively. DOD and OMB also commented that the report is based on very limited data involving only 12 competitions and that our finding of minimal impact cannot be used to predict the impact on future competitions. However, we did not focus on assessing what impacts the provision could potentially have on DOD’s competitive sourcing program in the future. Rather, we assessed the impacts the health benefits provision was having on DOD’s fiscal year 2005 competitive sourcing program. Our finding that the provision had minimal impact is based not only on the 12 competitions in which the provision was applicable, but also on analysis of 42 other public- private competitions where the provision did not come into play for various reasons. In addition, we reviewed other information and obtained the views of DOD officials involved with the competitions and representatives for private offerors who submitted health benefit cost data for DOD’s consideration. We are sending copies of this report to interested congressional committees, the Secretary of Defense, and the Director of OMB. We will also provide copies to others on request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-8214; or WoodsW@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report were Carolyn Kirby, Assistant Director; John Dicken, Rosa Johnson, Charles Perdue, Russ Reiter, Sylvia Schatz, Natalie Schneider, Bob Swierczek, Ann Marie Watt, and Anthony Wysocki. This appendix provides information on the availability of and employer contributions for health benefits in the private sector based on our review of recently published research. Information is also presented on the transitional benefit corporation concept that has received attention as a mechanism for minimizing the loss of health insurance and other benefits for civilian federal employees affected by conversion of commercial activities performed by government employees to private sector performance. Recent government and private sector studies indicate that a variety of changes have taken place with employer-sponsored health insurance plans in the last 5 years, including a decrease in the percentage of small firms offering health benefits and an increase in the cost of the health benefit premiums for all employers. According to recent Current Population Survey data, 81 percent of all individuals aged 18 to 64 years with health insurance in 2004 received coverage through employment-based insurance. From 2000 through 2005, the percentage of all firms offering health benefits fell from 69 percent to 60 percent according to the Kaiser Family Foundation’s annual survey in 2005 of employer benefits. This decline is largely due to the decline in the percentage of small firms that offer health insurance because small firms represent the majority of all employers. However, nearly all larger firms (with 200 or more employees) offer employer-paid health benefits—98 percent in 2005 according to Kaiser’s survey. This is consistent with government data from the 2003 Medical Expenditure Panel Survey (MEPS), which indicates that as establishment size increases in terms of the number of employees, the percentage of employers offering health insurance increases. (See fig. 2.) According to Kaiser’s annual survey in 2005, the cost of health insurance premiums has increased dramatically from 1999 to 2005, rising by over 97 percent. Average annual premiums for employer-sponsored health insurance rose to $4,024 for self-only and $10,880 for self-and-family. Analysis of 2003 MEPS data indicates that private industry generally contributes at least as much towards employees’ health insurance plan premiums as the 72 percent average that the government contributes towards civilian employees’ health insurance premiums under FEHBP. According to MEPS, all size categories of private sector employers on average paid greater than 80 percent of the health benefit premiums for self -alone coverage and between 69 percent and 78 percent for self-and- family coverage. (See fig. 3.) Under the transitional benefit corporation concept, if an agency determines that one of its commercial activities could be performed by nongovernmental employees, the employees currently performing that activity would be given the opportunity to incorporate as a new, more efficient business organization outside of the federal agency to continue performing the same type of activity. This new employee-formed corporation could obtain business by contracting with the private sector or partnering with other governmental, private sector, educational, or not- for-profit entities. The transitional benefit corporation concept includes a mechanism intended to minimize the immediate loss of federal health insurance and retirement benefits for those former government employees affected by the agency’s decision to convert work to private sector performance. Specifically, under the concept, the former government employees could temporarily keep their participation in federal health insurance and retirement benefit programs while transitioning from federal government to private sector employment status. Under this concept, during this transition, an agreement may be established allowing the government agency to continue to pay for the employee’s federal retirement and health insurance benefits, with the new private corporation eventually paying for those benefits. The concept has been suggested as an alternative to the government’s conducting Circular A-76 competitions for commercial activities. According to one analysis of this topic, the benefits of a transitional benefit corporation include Economic development and savings: The government would realize savings more quickly than through the A-76 competition process. For example, the estimated time period to develop a transitional benefit corporation is 6 months, with savings realized shortly thereafter. The current A-76 process may be much longer and therefore would not provide savings as quickly. Also, savings to the government would also result from no longer needing to maintain underutilized assets and personnel. Surge capability/readiness: The government could contract with the transitional benefit corporation in order to expand its workforce rapidly and draw on the former employees during times of increased government workload. Because the transitional benefit corporation is a private organization, it would be able to hire staff outside the constraints of traditional government hiring, which can slow the hiring process. “Soft landing” for former government employees: Government employees who would become part of the transitional benefit corporation would be guaranteed their job and allowed to retain their government benefits, such as pension and health insurance, for a certain time period. According to one analyst, for the transitional benefit corporation concept to be a viable alternative to A-76 and for the government to realize its potential benefits, three conditions must exist. First, displaced federal employees must have the appropriate skills to compete in private sector. Second, private sector competitors must be present within the same business area. Third, the agency proposing the creation of a transitional benefit corporation must have adequate knowledge about the current market conditions and whether or not workload would be sufficient for the new organization to be viable and maintain revenue. DOD competitive sourcing officials told us that they do not consider the concept as a viable alternative to competing commercial activities under the A-76 competitive sourcing process. DOD officials commented that the A-76 process is more appropriate because it emphasizes a competitive process to select a service provider, while the transitional benefit corporation concept would use a sole-source approach that preserves specific jobs and benefits for affected employees. DOD officials also questioned the feasibility of allowing former employees to retain and accrue federal benefits when they are no longer employed by the government. An OMB competitive sourcing official told us that while OMB officials are aware of the concept, they have no current plans to conduct an analysis for governmentwide implementation. To determine how DOD has implemented the health benefit cost comparability provision and the impact the provision is having on its fiscal year 2005 competitive sourcing program, we interviewed competitive sourcing officials with overall responsibility in the Office of the Deputy Undersecretary of Defense (Installations and Environment). We also interviewed DOD component competitive sourcing program and contracting officials in the Army, Air Force, Navy, Marine Corps, Army Corps of Engineers, Defense Logistics Agency, Defense Contract Management Agency, and the Department of Defense Education Activity involved with fiscal year 2005 public-private competitions involving the health benefit comparability provision. To determine the impact of the provision, we reviewed the 54 DOD public- private competitions that were in progress or completed (i.e., tentative or final sourcing decision announced) between October 1, 2004, and June 30, 2005. We identified and obtained data on these 54 competitions from DOD’s automated system used to manage the program across the department—the Commercial Activities Management Information System (CAMIS). CAMIS contains certain data elements for individual A-76 cost comparisons, including numbers and length of individual competitions; numbers of positions to be affected; comparisons of agency and contractor estimated costs; and solicitation, sourcing decision, and contract award dates. We have previously reported some concerns about the accuracy and completeness of data contained in CAMIS. A recent DOD Office of Inspector General report concluded that DOD has not effectively implemented its CAMIS system to track and assess the cost of the performance of functions under the competitive sourcing program. To check the quality of the CAMIS data on the 54 competitions we identified that were in progress or completed between October 1, 2004, and June 30, 2005, we asked cognizant DOD and competitive sourcing officials in the components to verify the accuracy and completeness of the CAMIS data we used for each of the 54 competitions. Based on the results of our verification of the data with these cognizant officials, we believe that the data are sufficiently reliable for purposes of this report. We reviewed Circular A-76 policies and procedures regarding agency cost estimates for personnel and benefits in public-private cost comparisons. We also discussed DOD’s implementation of the health benefit comparability provision with OMB officials responsible for governmentwide competitive sourcing policy and procedures under Circular A-76. We reviewed DOD’s policies, procedures, and guidance and analyzed public-private competitive sourcing and other documents pertaining to the implementation of the health benefit comparability provision in DOD’s fiscal year 2005 competitive sourcing program. We reviewed this material to document actions taken by DOD to implement the health benefit comparability provision in fiscal year 2005 public-private competitions and the impact the provision had in terms of administrative difficulty, competitive sourcing decision outcomes between agency or contractor performance, and any disincentives for private sector participation in DOD’s competitive sourcing program. We also obtained views and information about the implementation and impact of the health benefit comparability provision by interviewing representatives for six private offerors that submitted health benefits cost data for a public-private competition where DOD reached a sourcing decision between October 1, 2004, and June 30, 2005. We reviewed DOD competitive sourcing documents and interviewed contracting officials for another competition, but did not contact the offeror for an interview due to a pending appeal of the agency’s tentative sourcing decision. For background purposes to gather information on the health benefit comparability provision, we also interviewed representatives of a federal labor union, government contractor associations, and researchers on government competitive sourcing. To provide information on the availability of health benefits and employer contributions in the private sector, we reviewed recently published research from selected government and nongovernmental health benefits research organizations. To provide information on the transitional benefit corporation concept, we reviewed relevant literature. We interviewed one legal analyst who has published an article about governmentwide adoption of the transitional benefit corporation concept as an alternative to Circular A-76 public-private competitions. We also interviewed DOD and OMB competitive sourcing policy officials to obtain their views on the concept and prospects for implementation as an alternative to conducting A-76 public-private competitions for commercial activities. We conducted our review from February 2005 through October 2005 in accordance with generally accepted government auditing standards. The health benefit cost comparability provision is a requirement for DOD under Section 8014 of the Department of Defense Appropriations Act, 2005 (Public Law 108-287, enacted August 5, 2004). See italicized text below for the Section 8014 (a)(3) provision. SEC. 8014. (a) LIMITATION ON CONVERSION TO CONTRACTOR PERFORMANCE.—None of the funds appropriated by this Act shall be available to convert to contractor performance an activity or function of the Department of Defense that, on or after the date of the enactment of this Act, is performed by more than 10 Department of Defense civilian employees unless— (1) the conversion is based on the result of a public-private competition that includes a most efficient and cost effective organization plan developed by such activity or function; (2) the Competitive Sourcing Official determines that, over all performance periods stated in the solicitation of offers for performance of the activity or function, the cost of performance of the activity or function by a contractor would be less costly to the Department of Defense by an amount that equals or exceeds the lesser of— (A) 10 percent of the most efficient organization’s personnel- related costs for performance of that activity or function by Federal employees; or (B) $10,000,000; and (3) the contractor does not receive an advantage for a proposal that would reduce costs for the Department of Defense by— (A) not making an employer-sponsored health insurance plan available to the workers who are to be employed in the performance of that activity or function under the contract; or (B) offering to such workers an employer-sponsored health benefits plan that requires the employer to contribute less towards the premium or subscription share than the amount that is paid by the Department of Defense for health benefits for civilian employees under chapter 89 of title 5, United States Code. (b) EXCEPTIONS.— (1) The Department of Defense, without regard to subsection (a) of this section or subsections (a), (b), or (c) of section 2461 of title 10, United States Code, and notwithstanding any administrative regulation, requirement, or policy to the contrary shall have full authority to enter into a contract for the performance of any commercial or industrial type function of the Department of Defense that— (A) is included on the procurement list established pursuant to section 2 of the Javits-Wagner-O’Day Act (41 U.S.C. 47); (B) is planned to be converted to performance by a qualified nonprofit agency for the blind or by a qualified nonprofit agency for other severely handicapped individuals in accordance with that Act; or (C) is planned to be converted to performance by a qualified firm under at least 51 percent ownership by an Indian tribe, as defined in section 4(e) of the Indian Self-Determination and Education Assistance Act (25 U.S.C. 450b(e)), or a Native Hawaiian Organization, as defined in section 8(a)(15) of the Small Business Act (15 U.S.C. 637(a)(15)). (2) This section shall not apply to depot contracts or contracts for depot maintenance as provided in sections 2469 and 2474 of title 10, United States Code. (c) TREATMENT OF CONVERSION.—The conversion of any activity or function of the Department of Defense under the authority provided by this section shall be credited toward any competitive or outsourcing goal, target, or measurement that may be established by statute, regulation, or policy and is deemed to be awarded under the authority of, and in compliance with, subsection (h) of section 2304 of title 10, United States Code, for the competition or outsourcing of commercial activities. This appendix presents information on the 42 competitions in which the health benefit cost comparability provision was not a factor in sourcing decisions between October 1, 2004 and June 30, 2005, for various reasons. In one competition decided in October 2004, the health benefit comparability provision was not a factor (and is not included in the tables below). In this case, the Navy decided to retain the Naval Education and Training Command support services (involving 276 FTEs) within the agency because no cost proposals were submitted by private offerors in response to the Navy’s solicitation. Table 5 presents information on the 14 competitions that were in progress as of June 30, 2005. In these competitions, DOD had yet to make a sourcing decision, and thus DOD contracting officers had not yet needed to implement the health benefit cost comparability provision. Table 6 presents information on 14 streamlined competitions that—as a result of market research completed through June 30, 2005—contracting officers determined that the agency cost estimate was the lowest. Thus, in these decisions, DOD contracting officers did not need to request health benefit data because no private offerors were being considered for the work. Finally, table 7 presents information on 13 competitions involving 10 or fewer FTEs. In these cases, contracting officers did not need to implement the health benefit cost comparability provision, since the requirement applies only to competitions involving more than 10 FTEs.

Competitive sourcing is a management tool where federal agencies conduct competitions between federal employees and private companies to determine the best source to provide commercially available services. Concerns have been raised in the Congress that differences in the costs of federal and private health insurance benefits could disadvantage the federal workforce in public-private competitions. A health benefit cost comparability provision in the 2005 Defense Appropriations Act prohibited any advantage for private offerors that provide no health benefits or contribute less for them than the Department of Defense (DOD) contributes for its civilian employees. Legislation is pending to extend the provision for another year. GAO, in response to a mandate, determined (1) how DOD implemented the provision, and (2) what impact the provision had on DOD's fiscal year 2005 competitive sourcing program. Most DOD components implemented the health benefit cost provision using a process designed to ensure that private sector proposals include an amount for employee health benefits at least equal to the amount that Office of Management and Budget Circular A-76 requires to be added to agency cost estimates to account for employee health benefits. Under Circular A-76, this amount is 5.5 percent of direct labor costs. The Defense Logistics Agency (DLA), however, used a different process designed to determine whether a private sector offeror's monthly health benefit premium contributions are at least equal to DOD's. While DOD's and DLA's processes are both reasonable approaches, the use of different processes could result in different competitive sourcing outcomes in some cases. The health benefit cost provision had minimal impact on DOD's fiscal year 2005 competitive sourcing program. Of the 54 public-private competitions we reviewed, the health benefit provision was applicable in only 12 sourcing decisions. In 7 of these 12 competitions, DOD collected health benefit cost data from private sector offerors and found that most of their health benefit costs exceeded 5.5 percent of direct labor costs. This is largely due to the requirements of the Service Contract Act--which mandates minimum wages and fringe benefits (which could include health insurance) for employees on government service contracts. Although the processes used by DOD and DLA resulted in increasing two private offerors' cost proposals, the adjustments did not alter the outcome of the competitions. Contracting officials and the private sector offerors told us that complying with the health benefit cost provision was not unduly burdensome.

The tax code allows individuals and businesses to make noncash contributions (e.g., vehicles, paintings, used clothing, and household goods) to qualifying charities by allowing taxpayers to claim deductions for their donations on their tax returns. However, not all organizations are granted nonprofit or tax-exempt status by the IRS that qualifies taxpayers for tax deductions for items donated to them. Table 1 provides examples of organizations that do and do not qualify donors for noncash deductions, including vehicle donations. IRS guidance instructs donors to establish the value for their donation based on its “fair market value” for donated vehicles, that is, what the item would sell for on the market, taking into account its condition, including mileage in the case of vehicle donations. As is the case for all noncash contributions, the IRS does not require donors to obtain an independent appraisal for a vehicle’s value unless they claim over $5,000 for the donated property. IRS guidance suggests that donors use used car guides, comparable sales, and other sources to assist in establishing the fair market value for their donated vehicles. Regulatory oversight over charities and their vehicle donation programs is diffused, shared between the IRS and state agencies. The IRS decides which charities are granted nonprofit status and whether the charity meets tax-exempt requirements and complies with federal laws. Many states require charities soliciting within the states to register with the state attorney general’s office or the secretary of state’s office. Figure 1 shows the 39 states and the District of Columbia that require charitable organizations to register with state charity offices. In general, states prohibit unregistered organizations from soliciting for donations in their state. Some state agencies also review vehicle donation advertisements in response to consumer complaints, or when they discover a charity is soliciting for donations in their state without being registered. In addition to oversight by the IRS and state agencies, some private sector organizations develop standards to promote ethical charitable practices and collect information on charitable organizations. Charity “watchdog” organizations, such as the Better Business Bureau’s Wise Giving Alliance, Council of Better Business Bureaus, American Institute of Philanthropy, Association of Fund-Raising Professionals, and the Independent Sector, provide insight to the public on various fund-raising activities. These organizations collect information on charitable organizations and develop standards to promote ethical practices. They disseminate these standards in an effort to “inspire public confidence.” These standards include the voluntary disclosure of an organization’s activities, finances, fundraising practices, and governance. Based on our national survey of charities, few charities reported having a vehicle donation program. Correspondingly, a small percentage of taxpayers claimed tax deductions for donated vehicles. Despite frequent advertisements soliciting vehicle donations to charities, few charities reported having vehicle donation programs. Of U.S. charities with revenues of $100,000 or more, we estimate that 2.7 percent, or about 4,300 charities nationwide, have vehicle donation programs. This projection is based on our survey of 600 charities, of which 16 reported having a vehicle donation program. While the small number of charities with vehicle donation programs does not allow us to make national estimates, we found that most of the 16 vehicle donation programs identified by the national survey were relatively new programs, as shown in figure 2. Only 4 of the 16 charities had vehicle donation programs prior to 1998. Our analysis of IRS tax return data for tax year 2000 showed that a small percentage of taxpayers claimed deductions for vehicle donations. We reviewed a representative sample of taxpayer returns that claimed noncash contributions of over $500 for tax year 2000. We found that of the 129 million returns filed that year, an estimated 0.6 percent, or 733,000 returns, contained tax deductions for vehicle donations. The 733,000 returns represented about 17 percent of the 4.4 million returns filed with noncash contribution deductions over $500. We estimate that vehicle donation deductions lowered taxpayers’ income tax liability by an estimated $654 million. The dollar amount of vehicle donation deductions totaled about 6 percent of the noncash contributions claimed, while stocks and thrift store donations accounted for most of the deductions for noncash charitable contributions over $500, as shown in figure 3. While few taxpayers claim tax deductions for donated vehicles, 2 charities we contacted conducted surveys of their donors and found that the ability to claim a tax deduction was one of the important reasons individuals donated their vehicles to charity. Other important reasons cited in the surveys for donating vehicles were to help a charitable cause and to easily dispose of an unwanted vehicle. The vehicle donation process, for the charities we reviewed generally consisted of four steps: (1) solicitation/donor contact, (2) vehicle pick-up, (3) vehicle sale, and (4) distribution of proceeds. Forty-five of the 65 charities we interviewed reported using third-party agents for some or all of these steps rather than relying on in-house resources, and some had arrangements with more than one agent. About half of the 45 charities used third-party agents to run the entire program, while other charities used a third-party agent for only certain functions. The vehicle donation process is depicted in figure 4. Step 1 – Solicitation/donor contact. The vehicle donation process generally begins with solicitations for donated vehicles through advertisements. Vehicle donations may be solicited directly by charities, third-party agents, or both, depending on the agreement between the charities and third-party agents. Of the 45 charities we interviewed that discussed their advertising practices, 26 reported that advertising was handled solely by the charity. Some third-party agents solicited donated vehicles for several charities using a common advertisement. Some of the most common mediums for vehicle donation advertisements include the radio, newspapers, and the Internet. (For Web version of this report, click here (www.gao.gov/media/audio/donatecar.mp3) to hear a radio vehicle donation advertisement, or see app. II for the transcript.) Vehicle donations are also solicited through advertisements on billboards, truck banners (see fig. 5), and television, as well as in newsletters and even on small paper bags. Also during this step, donors initiate contact with the charity and or third- party agent to donate their vehicle. Either charities or third-party agents may take the initial call from a potential donor, asking the donor questions that may be used to screen vehicles, such as the vehicle’s make, year, and condition, and if the donor has the title to the vehicle. Twenty-four of the 65 charities we interviewed reported that they accepted donor calls in- house, while 23 said that they used third-party agents to accept calls or shared this responsibility. Some charities or their agents limited the vehicles they accepted to those they anticipated would produce a profit after towing and other expenses. However, some charities reported accepting vehicles regardless of condition, as suggested in figure 6. One charity official stated that accepting vehicles with little value was a way of generating goodwill for future donations. Step 2 – Vehicle pickup. After the donor makes the initial call to donate a vehicle, arrangements are made to pick up the vehicle and deliver it to wherever it will be stored until it is sold. Vehicles are generally towed, according to a third-party agent, due to safety and liability concerns. A majority of charities we contacted used third-party agents to pick up vehicles. Once vehicles are picked up, donors are generally provided with a receipt to document the donation for tax purposes. At this time, the charity or third-party agent also obtains the title of the vehicle from the donor. Some charities may provide the donor with state-required forms (e.g., release of liability), or references for establishing the tax deductible value of their donated vehicle (e.g., car guides or IRS guidance). Step 3 – Vehicle sale. Once collected, donated vehicles are most often sold. Charities or third-party agents typically sell donated vehicles through auctions to auto dealers, to the public, or to vehicle salvagers. The majority of charities we contacted said that charities do not handle the selling of vehicles themselves, but instead rely on a third-party agent. Charities and third-party agents said that they generally sold donated vehicles at auto auctions because (1) auctions allow high volume of auto sales and (2) charities do not have the resources, such as staff, storage space, or licenses required to sell vehicles themselves. Of the 65 charities we interviewed, 43 charity officials said they sold all of their donated vehicles, while officials at 16 charities said they used some donated vehicles for clients, charity staff, or other purposes. For example, 1 charity official said that the charity used donated vehicles for student training for a community college auto course. Step 4 – Distribution of proceeds. After vehicles have been liquidated, the proceeds are distributed. Charities with in-house vehicle donation programs keep proceeds that remain after deducting costs associated with processing the vehicles. When charities use third-party agents, the financial agreement between the charity and the third-party agents dictates the proceeds that the charity and fund-raiser will receive from the sale. In addition to the in-house and third-party arrangements, we identified some variations in how vehicle donation programs operate. In one case, a consortium of 14 charities jointly runs a vehicle donation program in conjunction with a wrecking yard. The charities share in oversight of the operations, such as inspecting donated vehicles and monitoring vehicle donation reports. Donors can select 1 charity to receive the proceeds, or if no charity is designated, proceeds are split among members of the consortium equally. In another case, 1 large charity runs a national vehicle donation program and serves regional offices as a third-party agent would, charging its regions vehicle processing costs. However, some of the charity’s affiliates choose other third-party agents that are not part of the national program to run their program. Finally, in still another case, a large charity runs a national program and serves charity affiliates, but also has a nonprofit vehicle donation program for other smaller charities. Although proceeds from vehicle donations are a welcomed source of revenue, it was not a crucial source of income for the majority of the charities we reviewed. The proceeds charities received from vehicle donations varied in the 54 cases we tracked, but were generally considerably less than the amount donors claimed on their tax returns for the donated vehicles. Based on information from charities we spoke with, this difference is due in part to donated vehicles being often sold at auto auctions at wholesale prices, and processing expenses and third-party fees reducing the amount of proceeds charities receive. We could not verify the accuracy of taxpayer claims regarding the value of their donated vehicle. The annual net proceeds from vehicle donations for 2002 reported by the charities we interviewed ranged from as little as $1,000 for 2 vehicles donated to a senior center, to over $8.8 million for 1 national charity that received over 70,000 vehicles. The charities considered the proceeds received as a welcomed, but rarely crucial source of income to sustain their operations. Although the dollar amount received from vehicle donations was over $1 million for several charities we spoke with, for many, the revenue was a small share of total charity revenue. Charity proceeds constituted less than 2 percent of the total annual budget for 15 of the 30 charities providing budget information; however, 2 of the charities stated that vehicle donation proceeds provided 90 percent or more of their annual revenue. Many of the charities we interviewed stated that their vehicle donation program provided benefits beyond revenue by providing an expanded donor base and name recognition for the charity. In the 54 specific vehicle donations we tracked, charity proceeds from vehicle donations were much less than the value deducted by donors on their tax returns. Based on charity and the third-party agent we contacted, two factors contributed to this difference: (1) vehicles are often sold at auto auctions for salvage or at wholesale prices, which are typically lower than prices that would be received if the donor sold the vehicle themselves and (2) processing costs and fees are deducted from gross sales revenue, further reducing charity proceeds. Figure 7 illustrates the amount a charity received from 1 of the 54 vehicle donations we tracked. In this case, a 1983 GMC Jimmy truck was donated in 2001 to a charity whose vehicle donation program is operated by a third- party agent. The gross sale price for the truck, which sold at an auction, was $375. After deducting third-party and advertising expenses, net proceeds from the vehicle sale totaled $62.00. This amount was split 50/50 between the third-party agent and charity, leaving the charity with $31 from the vehicle donation. The taxpayer claimed a $2,400 tax deduction for the donated vehicle on his/her tax return, based on the fair market value of the vehicle listed in a used car guidebook. Appendix III details the vehicle donation transactions for all 54 tracked cases. Donated vehicles are often sold at auto auctions for lower prices than what a seller might receive if the vehicle were sold to a private party. For the 54 donated vehicles we tracked, sale prices for donated vehicles ranged from 1 percent to 70 percent of donor tax deduction claims, and over half of the cases were 10 percent or less of what donors’ claimed. (See app. III) As one third-party agent stated, it is unfair to compare auction sale prices for donated vehicles to deduction claims because most donated vehicles are sold at auctions that cater predominantly to wholesalers who then resell the vehicle at higher prices. Of the 59 charities we contacted during our review that said they sell some or all donated vehicles, 42 used auctions to dispose of the vehicles. Another reason for the difference between the amounts deducted by donors for donated vehicles and the proceeds charities receive from vehicle sales is that sales proceeds are reduced by vehicle processing costs, such as towing, advertising, program administration, and third-party agent fees. California is the only state that collects data on the proceeds received by charities from vehicle donation programs. According to the California Attorney General’s records, 145 charities using third-party agents who had filed the required financial reports received approximately $16 million, or 35 percent of the $45.8 million raised from reported donated vehicle sales, during 2001. The amount of proceeds these charities received in California ranged from 2 percent to 80 percent of proceeds after third-party costs were deducted. Taking both the lower sales price and deductions for processing costs into account, the proceeds received by charities from donated vehicles were much lower than the donor-claimed value for the vehicles in the 54 donated vehicle cases we tracked. Charities received between 0 and 54 percent of the value claimed by donors, with most receiving 5 percent or less, as shown in figure 8. For some vehicle donation sales, charities receive no proceeds after the costs of vehicle donations are deducted. For 6 of the 54 donated vehicles we tracked, the processing costs exceeded the sales price for the vehicle (see app. III). For charities using third-party agents, whether the loss is absorbed by the third-party agent or deducted from charity proceeds from another higher value donated vehicle depends on the agreement between the parties. Because third-party agents and other donated vehicle processing costs vary among charities, comparing net vehicle donation proceeds between charities can be misleading. One third-party agent said that programs claiming a high percent of proceeds as profit are not including their full costs, that is, they are only counting towing costs and may not include space, advertising, or staff costs. Similarly, 1 charity managing its own vehicle donation program stated that its proceeds may be lower than some other charities, but they are accurately capturing the true program costs while others may not be. Proceeds also differ based on different agreements between charities and third-party agents for paying for the third-party’s services. Of the 45 charities we contacted during our review that reported using third-party agents, 20 said third parties were paid a percentage of net proceeds for their vehicle donation services, many splitting the net proceeds 50/50. Other charities reported paying third- party agents a flat fee per vehicle or had some other arrangement, such as paying for towing expenses incurred by third-party agents. A number of charities interviewed had little insight into vehicle donation processing costs incurred by third-party agents because they received limited, or in a few cases, no information on charges for vehicle donations. Although most charities reported receiving an itemized list of revenue and costs of donated vehicles from their third-party agents, the detail was not always provided. For example, 1 charity received an itemized list of each vehicle sold that listed all costs under “cost of sales” except for the third- party fee. Another received an itemized list of sold vehicles with all costs under “tow fees” and “expenses.” Thirteen of the charities we contacted reported that they received a check from a third-party agent accepting vehicles for the charity without their knowledge, and only 4 of the 13 said that information on processing costs was provided. One well-known charity with its own vehicle donation program stated that they repeatedly contacted two third-party agents to stop them from accepting vehicles on their behalf. The charity estimated that in 1 year one of the third parties had deducted approximately $25,000 from proceeds to the charity over what it would have cost the charity itself to process the donated vehicles. An additional factor that may explain the difference between the proceeds charities receive from vehicle donation sales and what donors claim as the value of the donated vehicle may be an inaccurate assessment by donors of the vehicle’s value. Although many charities we spoke with said they try to limit vehicle acceptance to those in running condition, some charities accept vehicles in poor condition. Charities stated that a number of the vehicles donated are sold for scrap, and some said donor claims about vehicle value might be inflated. We could not determine in the 54 cases we tracked whether donors appropriately claimed deductions for donated vehicles. IRS guidance suggests that taxpayers consider using used car guides when estimating the fair market value for donated vehicles, while also considering the vehicle’s condition and mileage. Of the 54 cases, 25 assessments were based on nationally recognized used car guides. However, since we did not have additional information, such as the vehicle’s condition and mileage, we could not determine whether the reported valuations claimed by donors accurately reflected fair market value. The IRS has one compliance program that produces audit leads on potentially overstated noncash contributions, but it does not follow up on these leads. According to IRS officials, it does not audit cases with potentially overstated noncash contributions because it has higher priority compliance issues to address. Also, IRS data on its returns processing procedures that are directed to disallowing deductions for inappropriately claimed noncash contributions show that a small percent of returns are subject to these procedures and that few deductions are disallowed. As part of its National Research Program, IRS plans to gather information on noncash contribution compliance issues, which could provide it with data to determine how best to address noncash contribution compliance issues. An IRS donated property task force has drafted several recommendations that could lead to more emphasis being directed to vehicle donation programs and deductions. Data from the 11 states we contacted were limited regarding noncompliance by charities and third-party agents involved with vehicle donation programs. However, several states’ agencies have identified and initiated legal actions against individuals and organizations that have not complied with laws or regulations related to vehicle donations. These include instances in which an organization posed as a charity to receive donations, a third party inappropriately kept vehicle donation proceeds, and a charity was not following state requirements for processing vehicles. One of IRS’s compliance programs is designed to detect individual taxpayers who may overstate noncash contributions on their tax returns, including donated vehicles. However, due to higher priority demands for determining compliance with tax laws, IRS has not been following up on leads of potential noncompliance that are generated from this program. Under this program, about two full-time equivalent employees at IRS’s Ogden Submission Processing Center are used to compare the proceeds charities received from donated property shown on property disposal forms (Form 8282 Donee Information Return) with the amounts claimed by taxpayers on their tax returns. If there is a wide discrepancy between the charity’s revenues for the property and the amount claimed by the taxpayer on their tax return, the case is referred to field offices for possible audit. An Ogden Campus official estimated that on average about 20,000 Form 8282s are received annually, and that 4,000 to 5,000 individual tax returns are reviewed to determine whether they should be audited. An Ogden official estimated that the cost to retrieve a tax return from IRS files is between $100 and $150. Ogden officials stated that they do not track the number of cases that are referred for possible audit or the types of donated property involved in the cases, but estimated that 30 percent to 40 percent of the returns reviewed are referred. An IRS official estimated that most referred cases related to donated land or boats, which generally have higher potential tax assessments than do vehicles. According to our analysis of the IRS audit data, none of the returns that were referred during fiscal years 2001 and 2002 were audited. IRS officials stated that the returns were not audited because the potential tax assessment yield from these cases was substantially smaller than from other types of compliance issues handled in the field. IRS also established processing procedures for returns to identify and disallow deductions for noncash contributions either when taxpayers claim noncash contributions over a certain amount or when they do not attach required Form 8283, Noncash Charitable Contributions to their returns. These processing procedures cover relatively few noncash contribution deductions. For example, we estimate that for tax year 2000 returns, IRS’s returns processing threshold for these deductions would account for about 1 percent of the returns where noncash contributions of over $500 were claimed. According to IRS, returns that meet the threshold are given a special code and are reviewed by the examination staff to determine whether they have audit potential. Returns with audit potential are put in the audit inventory for possible audit selection by field agents. IRS found that these returns were not being selected for audit because field agents had other higher priority work. IRS expects this higher priority work to continue into the foreseeable future, and as a result, beginning in January 2004, returns processing staff will discontinue coding these returns for review by examination staff. IRS does not have data on the number of noncash contribution deductions that have been disallowed because of missing Form 8283s, but IRS officials estimate that few were disallowed. IRS also has returns processing procedures to identify and disallow noncash donations to individuals or nonqualifying organizations, such as political organizations. According to IRS, in 2002 it disallowed noncash contributions of about $21.8 million on 154 tax returns for donations made to individuals and nonqualifying organizations. In addition to the above compliance activities that focus on taxpayers’ deductions for donated vehicles and other types of noncash contributions, IRS’s Exempt Organization Division has an examination program that focuses on whether charities meet tax-exempt requirements and complies with federal law, such as those governing the use of funds for a charitable purpose rather than private gain. IRS had little information on whether its examinations identified compliance problems with charities operating vehicle donation programs. At the time of our review, IRS officials informed us that IRS had seven vehicle donation program examinations in progress and had completed two cases. According to an IRS official, in one recent case, IRS revoked the exemption status for one Florida organization whose charitable purpose was to provide research, education, and technical training on the marine environment. The charity raised funds through the solicitation and sale of boats. IRS found that the organization’s charitable activities were insubstantial, and that private parties were benefiting from the substantial economic benefit of the organization’s activities. While more compliance resources are being devoted to higher priority audit issues such as abusive tax shelters and high-income nonfilers, IRS’s National Research Program is to provide data on compliance problems associated with noncash contributions, including deductions for donated vehicles. Under the program, the IRS randomly selected about 47,000 tax year 2001 returns to determine whether taxpayers complied with statutory income, expense, and tax reporting requirements. Returns with noncash contributions, including donated vehicles, could be subject to audit to verify donation claims. Once this project is completed in December 2004, IRS plans to assess individuals’ compliance related to deductions for noncash contributions and determine what actions are needed to help ensure proper reporting in this area. In 2001, IRS established a donated property task force that examined various issues relating to such topics as property appraisals and valuations and coordination of compliance activities between various IRS organizational units. In July 2002, the task force developed several draft recommendations for improving IRS’s oversight of donated property programs and deductions. The recommendations included revising Form 8283 to add a separate category for donations of motor vehicles on the portion of the form that identifies the type of property donated. Another recommendation made was to establish procedures to ensure that IRS records and maintains copies of Form 8282s that are filed with the Ogden Submission Processing Center. The task force noted that without such procedures, IRS could not verify the accuracy of the forms or determine whether charities filed them. The IRS’s audit procedures instruct auditors to determine whether charities submit required Form 8282 when disposing of donated vehicles. Auditors may assess penalties if they find that the charity did not submit required Forms 8282. However, determining whether charities filed the forms may be difficult because the forms are destroyed if they are not used in Ogden’s noncash contribution audit referral program discussed above. According to IRS officials, at the time of our review, IRS had not taken action on this recommendation. Many states oversee charities to protect the public, and 39 states and the District of Columbia require charities to register with the state attorney general or the secretary of state offices. States have an interest in whether charitable fund-raising is fraudulent and whether charities are using funds to meet the charitable purpose for which they were created. We contacted 11 state attorney general offices or secretaries of state to identify information related to vehicle donation programs. Only 1 state reported having data to identify charities with vehicle donation programs; however, several said that they would investigate a charity vehicle donation program if they received complaints from the public. As discussed in the following items, several of these states uncovered problems with vehicle donation programs, including instances in which an organization posed as a charity to receive donations, third-party agents inappropriately kept vehicle donation proceeds, a charity was not following state requirements in processing vehicles, and individuals solicited vehicle donations for fictitious charities. In Massachusetts, a for-profit company representing itself as a charity solicited cars through newspaper ads leading potential donors to believe that the organization was a charity and that all, or a substantial portion of the proceeds would go directly to providing counseling to children and parents in Massachusetts. In May 2002, Massachusetts brought and won an enforcement action in which the company’s president agreed with state officials to cease all further activity related to the car donation operation. Connecticut officials filed suit in July 2003 against a used auto dealership and a bogus charity that was created by the dealership’s owner. Vehicles were solicited, supposedly to help abused and abandoned animals, but virtually all of the proceeds were retained by the auto dealership, which maintained one checking account for both organizations, according to state officials. Ohio’s Attorney General filed a complaint in 2003 against a nonprofit organization that solicited over 800 vehicles in the name of donor designated charities, but at least $258,000 in vehicle proceeds was not provided to the designated charities, according to the complaint. The California Attorney General’s office filed a civil action against the incorporator of a nonprofit that solicited vehicles for charity, but there was never a charitable program, only a used car lot. The Attorney General’s office estimated that over $1 million was raised by the operation, none of which benefited charity. Criminal charges against the defendant resulted in a 5-year jail sentence. In a case filed in June 2003, the California Attorney General’s office filed a case against a nonprofit organization selling donated vehicles to the public that had not met California’s safety requirements. Some vehicles sold for export to locations such as Belize and Mexico, which do not require the same state safety standards, were not actually exported. An indictment filed in a U.S. District court in October 2002, charges that defendants established over 100 toll-free numbers with sound– alike charity names, such as the National Mental Health Association, Cancer Society, or National Diabetes Association. According to the case filing, potential donors were fraudulently told that the sound-alike organizations were the national charities whose names they approximated or were affiliated with nationally known charities; however, according to the case records, the defendants kept all proceeds. A number of government and consumer organizations provide guidance to donors to assist them in making informed decisions about donating their vehicles. Guidance is also available to assist charities in accepting noncash contributions and in selecting, hiring, and managing third-party agents. Charitable donors in general, including those that donate their vehicles, can rely on guidance available from the federal government and other sources to ensure that they make informed donations. For example, IRS publishes guidance on claiming deductions for donations, and establishing fair market value for items donated. In addition, a Federal Trade Commission-led initiative highlights legal actions taken against individuals and organizations that engage in phony philanthropic activities, and provides tips on how to recognize and avoid fraudulent solicitations. Some states also offer guidance for potential donors and may have financial information on specific charities. A number of nongovernmental sources offer donors similar advice. For example, the Better Business Bureau’s Wise Giving Alliance, Guidestar, Charity Navigator, and the American Institute of Philanthropy offer tips for charitable giving or information on specific charities. Appendix IV lists specific sources for donor guidance. The guidance generally identifies steps donors should take when donating vehicles and claiming associated tax deductions. These steps are listed as follows: Verify that the recipient organization is a tax-exempt charity. Churches, synagogues, temples, mosques, and governments are not required to apply for this exemption in order to be qualified. Determine whether the charity is properly registered with the state government agency that regulates charities. The state regulatory agency is generally the state attorney general’s office or the secretary of state. Ask questions about how the donated vehicle will be used to determine whether it will be used as intended. Such questions include: Will the vehicle be fixed up and given to the poor and needy? Will it be resold and, if so, what share of the proceeds will the charity receive? Itemize deductions in order to receive a tax benefit from the donation. The decision to itemize should be determined by whether total itemized deductions are greater than the standard deduction. Deduct only the fair market value of the vehicle. The fair market value takes into account many factors, including the vehicle’s condition, and can be substantially different from the value listed in used car guides. Document the charitable contribution deduction. IRS Publication 526 identifies requirements for the types of receipts taxpayers must obtain and the forms they must file. Follow state law regarding the car titles and license plates. Generally, the donor should ensure that the title of the vehicle is transferred to the charity’s name by contacting the state department of motor vehicles, and keep a copy of the title transfer. Donors are also advised to remove the vehicle’s license plate if allowed by the state. Some guidance is also available to charities with vehicle donation programs regarding accepting noncash contributions and hiring a third- party agent for fund-raising purposes. For example, IRS Publication 1771, Charitable Contributions, Substantiation, and Disclosure Requirements, explain federal law for organizations that receive tax- deductible contributions. The IRS also plans to publish, by March 31, 2004, a brochure advising charities on how to avoid problems raised by vehicle donation programs. State guidance to charities we reviewed was generally not specific to vehicle donation programs, but rather provided general guidelines for selecting, hiring, and managing third-party agents. In addition, a number of nongovernmental sources offer charities similar advice. The Better Business Bureau, for example, publishes standards for charities in areas such as fund-raising activities, and issuing informational materials to donors. Appendix IV lists specific sources for charitable guidance. Some of the general guidance provided to charities that are relevant to donation programs are listed as follows. Consult and adhere to IRS’s publications explaining federal law for organizations that receive tax-deductible contributions, and review IRS’s annual Exempt Organizations Implementing Guidelines. Comparison shop for fund-raising agents, insist on a written contract, and do not relinquish control of a program to the fund-raiser. Follow standards published by various nongovernmental sources for governance, oversight, fund-raising activities, and issuing informational materials. IRS will not have data on whether taxpayers are appropriately claiming tax deductions for noncash contributions, including donated vehicles, until it completes its National Research Program study at the end of 2004. In the meantime, IRS is using resources to produce audit leads on overstated noncash contributions that are not being audited because of higher priority compliance demands. To determine whether these resources could be used more productively, IRS could assess the merits of the noncash contribution compliance program to determine whether it could be modified to take advantage of the leads that are generated, such as auditing taxpayers through correspondence instead of sending them to field offices. The IRS has audit procedures that instruct auditors to assess penalties against charities that do not file the required Form 8282 when they dispose of property. However, as reported by its donated property task force, IRS does not have a way of verifying whether charities file the required Form 8282 or whether forms that are filed are accurate. The task force’s draft recommendation to maintain and record Form 8282 is worth considering. This recommendation was made in July 2002, and has not been acted upon. In conjunction with IRS’s ongoing National Research Program study, which is to be completed in December 2004, we recommend that the Commissioner of the Internal Revenue assess (1) whether the Ogden compliance program should be modified to take advantage of the leads generated by the program and (2) the feasibility and usefulness of maintaining and recording the receipt of Form 8282 as recommended by IRS’s donated property task force. We received written comments on a draft of this report from the Commissioner of Internal Revenue (see app. V). The Commissioner agreed with our recommendations and identified some alternatives and actions that IRS is considering to ensure compliance by charities with reporting requirements. With regard to our recommendation that IRS assess its compliance program for generating audit leads on taxpayers that may have overstated their noncash contributions, the Commissioner stated that IRS actions related to the compliance program will be based on its review of the level of noncompliance in reporting noncash contributions. The Commissioner also agreed with our second recommendation to consider whether the Form 8282s that charities submit when disposing of donated property should be recorded and retained. He stated that decisions on the handling of the forms would be made in conjunction with decisions on the first recommendation and with other changes IRS has underway, particularly the redesign of the Form 990. IRS is considering changing the Form 990 to include information on the filing of Form 8282 as an alternative to retaining the Form 8282. The Commissioner noted that some steps have already been taken to improve the reporting of vehicle donation programs, such as the revision of Form 990 filing instructions for 2003 to provide organizations with an example of a vehicle donation, and other actions noted in our report. As arranged with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after the date of this report. We will then send copies of this report to the Secretary of the Treasury; the Commissioner of Internal Revenue; the Director, Office of Management and Budget; and other interested parties. We will also make copies available to others on request. The report is also available on GAO’s Web site at http://www.gao.gov. If you or your staff have any questions on this report, please contact me at (202) 512-8777 or Ralph Block at (415) 904-2150. Key contributors to the report are listed in appendix VI. Our objectives were to determine: (1) the number of charities with vehicle donation programs, and the number of taxpayers claiming deductions for vehicle donations; (2) the vehicle donation process; (3) proceeds received by charities from vehicle donations to what donors claim for vehicle donation deductions; (4) the Internal Revenue Service’s (IRS) and state compliance activities directed at vehicle donations and incidents of noncompliance; and (5) guidance available to taxpayers and charities to help them make informed decisions regarding vehicle donations. To satisfy these objectives, we relied on two sources of nationally representative information and several anecdotal sources of information. Table 2 summarizes the primary data sources used to address each of the objectives. To estimate the number of charities that have vehicle donation programs, we conducted a national telephone survey with a statistically representative sample of organizations registered with IRS as 501(c)(3) organizations (charities) with an annual income of at least $100,000. The sample was drawn from the 2002core data set (tax year 2001) of the National Center for Charitable Statistics (NCCS), which contains the IRS Form 990 data for all nonprofit organizations in the United States. A simple random sample of 600 charities was drawn from the population of 157,512 charities. The NCCS data were determined to be sufficiently reliable for the purposes of our report, based on interviews with NCCS officials regarding reliability procedures and observation of the sampling process. Valid telephone numbers could not be obtained for 11 percent of the charities in the sample, and 6.8 percent of the charities did not answer the telephone after several contact attempts or did not return calls, yielding a total rate of nonresponse of 17.8 percent. We took steps to reduce errors in our estimates by pretesting the survey with charities of varying characteristics and performing computer analyses to identify inconsistencies and other indicators of errors. We received valid responses from 493 of the 600 charities, for a response rate of 82.2 percent. Interview responses were weighted to account for the initial sampling rate and response rate. As with most surveys, our estimation assumes that nonrespondents would have answered like the survey respondents, and we do not know of any evidence about whether our respondents were different from nonrespondents. Our confidence in the precision of the results from this sample is expressed in 95 percent confidence intervals. We are 95 percent confident that the results we would have obtained had we studied the entire population are within +/- 2 percentage points of this result. To estimate the number of taxpayers that claimed deductions for donated vehicles, we analyzed a random sample of tax year 2000 individual tax returns from IRS’s Statistics of Income (SOI) individual tax return file. The SOI file is a stratified probability sample of income tax returns filed with the IRS. The tax year 2000 sample represented about 129 million tax returns. The SOI sample contained information on 34,942 returns where taxpayers itemized deductions and claimed a noncash contribution of over $500. We randomly sampled 600 cases from this sample population and requested the tax returns from the IRS. Weights were adjusted to represent all taxpayers claiming noncash contributions exceeding $500. Of the 600 cases in our sample, actual returns were available for 509 of the cases. We determined that the SOI data were sufficiently reliable for the purposes of our report based on interviews with IRS officials and testing for bias in our sample. No bias was identified in a comparison of available returns and nonavailable returns in terms of taxable income, total contributions, and several other factors. To obtain information on vehicle donation program processes and proceeds from vehicle sales, we contacted 65 charities throughout the country. The 65 charities included: 16 charities in our national charity survey that reported having a 8 charities interviewed as pretests to the survey, and 41 charities identified in advertisements or on taxpayer deduction claims. Not all 65 charities provided information on all topics discussed in the report. Where information is reported, the specific number of charities that provided information related to the topic discussed is included. These examples cannot be generalized beyond the charities responding. We interviewed six third-party agents that provide vehicle donation program services to charities for examples of how the vehicle donation process works for their organizations and the charities they served. Some of the agents represented more than 100 different charities; however, the information we received from these agents cannot be generalized beyond the agents responding. We relied on a number of sources to identify vehicle donation third-party agents, including state attorney generals’ office officials, donated vehicle advertisements, and charity officials. To obtain information on guidance provided to donors and charities, we interviewed officials or reviewed materials from several organizations involved with charity oversight or charity fund-raising, including the Better Business Bureau, Chronicles of Philanthropy, the Independent Sector, and the American Institute of Philanthropy. In addressing all of our objectives, we spoke with IRS headquarters officials from the Tax Exempt and Government Entities Operating Division, the Wages and Investment Operating Division, and the Small Business/Self-Employed Operating Division. We also conducted an on-site visit to IRS’s Ogden Campus to obtain information on its Form 8282 compliance program. We also interviewed the Chairperson of the IRS Vehicle Donations Working Group. We contacted state officials based on referrals from the IRS and the National Association of State Charity Officials. We also interviewed state officials in California, Michigan, Massachusetts, Florida, Pennsylvania, South Carolina, Ohio, Oregon, Connecticut, Washington, and Texas. To determine the amount of proceeds charities receive from donated vehicles relative to the amounts taxpayers claimed as tax deductions, we tracked a judgmental set of donated vehicles obtained from 4 charities in 4 states. Two charities were local charities that used third-party agents to manage their vehicle donation programs; 1 was a national charity that used a third-party agent to handle processing (but not advertising) for their vehicle donation program; and the fourth was a national charity that managed its vehicle donation program in-house. The information we obtained from the charities included the vehicle description, selling price, net amount received by the charity after expenses, and donor information. Using the donor information, we requested tax return data from the IRS to determine the amounts donors claimed as deductions for their vehicles. Not all of the vehicle donors claimed deductions for their donations. Of the 70 donors the IRS was able to identify as filing returns, 12 did not itemize their taxes. Four additional cases were dropped due to data errors or missing information. We were able to track the remaining 54 vehicle donations. The individual cases or cases in aggregate are for illustration only, and cannot be used to generalize to vehicle donations overall. Our analysis of the 54 tracked vehicles is shown in appendix III. We reported summary data from the California Office of the Attorney General regarding the percent of proceeds received by charities from vehicle donation programs using commercial fundraisers. We discussed data reliability issues with state officials and determined that the data were sufficiently reliable for the purposes of our report. Based on a review of a nonprojectable sample of vehicle donation advertisements, we found that vehicle donation advertisements most often stated that individuals could claim a tax deduction for the donation, if the donation served a charitable purpose, and the donor’s vehicle would be towed free of charge. We analyzed 147 advertisements, including 69 newsprint advertisements from a sample of 50 newspapers nationwide, 33 radio advertisements from 19 radio stations in the top 10 U.S. markets, and 44 Internet advertisements. Figure 9 identifies the most common claims made in the newspaper, radio, and Internet advertisements we reviewed. According to an IRS official, advertisement claims are potentially misleading when they do not specify that taxpayers must itemize their deductions to claim a deduction for vehicle donations, since many taxpayers do not itemize their deductions. Of the 147 advertisements we reviewed, 117 indicated that taxpayers could claim a tax deduction, but only 7 advertisements specified that donors must itemize in order to claim a deduction. In addition, IRS officials stated that advertisements could be misleading when they claim donors can value their vehicles at full, or maximum, market value when claiming a tax deduction, particularly when the same advertisements claim that vehicles are accepted whether they are running or not. Full or maximum market value, although not clearly defined, implies that a vehicle is in good running condition with no major mechanical defects. Fair market value equates to what a vehicle would sell for on the market, and takes into account a vehicle’s condition and mileage, among other factors. Of the 147 advertisements we reviewed, 8 identified that a donor could claim full or maximum market value, while more commonly, advertisements identified that donors could claim fair market value on their tax returns when donating their vehicles. Responsibility for oversight of advertisements is diffused. The Federal Communications Commission defers regulatory authority regarding false advertising on radio or television to the Federal Trade Commission (FTC). The FTC does not have specific jurisdiction over charities, but may become involved in cases of fraud. State officials are primarily responsible for false advertising by charitable organizations. Officials we interviewed from 2 states said that limited resources prevent them from providing broad oversight over advertisements, and that they generally review advertisements in response to consumer complaints, or when they discover that charities or third-party agents are soliciting in their state without being registered. Table 3 identifies the number of advertisements we reviewed that specified taxpayers must itemize their deductions to claim a vehicle donation; that taxpayers could claim full, maximum, or fair market value for their deduction; and whether the advertisement referred potential donors to the IRS Web site, an accountant, the Kelley Blue Book, or other source for guidance on claiming a tax deduction. The transcript from an actual radio advertisement identifying some of the benefits of vehicle donation programs is shown in figure 10. What to do with that car? Donate it to the Council of the Blind! Donating a car is trouble free; you get a tax write off, and do your part for a worthy cause. The California Council of the Blind has helped in the workplace since the thirties. Simply call, we’ll pick up that car, running or not, in most cases, plus boats--even real estate can be donated for a write off. Call our live operators for your free pick up now-- 800 xxx-xxxx. 800 xxx-xxxx, that’s xxx-xxxx. Table 4 details 54 specific vehicle donations identifying the amount of proceeds charities receive from donated vehicle sales relative to the amounts taxpayers claim as tax deductions for donated vehicles. Information on this judgmental set of 54 vehicle donations were obtained from 4 charities in 4 states. The individual cases or cases in aggregate are for illustration only, and cannot be used to generalize vehicle donations overall. Government and consumer organizations provide guidance to donors to assist them in making informed decisions about donating vehicles. Guidance is also available to charities to assist them in selecting, hiring, and managing third-party agents. A partial list of resources is included in table 5. Leo Barbour, Carl Barden, Keira Dembowski, Michele Fejfar, Tre Forlano, Lemuel N. Jackson, Monica Kelly, Rosa Leung, Brittni Milam-Bell, Amy Rosewarne, Sam Scrutchins, Addie Spahr, and Wendy Turenne made key contributions to this report.

Donating a vehicle to charity enables a donor to support a charitable cause, dispose of an unwanted vehicle, and receive a tax benefit. More charities are turning to vehicle donation programs as a means for raising funds. As a result, our objectives were to: (1) determine the number of charities with vehicle donation programs, and the number of taxpayers claiming deductions for vehicle donations; (2) compare the proceeds received by charities from vehicle donations to what donors claimed for those deductions; and (3) describe related Internal Revenue Service (IRS) and state compliance activities. An estimated 4,300 charities have vehicle donation programs, based on a GAO survey of 157,500 charities with revenue of $100,000 or more. Taxpayers claimed deductions for donated vehicles on about 733,000 of the 4.4 million tax year 2000 returns filed with noncash deductions over $500, lowering taxpayer liability by an estimated $654 million. For the charities surveyed, proceeds from vehicle donations ranged from $1,000 for one charity, to $8.8 million for another. However, proceeds generally constituted a small share of total charity revenue for the majority of charities GAO reviewed. In addition, for two-thirds of the 54 specific vehicle donations GAO examined, charities received 5 percent or less of the value donors claimed as deductions on their tax return. Differences in proceeds received by the charity and value claimed by a taxpayer were due in part, to vehicles being sold at auctions at wholesale prices, and proceeds being reduced by vehicle processing and fundraising costs. Due to a lack of available data on the condition of donated vehicles, GAO could not determine whether taxpayers appropriately valued their vehicles when claiming associated tax deductions. The IRS has some activities designed to detect noncompliant claims for noncash deductions, including vehicle donations. However, the IRS has not pursued potential leads from these activities because tax revenue yields are less than other potential noncompliance cases, such as abusive tax shelters. IRS's ongoing National Research Program study may provide information on how to deal with donated vehicle compliance issues. Also, an IRS task force drafted recommendations for improving IRS's oversight of charities' donated property programs. State officials have filed legal actions in a number of cases involving problems with vehicle donation programs, such as an individual soliciting vehicle donations for fictitious charities.

CCOs are congressionally created entities that are in part privately funded but operate under some level of government oversight, generally through the appointment of their leadership, management oversight, or The missions of these entities vary substantially, additional regulation.but individually they are narrowly defined. Table 1 provides attributes of the four CCOs selected for our review. The Smithsonian Institution was established with funds from James Smithson, a British scientist who, on his death in 1829, left his estate to the United States to found “at Washington, under the name of the Smithsonian Institution, an establishment for the increase and diffusion of knowledge.” The Smithsonian Institution opened to the public in 1855, and has since become the world’s largest museum and research complex with 19 museums, the National Zoo, and nine research facilities in Washington, D.C., 7 states, and Panama. There were 30.3 million visitors to the institution’s museums and zoo in 2012. The National Gallery of Art was created in 1937 for the American people by Congress, accepting the gift of financier and art collector Andrew W. Mellon, who wished that his private art collection be the basis of a national art museum. Funds for the construction of the original building were provided by the A. W. Mellon Educational and Charitable Trust. The President accepted the completed building and the Mellon collection on behalf of the American people in 1941. There were 4.2 million visits to the museum in fiscal year 2012. In 1979, the newly-formed President’s Commission on the Holocaust recommended that a living memorial be established to honor the victims and survivors of the Holocaust and to ensure that the lessons of the Holocaust would be taught in perpetuity. The Holocaust Memorial Museum was chartered in 1980 and opened to the public in 1993. The museum is situated on federal land on the national mall and was built entirely with private funds. There were 1.6 million visitors to the museum in fiscal year 2012. Spain established the Presidio of San Francisco as its northern most military outpost in the New World in 1776. The U.S. Army took control of the military base in 1847 and later transformed it into part of the nation’s coastal defense system. The Presidio served as an active military installation until 1994, when it was transferred to the National Park Service. In 1996, Congress created the Presidio Trust and mandated that it preserve the site’s natural, cultural, scenic, and recreational resources. Congress assigned 80 percent of the park’s 1,500 acres to the Presidio Trust; the National Park Service continues to manage the remaining 20 percent (coastal areas). Congress directed that the Presidio Trust attain financial self-sufficiency 15 years after the first meeting of the Trust’s board of directors. The Presidio Trust achieved this goal and, since October 1, 2012, has sustained itself through a combination of philanthropic sources and rental income from residential and commercial buildings on its grounds. An estimated 4 million people visited the park in 2012. Congress created the Valles Caldera Trust in 2000 to preserve a unique volcanic parcel of land in New Mexico. Modeled after the Presidio Trust, the Valles Caldera Trust was considered a 20-year public-private land management experiment. The Valles Caldera Preservation Act authorized the Secretary of Agriculture to purchase about 89,000 acres, known as the Baca Ranch. The act charged the Valles Caldera Trust with managing the land to achieve a number of goals, including becoming financially self-sustaining by the end of fiscal year 2015. We found in 2009 that while the Valles Caldera Trust had taken steps to establish and implement a number of programs and activities to position it to achieve the Preservation Act’s goals, it was at least 5 years behind the schedule it set to achieve in meeting those goals. We also found that its biggest self- identified challenge was to achieve financial self-sustainability. As such, we recommended that the Chairman of the Board and Executive Director work with the relevant congressional committees to seek legislative remedies, as appropriate, for the legal challenges confronting the Valles Caldera Trust, such as prohibition from entering into long-term leases or acquiring property. When we followed up on our recommendations in February 2013, we found that New Mexico’s Senators had reintroduced legislation that would transfer ownership of the Valles Caldera National Preserve to the National Park Service. Because this would, in effect, eliminate the legal challenges we cited, as well as the goal of self- sufficiency for the Valles Caldera Trust, we have closed this recommendation as being implemented. Congress has given these CCOs a unique structure and range of authorities. For example, they are managed by boards of directors whose membership make-up is designated by statute. CCOs’ personnel and procurement flexibilities allow them to achieve public goals by leveraging nonfederal resources and increasing their capacity to respond more nimbly to the needs of the organization and attract and retain talent. The four CCOs we studied partnered with nonfederal entities ranging from nonprofit organizations and research institutes to for-profit businesses, international entities, and individuals. These partnering arrangements aimed to generate business revenues, further research goals, enhance programmatic offerings, strengthen internal operations, and raise private funds. These efforts are consistent with our previous work on collaborative efforts, which recognized that the federal government must identify ways to deliver results more efficiently and in a way that is consistent with its multiple demands and limited resources. Further, we found that the federal government could work together with the nonfederal sector to generate more public value than could be See appendix 1 for detailed produced when agencies act alone.examples of how the four CCOs we studied worked with nonfederal partners. Officials from the CCOs we studied cited four factors that facilitate their ability to leverage nonfederal resources: (1) unique authorities that provided legal flexibilities; (2) benefits received from being part of the federal government; (3) governing boards that provided management and oversight; and (4) informal networks that enabled CCOs to share lessons. Critical among the authorities mentioned above is the ability to accept gifts and solicit private donations, but the CCOs we studied are not required to, and did not always provide, a complete picture of their nonfederal resources to Congress. A critical factor that facilitates CCOs’ ability to leverage nonfederal resources are unique legal authorities exempting them from certain federal regulations on (1) soliciting and accepting private funds and (2) using those funds to hire staff and procure goods. Officials from all four CCOs we studied said that these authorities and flexibilities allowed them to act more nimbly and to adapt more readily to the needs of their organization. Authority to solicit and accept private funds. The four CCOs we studied have the statutory authority to solicit and accept private gifts and donations. They are also able to retain and use these funds without fiscal year limitation or further congressional approval. This authority provides CCOs with additional financial resources beyond directly appropriated federal funds. Officials from the Smithsonian Institution told us that their donors value the institution’s flexibility to retain and use nonfederal funds to carry out its mission. Further, they noted that donors view their contributions as an addition to, not replacement for, federal funding. CCOs used private funds, for example, to build new buildings and expand their collections. An official also explained that these authorities allow CCOs to accept gratis services from volunteers. Officials from all four CCOs we studied told us that this authority was critical to helping ensure they have appropriate resources to meet their organization’s mission and goals. Further, all four of the CCOs we studied are tax exempt under section 501(c)(3) of the Internal Revenue Code. The status as a charitable organization can facilitate fundraising efforts because donors are potentially eligible for tax benefits based on their donations. Officials told us that this status is critical to their ability to raise private funds. Hiring and procurement flexibilities. The four CCOs we studied also benefit from some exemptions from federal hiring and procurement requirements. For example, some CCO staff are not subject to many of the civil service laws in Title 5, which govern the hiring of federal employees. CCO officials told us they use this flexibility to compete with the private sector to attract and retain certain staff—such as fundraisers and business development managers—whose specialized skills and abilities play a key role in attracting and leveraging nonfederal funds. Officials also said that the ability to pay these staff in accordance with market rates (i.e., above the federal pay scale) allowed them to attract and retain highly qualified individuals to serve in critical positions. In addition to being able to compete with the private sector in this manner, CCO officials told us that they are authorized to terminate staff when their specialized skills are no longer needed. Lastly, CCO officials noted that the general exemption from federal procurement laws provides CCOs with the flexibility to procure goods and specialized services more quickly and efficiently than they would otherwise be able to. CCO officials reported they implement these managerial flexibilities in a way that maximizes financial resources while safeguarding the principles these federal laws are created to uphold. For example, CCO officials told us that they develop policies for staff paid with nonfederal funds that follow the spirit of federal employment laws, when appropriate. In this way, they see this as helping to ensure transparency and accountability while still maintaining flexibility. Specifically, the Human Resources Director at the Presidio Trust said that its personnel policies reflect principles consistent with federal equal employment opportunity guidelines. Further, while these flexibilities are critical to their organizations, CCO officials said that it is often not immediately apparent from their enabling legislation whether a CCO is considered a federal or private entity for employment, benefits, insurance, federal torts, copyright laws, and administrative procedures, and that making these legal determinations can be time consuming. Some of the most valuable resources of the four CCOs we studied have been derived from their federal status. CCO officials told us that this status provided them with significant assets and facilitated their ability to leverage nonfederal resources. Specifically, the Presidio is located on a former military reservation and the Holocaust Memorial Museum, National Gallery of Art, and Smithsonian Institution all enjoy exhibition and office space on federal land near the National Mall in Washington, D.C. Further, some CCOs received private collections and objects that were intended to be donated to the people of the United States and available as national assets. In addition to these physical assets, CCO officials cited intangible benefits that are derived from their federal status. For example, officials at the four CCOs we studied said that federal status signaled their entity’s permanence and therefore helped to secure nonfederal resources. One official explained that many private donors and partners are attracted to the perceived financial stability that federal ties bring, even in these times of shrinking federal budgets. Other officials noted that these nonfederal donors’ and partners’ views are tied to the belief that the federal government will “be around” and so their gifts will be there for future generations. Another official told us that partners and donors see federal status as an implicit recognition of the CCO’s credibility, which helps to bring in more donations and build relationships with other organizations. Officials at the four CCOs we studied also noted that private citizens’ willingness to contribute financial and nonfinancial resources to CCOs is based, in part, on their belief that these entities contribute to a greater public purpose. For example, Presidio Trust officials noted the its public mission combined with the assurance of sustained revenue created a powerful incentive for private citizens to donate and invest in the park. Another factor that facilitated CCOs’ ability to leverage nonfederal resources is access to the support and expertise of federal agencies, specifically the Departments of Justice and State. The Department of Justice represents CCOs in lawsuits and other legal matters. Additionally, officials from the Holocaust Memorial Museum said that the Department of State’s international contacts, expertise in negotiating with entities outside of the US, and diplomatic channels have occasionally facilitated the museum’s ability to resolve issues with international partners. For example, in negotiating agreements with foreign governments to obtain Holocaust-era objects and information from state-owned archives, the Holocaust Memorial Museum follows a template agreement developed and approved by the Department of State, allowing the museum to enter into specific types of agreements on its own authority. Finally, CCOs also found that the federal government immunity to state and local requirements facilitated their ability to leverage nonfederal resources. For example, Presidio Trust officials said that having exclusive federal jurisdiction enabled them to avoid jurisdictional confusion with other government entities. Specifically, these officials said that not having to adhere to local zoning laws helped to decrease the time needed to rehabilitate the Presidio Trust’s buildings. Further, they noted that the Presidio Trust benefited from exemption from certain state and local laws such as rent control. Holocaust Memorial Museum officials noted that the museum is exempt from having to register with each state attorney general for fundraising purposes because of its federal government status. One official added that being exempt from 50 different sets of fundraising restrictions relieves the museum of a significant administrative burden. A third factor officials reported to be helpful in leveraging nonfederal resources is the valuable expertise provided by governing boards. The size, appointment, and duties of these boards are specified in law and vary among CCOs. The size of these boards varies greatly among the four CCOs we studied: The smallest has 7 members and the largest 65.Board members include a mix of presidentially appointed members, ex- officio members based on a government position; such as the Chief Justice of the United States and the Secretary of State; and others including private citizens. Some boards have specific expertise and residency requirements for their members. For example, members of the Presidio Trust Board are required to have extensive knowledge of finance, real estate development, planning, and resource conservation or have expertise in these areas. Further, at least 3 board members must reside in the San Francisco Bay Area. CCO board member term lengths are also specified by law. Some of the CCOs we reviewed provided examples of the help they received from their governing boards. According to officials, the National Gallery of Art’s Board of Trustees plays an active role in fundraising and the Presidio Trust Board provides strategic advice on real estate, financial, and operational management. Additionally, the Smithsonian Board of Regents approves the Smithsonian Institution’s strategic plan, budgets, and other key documents. The Board of Regents also evaluates the performance of top executives and sets their pay. The importance of a stable board is evident when comparing the Valles Caldera experience with that of the other four CCOs we studied. Our previous work on the Valles Caldera Trust noted that board member turnover contributed to challenges, such as delays in decision making and false starts to programs.National Gallery of Art said that their boards have provided consistent guidance and leadership. In contrast, officials at the Presidio Trust and Lastly, some CCOs used informal networks to leverage nonfederal resources. Officials at the Smithsonian Institution and the Holocaust Memorial Museum meet periodically with their counterparts at other related organizations in informal settings to exchange lessons learned, seek advice on shared issues, and discuss questions specific to their unique legal and financial statuses. For example, the Holocaust Memorial Museum’s General Counsel said that he has sought advice from legal staff at other CCOs when considering whether certain laws and regulations apply to the museum. Officials told us they find these meetings to be particularly helpful because they provide a network for organizations that face similar challenges to learn from each other. The Presidio Trust’s Executive Director expressed interest in joining such a network. Further, Presidio Trust officials said that it would be useful to have a regular forum to exchange information and ideas between organizations responsible for former military bases or other public landholdings that seek to revitalize their organization through partnering arrangements. They noted that it may help to meet twice a year so officials could discuss best practices. The information about nonfederal funds presented to Congress in annual budget requests varies among the CCOs we studied. For example, the Smithsonian Institution and Holocaust Memorial Museum typically present some information about these funds in their annual budget requests to Congress, but the type of information differs. For example, the Smithsonian Institution summarizes the balance, source, and uses of its nonfederal nonappropriated funds. Further, it describes funding sources—including those from nonfederal resources—for its museums, research centers, and departments. In contrast, the Holocaust Memorial Museum provides a high-level summary of the uses of its nonfederal funds broken down by restricted and nonrestricted donations. While the National Gallery of Art reported on the nonfederal funds used for special exhibitions in its budget requests, it does not include information on nonfederal funds used for other purposes, which represents the majority of these funds. CCO officials told us that they are not required to report on their nonfederal funds and congressional staffers confirmed this. However, federal internal control standards note that financial information is needed, among other things, to make operating decisions, monitor performance, and allocate resources. These standards further note that pertinent information should be identified, captured, and distributed in a form and time frame that permits people to perform their duties efficiently. The federal budget process is the primary means by which the President and Congress select among competing demands for federal funds; as such, it is essential that budget information be comprehensive and clear. Consistent and timely information about CCOs’ complete financial picture—including both appropriated and nonfederal funds— could provide the Congress with important context for understanding both the relative tradeoffs among appropriation decisions and the implications of such decisions for these entities. While Congress does not direct the CCOs’ use of nonfederal funds, the lack of consistent information on nonfederal funds inhibits the ability to understand how CCOs leverage federal funds to meet their missions. Better information on nonfederal funds may also make clearer the donor and endowment restrictions on some of those funds and would also provide more context about the total financial resources available to CCOs. Although some information about nonfederal funds is available in the CCOs’ audited financial statements, annual reports, and publicly available tax returns, the timing and availability of those reports do not align with the typical time frames of congressional budget deliberations. We compiled six key principles and related key questions that CCOs can use to guide management decisions about partnering with the nonfederal sector. Each principle has corresponding elements that are intended to enhance or facilitate CCOs’ ability to achieve each principle (see table 2). As previously discussed, these key principles are grounded in relevant literature, including our prior work; the knowledge and experience of internal and external subject matter specialists; and the experiences of the four CCOs we studied. Incorporated throughout this section are examples of how CCOs used elements of these principles to manage their partnering arrangements. Additionally, many of these practices are consistent with the key issues we have identified when agencies or other organizations work collaboratively. These principles can be tailored to suit different types of arrangements, partners, and CCO needs, as appropriate. Organizations that leverage partnering arrangements have clear, well- articulated missions; the strategic goals to achieve them; and a defined process for assessing whether partnering arrangements are aligned with and further the organizations’ missions and goals. This is consistent with our past work noting that organizations must have a clear and compelling rationale to work together to overcome significant differences in missions, cultures, and established ways of doing business. A clear, well-articulated mission and supporting goals can help facilitate CCOs’ decision making about partnering. For example, the Presidio Trust’s efforts to clearly determine and articulate its mission have helped it make partnering decisions. Although the Presidio Trust’s authorizing legislation set out two broad goals—one to preserve the park and the other to achieve financial self-sufficiency in 15 years—the Presidio Trust made a strategic decision that self-sufficiency was to be its primary goal since it was a necessary condition for preserving the park’s beauty and natural resources. That is, officials reasoned that the Presidio Trust could only be preserved if the park was still in existence and well managed. To achieve the goal of self-sufficiency, the Presidio Trust worked with real estate and construction firms to plan, develop, and manage the rehabilitation and rental of structures. These partnering efforts helped the Presidio Trust successfully meet its goal of financial self-sufficiency by fiscal year 2012, within the mandated deadline, while also making significant progress in historical preservation. Organizations that are unable to articulate a clear and well-defined mission and to prioritize multiple goals in support of that mission are not well-positioned to make strategic partnering decisions. The Valles Caldera Trust was mandated by law to achieve a number of goals, including self-sufficiency by the end of 2015. We previously found that officials set out to achieve an acceptable balance in the pursuit of all six goals instead of prioritizing them and had did not have a strategic plan laying out those goals nor a performance plan that could measure progress in achieving those goals. Based on our analysis and the principles we compiled, we believe that because the Valles Caldera Trust lacked these plans, it would have also found it difficult to make partnering decisions. Complementary goals and missions are important to a successful partnership. Officials at both the Smithsonian Institution and Holocaust Memorial Museum described processes they have in place to make such determinations. For example, officials at Smithsonian Enterprises, the unit that oversees the Smithsonian Institution’s revenue generating activities, implemented a process to help ensure that business partners and projects have missions that are compatible with that of the institution. Smithsonian Enterprises has a strategic advisory committee composed of 15 to 18 staff from across the institution that meets quarterly to discuss new activities, updates, and ideas. These quarterly meetings have included substantive discussions about whether particular proposed business relationships are consistent with the Smithsonian Institution’s mission. Outside Smithsonian Enterprises, the Smithsonian Institution also carefully assesses whether potential partnering arrangements and partners are in line with its mission. For example, when making partnering decisions, officials rank potential projects based on the extent to which they might help to advance the institution’s mission. Specifically, officials assess the extent to which potential partners have complementary missions and beliefs. For example, officials noted that the institution has declined corporate sponsorship offers from private companies who have publicly espoused beliefs that are different from the institution’s core beliefs and principles, especially key scientific theories, such as evolution, that are important to its mission. Similarly, Holocaust Memorial Museum officials said they carefully assess partner organizations’ goals to help ensure that they align with the museum’s strategic vision and that potential partners’ causes do not conflict with or overshadow the museum’s own mission. To do so, officials consider how partners could further the Holocaust Memorial Museum’s mission. For example, the museum has partnered with the company Ancestry.com to develop software that could be used to index and access information on victims of the Holocaust. In this instance, according to museum officials, the partners have leveraged their complementary missions—to track individual and family heritage information—to provide this resource to more than 1 million people each year. Are the CCO’s mission and goals well-defined and clearly articulated? Are the partner’s goals clearly articulated and well aligned with the CCO’s mission and goals? Will partners understand how key activities, core processes, and resources link to shared mission and goals? How will the partner contribute to the CCO’s mission to deliver programs and services? Top leadership support for partnering arrangements is critical to successfully pursuing and engaging partners. The tone at the top— management’s philosophy and operating style—sets the stage for how the organization will make management decisions, including decisions related to partnering with the nonfederal sector. As a champion for these types of arrangements, leaders can encourage their staff and stakeholders to see the value in creatively building on the assets and resources of partner organizations. This is consistent with our past work noting that top-level commitment and continuity in leadership is a key issue when organizations work to collaborate with each other. Officials from all four CCOs we studied told us their top leaders have encouraged their organization to leverage external resources. To that end, they have cultivated an environment that has facilitated staffs’ ability to form, pursue, and engage with partners. Specifically, a Smithsonian Institution official told us that partnerships at the institution have benefited from extensive top leadership support, and that the Secretary has greatly increased the emphasis on partnerships and collaborations in general. For example, leaders at the Smithsonian Institution have championed the Grand Challenges Consortia Program, which offers competitive funding for interdisciplinary projects through internal grants to researchers that work with interdisciplinary partners to conduct research that aligns with the institution’s strategic goals. Leadership support of this program has sent a clear message about the importance of partnering across interdisciplinary boundaries, and officials noted such partnering has increased since the Consortia was established. In another example, Presidio Trust officials said their top leaders’ efforts to encourage the leveraging of nonfederal resources have resulted in $4 of private investment for every federal dollar received. In contrast, the lack of consistent top leadership can hinder CCOs’ partnering efforts. We previously found that leadership turnover at the Valles Caldera Trust caused management challenges. Specifically, the first executive director served for only 18 months. The position remained vacant for about 7 months and the next executive director resigned after 10 months. Other key positions became vacant in 2004 and 2005, including the controller, business manager, programs director, chief administrative officer, communications manager, and cultural program coordinator. In our past work, we found that the lack of consistent leadership and progress in organizational and program development contributed greatly to staff turnover. This frequent turnover led both to delays in partnering decisions about partnerships and false starts to programs. When we followed up on our prior work in February 2013, we learned that, just in the past year, the Valles Caldera Trust had appointed as its new executive director an employee who served in various roles at the preserve for over 10 years. This official’s long tenure and hands-on experience at the Valles Caldera can, among other things, help the Valles Caldera Trust engage partners in a way that could leverage nonfederal resources. CCOs institutionalize leadership support for partnering through their strategies and guiding principles and the actions of their top leaders. One way to institutionalize this support is to clearly document how partnering arrangements can be used to achieve organizational missions and goals. For example, the Smithsonian Institution’s strategic plan identifies the use of federal and nonfederal partners as a strategy to pursue the institution’s key goals. The plan also identifies steps to achieve this goal. Specifically, the plan states that the institution will enhance its research capacity by leveraging resources from a range of partners including federal agencies such as the National Aeronautics and Space Administration, universities, nongovernmental organizations, industry, and other domestic and international agencies. Are top leaders and managers committed to a common vision of success in partnering arrangements? Are specific responsibilities and accountability for the partnering arrangement clearly defined and established? Do open and candid communications with all stakeholders occur to minimize potential resistance to establishing the partnering arrangement? How will partners operate across different organizational cultures to accomplish respective partner missions and goals? Partnering decisions should reflect both the likely risk and the organization’s tolerance for risk in partnering. Incorporating risk assessment and risk management practices into partnering decisions can help ensure that the organization recognizes and is prepared to manage explicit risks (e.g., financial and physical) and implicit risks (e.g., reputational). The specific risk mitigation and management methodology used will likely vary by organization because of differences in missions and varying tolerance for risk. This is consistent with our prior work noting that risk management helps organizations assess risk, strategically allocate finite resources, and take actions under conditions of uncertainty. The Smithsonian Institution developed a formalized partner selection process to assess risks associated with its Affiliations Program, which shares institution resources, including artifacts, with museums nationwide. As part of the application process, the Smithsonian Institution implemented a process to verify the applicant’s nonprofit status, mission statement, and organization chart. Additionally, applicants are required to be in good standing under state laws and adhere to certain industry standards for managing and storing collections. Through this risk assessment process, the Smithsonian Institution evaluates the management abilities of potential partners and determines whether those partner institutions will be appropriate stewards of the loaned artifacts and resources. Similarly, the National Gallery of Art has assessed partnering risks through a formalized process to identify partners for its privately-funded Art Around the Corner program. This program brings District of Columbia public school classes to the museum to experience art through discussion, role-playing, sketching, art making, and creative writing. To select classes for the program, program staff first communicated with school principals and then invited teachers and their students to visit the museum as a class. During these visits, National Gallery of Art staff observed teacher and student interactions, the teacher’s enthusiasm for the program, and general class dynamics. Further, Art Around the Corner officials make themselves available to provide more information about the program as needed so that teachers understand their role in the program and the expected outcome for their students. During this observation process, National Gallery of Art program staff and teachers can mutually assess the program’s benefits and consider whether or not to be involved. Risks—both explicit and implicit—should be assessed and managed when partnering. For example, the way that the Smithsonian Institution funds projects as part of its Grand Challenges Consortia program takes explicit risks into account by minimizing the institution’s financial exposure. Specifically, the Consortia program staff review grant proposals and provide seed funding of $20,000 and then later up to $100,000 after a project has demonstrated increased capacity. This approach allowed the Smithsonian Institution to take calculated risks on new projects in a manner that mitigates extensive financial losses. Similarly, the Holocaust Memorial Museum has also assessed and managed explicit risks in its traveling exhibitions program by establishing requirements based on industry best practices regarding potential partners’ physical space. Specifically, partners that would like to borrow the museum’s artifacts are required to meet certain facility, security, and preservation requirements to reduce the risk that artifacts are damaged or lost. It is also important to manage implicit risks, such as the likelihood that partners could potentially damage a CCO’s reputation. For example, the Smithsonian Institution allows the use of its space in connection with substantial donations under the condition that the use is consistent with certain policies to help ensure the institution can retain appropriate control over its facilities. These policies include, for example, a prohibition on the use of facilities for such events as weddings or birthdays; partisan, political, or religious gatherings; fundraising; and marketing activities. Another type of implicit risk which can affect a CCO’s reputation is community and stakeholder opinion. Presidio Trust officials received a high degree of neighborhood opposition to a plan to build a contemporary art museum on its grounds. Preservationists and others fiercely opposed the scale of the museum as being inconsistent with the overall historical character of the park and criticized the Presidio Trust for failing to consider and plan for traffic impacts on local streets. Faced with adamant resistance from these outside groups, a decision was made not to pursue plans for the museum. Have explicit and implicit risks—both between and among all partners and internally between staff offices—been identified, analyzed, and allocated? Have the likelihood and significance of risks as well as strategies to manage those risks been identified? Selecting appropriate partners and projects is central to a successful partnering arrangement. Partners should bring complementary resources, skills, and financial capacities to the relationship. Further, a systematic approach helps to identify projects that are well-suited for partnering opportunities and helps to achieve an organization’s mission. This is consistent with our past work noting the importance of ensuring that relevant participants have been included in the collaborative effort. Further, we have found that it is helpful when the participants have full knowledge of their resources; the ability to commit these resources and make decisions; the ability to regularly attend all activities of the collaborative mechanism; and the knowledge, skills, and abilities to contribute to the outcomes of the collaborative effort. Partners have broadened CCOs’ audiences. The CCOs we met with have used partners to expand their audiences locally and throughout the world. For example, the National Gallery of Art has expanded its audience through its Art Around the Corner program which partners with District of Columbia public schools to bring fourth and fifth grade students to the museum who were otherwise unlikely to visit. As part of this program, students visit the museum up to 14 times over two school years to view, discuss, and create art. According to Art Around the Corner officials, the program also has provided an opportunity for the museum to engage with students’ families, some of whom have never visited the museum. Some CCOs reached a broader global audience through international partners. For example, the Holocaust Memorial Museum and the Mémorial de la Shoah in Paris convened an international symposium in 2010 to assess governments’ capacities to effectively respond to genocide and mass atrocities. The symposium—which was attended by leading genocide prevention and human rights officials and experts worldwide—highlighted and examined core issues in genocide prevention to governments all over the globe. Further, strategies were recommended to enhance international cooperation on this issue. Similarly, the National Gallery of Art has collaborated with foreign museums to broaden its international audience. For example, the museum organized an exhibition on Victorian art and design with the Tate Britain museum in London, England. The exhibition is also scheduled to travel to the State Pushkin Museum of Fine Arts in Moscow, Russia. Partners have provided technical support. The CCOs we studied worked with partners who have provided technical support in various areas of business expertise. For example, the Smithsonian Institution and George Mason University leveraged their respective expertise and resources to develop the Smithsonian-Mason School of Conservation. This school offers a semester-long residential academic program for undergraduate students in conservation biology located at the Smithsonian Conservation Biology Institute in Fort Royal, Virginia. A program official explained that George Mason University had previously offered conservation studies at the graduate level only and that this new program provides undergraduates access to prominent scientists and educators earlier in their academic careers. George Mason University provided academic and business expertise to help manage the program and conducted a market analysis to demonstrate the program’s economic sustainability. The Smithsonian Institution contributed scientific resources, such as experienced scientists, laboratories, and connections with other programs. The institution also made property available to house new dormitories and dining facilities. A program official noted that leveraging the university’s complementary areas of expertise was critical to the school’s success and economic self-sufficiency. The Smithsonian Institution has also contracted with companies to provide specialized services in connection with certain business opportunities, including managing food and beverage services, distributing Smithsonian books, and creating Smithsonian-branded products. Specifically, the Smithsonian Enterprises unit works with private sector partners to provide various business services, and the resulting revenues are used to fund programs throughout the institution. For example, the home shopping television channel, QVC, which has a large viewership and marketing expertise, helped the Smithsonian Institution sell jewelry with designs based on the institution’s gem collection. In another instance, Smithsonian Enterprises leveraged Mattel’s product development expertise to design, produce, and sell a paleontologist Barbie doll. In both arrangements, the Smithsonian Institution received a portion of the revenue. Partners have provided operational support. Volunteers and organizations have provided critical operational support to the CCOs we studied. For example, the Smithsonian Institution’s 6,500 volunteers outnumber its paid employees. Volunteers have led tours, conducted field work, assisted with research, provided administrative support, and staffed information desks, among other wide-ranging services. Volunteers have also provided specific technical services to the four CCOs we studied. For example, the Holocaust Memorial Museum worked with a law firm whose lawyers conducted legal research on genocide issues free of charge. Operational support may also come in the form of what is more traditionally viewed as a public-private partnership. For example, prior to the establishment of the Presidio Trust in September 1995, the National Park Service entered into a 55-year ground lease with Thoreau Center Partners, a for-profit California real estate company, to rehabilitate part of an old hospital. Thoreau Center Partners leased, developed, and operated some of these buildings, now known as the Thoreau Center for Sustainability, and then subleased the improved office space to a variety of subtenant organizations. Since the Presidio’s transfer from the National Park Service in 1998, the Presidio Trust has overseen this public-private partnership. When we followed up with Presidio Trust officials about this project in May 2013, they noted that this lease resulted in a good outcome for the National Park Service by providing for the rehabilitation of the building with outside funds.rates had not kept pace with market changes. However, the official said that rental According to Presidio Trust officials, the Presidio Trust continues to pursue a development strategy that includes rehabilitating and leasing buildings by executing long-term building and ground leases with tenants who independently fund building improvements. Officials noted that this approach provides a mix of revenue sources that balances low-risk, long- term, and market-driven rents which provide greater certainty of revenues during economic downturns. Further, officials said it allows tenants to make significant investments toward operational and rehabilitation costs. For example, the Presidio Trust replaced the previously vacant Letterman Army Medical Center with a new 850,000 square foot campus by partnering with filmmaker George Lucas whose company invested $300 million to rehabilitate the complex. This project—known as the Letterman Digital Arts Center—is home to Lucasfilm, Ltd and a number of its subsidiary companies including Industrial Light and Magic, LucasArts, and the George Lucas Educational Foundation. The Presidio Trust earns an revenue of approximately $6 million from the ground lease each year. Other examples of projects that relied on nonfederal partners to rehabilitate and lease Presidio Trust buildings include the Walt Disney Family Museum and the Presidio Landmark residential apartments. Between 2005 to 2011, the Presidio Trust executed 226 ground leases for nearly 1 million square feet that have provided rental revenues valued at $240 million over the duration of these leases. Partners have provided financial support. Nonfederal financial support is critical to CCOs’ ability to further their missions. Nonfederal partners have provided millions of dollars of support to each of the CCOs we studied. These funds allowed CCOs to purchase art and artifacts, construct new buildings, develop exhibitions, enhance program offerings, conduct research, and otherwise further the CCO’s mission. CCOs have worked with donor or member groups to solicit funds from individuals and corporations. For example, the Presidio Trust partnered with the Golden Gate National Parks Conservancy to raise private philanthropic funds. Projects are proposed internally and externally and some CCOs have developed internal processes to help ensure they select appropriate projects for partnering opportunities. Specifically, some CCOs have used committees to evaluate proposals against set criteria, including the extent to which the proposal would be consistent with the CCO’s mission and goals. For example, the Smithsonian Grand Challenges Consortia program assembles a review committee that rates potential projects based on a set of criteria that includes scholarly merit, ability to meet Consortia goals, ability to build coalitions, social impact, team member qualifications, ability to finish the project on time with the provided resources, and potential to garner additional funds. What resources will each partner contribute and how will the CCO leverage those resources? To what extent does the CCO have the key skills needed to create, manage, and monitor partners and projects? How will the potential partners and projects be selected and evaluated? Partnering arrangements are relationships between or among different parties that should be managed actively. Technology can enable information sharing between partners to facilitate the leveraging of resources. Formalizing collaborations between the partners, including documenting dispute resolution processes, can enable productive partner interactions. Further, it is important to have the staff with the right skills and experience to manage these opportunities. CCOs have created online tools that help partners share resources. For example, the Holocaust Memorial Museum leveraged the time and skills of individuals through the power of crowdsourcing on the Internet as part of the World Memory Project. Thousands of volunteers have helped to transcribe more than 2 million records of Holocaust-related historical documents into an online database hosted by the company Ancestry.com. The software developed by Ancestry.com allowed volunteers across the globe to easily access information and perform tasks that helped achieve the Holocaust Memorial Museum’s mission. In another example, the National Gallery of Art developed online educational resources to help it connect with students and teachers, who can help achieve the museum’s mission of fostering an understanding of works of art. These free resources include teaching packets and online interactive lessons. Specifically, the website includes lesson plans, worksheets, and other educational materials on topics such as self- portraits, 19th century American art, and art and ecology. The National Gallery of Art also has a website with activities and games to engage children in art. To facilitate collaboration, it is important that partners agree on roles and responsibilities and that there is a process in place to resolve disputes. One subject matter specialist noted that processes to mediate and resolve disputes and conflicts can help CCOs avoid confusion about partner expectations and may contribute to partners’ willingness to invest resources in the project. This is consistent with our recent work on collaborative mechanisms, in which we found that articulating these agreements in formal documents can strengthen organizations’ commitment to working collaboratively, and that it is important to address diverse organizational cultures to enable a cohesive working relationship and to create the mutual trust required to enhance and sustain the collaborative effort. Another subject matter specialist made the related point that it is also important to think ahead about whether other suitable partners exist if an arrangement fails. The Smithsonian-Mason School of Conservation used written documentation to foster partner collaboration. Specifically, the Smithsonian Institution and George Mason University signed a memorandum of understanding outlining roles and responsibilities, describing the financial commitments expected of each party, allocating financial risk between the partners, and setting terms for renewing the agreement. The document also established a dispute resolution process for discussing and negotiating conflicts. Specifically, the memorandum dictated when one partner has the final authority and, in other cases, when and how a consensus will be reached. Smithsonian Institution officials credited the memorandum of understanding with their ability to manage the whole academic, residence, and dining complex as one unified program, which has facilitated problem solving and de- emphasized the distinction among staff from both institutions. In another example, the National Gallery of Art has used written contracts outlining the teachers’ roles and responsibilities to encourage teacher engagement for its Art Around the Corner program. The National Gallery of Art provides extensive materials to teachers and students, including curricula, workbooks, sketchbooks, art reproductions, art materials, and children’s books. According to officials, integrating program-developed activities into the classroom has helped reinforce students’ critical thinking skills and therefore it has been important for teachers to follow through in carrying out the curriculum in their classrooms. To help ensure that teachers adhere to their contractual responsibilities, the teachers are paid an annual stipend only after their classes successfully completed the program. In addition to written agreements, establishing mutual expectations facilitated successful partnering arrangements. As noted by one subject matter specialist, developing such an understanding takes time and is predicated on coordinating, communicating, and learning how partners operate. Smithsonian Institution officials discussed the importance of setting expectations about how much time is needed for the partnering arrangement. Specifically, the institution has worked to set realistic expectations internally and externally about the time frames needed for its business enterprises. To assist with this process internally, Smithsonian Enterprises maintains a flow chart of the key dates for review. Externally, officials have managed private partners’ expectations by explaining that the development of licenses takes longer than in the private sector. Each of the four CCOs we studied has made use of their previously described legal authorities and exemptions to attract and retain appropriate staff to facilitate partnering. CCOs have used these authorities for staff in areas such as business development, philanthropy, and key management positions that require skills to facilitate and manage partnering arrangements. Further, one CCO used this exemption to be nimble and responsive when its partnering needs changed. Specifically, when working with private contractors, the Presidio Trust was able to downsize construction staff when the 2008 decline in real estate values resulted in less demand for housing in the San Francisco area. CCOs also have had a strong focus on recruiting and retaining staff skilled at facilitating partnering arrangements. This is consistent with our prior work, in which we found that strategic human capital management allows agencies to perform their missions economically, efficiently, and effectively and facilitates agencies’ ability to deploy the right skills, in the right places, at the right time. For example, the Presidio Trust has offered a rent credit program for selected employees who live on the park. It has also conducted an employee survey to assess personal work experience; recruitment, development, and retention strategies; job satisfaction; and feedback on the performance culture and leadership. These practices have helped the Presidio Trust retain staff with, among other skills, expertise in partnering. How will roles and responsibilities surrounding the partnering arrangement be delineated, agreed upon, and documented? Are there clear lines of authority to coordinate and elevate decision making discussions? How will programmatic decisions be made and disputes be resolved? How will the CCO ensure or promote effective and open communications between partners and what tools, if any, would facilitate this communication? What key knowledge, skills, and abilities are needed to manage the partnering arrangement? If the organization does not have staff on board with the necessary skills to manage partnering arrangements, how will it attract and retain that talent? Information about how well existing partnering arrangements leverage nonfederal resources is important to inform decisions about continuing arrangements or entering into new ones. Gathering this information also presents an opportunity to evaluate progress toward a project’s intended goals. It is important to evaluate the role of partners at an organizational level to help ensure nonfederal resources are effectively leveraged. To better understand the role of partners at the Smithsonian Institution and to promote internal and external collaborations, officials have completed various reports over time that identify the strengths, weaknesses, opportunities, and threats to partnering. These reports also focused on how the Smithsonian Institution can better develop a collaborative culture. In a February 2012 report, officials compiled an inventory of active collaborations with external partners. A separate April 2009 report recommended steps the Smithsonian Institution could take to facilitate collaboration. In addition, a 2008 Smithsonian Institution task force report evaluated the management of revenue-generating activities and recommended restructuring Smithsonian Enterprises to improve capacity, cost, and focus. Evaluating the effectiveness of specific partnering efforts can serve an important role in improving partnering arrangements. In past work, we found that agencies that create a means to monitor, evaluate, and report the results of collaborative efforts can better identify areas for improvement. Evaluations can also serve as a means to provide feedback to private donors that funded a program. For example, the National Gallery of Art conducted an external evaluation of its Art Around the Corner program’s outcome and participant impact. The December 2012 evaluation assessed the program’s mission and participant impact through interviews, and identified potential program improvements. National Gallery of Art officials have begun discussions about how to address the recommendations identified in the evaluation and also plan to use this as a means to report back to the private donors that funded the program. CCOs have also evaluated the success of potential programs by conducting pilot programs. For example, the Smithsonian-Mason School of Conservation used a pilot to determine the effectiveness of its business model and academic curriculum before fully implementing the program. Five pilot programs were conducted between 2008 and 2011 to evaluate the school’s curriculum and determine whether the program could become financially self-sustaining. These pilots allowed officials to evaluate the program’s financial and operational feasibility before fully committing resources. Upon the successful completion of these pilots, the school has been fully operational and has significantly increased the size of its program. To what extent are partnering arrangements used to better leverage nonfederal resources? How are partners and their performance evaluated? What data-based tools are available to determine whether a partnering arrangement is leveraging nonfederal resources effectively? What lessons learned from other partnering arrangements are used to inform new partnering decisions? Increasingly, the federal government relies on networks of partners to achieve critical results, often including multiple federal agencies, sectors, and levels of government. CCOs are in a unique position to leverage nonfederal resources by working with partners to produce greater public value than they can achieve alone, especially given the special managerial flexibilities and legal authorities they enjoy. Chief among these flexibilities is the ability to solicit private funds and accept gifts. These nonfederal resources are particularly valuable in light of today’s constrained budget environment, in which agencies may no longer expect regular increases in their budgets. While the four CCOs we studied—the Holocaust Memorial Museum, the National Gallery of Art, the Presidio Trust, and the Smithsonian Institution—benefit from millions of nonfederal, nonappropriated dollars that provide programmatic, research, and operational support, information about how these CCOs leveraged nonfederal funds are not reported in a timely, consistent manner. Congress does not direct the CCOs’ use of nonfederal funds. However, because the federal budget process is the primary means by which Congress evaluates competing demands for federal funds, it is essential that Congress has timely, sufficient information about the nature and scope of all resources—federal and nonfederal—being used to serve a public purpose. Absent a requirement to present this information in a timely, transparent fashion and make it available for use in congressional budget deliberations, Congress will lack complete information about CCOs’ federal and nonfederal financial resources. Further, CCOs may be missing an opportunity to share good practices and strategies for leveraging resources and strengthening relationships with private and nonprofit partners in new, more cost- effective ways. To provide more timely, complete information about CCOs’ fiscal health, and increase awareness about good practices and strategies for leveraging resources from nonfederal partners, congressional committees should consider requiring CCOs under their jurisdiction to report on their total nonfederal funds—including a breakdown of the amounts and uses of these funds—in their annual budget requests. In requiring this information, Congress should also consider what types of information on CCOs’ nonfederal funds would be helpful to them, how that information should be reported and at what level of detail, and whether the information should be presented consistently across CCOs. We provided the Executive Director of the Holocaust Memorial Museum, the Director of the National Gallery of Art, the Executive Director of the Presidio Trust, and the Secretary of the Smithsonian Institution with a draft of this report for review and comment. All of the CCOs generally agreed with the findings and conclusions in this report. They also provided technical comments, which we have incorporated, as appropriate. The Director of the National Gallery of Art provided written comments that we have reprinted in appendix II. In written comments, the Director of the National Gallery of Art agreed that Congress should have timely information about the nature and scope of nonfederal resources. However, he noted that Congress should not direct or restrict the use of private funding. Instead, the Board of Trustees has full fiduciary responsibility over nonfederal funds. We agree that Congress does not direct CCOs’ use of nonfederal funds. We have clarified this point in the report, as appropriate. We are sending copies of this report to the Holocaust Memorial Museum, the National Gallery of Art, the Presidio Trust, the Smithsonian Institution, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6806 or irvings@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix II. Smithsonian Enterprises, the institution’s revenue-producing organization, works with nonfederal partners in its three major business divisions: (1) The media division manages the institution’s magazines, books, and the Smithsonian Channel, (2) the retail operation manages the museum and airport stores, three IMAX theaters, and food concessions, and (3) the consumer products division manages agreements with more than 75 companies permitting the use of the Smithsonian name following collaboration with staff and curators. Smithsonian Enterprises works closely with the private sector to provide the services in each of their business divisions. The Smithsonian Grand Challenges Consortia help foster a spirit of interdisciplinary collaboration to stimulate intellectual exchange within the institution and beyond. The Consortia develop and launch collaborations, research centers, and programs that address the Smithsonian Institution’s four Grand Challenges: (1) Unlocking the Mysteries of the Universe, (2) Understanding and Sustaining a Biodiverse Planet, (3) Valuing World Cultures, and (4) Understanding the American Experience. The institution partners with George Mason University to create, manage, and operate the Smithsonian-Mason School of Conservation. The program offers residential, hands-on, interdisciplinary programs in conservation biology for undergraduate and graduate students and professionals at a jointly developed facility in Front Royal, Virginia. The Smithsonian Affiliations program shares the institution’s collections, scholarship, and exhibitions with Americans in their own communities by collaborating with museums and cultural and educational organizations. The program brings the institution’s resources to local communities through its 177 affiliates in 41 states, Puerto Rico, and Panama. Through the Encyclopedia of Life the Smithsonian Institution partners with other research organizations to compile and make available scientific research, data, and information to users worldwide. Its goal is to create “a webpage for every species” by bringing together trusted information from resources across the world such as museums, learned societies, expert scientists, and others into one database and a single, easy-to-use online portal. Volunteers provide office and event support by offering programmatic and administrative services, such as working on special projects, staffing information desks, providing docent tours, and caring for animals. The National Gallery of Art collaborates with conservation scientists and conservators, conservation laboratories, and universities both nationally and internationally on art conservation. The Center for Advanced Study in the Visual Arts collaborates with historians, critics, art theorists, and social science and humanities scholars to study the production, use, and cultural meaning of art, artifacts, architecture, urbanism, photography, and film, from prehistoric times to the present. Art Around the Corner partners with the District of Columbia Public School System’s elementary schools to bring fourth and fifth graders to the National Gallery of Art up to 14 times over 2 years to experience art through personal and interdisciplinary connections. In small groups, students look closely at works of art, engaging in open-ended discussion, role-playing, sketching, art making, and creative writing. The National Gallery of Art partners with educators to disseminate information and foster an understanding and appreciation of art. It provides resources for educators and students such as free interactive lesson units online that include lesson plans, worksheets, and student activities. Volunteers assist the public by staffing the information desks at the museum as well as providing docent tours of the art for the public. The museum’s Center for Advanced Holocaust Studies worked with international academics to support scholarship and publications in the field of Holocaust studies, promotes the growth of Holocaust studies at American universities, seeks to foster strong relationships between American and international scholars, and initiates programs to ensure the ongoing training of future generations of scholars specializing in the Holocaust. The Holocaust Memorial Museum’s World Memory Project program leveraged individual volunteer’s efforts to identify photographs of thousands of children. Further, the public helps to piece together information about the children’s wartime and postwar experiences and facilitate renewed connections among these young survivors, their families, and other individuals who were involved in their care during and after the war. The Genocide Prevention Task Force was jointly convened by the Holocaust Memorial Museum, U.S. Institute of Peace, and the American Academy of Diplomacy and funded through private foundations. Its goals included spotlighting genocide prevention as a national priority and developing practical policy recommendations to enhance the capacity of the U.S. government to respond to emerging threats of genocide and mass atrocities. The traveling exhibitions program allows different organizations to borrow exhibitions and thereby extend its educational activities to a broader audience. Since 1991, the museum’s traveling exhibitions have been to 150 cities in 45 U.S. states, as well as Canada and Germany. Volunteers worked to assist the Holocaust Memorial Museum’s operations such as visitor services special projects, clerical work, research, translation, and transcription. To augment the work of the museum, the Office of General Counsel said that law firms are periodically engaged to provide pro bono and discounted paid legal services, such as research on international law, advice on specific aspects of intellectual property law, and advice on major mediations. The Presidio Trust employs a firm specializing in residential leasing to manage residential homes in 21 neighborhoods, and a nonresidential management firm to manage leases for nonresidential building space. The Presidio Trust oversaw the rehabilitation of a historic building into the Inn at the Presidio and entered into a management service agreement with a hotel management company to manage and operate the inn. Presidio Trust Special Events issues permits for recreational uses. The Presidio Trust hospitality department rents out seven event venues for meetings and private events. The Presidio Trust partners with the National Park Service, academics, and researchers on archeology, historical preservation, and open space restoration. The Presidio Trust partners with the National Park Service, Golden Gate National Parks Conservancy, Presidio YMCA, and other organizations to provide cultural, recreational and natural resource programming. In addition to the contact named above, Jacqueline M. Nowicki, Acting Director, Melissa Emrey-Arras, Shirley S. Hwang, Melissa L. King, and Catherine H. Myrick made major contributions to this report. Also contributing to this report were Mallory Barg Bulman, Amy R. Bowser, Elizabeth Erdmann, Robert L. Gebhart, Mehrzad Nadji, Amy J. Radovich, Cynthia M. Saunders, Sabrina C. Streagle, and Sarah Veale.

Through congressional charters, Congress has created independent organizations which receive support from federal and nonfederal sources. These organizations, known as CCOs, are authorized to receive and retain financial and nonfinancial resources from nonfederal partners to help meet their core mission and goals. In 2012, GAO was directed to study CCOs. To determine whether selected CCOs offer lessons learned to facilitate the leveraging of nonfederal resources, GAO studied (1) factors, if any, that facilitated selected CCOs' ability to partner with the nonfederal sector and (2) key principles to better leverage resources through nonfederal partners. To do this, GAO reviewed relevant federal laws, regulations, and policies; analyzed relevant legal authorities, agency documents, and prior GAO reports; conducted site visits to the four CCOs; and reviewed literature on partnerships. GAO compiled key principles, discussed and validated them with subject matter specialists and the four CCOs, and incorporated their feedback, as appropriate. Four factors facilitated the ability of the U.S. Holocaust Memorial Museum, National Gallery of Art, Presidio Trust, and Smithsonian Institution to leverage nonfederal resources: (1) unique legal authorities and management flexibilities; (2) benefits received from these congressionally chartered organizations' (CCO) federal status; (3) governing boards that provided management and oversight; and (4) informal networks that enabled CCOs to share lessons. A critical flexibility is the ability to accept gifts and solicit private donations, but the CCOs in this study are not required to, and did not always provide, a complete picture of nonfederal resources to Congress. The federal budget process is the primary means by which the President and Congress select among competing demands for federal funds; as such, it is essential that budget information be comprehensive and clear. While Congress does not direct the CCOs' use of nonfederal funds, consistent and timely information about CCOs' total resources could provide important context for understanding both the relative tradeoffs among funding decisions and the implications of such decisions. GAO compiled six key principles to guide CCOs' management decisions about leveraging resources through nonfederal partners. 1. Make partnering decisions in line with mission. Organizations that leverage partnering arrangements have clear, well-articulated missions; strategic goals to achieve them; and a defined process for assessing whether partnering arrangements complement their missions and goals. 2. Ensure top leadership support for partnering arrangements. Top leadership support is critical to successfully pursuing and engaging partners. The tone at the top--management's philosophy and operating style--sets the stage for how the organization will make management decisions related to partnering. 3. Assess and manage risks . Partnering decisions should reflect both the likely risk and the organization's tolerance for risk in partnering. Incorporating risk assessment and risk management practices into partnering decisions can help ensure that the organization recognizes and is prepared to manage explicit risks (e.g., financial and physical) and implicit risks (e.g., reputational). 4. Select complementary partners and appropriate projects. Partners should bring complementary resources, skills, and financial capacities to the relationship. A systematic approach helps to identify projects that are well-suited for partnering opportunities and helps to achieve an organization's mission. 5. Manage partnering arrangements. Partnering arrangements are relationships that should be managed actively. Formalizing collaborations between the partners, including documenting dispute resolution processes, can enable productive partner interactions. Further, it is important to have the staff with the right skills and experience to manage these opportunities. 6. Evaluate partnering arrangements. Information about how well existing partnering arrangements leverage nonfederal resources could inform decisions about continuing arrangements or entering into new ones. Gathering this information also presents an opportunity to evaluate progress toward a project's intended goals. To provide more complete information about CCOs' fiscal position and strategies for leveraging resources from and strengthening relationships with nonfederal partners, congressional committees should consider requiring that the CCOs under their jurisdiction report on their total nonfederal funds--including a breakdown of the amounts and uses--in their annual budget requests. All of the CCOs GAO studied generally agreed with the report's findings and provided technical comments, which GAO incorporated, as appropriate.

VA arranges for the delivery of home health care services to veterans through two methods. Under its HBHC programs, teams of VA hospital staff deliver primary care services directly to veterans in their homes. Under VA’s second method, community-based providers deliver home health care services to veterans. VA pays for these community-based provider services through its fee-based home health services program unless the veteran is covered by Medicare’s home health care benefit. Generally, for these veterans, VA hospital staff facilitate the delivery of these services and Medicare pays for them. Begun in July 1972, VA’s HBHC program consists of individual HBHC programs affiliated with hospitals and medical centers around the country. HBHC is an extended care program designed to meet the long-term care needs of veterans who have chronic multiple medical and psychosocial problems, a terminal illness, or a need for post-hospital rehabilitation or monitoring. The objectives of the program are to provide primary care services to homebound patients in their homes; create a therapeutic and safe home environment; support the caregiver—the veteran’s spouse, other family member, or friend—in caring for the patient; reduce the need for, and provide an alternative to, hospitalization or other institutionalization; promote timely discharge of patients from hospitals or nursing homes; and provide an academic and clinical setting for students of the health professions. Veterans may or may not be charged a fee for HBHC services, depending upon their eligibility for outpatient services. In fiscal year 1994, VA served 9,953 veterans under this program. VA’s fee-based home health services program pays community-based providers to provide home health care services to veterans who received inpatient care for an acute condition at a medical center and have been discharged. It also pays for skilled medical treatment for other veterans entitled to VA medical care. Whether veterans are charged fees for services under this program depends upon their eligibility for outpatient services. In fiscal year 1994, VA served 12,800 patients under this program. In addition to the veterans receiving hospital-based and fee-based care, VA facilities referred at least 19,000 Medicare-eligible veterans to Medicare-certified health care agencies in fiscal year 1994. For hospitals with an HBHC program, the decision about whether a patient receives HBHC or community-based care is made by hospital staff on the basis of such factors as the patient’s medical condition and the types of services needed. If HBHC would best meet the patient’s needs, staff would try to use that program to provide home health care. If the hospital does not have an HBHC or if staff determine that community-based care would meet the veteran’s needs, the next step is determined by the veteran’s Medicare eligibility. Generally, if a veteran’s care would not be covered by Medicare, VA will pay for a veteran’s home health care services under the fee-based program. VA staff facilitate the delivery of home health care to veterans with Medicare coverage by referring them to community-based providers with the understanding that Medicare will pay for their services. However, VA continues to provide all medical care follow-up, drugs, and supplies that are not paid for by Medicare. VA hospitals are not required to have HBHC programs, and most do not. In fiscal year 1995, VA operated 74 HBHC programs: 73 of its 173 hospitals had an HBHC program, and one additional program was operated by an outpatient clinic. A VA hospital can operate an HBHC program if it meets certain criteria and receives VA approval. For example, guidance from VA’s Central Office requires hospitals to demonstrate that they discharge a required number of veterans with specified medical conditions and that they have staff to provide services to the veterans once they are discharged. Hospitals may also terminate their HBHC programs in response to other demands on their staff and budget. Between fiscal years 1990 and 1995, the number of programs increased from 72 to 74, with five VA hospitals initiating HBHC programs and three terminating their programs during this time. HBHC patients tend to be chronically ill and in need of long-term care, although some may be terminally ill or need short-term care following hospitalization. In fiscal year 1994, half of the HBHC programs reported to VA that their patients had an average age of 71 years or less. There is no limit on how long veterans may stay in the program. During fiscal year 1994, half the questionnaire respondents that had an HBHC program indicated that their patients stayed in the program an average of 7 months or less. As of March 31, 1995, about 44 percent of the patients had been in the program 1 year or longer. In order to qualify for HBHC services, veterans must be homebound; should have a caregiver—such as a spouse, family member, significant other, or friend—to assist with their care; must live within a defined geographic area—usually a 30-mile radius of the must generally only need services Monday through Friday during normal must be entitled to VA medical care. Under the HBHC program, an interdisciplinary team of VA hospital staff delivers primary care services to veterans in their homes. A typical program consists of several nurses, a social worker, a coordinator, a clerk, and the part-time services of a physician and dietician. Some programs also have a physical therapist, occupational therapist, or home health technician. Skilled nursing is the predominant service of HBHC programs: 99 percent of the HBHC programs report providing this service. Skilled nursing includes such activities as changing dressings, teaching patients how to manage their medical problems, administering medications, and drawing blood samples for laboratory analysis. Nursing services are provided in accordance with the patient’s plan of care. HBHC social workers provide psychosocial services, such as counseling and resolution of social and emotional problems that affect treatment or impede medical recovery. Ninety-six percent of the HBHC programs provide these services to veterans in their homes, according to questionnaire respondents from hospitals with HBHC programs. HBHC physicians usually work in the programs on a part-time basis and have primary medical responsibility for all HBHC patients. Their responsibilities include identifying the patients’ medical problems, defining the medical management of the problems, and determining whether to admit HBHC patients to the medical center. Approximately 80 percent of the respondents said their HBHC program provides physician services in veterans’ homes. Dietician services are another important component of HBHC programs, with almost 9 out of every 10 programs reporting that their dieticians make home visits. HBHC dieticians assess patients’ nutritional needs over time in relation to changes in their condition. They also teach patients and their caregivers how to adapt and modify their food preparation practices. Responses to our questionnaire showed that most HBHC programs in fiscal year 1995 did not provide skilled physical, occupational, or speech therapy; home health aide services; or pharmacy services to veterans in their homes. In fiscal year 1994, VA reported expenditures of $36.6 million on this program. Most veterans receive home health care services from community-based providers through either VA’s fee-based program or Medicare’s home health care benefit. These veterans tend to need short-term home health care associated with an acute medical condition, although some have longer-term needs. Many of the services commonly provided by the fee-based program and Medicare are the same, but more types of services are generally available to veterans covered by Medicare. Nearly all VA hospitals used the fee-based program in fiscal year 1995 to purchase skilled home health care services from community-based providers. Most veterans in this program receive short-term home health care services to address acute medical conditions, such as hip fractures or surgical wounds. Some veterans, however, receive long-term home health care services to address chronic conditions, as in the case of a patient with Parkinson’s disease, for example, who has an ongoing need for skilled services, such as intramuscular injections. Half of our questionnaire respondents reported that their fee-based patients in fiscal year 1994 had an average age of 61 years or less. VA will authorize payment to community providers for a maximum of 12 months following a veteran’s hospital discharge, and reauthorization can be extended upon the approval of a VA physician. During fiscal year 1994, half of our questionnaire respondents indicated that their fee-based patients had an average length of stay in the program of 90 days or less. To qualify for this program, veterans must be entitled to VA medical care. Veterans are not required, however, as in the HBHC program, to have a caregiver at home, live within a certain distance of the hospital, or generally need services only during certain hours of the work week. Skilled nursing is the predominant service covered by the fee-based program. Nearly all respondents to our questionnaire said that they usually purchase skilled nursing services for veterans. Physical, occupational, and speech therapy services are each purchased for patients in the fee-based program in over half of the medical centers, according to questionnaire respondents. Physical therapists work with patients to improve their capacity to perform simple daily activities, and occupational therapists assess patients’ rehabilitation needs, develop plans of care, and provide training. Speech therapists help patients such as stroke victims improve their ability to communicate. The fee-based program also covers the services of physicians and psychologists. However, only about one-fifth or fewer of the respondents said that they usually purchase these services under the fee-based program. Although VA cannot use fee-based program funds for home health aide services, 22 percent of the respondents said they provide this service to veterans in that program. Medical centers can pay for these services through VA’s Homemaker/Home Health Aide program. In fiscal year 1994, VA reported payments of $27.3 million on fee-based home health care services. Most respondents to our questionnaire indicated that they refer veterans covered by Medicare’s home health care benefit to community-based providers. Most people who receive Medicare home health care benefits do so for services associated with an acute medical condition, often following hospitalization. Since 1989, however, Medicare has been providing more long-term home health care services to chronically disabled elderly beneficiaries. VA officials told us that Medicare-eligible veterans follow the same basic pattern, with most receiving short-term home health care services around an acute medical condition requiring hospitalization. We were unable to determine how long veterans referred by VA hospitals received Medicare-funded home health care. However, in 1992, about three-quarters of the general Medicare population that used the home health care benefit received services for fewer than 120 days; the average duration of services for these beneficiaries was 42 days. To qualify for Medicare home health care coverage, beneficiaries must be eligible for Medicare (almost all elderly and some disabled people), homebound, in need of skilled nursing or therapy services on a part-time or intermittent basis, and under the care of a physician who prepares and periodically reviews their care plan. The VA physician usually fulfills Medicare’s physician requirement for veterans discharged from a VA hospital. Unlike the HBHC program, Medicare does not require beneficiaries to have a caregiver at home or live within a certain radius of the hospital that discharged them. Skilled nursing and home health aide services—along with physical therapy, occupational therapy, speech therapy, and medical social services—are the home health care services that VA usually facilitates for veterans covered by Medicare. Approximately 70 percent or more of the respondents to our questionnaire indicated that they facilitate the delivery of these services. Physician, dietician, and pharmacy services are not covered under Medicare’s home health care benefit. However, physician services are covered by other parts of the Medicare program; pharmacy and dietician services may be covered by Medicare under some circumstances. Medicare, and not VA, pays for community-based providers to deliver home health care services to veterans covered by Medicare’s home health care benefit. VA incurs some administrative costs in referring patients to Medicare, as well as the costs of VA physicians’ developing and reviewing plans of care, but VA does not separately identify these costs. Data on the costs of VA’s home health care programs are reported differently, both among HBHCs and between HBHCs and fee-based programs. As a result, to the extent that VA administrators make decisions about whether to have an HBHC program on the basis of the relative costs of HBHCs and fee-based programs, they do so on the basis of their perceptions of cost rather than comparable data. HBHC program costs are based on data developed by the hospitals that support the programs. VA Central Office officials told us that hospitals have wide latitude in deciding which costs to charge to their HBHC programs. This results in different cost charges among the 74 HBHC programs. For example, some HBHC programs include costs of librarians or chaplains, while others include costs of anesthesiologists or optometrists. In addition, hospitals commonly charge costs to their HBHC programs for certain administrative support functions, such as costs for a portion of one full-time-equivalent staff person in the Office of the Chief of Staff. One hospital, for example, charged $8,600 for support from the Chief of Staff’s Office in fiscal year 1994. However, we found another case in which a hospital charged $80,500 for approximately 2 full-time-equivalent staff from the Chief of Staff’s Office. A VA Central Office official agreed that a charge this high was an error. Central Office officials further stated that they discuss questionable charges that appear in VA’s cost reports with hospital staff but that it is up to the hospitals to appropriately allocate costs. VA’s reported costs of its fee-based program, on the other hand, represent payments made to community-based providers but exclude costs such as program administration and other indirect costs associated with caring for veterans in this program. For example, costs for staff who administer the program are included in the operating costs of the hospitals where the staff work and are not identified as a cost of the fee-based program. In addition, approximately 70 percent of our questionnaire respondents stated that they case-manage fee-based patients, yet costs associated with case management are not included in the fee-based program. Since VA reports the costs of its programs differently, VA hospital officials are left to make decisions on whether or not to have an HBHC program based on their perceptions of the relative cost of HBHC and fee-based programs. These perceptions vary widely. For example, about 3 years ago, the Tampa HBHC program began treating patients who previously would have received fee-based care. One reason for doing so was to reduce the costly fee-based payments for nursing services. The Kansas City Missouri hospital, on the other hand, terminated its HBHC program in 1994 and referred some of its HBHC patients to community-based providers. A former HBHC official told us that hospital administrators believed that it was less costly to pay for community-based services than for an HBHC program. Respondents to our questionnaire also expressed very different views regarding costs and why they either have or do not have an HBHC program. Over three-fourths of the questionnaire respondents with an HBHC program stated that one reason their medical center has an HBHC program is that it is less costly than purchasing fee-based services. Conversely, approximately half of the respondents without an HBHC program said that one reason they do not have an HBHC program is that it would be more costly than community-based care purchased under the fee-based program. Respondents to our questionnaire also expressed very different views on whether HBHC was more cost effective as compared with community-based care. In this instance, cost-effectiveness refers not only to the actual costs incurred in treating a veteran but also to the effectiveness of the care. For example, who has fewer hospital admissions and shorter hospital stays: HBHC patients receiving primary care services or patients receiving skilled services from community-based providers? About 40 percent of our respondents said they had no basis to judge whether HBHC was more or less cost effective than community-based care. The remaining respondents were evenly divided on which method of providing home health care to veterans was the more cost effective. VA monitors the quality of care provided by its home health care programs, but it is more directly involved in monitoring the care its own employees provide, through HBHC, than the care delivered by community-based providers. Licensing and certification assessments of community-based providers conducted by independent organizations provide VA some assurance that veterans in the fee-based program and those covered by Medicare’s home health care benefit receive care from qualified home health care providers. HBHC programs are assessed by outside organizations as well, but in addition, they use case management and quality indicators to evaluate the care they deliver. Medical centers are less likely to use these additional means to monitor the quality of care delivered by community-based providers. Medicare has the primary responsibility for ensuring that quality care is furnished under its home health care benefit. Medicare requires community-based providers to have internal quality assurance programs. Moreover, Medicare requires VA physicians referring patients to prepare and review plans of care. All HBHC programs are accredited by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and are subject to a performance review every 3 years. JCAHO staff apply standards contained in their Accreditation Manual for Home Care to evaluate how well the home health care provider assessed the patient’s service needs, planned for the patient’s care, and monitored the patient’s response to the care provided. Before 1995, JCAHO’s reviews did not measure actual outcomes of care but instead focused on processes and the capacity of a provider to deliver quality care. JCAHO’s 1995 standards, however, place more emphasis on outcomes of care. Community-based home health care providers that are certified to treat Medicare beneficiaries are assessed at least once every 15 months by a state survey agency (usually a component of the state health department), by JCAHO, or by the National League for Nursing’s Community Health Accreditation Program (CHAP). These surveys are intended to ensure that Medicare beneficiaries receive care from qualified home health care providers. State survey agencies, under contract with the Health Care Financing Administration (HCFA), which administers Medicare, survey community-based providers to determine if they meet Medicare’s conditions of participation. These conditions cover such topics as acceptance of patients, medical supervision, and skilled nursing services. Survey staff visit home health care providers and examine organizational, functional, personnel, and patient records and visit with patients in their homes to evaluate the quality and scope of services provided. Home health care providers that meet JCAHO’s or CHAP’s standards are also deemed to have met Medicare’s conditions of participation. Thus, veterans in HBHC programs and those receiving home health care services covered by Medicare are assured of receiving care from Medicare-certified providers. Most, but not all, fee-based patients also receive care from Medicare-certified providers. About 83 percent of the respondents to our questionnaire said they purchase care from Medicare-certified providers most, almost all, or all of the time for their fee-based patients. For example, the three hospitals we visited use only Medicare-certified home health care providers. These hospitals select the providers on the basis of factors such as which providers serve the area where the veteran needing home health care lives, whether the provider is Medicare-certified, and whether the hospital’s past experience with the provider has been positive. However, fee-based patients who are paralyzed do not always receive care from Medicare-certified providers because the fee-based program allows family members who have been trained and certified by VA to deliver bowel and bladder care to quadriplegic veterans. Also, a VA Central Office official told us that there are not enough Medicare-certified providers in some areas of the country. VA’s programs also employ case management to ensure quality care. Case management, in general, involves coordinating the services needed by and provided to a patient. Case management may address only a veteran’s health needs or the total needs of a veteran and his or her family. Similarly, case management may be limited to the process of arranging initial services or may be an ongoing process for the duration of an illness. Not everyone discharged to home health care from a hospital needs case management; some veterans and their families may act as their own case managers. When case management is appropriate, it is seen as important to the adequacy of the home health care veterans receive: 68 percent of the questionnaire respondents said that case management greatly improves the adequacy of care patients receive. The way in which a veteran’s care is managed by VA may differ somewhat depending upon the program that is providing for the veteran’s services. Nearly all (96 percent) of the respondents with HBHC programs said that they typically case manage patients and that nurses are usually the primary case managers. For veterans receiving fee-based services, most respondents (73 percent) said the VA physician who orders the home health care reviews the provider’s periodic reports on the patient’s health status most, almost all, or all of the time, and 62 percent said someone (most often a nurse) serves as a case manager for patients. For veterans receiving care paid for by Medicare, about the same number of respondents (72 percent) said the VA physician who orders the home health care reviews the provider’s periodic reports most, almost all, or all of the time, but fewer (49 percent) said that someone (usually a nurse) case manages the patient’s care. Respondents also described different primary functions of case managers in the different programs, as shown in table 1. Nearly all respondents with an HBHC program said that their case managers performed the primary functions listed in the table, while respondents without an HBHC program indicated that their case managers were much less likely to perform these functions. The greatest difference among respondents was that evaluating the patient’s home environment was much less likely to be a primary function of the community-based program case manager than it was to be a function of the HBHC case manager. The HBHC and community-based programs we visited replicated the difference in primary case management functions described in the questionnaires. Case management at the Boston and Tampa HBHC programs involves each of the six functions cited in table 1. For example, each program holds team meetings at least weekly to discuss veterans’ status and needs. The Boston, Tampa, and West Roxbury hospitals also case manage patients in their community-based programs, but the management is not as extensive as that conducted by HBHC programs. For example, case management at West Roxbury is limited to defining a plan of care and arranging for the veteran to receive it. At the Tampa hospital, case managers participate with community provider staff in assessing the care needs of patients, coordinate development of patient care plans, and periodically review patients’ medical conditions with community provider staff. However, case managers at Tampa do not have the primary functions of directly monitoring and managing the overall delivery of home health care services provided patients under the fee-based or Medicare programs or, for those patients covered by Medicare, directly evaluating the patient’s home environment and ability of the caregiver to meet the care needs of patients. Our discussions with medical center staff in the three locations we visited suggest some additional reasons why the primary functions of case management in the community-based programs may differ from those in HBHC. First, case managers for veterans receiving community-based care may also be responsible for a variety of other functions, leaving less time to devote to case management. For example, a Tampa medical center nurse who manages veterans’ cases in the fee-based and Medicare programs told us that she coordinates services for two other sets of veterans as well: those that need hospice services and those that are in contract nursing homes. None of the HBHC case managers we spoke to told us that they had similar responsibilities for veterans outside of the HBHC program. Second, medical center case managers may be responsible for arranging home health care services for a large number of veterans as compared with the number of veterans managed by HBHC case managers, which would also leave them less time to devote to case management functions. The two nurse case managers at the Boston medical center, for example, referred 1,074 veterans to community providers in fiscal year 1994. In contrast, Boston’s three HBHC case managers were responsible for case managing and providing care to 112 veterans that same year. Another way VA assesses veterans’ care is by monitoring performance indicators that VA believes are related to the quality of care provided. For example, if a provider frequently used by a VA hospital has high rates of patient deaths or patients’ being readmitted to the hospital, visiting an emergency room, falling, having impaired skin integrity, or getting an infection, this may reflect a problem with the care being provided. Patient satisfaction is also useful as a way of assessing the quality of care. For example, one HBHC program we visited set a standard that at least 90 percent of its veterans would be satisfied with their care, as measured by a patient satisfaction survey. In the fourth quarter of fiscal year 1994, 99.5 percent of the veterans surveyed said that they were satisfied with their care. Indicators such as these are useful to assess performance, identify problems, develop corrective actions, and monitor the effectiveness of the changes made. As table 2 shows, more medical centers track selected quality indicators for their HBHC programs than for their community-based programs. We did not ask how often the community-based providers themselves track these indicators for their patients. Although VA hospitals less often track these quality indicators for community-based providers and patients receiving their care from those providers, they do take other steps to ensure that veterans in community-based programs receive quality care. Eighty percent of questionnaire respondents stated that they have periodic telephone or personal contacts with provider staff most, almost all, or all of the time to discuss the health status of veterans in the fee-based program, and 65 percent said that they have similar contacts for veterans that are covered by Medicare’s home health care benefit. About 80 percent said that they require providers to submit periodic written reports regarding the health status of veterans in the fee-based program most, almost all, or all of the time, and about 70 percent said they ask for similar reports for those veterans in the Medicare program. Further, VA officials told us that hospitals evaluate patients’ medical conditions when they have an inpatient or outpatient visit at the hospital. Because VA’s home health care programs provide different arrays of services to veterans who generally have different home health care needs and because consistent program cost data are not available, it is difficult to compare the relative costs of VA’s methods of meeting veterans’ home health care needs. And although VA itself more directly monitors care provided under its HBHC program, the quality of care furnished by community-based providers paid for by both VA and Medicare is evaluated in other ways—including by HCFA as part of its responsibility for administering the Medicare program. We obtained comments on a draft of this report from VA officials, including the Deputy Under Secretary for Health. The officials noted that the lack of consistent cost data, in this case for the various types of home health care provided, is a problem not unique to VA and is a challenge for all health care providers. They said that VA, in making improvements to its financial management and information systems, is attempting to better identify costs associated with all of its programs, including each component of its home health care programs. The officials also told us that VA is developing performance measures that will allow managers at multiple levels to understand and identify desired program outcomes. Once these outcomes are in place, managers will be accountable for meeting them in the most cost-effective and efficient manner. VA officials also said that they intend to do cost-benefit analyses for home health care and other programs, once enough data are available. The VA officials additionally suggested some technical changes, primarily for clarification, which we incorporated as appropriate. As arranged with your staff, unless you announce its contents earlier, we plan no further distribution of this report for 7 days after its issue date. At that time, we will send copies to the Secretary of Veterans Affairs, the Senate and House Committees on Appropriations, and other interested parties. We will also make copies available to others upon request. This report was prepared under the direction of James Carlan, Assistant Director, Health Care Delivery and Quality Issues. If you or your staff have any questions concerning this report, please contact Robert Dee, the evaluator-in-charge, at (617) 565-7470. Other staff contributing to this report were Sally Coburn, Patricia Jones, Clarita Mrena, Joan Vogel, and Leonard Hamilton. To develop our description of the ways VA provides home health care, we obtained both nationally descriptive data and additional data at three locations. Collecting these data required audit work at VA’s Central Office as well as at the Boston, Massachusetts, and Tampa, Florida, medical centers and at the West Roxbury division of the Brockton Medical Center in Massachusetts. In addition, we reviewed VA’s policies and procedures for operating its HBHC and fee-based programs as well as Medicare’s regulations concerning its home health care benefit. We also obtained various VA reports detailing operations of its HBHC and fee-based programs. We selected the three hospitals we visited to give us examples of hospitals that have an HBHC program (the medical centers in Boston, Massachusetts, and Tampa, Florida) and one that does not (West Roxbury, a division of the Brockton Medical Center). At these locations, we interviewed staff and obtained documents about their programs. Additionally, we visited 30 veterans in their homes to understand better how their home health care services were provided; 12 were receiving HBHC services, while 18 were receiving community-based services. Of the 12 veterans receiving HBHC services, 6 were cared for by the Boston HBHC program and 6 by Tampa’s program. Of the 18 veterans receiving community-based care, we visited 6 discharged by each of the three VA hospitals. Seventeen of the 18 veterans were covered by Medicare’s home health care benefit, and 1 was covered by VA’s fee-based program. During our visits, a registered nurse on our staff interviewed veterans and discussed their care and activities, observed their medical conditions, and reviewed information from their medical files. We then discussed our observations with appropriate officials at the three medical centers. To obtain nationally representative data about the three home health care programs, we sent a detailed questionnaire to 158 VA medical center directors and 7 other VA health care facilities. We asked respondents to answer questions about their HBHC program, if they had one, and about their community-based program for health care. The questions covered general descriptive information, program staffing, patient admissions and discharges, program services, case management, quality assurance measures, reasons why they did or did not have an HBHC program, and other issues. We pretested the questionnaire at three medical centers, obtained comments from VA officials, and revised it accordingly. A total of 151 medical centers and 6 other VA health care facilities responded to the questionnaire. We were unable to verify independently most of the information provided through the questionnaire. However, questionnaire responses from programs at the three hospitals we visited were consistent with the information we obtained at those locations. In addition, questionnaire responses were consistent with selected aggregate information provided by the VA Central Office. Chicago (Lake Side), IL Chicago (West Side), IL (continued) Salt Lake City, UT West Los Angeles, CA (continued) White River Junction, VT The medical center did not respond to our questionnaire. Information was not provided on the questionnaire received from the medical center. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO provided information on how the Department of Veterans Affairs (VA) meets veterans' home health care needs, focusing on: (1) the characteristics and services of the home health care programs VA uses; (2) the available data describing program costs; and (3) how VA ensures that veterans receive quality home care services. GAO found that: (1) most veterans receive home health care services from community-based providers through either the VA fee-based program or Medicare's home health care benefit; (2) most veterans in these programs receive short-term home health care services for acute medical conditions, while some veterans receive long-term care for chronic conditions; (3) VA provides in-home physician, nursing, social work, and dietician services to veterans with chronic conditions through its Hospital-Based Home Care (HBHC) program; (4) VA makes decisions about using HBHC programs based on its perception of relative costs, since comparable cost data are not available; (5) HBHC program costs are based on data developed by hospitals that support the programs, while VA reported fee-based program costs represent payments made to providers and exclude certain administrative costs; (6) VA monitors the quality of care provided by HBHC programs more directly than it does community-based care; and (7) licensing and certification assessments of community-based providers provide VA assurance that care is provided by qualified sources, but VA is ultimately responsible for ensuring the quality of care in its programs.

State derives its authority to grant leave and travel reimbursements to its foreign service employees from the Foreign Service Act of 1980. To implement provisions of the act, the department issued the FAM and the FAH. Travel by State’s civil service employees is generally governed by the General Services Administration’s (GSA) Federal Travel Regulation (FTR), but in some cases is also governed by the FAM. State’s general policy is for its foreign and civil service employees to travel using coach-class accommodations provided by common carriers. However, regulations governing foreign service and civil service travel authorize the use of premium-class travel under specific circumstances. Both foreign service and civil service travel regulations require the agency head or his or her designee to authorize first-class travel in advance. These regulations also require the authorizing official at a post abroad or the executive director of the funding bureau or office domestically to authorize premium-class travel other than first class. Further, in September 2004, the Assistant Secretary of State for Administration sent a memorandum to all State executive directors emphasizing “that it is wrong to authorize premium-class travel on a blanket basis” and “that a separate justification for premium-class travel is required for each trip.” Federal and State travel regulations authorize premium-class accommodation when at least one of the following conditions exists: no space is available in coach-class accommodations, regularly scheduled flights provide only premium-class an employee with a disability or special need requires premium-class security issues or exceptional circumstances, travel lasts in excess of 14 hours without a rest stop, foreign-carrier coach-class air accommodations are inadequate, overall cost savings, such as when a premium-class ticket is less expensive than a coach-class ticket or in consideration of other economic factors, transportation costs are paid in full through agency acceptance of payment from a nonfederal source, or required because of agency mission (e.g., courier). The regulations also allow for the traveler to upgrade to premium-class accommodations, at the traveler’s expense or by using frequent traveler benefits, but the upgrade cannot be charged to the centrally billed account. State has the second largest centrally billed travel card program in the federal government. During fiscal years 2003 and 2004, State used 155 different centrally billed accounts-–143 international and 12 domestic-–to purchase more than $360 million in transportation services, such as airline tickets, train tickets, and bus tickets, for State and other foreign affairs agencies. Each bureau has its own travel budget and is responsible for obligating its travel expenses. The local travel-authorizing official or the executive director of the funding office is responsible for determining the necessity of travel, issuing the travel order, certifying the availability of funds, and recording an obligation against a unit’s appropriated funds. State’s travel management centers (TMC) make airline reservations, issue airline tickets charged to the centrally billed account upon receipt of a signed travel order, and perform a reconciliation between the tickets it issued and tickets charged on the Citibank invoice. To complete this reconciliation process, TMCs are responsible for associating each charge with a specific travel order. The financial management officer (FMO) at overseas posts and resource management’s Global Financial Operations in Charleston, South Carolina, for domestic activity, are generally responsible for reviewing a TMC’s monthly reconciliation, making appropriate changes, and certifying or authorizing Citibank’s invoice for payment. Upon receipt of the TMC’s reconciliation, billed transaction report (BTR), and supporting files, State pays Citibank for the tickets purchased on the centrally billed account. State also pays travelers for nontransportation costs claimed on their individual travel voucher. Figure 1 shows the design of the processes used to issue an airline ticket on centrally billed accounts and reimburse travelers for travel expenses. It also explains the roles of different offices in providing reasonable assurance that airline tickets charged to these cards are appropriate and meet a valid government need. Premium-class travel accounted for almost half of travel expenditures charged to State’s over 260 centrally billed accounts during most of fiscal years 2003 and 2004, including domestic and overseas operations, and this trend continued for fiscal year 2005. On the basis of our statistical sample, we estimate that 67 percent of premium-class travel during April 2003 through September 2004 for State and other foreign affairs personnel was improper--either not properly authorized or properly justified because of breakdowns in key internal controls. Examples of breakdowns in key controls include travelers flying premium-class travel when the travel orders did not authorize premium-class travel; subordinates authorizing their supervisors to take premium-class flights; and travel orders authorizing premium-class travel using criteria of a total flight time of more than 14 hours, even though the actual flight time, including layovers, was less than 14 hours. Also, State’s diplomatic couriers used premium-class travel even when it was not justified. In addition, we found that State’s top executives, including under secretaries and assistant secretaries, often used premium-class travel regardless of the length of the flight. Further, senior State officials told us that the department offered premium-class travel as a benefit to its employees, as part of their human capital initiative, for all flights lasting over 14 hours, which is allowed by federal and State regulations but is costly to taxpayers. However, State did not perform a cost-benefit analysis before offering this benefit. In comparison, agencies— such as DOD—attempt to avoid the significant additional cost associated with premium-class travel on flights lasting more than 14 hours by encouraging employees to take a rest stop en route to their final destination, saving hundreds, sometimes thousands, of tax dollars per trip. Prior to 2002, State policy prohibited the use of premium-class accommodations for permanent change of station travel even when the duration of the travel exceeded 14 hours—a prohibition established by many other agencies with staff stationed overseas. However in 2002, State eliminated that prohibition. Between April 2003 and September 2004, State and other foreign affairs agencies purchased over 32,000 airline tickets costing about $140 million that contained at least one leg of premium-class travel for State and other foreign affairs personnel using State’s centrally billed account travel cards. In addition, we determined that premium-class travel continues to be significant for fiscal year 2005. As discussed later in this report, because State does not obtain or maintain any information on premium-class travel, it cannot monitor its proper use, identify trends, or determine alternate, less expensive means of transportation. As shown in figure 2, premium- class travel represents about 19 percent of the tickets issued, and State’s and other foreign affairs agencies’ spending on premium-class travel represented about 49 percent of the $286 million spent on airfare charged to the centrally billed accounts during the period April 2003 through September 2004. Our analysis excluded all travel transactions at overseas posts that were not procured through the centrally billed travel accounts because it was outside the scope of our request. State told us that at some overseas posts travelers purchase airline tickets using Government Travel Requests (GTR) and purchase orders. Further, the information State provided for some tickets purchased with GTRs did not distinguish between premium- and coach-class tickets. Breakdowns in key internal control activities led to significant numbers of transactions lacking proper authorization and justification for premium- class travel. On the basis of our sample of premium transactions, an estimated 67 percent of premium-class travel was not properly authorized, justified, or both. Specifically, 39 percent of the premium-class airline tickets charged to State’s centrally billed account from April 2003 through September 2004 were not properly authorized. In addition, 28 percent of premium-class transactions that were authorized were not justified in accordance with either federal or State regulations. (See app. I for further details of our statistical sampling test results.) Further, State did not maintain accurate and complete data on the extent of premium-class travel and thus had a lack of controls in place to oversee and manage this travel. Each fiscal year State is required to report to GSA on first-class travel taken by all State and other foreign affairs personnel. However, we found 23 roundtrip first-class tickets valued at more than $85,000, obtained for State or other foreign affairs agencies, that were not reported by State to GSA as required in fiscal years 2003 and 2004. Further, we saw no evidence of external or internal audits of State’s centrally billed travel program. Requiring premium-class travel to be properly authorized is the first step in preventing improper premium-class travel. Federal and State regulations require premium-class travel to be specifically authorized. State travel regulations specify that premium-class travel must be authorized in advance of travel, unless extenuating circumstances or emergencies make prior authorization impossible, in which case the traveler is required to request written approval from the appropriate authority as soon as possible after the travel. Using these regulations, we found that transactions failed the authorization test in the following two categories: (1) the documentation did not specifically authorize premium-class travel or a blanket travel authorization was used to authorize premium-class travel and (2) the travel order authorizing premium-class travel was not signed. Premium-class travel was not specifically authorized. On the basis of our statistical sample, we estimated that the travel orders and other supporting documentation for 13 percent of the premium-class transactions did not specifically authorize the traveler to fly premium class, and thus the travel management center should not have issued the premium-class ticket. We estimated that an additional 17 percent of the transactions were authorized by a blanket authorization, including all diplomatic courier travel. A blanket authorization is not an appropriate vehicle for authorizing premium-class travel because federal and State travel regulations require that all premium-class travel be authorized on a trip-by-trip basis. In September 2004, State issued a memorandum to all executive directors reminding them about the use of blanket orders, emphasizing that it is wrong to authorize business-class travel on a blanket basis and also reminding the executive directors that a trip-specific justification must be provided for each business-class authorization. Travel order was not signed. We estimated that 5 percent of premium- class transactions did not have signed travel authorizations. Ensuring that travel orders are signed, and signed by an appropriate official, is a key control for preventing improper premium-class travel. If the travel order is not signed, or not signed by the individual designated to do so, State cannot guarantee that the substantially higher cost of the premium-class tickets was properly reviewed to ensure it represented an efficient use of government resources. Another internal control weakness identified in the statistical sample was that the justification used for premium-class travel was not provided, not accurate, or not complete enough to warrant the additional cost to the government. To determine whether premium-class travel was justified, we looked at whether there was documented authorization and, if there was, whether the authorization for premium-class travel was supported by a valid reason. Thirty-nine percent of premium-class transactions were not authorized and, therefore, could not have been justified. State asserts that even if business-class authorization for some trips was not properly documented, the premium travel was nevertheless justified so long as the trips were in excess of 14 hours. However, without properly documented authorization, we cannot assess the propriety of such travel notwithstanding the 14-hour travel rule and therefore must conclude that it was unjustified premium-class travel. In addition, 28 percent of premium- class transactions were authorized but were not supported by valid justification. Federal and State travel regulations provide that travel in excess of 14 hours, without a rest stop en route or a rest period on arrival is justification for premium class. We found premium travel included trips with such rest stops for flights lasting under 14 hours. Table 1 contains specific examples of both unauthorized and unjustified travel from both our statistical sample and data mining work. These examples illustrate the improper use of premium-class travel and a resulting increase in travel costs. More detailed information about some of the cases follows the table. Traveler #1 flew from Washington, D.C., to Honolulu, Hawaii. The total cost of the trip was $3,228. In comparison, the unrestricted government fare from Washington, D.C., to Honolulu was $790. According to State regulation, travelers using premium-class travel are not entitled to an overnight rest stop en route. Furthermore, the travel was authorized by a blanket premium-travel authorization signed by a subordinate of the traveler and a separate trip authorization was not included to specifically authorize this trip, as required. The travel authorization did not provide specific justification for business-class travel and the travel was not more than 14 hours. Therefore, the transaction failed authorization and justification. Travelers #2 and #3 traveled from Johannesburg to Asmara through Frankfurt, at a cost of about $8,353 each, a total of $16,706. Although they traveled business class for the entire trip, they were reimbursed for a hotel room during the layover in Frankfurt on the return visit, at a cost of about $171 each. According to State regulation, travelers using premium-class travel are not entitled to a government-funded rest stop en route. If the travelers had flown coach for this round trip and taken a rest stop en route, the airfare would have cost about $2,921 and State could have saved about $11,000 for the two tickets. One of these travelers approved the travel authorizations for both himself and the other traveler. Traveler #4 flew first class from Washington, D.C., to Hawaii on a blanket travel order that only authorized travel within Europe. Although the travel was less than 14 hours, State provided no justification for first class, and State did not report the first-class travel to GSA. We found that State issued a first-class airline ticket to Hawaii using a blanket travel authorization that authorized premium-class accommodations. State issued the ticket to an unauthorized destination–Hawaii–because the blanket travel order authorized travel to Europe and State’s travel officials did not review the blanket authorization to ensure that the travel authorization was current, valid, and the trip was to an authorized destination. Because State did not follow its own policies for authorization and review of travel, the government paid $4,155 for an unauthorized trip. State’s management allowed top State and other foreign affairs executives to use premium-class travel by approving blanket travel orders, similar to a blank check. State also allowed premium-class travel as a benefit–without considering less expensive alternatives–to other employees for flights lasting over 14 hours and for permanent change of station travel, costing taxpayers tens of millions of dollars. Further, State’s practice is for diplomatic couriers to use premium-class travel accommodations to escort diplomatic pouches. State’s top executives, including under secretaries and assistant secretaries, often used premium-class travel regardless of the length of the flight. Our data mining of frequent premium-class travelers showed that many of these travelers were senior foreign affairs executives. On the basis of this information, we expanded our data mining to include trips taken by selected presidential appointees and SES-level foreign affairs staff to determine if their travel was authorized and justified according to federal and State regulations. In addition to the federal and State regulations, we also applied the criteria set forth in our internal control standards and sensitive payments guidelines in evaluating the proper authorization of premium-class travel. For example, State travel regulations and policies do not restrict subordinates from authorizing their supervisors’ premium-class travel, a practice which our internal control standards consider to be flawed. Therefore, a premium-class transaction that was approved by a subordinate would fail the control test based on our internal control standards. State and other foreign affairs agencies paid over $1 million for 269 premium-class tickets for flights taken by 17 foreign affairs executives during April 2003 through September 2004. We found 65 tickets containing business- and first-class segments costing about $300,000 that were under 14 hours. Most of these flights were to destinations within the United States, South America, and Western Europe. Further, over $860,000 in premium-class trips taken by executives were obtained using blanket authorizations. For each premium-class trip, State regulation requires specific authorization to fly premium class. In most cases, the blanket travel orders authorized premium-class travel for an entire year and were signed by subordinates. State officials told us that because the blanket authorization allowed premium class, the executives obtained premium- class tickets even when the trip was under 14 hours. The subordinate authorizers told us they could not challenge an under secretary or an assistant secretary. Examples of premium-class trips associated with improper accommodation and their additional cost to taxpayers are included in figure 3 to illustrate the issues associated with executive premium-class travel found through our data mining. State also made a management decision to offer premium-class travel to its employees as a benefit, resulting in increased costs to taxpayers. Although State officials were aware that offering employees rest stops on longer flights was often less expensive than premium-class travel, they offered the more expensive premium-class travel to employees for all flights lasting over 14 hours, which increased costs. For example, one individual in our statistical sample flew premium-class roundtrip from Washington, D.C., to Tel Aviv at a cost of over $6,000. Although the trip lasted over 14 hours, as an alternative to paying the premium-class rates, State could have flown this employee coach and paid the cost of an overnight rest stop in London, for a total cost of about $2,300 (about $1,600 for the GSA contract airfare and $700 in lodging and per diem expenses). Overall, this option could have saved taxpayers over $3,700. State officials explained that they made these decisions about premium-class travel to improve morale and retain highly qualified foreign-service personnel. State officials also believed that, among other factors, their decisions about premium-class travel for trips in excess of 14 hours have led to increased morale, as reflected in “The Best Places to Work” survey. However, State could not provide any empirical evidence that showed a direct correlation that offering premium-class travel increased its scores on the survey or increased retention of foreign-service personnel, and could not provide evidence that travel was a metric in the “Best Places to Work” survey. In contrast, agencies, such as DOD, attempt to avoid the significant additional cost associated with premium-class travel on flights lasting more than 14 hours by encouraging employees to take a rest stop en route to their final destination, saving hundreds, sometimes thousands, of tax dollars per trip. Finally, our testing showed that all State employees, not just those in the foreign service that are governed by State regulations, were authorized to use premium-class, without constraint, when the trip was over 14 hours. State also decided to offer premium-class travel to foreign service employees for permanent change of station moves for all flights that exceeded 14 hours, in accordance with federal and State regulations. However, State’s decision resulted in increased costs to taxpayers. Permanent change of station and similar moves accounted for about $17 million (12 percent) of State’s and other foreign affairs agencies’ premium-class travel for April 2003 through September 2004. Prior to 2002, State policy prohibited the use of premium-class accommodations for permanent change of station travel, even when the duration of the travel exceeded 14 hours—a prohibition established by many other agencies with staff stationed overseas, including DOD. However, in 2002, State eliminated that prohibition at a significant cost to taxpayers. We found numerous examples in our statistical sample in which premium-class travel was properly authorized, and as such these transactions were among the 33 percent of transactions that were considered to be properly authorized and justified. However, it is important to note that because of State’s decision to treat premium-class travel as a benefit, State did not consider having the travelers take alternative, less expensive forms of travel. As mentioned, we did not evaluate whether couriers were necessary or appropriate or if there were any security issues associated with courier service procedures. same travel regulations explained earlier as all State and other foreign affairs employees. We tested diplomatic courier transactions in our statistical sample of premium-class transactions and performed data mining of fiscal year 2003 and 2004 transactions. In total, we tested over 20 diplomatic courier premium-class transactions. We found control breakdowns similar to those described above with blanket authorization and justification of courier premium-class travel. Blanket travel orders were used to authorize premium-class courier travel for all courier transactions that we tested but, as stated, blanket orders do not specifically authorize premium travel as required by State regulations. Although the Courier Service used mission security requirements to justify premium-class travel by its couriers, we found examples of premium-class travel when couriers were returning empty-handed, commonly referred to as “deadheading.” In response to these findings, Courier Service officials acknowledged that the use of premium class is not justified when couriers return empty-handed unless the 14-hour rule applies. Courier Service officials also told us that couriers may not know when they will be returning empty-handed until they arrive at an airport and are told that the post did not complete the expected outgoing pouch. By that time, they may not be able to downgrade their return ticket to economy class because a foreign airline is unwilling to do so, or time does not permit them to return to the gate to change their ticket. However, the Courier Service did not indicate on the documentation that it provided to us any attempts to downgrade their tickets in a deadheading or any other situation where premium-class travel was not justified. Further, the Courier Service Deputy Director told us that because there are still some problems in this area, they routinely check courier trip reports to identify and address any noncompliance. We found that State’s Courier Service has begun to institute cost-saving measures that, if expanded, could save taxpayer dollars. These measures include the expanded use of cargo carriers (e.g., FedEx), which do not require the couriers to purchase passenger tickets and charge lower freight costs than the commercial airlines. Our analysis of a FedEx study performed for the Courier Service showed that substantial air cargo savings and benefits could be achieved through direct cargo flights with multiple stops along a designated route. Although the Courier Service initiated the use of cargo carriers in late 2004, expanding this approach to the extent practical could achieve substantial savings. However, to achieve the additional savings, the Courier Service would need to overcome foreign mission resistance to meeting cargo aircraft outside of business hours. According to Courier Service officials, foreign mission personnel have been unwilling to meet air cargo shipments that arrive outside normal business hours and at cargo airports outside city limits. According to State, Mexico City has recently indicated a willingness to support cargo flight arrivals at Toluca airport. Courier Service officials also told us that while all agencies receiving diplomatic pouches should share responsibility for meeting and taking custody of diplomatic pouch shipments, the burden has generally fallen on State employees. Ineffective oversight and breakdowns in controls also led to problems with State’s other centrally billed travel activities. For example, although federal agencies are entitled to recover payments made to airlines for tickets that they ordered but did not use, State and other foreign affairs agencies paid for about $6 million in airline tickets that were not used and not processed for refund. We found paper and electronic unused tickets for both domestic and international flights. State was unaware of this problem before our audit because it did not monitor employees’ adherence to travel regulations and did not have a systematic process in place for TMCs to identify and process unused tickets. State also failed to reconcile or dispute over $420,000 of unauthorized and potentially fraudulent charges before paying its account. Instead of disputing these charges with Citibank, State simply deducted the amounts from its credit card bill. This action had the unanticipated consequence of substantially reducing the amount of rebates that State would have been eligible to receive. Thus, State earned only $700,000 out of a possible $2.8 million in rebates that could have been earned if State disputed unauthorized charges and paid the bill in accordance with the terms of the contract with Citibank. We asked for data on unused tickets purchased on State’s centrally billed accounts from the top six domestic airlines—United, Continental, American, Delta, Northwest, and U.S. Airways. All airlines except U.S. Airways directly provided us electronic data on unused tickets. Data provided by the five airlines and verified against Citibank’s data showed that over 2,700 airline tickets with a face value of about $6 million purchased with State’s centrally billed accounts were unused and not refunded. The airline tickets State purchased, for State and other foreign affairs personnel, through the centrally billed accounts are generally acquired under the terms of the air transportation services contract that GSA negotiates with U.S. airlines. Airline tickets purchased under this contract have no advance purchase requirements, have no minimum or maximum stay requirements, are fully refundable, and do not incur penalties for changes or cancellations. Under this contract, federal agencies are entitled to recover payments made to airlines for tickets that agencies acquired but did not use. While generally there is a 6-year statute of limitation on the government’s ability to file an action for financial damages based on a contractual right, the government also has up to 10 years to offset future payments for amounts it is owed. During fiscal years 2003 and 2004, State did not implement controls to monitor State’s and other foreign affairs employees’ adherence to travel regulations requiring notification of TMC or the appropriate State officials about unused tickets. Federal and State travel regulations require a traveler who purchased a ticket using the centrally billed account either to return any unused tickets purchased to the travel management center that furnished the airline ticket or to turn in unused tickets immediately upon arrival at their post to the administrative officer or, upon arrival in Washington, D.C., to the executive officer of the appropriate managing bureau or office. This notification of an unused ticket initiates a process to submit requests to the airlines for refunds. Figure 4 illustrates where control breakdowns can occur if travelers do not adhere to State requirements. As shown, once a ticket is charged to the centrally billed account and given to the traveler, State has no systematic controls to determine independently if the ticket was used—or remains unused—unless notified by the traveler. If the traveler does not report an unused ticket, the ticket would not be refunded unless TMC monitored the status of airline tickets issued electronically and applied for the refunds. Figure 4 shows that the failure of the traveler to notify the appropriate official of an unused paper ticket would result in the ticket being unused and not refunded. Although bank data indicate that State received some credits for airline tickets purchased, State did not maintain data in such a manner as to allow it to identify the extent of unused tickets and to determine whether credits were received. State did not have a systematic process in place to monitor whether TMCs were consistently identifying and filing for refunds on unused tickets. For instance, State contractually required the domestic TMC to identify and process all unused electronic tickets. In exchange, the TMC received a fee for each refund received for an unused ticket. However, State did not implement procedures to determine whether unused tickets were being identified and credits were being received. Instead, State officials took the TMC’s monthly report indicating only the total dollar amount of refunds submitted to the airlines as evidence of contractual compliance. Unless State implements control procedures to verify whether TMCs were identifying and filing for refunds on the unused tickets consistently, State cannot provide reasonable assurance that all requests for refunds resulted in a credit to the government. Even when a TMC had procedures in place to identify and process unused electronic tickets, State was still unable to identify unused paper tickets. For example, by fiscal years 2003 and 2004, State’s domestic TMC and TMCs at both of the overseas locations we visited had the capability to identify or search the databases of the airlines that participate in electronic ticketing or to receive notification from the airlines of unused tickets, and subsequently obtain refunds. However, even though the TMCs can identify electronic tickets, they cannot independently identify paper tickets, which are typically used for international travel. State has not implemented a systematic process to verify whether a significant portion of airline tickets are unused, such as matching tickets issued by TMC with travel vouchers submitted by travelers upon completion of their trip. Without such a process State will not have reasonable assurance that tickets purchased through the centrally billed accounts are used or refunded. In addition to the $6 million dollars of unused tickets or trip segments we identified using the airline data, we estimated that, based on the statistical sample, 3 percent of premium-class airline tickets were unused and not refunded. This 3 percent estimate is for premium-class tickets only and excludes coach accommodations. Table 2 contains specific examples of tickets that the airlines identified as unused that we tested as a part of our statistical sample of premium class transactions and data mining selections. Since these tickets were not used, they resulted in waste and increased costs to taxpayers. State did not dispute over 320 unauthorized transactions, totaling over $420,000, associated with its two primary domestic centrally billed accounts during fiscal year 2003 and fiscal year 2004. TMCs reconcile transactions on the monthly credit card invoice to the tickets issued by the TMC and recorded in the airline reservation system. Disputes are typically filed for transactions that neither the TMC nor State identified as having issued or authorized. Tickets that do not match could occur for many reasons, such as an airline charging the ticket to the wrong credit card account, an individual fraudulently obtaining an airline ticket, or the merchant or credit card vendor failing to provide enough information to allow the transaction to match. State did not have processes or procedures in place to file disputes for transactions that failed to reconcile between the bank invoice and the computer reservation system. We provided State a list of 219 travelers’ names associated with the over 320 unauthorized transactions to verify that they were State employees or otherwise authorized by State or other foreign affairs agencies to travel. According to State, 38 of the 219 travelers were individuals for whom State had no record of ever working for State as an employee, contractor, or being authorized to travel as an invited guest. Thus, these transactions could be potentially fraudulent charges. As for the remaining 181 travelers, State informed us that while the airline tickets purchased were for individuals who are either current or former State employees, contractors, or invited guests, State has no evidence that the trips had been authorized. Thus, these trips also could represent potentially fraudulent charges. As a result of not disputing unauthorized charges and not paying its bill in accordance with the contract, State faced the unanticipated consequence of substantially reducing the amount of rebates that it would have been eligible to receive. For example, if State had effectively managed the domestic accounts and disputed these charges, State could have earned over $1 million in rebates. Instead, State earned only about $174,000 in performance rebates for its domestic accounts. In contrast, at two overseas posts that we visited, State was properly disputing transactions. However, as previously noted, State still did not effectively manage its centrally billed accounts departmentwide and, consequently, earned only $700,000 out of a possible $2.8 million in performance rebates from Citibank. The contract that State entered into with Citibank to issue centrally billed account travel cards enables State to earn performance rebates based on how quickly State pays the monthly bill. To earn the performance rebate, State must pay the bill within 30 calendar days from the statement date. State earns the maximum performance rebate if it pays the centrally billed account—less any disputed charges—on the statement date; for unpaid bills, the amount of the rebate decreases each day thereafter. If State pays the centrally billed account more than 30 days after the statement date, State does not earn a performance rebate. Throughout the audit period, State generally submitted payment for its domestic centrally billed accounts within the 30 day window; however, State frequently failed to pay the entire amount of the bill, leaving potentially unauthorized charges unpaid, but not properly disputed. During fiscal years 2003 and 2004, State did not dispute any of the previously mentioned over 320 unauthorized charges applied to its domestic centrally billed accounts, and instead simply deducted the amounts due from its credit card bill. If State had disputed these charges, Citibank would have given State a 60-day grace period to investigate whether the charges were appropriate and the disputed amounts would not have to be paid until the investigation was completed. An average person cannot simply determine which charges on their credit card bill they are going to pay but must notify the bank of any unauthorized charges. Since State did not dispute the charges, it was still liable for the amounts associated with these charges and simply deducting them from the credit card bill did not relieve State of its responsibility for these charges. Consequently, State was not only paying for potentially fraudulent charges, but it also lost the performance rebates it could have earned by promptly paying its monthly centrally billed account bill. The State department serves a critical role for the federal government and in that role State and other foreign affairs employees are required to travel extensively, often internationally. However, travel regulations state that employees on official government travel must follow published requirements and exercise the same standard of care in incurring expenses that a prudent person would exercise when traveling on personal business. Our work shows that travelers using State’s over 260 centrally billed travel accounts often do not meet that standard, which has resulted in millions of dollars of unnecessary costs to taxpayers. With the serious fiscal challenges facing the federal government, agencies need to do everything they can to operate as efficiently as possible. Improved management and oversight of the State department’s centrally billed travel program would save taxpayers tens of millions of dollars annually. We are making the following 18 recommendations to improve internal control over the authorization and justification of premium-class travel and to strengthen the control environment as part of an overall effort to reduce improper premium-class travel and unnecessary or inappropriate State costs. Because of the substantial cost and sensitive nature of premium- class travel, we recommend that the Secretary of State direct the appropriate officials to implement specific internal control activities over the use of premium travel and establish policies and procedures to incorporate federal and State regulations as well as guidance specified in our Standards for Internal Control and our Guide for Evaluating and Testing Controls Over Sensitive Payments. While a wide range of activities can contribute to a system that provides reasonable assurance that premium-class travel is authorized and justified, at a minimum, the internal control activities should include the following: Develop procedures to identify the extent of premium-class travel, including all business-class travel, and monitor for trends and potential misuse. Develop procedures to identify all first-class fares so that State can prepare and submit complete and accurate first-class travel reports to GSA. Require State to develop a management plan requiring that audits of State’s issuance of premium-class travel are conducted regularly, and the results of these audits are reported to senior management. Audits of premium-class travel should include reviews of whether travel management centers adhere to all governmentwide and State regulations for issuing premium-class travel. Periodically provide notices to travelers and supervisors/managers that specifically identify the limitations on premium-class travel, the limited situations in which premium-class travel may be authorized, and how the additional cost of premium-class travel can be avoided. Require that premium-class travel be approved by individuals who are at least of the same grade as the travelers and specifically prohibit the travelers themselves or their subordinates from approving requests for premium-class travel. Prohibit the use of blanket authorization for premium-class travel, including management decisions offering premium-class travel as a benefit to executives and other employees. Encourage State department personnel traveling as a result of a permanent change of station to take a rest stop en route to their final destination to avoid the significant additional cost associated with premium-class flights and thus save the taxpayer thousands of dollars per trip. Urge other users of State’s centrally billed travel accounts to take parallel steps to comply with existing travel requirements. To promote the economy and efficiency of Courier Service operations, we recommend that the Secretary of State direct the Courier Service to take the following actions: Expand the use of cargo carriers, such as FedEx, to the extent practicable. Direct foreign missions to assure that organizations using diplomatic courier services share responsibility for meeting and accepting air cargo shipments of diplomatic pouches. Clarify written policy to clearly state that diplomatic couriers must use economy class accommodations when in a “dead-head” capacity unless relevant exceptions (e.g., 14-hour rule) exist, and enforce the requirement. To recover outstanding claims on unused tickets, we recommend that the Secretary of State initiate the following actions: Immediately submit claims to the airlines to recover the $6 million in fully and partially unused tickets identified by the airlines and discussed in this report. Work with the five airlines identified in this report and other airlines from which State purchased tickets with centrally billed accounts to determine the feasibility of recovering other fully and partially unused tickets, the value of the unused portions of those tickets, and initiate actions to obtain refunds. To enable State to systematically identify future unused airline tickets purchased through the centrally billed accounts, and improve internal controls over the processing of unused airline tickets for refunds, we recommend that the Secretary of State direct the appropriate personnel within services and agencies to take the following actions: Evaluate the feasibility of implementing procedures to reconcile airline tickets acquired using the centrally billed accounts to travel vouchers in the current travel system. Enforce employees’ adherence to existing travel regulations requiring notification of unused tickets. Modify existing travel management center contracts to include a requirement that the international travel management centers establish a capability to systematically identify unused electronic tickets in their computer reservation systems and file for refunds on the tickets identified as unused. Routinely compare unused tickets processed by the travel management centers to the credits on the Citibank invoice. To provide assurance of accurate and timely payments of the centrally billed accounts and to maximize rebates, we recommend that the Secretary of State establish procedures to ensure that all transactions on the Citibank invoice are either paid in accordance with the contract or properly disputed. In written comments on a draft of this report, State concurred with all 18 of our recommendations and said that it is firmly committed to aggressive stewardship of the taxpayers’ resources entrusted to the department. However, State also commented that our report overstates the problem, fails to identify improper travel conducted for other than official government travel, identifies only a few instances of unjustified travel, and implies incorrectly that State carelessly implemented business-class regulations without regard to the increased cost. We disagree. We do not agree with State’s position that we overstate the nature and extent of its control breakdowns and ineffective oversight. State and other foreign affairs travelers charged almost $140 million on premium-class travel from April 2003 through September 2004. On the basis of our statistical sample, 67 percent of premium-class travel was not properly authorized, justified, or both. This failure rate and the associated dollars spent on premium class travel shows that taxpayers lost tens of millions of dollars on improper travel. For example, State issued premium-class tickets to a family of four traveling from Washington to Moscow for a permanent change of duty station. Although this trip was well under the required 14 hours to justify premium-class travel, State purchased the premium class accommodations for almost four times the cost of coach seats. In addition to the waste exemplified here and elsewhere in our report, taxpayers lost millions more because State failed to recover payments made to airlines for tickets issued but never used and failed to reconcile and dispute other charges properly. For example, State paid for a premium-class ticket for roundtrip travel between New Mexico and Ethiopia that was neither used nor refunded. These specific examples and our overall analysis clearly show how ineffective oversight—not just procedural problems—resulted in substantial waste of taxpayers’ dollars. As our report clearly explains, we did not specifically question whether travel charged to State’s centrally billed travel accounts were necessary. Therefore, we purposely did not identify improper travel conducted for other than official government travel and thus our report makes no conclusions on this matter. State’s position that our findings of improper travel are simply the result of “procedural problems” and that “only a few instances” of travel were conducted outside of the regulations are inconsistent with the facts. In this regard, over half of the transactions we tested—not just a few instances— were not simply the result of procedural problems (e.g., not properly authorized), they were unjustified because the travel was conducted outside of the regulations. Over half of the travelers improperly flew premium-class on trips lasting shorter than 14 hours or flew business class and also took a rest stop, which is to be used in lieu of using premium-class accommodations to economize travel. For example, one State traveler flew premium-class between points in Europe on a trip lasting well short of 14 hours and also took an unjustified rest stop, which further added lodging and subsistence expenses to the total cost of travel. Another traveler flying short of 14 hours on a premium-class ticket enjoyed 3 nights of rest upon her return. These and other examples of unjustified travel underscore problems beyond what State says are simply “deficient procedural protocols” and demonstrate how State’s ineffective oversight of premium- class travel resulted in substantial losses to taxpayers. Finally, State takes exception with our characterization that it treated premium-class travel as an employee benefit. This position, however, is in stark contrast to the representations State made throughout our review. For example, although State prohibits blanket authorizations for premium- class travel, many of State’s top executives consistently flew on blanket travel orders improperly authorizing premium class from Washington to numerous domestic and other destinations that were well below the 14 hours required to justify such travel. For example, one senior State executive completed 45 premium-class trips costing $213,000, many of which were under 14 hours, using a blanket travel order. These executive travelers set a tone at the top that premium-class travel was in fact a benefit to the traveler and not something that should be minimized or used sparingly. In addition, during our review, State said that it indeed offered premium-class travel as a benefit to its employees and that such travel contributed to their improved employee feedback provided to “The Best Places to Work” survey. However, State could not provide evidence that travel was a metric in that survey. Moreover, regardless of the increased cost associated with such moves, State began in 2002 and continues today to offer premium-class travel for permanent change of station moves as a benefit to its employees and their families. We believe these examples, especially the top State executives who gave themselves the benefit of flying premium class when federal law and regulations did not allow such travel, demonstrate that the tone at the top of the department indicates that premium-class travel is in fact a benefit, without specific regard to cost. State’s comments are reprinted in appendix II. As agreed with your offices, unless you announce the contents of this report earlier, we will not distribute it until 30 days from its date. At that time, we will send copies to interested congressional committees; the Secretary of State, the Director and Deputy Director of the Diplomatic Courier Service, and the Director of the Office of Management and Budget. We will make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. Please contact me at (202) 512-7455 or kutzg@gao.gov if you or your staffs have any questions concerning this report. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors are listed in appendix III. This report responds to your request that we audit and investigate internal controls over State’s centrally billed travel accounts, which include travel related to the Department of State, other U.S. government agencies principally engaged in activities abroad, and other domestic departments and agencies with international operations. The objectives of our audit were to determine the effectiveness of the Department of State’s internal controls over its centrally billed travel card program and determine whether fraudulent, improper, and abusive travel expenses exist. Specifically we evaluated the effectiveness of State’s internal controls over (1) the authorization and justification of premium-class tickets charged to State’s centrally billed travel accounts and (2) monitoring unused tickets, reconciling monthly statements, and maximizing performance rebates. To assess the effectiveness of internal controls over State’s use of the centrally billed accounts, we obtained an understanding of the travel process, including premium-class travel authorization, unused ticket identification, and overall travel card management and oversight, by interviewing State officials from Resource Management, Travel and Transportation Management Division; Diplomatic Security, Overseas Building Operations; Educational and Cultural Affairs; U.S. Consulate, Frankfurt, Germany and U.S. Embassy, Pretoria. We also interviewed key officials from the American Express, Carlson Wagonlit, and Concorde travel management centers. We reviewed General Services Administration’s (GSA) Federal Travel Regulations (FTR) and State’s Foreign Affairs Manual (FAM) and Foreign Affairs Handbook (FAH). We reviewed State’s internal department notices and other travel-related guidance. Finally, we conducted “walk-throughs” of the domestic and overseas travel processes. We audited controls over the authorization and issuance of premium-class travel during fiscal years 2003 and 2004. State’s credit card vendor, Citibank, could not provide the first 6 months of fiscal year 2003 (October 2002–March 2003) level III data due to limitations in its archiving capabilities. The level III data indicate whether a transaction is premium or coach. Therefore, we used 18 months of data from April 2003 through September 2004 to select a probability sample of premium-class transactions and also used this same time period for our data mining and analysis of premium-class transactions. Our assessment covered the following: The extent to which State used the centrally billed accounts to obtain premium-class travel was determined. Testing a statistical sample of premium-class transactions to assess the implementation of key management controls and processes for authorizing and issuing premium-class travel, including approval by an authorized official and justification in accordance with regulations. We also used data mining to identify other selected transactions throughout the premium-class travel transactions to determine if indications of improper transactions existed. State’s management policy towards the use of premium-class travel was determined. To assess the magnitude of premium-class travel by State and other foreign affairs agencies, we obtained from Citibank a database of fiscal years 2003, 2004, and 2005 travel transactions charged to State’s centrally billed and individually billed travel card accounts. The databases contained transaction-specific information, including ticket fares, codes used to price the tickets—fare basis codes—ticket numbers, names of passengers, and numbers of segments in each ticket. We reconciled these data files to control totals provided by Citibank and to data reported by GSA on State’s centrally billed account activities. We queried the database of positive debit transactions (charges) for fare codes that corresponded to the issuance of first- and business-class travel, identifying all airline transactions that contained at least one leg in which State and other foreign affairs agencies paid for premium-class travel accommodations. We further limited the first- and business-class transactions to those costing more than $750 because many premium-class tickets on intra- European flights cost less than $750 and the corresponding coach-class tickets were not appreciably less. By eliminating from our population first- and business-class transactions costing less than $750, we avoided the possibility of identifying a large number of transactions in which the difference in cost was not significant enough to raise concerns of the effectiveness of the internal controls. The total number of transactions excluded was 1,067, costing approximately $532,000. While we excluded premium-class transactions costing less than $750, we (1) did not exclude all intra-European flights and (2) potentially excluded unauthorized premium-class flights. Limitations of the database prevented a more precise methodology of excluding lower-cost first- and business-class tickets. Table 3 summarizes the population of State and other foreign affairs agencies’ airline travel transactions containing at least one premium-class leg charged to State’s centrally billed accounts from April 2003 through September 2004 and the subpopulation subjected to testing. To assess the implementation of key controls over the authorization and issuance of premium-class travel, we tested a probability sample of premium-class transactions. In general, the population from which we selected our transactions for testing was the set of positive debit transactions totaling $750 or more for both first- and business-class travel that were charged to State’s centrally billed accounts during April 2003 through September 2004. Because our objective was to test controls over travel card expenses, we excluded credits and miscellaneous debits (such as fees) that would not have been for ticket purchases from the populations tested. We further limited the population of first- and business-class transactions to those without a matching credit. By eliminating transactions with matching credits, we avoided selecting a large number of transactions in which the potential additional cost of the premium-class ticket was mitigated by a credit refund so as not to raise concerns about the effectiveness of the internal controls. The total number of transactions excluded was 2,799, totaling approximately $11.7 million. While we excluded premium-class transactions with a matching credit, we did not exclude all transactions with a matching credit because sometimes the data did not always identify the fare basis codes to allow us to determine if the travel was premium or coach. To test the implementation of key control activities over the issuance of premium-class travel transactions, we selected a probability sample of transactions. Specifically, we selected 107 premium-class transactions totaling about $467,000. For each transaction sampled, we requested that State provide us the travel order, travel voucher, travel itinerary, and other related supporting documentation. We used that information to test whether documentation existed that demonstrated that State had adhered to key internal controls over authorizing and justifying premium-class tickets. On the basis of the information State provided, we determined whether a valid official approved the premium-class travel and whether the premium-class travel was justified in accordance with State regulations. We also applied criteria set forth in our internal control standards and sensitive payments guidelines in evaluating the proper authorization of premium- class travel. For example, while State travel regulations and policies do not address subordinates authorizing their supervisors’ premium-class travel, our internal control standards consider such a policy to be flawed; therefore, a premium-class transaction that was approved by a subordinate would fail the control test. The results of the samples of these control attributes can be projected to the population of transactions at State and other foreign affairs agencies as a whole, but not to individual bureaus or posts. With our probability sample, each transaction in the population had a nonzero probability of being included, and that probability could be computed for any transaction. Each sample element was subsequently weighted in the analysis to account statistically for all the transactions in the population, including those that were not selected. Because we followed a probability procedure based on random selections, our sample is only one of a large number of samples that we might have drawn. Since each sample could have provided different estimates, we express our confidence in the precision of our particular sample’s estimates as 95- percent confidence intervals (e.g., plus or minus 10 percentage points.) These are intervals that would contain the actual population value for 95- percent of the samples we could have drawn. As a result, we are 95-percent confident that each of the confidence intervals in this report will include the true values in the study population. All percentage estimates from the sample of premium-class air travel have sampling errors (confidence interval widths) of plus or minus 10 percentage points or less. Table 4 summarizes the premium-class statistical sample results. In addition to our statistical sample, we selected other transactions identified by our data mining efforts for review. Our data mining identified individuals who frequently flew using first- or business-class accommodations. For data mining transactions, we also requested that State provide us the travel order, travel voucher, travel itinerary, and any other supporting documentation that could provide evidence that the premium-class travel was properly authorized and justified in accordance with State policies. If the documentation provided indicated that the transactions were proper and valid, we did not pursue the matter further. However, if the documentation was not provided, or if it indicated further issues related to the transactions, we obtained and reviewed additional documentation about these transactions. Our initial data mining efforts identified executives that frequently flew first and business class. On the basis of our findings, we expanded our selection of high-level officials to include most of State’s top executives, including presidential appointees and senior executives. We evaluated these transactions in the same manner as described above. Based on the statistical sample of premium class transactions, we estimate that 6 percent of the transactions in the sample population represent travel by diplomatic couriers. We also identified courier transactions by data mining for travelers that frequently flew first and business class. We found six courier transactions in our statistical sample and an additional 16 transactions identified during data mining for proper authorization and justification. We reviewed pertinent laws, federal regulations, and State department policies and procedures and interviewed current and former Diplomatic Courier Service staff. We also conducted an on-site inspection of classified pouch procedures at the Logistics Operations Center and observed the FedEx process for inventory, pouching, and packaging of classified materials for shipment to London, Paris, and Frankfurt. We did not have authorization to open, inspect, and verify that classified pouches contained only classified materials. Also, we did not observe and assess courier procedures at foreign airports related to accessing the tarmac to take custody of outgoing and incoming diplomatic pouch materials. During the course of our work, we interviewed Department of State Inspector General, Diplomatic Courier Service, and Administrative Logistics Management officials and Department of Homeland Security officials responsible for customs and border protection. We also audited the controls over other centrally billed account activities, including the identification and processing of unused tickets and disputing of unauthorized transactions, during fiscal years 2003 and 2004. Our assessment covered the magnitude of centrally purchased tickets that were not used and not processed for a refund, and the extent of unauthorized transactions that were not disputed and of the rebates lost, as a result. To assess the internal controls over these other CBA activities, we first applied the fundamental concepts and standards set forth in our Standards for Internal Control in the Federal Government to the practices followed by these units to manage unused tickets and to dispute transactions that did not match or that the reconciliation process determined were unresolved. Because we determined that controls over unused tickets were ineffective, we did not assess these controls. To assess the magnitude of tickets charged to the centrally billed accounts, which were unused and not refunded, we requested that the six airlines that State and other foreign affairs agencies used most frequently provide us with data relating to tickets State and other foreign affairs agencies purchased during fiscal years 2003 and 2004 that were unused and not refunded. These six airlines—American, Delta, Northwest, Continental, United, and U.S. Airways—together accounted for about 80 percent of the value of total airline tickets State and other foreign affairs agencies purchased. To obtain assurance that the tickets the airlines reported as unused represented only airline tickets charged to State centrally billed accounts, we compared data provided by the airlines to transaction data provided by Citibank. Because State does not track whether tickets purchased with centrally billed accounts were used, we were unable to confirm that the population of unused tickets that the airlines provided was complete in that it included all State and other foreign affairs agencies’ tickets that were unused and not refunded. While American, Delta, Northwest, and United provided data that allowed us to identify the centrally purchased tickets that were fully unused and not refunded and partially used and not refunded, Continental could only provide data on fully unused and not refunded tickets and U.S. Airways did not provide any data. Because none of the airlines provided data sufficient for calculating the exact unused value (residual value), we were limited to reporting the amount charged to the centrally billed accounts related to both fully unused and partially unused tickets. To determine the extent of airline tickets that did not reconcile between the tickets issued by State’s travel management center and the Citibank invoice of tickets purchased on the centrally billed account, we (1) obtained unresolved transaction reports for State’s largest domestically managed centrally billed accounts and (2) verified that the transactions were charged to a State centrally billed account using the Citibank transaction data. To identify the potential rebates lost on State’s centrally billed accounts, we requested that Citibank provide (1) the total amount of rebates earned by State on its centrally billed account program for fiscal year 2003 and fiscal year 2004, (2) the volume of transactions used by Citibank to compute the rebate amounts, and (3) the rebate pricing schedule Citibank used to determine the amount of rebates. Using the volume of transactions and the rebate pricing schedule provided by Citibank, we calculated the highest potential rebate that State could have earned on the centrally billed account program. We then compared the potential rebate amounts to the actual rebates earned. We assessed the reliability of the Citibank centrally billed account data by (1) performing various testing of required data elements, (2) reviewing existing information about the data and system that produced them, and (3) interviewing Citibank officials knowledgeable about the data. In addition, we verified that totals from the databases agreed with the centrally billed account activity reported by GSA. We determined that data were sufficiently reliable for the purposes of our report. To assess the reliability of the unused ticket data provided to us by American, Continental, Delta, Northwest, and United Airlines, we (1) consulted airline officials knowledgeable about the data and (2) performed testing on specific data elements. In addition, we validated that the tickets reported as unused by each airline represented tickets centrally purchased by State by comparing each airline’s data to the Citibank centrally billed account. We also reviewed the 2003 and 2004 Notes to the Consolidated Financial Statements for each airline to verify that amounts related to unused tickets were included as a liability. We concluded that the data were sufficiently reliable for the purposes of this report. Key contributors to this report include Cindy Barnes, Felicia Brooks, Norman Burrell, Beverly Burke, Jennifer Costello, Francine DelVecchio, Abe Dymond, Aaron Holling, Jason Kelly, John V. Kelly, Andrea Levine, Barbara Lewis, Jenny Li, Katherine Peterson, Mark Ramage, John Ryan, Sidney H. Schwartz, and Michael C. Zola.

The relative size of the Department of State's (State) travel program and continuing concerns about fraud, waste, and abuse in government travel card programs led to this request to audit State's centrally billed travel accounts. GAO was asked to evaluate the effectiveness of internal controls over (1) the authorization and justification of premium-class tickets charged to the centrally billed account and (2) monitoring of unused tickets, reconciling monthly statements, and maximizing performance rebates. Breakdowns in key internal controls, a weak control environment, and ineffective oversight of State's centrally billed travel accounts resulted in taxpayers paying tens of millions of dollars for unauthorized and improper premium-class travel and unused airline tickets. State's over 260 centrally billed accounts are used by State and other foreign affairs agencies to purchase transportation services, such as airline and train tickets. GAO found that between April 2003 and September 2004 State's centrally billed accounts were used to purchase over 32,000 premium-class tickets costing almost $140 million. Premium-class travel--primarily business-class airline tickets--represented about 19 percent of the tickets issued but about 49 percent of the $286 million spent on airline tickets with State's centrally billed account travel cards. GAO determined that this trend continued for fiscal year 2005. GAO found that 67 percent of this premium-class travel was not properly authorized, justified, or both. Because premium-class tickets typically cost substantially more than coach tickets, improper premium-class travel represents a waste of tax dollars. Most of these blanket premium-class travel authorizations were signed by subordinates who told us they couldn't challenge the use of premium-class travel by senior executives. Ineffective oversight and control breakdowns also contributed to problems with monitoring unused tickets, reconciling monthly statements, and maximizing performance rebates. Although federal agencies are authorized to recover payments made to airlines for tickets that they ordered but did not use, State failed to do so and paid for about $6 million for airline tickets that were not used or processed for refund. State was unaware of this problem before our review because it neither monitored travelers' adherence to travel regulations nor systematically identified and processed all unused tickets. State also failed to reconcile or dispute over $420,000 of unauthorized charges before paying its monthly bank invoice and instead deducted the amounts from its bill. Because these amounts were not properly disputed under the contract terms, State underpaid its monthly bills and was thus frequently delinquent. Handling questionable charges in this ad hoc manner sharply reduced State's eligible rebates. Overall, State earned only $700,000 out of a possible $2.8 million in rebates that could have been earned if it had properly disputed unauthorized charges and paid the bill in accordance with the contract.

FDA’s authority to regulate tobacco products under a public health standard is unique among its regulatory responsibilities. CTP is the FDA center with primary responsibility for executing this regulatory responsibility, and its offices conduct work in several areas, including reviewing submissions for new tobacco products to determine if such products can be legally marketed in the United States, and responding to meeting requests from manufacturers and other entities. All of CTP’s activities are funded through tobacco manufacturer user fees, as required by the Tobacco Control Act. FDA—primarily through CTP—undertakes four broad categories of activities in carrying out its responsibilities and authorities under the (1) reviewing submissions for marketing new Tobacco Control Act:tobacco products and setting scientific standards for tobacco products; (2) enforcing statutory and regulatory requirements prohibiting the sale, marketing, and distribution of certain tobacco products; (3) developing and issuing regulations and guidance, conducting compliance checks, and removing violative products from the market pursuant to the Tobacco Control Act; and (4) engaging in public education and outreach activities about the risks associated with tobacco product use, and promoting awareness of and compliance with the Tobacco Control Act. CTP is organized into seven offices. (See table 1.) Within CTP, OS is the office primarily responsible for conducting reviews of new tobacco product submissions; however, OS staff duties are not limited to reviewing new tobacco product submissions. Under the Tobacco Control Act, a manufacturer may make a submission to FDA for CTP’s determination of whether the manufacturer may introduce a new tobacco product to the market in the United States. CTP reviews submissions made by manufacturers through one of three pathways: Substantial Equivalence (SE) pathway: Manufacturers make a submission under the SE pathway if either (1) a new tobacco product has the same characteristics as a predicate tobacco product—that is, a product commercially marketed in the United States on February 15, 2007, or a product previously found by CTP to be substantially equivalent; or (2) the new tobacco product has different characteristics from a predicate tobacco product, but does not raise different questions of public health. There are two types of submissions made under the SE pathway—provisional and regular— that are defined by the date that the product came on the market and when the manufacturer made the submission. For provisional SE submissions, a manufacturer may market the new product that is the subject of the submission while CTP conducts its review of the submission, but for regular SE submissions, a manufacturer may not market the new product until CTP completes its review and determines that the product meets the SE requirements. (See table 2.) Exemption from SE pathway: Manufacturers make a submission under the Exemption from SE pathway if (1) the new product is a minor modification (adding, deleting, or changing the quantity of an additive) of another tobacco product marketed by the same manufacturer; (2) an SE submission is not necessary to ensure that permitting the tobacco product to be marketed would be appropriate for the protection of public health; and (3) an Exemption from SE is otherwise appropriate. Premarket Tobacco Product Application (PMTA) pathway: Manufacturers make a submission under the PMTA pathway if the new tobacco product does not meet the criteria of the SE or Exemption from SE pathways—that is, the new tobacco product is not substantially equivalent to a predicate product or is not a minor modification of an appropriate product for modification. The PMTA submission must include, among other things, full reports of investigations of health risks, and must meet the public health standard described under the Tobacco Control Act (that is, would be appropriate for the protection of public health). The Tobacco Control Act does not mandate a time frame for CTP’s review of new tobacco product submissions with the exception of PMTA submissions. For PMTA submissions, the act requires CTP to issue an order stating whether the product may be marketed as promptly as possible, but not later than 180 days after FDA’s receipt of a submission. CTP reviews of SE submissions—primarily conducted by OS—include three key steps: (1) jurisdiction review to determine if the product is regulated by FDA, (2) completeness review to determine if the submission is missing information, and (3) scientific review to determine if the product is substantially equivalent or not (see fig.1). According to CTP officials, project managers determine whether the product (including any component, part, or accessory of the product) is made or derived from tobacco; whether it is a drug or medical device; and whether it meets established definitions for any type of FDA-regulated tobacco product. respond to administrative AI letters, but in April 2012, CTP began giving manufacturers 30 days to respond to an administrative AI letter. After OS finishes these initial two steps in the SE review process, the next step is a scientific review, which involves an assessment of the product by scientists in different disciplines (such as chemistry and toxicology). These scientists work to determine whether the product is substantially equivalent to a product already on the market—that is, has the same characteristics as a predicate tobacco product, or has different characteristics but does not raise different questions of public health. During scientific review, OS may issue scientific AI letters to request additional information that the scientists determine is needed to make a final determination (such as clarification of ingredients and additional testing results). In these letters, CTP officials told us that OS requests that manufacturers respond within 60 days. If OS determines that the SE criteria have been met, then CTP will issue an SE order, and the product may continue being marketed by the manufacturer (if it was a provisional SE submission) or may be legally introduced into the U.S. market (if it was a regular SE submission). If neither of these criteria is met, then CTP will issue an order that the product is not substantially equivalent and the manufacturer must remove the product from the market (if it was a provisional SE submission) or cannot introduce the product into the market under the SE pathway (if it was a regular SE submission). According to CTP officials, reviews of Exemption from SE and PMTA submissions also include jurisdiction, completeness, and scientific review steps. However, the specific activities within each review step for those pathways may differ from the specific activities involved in review steps for SE submissions. The Tobacco Control Act does not require CTP to conduct meetings with outside entities, but CTP officials reported that they are valuable because they increase knowledge of tobacco regulation among public health groups, promote compliance among manufacturers, and clarify information needed for new tobacco product submissions. However, each CTP office follows different processes for receiving and processing meeting requests. In the event that an outside entity—for instance, a manufacturer or a public health advocacy organization—wants to meet with CTP officials, it can request a meeting in various ways. For example, manufacturers can submit written requests to the Director of OS by mail, courier, or electronically to FDA’s document center. Manufacturers have requested meetings with OS to discuss their new tobacco product submissions, as well as study protocols and other scientific issues. Manufacturers, tobacco trade associations, and other entities have also proposed meetings with OS, OCD, OCE, and OP to educate CTP on tobacco industry operations (for example, current practices in tobacco product manufacturing), and to discuss industry’s views on FDA’s approaches to tobacco regulation (for example, industry feedback on published guidance documents). State, local, and tribal governments, as well as academic and scientific organizations, have requested meetings in order to coordinate public health efforts or share relevant knowledge. CTP officials told us that CTP follows FDA’s practice not to grant meetings for which the topic of discussion is in draft guidance. Additionally, according to officials, one office within CTP may transfer a meeting request to another office within CTP in order to provide the most knowledgeable and appropriate agency officials at the meeting. However, a request may not result in a scheduled meeting. The Tobacco Control Act requires FDA to assess user fees on manufacturers of FDA-regulated tobacco products based on their market share and specifies that the tobacco user fees can only be applied toward FDA activities that relate to the regulation of tobacco products. FDA bills and collects tobacco user fees from manufacturers on a quarterly basis and fees are generally collected the quarter after they are billed. For example, fees billed in the fourth quarter of fiscal year 2011 were collected in the first quarter of fiscal year 2012. The Tobacco Control Act specified the total amount of user fees authorized to be collected for each fiscal year beginning with fiscal year 2009, and authorized user fees to remain available until expended (which means that FDA may carry over user fees to subsequent fiscal years if they are not obligated by the end of the fiscal year in which they were collected). (See table 3.) As of January 7, 2013, the vast majority of new tobacco product submissions FDA received from manufacturers were made under the SE pathway. CTP has finished initial review steps (jurisdiction and completeness reviews) for most SE submissions, but CTP has not made final decisions for most submissions. For the majority of provisional SE submissions, CTP took over a year and a half to complete these initial review steps. In late June 2013, CTP made a final decision on 6 of the 3,788 SE submissions, finding that 2 of the products were substantially equivalent and that 4 were not; the remaining submissions were still undergoing CTP review. Several factors contributed to the significant amount of time it took for review of new tobacco product submissions, according to officials from CTP and tobacco manufacturers. CTP officials reported taking steps to address factors that contributed to the length of time the center has taken to review submissions, but the center has not established review time frames by which to assess progress. As of January 7, 2013, nearly all new tobacco product submissions FDA received from manufacturers (99 percent) were SE submissions, most of which were provisional SE submissions. FDA received a total of 3,788 SE submissions and 23 Exemption from SE submissions from manufacturers. FDA did not receive any PMTA submissions. (See fig. 2.) As shown in figure 2, of the 3,788 SE submissions received by FDA as of January 7, 2013, 3,165 (84 percent) were provisional SE submissions Almost all of the and 623 (16 percent) were regular SE submissions.provisional SE submissions were received in the second quarter of fiscal year 2011—3,115 of the provisional SE submissions were received within the 3 weeks prior to the statutory deadline of March 22, 2011. The number of regular SE submissions received in a quarter ranged from 19 (in the third quarter of fiscal year 2011) to 192 (in the third quarter of fiscal year 2012). (See fig. 3.) In addition to the 3,788 SE submissions, FDA received 23 Exemption from SE submissions from manufacturers as of January 7, 2013. Eligibility for the Exemption from SE pathway is limited to new tobacco products that are minor modifications of an existing tobacco product (adding, deleting, or changing the quantity of an additive) already marketed by the same manufacturer. According to CTP officials, a key factor contributing to the relatively small number of submissions is that it is not common for a manufacturer to change only additives when making a change to an existing tobacco product. According to industry representatives, a key reason for the relatively small number of submissions under this pathway is insufficient guidance from CTP about what exactly constitutes a minor modification of another commercially marketed tobacco product. FDA did not include a definition of the term “minor modification” in its final rule to establish procedures for the Exemption from SE pathway because the agency did not have the In the rule, FDA stated experience needed to provide a useful definition.that as it gains experience in evaluating Exemption from SE submissions, it will consider establishing a definition for minor modifications. CTP officials also reported that no submissions were received by FDA from January 8, 2013, through June 25, 2013. that current tobacco users will stop using tobacco products. According to industry representatives, meeting the standards under the PMTA pathway may not be feasible for some manufacturers—in particular, for small manufacturers (which are manufacturers that have fewer than 350 employees). Industry representatives reported that small manufacturers do not have the research and development resources to design or initiate clinical trials that would be needed to support a PMTA submission. As of January 7, 2013, CTP finished jurisdiction and completeness reviews for over two thirds of the provisional and regular SE submissions received since June 2010, but had not made a final decision on any of the 3,788 SE submissions. CTP finished both jurisdiction and completeness reviews for about 69 percent of provisional SE submissions (2,191 out of 3,165), and about 67 percent of regular SE submissions (415 out of 623). Almost all of the remaining 974 provisional SE submissions and about half of the remaining 208 regular SE submissions were through jurisdiction review but not completeness review. (See fig. 4.) Provisional SE submissions and regular SE submissions were pending in completeness review for as long about 1.5 years and 1 year, respectively.review for any of the SE submissions. As of January 7, 2013, CTP had not finished scientific According to CTP officials, prioritization of provisional SE On June 25, 2013—about 3 years after FDA’s receipt of the first SE submission—CTP made a final decision on 6 of the 3,788 SE submissions. CTP concluded that the new tobacco products in two of the submissions were substantially equivalent and that the products in the four other submissions were not. These six submissions were regular SE submissions received by FDA in fall 2011 (about 1 year and 8 months prior to CTP’s final decisions). For each of the two substantially equivalent products, CTP found that the new product had different characteristics than the predicate tobacco product but did not raise different questions of public health. CTP found that four new tobacco products were not substantially equivalent to predicate tobacco products due to factors such as inadequate evidence that the products to which the new products were being compared were valid predicate products and lack of complete information on tobacco product characteristics. CTP took over a year and a half from FDA’s receipt of a submission through the end of initial review steps for more than half of provisional SE submissions, and 6 months for more than half of the regular SE submissions. As of January 7, 2013, the median length of time to finish initial review steps—from FDA’s receipt of a submission through the end of completeness review—for provisional SE submissions was about 1 year and 9 months, and the length of time ranged from about 9 months to about 2.5 years (see fig.5). The median length of time to finish initial review steps for regular SE submissions was about 6 months, ranging from about 1 month to about 2 years (see fig. 6). Several factors have contributed to the significant amount of time it took for review of SE submissions, according to CTP officials and industry representatives. These officials identified factors such as insufficient information provided by manufacturers in submissions; the prioritization of regular SE submission reviews over provisional SE submissions; and other factors. CTP officials told us that insufficient information from manufacturers in SE submissions has had the most significant impact on review times for those submissions. According to CTP officials, the majority of SE submissions were incomplete and required follow-up with manufacturers to obtain additional information, such as a full description of both the new tobacco product and the predicate tobacco product. CTP officials reported that they spent significant time sending out AI letters requesting missing information from manufacturers and awaiting the manufacturers’ responses. Our analysis found that administrative AI letters were associated with 2,559 SE submissions, and CTP officials told us that some submissions had more than one administrative AI letter. In these letters, CTP officials requested that manufacturers respond to requests within 60 days or 30 days. In addition, our analysis found that scientific AI letters were associated with 81 SE submissions. In these letters, CTP requested that manufacturers respond to requests within 60 days, but CTP officials reported that it had granted extensions of up to 4 months. Industry representatives agreed that the lack of completeness of submissions had an impact on reviews, but they told us that guidance provided by CTP was neither timely nor adequate for manufacturers to provide what CTP would consider SE submissions with sufficient information. Manufacturers we interviewed said they were not able to include all information indicated in CTP guidance that was issued on January 5, 2011, for provisional SE submissions, which needed to be submitted by March 22, 2011, in order for those products to remain on the Some industry representatives indicated that the market provisionally.time it took to prepare a submission was more than CTP estimated, and that the deadline for provisional SE submissions was not enough time to incorporate all of the requirements in the guidance in their submissions. Additionally, industry representatives we interviewed reported that the January 2011 guidance did not direct manufacturers to include some information by the March 22, 2011, submission deadline that CTP later requested in its September 2011 draft guidance or AI letters, such as an environmental assessment. CTP placed a higher priority on its review of regular SE submissions than on its review of provisional SE submissions, which contributed to longer review times for provisional SE submissions when compared to regular SE submissions. Specifically, according to OS officials, in the summer of 2011 CTP prioritized completeness reviews for regular SE submissions over provisional SE submissions, so resources were shifted away from provisional SE submissions. As a result of this decision—coupled with the fact that provisional SE submissions were received earlier than regular SE submissions—completeness review times for provisional SE submissions were longer than for regular SE submissions. CTP officials said that there were three reasons for placing a higher priority on its review of regular SE submissions over provisional SE submissions: (1) tobacco products in provisional SE submissions could remain on the market legally (unless and until CTP issued an order of not substantially equivalent), (2) FDA received a large number of provisional SE submissions on March 21, 2011 (the day before the statutory deadline for submitting provisional SE submissions), making it impractical to prioritize reviews by the date the submission was received, and (3) CTP required time to assess which approach to reviewing provisional submissions would be the most effective at addressing the public health burden of tobacco use. Two more factors that had a significant impact on review times were a shortage of experienced tobacco product review staff and slow IT systems, according to CTP officials. These officials reported that when they started reviews of SE submissions the center had a shortage of experienced staff and that finding qualified staff was challenging. Additionally, CTP officials said that initial training of review staff contributed to review times as new staff were unable to review submissions until receiving the necessary training. CTP officials also told us that a slow IT system impacted the rate at which project managers could enter data during jurisdiction and completeness reviews of SE submissions, which slowed down those review times. CTP has taken action to address the factors CTP officials identified as contributing to the significant amount of time the center has taken to review submissions. CTP has provided additional direction to manufacturers in an attempt to decrease delays due to agency requests for more information through AI letters. Specifically, it has held webinars and published frequently asked questions to provide more guidance to manufacturers that prepare submissions. Additionally, CTP officials told us that in November 2012 CTP began alerting manufacturers of upcoming scientific review of their submissions by issuing a notification to manufacturers 45 days prior to starting scientific review. According to CTP officials, this notification reminds manufacturers of the option to amend their submissions as needed prior to the start of scientific review, to facilitate higher quality submissions, and potentially avoid delays in scientific review due to the issuance of scientific AI letters. CTP also noted that it is working on a standardized form for manufacturers to use when submitting new tobacco product information for review. According to CTP officials, this form may take time to develop as it will require FDA to issue regulations, but CTP officials anticipate that, when implemented, a standardized form should improve review times. To address the shortage of staff available for reviews, CTP officials told us they have increased OS staff from 12 staff in June 2010 to more than 100 staff in January 2013, including scientists and project managers involved in submission reviews. Also in 2012, CTP drafted a reviewers’ guide to help train staff on aspects of the SE review process. According to CTP officials, the center plans to continue to revise its draft reviewer’s guide as it further refines its new tobacco product review process. CTP officials also reported that CTP had upgraded its IT system as of early 2013, which has improved the time taken for data entry on SE submissions. They also reported that CTP plans to transition to a new IT system in late 2013. Our analysis of data provided by CTP found that for regular SE submissions the length of time from the end of jurisdiction review through the end of completeness review improved over time. Among regular SE submissions received by FDA in fiscal year 2011 and for which CTP had finished completeness review as of January 7, 2013, the length of time from the end of jurisdiction review to the end of completeness review ranged from about 3 months to 1.5 years, with a median length of time of about 8 months. In contrast, the length of time for these steps for regular SE submissions received in fiscal year 2012 ranged from less than 1 day to 11 months, with a median of about 2 months. CTP officials reported that actions such as hiring review staff and providing training for review staff have resulted in improved review times. While CTP is moving forward with its reviews of SE submissions and efforts to improve review times, CTP does not have time frames for reaching a final decision on submissions. Time frames would allow CTP to evaluate its efficiency and effectiveness and help it make appropriate adjustments. Under federal standards for internal control, control activities that establish performance measures, such as time frames, and the monitoring of actual performance against measures are an integral part of operating efficiently, achieving effective results, and planning appropriately. There are no time frames set by statute for the SE pathway, and CTP has not established performance measures that include time frames for making final decisions on the review of SE submissions. Although CTP officials agreed that establishing time frames would be useful for performance evaluation, CTP has not identified specific plans to establish such time frames. According to CTP officials, they have not yet established time frames because they first need to collect and analyze information about how long each review step should take. Yet without time frames, CTP is limited in its ability to evaluate policies, procedures, and staffing resources in relation to its review process and this, in turn, limits CTP’s ability to reasonably assure efficiency and effectiveness. As a result, CTP is limited in its ability to determine the adjustments needed to make improvements. For example, CTP is limited in its ability to evaluate whether OS staff are performing efficiently and effectively in relation to specific review steps, and as a result, CTP may not appropriately make adjustments such as changing an individual staff member’s responsibilities or increasing the number of available staff. As of January 7, 2013, CTP granted more meetings than it denied. The number of calendar days from the date a meeting request was received to the date a meeting was held varied widely, and CTP officials reported that logistics and subject matter contributed to these variations. As of January 7, 2013, CTP’s offices had responded—granted, denied, or transferred—to over 93 percent of the meeting requests they received through January 7, 2013. Based on the data provided by CTP officials from the four offices that received meeting requests from outside entities, CTP’s offices responded to 108 of the 116 meeting requests received as of January 7, 2013 (see table 4). Of these 108 responses, 72 of the meeting requests were granted, 22 were denied, and 14 were transferred According to CTP officials, in some cases, to another office within CTP.the CTP office denied a meeting request because the office was able to address the entity’s questions by telephone and a formal meeting was no longer necessary. The remaining eight meeting requests were pending or withdrawn as of January 7, 2013. CTP officials told us that since January 7, 2013, they responded to three of the five pending meetings by granting two meetings and denying one. According to CTP officials, as of July 2013, the other two meetings were still pending because the meeting requester had not responded to CTP. Of the 116 meeting requests from outside entities, most (74) were requested by tobacco manufacturers. Public health advocacy organizations had the second highest number with 19 meeting requests (see fig. 7). The data compiled by the CTP offices did not include data on whether the transferred meeting requests were either granted or denied by the office receiving the transferred request. As a result, a transferred meeting request may also be counted as granted or denied in the office that received the transferred request. requested meetings in order to provide information to CTP that may be useful for CTP’s work. The number of calendar days taken from the date a CTP office received a meeting request to the date the meeting was held varied widely. For example, in OP, the number of days from the date a meeting request was received to the date a meeting was held ranged from 3 days to almost five months, with half of the responses to meeting requests taking more than about 1.5 months. Further, for OCD, the number of days from the date a meeting request was received to the date a meeting was held ranged from 9 days to more than 8 months with at least half of the responses to meeting requests taking over 2.5 months. (See table 5.) For tobacco manufacturers, the type of entity with the most meeting requests, the amount of time taken from the date the meeting request was received to the date the meeting was held also varied by office. For example, the minimum number of days from a meeting request to the date the meeting was held for OS was about a month, and the maximum was about 5 months, with half of the responses to meeting requests taking more than about 3 months. The minimum number of days from a meeting request to the date the meeting was held for OP was 3 days, and the maximum was almost 4 months, with half of the responses to meeting requests taking more than about 1.5 months. According to CTP officials, logistics for scheduling meetings and the subject of the request contributed to the wide variation in time taken from the date of the request to the date the meeting was held. For example, OP officials said that the entity requesting the meeting may have to coordinate travel for several people across many locations in order to schedule a meeting and this coordination may contribute to a longer period of time before the meeting will take place. In addition, the subject matter of the request was another factor that CTP officials reported as contributing to the time taken by CTP offices to hold a meeting. For example, officials from OS said that CTP is a new regulatory agency and, as a result, it sometimes receives meeting requests on subject matters with which the center is unfamiliar and officials must involve many entities within both CTP and FDA to determine several things, including which office within CTP should host the meeting and what information the requested entity should prepare. As of the end of fiscal year 2012, FDA had spent less than half of the tobacco user fees collected and CTP had spent less than planned. CTP officials reported that issues related to contracting contributed to lower than expected spending. As of the end of fiscal year 2012, FDA had spent less than half of the $1.1 billion in tobacco user fee funds collected (46 percent) from fiscal year 2009 through fiscal year 2012, leaving more than $603 million (54 percent) unspent. (See fig. 8.) Of the almost $513 million spent during this time, CTP spent almost $468 million. The remaining funds were spent by other FDA entities, such as the Office of Regulatory Affairs. In fiscal years 2011 and 2012, CTP spent less than the amounts it identified in its spend plan—that is, spent less than planned. According to CTP officials, the center’s spend plan identifies plans for spending CTP’s user fee funds on staffing, acquisitions, and operational needs. The spend plan is based on user fee funds anticipated to be collected by FDA and user fee funds that CTP did not spend in the previous fiscal year. Based on the spend plan for fiscal year 2011, all seven CTP offices had planned on spending a total of $225.4 million for fiscal year 2011, and these offices spent $106.4 million for that year. CTP continued to spend less than planned for fiscal year 2012. (See table 6.) CTP officials reported that based on spending through the third quarter of fiscal year 2013, the difference between the amount of planned spending and the amount of actual spending in fiscal year 2013 will be less than the differences between planned and actual spending in previous years. CTP planned to spend more than $810 million in fiscal year 2013, and as of June 30, 2013, CTP has spent or is committed to spend over $712 million. Specifically, six of the seven CTP offices spent less user fee funding than CTP planned for fiscal years 2011 and 2012. For example, for fiscal year 2011, CTP’s Office of Health Communication and Education, OCE, and OS planned to spend about $30 million more than they actually spent; and the Office of Management was the only CTP office that planned to spend less than it actually spent—it planned to spend about $1 million less than it spent. (See fig. 9.) CTP officials told us that issues related to contracting accounted for most of the difference between the amounts spent and planned spending. Specifically, they reported that the time it took to award contracts resulted in CTP not spending the funds that the center planned to spend for a given fiscal year. For example, according to CTP officials, CTP’s Office of Health Communication and Education had planned to award a $55 million contract for communications support services for part of its public education campaign for fiscal year 2011. This office also planned to award a related $145 million contract in fiscal year 2012 for a public health education campaign. However, most of the planned $200 million total was not awarded until the first quarter of fiscal year 2013.officials told us that both contracts were not awarded at these amounts in fiscal year 2011 or 2012 as planned because CTP and FDA spent significant amounts of time to determine the structure of the contract as FDA had never conducted a public education campaign of this magnitude. Spending for other contracts for both fiscal years 2011 and 2012 was lower than expected for a number of reasons, according to CTP officials: fewer than expected contracts were awarded, the scope of a contract changed, or CTP was short of staff to support the work of the contract. For fiscal year 2011, CTP’s OCE had planned to award $55 million in contracts with states to ensure compliance with tobacco regulations, but CTP awarded a total of $24 million for that fiscal year because fewer states participated than expected. For fiscal year 2012, CTP’s OS entered into an interagency agreement with the Centers for Disease Control and Prevention to develop analytical methods and establish baseline levels of harmful or potentially harmful constituents in tobacco products for $20 million less than planned because of a change in scope of the activities for this contract. For fiscal year 2011, CTP’s Office of Health Communication and Education entered into an interagency agreement with the National Institutes of Health to support regulatory communications activities. The agreement was $3.5 million less than initially planned because the Office of Health Communication and Education was just being established at the time and it did not have enough staff to support this joint effort. As a result, the office reduced the scope of the contract. In addition to issues related to contracting, CTP officials said that plans to hire more staff than it did and planned management related activities that were not undertaken were other reasons why the amounts spent were lower than planned. According to CTP officials, for fiscal years 2011 and 2012, CTP had planned to hire more staff than it did and this accounted for $6 million and $10 million of the differences between amounts planned to be spent and spent, respectively. Further, according to CTP officials, lower than planned spending for other management activities (such as computer updates and planning potential reorganization) is another reason why the amounts spent by CTP were lower than planned. For example, for fiscal year 2011, the CTP spend plan included $35 million for planning associated with establishing two new offices within CTP. According to CTP officials, this amount was expected to cover contingencies, such as computer updates or management development, if they were needed. However, the officials reported that this reserve was not used because funds were available in the Office of Management to handle any issues related to the addition of these new offices. Four years after the Tobacco Control Act established CTP and about 3 years after the first new tobacco product submission, FDA has received about 4,000 submissions and collected over $1.1 billion in tobacco user fee funds. Although CTP has finished initial review steps for most of these submissions, as of June 2013, the center made a final decision on only 6 submissions and the time taken on reviews has been significant. Certainly, insufficient information provided by manufacturers in submissions, the prioritization of regular SE submission reviews over provisional SE submissions, and other factors have contributed to the time CTP has taken in its reviews. Yet, as CTP moves forward with its work, the lack of performance measures like time frames for reviews of SE submissions will limit CTP’s ability to evaluate policies, procedures, and staffing resources in relation to CTP’s submission review process and, in turn, limit CTP’s ability to reasonably assure efficient operations and effective results. An entity that is limited in its ability to evaluate its performance will be hard-pressed to determine what adjustments it should make to its operations or how to plan for the future. To improve CTP’s ability to operate efficiently, achieve effective results, and plan appropriately, we recommend that the Secretary of Health and Human Services direct the Commissioner of FDA to establish performance measures that include time frames for making final decisions on SE submissions and Exemption from SE submissions, and monitor FDA’s performance relative to those time frames, such as evaluating whether staff are performing reviews of these submissions efficiently and effectively. We provided a draft of this report to HHS for comment. In its written comments, reproduced in appendix II, HHS agreed with our recommendations. Specifically, HHS stated that FDA will identify performance measures and time frames for regular SE and Exemption from SE review processes within 6 months of our report’s publication and that FDA will monitor its progress to determine if subsequent SE reviews meet the identified time frames. In addition, HHS commented that FDA will identify performance measures and time frames for the provisional SE review process as FDA gains more experience reviewing these SE submissions. HHS further stated that based on the actual performance of meeting the identified time frames, FDA will make modifications to the review process, if appropriate, in order to meet agency objectives. HHS also provided additional information on CTP activities in its comments. For example, HHS stated that CTP is working to reach determinations on SE and Exemption from SE submissions as expeditiously as possible, and that CTP has continued to make progress on conducting product reviews and in its process and timeliness for responding to requests for meetings with CTP offices. Regarding tobacco user fee funds, HHS commented that CTP is projecting that it will decrease the amount of unspent tobacco user fee funds to carry over at the end of fiscal year 2013 to the mid-$200 millions, which is less than half of the amount carried over at the end of fiscal year 2012. HHS also suggested that our report should include information on all user fee spending, including spending by FDA entities other than CTP. We do report total user fees spent and not spent by FDA, including spending by both CTP and other FDA entities, through fiscal year 2012. In comparing spend plans with actual spending, we reported on spending by CTP, which comprised more than 90 percent of the $513 million spent by FDA through fiscal year 2012. In reporting on CTP spending, we clearly note that other FDA entities, including the Office of Regulatory Affairs, Headquarters, and the Office of the Commissioner, spend tobacco user fee funds, and that these entities spent $11 million in fiscal year 2011 and $24 million in fiscal year 2012. HHS also provided technical comments that were incorporated, as appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the Secretary of Health and Human Services, the Commissioner of FDA, and other interested parties. In addition, the report will be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or at crossem@gao.gov. Contact points for our Office of Congressional Relations and Office of Public Affairs can be found on the last page of this report. Other major contributors to this report are listed in appendix III. As of January 7, 2013, the Office of Science (OS)—the only Center for Tobacco Products (CTP) office involved in all steps of reviewing new tobacco product submissions—had 124 staff members on board, and the majority of the staff (102 or 82 percent) reported spending some portion of their time reviewing new tobacco product submissions. OS has other responsibilities in addition to reviewing new tobacco product submissions, including research to meet regulatory science needs and to evaluate the population and public health impact of tobacco products. According to OS officials, of the 102 staff who reported spending time on reviewing submissions, 60 percent or 61 staff reported that in general they spent at least half of their time working on reviews of new tobacco product submissions. The remaining 41 staff reported generally spending less half of their time on reviews of new tobacco product submissions. (See fig. 10.) The amount of time an OS staff person reported spending on new tobacco product submissions varied by job title. Specifically, the 23 project managers, the OS officials responsible for coordinating the reviews of new tobacco product submissions, and 17 scientists (such as chemists and toxicologists) reported spending at least half of their time working on reviews of new tobacco product submissions. Meanwhile, the Deputy Director for Research and the Special Assistant to the Director reported spending less than half of their time on the review of new tobacco product submissions. (See table 7.) In addition to the contact named above, Kim Yamane, Assistant Director; Danielle Bernstein; Hernán Bozzolo; Britt Carlson; Cathleen Hamann; Richard Lipinski; and Lisa Motley made key contributions to this report.

In 2009, the Family Smoking Prevention and Tobacco Control Act granted FDA, an agency within the Department of Health and Human Services (HHS), authority to regulate tobacco products such as cigarettes. The act requires that tobacco manufacturers submit information to be reviewed by FDA in order to market new tobacco products and established tobacco user fees to fund FDA's tobacco-related activities. The act represents the first time that FDA has had the authority to regulate tobacco products. Manufacturers have raised concerns about the progress of CTP, the FDA center established by the act to implement its provisions. GAO was asked to examine CTP's review of new tobacco product submissions, responses to meeting requests, and use of funds. This report examines (1) the status of CTP's reviews of new tobacco product submissions; (2) how CTP responded to manufacturers' and other entities' meeting requests, and the length of time CTP took to hold the meetings; and (3) the extent to which FDA has spent its tobacco user fee funds. GAO analyzed data regarding submissions received by FDA as of January 7, 2013; reviewed data on meeting requests, spending plans, and amounts obligated; and interviewed CTP and tobacco industry officials. As of January 7, 2013, the Food and Drug Administration's (FDA) Center for Tobacco Products (CTP) had finished initial, but not final, review steps for most of about 3,800 submissions for new tobacco products (those not on the market on February 15, 2007). Ninety-nine percent of the submissions received by FDA were made under the substantial equivalence (SE) pathway. CTP determines whether the new tobacco product in an SE submission has the same characteristics as a predicate tobacco product (a product commercially marketed in the United States on February 15, 2007, or previously found by FDA to be substantially equivalent) or has different characteristics that do not raise different questions of public health. Initial review steps include CTP's determination of whether the new product is a type regulated by FDA and whether the submission is missing information. For most SE submissions, CTP took more than a year and a half from the date a submission was received to the date these initial steps were completed. Of the 3,788 SE submissions, 3,165 were received by FDA prior to a statutory deadline (March 22, 2011) allowing the product to be marketed unless CTP finds that they are not substantially equivalent. SE submissions received after that date cannot be marketed until CTP determines they are substantially equivalent. In late June 2013, CTP made a final decision on 6 of the 3,788 SE submissions, finding that 2 of the products were substantially equivalent and that 4 were not; the remaining submissions were still undergoing CTP review. CTP officials and manufacturers told GAO that several factors (such as CTP requests for additional information from manufacturers for submissions and having to hire and train new staff) impacted the time it took CTP to review SE submissions. While CTP is working to address these factors by, for example, disseminating information to manufacturers to improve submission quality and developing training for staff, CTP does not have performance measures that include time frames for making final decisions on submissions by which to assess its progress. Without time frames, CTP is limited in its ability to evaluate policies, procedures, and staffing resources in relation to its review process and, in turn, is limited in its ability to reasonably assure efficiency and effectiveness. A variety of outside entities (such as manufacturers) have requested meetings with CTP to discuss new tobacco product submissions, public health activities, and other issues, and four CTP offices have received meeting requests. Those offices granted more meetings (72) than they denied (22) of all the meeting requests they received through January 7, 2013. The number of calendar days from the date a meeting was requested to the date it was held ranged from 1 to 262 days, and the averages among the four offices ranged from 51 to 97 days. FDA spent (obligated) less than half of the nearly $1.1 billion in tobacco user fees it collected from manufacturers and others through the end of fiscal year 2012; $603 million of these user fees remained unspent and, thus, remained available to CTP. CTP spent substantially less than planned in fiscal years 2011 and 2012. CTP had planned on spending a total of $611 million for fiscal year 2012; instead, the center spent $272 million for that year. CTP officials told GAO that the time it took to award contracts contributed to the center spending less than planned. For example, CTP planned to award a $145 million contract in fiscal year 2012 for a public health education campaign, but most of that amount was not awarded until the first quarter of fiscal year 2013. GAO recommends that FDA establish performance measures that include time frames for making decisions on new tobacco product submissions and that the agency monitor performance relative to those time frames. HHS agreed with GAO's recommendations.

This section describes (1) DOE’s project management requirements for capital asset projects and cleanup projects, (2) TRU waste operations at Area G, (3) the TRU waste removal project, (4) the TWF construction project, and (5) GAO’s Cost Guide. DOE has established separate project management requirements for its capital asset projects and certain cleanup projects defined as operations activities. DOE’s project management order for capital asset projects, Order 413.3B, establishes the requirements for managing capital asset projects, and EM’s Operations Activities Protocol establishes the requirements for managing cleanup projects defined as operations activities. The TRU waste removal project is a cleanup operations activity and is subject to EM’s Operations Activities Protocol, whereas the TWF is a capital asset construction project and must be carried out in accordance with Order 413.3B. DOE’s project management order for capital asset projects, Order 413.3B, applies to all capital asset projects with a total project cost greater than or equal to $50 million. Capital asset projects include construction projects that build large complexes that often house unique equipment and technologies such as those that process TRU waste or other radioactive material. DOE’s order establishes a process for NNSA and other DOE offices to manage projects, from identification of need through project completion, with the goal of delivering projects within the original performance baseline that are fully capable of meeting mission performance and other requirements, such as environmental, safety, and health standards. In particular, the order defines five major milestones— or Critical Decision (CD) points—that span the life of a project: CD-0: Approve mission need. DOE identifies a credible performance gap between its current capabilities and capacities and those required to achieve the goals defined in its strategic plan. The mission need translates this gap into functional requirements. DOE formally establishes a project and begins the process of conceptual planning and identifying a range of alternative approaches to meet the identified need. CD-1: Approve alternative selection and cost range. DOE completes the conceptual design and selects its preferred approach based on analysis of life-cycle costs, and approves the project’s preliminary cost range to complete the project’s design and construction. CD-2: Approve the performance baseline. A project’s performance baseline consists of key cost, scope, schedule, and performance parameter targets. The project’s scope defines the technical goals and requirements that the project is to deliver at completion. The performance baseline cost includes the entire project budget, or total project cost, and represents DOE’s commitment to Congress. At this milestone, DOE completes its preliminary design and develops a definitive cost estimate that is a point estimate and no longer a range. CD-3: Approve the start of construction. Design and engineering are essentially complete and have been reviewed, and project construction or implementation begins. CD-4: Approve the start of operations or project completion. For construction projects, at this milestone, DOE completes the project and begins the transition to operations. DOE’s project management order for capital asset projects specifies the requirements that must be met, including for developing project cost estimates, along with the documentation necessary, to move a project past each CD point. In addition, the order requires senior management to review the supporting documentation and decide whether to approve the project at each CD point. DOE also provides suggested approaches for meeting the requirements contained in the order through a series of guides, such as guides for cost estimating and project reviews. The life-cycle cost estimate data are maintained in EM’s Integrated Planning, Accountability, and Budgeting System and are used to develop DOE’s reported environmental liability. performance and, if practical, how that performance affects the life- cycle cost estimate and contract period of performance baseline. Before 1970, TRU waste generated at LANL was managed as low-level radioactive waste and was disposed of at Area G in pits and trenches along with hazardous waste. In 1970, in response to concerns that TRU waste remained radioactive for an extremely long time and scientific research recommending deep geologic disposal for this waste, the Atomic Energy Commission––a DOE predecessor––directed sites that generated TRU waste to begin segregating it from other waste and storing it in retrievable packages for an interim period, pending disposal in a repository. As a result of the directive, starting in the early 1970s, the TRU waste generated at LANL was stored in segregated TRU waste pits and trenches and aboveground in fabric domes so that it could be more easily retrieved when a permanent repository site opened. (see fig. 1). Today, Area G serves as LANL’s primary location for storing and processing TRU waste. Both legacy and newly generated TRU waste are stored and processed for shipping to WIPP at facilities in Area G. TRU waste operations at Area G include the following processes: packaging waste into 55-gallon drums or other approved containers following DOE standards, called waste acceptance criteria, to protect human health, safety, and the environment during the waste’s transport to and disposal in WIPP; repackaging containers if they are found to not meet WIPP’s waste acceptance criteria; resizing large waste using methods such as cutting it into smaller pieces so that it can be placed into approved containers; characterizing the waste by using specialized scanning equipment to assess the contents of each waste container and the amount of radioactivity it contains; and certifying the waste to declare that it meets WIPP’s waste acceptance criteria. Starting in 2011, NNSA and the New Mexico Environment Department agreed to significant changes in the strategy for completing the TRU waste removal project. In that year, a wildfire occurred near Area G, resulting in increased public concern about the risk posed by the TRU waste stored aboveground at Area G. To address this risk, in 2012, NNSA and the New Mexico Environment Department reached a voluntary agreement, called the Framework Agreement, which established a June 2014 deadline for the accelerated removal of 3,706 cubic meters of aboveground TRU waste at a high risk from wildfires. To meet this deadline, NNSA initiated an effort know as the “3706 Campaign.” To facilitate this campaign, the New Mexico Environment Department and DOE agreed to extend other deadlines established under the 2005 Consent Order governing hazardous waste cleanup activities for locations across LANL. With these deadlines extended, NNSA was able to reallocate EM funding for environmental cleanup activities at LANL to focus on the 3706 Campaign. Using the additional funds, NNSA increased the TRU waste processing capacity at LANL by constructing more facilities for repackaging waste and hiring additional contractors to operate the facilities 7 days a week. Under the Framework Agreement, NNSA also agreed to remove the remaining aboveground TRU waste at Area G and developed a schedule for removing the belowground TRU waste at Area G by September 30, 2018. According to NNSA officials, part of the plan in establishing the Framework Agreement was that, once the 3706 Campaign was completed, NNSA and the New Mexico Environment Department would discuss renegotiating the final completion date for the Consent Order. On May 30, 2014, DOE announced that NNSA had completed the removal of about 93 percent of the TRU waste included in the 3706 Campaign but would not meet the June 2014 deadline because of the department’s decision to halt the TRU waste removal project. As noted previously, DOE halted the project at LANL in February 2014, in response to an incident at WIPP that involved a LANL TRU waste container that ruptured and leaked radioactive material while in storage. The TWF project is to replace NNSA’s existing capabilities that reside at Area G for storage, characterization, and certification of newly generated TRU waste at LANL. The TWF design includes multiple buildings for waste storage, waste characterization, and operational support. The facility will also have space and utility hookups for three mobile trailers to be provided by WIPP that will contain additional characterization capabilities needed to certify that TRU waste containers meet WIPP waste acceptance criteria (see fig. 2). The TWF is being designed and constructed as a high-hazard nuclear facility, which must meet nuclear safety standards for storage and handling of nuclear waste. Nuclear safety design features of the TWF include a barrier to prevent large vehicles from crashing into the facility, a seismic power cutoff switch designed to reduce possible sources of fire that could result from an earthquake, and a tank to store water to help suppress any earthquake- initiated fire. NNSA’s Office of Defense Programs, which is the program office responsible for maintaining the nation’s nuclear weapons stockpile, is sponsoring the TWF project. The office provides the annual funding for planning and construction and approves the project’s milestones. NNSA’s Office of Acquisition and Project Management is responsible for overseeing the construction of the TWF within NNSA’s approved cost and schedule estimates. To do so, the office provides direction and oversight of NNSA’s management and operating contractor at LANL. NNSA divided the TWF project’s design and construction into two subprojects: site development and facilities construction. NNSA completed the site development activities, which included relocation of utility lines, as well as excavation and grading to prepare the site for the facility’s construction, in December 2012, at a cost of $7.7 million. In February 2013, NNSA approved the project’s CD-2 performance baseline estimate of $99.2 million to construct the TWF with a completion date between April 30, 2016, and January 31, 2018. As mentioned previously, NNSA also estimated that the facility will cost $300 million to operate and maintain for its projected useful life of 50 years, spanning 2018 through 2068. In combination, NNSA’s estimated life-cycle costs for the TWF when it approved the project’s performance baseline to complete construction at CD-2 totaled about $406.9 million. Drawing from federal cost-estimating organizations and industry, the Cost Guide provides best practices about the processes, procedures, and practices needed for ensuring development of high-quality—that is, reliable cost estimates.characteristics of a high-quality, reliable cost estimate: The Cost Guide identifies the following four Comprehensive when it accounts for all life-cycle costs associated with a project, is based on a completely defined and technically reasonable plan, and it contains a cost estimating structure in sufficient detail to ensure that costs are neither omitted nor double- counted; Well-documented when supporting documentation explains the process, sources, and methods used to create the estimate and contains the underlying data used to develop the estimate; Accurate when it is not overly conservative or too optimistic and is based on an assessment of the costs most likely to be incurred; and Credible when a sensitivity analysis has been conducted, the level of confidence associated with the point estimate has been identified through the use of risk and uncertainty analysis, and the point estimate has been cross-checked with an independent cost estimate (ICE). To develop a cost estimate that embodies these four characteristics, our Cost Guide lays out best practice steps. For example, one step in developing an accurate estimate is to identify and document ground rules that establish a common set of agreed-on estimating standards and solid assumptions that are measurable, specific, and consistent with historical data. According to the Cost Guide, it is imperative that cost estimators brief management on the ground rules and assumptions used for an estimate so that management understands the conditions the estimate was structured on and can avoid overly optimistic assumptions. NNSA’s TRU waste removal project at LANL did not meet its 2006 cost estimate and is not expected to meet the 2009 cost estimate established for the completion of the project. During our review, NNSA and EM were in the process of developing a new cost estimate for the project. The TRU waste removal project has not met its past cost estimates, partly because the 2006 and 2009 cost estimates were based on aggressive funding assumptions to meet the deadlines of the Consent Order. In addition, because NNSA did not maintain or use two of the three project baselines outlined in its cleanup project requirements, it could not measure the progress of the total project. As of the end of fiscal year 2014, NNSA had spent about $931 million on the project, which exceeded the 2006 cost estimate of $729 million by $202 million (see fig. 3). The amount expended by the end of fiscal year 2014 did not exceed the $1.2 billion upper range of the 2009 cost estimate, but it did exceed the $848 million lower range of the estimate by $83 million. As of July 2014, the most recent date for which data were available, NNSA had removed approximately 79 percent of the TRU waste at LANL; however, the remaining 21 percent includes the waste buried belowground, which will be the most difficult and expensive to address, according to NNSA officials. The new fiscal year 2015 draft estimate, currently under review by NNSA and EM, projects that the final project costs will be approximately $1.6 billion, or $400 million above the upper range of the 2009 cost estimate. NNSA’s TRU waste removal project exceeded the 2006 cost estimate and is not expected to meet the 2009 cost estimate, in part, because NNSA and EM developed the cost estimates using aggressive project funding assumptions that were based on the need to meet the Consent Order requirement for closing Area G by 2015, according to NNSA and EM officials. For the 2006 cost estimate, NNSA officials overseeing the TRU waste removal project developed the parameters of the project estimate based on the need to remove almost all of the TRU waste by 2012. In particular, to meet these deadlines, NNSA based its cost estimate on funding projections provided by EM that were consistent with meeting the Consent Order deadline and that assumed that EM would increase the yearly funding for environmental cleanup projects at LANL. According to NNSA and EM officials, they recognized that the funding assumptions used in the cost estimate were aggressive and that significant funding shortfalls would inhibit the TRU waste removal project’s ability to remain on schedule. From fiscal year 2006 through fiscal year 2008, EM provided approximately $457 million—$284 million (38 percent) less than the amount requested by NNSA officials for all cleanup activities at LANL—and, as a result, it was not possible to fund the TRU waste removal project at the levels established in the 2006 estimate. According to NNSA officials, EM was unable to increase the funding for LANL cleanup projects due to limited budget flexibility and competing demands from cleanup projects at other DOE sites. In a 2008 report on LANL’s cleanup efforts, DOE’s Inspector General found that EM did not have enough money to address all the milestones in the NNSA officials said that, by environmental agreements they signed.2009, the TRU waste removal project had fallen behind schedule and could not be completed by 2012, in part, because of the shortfall in funding. The extension of the project’s completion date beyond 2012 resulted in additional costs, which contributed to the total cost of the project exceeding the 2006 estimate. In 2009, NNSA and EM developed a new cost estimate for the TRU waste removal project with a completion date in 2018 but again used aggressive funding assumptions to complete the project as close to the Consent Order’s 2015 deadline as possible. Similar to the 2006 estimate, the 2009 cost estimate used funding projections provided by EM that assumed an increase in the yearly levels of funding for LANL cleanup activities that were necessary for NNSA to complete the TRU waste removal project by July of 2018—2 and a half years after the Consent Order deadline in 2015. However, due to the same budget restrictions that affected cleanup project funding previously, actual funding levels provided by EM for fiscal years 2009 to 2012 for all cleanup projects at LANL again came in below the levels requested by NNSA officials at LANL. Specifically, LANL received approximately $1 billion over these years, which was $240 million (19 percent) less than the levels requested by NNSA officials at LANL for cleanup projects at the site. As a result, funding for the TRU waste removal project was reduced, and this reduction caused the project to fall behind the schedule set in the 2009 cost estimate, according to NNSA officials. Moreover, the 2009 cost estimate was never officially approved by NNSA and EM because the date estimated for project completion was not consistent with the requirements of the Consent Order. According to NNSA officials, they could not formally approve the 2009 estimate because it included a 2018 estimated completion date for the TRU waste removal project, which conflicted with the required 2015 closure date established in the Consent Order. In addition to developing estimates using aggressive funding assumptions, NNSA did not maintain or use two of the three project baselines outlined in the Operations Activity Protocol, so the agency could not measure the progress of the total project. As discussed previously, NNSA was to manage the TRU waste removal project using EM’s Operations Activities Protocol, which is intended to provide the framework for managing and reporting on the progress of cleanup projects through the use of three performance baselines: life-cycle cost, contract period of performance, and fiscal year work plan. However, for the TRU waste removal project, because NNSA did not have an updated life-cycle cost baseline and did not establish a contract period of performance baseline, it only used the fiscal year work plan baseline to manage the project, as discussed below: Life-cycle cost baseline. NNSA has not updated the life-cycle cost baseline for the project since 2009, even though agency officials told us they were aware that the estimate has been out-of-date since about 2012. Since that year, the project has undergone significant changes that affected its estimated cost and completion date. For example, by initiating the 3706 Campaign in 2012, NNSA altered the scope of work from what was planned in the 2009 cost estimate. NNSA and EM officials told us that, although they recognized in 2012 that the 2009 estimate was no longer valid, they did not see the purpose in completing a new estimate before completion of the campaign and the expected renegotiation of the Consent Order deadlines. According to these officials, a new cost estimate for the project would have either needed to use unreasonably high funding assumptions to achieve project completion by 2015 or, if it used more reasonable funding assumptions, it could not have been approved because of the political issues associated with a completion date beyond the 2015 Consent Order deadline. The Operations Activities Protocol requires that EM or NNSA Site Office Managers develop cost estimates that cover the full life-cycle of a cleanup project, but it leaves the Site Office Manager discretion to determine whether an update to the life-cycle cost baseline is required. Because NNSA’s LANL Site Office Manager, in consultation with EM officials, decided not to update the life-cycle cost baseline, NNSA could not measure project performance to determine the impact of management actions. For example, from fiscal year 2012 to 2014, NNSA used additional funding that was reallocated to the TRU waste removal project to increase the pace of TRU waste packaging and removal; however, because they did not have an updated cost estimate to measure against, NNSA managers were unable to identify the effect these actions had on the total cost of the project. Contract period of performance baseline. NNSA has not managed the TRU waste removal project using a contract period of performance because, according to NNSA officials, the project, like other projects at LANL, is being conducted through NNSA’s management and operating contract. According to the Operations Activities Protocol, a contract period of performance baseline is required for those cleanup projects that are executed through a contract that establishes a performance baseline for cost and scope over the duration of the contract. The management and operating contract covers all work performed at LANL, but it does not establish a cost estimate for specific projects such as the TRU waste removal project, according to NNSA officials. In contrast, for cleanup projects at EM-managed sites, the scope of the contract is typically limited to the cleanup project and would not include other site activities that were unrelated to the project. According to NNSA officials, executing the TRU waste removal project through the management and operating contract does not provide the baseline necessary for establishing a contract period of performance, so NNSA is not required to manage to a contract period of performance baseline for this project. According to EM officials, DOE determined that as part of EM’s transition to direct oversight of the legacy cleanup work at LANL, EM will transition away from using a management and operations contract to manage the remaining cleanup work. When this change in contract type is completed, the officials stated that a contract period of performance baseline will be available for monitoring performance of the work, including the TRU waste removal project. Fiscal year work plan baseline. NNSA has used the fiscal year work plans outlined in the Operations Activities Protocol to monitor the performance of portions of the TRU waste removal project and to manage and assess the performance of the entire TRU waste removal project in the absence of an accurate life-cycle cost estimate baseline or a contract period of performance baseline. According to NNSA officials, a new fiscal year work plan is developed each year for the TRU waste removal project using an integrated priorities list for remaining TRU removal work and the projected funding amount allocated for LANL cleanup. However, while NNSA and EM site and headquarters officials monitored progress against the fiscal year work plan, the agency was unable to evaluate the performance of the entire TRU waste removal project and identify potential cost overruns because its life-cycle cost estimate was out-of-date, and it did not manage to a contract period of performance. The Operations Activities Protocol requires that the Site Office Manager report yearly on any variances between total project costs to date and the estimated costs for the entire project as part of the fiscal year work plans. However, because NNSA did not have an accurate cost estimate for the TRU waste removal project, the 2012, 2013, and 2014 fiscal year work plans used the outdated 2009 cost estimate to report on the variances between the current and estimated costs for the project. As a result, these fiscal year work plans did not provide accurate information on project performance to date to help managers measure total project performance and manage costs. At the time of our review, NNSA and EM were in the process of developing a new cost estimate for the TRU waste removal project that they expect to complete in fiscal year 2015; however, this estimate may quickly become inaccurate due to assumptions related to funding and the status of WIPP that could be invalidated. As discussed previously, the new draft cost estimate increases the estimated total cost of the project to $1.6 billion, with a completion date in fiscal year 2023 (i.e., October 2022). According to an NNSA official, this draft estimate uses a more conservative approach than past estimates by expanding the scope of the project estimate to include the costs to remove additional TRU waste that was not included in past estimates.concerning when WIPP will reopen and the need to address the impending Consent Order deadlines in 2015 that have not yet been renegotiated with New Mexico, the same NNSA official told us that several of the new estimate’s assumptions are no longer valid or may not be valid after a few years. For example, the funding assumptions in the new estimate may already be invalid. With the TRU waste removal project on hold as a result of the WIPP closure, the New Mexico Environment Department is no longer providing deadline extensions for some of LANL’s hazardous waste cleanup work and, as a result, NNSA officials managing the cleanup work have had to reprioritize their funding to attempt to meet those deadlines. Restoring funding to hazardous waste cleanup projects reduces the funding available for TRU waste removal, which would invalidate the funding assumptions used in the new cost estimate. In addition, the NNSA official told us that the new cost estimate also assumes that LANL will resume TRU waste shipments to WIPP in fiscal year 2017 based on an unofficial estimate from the WIPP manager that WIPP will reopen between 18 and 30 months after the initial However, in light of the uncertainty assessment of the incident.shipments from LANL in this time frame, the TRU waste removal project at LANL would be delayed, resulting in additional costs not accounted for in the new estimate, such as the costs for maintaining the project workforce longer than anticipated. If WIPP does not start accepting TRU waste To objectively measure the performance of the TRU waste removal project and take action to manage project costs, NNSA managers of the project need an updated cost estimate. EM’s Operations Activity Protocol leaves it to the discretion of the Site Office Manager to update cleanup project estimates, and NNSA’s LANL Site Office Manager chose to delay revising the outdated 2009 estimate. As a result, DOE does not have an estimate of the total cost or completion date of the TRU waste removal project that uses updated assumptions based on the current understanding of project conditions. NNSA and EM are developing a new cost estimate for the project; however, the new estimate may quickly become inaccurate because of changes in funding, and the status of the WIPP may soon invalidate its assumptions. According to best practices for cost estimating, maintaining an updated cost estimate is critical so that officials making decisions about the future management of a project have accurate information for assessing their alternatives. By revising the TRU waste removal project’s estimate to include the current understanding of project conditions, NNSA program managers could more accurately identify cost overruns. NNSA’s cost estimate for the TWF partially met best practices. More specifically, NNSA’s cost estimate—which consisted of separate cost estimates for completing construction and for operations and maintenance, as the TWF’s life-cycle costs—partially reflected each of the four characteristics of a reliable estimate (comprehensive, well- documented, accurate, and credible) as established by best practices. In developing the construction estimate, NNSA took several steps that conformed to best practices, such as validating the construction estimate by completing an ICE. Although DOE’s project management order for capital asset projects does not require the use of all best practices in developing a cost estimate, DOE’s related cost-estimating guidance, which is optional, describes most of them. In contrast, in developing its operations and maintenance cost estimate, NNSA did not take steps that conformed to best practices. In particular, NNSA did not sufficiently document the approach (i.e., data sources and methodologies) used to develop the estimate, even though operations and maintenance costs represented about 74 percent of the total life-cycle costs of the facility. The reason was that operations and maintenance cost estimates for a construction project do not need to be updated and documented at CD-2 under DOE’s project management order, as funds for these costs do not need to be specifically requested from Congress to complete the project. By not sufficiently documenting the approach used to develop the estimate, NNSA may not have reliable information to support budgetary decisions for funding the TWF’s operations and maintenance in the future. For example, the contractor’s representatives told us that they estimated the operations and maintenance costs to be $6 million annually from 2018 through 2068, for a total of $300 million but did not use an inflation rate in the calculations. Thus, although $6 million may be an accurate estimate for the first year, without documentation of the approach used to develop the estimate, we could not determine its reliability for the first or future years. Appendix II provides a summary description of our assessment of NNSA’s cost estimates for the TWF project’s construction and operations and maintenance. The following are examples from our assessment, by best practice characteristic: Comprehensive. The TWF estimates partially reflected the characteristics of comprehensive cost estimates. For example, NNSA partially followed the best practice for completely defining the program, as NNSA’s contractor based the TWF construction cost estimate on a mature design plan that detailed the technical requirements and characteristics for the TWF. In contrast, for the TWF operations and maintenance cost estimate, NNSA’s contractor was not able to provide us with definitions of the technical requirements and characteristics that would have formed the basis of the estimate, although the $300 million estimate represented a substantial change from a $642 million estimate that NNSA’s contractor produced in June 2010 for CD-1 (approve alternative selection and cost range). As a result, we could not determine whether the $300 million estimate reflected the most recent TWF design approved at CD-2 to complete construction. NNSA officials told us they did not develop an updated and documented basis to support the operations and maintenance cost estimate because DOE’s project order does not require updated and documented estimates of all life-cycle costs at CD-2. Instead, at CD-2, the order focuses on the need for the baseline cost estimate to complete construction because funding for construction needs to be specifically requested from Congress to complete the project. According to best practices, clearly defining the technical requirements would help to ensure that managers have an adequate understanding of the facility and where information was limited and assumptions were made in developing the estimate. Well-documented. The TWF estimates partially reflected the characteristics of well-documented cost estimates. For example, regarding the construction estimate, NNSA’s contractor documented the data sources and the methodology used to calculate the construction estimate so that a cost analyst unfamiliar with the project could understand what was done and replicate the estimate. In contrast, NNSA did not document the approach (data sources and methodologies) used to develop the operations and maintenance estimate, even though the operations and maintenance costs represented about 74 percent of the TWF’s life-cycle costs. As mentioned previously, DOE’s project management order does not require documentation of the operations and maintenance costs at CD-2. NNSA was required to report the operations and maintenance cost estimate by following a DOE budget formulation guidance that did not specify requirements for documenting the estimate. Because NNSA did not document the approach used, we could not determine whether it was appropriate for developing the operations and maintenance estimate; whether NNSA management reviewed the estimate, including its risks and uncertainties, or whether NNSA management approved the estimate. Accurate. The TWF estimates partially reflected the characteristics of accurate cost estimates. For example, NNSA’s contractor partially followed the best practice for properly adjusting the estimates for inflation. Regarding the construction estimate, NNSA’s contractor developed the estimate using pricing data that were adjusted for inflation. However, the contractor did not then normalize the data to remove the effects of inflation. According to representatives from the contractor, they did not believe data normalization was an applicable step for the TWF construction estimate because the data set was too small. According to cost-estimating best practices, data normalization is often necessary to ensure comparability of data sets because data can be gathered from a variety of sources and in different forms that need to be adjusted before being used. Regarding the operations and maintenance estimate, NNSA’s contractor did not properly adjust the estimate for inflation over the 50-year useful life of the TWF. Specifically, the contractor’s representatives told us that they estimated the operations and maintenance costs to be $6 million annually from 2018 through 2068, for a total of $300 million, but they did not use an inflation rate in these calculations. Adjusting for inflation is an important step in developing an estimate because if the inflation amount is not correct, the estimate is not accurate. Applying the wrong inflation rate will either result in a higher cost estimate or estimated costs that are not sufficient to keep pace with inflation. Credible. The TWF estimates partially reflected the characteristics of credible cost estimates. For example, NNSA followed the best practice to have an ICE completed in January 2013 to validate the TWF construction costs. However, because DOE’s project management order does not define the operations and maintenance costs as project costs, NNSA was not required to include these costs in the ICE. By not including the TWF operations and maintenance costs in the ICE, NNSA managers may lack insight into these future costs. According to best practices, an ICE can provide NNSA managers with additional insight into the TWF’s potential operations and maintenance costs—in part, because ICEs frequently use different methods and are less burdened with organizational bias. Therefore, according to best practices, an ICE can be used as a benchmark to assess the reasonableness of the contractor’s proposed operations and maintenance costs, improving NNSA management’s ability to make sound investment decisions, and accurately assess the contractor’s performance. Moreover, because DOE’s project management order does not require it, NNSA’s contractor did not follow the best practice to complete a sensitivity analysis to quantify the extent to which either the construction or operations and maintenance cost estimates could vary because of changes in key assumptions and ground rules. Such an analysis is a best practice because uncertainty cannot be avoided and, therefore, it is necessary to identify the cost elements that represent the most risk and, if possible, quantify them. According to cost-estimating best practices, doing a sensitivity analysis increases the chance that decisions that influence the design, production, and operation of the TWF will be made with a focus on the elements that have the greatest effect on cost. According to NNSA officials who oversee the TWF project, combining the TWF construction estimate with the operations and maintenance estimate to reflect the life-cycle costs and assessing the combined estimate obscured the positive steps NNSA took in developing the construction estimate. We agree that the TWF construction estimate conformed to several best practices. Examples are as follows: NNSA developed the construction estimate based on a mature design plan for the facility to ensure that it was based on the best available information at the time, NNSA documented the data sources and the methodologies used to calculate the construction estimate so that a cost analyst unfamiliar with the project could understand what was done and replicate it, and NNSA had an ICE completed to provide an unbiased test of the reasonableness of the TWF construction costs. Nonetheless, as described above, and in appendix II, we also identified examples where NNSA’s construction estimate did not conform to best practices. Because NNSA did not follow all best practices for cost estimating, particularly for the TWF operations and maintenance cost estimate, NNSA may not have reliable information to support budgetary decisions for funding the TWF’s operations and maintenance. NNSA expects the TWF may be ready to start operations as early as April 2016, and when it does, NNSA will need to balance funding for the TWF with the operations and maintenance costs for other nuclear infrastructure and facilities at LANL and other NNSA sites that make up the national nuclear security enterprise. In December 2013, we found that, as the facilities and infrastructure that support the nuclear security enterprise continue to age, maintenance costs are likely to grow. In that report, NNSA officials said that deferred maintenance projects will have to compete against programmatic priorities for funding within the overall pool of maintenance funds available. By having reliable information on the TWF’s costs, including operations and maintenance, well before the project starts operations, NNSA managers would be better able to plan for, and manage the costs of the TWF in balance with other infrastructure in the national nuclear security enterprise. GAO, Project and Program Management: DOE Needs to Revise Requirements and Guidance for Cost Estimating and Related Reviews, GAO-15-29 (Washington, D.C.: Nov. 25, 2015). Specifically regarding DOE’s cost-estimating guide, we found, in November 2014, that for two best practice steps—determining the estimating structure and conducting sensitivity analysis—the guide only partially or minimally describes those steps. operations and maintenance estimate until the project is constructed and undergoes operational readiness reviews in preparation for CD-4 (to approve the start of operations or project completion). Updating the TWF’s cost estimate to include all life-cycle costs and needed analyses would provide NNSA more reliable information for better managing the TWF as it prepares for the start of operations. Safely removing the TRU waste stored at LANL is a critical part of NNSA’s efforts to clean up the legacy environmental contamination from decades of nuclear weapons activities. NNSA has made progress in removing the TRU waste from LANL’s Area G and has monitored the project’s recent performance through fiscal year work plans. But NNSA has consistently used cost estimates for completing the project that it could not meet because the estimates were developed based on aggressive and unrealized funding assumptions, and the agency chose to delay revising the 2009 estimate when it was determined to be outdated, which was not consistent with the intent of DOE’s cleanup project requirements for maintaining an updated life-cycle cost baseline. While NNSA and EM are developing a new cost estimate for the project, the new estimate may quickly become inaccurate because changes in funding and the status of the WIPP may soon invalidate its assumptions. By revising the estimate to include the current understanding of project conditions, including the uncertainty at WIPP, NNSA program managers can more accurately identify cost overruns consistent with best practices. When completed, NNSA expects the TWF to provide TRU waste capabilities at LANL to support NNSA’s nuclear weapons mission for the next 50 years and has taken several steps that conformed to best practices in developing the TWF’s construction estimate. However, NNSA has not developed a reliable estimate of its operations and maintenance costs by, for example, not sufficiently documenting its approach and not using an inflation rate in its calculations because DOE’s project management order for capital asset projects does not require the use of all cost-estimating best practices in developing estimates of all life-cycle costs. Thus, although the operations and maintenance costs were estimated to be $6 million annually from 2018 through 2068, for a total of $300 million, without documenting the approach used to develop the estimate, and not using an inflation rate in these calculations, we could not determine the reliability of the estimate for future years. DOE agreed with the recommendations in our November 2014 report to revise its order to require that DOE, NNSA, and its contractors develop cost estimates in accordance with all best practices. However, opportunities exist currently to enhance the reliability of the TWF cost estimate. Updating the TWF’s cost estimate to include all life-cycle costs, as well as needed analyses, would provide NNSA more reliable information for better managing the TWF as it prepares for the start of operations, which NNSA expects could be as early as April 2016. To develop reliable cost estimates for the TRU waste removal project and for the TWF construction project at LANL, we recommend that the Secretary of Energy take the following two actions: Direct NNSA and EM to revise the cost estimate for the TRU waste removal project to ensure that it uses updated assumptions based on the current understanding of project conditions, such as the status of WIPP. Direct NNSA to revise and update the TWF project’s cost estimate by following all best practices for developing a reliable cost estimate that covers all life-cycle costs for better managing the project going forward. We provided DOE with a draft of this report for its review and comment. In written comments, reproduced in appendix III, NNSA provided a joint response to our draft report for itself and DOE’s EM, which generally agreed with both of the report’s recommendations. In its comments, NNSA stated that it will update its cost estimates for both the TRU waste removal project at Area G and the TWF’s operations and maintenance and provided details of the specific actions planned or taken to address both recommendations and timelines for completing these actions. NNSA also provided general and technical comments that we incorporated into the report, as appropriate. In regard to the report’s first recommendation to revise the cost estimate for the TRU waste removal project to ensure that it uses updated assumptions based on the current understanding of project conditions, NNSA stated actions have been taken to address this recommendation. Specifically, a comprehensive life-cycle baseline revision was submitted and reviewed, and EM is currently taking steps to revise and finalize this new baseline cost estimate for the project in light of realistic out-year funding profiles to support a planned renegotiation of the Consent Order with the New Mexico Environment Department. In addition, the pending changes to the type of contract used to manage the legacy cleanup work at LANL will be factored into the baseline revision. DOE plans to complete the revised cost estimate by September 30, 2015. We are pleased that DOE plans to address this recommendation and has actions under way to do so. In regard to the report’s second recommendation to revise and update the TWF project’s cost estimate by following all best practices for developing a reliable cost estimate that covers all life-cycle costs, NNSA stated in its written comments that it will update the TWF’s operations and maintenance cost estimate to ensure effective management of the facility once it is operational. Regarding the TWF’s construction estimate, as we described in the report and in appendix II, NNSA’s TWF construction estimate conformed to several but not all best practices. In particular, NNSA validated the project team’s estimate through an ICE to provide an unbiased test of the reasonableness of the TWF construction costs. Regarding the TWF’s operations and maintenance estimate, NNSA stated it will prepare the updated estimate as part of the programming process for the fiscal year 2017 budget, which takes place in fiscal year 2015, to support postconstruction activities and operations. Further, NNSA stated that the estimate will reflect operational costs for a 7-year window and incorporate applicable best practices, including documentation of any significant deviations and uncertainties impacting the estimate. The estimated completion date for these activities is March 30, 2015. We are encouraged by NNSA’s planned actions to update the TWF operations and maintenance estimate using applicable best practices. However, it will be particularly important for NNSA to document its decisions on which best practices are being followed and the reasons practices not being followed are not applicable. As we noted in the report, one of the key weaknesses we found was that NNSA did not document the approach (data sources and methodologies) used to develop the $300 million operations and maintenance estimate for the TWF, even though the operations and maintenance costs represented about 74 percent of the TWF’s life-cycle costs. Because NNSA did not document the approach used, we could not determine whether it was appropriate for developing the operations and maintenance estimate. With regard to the time frames covered by this estimate, NNSA plans to update the TWF operations and maintenance cost estimate to cover a 7-year period. Given the need now for reliable information on the estimated costs for operations, NNSA’s plan to update the TWF operations and maintenance cost estimate following applicable best practices, and covering a 7-year period, by March 30, 2015, would provide it more reliable information for managing the facility as it prepares for the start of operations. As we noted in the report, however, NNSA expects the TWF to provide TRU waste capabilities at LANL to support NNSA’s nuclear weapons mission for the next 50 years. By having a reliable and updated life-cycle estimate for the TWF that covers the estimated useful life of the facility, NNSA managers would be better able to plan for the TWF costs in balance with other infrastructure in the national nuclear security enterprise. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees, the Secretary of Energy, and other interested parties. In addition, this report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact me at (202) 512-3841 or trimbled@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix IV. Our objectives were to examine (1) the extent to which the National Nuclear Security Administration (NNSA) has met its cost targets for the transuranic (TRU) waste removal project at Los Alamos National Laboratory (LANL) and (2) the extent to which NNSA’s cost estimate for the TRU Waste Facility (TWF) project at LANL met best practices for a reliable cost estimate. To examine the extent to which NNSA’s TRU waste removal project at LANL has met its cost estimates, we reviewed documentation of NNSA’s total project cost estimates from 2006 and 2009. We focused on these cost estimates because they were developed after the establishment of the Consent Order in 2005, and they were the most recently completed estimates for the total cost of the project. During our review, NNSA was in the process of developing a new cost estimate, the draft fiscal year 2015 cost estimate, for the TRU waste removal project. We interviewed the NNSA officials and representatives from its LANL contractor who were working on the draft estimate to understand the cost estimation process and preliminary results. We reviewed data provided to us by NNSA from the Department of Energy’s (DOE) Integrated Planning, Accountability, and Budgeting System on the annual dollars spent for fiscal years 2006 through 2013 for the TRU waste removal project, as well as NNSA’s estimate of fiscal year 2014 year-end spending for the project. We compared the project spending data with the 2006 and 2009 cost estimates, and with the new draft 2015 cost estimate. We also reviewed data provided to us by NNSA from LANL’s Waste Compliance and Tracking System on total volumes of TRU waste removed from LANL from 1999 through July 2014, as well as NNSA’s estimate of the total volume of TRU waste remaining. We assessed the reliability of the project data we reviewed and analyzed, and we determined that the data for this period were sufficiently reliable to examine the extent to which NNSA’s TRU waste removal project at LANL has met its cost targets. To assess the reliability of the project data, we reviewed information provided by NNSA on the data systems used for managing and reporting the data, including the systems’ controls and checks that ensure the accuracy and completeness of the data, as well as procedures that were in place to review and certify the reliability of the data such as inspector general or internal audit reports of the quality of the data. In addition, we reviewed NNSA’s fiscal year work plans for the years after NNSA adopted the Operations Activities Protocol for managing the TRU waste removal project, fiscal years 2012 through 2014. We compared the total project cost estimates and the fiscal year work plans to the requirements for developing and using cost estimates found in DOE’s Operations Activities Protocol, which sets the requirements for managing cleanup projects at DOE defined as operations activities. To refine our analysis, we interviewed officials from NNSA’s Office of Environment, Health, and Safety in headquarters, the Environmental Project’s Office in NNSA’s LANL Site Office, and the DOE Office of Environmental Management’s (EM) Office of Disposal Operations. We also met with the contractors working on the TRU waste removal project at LANL during a visit to the site where we toured Area G and the buildings conducting TRU waste processing. average of the individual assessment ratings to determine the overall assessment rating for each of the four characteristics as follows: Not Met = 1.0 to 1.4, Minimally Met = 1.5 to 2.4, Partially Met = 2.5 to 3.4, Substantially Met = 3.5 to 4.4, and Fully Met = 4.5 to 5.0. We consider a cost estimate reliable if the overall assessment ratings for each of the four characteristics are substantially or fully met. If any of the characteristics are not met, minimally met, or partially met, then the cost estimate does not fully reflect the characteristics of a high-quality estimate and cannot be considered reliable. We conducted this performance audit from July 2013 to February 2015 in accordance with generally accepted government standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Best practice The cost estimate includes all life-cycle costs. Detailed assessment Partially met. The life-cycle cost estimate consisted of a design and construction (construction) estimate and a 50-year operations and maintenance estimate. The operations and maintenance cost estimate was not updated at Critical Decision (CD)-2 when the construction estimate was approved. The life-cycle estimate did not include retirement of the facility. The cost estimate completely defines the program, reflects the current schedule, and is technically reasonable. Partially met. The construction estimate was based on the technical requirements considered 90 percent mature. The operations and maintenance estimate was not supported by definitions of the technical requirements that would have formed the basis of the estimate. The cost estimate work breakdown structure is product-oriented, traceable to the statement of work/objective, and at an appropriate level of detail to ensure that cost elements are neither omitted nor double-counted. Partially met. The construction estimate work breakdown structure covered the major work for the end product of the TWF project. The work breakdown structure did not present all cost elements at a clear level of detail, and it was not standardized so that cost data can be collected and used for estimating future programs. The operations and maintenance estimate did not reflect a documented work breakdown structure. The estimate documents all cost-influencing ground rules and assumptions. Partially met. The construction estimate included ground rules and assumptions, such as technical specifications, vendor quotes, and registered risks. The operations and maintenance estimate did not document ground rules and assumptions. The documentation captures the source data used, the reliability of the data, and how the data were normalized. Partially met. The construction estimate documented the data sources used but not data reliability and how the data were normalized. The operations and maintenance estimate was not supported by detailed documentation. The documentation describes in sufficient detail the calculations performed and the estimating methodology used to derive each element’s cost. Partially met. The construction estimate used an engineering buildup approach for individual cost elements. The operations and maintenance estimate did not include documentation on how cost elements were derived. The documentation describes, step by step, how the estimate was developed so that a cost analyst unfamiliar with the program could understand what was done and replicate it. Partially met. The construction estimate documentation explained how work breakdown structure elements were estimated and the documentation was mathematically sensible and logical. The documentation explains how management reserve and contingency were calculated and was composed of cost and schedule uncertainty. The operations and maintenance estimate did not include documentation that detailed how the estimate was developed. Best practice The documentation discusses the technical baseline description, and the data in the baseline is consistent with the estimate. Detailed assessment Partially met. The documentation for the construction estimate matched the technical requirements document. The operations and maintenance estimate was not supported by a technical baseline document. The documentation provides evidence that the cost estimate was reviewed and accepted by management. Partially met. NNSA approved the construction estimate. The approval memo did not detail recommendations for changes, feedback, and the level of contingency reserves decided upon to reach a desired level of confidence. The operations and maintenance estimate did not document management review and approval. The cost estimate results are unbiased, not overly conservative or optimistic, and based on an assessment of most likely costs. Partially met. The construction estimate included risk and uncertainty analysis, the results included S curve cumulative probabilities. The risk and uncertainty is quantified as management reserve and contingency. The cost estimate was in range when compared with metrics that benchmark the TWF estimate to similar nuclear projects. The operations and maintenance estimate did not include documentation to help determine whether it was unbiased and not overly conservative or optimistic. The estimate has been adjusted properly for inflation. Partially met. The construction estimate was adjusted for inflation for the period of the construction schedule, but the cost data were not normalized. The operations and maintenance estimate was not properly adjusted for inflation. The estimate contains few, if any, minor mistakes. Partially met. The construction estimate included minor calculation errors in the cost summary table. We were not able to perform random sampling to check calculations for accuracy because the electronic cost model provided to us did not identify the formulas for calculations. The operations and maintenance estimate did not include detailed calculations that we could use to check its accuracy. The cost estimate is regularly updated to reflect significant changes in the program so that it always reflects current status. Minimally met. The construction estimate was updated to reflect changes in technical or program assumptions at the CD-3 (approve the start of construction) milestone but is not regularly updated with actual costs on an ongoing basis. The operations and maintenance estimate is not regularly updated to reflect changes in the project and has not been updated since June 2010. Variances between planned and actual costs are documented, explained, and reviewed. Minimally met. The construction estimate did not explain variances between planned and actual costs. The CD-2 estimate included a summary level reconciliation with the CD-1 estimate. Best practice The estimate is based on a historical record of cost estimating and actual experiences from other comparable programs. Detailed assessment Partially met. The construction estimate was based on the contractor’s market price data for the scope of work required to complete the project. The estimate also uses metrics from similar nuclear projects but did not include documentation on the reliability of the metrics data. The operations and maintenance estimate did not document whether it was based on historical or other data. The estimating technique for each cost element was used appropriately. Partially met. The construction estimate was based on engineering buildup approaches appropriate to each cost element. The operations and maintenance estimate did not include documentation of the techniques used for each cost element. The cost estimate includes a sensitivity analysis that identifies a range of possible costs based on varying major assumptions, parameters, and data inputs. Minimally met. The construction estimate identified and examined key cost drivers but did not include a formal sensitivity analysis. The operations and maintenance estimate did not include a sensitivity analysis. A risk and uncertainty analysis was conducted that quantified the imperfectly understood risks and identified the effects of changing key cost driver assumptions and factors. Partially met. The construction estimate included risk and uncertainty analysis for cost and schedule and quantified the cost of these risks as management reserve and contingency reserve. The estimate did not document how correlation of cost elements was accounted for in the risk and uncertainty analysis. The operations and maintenance estimate did not include risk and uncertainty analysis. Major cost elements were cross-checked to see whether results were similar. Partially met. The construction estimate included cross-checks with metrics that benchmark the TWF to similar nuclear projects. The operations and maintenance estimate did not include documentation of cross-checks. An independent cost estimate (ICE) was conducted by a group outside the acquiring organization to determine whether other estimating methods produce similar results. Partially met. An ICE for the project’s construction phase was performed by DOE’s Office of Acquisition and Project Management. The ICE appears to have been based on a similar technical baseline to the program office estimate. However, the program estimate was 13 percent higher than the ICE. NNSA did not document how it reconciled the two estimates. The ICE did not cover the operations and maintenance costs of the facility. The ratings we used in this analysis are as follows: “Not met” means the cost estimate provided no evidence that satisfies the best practice. “Minimally met” means the cost estimate provided evidence that satisfies a small portion of the best practice. “Partially met” means the cost estimate provided evidence that satisfies about half of the best practice. “Substantially met” means the cost estimate provided evidence that satisfies a large portion of the best practice. “Fully met” means the cost estimate provided complete evidence that satisfies the entire best practice. In addition to the individual named above, Diane LoFaro, Assistant Director; Mark Braza; Richard P. Burkard; Brian M. Friedman; Abishek Krupanand; Eli Lewine; Cynthia Norris; Katrina Pekar-Carpenter; and Karen Richey made key contributions to this report.

Nuclear weapons activities at LANL have generated large quantities of TRU waste that must be disposed of properly. To address a 2005 cleanup agreement with the state of New Mexico requiring DOE to close LANL's TRU waste site, NNSA is to oversee two TRU waste projects. The first is to remove the waste stored at LANL and ship it to WIPP for permanent disposal. The second is to construct a facility—the TWF—to provide new capabilities for managing newly generated TRU waste at LANL. NNSA has developed cost estimates for both projects. GAO was asked to review cost estimates for the TRU waste projects at LANL. This report examines (1) the extent to which NNSA's TRU waste removal project at LANL has met its cost estimates and (2) the extent to which NNSA's cost estimate for the TWF met best practices for a reliable estimate. GAO reviewed spending data for the TRU waste removal project for fiscal years 2006 through 2014 and the cost estimates for both projects, compared the cost estimate for the TWF with best practices, and interviewed agency officials. The National Nuclear Security Administration's (NNSA) project to remove transuranic (TRU) waste—primarily discarded equipment and soils contaminated with certain radioactive material—at Los Alamos National Laboratory (LANL) did not meet its cost estimates. At the end of fiscal year 2014, NNSA had spent about $931 million on the project, exceeding its 2006 estimate of $729 million by $202 million. Under current plans, the project is also expected to exceed its 2009 estimate. NNSA did not meet its cost estimates, in part, because they were based on aggressive funding assumptions designed to meet the completion dates agreed to in a 2005 cleanup agreement, which the Department of Energy (DOE) did not fully fund. At the time of GAO's review, NNSA was developing a new project completion cost estimate of about $1.6 billion, with completion projected for October 2022. NNSA had not revised the project's cost estimate since 2009 because the agency was reluctant to approve an estimate with a completion date that conflicted with the 2005 cleanup agreement. However, according to an NNSA official, NNSA's new estimate may not reflect current conditions—partly because of uncertainty created by funding and the indefinite suspension of shipments of TRU waste to the permanent repository at DOE's Waste Isolation Pilot Plant (WIPP) after a radioactive release closed WIPP in February 2014. By revising the estimate to include the current understanding of project conditions, including the uncertainty at WIPP, NNSA program managers can, for example, more accurately identify cost overruns. NNSA's cost estimate for the TRU Waste Facility (TWF), which consisted of separate cost estimates for completing construction and for operations and maintenance, partially reflected each of the four characteristics of a reliable estimate (comprehensive, well-documented, accurate, and credible) as established by best practices. For example, NNSA's estimate was partially well-documented by clearly documenting the data sources and methodology used to develop the construction estimate. However, NNSA did not sufficiently document the approach used to develop the operations and maintenance estimate, which represented about 74 percent of the TWF's life-cycle costs, because DOE's project management order does not require these costs to be documented when a project is approved to request funding from Congress for construction. As a result, GAO could not determine whether the cost-estimating approach was appropriate. In addition, NNSA's estimate was partially credible because NNSA completed an independent cost estimate (ICE) that provided an unbiased cross-check of the construction estimate consistent with best practices, but it did not include the operations and maintenance costs in the ICE because it was not required by DOE's project management order. Moreover, NNSA did not conduct a sensitivity analysis to quantify variations in the TWF's cost estimates due to changes in key assumptions because it was not required by DOE, which also affected the estimate's credibility. Doing a sensitivity analysis increases the chance that decisions for the TWF will focus on the elements that have the greatest effect on cost, according to best practices. Updating the TWF's cost estimate to include all life-cycle costs and needed analyses, would provide NNSA more reliable information for better managing the TWF as it prepares for the start of operations, which NNSA expects could be as early as April 2016. GAO recommends that DOE revise the cost estimate for the TRU waste removal project to reflect the current understanding of project conditions and update the TWF's cost estimate to allow better management of the project's life-cycle costs going forward. DOE generally agreed with GAO's recommendations.

On August 10, 1993, the Congress enacted the Omnibus Budget Reconciliation Act of 1993 (OBRA 1993, P.L. 103-66), which established the EZ/EC program’s eligibility criteria, designation procedures, and benefits. The act specified that an area could not be selected for the program unless it (1) met specific criteria for characteristics such as geographic size and poverty rate and (2) prepared a strategic plan for implementing the program. The act also authorized the Secretary of Housing and Urban Development and the Secretary of Agriculture to designate the EZs and ECs in urban and rural areas, respectively; limited the number of designations that could be made; set the length of the designation at 10 years; required that nominations be made jointly by the local and state governments; and authorized the Secretaries to prescribe any regulations needed to carry out the program. The act also amended title XX of the Social Security Act to authorize the special use of SSBG funds for the EZ program. The use of SSBG funds was expanded to cover a range of economic and social development activities. Like other SSBG funds, the funds allotted for the EZ program are granted by HHS to the state, which is fiscally responsible for the funds. HHS’ regulations covering block grants (45 C.F.R. part 96) provide maximum fiscal and administrative discretion to the states and place full reliance on state law and procedures. HHS encouraged the states to carry out their EZ funding responsibilities with as few restrictions as possible under the law. After the state grants the funds to the EZ or the city, the EZ can draw down the funds through the state for specific projects over the 10-year life of the program. In January 1994, the Clinton administration announced the nominating procedure and required that nominations be received by June 30, 1994. After collaborating with other federal agencies including HHS, HUD and the U.S. Department of Agriculture (USDA) jointly issued an application guiderequiring each applicant to submit its nomination along with a strategic plan that had been developed with input from community stakeholders, such as residents, businesses, financial institutions, service providers, and state and local governments. The plan was to describe the community’s overall vision for revitalization, link this vision to the program’s four key principles, identify other governmental and private resources that would be committed to this program, and describe potential barriers to the successful implementation of the plan. HUD also published guidelines for developing strategic plans, conducted technical assistance workshops around the country, provided advice and technical assistance through federal employees, and contracted for technical assistance in fields such as planning and community development. The federal government received over 500 nominations, including 290 from urban communities. The nominations were reviewed by the EZ/EC task force, which consisted of federal employees detailed from many federal agencies. This task force reported on the urban applications to a review panel that consisted of three senior officials from HUD’s Office of Community Planning and Development. This panel recommended potential urban designees to the Secretary of Housing and Urban Development. The Secretary then sent his preliminary selections to the Community Empowerment Board for review. This board is chaired by the Vice President and its members include the heads of cabinet-level and other federal agencies. The board was established to offer a single point of federal coordination for communities and to facilitate one-stop access to federal resources. On December 21, 1994, the Secretaries of Housing and Agriculture designated 104 EZs and ECs—6 urban EZs, 3 rural EZs, 65 urban ECs, and 30 rural ECs. On the same day, HHS awarded the first half of the SSBG to the EZs and ECs, as provided for in the authorizing legislation.Subsequently, HHS provided information to the designees to clarify the uses of and controls on the EZ/EC SSBG funds and advised the EZs and ECs about other HHS grants that have been awarded or are available to the EZs. All of the designated communities will receive federal assistance; however, as established by OBRA 1993 and HUD’s implementing regulation (24 C.F.R. part 597), the EZs are eligible for more assistance than the ECs. Each urban EZ was allocated $100 million, and each rural EZ was allocated $40 million, in EZ/EC SSBG funds for use over the 10-year life of the program. In addition, up to $20 million in state and local bonds—whose proceeds were to be used to provide facilities and land for businesses in the zone—would be tax-exempt. Furthermore, businesses located in the EZ would be eligible for (1) tax credits on wages paid to employees who live in the EZ and (2) increased deductions for depreciation. Each urban and rural EC was allocated just under $3 million and qualified only for the tax-exempt bonds. The federal government also made a commitment to all of the EZs and ECs to (1) give them special consideration in competitions for funds from other federal programs, (2) work cooperatively with them in overcoming regulatory impediments, and (3) allow them to make more flexible use of existing federal funds. After making the designations, HUD issued implementation guidelinesdescribing the program as one in which (1) solutions to community problems are to originate from the neighborhood up rather than from Washington down and (2) progress is to be based on performance benchmarks established by the EZs, not on the amount of federal money spent. The benchmarks are to measure the results of the activities described in each EZ’s strategic plan. These benchmarks became part of the agreement that was signed by HUD and state and local government officials for each zone. The EZ/EC task force’s members were available to assist the EZs in preparing their benchmarks. HUD also uses contractors, which HUD refers to as generalists, to provide day-to-day assistance to the EZs. All six of the urban EZs have met the criteria defined in OBRA 1993,developed a strategic plan, signed an agreement with HUD and the state for implementing the program, signed an agreement with the state for obtaining the EZ/EC SSBG funds, drafted performance benchmarks, and set up a governance structure. In addition, all of the EZs have included public housing officials and public housing residents in planning and implementing the program. However, the EZs differ in their demographic characteristics, organizational structure, and plans for using their EZ/EC SSBG funds. Also, a few public housing officials indicated that their EZ had not done enough to involve the public housing authority or public housing residents. The EZs differ in their geographic and demographic characteristics, reflecting the selection criteria. For example, Detroit’s EZ covers about 18 square miles and is over four times as large as the Philadelphia-Camden EZ. In Baltimore, Chicago, and Philadelphia-Camden, the areas included in the EZ are not contiguous, while in Atlanta, Detroit, and New York, they are contiguous. Furthermore, the Philadelphia-Camden EZ is unique because it is located in two cities and states. The population of the EZs ranges from about 50,000 in Atlanta and Philadelphia-Camden to nearly 200,000 in Chicago and New York. The poverty and unemployment rates also vary across the EZs. The overall poverty rate for the Atlanta EZ was the highest, encompassing 55 percent of the residents, while the New York and Baltimore EZs reported an overall poverty rate of about 40 percent. The poverty rates for all of the EZs are high compared with the national poverty rate of about 14 percent. Similarly, the unemployment rates in the EZs ranged from 15 percent in Baltimore to 29 percent in Detroit, while the national rate was about 6 percent. (See table 1 for details on each EZ.) Local governments have chosen different approaches to implementing the EZ program. Atlanta, Baltimore, Detroit, New York, and Camden have each established a nonprofit corporation to administer the program, while Chicago and Philadelphia are operating through the city government. At the state level, the types of agencies involved and the requirements for drawing down the EZ/EC SSBG funds differ. HHS awarded the funds to the state agency that managed the regular SSBG program unless the state asked HHS to transfer the responsibility to a state agency that deals primarily with economic development. Consequently, the funds for Atlanta and New York pass through their state’s economic development agency, while the funds for the other EZs pass through the state agency that manages the regular SSBG program. Some states, as the entities with fiscal responsibility for the EZ/EC SSBG funds, identified additional requirements that the EZ must meet before it can draw down funds. For example, one state requires the EZ to follow the guidelines established in the Office of Management and Budget’s Circular A-87, Cost Principles for State, Local, and Indian Tribal Governments. The federal government does not require the recipients of SSBG funds to follow these guidelines. Finally, each EZ has planned diverse activities to meet its city’s unique needs. All of the EZs have planned activities to increase the number of jobs in the EZ, improve the EZ’s infrastructure, and provide better support to families. However, the specific activities vary, reflecting decisions made within each EZ. According to HUD, the EZs have obligated over $170 million. However, the definition of obligations differs across the EZs. For example, one EZ defines obligations as the amount of money awarded under contracts. Another EZ defines obligations as the total value of the projects that have been approved by the city council, only a small part of which has been awarded under contracts. As of October 31, 1996, the six EZs had drawn down about $5 million from the EZ/EC SSBG funds for administrative costs, as well as for specific activities in the EZs. Administrative costs covered salaries, office equipment, supplies, audits, and consultants’ fees. Individual EZs had also provided funds for activities such as initiating a project to reduce alcohol- and drug-related violence among high-risk youth, acquiring sites for a supermarket and retail stores, and creating an industrial ecological park. The EZs have used very little of the federal funding available to them because they have been involved in other activities, such as setting up their governance structures, establishing procedures for obtaining funds, and encouraging businesses to invest in them. (See app. I for details on each EZ’s governance structure, use of EZ/EC SSBG funds, and planned activities.) The EZ program requires the participation of various segments of the community, including the residents. Although the program does not explicitly require the involvement of public housing authority (PHA) officials or public housing residents, all of the EZs contain public housing units. We interviewed PHA officials in the six urban EZs to obtain information on the number of public housing residents in each EZ and the participation of PHAs and public housing residents in their EZ’s activities. According to the information we obtained, about 50 percent of the residents in the Atlanta EZ live in public housing, followed by 42 percent in New York, 18 percent in Baltimore, 15 percent in Chicago, 10 percent in Philadelphia-Camden, and 6 percent in Detroit. PHA officials in all six EZs said that they and the residents they serve participated in their EZ’s activities. Initially, PHAs and residents helped to develop the EZ’s applications and benchmarks, organized community meetings, and served on housing committees and local task forces. More recently, PHA officials and residents have served on governance boards and housing councils and have been active in human service and job training programs. Officials from three of the EZs, including one with a large number of public housing residents, told us that their EZ had not done enough to include either the PHA or the public housing residents in the EZ’s activities after designation. These officials suggested that greater involvement is needed because a large proportion of their city’s public housing is in or near the EZ. The officials also reported that their involvement in the EZ program could maximize their city’s use of the federal resources allocated to public housing and the EZ’s activities. For example, one official saw an opportunity to coordinate HUD’s Hope VI program with the EZ’s activities. Overall, PHA officials are optimistic about their involvement in the EZ program and believe that it will continue or increase. In addition to serving on various boards and councils, the PHAs expect to expand their role in home ownership and housing rehabilitation initiatives and in job creation and training programs. We interviewed participants in the EZ program across all six EZs—including EZ directors and governance board members, state officials involved in drawing down the EZ/EC SSBG funds, contractors who provide day-to-day assistance to the EZs, and HUD and HHS employees—and asked them to identify what had and had not gone well in planning and implementing the program. To obtain reactions to all of the factors that these individuals identified, we listed the factors in a questionnaire and mailed it to 32 program participants, including those we had already interviewed. The questionnaire asked the survey recipients to indicate the extent to which each factor had helped or hindered the program’s implementation. The survey also provided space for the respondents to give examples or suggest solutions. While the survey respondents’ views cannot be generalized to the entire EZ/EC program, they may be useful to HUD as a starting point for communicating with the EZs to improve the current program. These views also can form a basis for framing future initiatives with goals similar to those of the EZ program. (App. II lists all of the factors identified in the telephone interviews and in the survey. App. III contains a more detailed discussion of our methodology.) In the 27 surveys that were returned to us, five factors were frequently identified as having helped, and six factors were frequently identified as having constrained, the program’s implementation. Five factors were identified by more than half of the survey respondents as having helped them plan and implement the EZ program: community representation on the EZ governance boards, assistance from HUD’s contractors (called generalists), enhanced communication among stakeholders, support from the city’s mayor, and support from White House and cabinet-level officials. Having community representatives on the governance boards created a shared responsibility for the program’s success and helped to break down barriers between the residents and other segments of the community that were represented on the board, such as the local government and businesses. The generalists hired by HUD to work on a daily basis with the zones have been accessible and have provided important assistance to the zones. Respondents said that the generalists’ assistance included providing information, negotiating with elected and public officials, forging relationships with the private sector, and arranging meetings. One respondent said that the generalists are seen as people who can cut through bureaucratic red tape and get things done in the community. The EZ program has brought community stakeholders together. People from the neighborhoods, the private sector, and the city and state governments who did not previously interact are now discussing community revitalization. These stakeholders have created partnerships that respondents believe have improved relationships between government officials and community leaders; stimulated revitalization throughout the city, not just within the EZ; increased coordination across economic and human development activities, as well as among the public, private, and nonprofit sectors, thereby expanding the availability of funds; and created a basis for ensuring sustainable results. Strong mayoral support, shown by activities such as reviewing proposed benchmarks and providing needed resources, produced benefits that included obtaining a high level of involvement from the private sector, resolving issues of distrust, effectively conveying the city’s concerns to federal officials, helping attract economic development to the EZ, and increasing the coordination with city departments whose assistance was critical to the program’s success. The participation of White House and cabinet-level officials enhanced the program’s credibility at the community level. For example, one city official said that the involvement of White House officials gave the community hope that their issues would be heard and that federal regulations would be eliminated. Six factors were frequently identified by survey respondents as having constrained their efforts to plan and implement the EZ program: difficulty in selecting an appropriate governance board structure, the additional layer of bureaucracy created by the state government’s involvement, preexisting relationships among EZ stakeholders, pressure for quick results from the media, the lack of federal funding for initial administrative activities, and pressure for quick results from the public and private sectors. Several respondents noted that selecting the governance structure and deciding on the size and composition of the executive board was time-consuming, taking in at least one case more than a year to resolve. Respondents suggested that HUD could have (1) provided examples of governance structures in its application guidelines and (2) set a time limit for the EZs to adopt a governance structure. The state government’s involvement has created an unnecessary layer of bureaucracy, according to some respondents. One respondent said that the state government is requiring the EZ to obtain the state’s approval for funding requests. In the respondent’s view, this requirement exceeds the state’s responsibility to provide fiscal oversight. Some federal and city government respondents said that the approvals at the state level add an unnecessary layer of bureaucracy. At least one state respondent agreed but said the states have no alternative because the SSBG program’s regulations, which were not revised for the EZ program, require the states to oversee the use of EZ/EC SSBG funds. He added that even without the fiscal oversight role, state agency officials should be involved in activities such as reviewing the EZ’s plans or ensuring consistency with other programs. Suggestions for preventing cumbersome state reviews included (1) clarifying early in the program that the state’s role is minimal, (2) eliminating the state’s fiscal responsibility for the EZ/EC SSBG funds allocated to the EZ program, and (3) financing the program with funds that, unlike SSBG funds, do not have to flow through the state. In some communities, a history of antagonism and ineffective communication among state and local government representatives, community leaders, residents, and private-sector representatives has impeded consensus-building and teamwork. One respondent suggested that increased team-building efforts—such as off-site team-building training for governance board members—and more effective communication programs could help. Pressure for quick results from the media has created unrealistic expectations about how quickly progress can be achieved. Respondents wrote that some media representatives may not understand the program and that media attention stemmed from the way the federal government initially described the program. Suggested remedies included having (1) local public information officers clarify the program’s goals for the media and (2) the federal government stress that this is a 10-year program. Immediately after designation, some EZs did not have the financial and/or human resources that they needed to perform the program’s initial administrative activities. The federal government initially told the EZs that they could not access the EZ/EC SSBG funds until agreements were signed by the federal, state, and city governments. The earliest of these was signed in July 1995. In some cases, city and state governments provided funding and assigned staff to the EZ; however, one respondent noted that having work performed by government employees created doubts about who controlled the program—the government or the community. Respondents’ suggestions for reducing confusion included (1) making an EZ’s designation contingent on a commitment by the city and/or state government to provide funds for hiring administrative staff or (2) providing a portion of the EZ/EC SSBG funds to the EZ for administration immediately after designation. Pressure for quick results from officials at the federal, state, and/or local levels, as well as from the public, was also identified as an impediment. One respondent suggested that before money could be spent on revitalizing the community, the EZs had to develop team-building and decision-making processes. These processes were not in place when the program began, took time to develop, and were necessary to ensure sustainable results. Some respondents noted that, in some instances, rushing the planning and implementation steps resulted in mistakes that took time to correct. Furthermore, the community-based approach used in this program involves a larger group of people than does more traditional decision-making; thus, reaching decisions takes more time. Respondents suggested that the federal government should (1) measure short-term success by the development of capacity in the zones, such as the establishment of participatory processes, not by the amount of money spent and (2) emphasize and encourage the replication of best practices and the transfer of technology among the zones. From the beginning, the Congress and HUD have made evaluation plans an integral part of the EZ program. OBRA 1993 required that each EZ applicant identify in its strategic plan the baselines, methods, and benchmarks for measuring the success of its plan and vision. In its application guidelines, HUD amplified the act’s requirements by asking each urban applicant to submit a strategic plan based on four principles: (1) creating economic opportunity for the EZ’s residents, (2) creating sustainable community development, (3) building broad participation among community-based partners, and (4) describing a strategic vision for change in the community. These guidelines also stated that the EZs’ performance would be tracked in order to, among other things, “measure the impact of the EZ/EC program so that we can learn what works.” According to HUD, these four principles serve as the overall goals of the program. Furthermore, HUD’s implementation guidelines required each EZ to measure the results of its plan by defining benchmarks for each activity in the plan. HUD intended to track performance by (1) requiring the EZs to report periodically to HUD on their progress in accomplishing the benchmarks established in their strategic plans and (2) commissioning third-party evaluations of the program. HUD stated that information from the progress reports that the EZs prepare would provide the raw material for annual status reports to HUD and long-term evaluation reports. HUD is reviewing information on the progress made in each EZ and EC to decide whether to continue each community’s designation as an EZ or an EC. All six of the urban EZs prepared strategic plans that include a section on evaluation. They also prepared benchmarks that comply with HUD’s guidelines and describe activities that they have planned to implement the program. In most cases, the benchmarks indicate how much work, often referred to as an output, will be accomplished relative to a baseline. For example, a benchmark for one EZ is establishing a single point of access to substance abuse treatment for 1,800 EZ residents. The baseline associated with this benchmark is that 5,400 EZ residents lack access to substance abuse treatment. A benchmark for another EZ states that the EZ will assist businesses and entrepreneurs in gaining access to capital resources and technical assistance through the establishment of a single facility called a one-stop capital shop. The associated baseline is that there is currently no one-stop capital shop to promote business activity. The performance measures for this benchmark include the amount of money provided in commercial lending, the number of loans made, the number of consultations provided, and the number of people trained. As we have previously reported, the Congress, the executive branch, and the public are beginning to hold agencies accountable for the outcomes of their programs—the results as measured by the differences that the programs make, for example, in participants’ lives. Specifically, the Government Performance and Results Act requires federal agencies to clearly define their missions and to establish long-term strategic goals, as well as annual goals linked to the strategic ones. Our previous report found that results-oriented organizations follow three steps: (1) define the mission and desired outcomes, (2) measure performance to reinforce the relationship of daily activities to the long-term mission and outcomes, and (3) use information on performance as a basis for decision-making. In the EZ program, HUD has followed this format to some extent by (1) defining the four key principles, which serve as missions and goals for the EZs; (2) requiring baselines and performance measures for benchmarks in each EZ to help measure the EZ’s progress in achieving specific benchmarks; and (3) developing procedures for including performance measures in HUD’s decision-making process. However, the measures being used generally describe the amount of work that will be produced (outputs) rather than the results that are anticipated (outcomes). For example, for the second benchmark cited above (establishing a one-stop capital shop), the EZ has not indicated how the outputs (the amount of money provided in commercial lending, the number of loans made, the number of consultations provided, and the number of people trained) will help to achieve the desired outcome (creating economic opportunity, the relevant key principle). To link the outputs to the outcome, the EZ could measure the extent to which accomplishing the benchmark increases the number of businesses located in the zone. Without identifying and measuring desired outcomes, HUD and the EZs may have difficulty determining how much progress the EZs are making toward accomplishing the program’s overall mission. HUD officials involved in the EZ/EC program told us that HUD has been working with the EZs to ensure that they can measure their accomplishment of the individual benchmarks. The benchmarks are revised, as needed, to reflect changes in the community and to include activities that will be performed after the first 2 years of the program. HUD officials agree that the performance measures used in the EZ program are output-oriented and believe that these are appropriate in the short term. They believe that the desired outcomes of the EZ program are subject to actions that cannot be controlled by the entities involved in managing this program. In addition, the impact of the EZ program on desired outcomes cannot be isolated from the impact of other events. Consequently, HUD believes that defining outcomes for the EZ program is not feasible. Concerns about the feasibility of establishing measurable outcomes for programs are common among agencies facing this difficult task. However, because HUD and the EZs have made steady progress in establishing an output-oriented process for evaluating performance in the EZ program, they could build on their efforts to incorporate measures that are more outcome-oriented. Specifically, HUD and the EZs could describe measurable outcomes for the program’s key principles and indicate how the outputs anticipated from one or more benchmarks will help to achieve those outcomes. HUD has taken commendable steps toward establishing results-oriented measures for the EZ program. Among other things, the EZs have developed benchmarks that describe planned activities, as well as the baselines and time frames against which progress toward accomplishing individual benchmarks can be measured. However, HUD and the EZs are not yet measuring performance in a way that allows them to assess how much progress is being made in satisfying the program’s four key principles because they have not yet (1) described measurable outcomes for the program’s key principles or (2) indicated how the outputs anticipated from one or more benchmarks will help to achieve those outcomes. Unless they can measure the EZs’ progress in producing desired outcomes, HUD and the EZs may have difficulty identifying activities that should be duplicated at other locations. In addition, HUD and the EZs may not be able to describe the extent to which the program’s activities are helping to accomplish the program’s mission. We recommend that the Secretary of Housing and Urban Development work with the EZs to establish a process for incorporating measurable outcomes for the program’s principles into any future revisions of the EZs’ strategic plans and benchmarks. Among other things, this process should describe the outcomes anticipated from the EZs’ activities, indicate how the outcomes will be measured, and identify the benchmarks helping to achieve each outcome. We provided copies of a draft of this report for review and comment to HUD and HHS. These agencies’ written comments and our responses appear in appendixes IV and V. In commenting on the draft, HUD said that it found the report to be a useful and accurate description and analysis of the status of the EZ/EC initiative. HUD also indicated two primary areas of concern about the report. First, HUD expressed concern that we did not understand its process for measuring performance for the EZ/EC program. This process uses the benchmarks as the basis for measuring quantitative progress in implementing the EZs’ and ECs’ strategic plans. We agree that HUD has established procedures for measuring individual activities within each EZ. However, more could be done to describe the program’s anticipated outcomes and link individual activities to those outcomes. Such additional efforts should allow HUD and the EZs to better measure performance over the program’s 10-year life. We revised the information in the report to include HUD’s concerns and to clarify the actions that HUD and the EZs can take to strengthen their evaluation efforts. Second, HUD felt that the amounts of EZ/EC SSBG funds that have been drawn down, viewed in isolation, could be misleading as measures of the program’s spending. HUD provided a list of amounts obligated by each EZ and asked us to include those obligations in table 1. However, the definition of obligations differs across the EZs. For example, one EZ defines obligations as the amount of money awarded under contracts. Another EZ defines obligations as the total value of the projects that have been approved by the city council, only a small part of which has been awarded under contracts. We chose not to include obligations in the table because including them would invite inappropriate comparisons across the EZs. However, we revised the report to indicate that higher amounts have been obligated by the EZs. HUD also commented on the sections of the report in which we summarized the concerns of program participants about the role of the states, the lack of early administrative funding, and the pressure from the federal government and others for quick results. HUD asked us to revise these sections of the report because it felt that the concerns were inaccurate or did not recognize actions taken by HUD. As stated earlier in this letter and in the appendix on our scope and methodology, we did not independently analyze the factors identified by the respondents or examine their applicability across all of the EZs. Although we did add a sentence specifying when HHS obligated the EZ/EC SSBG funds, we believe that the report accurately and sufficiently covers participants’ concerns about the topics mentioned, and we did not make other changes suggested by HUD. Finally, HUD suggested editorial and technical changes, which we incorporated when appropriate. In its comments on the report, HHS said that it found the report very well done and thought provoking. HHS also noted that Philadelphia and Camden operate as two separate entities in many ways and thought that the report should treat the two parts of this EZ separately, rather than as one EZ. We agree that, for the most part, the two segments of this zone operate autonomously; however, we believe that discussing them together is appropriate because Philadelphia-Camden is a single, bistate zone. Furthermore, the administration of this EZ was not significantly different from that of the New York EZ, which has split its operations between two corporations covering distinct parts of the designated zone. HHS also thought that the report should include additional background information about HHS’ role in the program and raised a number of editorial and technical comments on the report. We revised the wording of the report, as appropriate, to address these comments. (See apps. IV and V for HUD’s and HHS’ comments on the report and our responses to those comments.) We also sent the detailed information on each EZ to the applicable EZ director for review and comment. We received minor technical and editorial corrections from all of the EZs and incorporated these into the report as appropriate. To assess the status of HUD’s implementation of the six urban EZs and to describe the Department’s plans for evaluating the initiative, we interviewed officials from HUD and HHS who were responsible for the EZ program and reviewed documents obtained from HUD, HHS, and the EZs. To understand the role played by public housing officials and residents in the EZ program, we surveyed representatives of the PHAs in the seven cities included in the six urban EZs using a structured telephone survey. We used interviews with EZ program participants and responses to a mailed questionnaire to identify factors that have helped or hindered efforts to carry out the EZ program. We performed our work at HUD’s Office of Community Planning and Development and HHS’ EZ/EC Support Team in Washington, D.C. We also retained Dr. Marilyn M. Rubin, an expert with extensive knowledge in economic development and evaluation, to advise us on all aspects of our work. We performed our work in accordance with generally accepted government auditing standards from November 1995 through October 1996. Appendix III contains details on our scope and methodology. As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days from the date of this letter. At that time, we will send copies to the Secretaries of Housing and Urban Development and Health and Human Services, as well as to the people who participated in our survey. Copies will be made available to others upon request. Please call me at (202) 512-7631 if you or your staff have any questions. Major contributors to this report are listed in appendix VI. The Atlanta Empowerment Zone Corporation was formed as a nonprofit organization to oversee the implementation of the strategic plan. It is headed by a president who answers to the executive board and includes staff to administer the programs and service contracts. The 17-member executive board comprises representatives of public agencies, service providers, the private sector, and the community, as well as 6 residents. The board is the final decision-making body for the EZ. The 36-member Community Empowerment Advisory Board will carry advice from the community to the executive board. This board consists of 1 representative from each of the 30 neighborhoods in the zone, plus 6 representatives from 39 communities adjacent to the zone. The EZ/EC SSBG funds pass through the Georgia Department of Community Affairs to the city of Atlanta for use by the EZ. As of October 31, 1996, the EZ had obtained $1,535,605 for administrative expenses that include costs for EZ personnel, consultants, travel, office equipment and supplies, and printing. A portion of these funds was used to support the community advisory board’s office staff. Atlanta envisions the EZ as an “urban village” working cooperatively to improve the quality of life in its neighborhoods and emphasizing development that is economically and ecologically sound. Atlanta has organized its planned activities into five categories: expanding employment and economic investment by increasing jobs, training for jobs, and transportation to jobs; attracting businesses to the EZ; and increasing sources of funds for businesses; creating safe and livable communities by increasing public safety; improving streets, sidewalks, lighting, and parks; and promoting ecologically sustainable communities; lifting youth and families out of poverty by confronting and reducing drug and substance abuse, increasing learning opportunities to reduce the number of high school dropouts, creating food cooperatives and community gardens to feed the hungry, and providing comprehensive human development programs; providing adequate housing for all by increasing access to credit, improving the affordability and availability of housing, increasing home ownership, and meeting the needs of the homeless; and providing governance by creating the corporation to implement the EZ’s strategic plan. The Empower Baltimore Management Corporation, which has a president and chief executive officer, was formed as a public, nonprofit organization to oversee the implementation of the strategic plan. The corporation’s board of directors, which is headed by a chairman, consists of 30 members including community leaders, city agency heads, and representatives of the business community, foundations, and universities. Six village centers have been created as public, nonprofit organizations that will help create sustainable communities. The EZ/EC SSBG funds pass through the Maryland Department of Human Resources to the Empower Baltimore Management Corporation for use in the EZ. As of October 31, 1996, Baltimore had obtained $2,095,500 of its EZ funds for administrative costs and grants. The administrative costs have included costs for salaries, printing, and office supplies and equipment. The grants have been awarded to fund a business empowerment center; the Fairfield Ecological Industrial Park; and the village centers. Baltimore’s goal is to connect the EZ’s workforce with the area’s mainstream economy, rebuild all basic social and neighborhood systems simultaneously through comprehensive approaches, and solve problems and advance EZ initiatives through a highly mobilized citizen-resident force on a block-by-block basis. Baltimore has organized its planned activities into eight categories: community mobilization, which includes creating village centers, providing technical assistance to communities, and using information and community technology at the community level; community development, which includes developing a land-use plan and improving the area’s overall environment and quality of life; public safety, which includes enhancing community policing, addressing substance abuse enforcement, and designing safe neighborhoods; housing, which includes facilitating home ownership, increasing the availability of affordable rental housing, and improving substandard housing; health and family development, which includes linking human services to education, training, and literacy, which includes providing a full range of training opportunities for EZ residents, such as training in classrooms and in local career centers; economic development, which includes strengthening and expanding existing businesses, attracting new businesses, and developing entrepreneurs within the EZ; and evaluating and monitoring, which includes operating the Empower Baltimore Management Corporation, as well as evaluating and monitoring the program. The EZ/EC Coordinating Council is the governing body of the Chicago EZ. Its 39 members include representatives from businesses and communities in the zone and officials from the city, county, and state governments. The council also includes representatives from three city-designated neighborhoods that the city wants to benefit from the zone’s activities. The council’s responsibilities include developing zonewide policies, identifying support resources, and reviewing and recommending the approval of requests for funding projects aimed at implementing the strategic plan. Four committees established by the EZ/EC Coordinating Council manage the EZ. The Executive Committee calls meetings and establishes rules and procedures for the coordinating council. The Committee on Policy and Planning advises the council on the allocation of EZ funds, while the Committee on Finance monitors the receipt, use, and distribution of funds. Finally, the Committee on Community and Business Outreach promotes the benefits and services of the EZ to businesses and residents. The EZ/EC SSBG funds pass through the Illinois Division of Family Support Services to the city of Chicago for use by the EZ. As of October 31, 1996, the EZ had drawn down $279,000 to initiate projects that would rehabilitate office space for future use by businesses, establish a mechanism to link EZ residents with employers, promote home ownership, develop public schools into community learning centers, provide business training to 30 EZ residents with business potential, and create a partnership with a local college to prepare students for the General Educational Development tests. These projects are the first of 86 projects that were approved by the EZ in September 1996. Funding for these projects totals about $45 million, including $41 million in federal funds. Chicago’s two overall goals for the EZ are alleviating poverty and changing the way that the federal, state, county, and city governments interact with the EZ’s residents. Chicago’s strategic plan includes seven initiatives aimed at achieving these goals: building human and organizational capacity by developing programs that expand traditional job training projects to include life skills, job readiness, and apprenticeship and mentoring; linking health and human services by establishing a wellness system that encourages a healthy workforce through health screening, assessment, and medical referrals; improving public safety by increasing community security and providing opportunities for local residents to become more involved in making their environment safe; achieving economic empowerment by increasing investment in and by the community; developing affordable and accessible housing by expanding home ownership opportunities within the community and increasing the number of housing units for the elderly and persons with disabilities; enhancing youth futures by establishing youth training programs and building on cultural diversity by promoting tourism and increasing and fostering cultural sensitivity. The Empowerment Zone Development Corporation was formed as a nonprofit corporation by state and local legislation to oversee the implementation of the strategic plan. An executive director, who is hired by the board of directors, heads the corporation and is assisted by other staff. The board of directors is composed of 50 representatives of all sectors in the community. Sixty percent of the board’s members are community-based, including representatives of community development corporations, businesses, neighborhood councils, and places of worship. The remaining 40 percent represent the larger community, which includes government, corporations, banks, and foundations. An executive committee, which consists of 25 members selected from the board of directors, conducts the corporation’s business. Neighborhood review panels will provide a way for the EZ’s neighborhoods to convey new ideas or suggest revisions to the executive committee. The panels will consist of residential and business representatives from all three neighborhoods in the EZ. The EZ/EC SSBG funds pass through the Michigan Family Independence Agency to the city of Detroit, which disburses funds to agencies implementing the programs and projects in the approved strategic plan under contracts approved by the Detroit City Council. As of October 31, 1996, Detroit had drawn down $54,327 to begin a school-based program designed to reduce alcohol- and drug-related violence. In addition, Detroit has contracted with 18 agencies to implement projects totaling $29.4 million in EZ/EC SSBG funds. Detroit envisions healthy neighborhoods, strong families, and economic opportunities that provide well-paying jobs. Detroit has organized its planned activities into the following three categories: creating economic opportunity by improving businesses’ access to capital, attracting new businesses to the EZ, linking residents to jobs, and increasing international trade and tourism; sustaining families by improving public safety; building on existing programs to support productive, stable families; improving the quality of learning; and integrating technology into training and educational programs; and upgrading neighborhoods by preserving and developing affordable housing, creating housing alternatives for the homeless, making transportation more accessible, improving vacant land, increasing the reuse of contaminated land, and improving recreational facilities and programs. The New York Empowerment Zone Corporation is a local-city-state public benefit corporation. The corporation has a seven-member board composed of representatives from Harlem, the South Bronx, New York City, and New York State. The Bronx Overall Economic Development Corporation and the Upper Manhattan Empowerment Zone Development Corporation are nonprofit organizations that plan and implement zone activities in their respective parts of the EZ. Their functions include directing the execution of contracts with service providers, negotiating contracts, and awarding contracts. The EZ/EC SSBG funds pass through the state’s Empire State Development Corporation to the New York Empowerment Zone Corporation for use by the EZ. As of October 31, 1996, the EZ had obtained $511,202 of its funds. New York envisions revitalizing the economic, social, and physical infrastructure of the EZ’s neighborhoods. New York has organized its planned activities into five categories: creating economic opportunities by enhancing the small business base in neighborhoods, fortifying community-based organizations, providing comprehensive educational and job training programs, and aligning the EZ’s neighborhoods and residents with economic opportunities in the city and region; preparing children and youth for a productive future by expanding and upgrading early childhood development and day care programs; supporting families by ensuring the availability and quality of, and increasing the funding for, a variety of support services, such as primary health care programs and substance abuse prevention and treatment programs; restoring and maintaining the EZ’s infrastructure—its housing and open spaces—and making them safe; encouraging community involvement by increasing ways that residents can participate in decisions affecting their community, creating neighborhood planning centers, and connecting the EZ to the information superhighway through a communitywide network. The Philadelphia-Camden Bi-State Governance Board will provide oversight and guidance for regional aspects of the strategic plan. The Philadelphia segment of the EZ is divided into three neighborhoods—American Street, North Central Philadelphia and West Philadelphia—each of which has a Community Trust Board, whose members are responsible for overseeing the EZ’s activities in their neighborhood. The board members are elected by the zone’s residents or appointed by the mayor. The Philadelphia Empowerment Zone Office is part of the city of Philadelphia’s government. Its employees include people hired from the EZ’s communities and staff on loan from other city departments. The office organizes communities and works with the community trust boards and neighborhood committees to implement the benchmarks in the neighborhoods. The Camden segment of the EZ has set up the Camden Empowerment Zone Corporation. Its board consists of 35 representatives, at least 12 of whom are residents elected to sit on the board. The remaining members are appointed by the mayor. The mayor and the mayor’s department heads sit on the board as ex-officio members. The EZ/EC SSBG funds pass through the Pennsylvania Department of Public Welfare to the city of Philadelphia for use by the EZ and through the New Jersey Department of Human Resources to the corporation. The EZ entities have agreed that the $100 million will be split between the two cities, with Philadelphia receiving $79 million and Camden receiving $21 million. As of October 31, 1996, the EZ had obtained $570,943 to buy sites for a supermarket and retail stores in the Philadelphia part of the EZ, which the EZ hopes will create jobs for residents. The Philadelphia-Camden vision for the EZ is based on revitalizing the economic, social, and physical infrastructure of the zone’s neighborhoods. The EZ has organized its activities into five categories: producing economic growth by creating 10,000 jobs over the life of the creating affordable housing by establishing a housing trust fund to support the development of at least 1,000 affordable dwellings for home owners and renters; supporting families by establishing a center to provide one-stop-shopping for family support services. improving public health by expanding medical services to the community through activities such as providing additional intake facilities for drug and alcohol programs and people with AIDS. increasing safety by organizing and expanding community policing, developing 40 new Town Watch programs, and establishing community safety centers. To assess the status of HUD’s implementation of the six urban EZs and to describe the Department’s plans for evaluating the initiative, we interviewed officials from HUD and HHS who were responsible for the EZ program. We also reviewed HUD’s and HHS’ application and implementation guidance and policy memorandums, evaluation plans, and other relevant documents. We interviewed the EZ directors in all six urban EZs and visited the Atlanta, Baltimore, and Philadelphia-Camden EZs. We also reviewed the EZs’ strategic plans, benchmarks, status reports, and funding documents. We did not evaluate whether the EZ program will meet its objectives. We also did not examine the use of tax incentives or of funds other than the EZ/EC SSBG funds. To understand the role played by public housing officials and residents in the EZ program, we surveyed representatives of the PHAs in the seven cities included in the six urban EZs. Using a structured telephone survey, we asked the PHA representatives (1) to what extent, if any, the PHA and its residents were involved in the EZ’s activities before and after their city was designated as a federal EZ and (2) whether they felt their level of involvement was appropriate. We did not evaluate the adequacy of the PHA officials’ or the public housing residents’ participation in the EZ program. To identify factors that have helped or hindered efforts to carry out the EZ program, we interviewed 28 participants who represented all six EZs and included EZ directors, governance board members, state officials involved in drawing down the EZ/EC SSBG funds, generalists hired by HUD, and HUD and HHS employees. We selected these participants from a list provided by HUD. To obtain more consistent reactions to the factors that these individuals identified, we listed all of the factors in a questionnaire and mailed 34 questionnaires to 32 participants, including those we had already interviewed. One HUD generalist received three questionnaires—one for each of the cities with which he works. We did not independently analyze the factors identified by the respondents or examine their applicability across all of the EZs. The questionnaire asked the survey recipients to indicate the extent to which each factor had helped or hindered the program’s implementation. We also asked the recipients to elaborate on the factors that most extensively helped or hindered implementation and to identify possible remedies for the impediments. Finally, we asked the recipients to pick the three factors that had helped efforts the most and the three factors that had hindered efforts the most. Two people told us they preferred not to return the questionnaire. One was a private-sector representative on an EZ board who was concerned that completing the questionnaire could be perceived as an endorsement of the program. The other was a state government official who said that he was not sufficiently involved in the program to be able to complete the questionnaire. We received 27 completed surveys from respondents representing all of the EZs, as well as the federal and state governments. In general, the factors discussed in this report were the ones that were most often identified by respondents as (1) having a great or very great impact on the program’s implementation and (2) having helped or hindered implementation more than other similar factors. The results of this survey cannot be generalized to the entire EZ/EC program. We performed our work at HUD’s Office of Community Planning and Development and HHS’ EZ/EC Support Team in Washington, D.C. We also retained Dr. Marilyn M. Rubin, an expert with extensive knowledge in economic development and evaluation, to advise us on all aspects of our work. We performed our work in accordance with generally accepted government auditing standards from November 1995 through October 1996. The following are GAO’s comments on HUD’s letter dated November 22, 1996. 1. Throughout our assignment, we have included program evaluation as a topic in interviews that included the Office of Community Planning and Development’s General Deputy Assistant Secretary and Special Assistant to the General Deputy Assistant Secretary, and the Office of Economic Development’s Director and Deputy Director, as well as members of the EZ/EC task force. As recently as October 16, 1996, we asked the Deputy Director of the Office of Economic Development, under which the EZ/EC task force is located, for a copy of the reports on the EZs’ performance. We were told they were not available. After receiving HUD’s comments on this report, we talked with the General Deputy Assistant Secretary, who provided an example of a performance report from one EZ and told us that HUD’s analysis of the EZs’ reports and the results of short-term reviews that are being performed by a contractor were not yet available. The section of our report on evaluating the EZ program is not meant to detract from the efforts that HUD and the EZs have already made in setting up a system to track and measure activities in each zone. On the contrary, we believe that the steps taken so far are essential in building a set of outcome-oriented performance measures. We revised the report to clarify our support for measuring the program in results-oriented terms and to include HUD’s position on outcome measures. 2. We agree with HUD that the federal government obligated the EZ/EC SSBG funds in a timely manner and have added language to that effect. Although HUD sent a letter to the EZs on the use of funds for administrative operations, this section of the report restates comments made by people involved in the program at the federal, state, and local levels. As we stated in the beginning of that section of the report and in the appendix on our scope and methodology, we did not independently analyze the factors identified by the respondents or examine their applicability across all of the EZs. We believe the report accurately and sufficiently covers the program participants’ concerns about the lack of administrative funding. 3. We chose not to include obligations in the table because there is no standard definition for obligations below the federal level. Therefore, including the numbers in a table would invite comparisons of amounts that are not comparable. The EZs’ definitions of obligations range from obtaining approval for a project’s funding from the city council to awarding a contract. However, we have revised the report to indicate that higher amounts have been obligated by the EZs. 4. As noted in comment 2 above, this section of the report restates comments from people involved in the EZ program. As we said in the beginning of that section of the report and in the appendix on our scope and methodology, we did not independently analyze the factors identified by the respondents or examine their applicability across all of the EZs. Furthermore, in the background section of the report, we acknowledged that HHS encouraged the states to carry out their EZ funding responsibilities with as few restrictions as possible under the law, a statement that we based on the same letter that HUD quotes. We also note in the background section of the report that the state is fiscally responsible for the funds. This statement is related to another part of the letter from HHS that says (1) the states will technically be subject to possible recoupment actions by HHS if an EZ or EC uses the EZ/EC SSBG funds to finance an activity not allowed by the authorizing statute and (2) the state can hold localities accountable for the appropriate use of funds. We believe the report accurately and sufficiently covers the program participants’ concerns about the states’ roles. 5. We revised the report to indicate that HUD identified these two examples of programs that involve both the EZs and public housing officials. 6. As noted in comment 2 above, this section of the report restates comments from people involved in the EZ program. As we stated in the beginning of that section of the report and in the appendix on our scope and methodology, we did not independently analyze the factors identified by the program participants. We agree that HUD has a variety of methods for communicating with the EZs, a factor that was included in our survey. However, too few program participants indicated that this factor helped implementation efforts to a great or very great extent for us to include the factor in the report. Consequently, we feel that no change is needed. 7. We revised the report to include these suggestions. The following are GAO’s comments on HHS’ letter dated November 29, 1996. 1. We agree that HHS is and has been an important partner in this program. Among other things, HHS has provided information to the EZs to clarify the uses of and controls on the EZ/EC SSBG funds, advised the EZs and ECs about other HHS grants that have been awarded or are available to the EZs, and worked with HUD and USDA in preparing guidance issued to the EZs. We added some information on HHS to the background section of the report. 2. Although the two segments of this zone operate autonomously for the most part, they are still a single EZ. Furthermore, the administration of this EZ was not significantly different from that of the New York EZ, which has split its operations between two corporations covering distinct parts of the designated zone. Consequently, we believe that discussing the two segments of the Philadelphia-Camden zone together is appropriate because they form a single, bistate zone. 3. We have revised or added wording to the report to make the changes, when appropriate. HHS also felt that the report should include additional background information about HHS’ role in the program and raised a number of editorial and technical comments on the report. We changed the wording of the report, when appropriate, to address these comments. 4. We updated the report to include the amounts drawn down as of October 31, 1996. The amounts for Detroit and Chicago were provided by the EZ. A state official in Michigan verified that the drawdown had taken place and that the state had provided the funds to the EZ. They added that the amount would not yet show up on HHS’ records because of the timing of the state’s request for a drawdown. An HHS official told us that the process in Illinois was similar and that HHS had received a request for a drawdown in November 1996. Nancy A. Simmons Carole Buncher Merrie Nichols-Dixon Susan Beekman Gwenetta Blackwell Johnnie Barnes Elizabeth R. Eisenstadt Alice G. Feldesman Mitchell B. Karpman John T. McGrail The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed the Department of Housing and Urban Development's (HUD) Empowerment Zone and Enterprise Community (EZ/EC) Program, focusing on the: (1) status of the program's implementation in the urban EZs, including the extent to which public housing officials and residents have been involved; (2) factors that participants believe have either helped or hindered efforts to carry out the program; and (3) plans for evaluating the program. GAO found that: (1) the six urban EZs reviewed resemble each other in some ways, but also differ in ways that reflect the diversity of the communities; (2) all of the EZs have included public housing authority officials in planning and implementing the program; (3) many officials involved in implementing the program generally agreed on factors that have either helped or hindered their efforts; (4) in a survey GAO conducted of program participants at the federal, state, and local levels, over half of the 27 program officials who responded agreed that community representation on the EZ governance boards, technical assistance provided by HUD's contractors, enhanced communication among stakeholders, and support from the city's mayor and from White House and cabinet-level officials had helped the program's implementation; (5) conversely, the difficulty in selecting a governance structure, the additional layer of bureaucracy created by the state government's involvement, preexisting relationships among EZ stakeholders, pressure for quick results, and the lack of federal funding for the program's initial administrative activities were frequently identified as factors constraining implementation; (6) HUD required each EZ to prepare a strategic plan indicating how it would satisfy the EZ program's four key principles and required the strategic plan to include realistic performance benchmarks for measuring progress in implementing the program; (7) all six of the urban EZs prepared strategic plans that include benchmarks describing the activities that the EZ planned to accomplish during the first 2 years of the program; and (8) HUD and the EZs have not yet described measurable outcomes for the program's key principles or indicated how the outputs anticipated from one or more benchmarks will help to achieve those outcomes.

The definition of “elderly families” used to determine eligibility for certain public housing has evolved over time. Until 1992, the term encompassed low-income families whose head, spouse, or sole member was aged 62 or older, as well as low-income individuals with disabilities, regardless of age. All housing authorities were obligated by law to provide equal access to units in developments known as “elderly buildings” to both elderly persons and persons with disabilities. These developments usually consisted of efficiencies and one-bedroom units rather than the multiple-bedroom units typical of family housing. As the number of younger persons with disabilities residing in “elderly housing” increased, the complaints from their elderly neighbors also increased. Elderly residents cited the differences in values and lifestyles they had with the younger residents as impediments to safe and decent housing for the elderly. In 1992, we reported that nonelderly tenants with mental disabilities occupied between 8 and 10 percent of the units in public housing for the elderly, but were, according to public housing authorities (PHA), causing a disproportionate share of problems for the elderly residents as well as for the PHAs’ management and staff. Problems included complaints about noise, visitors, crime, and disrespectful attitudes of the younger residents toward the elderly residents. It was primarily this increase in the number of tenants with disabilities in elderly housing and the resultant complaints that led the Congress to promote designated housing in the 1992 act. First, the 1992 act established a definition of elderly persons that no longer included persons with disabilities under the age of 62. Second, the act allowed PHAs to seek approval for designating housing as elderly-only, disabled-only, or elderly and disabled through allocation plans submitted to HUD. The 1992 act laid out criteria for the contents of allocation plans and standards that HUD should use to approve the plans. Specifically, the law stipulated that approval is to be granted to only PHAs that demonstrate in their allocation plans that designation is necessary to achieve the housing goals for their jurisdictions and to meet the needs of the jurisdictions’ low-income population. PHAs that do not have approved allocation plans must continue to treat persons with disabilities and the elderly equally and allow both to live in elderly buildings on a first-come, first-serve basis. Finally, the act allowed owners of privately owned, HUD-assisted projects that were designed primarily for occupancy by elderly families to designate housing for the elderly through the establishment of elderly preferences or restrictions. Unlike PHAs that designate housing, however, owners of privately owned projects do not need to seek HUD’s approval prior to designation or to notify HUD once designation occurs. To offset the potential loss of housing for persons with disabilities and to provide them with greater housing choices, HUD set aside Section 8 rental housing certificates and vouchers for their use. Through appropriations for fiscal years 1997 and 1998 for incremental certificates and vouchers, the Congress augmented HUD’s ability to set aside Section 8 assistance for persons with disabilities displaced by public and privately owned housing designated for the elderly. Separately, HUD has also made Section 8 certificates and vouchers available to provide mainstream housing opportunities for persons with disabilities. The mainstream program is open to all housing authorities, not just those with approved allocation plans. It is popular with persons with disabilities who want to find housing in the private sector (known as “housing of choice”) rather than in public or project-based housing. Only 73 of the nation’s 3,200 PHAs had allocation plans as of November 1, 1997. These PHAs typically elected to designate units for the elderly or for the elderly and persons with disabilities combined. Of all the 73 PHAs with allocation plans, 64 designated a total of 24,902 housing units—approximately 36 percent of their housing stock for the elderly or persons with disabilities—as elderly-only. As a result, these units may no longer be available to persons with disabilities. However, persons with disabilities occupied only a portion of these units in the past, and most who resided in those units still do. Specifically, 53 of the 64 PHAs that designated units as elderly-only and provided complete occupancy data in our survey reported that when they submitted their plans, about 4,100 persons with disabilities were occupying units now designated for the elderly. The number had fallen to about 3,000 as of November 1, 1997. Overall, there has been little change in the number of persons with disabilities residing in units now designated as elderly-only, disabled-only, and elderly and disabled or in other undesignated units available to the elderly and persons with disabilities. One reason is that more than half of the PHAs with allocation plans had them approved as recently as 1996 or 1997, so the designations are relatively new. Another reason is that PHAs are not permitted to force residents with disabilities to move out of housing newly designated for the elderly and residents with disabilities may not wish to relocate. (See app. IV for our survey instrument and the complete responses to our survey.) According to HUD, 73 out of 3,200 PHAs had approved allocation plans as of November 1, 1997. HUD denied applications from another 22 PHAs, and 2 other PHAs withdrew their applications. Although some plans were approved as early as 1994, over half of them were approved in 1996 and 1997. The 73 PHAs with approved allocation plans represent a cross-section of housing authorities in terms of size and are located in 35 states. Twenty-four PHAs with approved allocation plans had fewer than 500 units of public housing, 16 had between 500 and 1,249 units, and 33 had 1,250 or more units. All but one of these PHAs—Toccoa, Georgia—reported designating housing units as part of their allocation plans. (See fig. 1.) Most of the 73 PHAs with allocation plans designated units as elderly-only. Specifically, 64 of the PHAs that have approved allocation plans reported in our survey that they designated 24,902 units as elderly-only, meaning that the units will no longer be rented to new tenants who are not at least 62 years old. Five PHAs accounted for more than a third of the units designated as elderly-only: Worcester, Massachusetts; Milwaukee, Wisconsin; Minneapolis, Minnesota; San Antonio, Texas; and Chicago, Illinois. Seven PHAs designated 50 or fewer units as elderly-only. (See fig. 2.) Fewer units were designated as disabled-only or elderly and disabled. (See table 1.) Among the 64 PHAs that designated units only for the elderly, 19 also designated units in which either the elderly or persons with disabilities could be housed, and 7 also designated units as disabled-only. Four of the PHAs used all three of the possible designations. One PHA with an approved allocation plan opted not to designate any housing units. Finally, nonelderly persons with disabilities and the elderly were eligible for an additional 23,870 units at the 73 PHAs with approved allocation plans. These units were not designated as elderly- or disabled-only in the PHAs’ allocation plans but were available to these persons nonetheless. Over half of these units belonged to three PHAs: Chicago had 5,320; Baltimore, 3,662; and Minneapolis, 3,572. Of the 24,902 units designated for elderly residents, 24,471 were previously available to both the elderly and to younger persons with disabilities. About 19 percent of the units that were previously available to both types of tenants were occupied by younger persons with disabilities when the PHAs submitted their allocation plans. An additional 367 units designated as elderly-only were newly constructed or acquired, and 64 were in other types of public housing units. Overall, PHAs reported in our survey that occupancy by both persons with disabilities and the elderly declined slightly across all units between the time the PHAs submitted their allocation plans and November 1, 1997.Specifically, the 56 PHAs that provided complete occupancy data reported that occupancy by persons with disabilities in units now designated for the elderly or persons with disabilities or available to either declined by 533 households—about a 5-percent decrease. The number of units occupied by the elderly declined by 432—not quite a 2-percent decrease. (See fig. 3.) Those PHAs that had specifically designated units as elderly-only reported the same pattern as did PHAs with allocation plans generally. While overall occupancy changed little, the occupancy of persons with disabilities varied depending on whether the units were designated as elderly-only or elderly and disabled. Changes in occupancy also varied by individual PHA. Of the 53 PHAs that designated units as elderly-only and provided complete occupancy data for those units in their survey responses, 25 reported a decline in the number of persons with disabilities residing in elderly-only units. The total decrease of 1,066 occupants represented about one-quarter of the persons with disabilities residing in those PHAs’ units designated as elderly-only. The Chicago Housing Authority accounted for much of this decline, reporting 419 fewer people with disabilities residing in units now designated elderly-only. Thirteen PHAs reported no change, and 15 reported increases. One PHA reported an increase of 52 nonelderly people with disabilities occupying units that had been designated elderly-only. According to an official at this PHA, the allocation plan permits the PHA to accept near-elderly persons—including those with disabilities—if there is an insufficient number of elderly persons to occupy designated units. The increase in persons with disabilities at this PHA is due to the large number of near-elderly persons with disabilities who moved into designated units after the PHA submitted its allocation plan. Forty PHAs that designated units as elderly-only reported that 618 people with disabilities had moved out of the elderly-only units. Some of these PHAs also reported where these persons with disabilities went: 80 moved into other public housing units; 138 moved into private housing using Section 8 certificates or vouchers; and 275 moved into other housing. The PHAs that designated units for the elderly and disabled combined or that had undesignated units available for the elderly and persons with disabilities reported a total increase in the number of persons with disabilities occupying those units. The 47 PHAs in this category that provided complete occupancy data reported 472 more of these units were occupied by persons with disabilities—about a 7-percent increase. The 13 PHAs that designated 788 of their units as disabled-only reported that occupancy of those units by persons with disabilities increased by 20 households—more than a 4-percent increase—between the time they submitted their allocation plans and November 1, 1997. (See table 2.) One reason the provisions in the 1992 act allowing PHAs to designate units for the elderly have had little impact on the availability of public housing for persons with disabilities is that so few PHAs have sought to use the provisions thus far. While we did not attempt to survey the more than 3,100 PHAs that have not submitted allocation plans to designate housing, HUD officials told us that most PHAs do not view developing the plans as a priority or do not believe that designated housing is necessary. Moreover, HUD agreed that designated housing has had little impact on housing opportunities for persons with disabilities. HUD cited the collaborative approach it uses to review plans, coupled with the availability of targeted Section 8 certificates and vouchers, as the factors that have minimized the potential impact. Two-thirds of the PHAs responding to our survey indicated that their designation of housing to date has neither helped nor hindered their ability to meet the housing needs of persons with disabilities. Officials at the three PHAs we visited that had allocation plans confirmed that designating units for the elderly had not had an impact on persons with disabilities. They noted that persons with disabilities were not required to move out of the designated units and that many of them had decided not to relocate. For example, an official at the Dallas Housing Authority said that many of the tenants preferred to stay where they were. At the Fall River Housing Authority, persons with disabilities residing in units designated as elderly-only were offered Section 8 certificates. Some of these residents were able to use these certificates to find housing in the private market, while the other residents remained in their newly designated units. The extent to which Section 8 rental certificates and vouchers for persons with disabilities have helped meet the demand for affordable housing for persons with disabilities is not yet clear. Fifty-nine housing authorities reported receiving 4,943 certificates and vouchers for persons with disabilities. Thirty-three of these 59 authorities also designated units as elderly-only. About two-thirds of the PHAs with elderly-only units reported that they received more certificates and vouchers than the number of persons with disabilities residing in their elderly-only units. However, as of November 1, 1997, only about one-fourth of the total 4,943 certificates and vouchers reported had been used successfully. Officials from HUD and housing authorities that we surveyed gave various reasons for certificates and vouchers not being used. For instance, because of the time involved in awarding certificates and vouchers to housing authorities, many had not been in a position to issue the new rental certificates and vouchers by November 1, 1997. Statutory restrictions on the funds have limited their use. Persons with disabilities seeking to use certificates and vouchers might have had difficulty finding affordable, privately owned housing in their communities. And many tenants with disabilities might not be interested in moving out of the public housing units they now occupy. The Section 8 Rental Certificate and Rental Voucher program is a federally funded affordable housing option administered by the local housing authorities, which issue the certificates and vouchers to eligible tenants.HUD has made Section 8 certificates and vouchers available for persons with disabilities in two ways: (1) in connection with designated public or privately owned, HUD-subsidized housing or (2) through a mainstream housing opportunities program. Since 1992, a total of $278.9 million has been earmarked for certificates and vouchers for persons with disabilities. As of November 1, 1997, HUD had made $190.4 million available through four notices of funding availability. As figure 4 shows, only the funds from the March 1995 notice and the April 1997 notice for the mainstream program have been totally awarded to the housing authorities. HUD told us that between November 1997 and March 1998, another $15 million made available through the October 1996 notice was awarded to PHAs with newly approved allocation plans. HUD expects all the funds from the October 1996 notice to be spent by the end of fiscal year 1998. The April 1997 funding notice making $50 million available to PHAs was the result of a set-aside in HUD’s fiscal year 1997 appropriation. The Congress set aside this money to fund Section 8 assistance to persons with disabilities affected by designated public and private housing. Half of the $50 million was earmarked for housing authorities with approved allocation plans, and the other half was earmarked for housing authorities that could identify the impact on persons with disabilities of elderly-only preferences established by privately owned projects in their communities. As of November 1, 1997, none of the $25 million for housing authorities with approved allocation plans had been spent. According to HUD officials, this $25 million will be used once all of the funding made available in the October 1996 notice is awarded. From the $25 million intended to offset the impact of elderly-only preferences established by privately owned projects, HUD had awarded $2.7 million to five PHAs. Appendix V provides further discussion of Section 8 certificates and vouchers for persons with disabilities. A total of 59 housing authorities reported receiving Section 8 certificates and vouchers for persons with disabilities. (Fig. 5 shows their locations.) Six of these housing authorities received awards for both designated housing and mainstream opportunities. They are Rochester, New York; Butler County, Pennsylvania; Kansas City, Missouri; Wilmington, Delaware; Greensboro, North Carolina; and Salem, Oregon. Fifty-nine of the 96 housing authorities we surveyed reported that they had received 4,943 certificates and vouchers for persons with disabilities from one or more of the four notices of funding availability. However, because of the length of the process used to make awards to housing authorities, many were not in a position to issue the new rental certificates and vouchers awarded from one of the 1997 notices by November 1, 1997. Of the approximately 3,000 certificates and vouchers that were available to housing authorities as of November 1, 1997, 1,558 had been issued to persons with disabilities who, in turn, had used 1,162 to obtain private rental housing and had turned back 174 unused, according to the authorities. Some housing authorities reported that their recipients were still in the process of searching for housing. Persons with disabilities generally have 60 days to use certificates and vouchers and may ask for extensions if necessary. Thirty-three PHAs reported that they designated units as elderly-only and received 2,982 certificates and vouchers, primarily in connection with their designation. About two-thirds of these PHAs with elderly-only units reported that they received more certificates and vouchers than the number of persons with disabilities residing in their elderly-only units.These PHAs issued 1,236 of these certificates and vouchers to persons with disabilities. Of the certificates and vouchers issued, 960 had been used by persons with disabilities to obtain housing and 172 had been turned back to the housing authorities unused. The remaining recipients were still looking for housing. HUD still has funds remaining for certificates and vouchers associated with designated housing, but all the funds for the mainstream program have been awarded. According to HUD officials, there is not much demand for the designated housing certificates and vouchers. The Section 8 Certificates and Vouchers Director told us that PHAs prefer to apply for the mainstream housing opportunities program for persons with disabilities because it is less restrictive, it does not require PHAs to submit allocation plans, and the funds have 5-year terms. HUD officials also told us that the restrictions placed on the $50 million the Congress set aside for persons with disabilities in the fiscal year 1997 appropriation made the funds difficult to award. Consequently, HUD worked with the Congress to make the language less restrictive for the $40 million set aside in the fiscal year 1998 appropriation. The language now states that if the funds cannot be awarded to PHAs that have designated housing or that have identified the impact on persons with disabilities of elderly preferences established by privately owned projects, then the Secretary of HUD may make the remaining funds available for the mainstream program. This language is retroactive to the remaining funds from the fiscal year 1997 appropriation. On April 30, 1998, HUD published a notice of funding availability to disseminate the $87.3 million from these fiscal years and the $48.5 million earmarked for the fiscal year 1998 mainstream program. Our survey found that the number of persons with disabilities living in public housing units managed by those PHAs that have designated housing and received certificates and vouchers actually increased by 198 between the time the PHAs submitted their allocation plans and November 1, 1997.Similarly, occupancy by persons with disabilities increased by 197 households at the PHAs that designated elderly-only units and received certificates and vouchers and that provided complete occupancy data. Most of the persons with disabilities using the certificates or vouchers were not tenants in newly designated units; rather, they had been on the PHAs’ waiting lists for either public housing or Section 8 assistance. As shown in figure 6, only 18 percent of the 1,147 certificates and vouchers whose recipients housing authorities were able to identify were used to move persons with disabilities out of public housing. Fifty-five percent were used by persons with disabilities who had been on the PHAs’ public housing waiting lists, and 27 percent were used by persons with disabilities who had been on the PHAs’ Section 8 waiting lists. The 33 PHAs that made elderly-only designations and received certificates and vouchers reported a very similar pattern of use. Our survey data support the view that tenants with disabilities may be reluctant to move from what they perceive to be a stable, known housing situation into an unknown situation. Those on the waiting lists, however, are seeking affordable housing opportunities and therefore may be more willing to use the certificates and vouchers to improve their housing situation. Depending on the resident and on the housing market, Section 8 certificates and vouchers may or may not be as preferable as public housing. Some PHA officials told us that not every person with disabilities residing in public housing that has been newly designated as elderly-only wants to move. Tenants may want to stay because they have a network of friends nearby, supportive services, and available transportation services. Moreover, moving to Section 8 housing means paying security deposits, which would be an added financial burden for these tenants. In addition, some tenants believe that a certificate or voucher would not be as permanent as public housing. Even at those PHAs that provide incentives, such as paying for moving expenses and phone service transfers, public housing residents with disabilities reportedly do not want to relocate. The Dallas Housing Authority, for example, found that even with incentives, only 20 of the 80 persons with disabilities residing in the newly designated elderly-only buildings were willing to use Section 8 certificates. As more PHAs continue to receive their allotments of certificates and vouchers and to educate their residents on how to use them, this situation may change. At other locations, certificates and vouchers may not be an appropriate option for persons with disabilities because of tight housing markets, rents that are above HUD’s fair market rents (FMR), or the prevalence of older housing that has not been adapted to the needs of people with disabilities.For example, because of the high demand for rental housing in the San Francisco market and the ease with which landlords can get rents higher than HUD’s FMR, landlords may not be willing to rent to subsidized tenants, especially to persons with disabilities for whom they would have to make accessibility accommodations. HUD’s FMR is also an issue in Cambridge, Massachusetts, which recently underwent rent decontrol. The Cambridge Housing Authority used a portion of its Section 8 administrative fees to hire a housing-search worker to help Section 8 recipients find housing. Another PHA we surveyed—in Westbrook, Maine—also cited a tight housing market and a predominance of older, inaccessible housing as obstacles faced by persons with disabilities seeking housing in the private market. Apartments in older buildings usually have small rooms and narrow entrances, making it difficult to improve their accessibility. Despite the difficulties encountered in some housing markets, when persons with disabilities were offered Section 8 certificates and vouchers, they appeared to be able to use them in most locations. Overall, only about 11 percent of the certificates and vouchers had been turned back to the housing authorities unused as of the time of our survey. Half of the housing authorities reported that the rate at which persons with disabilities turned back certificates and vouchers was about the same as that for recipients without disabilities. Moreover, they said that these users were taking about the same amount of time as the other recipients to find housing. However, the housing authorities reported in our survey that persons with disabilities required greater assistance to locate private housing. In our case study interviews, housing authorities’ managers told us that the extra assistance they provided included preparing lists of apartments that accepted Section 8 certificates or vouchers and transportation to the apartments. Where assistance was provided, certificates and vouchers had helped persons with disabilities. In our survey, the majority of housing authorities reported that it was still too early to determine how successful certificates and vouchers have been in helping persons with disabilities rent private housing. This observation was affirmed by our case study work. Four of the six housing authorities where we conducted case studies had received certificates or vouchers, but managers there said that it was too early to determine how successful this Section 8 assistance had been in providing housing options. The Housing Authority of Gloucester County, New Jersey, for example, had only recently hired a full-time person to administer the 130 vouchers it received after it identified privately owned projects that had established preferences for the elderly; as of January 1998, it had successfully utilized 10 vouchers to provide housing. Similarly, the Anaheim, California, housing authority had not yet issued any of the 150 vouchers it had received from the mainstream housing program. It expected that all 150 vouchers would be used, however, and planned to apply for more vouchers because it had 1,500 persons with disabilities on its waiting list. We provided HUD with a draft of this report for review and comment. We met with HUD officials to discuss their comments and our response. Specifically, we met with the following officials from the Office of the Assistant Secretary for Public and Indian Housing: the Senior Director of the Office of the Deputy Assistant Secretary for Policy, Program, and Legislative Initiatives; the Acting Director of Management and Planning; the Senior Program Manager, Real Estate and Housing Performance Division (Section 8); and the Senior Program Analyst, Customer Service and Amenities Division. We also met with officials from the Office of Fair Housing and Equal Opportunity and the Special Assistant to the Secretary on Disability Rights. There were four general categories of comments, which are listed below with our response. Comments on our analysis on the allocation plans: HUD requested that we include the most up-to-date information on the number of PHAs with approved allocation plans and the number of designated units. While we did not obtain survey responses from the PHAs that received approval for their allocation plans after November 1, 1997, we have made note of the most recent data provided by HUD. HUD also believed that our analysis showing that the 24,902 units designated as elderly-only represent 7.5 percent of these PHAs’ total units was misleading. HUD said the figure implies that only an insignificant percentage of the available public housing is no longer available to persons with disabilities in communities with approved allocation plans. We have changed the language in the report to reflect that the 24,902 units represent 36 percent of these PHAs’ housing stock for the elderly and for persons with disabilities. HUD requested that we include additional information on the statutory requirements of the designated housing law. Where appropriate, we have done so. Finally, we have incorporated HUD’s position that the reason designating housing for the elderly has had little impact on housing opportunities for persons with disabilities is that HUD worked closely with these PHAs to ensure that they were able to address the needs of the members of all the groups affected by designations. Comments on our analysis on Section 8 certificates and vouchers: HUD believed that because of the length of time taken by the submission and approval of applications and by the execution of annual contributions contracts, many PHAs were not in a position to issue the certificates and vouchers they were awarded from either of the two notices of funding availability published in April 1997 until late October or early November. HUD requested that we add language to state that it should not be surprising that only a small percentage of certificates and vouchers had been issued and used by persons with disabilities as of November 1, 1997. We have added, in the appropriate sections of the report, language to acknowledge the time needed for the submission and approval of housing authorities’ applications and for the issuance and execution of annual contributions contracts before certificates and vouchers are actually available for use. HUD also suggested that we include the fiscal year 1998 funding of $48.5 million for the mainstream certificates and vouchers program. We did so. However, we disagree with HUD’s suggestion that we delete from our analysis the $25 million of the fiscal year 1997 appropriations and the $20 million of the fiscal year 1998 appropriations that were directed at assisting nonelderly persons with disabilities affected by elderly preferences established by privately owned projects. We included these amounts because the Congress requested that our analysis include all new Section 8 certificates and vouchers for persons with disabilities. Comments on our survey instrument: HUD thought that our survey instrument should have included questions pertaining to housing demand, characteristics of persons on public housing waiting lists, and the number of efficiencies and one-bedroom units in family developments. Questions 24, 37, and 38 of the survey instrument asked PHAs to evaluate the impact that designating housing and issuing additional certificates and vouchers have had on their ability to meet the housing needs of persons with disabilities. We expected the PHAs to take demand for housing and the characteristics of the persons on their waiting lists into consideration in their responses to these subjective questions. Public housing waiting lists, we discovered in our pretests, are imperfect measures of demand for housing as they reflect demand only at the time the individual is placed on the waiting list or when the list is updated. Our survey found that 82 percent of the certificates and vouchers were used by persons with disabilities on public housing and/or Section 8 waiting lists. This is consistent with HUD’s point that persons on waiting lists are the group most helped by the targeted Section 8 rental assistance. Furthermore, the focus of the survey questions regarding occupancy was on the designated units and the units available to the elderly and persons with disabilities. These questions did not exclude the possibility of such units in family buildings. Finally, we disagree that our question on how many units PHAs had designated for only persons with disabilities did not differentiate between units designated in the allocation plan and physically accessible units that were not designated in the plan. Questions 11 and 12 of the survey used the language “units you designated in your allocation plan as ’disabled-only,’” which we believe is very clear. We had no indication from either our pretests or our callbacks to housing authorities for response clarification that they were incorrectly interpreting these questions. Comments on our population analysis: HUD believed that we needed to focus more on families with very low incomes (i.e., incomes at or below 50 percent of the median family income for their area) in our analysis, not just low-income families (i.e., those with incomes at or below 80 percent of the median income for their area). HUD stated that families with incomes between 50 and 80 percent of the area’s median income qualify for a smaller number of public housing units and generally do not qualify for Section 8 rental assistance. We concurred and made appropriate modifications. HUD also pointed out that our estimates of low-income persons with disabilities meeting HUD’s definition differ significantly from the conclusion HUD reached in its 1997 report to the Congress on worst-case housing needs. While we used the definition of disability used in HUD’s housing programs, HUD, in its report on worst-case housing needs, used a more restrictive definition of disability than it uses in its housing programs. We conducted our review from July 1997 through May 1998 in accordance with generally accepted government auditing standards. We are sending copies of this report to the appropriate Senate and House committees; the Secretaries of HUD and Health and Human Services; and the Director of the Office of Management and Budget. We will make copies available to others on request. Please call me at (202) 512-7631 if you or your staff have any questions about the material in this report. Major contributors to this report are listed in appendix VI. To estimate the number of households in which the head of the household or spouse meets the Department of Housing and Urban Development’s (HUD) definition of a low-income person with a disability, we used data from the National Health Interview Survey (NHIS) conducted for the Department of Health and Human Services’ (HHS) National Center for Health Statistics. We worked with HHS’ Office of the Assistant Secretary for Planning and Evaluation (ASPE) and the Special Assistant to the Secretary in HUD’s Office of Disability Policy to develop criteria for using data about income, age, and disabilities from the 1994 survey. HHS then applied these criteria to estimate the number of noninstitutionalized households in which the head of the household or spouse—or the sole person in the case of a single-person household—would be considered under HUD’s definition to be a person with a disability. NHIS, which was first conducted in 1957, is a continuing national survey of civilian, noninstitutionalized households. It is the principal source of information on the health of this population. In addition to the basic survey, NHIS in many years includes supplements covering special topics. For 1994, NHIS covered five special topics, including disability and family resources—both of which were used in the estimates of households meeting HUD’s definition of persons with disabilities. Because NHIS estimates are based on a sample of households, they may differ somewhat from the figures that would have been obtained from a complete census. In 1994, 45,705 households were interviewed, resulting in a sample of 116,179 persons. Since NHIS’s design is a complex multistage probability sample, the estimates provided do have sampling errors; however, HHS did not compute the sampling errors for each estimate. Where we could identify similar estimates published by others, we found that HHS’ estimates were generally in the same range. We did not perform a technical review of HHS’ programming. NHIS excludes members of the armed forces, U.S. nationals living abroad, nursing home residents, and institutionalized persons. According to the 1990 census, 3.3 million Americans lived in institutions, including 1.8 million persons in nursing homes. An additional 340,000 persons lived in other types of institutions, such as psychiatric hospitals and schools, hospitals, or wards for the mentally retarded. Some portion of the 340,000 persons in such institutions, as well as those in nursing homes, might be expected to have conditions that substantially impair their ability to live independently, according to HUD officials. However, these persons are not included in the estimates presented here of the population meeting HUD’s definition for persons with disabilities. The remaining persons in institutions include 1.1 million persons in correctional institutions and 104,200 in juvenile institutions. Low-income and very low-income families are eligible for housing assistance under HUD’s public and assisted housing programs. Generally, low-income families are those with adjusted incomes at or below 80 percent of the median income in their areas, as determined by HUD with adjustments for smaller and larger families. Very low-income families are those with adjusted incomes at or below 50 percent of their areas’ median income. Very low-income families qualify for public housing and Section 8 rental assistance. Families with incomes between 50 and 80 percent of their area’s median income, however, are eligible for HUD’s public housing programs but generally compete for a smaller number of units, and most do not qualify for Section 8 rental assistance. HUD’s public and assisted housing programs use a definition of disability that includes a measure of functional limitation that is due to a medical condition of a certain duration. Specifically, the programs generally consider a person to have a disability if that person (1) meets the definition used under section 223 of the Social Security Act; or (2) meets the definition for developmental disabilities found in the Developmental Disabilities Assistance and Bill of Rights Act; or (3) has a physical, mental, or emotional condition that is expected to be of long-continued and indefinite duration, substantially impedes the person’s ability to live independently, and is of such nature that the person’s ability to live independently could be improved by more suitable housing conditions. A low-income or very low-income family whose household head or spouse meets HUD’s definition of a person with a disability is ordinarily eligible for housing units available to the elderly, even though the household head or spouse may not be elderly. However, the Housing and Community Development Act of 1992 allows public housing authorities (PHA) that have approved allocation plans and owners of certain HUD-assisted projects to restrict occupancy in particular units to the elderly only. To identify low-income and very low-income families, we asked that HHS compare total household income—excluding income from persons under the age of 18—recorded in the NHIS data with HUD’s 1994 adjusted median family income for metro and nonmetro areas within the four census regions, adjusted for family size. HHS summarized responses to questions in the family resources supplement of NHIS on components of household income for adult family members—in the case of families—and for the sole person in single-person households. For NHIS, all persons in a household related to each other by blood, marriage, or adoption constitute a family. For about 16 percent of the households, there were no income data; these households were excluded from the estimates provided here. We do not know how the income of those households compared with the income of households for which the data were available. To identify households in which the household head or spouse (or the sole person in the case of a single-person household) was a person with a disability as defined by HUD, we utilized responses to questions about the receipt of Supplemental Security Income (SSI); any conditions associated with developmental disabilities; the presence of physical, mental, and emotional conditions; the duration of the condition; and the difficulty experienced with activities of daily living (ADL) and instrumental activities of daily living (IADL). Specifically, we estimated the number of households in which the household head or spouse or sole individual (1) received SSI; (2) met the definition of disability under the Developmental Disabilities Assistance and Bill of Rights Act;(3) had—without the use of equipment, help, or supervision—a lot of difficulty with or was unable to perform one or more ADL or IADL because of a physical, mental, or emotional condition expected to last another 12 months; or (4) had a mental or emotional condition that seriously interfered with his or her ability to work, attend school, or manage day-to-day activities without regard to the duration of this condition. We did not attempt to determine if the person’s ability to live independently could have been improved by more suitable housing conditions. Also, we did not attempt to determine the person’s housing needs or whether a person was currently residing in public or assisted housing. According to the estimates provided by HHS, in 1994 about 9 million low-income households (including about 7 million very low-income households) not living in institutions had as a household head or spouse a person who may have met HUD’s definition of a person with a disability. In almost half of these households, the household head or spouse was a nonelderly adult with a disability. There were an estimated 2.5 million low-income renter households that had as a household head or spouse a nonelderly person with a disability. The overall estimates provided by HHS were within the range of other estimates of the number of persons with disabilities. Specifically, when not considering income, HHS estimated that there were about 11 million households—or almost 14 percent of all households—that had as a household head or spouse a person who met HUD’s definition of a person with a disability. In comparison, a 1996 study estimated that for about 17 million families—or about 24 percent of all families in the 1990 NHIS—the household head, or in the case of “partnered families” one or both partners, had a disability. A study using data from the 1992 Survey of Income and Program Participation estimated that about 48.9 million noninstitutionalized civilians—or about 19.4 percent of this population—had a disability. Of these, about half had a severe disability. Considering that in 1992, households had an average of 2.62 persons, the number of households with a member that had a disability would have been about 19 million, and the number of households with a member that had a severe disability would have been about 9 million. According to the 1990 census, about 22 million noninstitutionalized persons aged 16 and older had a work disability, a mobility limitation, or a self-care limitation. This may equate to about 8.5 million such households. Of the HHS-estimated 46.3 million low-income households, 8.7 million, or about 19 percent, were households that were headed by persons or spouses with disabilities and that had incomes at or below 80 percent of HUD’s adjusted median family income. Of that 8.7 million households, 6.6 million had incomes at or below 50 percent of HUD’s adjusted median family income. A low-income household was more likely than other households to have a person with a disability as a household head or spouse. About half of these households had as a household head or spouse a person with a disability who was younger than age 62. Elderly households represented about 51 percent of the low-income households headed by a person or a spouse with a disability. However, elderly households were twice as likely as younger households to have a household head or spouse with a disability. That is, while 28 percent of low-income elderly households had as a household head or spouse a person with a disability, the figure was 14 percent for younger low-income households. About 45 percent of the low-income households headed by a person or a spouse with a disability rented, rather than owned, their homes. Table I.1 shows the estimated income and disability status of the households included in the 1994 NHIS. Sole person in the case of single-person households. Does not include secondary families and individuals. Head of household or spouse was aged 62 and older. Head of household or spouse was aged 18 to 61. The Subcommittee on VA, HUD, and Independent Agencies of the House Committee on Appropriations asked us to assess (1) the impact of the Housing and Community Development Act of 1992 on the availability of public housing for persons with disabilities and (2) how incremental Section 8 certificates and vouchers that were made available since the passage of the 1992 act were assisting persons with disabilities seeking affordable rental housing. As requested, we also developed estimates of the number of households that may meet HUD’s definition of persons with disabilities. To obtain information for this report, we conducted a survey of the 73 PHAs that, according to HUD, had approved allocation plans as of November 1, 1997. In addition, we sent the survey to 23 other housing authorities that did not have approved allocation plans but had been awarded Section 8 certificates and vouchers for the exclusive use of persons with disabilities as of November 1, 1997. We mailed a questionnaire to each of these 96 housing authorities and made follow-up calls as needed to stimulate responses or verify unclear answers. All 96 housing authorities responded to the survey. The results of the survey are summarized in appendix IV. To supplement the results of the survey, we conducted six case studies at housing authorities and their immediate communities in (1) Fall River, Massachusetts; (2) Dallas, Texas; (3) Corinth, Mississippi; (4) Gloucester County, New Jersey; (5) Anaheim, California; and (6) San Francisco, California. We neither evaluated these housing authorities’ provision of affordable housing for persons with disabilities nor compared one authority’s efforts with those of another. Rather, our aim was to provide a comprehensive picture of housing authorities facing different challenges in providing housing for persons with disabilities. Our criteria for selecting the locations included the following: Size of the housing authority—Our case studies included large and small housing authorities. Size of the housing authority’s community—Our case studies included large cities, suburban communities, and a rural community. Condition of the housing market—Our case studies ranged from a city with a housing market known to be particularly tight to a community with a surplus of affordable housing. Designated housing and Section 8 assistance—Of the six locations, three PHAs had allocation plans for designated public housing and Section 8 certificates and vouchers; two housing authorities had no designated public housing but received certificates and vouchers (one was a housing authority that had no public housing units but administered Section 8 assistance); and one PHA neither had an allocation plan nor received certificates or vouchers for the exclusive use of persons with disabilities. We visited these six locations and interviewed the local housing authority officials about their experience in housing persons with disabilities. We also interviewed representatives of a judgmentally selected number of HUD-subsidized, privately owned projects that provided housing for the elderly and persons with disabilities to determine whether they established preferences for the elderly. Finally, we spoke with advocates for both the elderly and persons with disabilities in the communities we studied, where we were able to identify them. See appendix III for detailed information about the six case studies. To conduct the analysis of the population of low-income persons with disabilities, we asked the Department of Health and Human Services’ Office of the Assistant Secretary for Planning and Evaluation (ASPE) to conduct population analyses of data from the agency’s National Health Interview Survey, using HUD’s definition of low-income persons with disabilities as parameters. Bob Clark, Senior Program Analyst, and Don Chontos, Programmer, of ASPE developed the analysis. See appendix I for information on the population analysis. Finally, we reviewed the legislative history of the provisions in the 1992 act that allowed housing authorities to seek approval from HUD to designate housing for the elderly, persons with disabilities, or both and the appropriations set-asides for Section 8 certificates and vouchers in fiscal years 1997 and 1998. We interviewed appropriate HUD officials about designated housing, other housing for the elderly and persons with disabilities, and Section 8 certificates and vouchers (see app. V for a discussion on Section 8 certificates and vouchers). We also interviewed representatives of the national associations for persons with disabilities and the elderly. We conducted our review from July 1997 through May 1998 in accordance with generally accepted government auditing standards. To better describe the impact of the Housing and Community Development Act of 1992 on the availability of housing for persons with disabilities, we visited six locations. Three of the six have housing authorities with approved allocation plans, and two of these also have Section 8 certificates for persons with disabilities. Two other authorities in our case study do not have allocation plans but received vouchers. We selected the final location to review the practicality of certificates and vouchers for persons with disabilities in an extremely tight housing market. Table III.1 shows the six housing authorities we visited. Five of the six locations we visited have housing authorities that manage public housing units. Officials from all five told us that mixing elderly and nonelderly persons with disabilities in the same buildings does not always work well. In some cases it was the lifestyles associated with youth, not necessarily disability, that caused the underlying problems, and in other cases, it was younger persons with mental disabilities and their friends and visitors who caused problems for the elderly. All three housing authorities with allocation plans—Fall River, Dallas, and the Tennessee Valley Regional Housing Authority (TVRHA) in Corinth—designated entire buildings for the elderly. The Fall River authority designated six of its federally aided elderly/disabled buildings as elderly-only. It also has one building primarily for younger persons with disabilities, but officials there said they will accept elderly applicants for these units as well. The Dallas Housing Authority designated two of its four elderly/disabled buildings as elderly-only. The other two buildings are for mixed elderly/disabled occupancy. TVRHA in Corinth designated its one high-rise elderly/disabled building for the elderly and near-elderly. HUD awarded Section 8 certificates for those younger persons with disabilities in Dallas who wanted to move out of the buildings designated for the elderly into privately owned rental units in the community. HUD also required the Fall River Housing Authority to offer certain certificates it had been awarded previously to younger persons with disabilities. TVRHA officials told us that they had not requested additional certificates or vouchers for persons with disabilities because there was no need for them. The authority had no elderly or persons with disabilities on its waiting list. Officials at all three housing authorities stated that the elderly and persons with disabilities may apply for appropriately sized units in any of the other buildings the authorities manage. TVRHA reported that elderly and persons with disabilities constitute over two-thirds of their public housing residents. Housing authority officials in Fall River said they constitute almost 39 percent. Table III.2 shows the number of units designated for the elderly and for persons with disabilities as a result of the allocation plans, and also shows the number of units that are not designated but are considered elderly/disabled units. The impact of the allocation plans on affordable housing for persons with disabilities was minimal, at the time of our visits. Occupancy in the buildings designated for the elderly had not changed substantially since the housing authorities submitted their allocation plans, and where certificates and vouchers had been made available, persons with disabilities on waiting lists were able to use them to find housing. TVRHA allowed near-elderly persons with disabilities—those between the ages of 50 and 62—to move into its elderly-only building. Most of the persons with disabilities living in the units designated for the elderly in Fall River and Dallas chose to remain there rather than use the Section 8 certificates allocated for them, according to housing authority officials. In Fall River, where 76 certificates had been set aside for the 77 persons with disabilities occupying units in elderly-only buildings, 22 chose to use a certificate to move. In Dallas, which had 80 certificates allotted, 20 tenants used them. As for the remaining certificates, the Fall River authority offered them to younger persons with disabilities on its other waiting lists, and all have been used. Dallas officials intend to offer them to people from other developments but, at the time of our visit, had not yet made them available. They also told us that 49 additional persons with disabilities had requested the certificates but had subsequently returned them to the authority unused. The officials said they believed these residents decided to stay in their current units because they felt comfortable there, but we did not verify this assertion. Most of the advocates for persons with disabilities we spoke with told us that their clients should be able to choose to live wherever they want. Some prefer to live independently, and some like the security and amenities offered in buildings designed for the elderly. An advocate in Fall River did not consider the building reserved for persons with disabilities an attractive alternative and told us that the trend now is for this population to live in the community and arrange for support services of their own choosing. Similarly, some advocates for the elderly told us that the elderly should have choices and that they prefer to live among other elderly and not among younger people. Officials at all three housing authorities with approved allocation plans told us that the length of time the elderly and persons with disabilities remain on the waiting list is about the same as it was prior to the allocation plans. They said that the average time an applicant remains on the waiting list is between 60 to 90 days in Dallas, about 2 to 3 months in Fall River, and about 30 days in Corinth. While three of the six authorities we visited—Gloucester County, Anaheim, and San Francisco—do not have allocation plans, housing authorities in Gloucester County and Anaheim received vouchers for nonelderly persons with disabilities. The San Francisco housing authority had neither an allocation plan nor special certificates or vouchers for persons with disabilities at the time of our visit, but said it had considered the issue in the past and is likely to submit an allocation plan at some future point. The housing authorities in Gloucester County and Anaheim requested vouchers for persons with disabilities on their waiting lists, but for different reasons. The Housing Authority of Gloucester County manages a Section 8 project that it designated as elderly-only in 1993 because the project was originally built for the elderly. To offset the loss of units to persons with disabilities in this building, the housing authority applied for 130 vouchers—roughly enough to handle all persons with disabilities on its waiting lists. The Anaheim Housing Authority manages no public housing units, so it will use its 100 vouchers for persons with disabilities on its Section 8 waiting list. At the time of our visits, it was too early to determine the impact of these additional vouchers on affordable housing for persons with disabilities, as neither housing authority had been able to issue them all, officials there said. HUD awarded funding for Gloucester County’s vouchers in September 1997. Officials from the authority said they then hired and trained a staff person to assist persons with disabilities. As of January 1998, the authority had interviewed 80 applicants, had successfully utilized 10 vouchers to provide housing, and hoped to utilize 30 more by February 1. Officials from the Anaheim Housing Authority said they had not issued any vouchers at the time of our visit because funding had been awarded only recently. They said they expected all the vouchers to be used and planned to apply for more in the future because of the demand. They also said that with existing Section 8 certificates and vouchers, persons with disabilities and the elderly have been relatively easy to place compared with low-income families because they are less transient than the families that apply for assistance. The only difficulty the officials anticipated was in placing persons with mental disabilities. But, according to its voucher application, the authority plans to help individuals using these vouchers find housing and also work with local agencies to provide additional support services to these tenants. Officials from the San Francisco Housing Authority told us that the use of certificates and vouchers by persons with or without disabilities has been problematic because of the tight housing market there. Rents are very high and usually exceed what a certificate or voucher will cover, making it difficult to find an appropriate apartment using a Section 8 certificate or voucher, they said. Moreover, they have not opened their Section 8 waiting list since 1986, when about 10,000 households applied. An official from the Housing Authority of Gloucester County said that persons with disabilities who apply for certificates or vouchers return them unused at about the same rate as other people. Anaheim authority officials told us that they usually do not have problems placing the elderly or persons with disabilities. However, a San Francisco Housing Authority official said persons with disabilities often have special problems using the certificates or vouchers in that area because apartments are scarce, the process of looking for an apartment is challenging in the hilly city, and discrimination against persons with physical and mental disabilities is believed to be pervasive. In Gloucester County and Anaheim, housing authority officials said that for every three Section 8 certificates or vouchers issued, two are returned and one is used. Gloucester County attributed this high turnback rate, in some cases, to applicants being unable to afford the security deposits or utility bills. They also said that some persons with disabilities who live with their families do not want to move away from home, but take a certificate or voucher to please their families and then return it unused. Fall River is a city of about 91,000 located in the southeastern corner of Massachusetts. The Fall River Housing Authority’s allocation plan, approved in February 1996, designated six public housing buildings for the elderly, with a seventh building primarily serving younger persons with disabilities. When the plan was approved, HUD required the authority to offer Section 8 certificates exclusively to persons with disabilities to compensate for the housing units designated elderly-only. Fall River has a relatively depressed economy. Although vacancy rates in the private housing market are relatively high and rents are relatively low, the area’s high unemployment rate and low income levels mean that many households still pay a large proportion of their income in rent, putting Section 8 certificates and vouchers in high demand. When the waiting list for Section 8 assistance was last opened in September 1997, over 1,100 people applied in a single day. Because vacancy rates are high and landlords generally welcome Section 8 tenants, there is little difficulty with using certificates or vouchers. The housing authority administers 2,181 certificates and vouchers. Fall River has a surplus of public housing, with more than 2,500 public housing units in buildings funded by the federal or state government and nearly 1,700 units in private developments subsidized by the federal or state government. The overall vacancy rate is 9 percent, and waiting times for public housing are short—usually 2 to 3 months—for those who are willing to take the first available unit in any building. The housing authority has recently begun a marketing effort to attract more applicants to public housing. Table III.3 shows the types of public and assisted housing in the community. The housing authority’s allocation plan designated six of its seven federally aided elderly/disabled projects as elderly-only. At the time the plan was approved, 77 younger persons with disabilities lived in those projects. A seventh project, Cardinal Medeiros Towers, remained categorized as elderly/disabled, primarily serving persons with disabilities under the age of 62. A survey by the housing authority found that of the younger persons with disabilities living in the buildings designated elderly-only, 56 percent had physical disabilities and 44 percent had mental disabilities, which included mental illness, retardation, and substance abuse. Officials at the authority said that they decided to submit an allocation plan because of complaints from elderly residents that many of the younger residents played loud music, had undesirable visitors at all hours of the night, or behaved in a generally threatening or disruptive manner. Officials at the authority and an advocate for the elderly said that active hostility between the elderly and the younger residents with disabilities ran both ways, resulting in a tense living environment. They also said that the issue was one of age and not disability. None of the 77 younger persons with disabilities occupying units in elderly-only buildings was forced to move. The authority said it offered each a Section 8 certificate or the opportunity to move to other public housing units, including Cardinal Medeiros Towers, and also offered staff assistance and money to help with the move. Twenty-two of the 77 had moved out of the elderly-only buildings by November 1997, most using Section 8 certificates. Housing authority staff said that it has not been difficult for persons with disabilities to find appropriate apartments in Fall River using the certificates. The rest of the certificates were used by persons with disabilities who had been on the waiting lists for public housing or for Section 8 assistance. Elderly residents are very happy with the results of the allocation plan, according to housing authority officials, who also said that there have been no complaints from younger persons with disabilities. We spoke with a representative of an advocacy group for persons with disabilities who said she was concerned that the allocation plan did not do enough to ensure the availability of accessible housing for persons with physical disabilities. She also did not consider the one federal public housing option left open to younger persons with disabilities, Cardinal Medeiros Towers, to be an attractive alternative. Dallas is the eighth largest city in the United States, with a population of more than 1 million people. In January 1995, the Dallas Housing Authority submitted an allocation plan to designate two of its four elderly/disabled buildings as elderly-only. HUD did not respond within the 90 days established by law, so the plan was approved by default. To provide an opportunity for persons with disabilities living in designated housing to move into the community, the housing authority requested and received 80 Section 8 certificates. Dallas has a tight rental market, with a vacancy rate around 5.5 percent and an estimated average rental cost of $696 for a two-bedroom unit. About 46 percent of low- and moderate-income households pay more than 30 percent of their income for rent, and that figure rises to about 72 percent among those with very low incomes, according to HUD data. As of January 1998, 726 of the housing authority’s 4,647 public housing units in four developments were designated for the elderly only or for the elderly/disabled (mixed housing). The authority owns another four buildings with a total of 572 units that receive project-based Section 8 assistance and that are reserved for residents who are elderly or have disabilities. The authority also administers funds to pay for 30 Section 8 units in a property run by the Deaf Action Center. Table III.4 shows the number of HUD-assisted housing units available. The authority developed its allocation plan because concerns arose about mixing elderly residents and younger people with disabilities. Staff told us that some of the elderly residents at the mixed developments had raised noise and security issues. In addition, a resident survey indicated strong support for elderly-only housing. The impact of the housing authority’s allocation plan on persons with disabilities has been minimal thus far. Occupancy in the two buildings designated as elderly-only remained generally the same before and after the designation. The two developments designated, Audelia Manor and Park Manor, had been mixed elderly/disabled housing and had a total of 319 units. The authority requested and received 80 certificates to assist the younger tenants with disabilities expected to move out of the designated units. The authority also provided incentives, offering to pay for moving expenses, moving supplies, and telephone reconnection fees, but did not require anyone to move. At the time of our visit, 20 of the younger tenants with disabilities in the designated developments had used certificates to rent apartments in the private sector. Another 49 had requested certificates but returned them unused. Authority staff told us that many of these residents were reluctant to leave the security of their surroundings. Most still reside in the buildings. The housing authority received permission from HUD to offer the remaining certificates to people with disabilities residing at other developments and, as of April 1998, were offering them to residents at Forest Green and Lakeland Manor. To assist the residents with disabilities eligible for the 80 certificates, the Dallas Housing Authority generally used its regular Section 8 program to locate housing in the community. Under this program, the authority provides lists of landlords willing to rent to Section 8 tenants and takes prospective Section 8 families on van tours to properties that will accept certificates. To further assist the residents moving out of the developments newly designated for the elderly, the housing authority contacted local landlords to find those willing to rent to people with disabilities. Housing authority staff told us that landlords were typically more willing to accept Section 8 individuals who were physically disabled than they were to accept families receiving Section 8 assistance. Advocates for people with disabilities in Dallas said that the housing issues their clients face are usually related to accessibility rather than to the availability of assisted housing. Although the allocation plan did not result in as many of the younger residents with disabilities moving out of the buildings designated for the elderly as the housing authority had anticipated, officials said complaints from elderly residents had declined even though occupancy had not changed dramatically. Designating housing for the elderly has not affected waiting time for units. The waiting time for those needing an elderly/disabled unit is about 60 to 90 days, though the wait may be as long as 6 months if the applicant needs a studio or an accessible unit. This contrasts with an average waiting time of about 60 days for a regular public housing unit. The longer wait for elderly/disabled units stems from the lesser availability of those units, however, not from the impact of the allocation plan. Corinth is a city with a population of fewer than 12,000 in Alcorn County in northeastern Mississippi. Public housing is administered by both the Housing Authority of the City of Corinth and the Tennessee Valley Regional Housing Authority (TVRHA), which serves Alcorn and nine other counties. TVRHA also administers Section 8 certificates and vouchers for its 10-county area. In 1995, TVRHA submitted an allocation plan to designate all 50 units in the development in Corinth known as Fort Robinett Manor as elderly-only. HUD approved the plan on March 30, 1995. TVRHA designated Fort Robinett for elderly residents because 96 percent of its units were occupied by the elderly and the development has amenities ideally suited for the elderly. TVRHA officials also told us that mixing elderly and younger persons with disabilities is not always a good idea. TVRHA has not requested Section 8 certificates or vouchers for the exclusive use of persons with disabilities but administers 168 regular certificates and vouchers in Alcorn County. Corinth has 673 units of public housing, managed by two housing authorities. About 66 percent of the units are occupied by either elderly tenants or persons with disabilities, according to the two authorities. Housing authority officials and managers of private projects said that subsidized housing units are overabundant in Corinth, and managers are concerned about a high vacancy rate of around 11 percent. Corinth has more subsidized housing for its population than any place else in the world, according to an official from the city’s housing authority. The fair market rent for a one-bedroom unit is $290 per month; most units administered by TVRHA rent for $77 plus utilities. Table III.5 shows the categories of public and assisted housing in Corinth. City of Corinth public housing Section 8 certificates, TVRHA (Alcorn County-wide) Section 8 vouchers, TVRHA (Alcorn County-wide) The impact of TVRHA’s allocation plan on housing opportunities for public housing residents with disabilities appears to have been insignificant thus far. At Fort Robinett, occupancy had remained nearly the same. At the time the allocation plan was submitted, 49 of the 50 units were occupied by elderly households. As of November 1, 1997, 44 units were occupied by elderly households, 3 by near-elderly persons with disabilities, and 2 by near-elderly persons without disabilities, for a total of 49 occupied units. When selecting applicants for handicapped-accessible units, TVRHA gives priority to families that include disabled persons who can benefit from those features. TVRHA officials told us they had not requested additional Section 8 certificates and vouchers to offset the impact of designating housing for the elderly because there were no persons with disabilities on the authority’s waiting list. In addition to the handicapped units at Fort Robinett, TVRHA has converted 17 more units for persons with disabilities in Corinth—10 are wheelchair-accessible and 7 are for those who have visual or hearing impairments. Authority officials said that these additional units, along with the normal turnover rate, provide sufficient housing for the younger persons with disabilities who are no longer eligible to live at Fort Robinett. Moreover, the allocation plan states that if any of the designated units are vacant for over 60 days, they will be made available to younger applicants. For Section 8 certificates and vouchers, TVRHA reported that persons with disabilities and the elderly are given priority over single applicants who are not elderly and who do not have disabilities. Renting units with Section 8 assistance is generally not a problem for persons with disabilities, as they usually have a landlord in mind who has been contacted and is willing to work with them, according to TVRHA. However, few landlords are willing to make expensive modifications for tenants they are unfamiliar with because they do not know how long the tenants will stay. A TVRHA official told us that as a result, it may sometimes take persons with disabilities a little longer to secure a lease if units need various changes to comply with HUD standards. TVRHA officials said that persons with disabilities usually prefer public housing instead of certificates or vouchers because the public housing units are better equipped for their needs. An official from the Housing Authority of the City of Corinth told us that the authority had no need to designate housing for the elderly. It operates a 100-unit development for the elderly and persons with disabilities. At the time of our visit, 17 of the units were occupied by persons with disabilities, but most of these tenants were between 50 and 58 years old and fit in well with the older residents. He also told us that younger persons with disabilities do not generally apply to live in this building. Were the situation to change and problems were to develop, the housing authority might consider submitting an allocation plan to designate housing, the official said. Gloucester County, New Jersey, with a population of about 230,000, is located within the highly urbanized Philadelphia metropolitan area, although many parts of the county are rural or small-town in nature. In 1990, about 35 percent of the county’s 79,000 households had low or moderate incomes, according to data provided by Gloucester County. Because housing costs had risen much faster than household incomes, almost 22 percent of all households paid over 30 percent of their gross income for housing costs. The overall housing vacancy rate in the county was 4.4 percent. In 1997, HUD awarded the Housing Authority of Gloucester County 130 Section 8 vouchers for persons with disabilities as an alternative to units in a building the authority had designated for the elderly in 1993. The authority submitted an allocation plan in 1997 to designate additional housing for the elderly, but HUD denied the plan because it lacked necessary information. However, in March 1998, HUD conditionally approved the authority’s revised allocation plan subject to certain revisions. The authority has 262 public housing units and manages another building with 199 units that receive Section 8 project-based rental assistance. With the exception of 20 units for people with mobility impairments, the authority designated this 199-unit building for the elderly in 1993. To compensate for any loss of housing for persons with disabilities, the authority applied for and received 130 additional vouchers because 133 younger persons with disabilities were on its waiting lists for other programs. Of the authority’s 1,485 regular Section 8 certificates and vouchers, 260 were used by persons with disabilities and 252 by elderly persons at the time of our visit. In addition to the Housing Authority of Gloucester County, the Housing Authority of the Borough of Glassboro also administers public housing and Section 8 certificates in Gloucester County. Table III.6 shows the number of subsidized units in the county. Public housing (Housing Authority of Gloucester County) Public housing (Housing Authority of the Borough of Glassboro) Section 8 certificates and vouchers, including 130 vouchers for persons with disabilities (Housing Authority of Gloucester County) Section 8 certificates (Housing Authority of the Borough of Glassboro) State agency certificates and vouchers (State of New Jersey) Besides the units already designated for the elderly, the allocation plan the authority submitted unsuccessfully in 1997 would have limited two high-rise buildings with a total of 200 units to elderly and near-elderly tenants aged 50 and over. The authority’s director told us he plans to modify the plan and resubmit it to HUD for approval. Advocates both for the elderly and for persons with disabilities said the housing problem in Gloucester County was a lack of affordable housing, not necessarily mixing the two populations. Overall, advocates for the disabled said that the younger persons with disabilities should have the freedom to choose where they live because not all of them want to live with the elderly or with other persons with disabilities. However, this group may want to select units in buildings for the elderly because many offer important amenities, such as security, transportation, and supportive services. An official from the county office for the disabled said that security is especially important for people who have visual or hearing impairments. At the same time, younger persons with disabilities who are capable of living independently usually prefer to do so. An official from the department on aging said that she believes the elderly prefer to live among other elderly because of the conflicting lifestyles of younger persons with disabilities. At the time of our visit, it was too soon to determine the impact of the special vouchers on the supply of affordable housing for persons with disabilities because the authority was just beginning to issue them. To administer the 130 vouchers after HUD awarded funding in September 1997, the authority hired a full-time staff person, who had then spent several months in training before working with persons with disabilities full time. As of January 1, 1998, 40 vouchers had been provided to persons with disabilities, 10 of which had been successfully used to obtain housing. An official said it was hoped the other 30 recipients would have housing by February 1. The number of persons with disabilities on the authority’s waiting list has increased in the time since the authority applied for the 130 vouchers. Staff said they would apply for additional vouchers if necessary. The turnback rate for Section 8 certificates and vouchers issued to persons with disabilities is about the same as for other Section 8 applicants—for every three certificates or vouchers issued, two are returned, according to authority officials. Among the reasons they cited were applicants’ not being able to afford security deposits or utility bills and some applicants’ requesting rental assistance as a token effort to please their families, then returning the certificates or vouchers unused. Anaheim is a city with a population of approximately 300,000 in Orange County in southern California. The Anaheim Housing Authority administers no public housing and therefore has no allocation plan, but it provides other housing assistance, including Section 8 rental certificates and vouchers. Housing authority officials told us they applied for 100 vouchers exclusively for persons with disabilities because they had a high number of these people on their waiting list for housing assistance—about 1,500 of the 6,000 individuals listed. The authority recently received those vouchers from HUD’s 1997 funding for Mainstream Housing Opportunities for Persons With Disabilities (the mainstream program). The need for affordable housing in Anaheim, particularly for the elderly and for persons with disabilities, is great, according to officials of the housing authority and of advocacy groups we spoke with. HUD data indicate that 70 percent of the city’s elderly residents pay more than 30 percent of their income for rent, with many paying more than 50 percent. According to an advocacy group for persons with disabilities, its clients’ incomes are about $600 a month, making the typical rents for a one-bedroom apartment of $580 to $640 per month unaffordable. Subsidized private housing projects for the elderly and for persons with disabilities provide 391 affordable units in Anaheim. Of the approximately 3,400 low-income households receiving Section 8 certificates and vouchers in Anaheim in 1995, about 1,300 were elderly or persons with disabilities. At the time of our visit, the Anaheim Housing Authority had not yet issued any of the 100 vouchers HUD had awarded under the mainstream program for persons with disabilities. Housing authority officials said they expected to use all of them, however, and planned to apply for more vouchers in the future because of the demand. The effects of the 1992 act in Anaheim have been positive for persons with disabilities, according to the housing authority officials, who believed it had increased the number of affordable housing opportunities for this group. Because of the mainstream program vouchers, they said they expected to be housing persons with disabilities on their waiting list at a faster rate than other applicants. On the basis of their previous experience with regular certificates and vouchers, housing authority officials said they do not anticipate difficulties or delays in using the mainstream program vouchers. They said persons with disabilities and the elderly have been relatively easy to place compared with low-income families, which tend to be more transient and often move out of the Anaheim area before they use their certificates or vouchers. The housing authority officials anticipated greater difficulty with placing persons with mental disabilities. According to these officials, evictions have been more common with this particular group than with others, but they believe that this problem can be averted if these tenants receive the necessary support services. To assist individuals in using the vouchers, the housing authority plans to assemble a list of accessible rental units, provide assistance to landlords willing to make modifications to make their properties accessible, assign a full-time housing counselor to help the mainstream program participants, and work with local nonprofit agencies to get tenants additional support services. The mainstream vouchers will be offered first to the persons with disabilities who have been on the housing authority’s waiting list the longest. Officials of the housing authority and the advocacy group for the disabled with whom we spoke generally agreed that “mainstreaming” is the best housing option for persons with disabilities, although they did not see a particular problem in housing elderly and persons with disabilities together. Housing authority officials qualified this by saying that mainstreaming is usually the best option for persons with physical disabilities, whereas persons with mental and developmental disabilities usually do better in group settings where they get the emotional support and the services that they need. Officials we spoke with at an advocacy group for persons with disabilities fully supported mainstreaming in all cases because they believed that segregation perpetuates stereotypes and prevents persons with disabilities from believing that they can take care of themselves or work. San Francisco is a racially and ethnically diverse city of about 760,000 people in northern California. The San Francisco Housing Authority has not submitted an allocation plan and has not obtained certificates or vouchers exclusively for persons with disabilities. While the authority administers almost 5,000 certificates and vouchers, San Francisco’s housing market is extraordinarily tight and finding landlords willing to accept Section 8 applicants—with or without disabilities—has become a serious problem, according to housing authority officials. The vacancy rate for rental units is very low, and rents have gone up substantially in the past 2 years, with the median market rent for a one-bedroom apartment in larger buildings at over $1,500. According to information provided by the Mayor’s Office of Housing, affordable housing is thus extremely scarce in San Francisco, and 55 percent of low-income and 73 percent of very low-income renter households pay more than 30 percent of their incomes in rent. Overcrowding and homelessness are also significant problems. Advocates for persons with disabilities with whom we spoke universally described the housing situation for persons with disabilities in San Francisco as extremely dire. The housing stock is generally old, and units tend to be small. The topography of the city, with its many hills, limits accessibility for persons with disabilities and makes apartment-hunting difficult. Less than 2 percent of the city’s private rental stock is estimated to be minimally accessible to people with physical disabilities. The situation is particularly difficult because discrimination against persons with both physical and mental disabilities is believed to be pervasive. The San Francisco Housing Authority has 43 developments with a total of 6,722 units of public housing serving 12,436 residents. Twenty-one of these developments, serving 2,165 residents, are intended to be elderly/disabled housing. Roughly 9 percent of the residents of these developments are persons with disabilities who are under the age of 60. The housing authority settled a class-action lawsuit in 1991 by agreeing to increase its number of handicapped-accessible units. A Special Master appointed by the court is assisting in the implementation of the settlement. At the time of our visit, the authority had only 144 accessible public housing units, but it was adapting several hundred more and expected to have 690 accessible or adaptable units within a few years. In addition to public housing, San Francisco has 9,892 project-based units in 87 developments that either have federally insured mortgages or receive federal subsidies. The San Francisco Housing Authority also administers 4,945 Section 8 certificates and vouchers. The authority has not applied for any certificates or vouchers since about late 1995 because of various internal problems, an official said. With the recent tightening of the housing market, finding landlords willing to rent to people with Section 8 certificates and vouchers has become problematic, in large part because HUD’s fair market rents are considered too low, according to both housing authority officials and advocates. Authority staff said that in their experience, about 15 percent of those receiving certificates or vouchers turn them back, mostly because they cannot find affordable rental units; another 30 percent or so have to move outside of San Francisco to find affordable housing. HUD is currently collecting data and considering modifications to its fair market rent in San Francisco. The waiting lists for all types of subsidized housing are very long, according to authority officials. About 11,600 people are on the waiting list for public housing, with an average wait of at least 2.5 to 3 years. For persons with disabilities, the wait is slightly longer because of a shortage of studio and one-bedroom apartments, though for persons requiring handicapped-accessible units the wait is slightly shorter. The waiting list for certificates and vouchers has not been opened since 1986, when about 10,000 people applied; as of February 1998, the authority was still working off that list. The wait for units in the project-based Section 8 developments we visited was at least 3 to 5 years. The authority had put together an exploratory committee in the early 1990s to consider the issue of designating certain buildings of its own as elderly-only, but the committee fell to the wayside because of other pressing concerns. The authority has been working with a HUD recovery team, and authority officials said that once the current restructuring is complete, the issue will be revisited and an allocation plan will likely be submitted to HUD. The authority first explored the possibility of submitting an allocation plan to designate housing because of a perceived conflict between elderly residents and younger persons with disabilities and a strong consensus among the elderly tenants that they would prefer to live among other seniors. Elderly residents complained that younger tenants played loud music; had guests at all hours of the night; and displayed intimidating, threatening, or bizarre behavior. The main source of concern was younger residents who were mentally ill. Housing authority officials estimated that 70 to 80 percent of the younger residents with disabilities had mental rather than physical disabilities. When the idea of designating buildings as elderly-only was initially discussed, many of the younger tenants with disabilities expressed strong concerns. We spoke with representatives of five advocacy organizations for persons with disabilities. Some said that the concerns about mixing the elderly and younger persons with disabilities were misguided and based largely on overreaction and misunderstanding. In general, they feared that elderly-only designations would reduce the housing options for persons with disabilities. This is a special concern in San Francisco, they said, where public housing and project-based Section 8 housing are the only viable alternatives for many persons with disabilities because it is so difficult to find affordable and accessible housing using a certificate or voucher. The advocates opposed segregating persons with disabilities in housing designated disabled-only, which they perceived as isolating, unfair, and antithetical to the ideal of integrating persons with disabilities into mainstream society. Since passage of the Housing and Community Development Act of 1992, HUD has made Section 8 certificates and vouchers for persons with disabilities available in two ways: in connection with designated housing or through a mainstream housing opportunities program. The funds for these certificates and vouchers, totaling $278.9 million, came from HUD’s administrative set-asides and from congressional set-asides in HUD’s appropriations. The purpose of Section 8 certificates and vouchers awarded in connection with designated housing is to offset the effects of reserving units in public and privately owned, HUD-subsidized housing for elderly residents. To receive certificates and vouchers in connection with designated public housing, a PHA must have a HUD-approved allocation plan. While HUD is more inclined to award certificates and vouchers to a PHA that has designated units for the elderly that were previously available to younger people with disabilities, HUD officials told us that they look at other factors as well in determining how many certificates and vouchers a PHA should receive in connection with designated housing. These factors include the PHA’s vacancy rates, the demand for affordable housing by persons with disabilities, and the availability of other types of housing to persons with disabilities, including efficiencies and one-bedroom units in the PHA’s family developments. The number of certificates or vouchers HUD awards a PHA is not a one-for-one replacement for the units that the PHA has designated as elderly-only. Moreover, designating units as elderly-only does not guarantee that a PHA will receive additional certificates and vouchers. These certificates and vouchers have budget authority terms of 1, 2, and 5 years, depending on the funding. In communities where privately owned, HUD-subsidized projects have established preferences for the elderly, the housing authorities can apply for certificates and vouchers if they can identify the number of nonelderly persons with disabilities who are not receiving housing assistance as a result of these preferences. To prepare their applications, the housing authorities may seek assistance from their local HUD field offices to identify these privately owned projects. These certificates and vouchers have 1-year terms. The mainstream housing program is intended to provide persons with disabilities greater opportunities to find affordable housing of their choice in their communities (hence, the term “mainstreaming”). Any housing authority that administers the Section 8 rental assistance program is eligible to apply for the mainstream certificates and vouchers and may issue them to any persons with disabilities eligible for Section 8 assistance. The requirements for the mainstream certificates and vouchers are less restrictive than those for certificates and vouchers connected with designated housing, and PHAs are not required to submit allocation plans to qualify. The mainstream program’s certificates and vouchers have terms of 5 years. As of November 1, 1997, HUD had made available $190.4 million of the total $278.9 million set aside since 1992 for certificates and vouchers for persons with disabilities. To do so, HUD issued four notices of funding availability: one in March 1995, another in October 1996, and two in April 1997. The remaining $88.5 million was from HUD’s fiscal year 1998 appropriation: $40 million of it to be used in connection with designated public and private housing and $48.5 million to be used for the mainstream program. HUD issued a fifth notice of funding availability on April 30, 1998, for these funds and for the funds remaining from the fiscal year 1997 appropriation of $50 million. As table V.1 shows, only the funds made available through two of the notices have been totally awarded: the $13.3 million made available through the March 1995 notice and the $48.5 million made available through the April 1997 notice for the mainstream program. Of the $78.6 million HUD made available through its October 1996 notice, $48.7 million had been awarded as of November 1, 1997, and HUD was continuing to make awards. HUD said that between November 1, 1997, and March 1, 1998, it had awarded another $15 million from the October 1996 notice to PHAs with newly approved allocation plans. HUD expected to award the remainder of the funds by the end of fiscal year 1998. Table V.1: HUD Notices of Funding Availability for Persons With Disabilities, as of November 1, 1997 Total funds available (millions) Total awarded (millions) Of the $50 million made available under HUD’s 1997 notice made in connection with designated public and private housing, $2.7 million had been awarded as of November 1, 1997. The funds made available through the notice were set aside by the Congress in HUD’s fiscal year 1997 appropriation. Half of the money was earmarked for PHAs with approved allocation plans, the other half for housing authorities that could identify the impact that elderly preferences established by privately owned projects in their communities had on persons with disabilities. As of November 1, 1997, none of the $25 million for PHAs with approved allocation plans had been awarded. HUD officials said the money will be used once all the funding from the October 1996 notice is awarded. Of the $25 million earmarked to offset elderly preferences established by privately owned projects, HUD had awarded $2.7 million to five PHAs that were able to identify the impact the preferences had on persons with disabilities. As mentioned previously, the funding made available under the other notice HUD issued in April 1997—the notice for the mainstream program—had been totally awarded by November 1, 1997. The $48.5 million made available through this notice came from the fiscal year 1997 funds appropriated for the Section 811 program for the capital development of privately owned projects for persons with disabilities. According to HUD, 281 PHAs submitted approvable applications for this disproportionately more popular program, but HUD was able to award funds to only 25, providing them with 1,756 certificates and vouchers. John T. McGrail The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a legislative requirement, GAO reviewed: (1) the impact of the Housing and Community Development Act of 1992 on the availability of public housing for persons with disabilities; (2) how incremental Section 8 certificates and vouchers that were made available since the passage of the 1992 act were assisting persons with disabilities seeking affordable rental housing; and (3) the number of households that may meet the Department of Housing and Urban Development's (HUD) definition for persons with disabilities. GAO noted that: (1) provisions of the Housing and Community Development Act of 1992 allowing public housing authorities to designate units as elderly-only have had little impact on the availability of public housing for people with disabilities; (2) 73 of the 3,200 public housing authorities had allocation plans approved by HUD as of November 1, 1997, allowing them to designate 24,902 of their units as elderly-only, approximately 36 percent of their housing stock for the elderly and persons with disabilities; (3) nearly all of these designated units had been available previously to tenants who were elderly or who had disabilities but were younger than 62; (4) the number of elderly residents and residents with disabilities in these and other housing units for which they were eligible had not changed substantially since the housing authorities began submitting allocation plans; (5) the number of younger tenants with disabilities living in housing designated for the elderly had declined by about 1,100 at the 53 housing authorities that provided complete occupancy data; (6) designating public housing units as elderly-only may have more impact in the future, depending on how many more housing authorities opt to do so and on what the housing alternatives are for younger people with disabilities; (7) it is too soon to determine the extent to which the Section 8 rental certificates and vouchers set aside for persons with disabilities have helped meet housing needs; (8) of approximately 3,000 certificates and vouchers that were available to housing authorities November 1, 1997, the authorities reported that they issued about 1,600 to persons with disabilities who used 1,162 to obtain private rental housing; (9) about 18 percent of the users had been living in public housing that had been designated for the elderly--indicating little movement by persons with disabilities residing in housing now designated as elderly-only; (10) how successful rental certificates and vouchers will be in providing housing alternatives for people with disabilities will be influenced by several factors, including statutory restrictions, local housing markets, and willingness of tenants with disabilities to use certificates and vouchers; and (11) according to housing authorities, those persons with disabilities who used certificates and vouchers required greater assistance than other recipients.

Every time responsibility for cargo changes hands along the global supply chain there is the potential for a security breach. As a result, vulnerabilities exist that terrorists could take advantage of by, for example, placing a WMD into a container bound for the United States. While there have been no known incidents of containers being used to transport WMD, criminals have exploited containers for other illegal purposes, such as smuggling weapons, people, and illicit substances. To address the potential security risks posed by the millions of containers that arrive in the United States each year, CBP has implemented a layered security strategy of related initiatives and programs that focus CBP’s limited resources on potentially high-risk cargo bound for the United States while allowing other cargo to proceed without unduly disrupting commerce. Key elements of CBP’s maritime cargo security initiatives and programs are described below. Automated Targeting System. Information on shipments destined for the United States is automatically fed into CBP’s Automated Targeting System (ATS)—an enforcement and decision support system that compares cargo information against intelligence and other law enforcement data. ATS consolidates data from various sources to create a single, comprehensive record for each U.S.-bound shipment. ATS uses a set of rules that assess different factors in the information to determine the risk level of a shipment. One set of rules within ATS, referred to collectively as the maritime national security weight set, is programmed to check for information or patterns that could be indicative of suspicious or terrorist activity. ATS uses this weight set to assess and generate risk scores for every cargo shipment as the shipment moves throughout the global supply chain and new information is provided or existing information is revised. CBP classifies the risk scores from the maritime national security weight set as low, medium, or high risk. ATS automatically places high-risk shipments on hold, and CBP officials use information in ATS to identify (target) which high-risk shipments should be examined or waived. To assist in its targeting efforts, CBP uses key information about shipments destined for the United States obtained through the 24-hour rule and the 10+2 rule. Through the 24-hour rule, CBP generally requires vessel carriers to electronically transmit cargo manifests to CBP 24 hours before cargo is loaded onto U.S.-bound vessels at foreign ports. Through the Importer Security Filing and Additional Carrier Requirements (known as the 10+2 rule), CBP requires importers and vessel carriers to provide data elements for improved identification of cargo shipments that may pose a risk for terrorism. Importers are responsible for supplying CBP with 10 shipping data elements—such as country of origin—24 hours prior to loading, while vessel carriers are required to provide 2 data elements—container status messages and stow plans—that are not required by the 24-hour rule. Container Security Initiative. CSI is a bilateral government partnership program operated by CBP that aims to identify and examine U.S.-bound cargo container shipments that are at risk of containing WMD or other terrorist contraband. As part of the program, CBP officers are stationed at select foreign seaports and review information about U.S.-bound containerized cargo shipments. CBP uses ATS to target U.S.-bound container shipments and request examinations of high-risk container shipments before they are loaded onto vessels. CSI is operational at ports in North America, Europe, Asia, Africa, the Middle East, and Latin and Central America. CBP estimates that, through the CSI program, it prescreens over 80 percent of all maritime containerized cargo imported into the United States. Secure Freight Initiative. In response to a requirement in the SAFE Port Act to scan 100 percent of U.S.-bound cargo containers, CBP established SFI. CBP uses radiation detection and non-intrusive inspection equipment to scan cargo containers before they are loaded onto vessels at select foreign seaports Radiation detection equipment, such as radiation portal monitors (RPM) and radiation isotope identification devices (RIID) detect the presence of radioactive material that may be in a container. RIIDs and certain types of RPMs can identify the specific radioactive isotope being emitted and whether the radiation is a threat or is naturally occurring, such as that found in certain ceramic tiles. The second type of equipment, referred to as non-intrusive inspection equipment, uses X-rays or gamma rays to scan a container and produce images of a container’s contents without having to open it. Customs-Trade Partnership Against Terrorism. C-TPAT is a voluntary, public-private sector partnership with private stakeholders in the international trade community that aims to secure the flow of maritime cargo bound for the United States. Through C-TPAT, CBP officials work with member private companies to review the security of their supply chains to ensure their security practices meet CBP’s minimum security criteria. In return, C-TPAT members receive various benefits, such as reduced scrutiny of their shipments. Figure 1 provides an overview of the global supply chain and the steps in the supply chain where CBP’s key initiatives and programs come into play. Our prior work has shown that CBP has made substantial progress in implementing various initiatives and programs that, collectively, have enhanced cargo security, but some challenges remain. Examples of progress and challenges in the areas of (1) using information for improving targeting and risk assessment of cargo shipments, (2) partnerships with foreign governments, and (3) partnerships with the trade industry are discussed below. In January 2015, we found, among other things, that CBP did not have accurate data on the number and disposition of each high-risk maritime cargo shipment scheduled to arrive in the United States. On the basis of our analyses of CBP data for fiscal years 2009 through 2013, we found that, on average each year, approximately 11.6 million maritime cargo container shipments arrived in the United States, and less than 1 percent of those shipments were determined by ATS to be high-risk. We found that CBP examined the vast majority of high-risk shipments, but CBP’s data on the disposition of high-risk shipments were not accurate because of various factors, such as the inclusion of shipments that were never sent to the United States. Further, our analyses found that CBP’s data overstated the number of high-risk shipments, including those that appeared not to be resolved (examined or waived) in accordance with CBP policy. We also found that when determining the disposition of high- risk shipments, CBP officers were inconsistently applying criteria to make some waiver decisions and were also incorrectly documenting the reasons for waivers. As a result, we concluded that CBP could not accurately determine the extent to which waivers were used consistently and judiciously across CBP targeting units, as required by policy. We recommended, among other things, that CBP define waiver categories and disseminate policy on issuing waivers for high-risk shipments. DHS concurred with our recommendations and, in December 2015, CBP issued a new policy, National Security Cargo Targeting Procedures, that includes criteria for waiving mandatory examinations of high-risk shipments (referred to as exceptions). The new policy also specifically identifies certain types of shipments that do not qualify for exceptions to examination requirements. In addition, CBP developed a new process for recording waivers and issued a memorandum to targeting units on how to apply the new procedures. CBP’s actions help ensure that all of its targeting units are correctly and consistently applying and documenting waivers. In October 2012, we found that more regular assessments of ATS were needed to enhance CBP’s targeting of maritime cargo and better position CBP to provide reasonable assurance of the effectiveness of ATS. We, therefore, recommended that CBP (1) ensure that future updates to the rules that identify risks are based on results of assessments that demonstrate the effectiveness of such updates; and (2) establish targets for CBP’s performance measures and use those measures to assess the effectiveness of ATS on a regular basis to better determine when updates to the rules that identify risks are needed. DHS concurred with the recommendations and, in May 2015, CBP revised its National Security Weight Set, Maritime Standard Operating Procedures (SOP) to address the new requirements for the maintenance, review, and update of the national security weight set in ATS. The SOP requires program managers to compare proposed versions of the national security weight set against the existing version as part of the process for determining whether to implement a proposed new version of the weight set. Doing so will help provide reasonable assurance that changes to the weight set will improve the effectiveness of CBP’s targeting of maritime cargo container shipments. The SOP also establishes a performance measure and an associated target that will assist CBP in determining whether the weight set is effectively targeting maritime cargo container shipments. The SOP requires CBP to review the national security weight set for revisions if the weight set does not meet the performance target in two consecutive quarters. By assessing the weight set regularly against a performance target, CBP will be better positioned to determine when updates to the weight set are needed to ensure continued effectiveness in targeting of high-risk maritime cargo container shipments. In September 2010, we reviewed CBP’s efforts to collect additional data through the 10+2 rule and utilize these data to identify high-risk shipments. We found that the 10+2 rule data elements were available for identifying high-risk cargo, but CBP had not yet finalized its national security targeting criteria to include these additional data elements to support high-risk targeting. We recommended that CBP establish milestones and time frames for updating the targeting criteria. In December 2010, CBP provided us with a project plan for integrating the data into its criteria, and in January 2011, CBP implemented the updates to address risk factors present in the 10+2 data. We are currently reviewing CBP’s implementation and enforcement of the 10+2 program and anticipate issuing our report in spring 2017. In September 2013, we reported on CBP’s progress in implementing CSI. Specifically, we found that CBP had not regularly assessed foreign ports for risks to cargo under the CSI program since 2005. While CBP took steps to rank ports for risks in 2009, we found that CBP did not use results from this assessment to make modifications to the locations where CSI staff are posted because of budget cuts. By applying CBP’s risk model to fiscal year 2012 cargo shipment data, we found that CSI did not have a presence at about half of the foreign ports CBP considered high- risk, and about one-fifth of the existing CSI ports were at lower-risk locations. We recommended that DHS periodically assess the supply chain security risks from all foreign ports that ship cargo to the United States and use the results of these risk assessments to inform any future expansion of CSI to additional locations and determine whether changes need to be made to existing CSI ports and make adjustments as appropriate and feasible. DHS concurred with our recommendation and, in response, CBP developed a CSI Port Risk Matrix and Port Priority Map. CBP officials stated that the matrix and map will be used, along with several other tools available to CSI, to assess whether changes need to be made to CSI ports worldwide. According to CBP, these tools are to be updated yearly and, if necessary, can be updated more frequently based on significant changes, emerging threats, and intelligence. As a result of developing and employing these new risk-assessment tools, CBP should be better positioned to ensure that it is allocating its resources to provide the greatest possible coverage of high-risk cargo to best mitigate the risk of importing WMD or other terrorist contraband into the United States through the supply chain. In October 2009, we reported that scanning operations at the initial SFI ports encountered a number of challenges—including safety concerns, logistical problems with containers transferred from rail or other vessels, scanning equipment breakdowns, and poor-quality scan images. Both CBP and GAO had previously identified many of these challenges, and CBP officials were concerned that they and the participating ports could not overcome them. Senior DHS and CBP officials acknowledged that most, if not all foreign ports, would not be able to meet the July 2012 target date for scanning all U.S.-bound cargo, and DHS would need to issue extensions to such ports to allow the continued flow of commerce in order to remain in compliance with relevant statutory requirements. We recommended that DHS, in consultation with the Secretaries of Energy and State, develop, among other things, more comprehensive cost estimates, conduct cost-benefit and feasibility analyses, and provide the results to Congress. In response to our recommendations, CBP stated it had no plans to develop comprehensive cost estimates or feasibility analyses since SFI is operating at one port and it had no funds to conduct such analyses. In July 2013, we closed these recommendations as not implemented. In May of 2012, 2014, and 2016, the Secretary of Homeland Security authorized a 2-year extension of the deadline for implementing the 100 percent scanning requirement for U.S. bound cargo before it is loaded onto vessels at foreign seaports. In May 2014, the Secretary of Homeland Security renewed the extension (until July 2016) and stated that “DHS’s ability to fully comply with this unfunded mandate of 100 percent scanning, even in long term, is highly improbable, hugely expensive, and in our judgment, not the best use of taxpayer resources to meet this country’s port security and homeland security needs.” The Secretary also stated that he instructed DHS, including CBP, to do a better job of meeting the underlying objectives of the mandate. In the most recent letter, dated May 2016, authorizing the extension until July 2018, the Secretary stated he has committed the Department to work towards meeting the mandated 100 percent scanning requirement. The Secretary also outlined steps DHS is taking to engage stakeholders to identify solutions by leveraging the private sector. DHS plans to assess the feedback it receives during the summer of 2016 and will subsequently seek to test viable solutions in operational environments. In April 2008, we reported, among other things, that CBP took steps to improve the process for validating C-TPAT applicants’ security practices and implemented numerous actions to address C-TPAT management and staffing challenges. However, we found challenges with the technology CBP used to help ensure that validation information is consistently collected, documented, and uniformly applied to decisions regarding the awarding of benefits to C-TPAT members, and that CBP lacked a systematic process to ensure that members take appropriate actions in response to security validation findings. We also found that C- TPAT’s performance measures were insufficient to assess the impact of C-TPAT on increasing supply chain security. We made recommendations to CBP to strengthen C-TPAT program management and oversight. Specifically, we recommended, among other things, that CBP document key data elements needed to track compliance with the SAFE Port Act and other CBP internal requirements and to identify and pursue opportunities in information collected during C-TPAT member processing activities that may provide direction for developing performance measures of enhanced supply chain security. CBP has since implemented these recommendations by, for example, creating an automated platform to track and capture the content and communication between CBP and C- TPAT members to ensure that C-TPAT validation report recommendations are implemented and identifying analytical tools and data for trend analysis to better assess C-TPAT’s impact on the supply chain. We are currently reviewing the C-TPAT program, specifically how CBP assesses member benefits and conducts security validation responsibilities. We anticipate issuing our report in late fall 2016. Thank you Chairman Hunter, Chairwoman McSally, Ranking Members Garamendi and Vela, and Members of the Subcommittees. This completes my prepared statement. I would be happy to respond to any questions you may have at this time. For questions about this statement, please contact Jennifer Grover at (202) 512-7141 or groverj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Individuals making key contributions to this statement include Christopher Conrad (Assistant Director), Carla Brown, Lisa Canini, Michele Fejfar, Eric Hauswirth, Heidi Nielson, Ashley Rawson, and Natarajan Subramanian. Key contributors for the previous work that this testimony is based on are listed in those products. This appendix describes the key initiatives and programs related to U.S. Customs and Border Protection’s (CBP) strategy for ensuring the security of maritime cargo. CBP has developed this strategy to mitigate the risk of weapons of mass destruction, terrorist-related material, or other contraband from being smuggled into the United States. CBP’s strategy is based on related initiatives and programs that attempt to focus resources on high-risk shipments while allowing other cargo shipments to proceed without unduly disrupting the flow of commerce into the United States. The strategy includes obtaining cargo information on shipments in advance of their arrival at U.S. ports to identify high-risk shipments, using technology to inspect cargo, and partnering with foreign governments and members of the trade industry. Table 1 provides a brief description of some of the key initiatives and programs that compose this security strategy.

The U.S. economy is dependent on the expeditious flow of millions of tons of cargo each day through the global supply chain—the flow of goods from manufacturers to retailers. Criminal or terrorist attacks using cargo shipments can cause disruptions to the supply chain and can limit global economic growth and productivity. Within DHS, CBP has responsibility for administering maritime cargo security measures and reducing the vulnerabilities associated with the supply chain. CBP has developed a layered security strategy that focuses its limited resources on targeting and examining high-risk cargo shipments that could pose a risk while allowing other cargo shipments to proceed without unduly disrupting commerce arriving in the United States. This statement discusses the progress and challenges associated with CBP's implementation of initiatives and programs responsible for enhancing the security of the global supply chain. The statement is based on reports and testimonies GAO issued from April 2008 through January 2015 related to maritime cargo security—with selected updates on how DHS has responded to GAO's prior recommendations. The Department of Homeland Security (DHS) and U.S. Customs and Border Protection (CBP) have made substantial progress in implementing initiatives and programs that, collectively, have enhanced cargo security, but some challenges remain. Examples of progress and challenges are discussed below. Risk Assessments of Cargo Shipments . In January 2015, GAO found that CBP did not have accurate data on the number and disposition of each high-risk shipment scheduled to arrive in the United States. Specifically, CBP's data overstated the number of high-risk shipments, including those that appeared not to be examined or waived in accordance with CBP policy. CBP officers inconsistently applied criteria to make some waiver decisions and incorrectly documented waiver reasons. GAO recommended that CBP define waiver categories and disseminate policy on issuing waivers. In response, CBP issued a new policy that includes criteria for waiving examinations of high-risk shipments and developed a new process for recording waivers and issued a memorandum. Partnerships with Foreign Governments. In September 2013, GAO reported that CBP had not regularly assessed foreign ports for risks to cargo since 2005. GAO recommended that DHS periodically assess the security risks from ports that ship cargo to the United States and use the results to inform whether changes need to be made to Container Security Initiative (CSI) ports. DHS concurred with the recommendation and CBP has since developed a port risk matrix and priority map to be used to help assess whether changes need to be made to CSI ports. These tools are to be updated yearly and can be updated more frequently based on significant changes, emerging threats, and intelligence. These tools should assist CBP in ensuring it is allocating its resources to provide the greatest coverage of U.S.-bound high-risk cargo. In October 2009, GAO reported challenges to scanning 100 percent of U.S.- bound cargo at foreign ports. DHS officials acknowledged that most, if not all foreign ports, would not be able to meet the July 2012 target date for scanning all U.S.-bound cargo, and DHS would need to issue extensions to allow the continued flow of commerce and remain in compliance with statutory requirements. Although the Secretary of Homeland Security has issued three 2-year extensions for implementing the 100 percent scanning mandate, which have extended the deadline to July 2018, DHS has not yet identified a viable solution to meet the requirement. Partnerships with the Trade Industry. Through the Customs-Trade Partnership Against Terrorism (C-TPAT) program, CBP officials work with member companies to validate the security of their supply chains in exchange for benefits, such as reduced scrutiny of their shipments. In April 2008, GAO found, among other things, that CBP lacked a systematic process to ensure that members take appropriate actions in response to security validations. GAO recommended that CBP document key data elements needed to track compliance. CBP has since implemented a process to ensure that C-TPAT validation report recommendations are implemented. GAO is currently reviewing the C-TPAT program, to include an assessment of CBP's ability to meet its security validation responsibilities. In prior reports, GAO has made recommendations to DHS to strengthen various maritime cargo security programs. DHS generally concurred with the recommendations and has taken actions, or has actions under way, to address many of these recommendations.

The DI program provides monthly cash benefits to insured, severely disabled workers; the SSI program provides monthly cash payments to aged, blind, or disabled people whose income and resources fall below a certain threshold. Claimants under either program file an application for disability benefits with one of SSA’s more than 1,300 field offices. Applications, along with supporting medical evidence, are then forwarded to state disability determination service (DDS) offices, which make the initial medical determination of eligibility in accordance with SSA’s policies and procedures. Claimants DDS examiners find ineligible have the right to appeal the decision to OHA, where cases are heard by administrative law judges (ALJ). A steadily increasing number of appeals has caused workload pressures and processing delays for OHA. Between 1985 and 1995, appeals increased more than 140 percent, and the number of appealed cases awaiting an OHA decision grew from about 107,000 to almost 548,000. During this period, average processing time for cases appealed to OHA—measured from the date a claimant files a request for a hearing to when a decision is issued— increased 110 percent, from 167 days to 350 days. In addition, “aged” appealed cases (those taking 270 days or more for a decision) increased from 5 percent of pending appealed cases to 39 percent during the same period. SSA has a long-term strategy—its Plan for a New Disability Claim Process—designed to address systemic problems contributing to inefficiencies in its disability programs and significantly reduce the time claimants must wait to receive a decision on their claim. STDP is SSA’s ongoing effort to achieve some reduction in OHA’s backlog of appealed cases. SSA began STDP in November 1994 to address the backlog crisis from an agencywide perspective and establish specific goals and time frames for reducing backlogs. STDP includes 19 temporary initiatives to expedite the disability determination process and reduce OHA’s backlog from 488,000 appealed cases in October 1994 to 375,000 by December 1996. SSA set its backlog target to equal one and one-half times the number of appealed cases that, in OHA’s opinion, constitutes an appropriate workload for its ALJs and staff—about 250,000 appealed cases. According to OHA, the 375,000 target does not relate to any processing time or waiting time goal— it simply is a target that SSA believed was achievable at STDP’s inception. To reach its aggressive backlog reduction goal, STDP relies heavily on a temporary reallocation of agency resources and process changes to reduce the number of appealed cases requiring an ALJ hearing. Although STDP has 19 temporary initiatives, OHA expects that its major effect will come primarily from expanding two pre-STDP initiatives to expedite the processing of appealed cases. These two initiatives—regional screening unit and prehearing conferencing activities—were designed to target for review specific kinds of appealed cases that are likely to result in ALJs’ approving the claim for payment (referred to as “allowance”). These reviews can result in possible allowance without the more costly and time-consuming process of an ALJ hearing. Before STDP’s implementation, SSA had established screening units in each region to help alleviate OHA’s backlog. Screening unit examiners, who were not OHA staff, reviewed certain appealed cases to determine if the evidence in the case file was sufficient to permit an allowance, eliminating the need for a hearing. SSA selected most cases for review by screening unit staff by using computer-generated case profiles to identify potentially incorrect claim denials by DDS staff. SSA officials believe that such profiling of appealed cases minimizes the risk of incorrect allowances. Under STDP, SSA expanded screening unit activities by assigning OHA attorneys to help examiners in all of SSA’s regional screening units to identify more appealed cases that could be allowed earlier in the process. According to SSA, the opportunity for screening unit examiners to discuss issues with OHA attorneys gave the examiners more insight into the adjudication process and enabled the examiners and attorneys, where appropriate, to recommend allowance in more cases. SSA’s pre-STDP efforts to reduce the backlog of appealed cases also included implementing a prehearing conferencing process. The purpose of prehearing conferencing was to shorten processing time for appealed cases by assigning experienced OHA attorneys to review and identify appealed cases that potentially could be allowed without a formal ALJ hearing. While screening unit activities focused on reviewing evidence already in the case file, prehearing conferencing enabled attorneys to review evidence in the case file, confer with claimant representatives, conduct limited case development, and draft decisions to be reviewed and approved by ALJs. Under STDP’s expanded prehearing conferencing initiative, OHA’s senior attorneys have been given quasi-judicial powers or the authority to issue allowance decisions without an ALJ’s involvement or approval. Under the initiative, OHA attorneys are to extensively develop the case record, which includes obtaining medical and vocational evidence, conducting conferences with claimant representatives as well as medical and vocational experts, and issuing allowance decisions. If they cannot allow the claim on the basis of their review of the evidence, the case is scheduled for an ALJ hearing. As in the screening unit initiative, SSA relied on computer-generated case profiles to select cases to be processed under this effort. Cases were selected on the basis of their likelihood to be allowed on the record by an ALJ. STDP is scheduled to be phased out in December 1996. Although OHA has proposed that SSA extend expanded screening unit activities through December 1997, as of September 1996 SSA had made no final decision on this. Expanded prehearing conferencing, however, will remain active until June 30, 1997, when regulatory authority for senior attorneys to allow appealed cases expires. In fiscal year 1997, SSA expects to implement certain features from its ongoing efforts to redesign the disability claims process. One of the features being tested is a new decision-making position to help expedite appealed claims through the process. Like activities under STDP’s expanded screening unit and prehearing conferencing initiatives, this position will enable someone other than an ALJ to review and allow some appealed cases, eliminating the need for an ALJ hearing. SSA acknowledges that it will not reach STDP’s goal of reducing the backlog of appealed cases to 375,000 by December 1996. In fact, OHA’s backlog of about 515,000 appealed cases as of August 1996—about 22 months into STDP—was 3 percent higher than the backlog of about 500,000 that existed at the plan’s inception. Although SSA will not reach STDP’s backlog reduction goal, the agency believes that the plan has helped to reduce the growth in the backlog of appealed cases awaiting a decision. Since peaking at about 552,000 in December 1995, OHA’s backlog decreased steadily by an average of about 4,600 appealed cases per month through August 1996 or by about 37,000 total appealed cases. As shown in figure 1, OHA’s backlog decreased during each of the last two fiscal quarters of 1996. As of August 31, 1996, the backlog was 515,009 appealed cases. OHA’s current projections indicate that its backlog of appealed cases will be approximately 498,000 at the end of calendar year 1996 or about 123,000 above STDP’s target. OHA is relying on increased productivity from its ALJs and attorneys to increase its ability to dispose of cases and facilitate reaching this revised target. According to OHA, its inability to reach STDP’s backlog reduction goal is due to start-up delays, overly optimistic projections on the number of appealed cases that could be processed, and an unexpected increase in the number of appealed cases. Figure 2 illustrates the disparity between the number of appealed cases OHA expected to allow under STDP through December 1996 and the actual number that have been allowed through August 1996—22 months since the plan was initiated. Start-up delays associated with prehearing conferencing—the initiative expected to have the greatest impact on reducing OHA’s backlog of appealed cases—have hindered SSA’s ability to reach STDP’s goals. To implement this initiative, SSA had to seek a regulatory change to give about 600 OHA senior and supervisory staff attorneys the authority to decide certain appealed cases that were formerly limited to ALJ jurisdiction. However, the process of obtaining regulatory change and defining the specific duties and responsibilities these attorneys would have under STDP was lengthy, and implementation did not begin until July 1995—or about 6 months after the projected start-up date. Overly optimistic allowance projections for STDP’s expanded prehearing conferencing and screening unit initiatives also contributed to OHA’s inability to reach the plan’s backlog reduction goal. SSA initially projected that expanded prehearing conferencing would result in 224,000 allowances by senior attorneys through the 2-year period ending December 1996. However, as of August 31, 1996—or about 22 months into STDP—these attorneys had allowed only 55,363 appealed cases or about 25 percent of the projected total. The aggressive projections for this initiative were based on the results of the prehearing conferencing pilot, which OHA conducted before STDP’s implementation, and the assumption that the use of profiling to select cases would result in a higher rate of cases that could be allowed without a hearing. On the basis of the prehearing conferencing pilot, which was conducted at 19 hearing offices that agreed to participate, OHA estimated that senior attorneys would be able to allow approximately 75 percent of the appealed cases selected for their review. However, data show that between August 1995 and August 1996 senior attorneys allowed only about 24 percent of the appealed cases reviewed under STDP. According to SSA, the lower allowance rate is primarily due to senior attorneys’ not conducting prehearing conferences with claimants as frequently as anticipated as well as not sufficiently developing evidence necessary to complete a claimant’s case record. To increase the number of allowances under this initiative, OHA has directed its hearing offices to ensure that all senior attorneys receive training to better familiarize themselves with OHA’s case development process. In addition, through directives and a series of conference calls with all its hearing offices, OHA has provided its senior attorneys with specific guidance that includes the kind of evidence that would adequately support an allowance decision. Like STDP’s prehearing conferencing initiative, expanded regional screening has not reached STDP’s allowance goals. Before STDP, screening units were expected to allow about 20,000 appealed cases annually. With STDP’s introduction of OHA attorneys to the process, SSA expected to allow 38,000 appealed cases annually or 76,000 over the 2 years the initiative was to be in place. In the 22 months since STDP was initiated, however, screening units had allowed a total of 26,022 appealed cases or about 34 percent of the projected total as of August 31, 1996. SSA expected that under STDP, regional screening units would allow 76,000 appealed cases or about 15 percent of those selected for review. To reach the initiative’s target of 76,000 allowances, screening units would have had to review a total of about 507,000 cases. Since STDP’s inception in November 1994, however, screening units had reviewed only about 258,000 cases as of August 31, 1996. According to SSA, the shortfall in the number of appealed cases processed by screening units is mainly due to SSA’s reassignment of some screening unit staff to other duties. Finally, an unexpected increase in the number of appeals also hindered OHA’s efforts to reduce its backlog to STDP’s goal. During fiscal year 1995, OHA received approximately 37,500 more appealed cases than it had initially projected for the year. According to OHA’s staff management officer, this unanticipated workload was due primarily to an increased number of cases processed by DDS staff. STDP has enhanced OHA’s ability to dispose of appealed cases, helped decrease the agency’s decision-writing backlog, and reduced processing time for some appealed cases. OHA estimates that as of August 31, 1996, STDP had resulted in a net increase of about 66,500 dispositions. This estimate is based on time savings associated with appealed cases allowed under STDP’s expanded screening unit and prehearing conferencing initiatives. To determine the net increase in dispositions attributable to STDP, OHA estimated the amount of ALJ time that could be saved through activities implemented under the plan’s two key initiatives and converted these time savings into the number of additional cases that could be disposed of by ALJs in that amount of time. OHA’s estimate that the number of dispositions through August 1996 increased by about 66,500 as a result of STDP is consistent with our estimate. On the basis of our analysis of ALJ productivity before STDP, had SSA not implemented the plan, OHA would have disposed of about 68,000 fewer cases between the beginning of fiscal year 1995 and August 1996. STDP has also helped to reduce the number of appealed cases awaiting a written decision. To increase OHA’s decision-writing capacity, staff from various SSA offices were temporarily detailed to OHA under STDP. Efforts made under STDP helped reduce the decision-writing backlog from 40,567 decisions—its level at STDP’s inception—to 20,293 as of August 31, 1996, or by about 50 percent. Finally, STDP has significantly reduced processing times for appealed cases allowed under its expanded screening unit and prehearing conferencing initiatives. On average, processing times for screening unit examiners’ decisions have averaged 39 days; processing times for senior attorneys’ prehearing conferencing decisions have averaged 121 days. These processing times are substantially shorter than the average monthly processing time of 264 days for similar cases decided by ALJs from May 1995 through May 1996. Some SSA and OHA officials had expressed concern to us that STDP’s aggressive processing goals could result in inappropriate benefit awards for some disability claimants and that STDP’s initiatives could cause OHA’s allowance rate to increase. However, the percent of appealed cases allowed by OHA since STDP’s inception has notably decreased. The allowance rate has decreased from about 75 percent in fiscal year 1994— the fiscal year preceding STDP’s implementation—to about 69 percent through the third quarter of fiscal year 1996. This allowance rate reflects cases decided by ALJs as well as those decided by screening unit staff and senior attorneys under STDP. As figure 3 shows, except for the third quarter of fiscal year 1996, the allowance rate has decreased during every quarter since the beginning of 1995. SSA has not completed any analyses of factors contributing to this decrease, however. STDP is SSA’s effort to achieve some reduction in what has been OHA’s growing backlog of appealed cases. Recent processing trends show that STDP has helped the agency reduce the backlog, which has decreased steadily in the past 8 months. In addition, concerns that STDP could result in inappropriate allowances and that OHA’s allowance rate could increase have not been substantiated. SSA is evaluating the accuracy of the decisions made under STDP to help determine the advisability of continuing with the plan. Because STDP has shown that it can help reduce the backlog of appealed cases, we recommend that—if SSA determines that accurate decisions are being made—the Commissioner of the Social Security Administration extend STDP until the agency institutes a permanent process that ensures the timely and expeditious disposition of appeals. In commenting on a draft of this report, SSA agreed with our conclusions and recommendation on the conditions for extending STDP. The agency stated that it recently found the accuracy of screening unit allowances to be acceptable and has decided to extend the initiative beyond the original December 1996 expiration date. The agency also stated that it is reviewing the accuracy of prehearing conferencing allowances and will soon decide whether to extend that initiative. We also received technical comments from SSA, which we incorporated where appropriate. SSA’s comments are reprinted in appendix I. We are providing copies of this report to the Director of the Office of Management and Budget and the Commissioner of the Social Security Administration. We will also make copies available to others upon request. Major contributors to this report are listed in appendix II. If you have any questions concerning this report or need additional information, please call me on (202) 512-7215. Michael T. Blair, Jr., Assistant Director, (404) 679-1944 Carlos J. Evora, Evaluator-in-Charge, (404) 679-1845 The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed the Social Security Administration's (SSA) Short-Term Disability Plan (STDP), focusing on: (1) progress made by SSA's Office of Hearings and Appeals (OHA) in meeting STDP backlog reduction and case processing goals; (2) the current OHA allowance rate for appealed cases compared with pre-STDP levels; and (3) the accuracy of OHA decisions made under STDP. GAO found that: (1) OHA has made progress in reducing its inventory of appealed cases, but SSA will not reach its goal of reducing this backlog to 375,000 by December 1996; (2) activities under the plan's key initiatives allowed OHA to dispose of about 66,500 more cases than it would have had STDP not been implemented, but despite OHA's increased productivity, as of August 1996 its backlog of appealed cases was about 3 percent higher than in November 1994; (3) since STDP was initiated, the allowance rate has decreased from about 75 percent in fiscal year (FY) 1994 to about 69 percent through the third quarter of (FY) 1996; and (4) SSA has not completed any analysis of the accuracy of STDP decisions or clearly established to what extent, if any, STDP has affected OHA's allowance rate.

Full and open competition is the preferred method for federal agencies to award contracts. This preference was established through the Competition in Contracting Act (CICA) of 1984, which required agencies to obtain full and open competition through the use of competitive procedures in their procurement activities unless otherwise authorized by law. Contracts awarded using full and open competition means that all responsible sources—or prospective contractors that meet certain criteria—are permitted to submit proposals. Agencies are generally required to perform acquisition planning and conduct market research to promote and provide for, among other things, full and open competition. However, Congress, by enacting CICA, also recognized that there are situations that require or allow for contracts to be awarded noncompetitively—that is, contracts awarded without full and open competition. Some of the permitted exceptions to full and open competition follow. Supplies and services are only available from one responsible source, such as unique services from one supplier with unique capabilities, or limited rights to data that make certain services available from one source. The government is under unusual and compelling urgency to procure a good or service, and delaying the award of a contract would result in serious injury, financial or other, to the government. A statute expressly authorizes or requires that the acquisition be made from a specific source or through another agency, such as sole source awards under the SBA’s 8(a) program—one of the federal government’s primary means for developing small businesses owned by socially and economically disadvantaged individuals. The terms of an international agreement between the United States and a foreign government, or written directions of a foreign government reimbursing a federal agency for the cost of an acquisition, preclude competition. The disclosure of the agency’s needs would compromise national security. This exception, however, is not to be used merely because the acquisition is classified or because access to classified matter is necessary. Noncompetitive contracts are not permitted in situations in which the requiring agency has failed to adequately plan for the procurement or in which there are concerns related to availability of funding for the agency, such as funds expiring at the end of the year. Generally, noncompetitive contracts must be supported by written justifications and approvals that contain sufficient facts and rationale to justify the use of the specific exception to full and open competition that is being applied to the procurement. These justifications must include, at a minimum, 12 elements specified by the FAR, for example, a description of the supplies or services required to meet the agency’s needs and their estimated value; identification of the specific statutory authority permitting other than full and open competition; a determination by the contracting officer that the anticipated cost to the government will be fair and reasonable; a description of market research conducted, if any; and a statement of the actions, if any, the agency may take to remove or overcome any barriers to competition before any subsequent acquisitions for the supplies or services required. The approval level for these types of noncompetitive contracts varies according to the dollar value of the procurement. Some contracts do not require written justifications, including those awarded on a sole source basis through the 8(a) program under the “authorized or required by statute” exception. Although full and open competition is the preferred method to award a contract, agencies can competitively award contracts after limiting the pool of available contractors—a process called “full and open competition after exclusion of sources.” An example of this is when agencies set aside procurements for small businesses. In fact, agencies are required to set aside procurements for competition among qualified small businesses if there is a reasonable expectation that two or more responsible small businesses will compete for the work. Federal agencies can establish indefinite delivery / indefinite quantity (IDIQ) contracts, or issue orders under them, using a number of different authorities. The following is a discussion of some of these authorities pertinent to the contracts included in our review and any provisions for exceptions to competition. The Federal Acquisition Streamlining Act (FASA) of 1994 provided competition requirements for task order and delivery order contracts, referred to as IDIQ contracts. IDIQ contracts can be single award or multiple award contracts, but FASA establishes a preference for multiple award contracts. Multiple award IDIQ contracts are awarded to multiple contractors through one solicitation. The number of contract holders depends on the number of contractors receiving the award, which could be from two contractors to thousands. Agencies are required to compete orders on multiple award contracts among all contract holders; however, agencies can award noncompetitive orders—through a process called an exception to a fair opportunity to compete—for reasons similar to those used for awarding contracts without full and open competition, such as only one contractor being capable of providing the supplies or services needed, or an urgent requirement. The FAR requires contracting officers to document, in the contract file, the rationale for awarding the order noncompetitively, but does not specify what should be included in these justifications. In addition, approval of the justifications for noncompetitively awarded orders is not required. One example of a large, multiple award IDIQ is the Navy’s Seaport Enhanced (Seaport-e) program with over 1,200 contract holders that can provide 22 different services, such as engineering, program, and logistics support. Orders may be issued under Seaport-e by Navy Systems Commands, and other Navy Commands and offices. Requirements must be competed among all contractors within a certain geographical area. Noncompetitive orders are only allowed if no alternative contract vehicle exists and written approval from the program manager of Seaport-e is obtained. GSA, under its schedules program, awards IDIQ contracts to multiple vendors for commercially available goods and services, and federal agencies place orders under the contracts. To compete orders over $3,000, agencies need only survey three schedule contractors that offer services that will meet their needs. For orders issued noncompetitively under the schedules program, however, the ordering agency must justify in writing— with specific content required by the FAR—the need to restrict competition and also obtain approval at the same dollar values and by the same officials as for contracts awarded without full and open competition. The Army has established BPAs under GSA schedule contracts with about 1,200 contractors–called the Express Program—to provide advisory and assistance services in four domains: business and analytical, programmatic, logistical, and technical. Requirements are generally competed within each domain, but orders can be placed noncompetitively. Contracts that are awarded using competitive procedures but where only one offer is received have recently gained attention as an area of concern. OFPP recently noted that competitions that yield only one offer in response to a solicitation deprive agencies of the ability to consider alternative solutions in a reasoned and structured manner. The Office of Federal Procurement Policy Act, as amended by CICA, required that agencies begin separating data collected on contracts that were awarded using competitive procedures where only one offer was received. The act stipulated that these contracts be recorded as “noncompetitive procurements using competitive procedures.” Currently, FPDS-NG distinguishes these contracts by recording how many offers were received on any procurement. Congress and the executive branch have recently taken actions that require or encourage more competition in federal contracting and that bring more scrutiny to noncompetitive contracts. For example, since 2008 Congress has enacted legislation that: requires justifications for certain noncompetitive awards to be publicly posted; enhances competition for task orders on multiple award contracts; requires acquisition strategies for major defense acquisition programs to include measures to ensure competition throughout the life cycle of the program; and requires justifications, approvals, and notices for sole source contract awards over $20 million awarded under the authority of SBA’s 8(a) program. The executive branch also has brought attention to the importance of competition. In May 2007, OFPP called for agencies to reinvigorate the role of the competition advocate, a position required by law at each executive agency to promote competition. Each competition advocate must, among other things, submit an annual report on competition to the agency’s senior procurement executive and chief acquisition officer and recommend goals and plans for increasing competition. In March 2009, the President called on federal agencies to examine their use of noncompetitive contracting as one of several important steps to improving the results achieved from government contractors. In July 2009, OMB instructed agencies to reduce dollars obligated to high-risk contracts— including noncompetitively awarded contracts and contracts competed with only one offer received—by 10 percent in fiscal year 2010. In October 2009, OFPP followed up with guidelines for agencies to evaluate, in part, the effectiveness of their agencies’ competition practices. Total obligations reported in FPDS-NG increased during fiscal years 2005 through 2009, from $430.6 billion to $543.6 billion. For the same 5-year period, the percentage of obligations reported for noncompetitive contracts decreased, from 35.6 percent to 31.2 percent of total obligations, while those reported under contracts that were competed with one offer received (noncompetitive procurements using competitive procedures) were steady, at about 13 percent of total obligations. To determine whether there was any variation in dollars obligated to noncompetitive contracts during the four quarters of the fiscal year, we analyzed dollars obligated in each quarter for fiscal years 2005 through 2009. We found that the fourth quarter consistently had the lowest percentage of obligations to noncompetitive contracts in each fiscal year, while the first quarter generally had the highest percentage of obligations to noncompetitive contracts, as shown in figure 2. In fiscal year 2009, among all federal agencies and DOD services that obligated over $1 billion, the Navy and Air Force had some of the highest percentages of total contract obligations that were not competed, at about 45 percent. The agencies with some of the lowest percentages of total contract obligations to noncompetitive contracts were the Department of Energy and the Office of Personnel and Management, with 7 percent and 5 percent respectively. Although our sample is not representative of all federal contract obligations, we found coding errors in FPDS-NG. Specifically, 19 of the 107 contracts and orders we reviewed, or about 18 percent, were coded incorrectly. See table 1. The 9 contracts and orders miscoded as noncompetitive had actually been competed. For example, one ICE contract had been coded in FPDS-NG as not competed, but the agency had in fact competed it and received proposals from 5 vendors. Another 5 of the 9 miscoded noncompeted contracts were actually orders under single-award IDIQ contracts that were competed, but the orders were coded as not competed. For example, three orders at Interior coded as noncompeted in FPDS-NG turned out to have actually been competed, since their base contracts were competed. When a single-award IDIQ contract is competed, the orders under that contract are considered competed. This type of error appears to have stemmed from a lack of understanding on the part of the person entering the data, as some agency officials we spoke with admitted that there was confusion among contracting officials about how to code these orders. In April 2008, the Department of the Interior issued guidance clarifying that orders awarded under single-award indefinite delivery contracts that were awarded under full and open competition should be coded as competed. However, to address the issue more widely, on October 31, 2009, systemwide changes were made to FPDS-NG. Now, coding of the extent of competition under the base contract is automatically pulled forward to subsequent orders. This action should mitigate such errors in the data going forward. Of the 10 contracts and orders that were incorrectly coded as competed with one offer received, 4 had not been competed at all. Two of the 4 were sole source contracts awarded on the basis of only one responsible contractor that could perform the work; one was a sole source contract award through the 8(a) program; and one was a sole source contract award on the basis of an international agreement with a foreign government. For the other 6 contracts and orders miscoded as competed with one offer received, documentation in the contract file indicated that they were actually competed with more than one offer received. For example, two Army contracts were labeled as competed with one offer received, but one had three offers and the other had four offers. It is not clear why these contracts and orders were miscoded. Accounting for the miscoded contracts and orders, our analysis going forward focused on 74 noncompetitive contracts and 19 contracts that were competed with one offer received. Agencies used a variety of exceptions to full and open competition, and ordering processes, to award the 74 noncompetitive contracts in our sample. Table 2 shows the spectrum of exceptions and processes that agencies used to award these contracts or orders. As indicated in the table, for 42 of 74 contracts—or 57 percent of the noncompetitive contracts in our sample—agencies determined, under FAR Part 6.3, that only one responsible contractor could meet the agency’s requirements. For example, the National Weather Service—through an interagency contract awarded by Interior—turned to the original provider of weather radios to obtain compatible spare parts. In another example at ICE, only one contractor could provide specified communications equipment, supplies and services being used in the field at the time. According to an ICE contracting official, this contractor essentially owns the market, and until other vendors or products are available, ICE is bound by the limited availability of items. The second most frequently used exception to competition—for 20 of the 74 noncompetitive contracts in our sample, or 27 percent—was the authority to award sole source contracts to qualified firms in SBA’s 8(a) business development program. Through the 8(a) program, agencies are encouraged to award sole source contracts under $3.5 million when procuring services, or $5.5 million for manufacturing, to participating 8(a) firms. In fact, the FAR encourages agencies not to compete under these thresholds, requiring agencies to obtain the approval from the SBA Associate Administrator for 8(a) Business Development for any competed procurements under the threshold, and this approval is to be given on a limited basis. One example was a sole source contract for $1.7 million to an 8(a) firm for lead abatement services for one of the Secret Service’s training facilities. Our sample also included large dollar value sole source contracts to 8(a) firms owned by Alaska Native Corporations (ANC) or tribal entities, such as American Indian tribes, which have special advantages over other 8(a) firms and can receive sole source contracts for any dollar amount. Some examples of these 8(a) contracts in our sample follow. The Air Force awarded a $75 million sole source award to an 8(a) firm owned by an American Indian tribe for analysis, integration and technical support services related to corrosion prevention and control. The Navy awarded a sole source contract to an 8(a) ANC firm for operation and management support and analysis and technical support for $131 million. In general, awarding noncompetitive contracts through the 8(a) program is an easy and quick way for agencies to award a contract, rather than using full and open competition. First, when awarding a sole source contract through the 8(a) program, a justification for awarding a sole source contract is generally not required. Second, the agency need only identify a qualified 8(a) firm and obtain approval from SBA to award it a contract. For example, a Secret Service contract estimated at $3 million for information technology services included a description in the contract file of the market research that had been conducted, which simply stated that the program office provided the source. In another example from Interior, a program staff person at the National Institutes of Health suggested a contractor for building repair services to the Interior contracting officer. The program staff informed us that although other contractors were available, he was most comfortable with the vendor he suggested, and therefore requested—and received—a noncompetitive award through the 8(a) program for approximately $3.5 million. SBA officials told us that agencies’ procurement activities are encouraged to direct all work to small businesses as long as they do not run afoul of the Small Business Act or federal acquisition regulations. The SBA takes the general position that a procuring agency does not need to document in a contract file any other prospective sources if the agency selects an 8(a) participant to perform the requirement, offers it to SBA, and SBA accepts the requirement into the 8(a) program. SBA officials note that it is the procuring agency’s responsibility to conduct market research to determine whether the requirements of the Small Business Act can be met, and then to determine the appropriate contracting vehicle to use. However, SBA considers market research requirements to be satisfied when a participant in the 8(a) program self-markets its abilities to a procuring agency and is subsequently offered a sole source 8(a) requirement. When we discussed this issue with procurement policy officials at DHS, they said that, while these activities may meet the regulatory requirements, in practice they like to see additional market research so that the offer to the 8(a) firm has a more solid basis. Agencies in our review used a number of other exceptions under FAR Part 6.3 (“Other than Full and Open Competition”) to award noncompetitive contracts, which includes orders issued under noncompetitively awarded IDIQ contracts. For example, Three orders at the Air Force were issued under separate sole source contracts using the justification that disclosure of information on the program would compromise national security. One of these orders was to provide spare parts and resolve system failures to sustain the fielded equipment and software for remote airborne sensors. These orders were justified as a sole source procurement using a class justification—meaning one justification is used for consolidated requirements across DOD activities and multiple programs, such as the U-2 program. Two contracts in our sample were awarded directly to one company on behalf of another country through an international agreement, referred to as foreign military sales. One example was an Army contract to install and configure software to modernize the logistics system for the Defense Forces of Saudi Arabia. One noncompetitive order we reviewed was issued under a contract to a federally funded research and development center (FFRDC) to look at work processes and work flow requirements for clinical research and their interoperability with those of disease-specific research networks, hospitals, institutions, industry, and government. The exception to competition that was used is that it was necessary to award the order to a particular source to establish or maintain an essential engineering, research, or development capability to be provided by an FFRDC. Although the contracting officer at Interior identified four FFRDCs that could do the work, the National Institutes of Health staff determined that this particular firm best met their needs because this work was a continuation of research that it was performing for them. Agencies also used different procedures under the FAR to issue noncompetitive orders under competitively awarded IDIQ contracts. Four of the orders at two different agencies were issued under GSA schedules contracts, using procedures under FAR 8.405-6. Agencies justified not competing the orders because the work was a follow-on to another requirement that the company performed, or because there was only one source that could perform the specific work. In one case, a requiring office at the Department of the Interior that provides financial services to other federal agencies (or “federal customers”) needed a contractor to help with the integration and execution of the financial services provided. The Interior contracting officer suggested a minicompetition among GSA schedule contractors; however, the limited sources justification noted that this procurement was a logical follow-on because this particular company had partnered with the requiring office to perform these same integration services with six different federal customers and that, because this company was already familiar with the customer, it was in the best position to provide these services. In two other cases from our sample, orders were awarded through a process called “exceptions to the fair opportunity process” under FAR 16.505(b)(2). These exceptions allow noncompetitive orders exceeding $3,000 issued under multiple award contracts using one of four reasons: (1) only one contractor is capable, (2) urgency, (3) the work is a logical follow-on to another task, or (4) there is a need to place an order with a particular contractor to satisfy a minimum guarantee. For example, an order was awarded for engineering support to redesign the B-1 aircraft main landing gear wheel and brake assembly because the contractor had previously worked on this airplane and had the expertise. The program official explained that when a new type of plane comes in for repair, requirements are typically competed between the two contractors on the multiple award contract, but once a contractor has built up expertise on that airplane’s system, it is logical to have the same contractor perform additional work on that system. A variety of factors affect competition, including reliance on contractor expertise and decisions made by officials in program and contracting offices. For services supporting DOD weapons programs, the government’s lack of access to proprietary technical data and a heavy reliance on specific contractors for expertise limit, or even preclude the possibility of, competition. Contracting officials pointed out that a program office may be comfortable with the incumbent contractor and presses the contracting office to remain with that contractor, thus inhibiting competition. Even for some procurements using competitive procedures, a strong incumbent coupled with overly restrictively written requirements can lead to only one offer—from the incumbent—being received. In other cases, groups of vendors have formed teams to compete for government requirements. Contracting officials and contractors told us that whereas previously several vendors might have submitted offers for more specific requirements, now only one offer—from the prime contractor on the team—is being received. However, we did find cases in which contracting and program officials were actively seeking opportunities to compete requirements. For 27 of the 47 noncompetitive DOD contracts we reviewed, the government was unable to compete requirements due to a lack of access to proprietary technical data. This situation, combined with a heavy reliance on certain contractors’ expertise built over years of experience, inhibits competition. Most of the contracting and program officials at DOD that we spoke with pointed to the lack of access to technical data as one of the main barriers to competition. Some contracting officers described this condition as essentially being “stuck” with a certain contractor. For example, a $46 million contract at the Navy for engineering services in the DOD’s Prowler/Growler aircraft programs could not be competitively awarded because the government had not procured the technical data package and only the original contractor, who was one of the developers of the system, has over 20 years experience and expertise to perform the work. Several officials pointed out that the situation the government is currently experiencing is a result of decisions made years ago, when first acquiring a weapon system, to not purchase critical technical data packages for reasons that include budgetary constraints or a push toward streamlined contracting processes by purchasing commercial items. For a couple of the contracts in our sample, the government had purchased some of the technical data, but, for budgetary reasons, has not kept those data packages current over time. Hence, only the original equipment manufacturer has the technical data needed for follow-on maintenance and engineering support contracts. Some contracting and program officials have inquired about the cost of obtaining the technical data, only to discover that the package is not for sale or purchase of it would be cost-prohibitive, especially the systems and equipment that have been contracted out for decades. In one instance, the Air Force requested an estimate of the cost to the government to purchase the technical data package for an aircraft program, and the contractor— the original equipment manufacturer that had been working on the system for over 30 years—replied that while it was not for sale, if they were to sell it, the estimated cost was $1 billon. On a $4.8 billion contract for sustainment and support for another Air Force program, the contractor estimated the cost to purchase the data rights to be more than $1.3 billion. However, the market research report noted that the contractor refused to sell the data, and because commercial contracting procedures under FAR Part 12 were used in this procurement, the contractor was able to retain strict control over data rights and the government did not have insight into the work performed by the major subcontractors. In yet another case, a contractor for an Army missile program informed the program office that they would charge approximately $30,000 just to put together a cost estimate for the technical data package, which the contractor later stated would be $31 million for selected technical data elements of the missile program, but excluding rights to critical contractor-specific software. The contractor was the original equipment manufacturer and sole producer of the missiles since the early 1960s. DOD procurement policy officials told us they view this issue as a long-standing problem and that any significant turnaround will need to occur with new programs. They also said they see refusal to share or sell technical data as a larger problem under commercial acquisitions, where the government lacks leverage. Recently, Congress has taken steps to address the lack of access to technical data. For example, the John Warner National Defense Authorization Act of Fiscal Year 2007 required DOD program managers for major weapons systems to assess the long-term technical data needs and establish corresponding acquisition strategies that provide for the technical data rights needed to sustain such systems over their life cycle. Further, Congress enacted legislation in May of 2009 that requires DOD to include in the acquisition strategy for each major defense acquisition program measures to ensure competition—or the option of competition— at both the prime contract level and subcontract level throughout the life- cycle of the program. This includes considering the acquisition of complete technical data packages, among other things. For almost a decade, we have reported on the limitations to competition when DOD does not purchase technical data rights for sustainment of weapon systems and the increased costs as a result. In 2001 and 2002, we reported that DOD had often failed to put adequate emphasis on obtaining needed technical data during the acquisition process, and noted officials’ concerns on the potential negative impact on competition and potential increase in costs. In 2004, we found that not obtaining technical data limited DOD’s flexibility to perform work in house or support alternate source development if necessary. In another report, in 2006, we noted that as a result of the limitations of not having technical data rights, the military services had to alter their plans for developing new sources of supply to increase production or to obtain competitive offers for the acquisition of spare parts and components to reduce sustainment costs. In that report, we also found that DOD’s acquisition policies did not specifically address long term needs for technical data rights and recommended that DOD require program managers to assess long-term technical data needs and establish corresponding acquisition strategies that provide for technical data rights needed to sustain weapons systems over their life cycles. Even when technical data are not an issue, the government may have little choice other than to rely on the contractors that were the original equipment manufacturers, and who, in some cases, designed and developed the weapon system. A few contracting and program officials we spoke with noted that for some DOD programs, the government is so reliant on the contractor that it is difficult for the government to even make decisions or set requirements anymore. Our prior work has noted the government’s increasing reliance on contractors and pointed to the challenges of this increasing reliance, such as identifying and distinguishing roles and responsibilities and ensuring appropriate oversight. Most noncompetitive DOD contracts in our sample indicated that the contractor was the only source of the expertise for the system, having developed that expertise and the infrastructure over time. For example: An engineering contract for the Army’s Hellfire missile program has current obligations at almost $72 million. According to the justification in the contract file, the technical data package has been developed, but since 1994, the contractor has been acquiring unique expertise that is not contained in any documentation. The contractor for the Army’s Patriot Missile program has been designing the missile since 1972. The contract for engineering and other support of the program was worth an estimated $122 million. The contractor that designed the program was the only contractor capable of performing the support work because the contractor also developed and manufactured the system. Over time, the contractor developed the technical expertise, experience, and the facilities needed for the contract. The Army awarded a $1.7 billion contract for engineering support and maintenance for the Chinook and Apache Helicopter Programs. The contractor for these two helicopters—since 1961 and 1984 respectively—is the only contractor with the needed skills, technical and engineering expertise, and the technical data to provide the full range of services needed. This contract was one of several billion-dollar sole-source Army contracts we reviewed that had been awarded to large prime contractors for depot maintenance requirements that had been previously performed by many small businesses. A business case analysis was performed, showing that these contracts were burdensome to manage and left the government without one entity to hold accountable. The contracts were bundled into one requirement (with the appropriate justifications and approvals) and the prime contractors’ subcontracting plans emphasized the need to compete among small businesses at that level. The Air Force awarded a contract for engineering support and software maintenance for satellite communication systems with an estimated value of $404.7 million. The contractor, in place since 1982, has 20 years of knowledge and experience with extensive hardware, software, and test facilities needed to support the system. Further, the cost, including time and money, of changing contractors can be relatively high. For instance, the sole-source justification for an almost $1 billion contract awarded in June 2008 for the overhaul and recapitalization of the Army’s Blackhawk helicopter included a $50 million estimate as the minimum investment needed to bring on another contractor and a lead time of 24 to 36 months. The justification further stated that the current contractor’s knowledge could not be easily duplicated, even with significant investment and that it was unlikely that the government would be able to recover the investment cost through competition. In another example, the sole source justification of an Air Force contract, estimated at $50 million, for general engineering support for Military Satellite Communication programs, included an estimate of $5 million for developing another contractor’s knowledge and skill, including the time to gain familiarity with the software tools and technical requirements. However, the justification also described that any delays would result in a cost increase of hundreds of millions of dollars for the program and any unplanned schedule delays would adversely affect the warfighter. Program officials play a significant role in the contracting process— developing requirements, performing market research, and interfacing with contractors. In their 2009 competition reports to OMB, several agencies in our review recognized the pressure that program offices place on the contracting process to award new contracts to a specific vendor without competition. Many contracting officials we spoke with recognized that program staff sometimes prefer a specific vendor, in some cases because a relationship had developed between the program office and the contractor, who understands the program requirements. We also heard this echoed in discussions with program staff. Program officials from two program offices at the National Institutes of Health, for example, described their comfort with certain contractors because of their level of understanding of requirements and because they could be relied on to complete the work. A Navy program official stated that, when one contractor has been performing a requirement for many years, it is easier to go back to the contractor personnel who understand the requirement rather than taking the time to find a new vendor. One contracting official described how, in his former role as a program manager, he did not want to change contractors for products and services once he found ones he liked. Program offices can also influence levels of competition through their roles in the acquisition planning process, in particular by having sufficient knowledge of the contract award process and providing contracting officials with enough time to compete requirements. However in their competition reports, some agencies in our review pointed to a lack of acquisition planning, and the role that the program office plays in it, as a barrier to competition. Further, several contracting officials from different agencies expressed concern about the fact that they receive short notices from program offices for acquisitions. Others noted that program offices sometimes do not allow them enough time to execute a sufficiently robust acquisition planning process that could increase opportunities for competition. They told us that program offices are insufficiently aware of the amount of time needed to properly define requirements or conduct adequate market research. According to an official at ICE, in one instance, he only had a couple of days to complete certain procurements, which he managed to do, but he believed that the customer would have received a better product if he had had enough time to obtain more high-quality proposals from the marketplace. Several contracting officials also pointed to the experience levels and staffing shortfalls of both the contracting staff and program staff as affecting the quality of the procurement processes, and, in turn, the extent of competition. For example, DHS contracting officials stated that high workloads for limited numbers of staff and inexperienced staff can hinder the acquisition planning, timing of the procurements, and market research. Some agency officials recognize that training on the acquisition process for program staff may help address some of these issues, but we found that training on competition issues is often directed to contracting officers and not necessarily the program staff. For example, DOD has developed formal training on enhancing competition awareness, but it is required only for contracting staff and just recommended for others in the acquisition community. The Navy, however, has made this training mandatory for Navy personnel engaged in the acquisition process, including program managers, program executive officers and logistics personnel. ICE contracting officials said that they regularly reach out to the program offices—through meetings and supervisor trainings and by making guidance available—to provide information on the acquisition process with the goal of increasing competition. They noted, however, that ICE program offices still struggle to understand the acquisition process. Other contracting officials, for instance at the location we visited at the Department of the Interior, stated that they train program staff about the benefits of competition during regular, informal interactions or do so only on issues pertaining to a specific procurement. From a practical standpoint, for contracts awarded using competitive solicitation procedures where only one offer is received, the government does not have the benefit of evaluating more than one competing proposal. As noted earlier, OMB’s July 2009 instruction to agencies to reduce dollars obligated to high risk contracts included contracts that had been competed but where only one offer was received. The government’s requirements can influence the number of offers received under competitive solicitations if requirements are written too restrictively. Some contracting officials noted the challenge of questioning program office requirements that are written so restrictively that they are geared towards the incumbent. These contracting officials informed us that their technical backgrounds and having the assistance of technical staff in evaluating the requirements can help them determine whether the requirements can be broadened. They noted that if they lack technical expertise in the specific area of requirements, it is more difficult to question whether a statement of work is too restrictive. The FAR does not require actions to be taken in circumstances where only one offer is received in response to a competitive solicitation, such as performing additional market research or determining if the requirements were overly restrictive. However, contracting officials at two of the locations we visited noted that they have a local requirement to document in the contract file the circumstances that may have led to only one offer being received and actions that will be taken to obtain more competition if there is a follow-on procurement. None of the contract files we reviewed where one offer was received included this information. Although the government is generally required to make every effort to obtain as much competition as possible, the contractors themselves make a business decision about when to submit an offer in response to a solicitation. The contractors we spoke with told us that they consider a wide variety of factors before submitting a proposal in response to a solicitation, such as the cost of developing proposals; their ability to provide the services, including key personnel; their knowledge and history of the requirement; rapport with the government personnel; ability to partner with small businesses to meet small business subcontracting requirements; and the potential financial gain from the procurement. There are also certain strategies that companies take when deciding whether to submit a proposal. For example, companies may submit a proposal to test the water or get their name recognized as a potential contractor for a particular requirement, then bid more aggressively for the follow-on procurement. Contractors may also intentionally not submit an offer on a certain procurement to retain their status as a small business— and thus remain eligible for procurements designated for small businesses. A predominant factor that contractors consider when deciding whether to compete for a contract is the performance of the incumbents. Contracting officers and contractor representatives explained that when an incumbent is known, contractors may not compete if the incumbent has historically provided the requirement and is identified as well-performing. The solicitations for all of the 11 orders we reviewed under the Navy’s Seaport- e Multiple Award Schedule and the Army’s Express Multiple Award Schedule listed previous contractors’ names and contract numbers, and only one offer was received for 9 out of the 11 solicitations. In talking to us about another contract in our sample, an Army program official said she believed that vendors other than the incumbent could have competed for a certain contract, but that a short time frame combined with the incumbent’s history on the contract caused many vendors to be disinclined to compete. Ultimately, the Army received only one offer in response to its solicitation. An Army contracting official also noted that when evaluation factors in the solicitations are based mostly on experience with the system and technical skills, other competitors may not submit offers because the cost of developing a proposal is too high to outweigh the risk of not winning the award. Furthermore, several contracting officials and contractors told us that some contractors find it necessary to team up with other contractors in order to fulfill certain government needs, which can also contribute to only one offer being received. For example, a Navy requirement for submarine engineering services was being performed by a large business. The Navy decided to set aside the follow-on contract, estimated at $34 million, for small businesses. Only one small business submitted a proposal, which included 10 subcontractors—1 of which was the large- business incumbent, and another of which was a small business that was originally identified as a possible competitor for the procurement. In another example, one order under the Army’s Express program, estimated at $122 million, was awarded to a small business after one offer was received, helping the Army meet its small business goals. Under a teaming arrangement involving a number of subcontractors, the small business prime contractor was going to perform only 7 percent of the work while one subcontractor was going to perform 87 percent of the work. Under another Express order, a small business was the prime contractor and a large business, which had been the prime contractor under a predecessor contract for the same requirement, was a subcontractor performing the bulk of the work. Under the Express program, the Army claims full small business credit for all obligations under these types of arrangements. In one contractor proposal we reviewed, submitted in response to an Express program solicitation, the small business prime contractor pointed out this benefit. According to SBA officials, as long as the procurement was awarded using full and open competition, the percentage of the work performed by the small business prime contractor is not relevant. Some agencies actively seek out opportunities to compete requirements and contracts that were originally awarded noncompetitively, such as by breaking out components of the requirement that can be awarded competitively. We encountered several contracting and program officials who told us that they broke out pieces of requirements from past sole- source procurements in order to compete them. For example, Coast Guard contracting officials informed us that they broke one requirement for aircraft maintenance and repair into one contract and three separate orders, which they believe will save the government approximately $13 million. In another case, a Navy contracting official informed us that certain requirements for submarine components had been separated from a sole-source procurement and are currently in second or third cycles of competition. We also found one noncompetitive contract at the Navy where the follow-on requirement was competed. The contract, for operations and maintenance support and engineering services, was a sole- source award to an ANC 8(a) firm, but the contracting officer who inherited the contract decided to compete the follow-on contract in the 8(a) program. In some cases, the government actively sought additional vendors for certain requirements. For instance, Air Force officials informed us that they encouraged a second vendor to compete against the incumbent contractor for production of aluminum pallets. These officials noted that at the threat of competition, the incumbent—who ultimately won the contract—”sharpened his pencil,” resulting in savings for the government through a lower price and improved delivery schedules. In another example, the Army’s Tube-launched, Optically-tracked, Wire-guided (TOW) missile has been with one contractor since its inception in the early 1960s. The Army recently made a business case for breaking out a portion of the requirement and competing it as a separate procurement because it had identified another capable contractor: one of the subcontractors with a long history with the program. But other officials noted that it is not always the best business decision to invest time and money into finding other vendors; each situation has to be evaluated on its own merits and future procurements or production lines must be sufficient to warrant the government’s investment in a second source. The C-130J engine has also been broken out of the overall Lockheed Martin contract, with a separate sole source contract to Rolls Royce. According to program and contracting officials, this decision was made to save money. Some contracting officials we spoke with recognized the importance of thorough market research for identifying possible vendors even when it appears that only one contractor is capable of doing the work. For example, during the market research phase for the awarding of two contracts for engineering services for military satellite communications at the Los Angeles Air Force base, the contracting officers requested that potential contractors provide information on their abilities to meet the government’s requirements in an attempt to identify other qualified contractors. Ultimately, however, the two contracts were awarded using a sole source justification that only one responsible contractor was capable of doing the work. The director of contracting at this location informed us that they typically reach out to the open market when they are not familiar with a requirement or when a requirement has been procured on a sole source basis for many years and they wish to test the marketplace to determine if it has changed over time. He also noted that there have been instances in which new contractors have expressed interest, but usually no new contractors come forward. In another case, ICE contracting officials informed us that the program office wanted a specific vendor for a requirement for rifle cases. The contracting officer pushed back against the program’s specific request, competed the requirement, and received numerous offers. The contract was awarded to a vendor that the program office was not aware of, and the contracting officer reported that they were very pleased with the results of the competition. In reviewing the contracts in our sample, we identified contracting approaches for nine contracts or orders that did not reflect sound procurement or management practices, in some cases not leveraging the effectiveness of the market place. These approaches included ambiguously written justifications for noncompetitive contracts, very limited documentation of the reasonableness of contractors’ proposed prices, instances where the contract’s cost grew significantly, and labor categories that were improperly authorized because they were not included in the contract. In addition, our sample contained undefinitized contract actions (UCA) that did not clearly follow UCA policies or did not meet the definitization requirements, which puts the government at risk because contractors lack incentives to control costs during this period. Finally, during our file review, we found an example of a noncompetitive contract awarded in an urgent situation that failed to follow sound procurement practices in several ways, such as drastic increases in ceiling prices, improper modifications to the contract, inappropriate communications between the program staff and the contractor, and a program official serving as the contracting officer’s technical representative (COTR) without the required training. We also found that sound management practices were not followed in the administration of this contract. For two contracts in our sample, the justifications for not competing cited exceptions to competition that were not supported by the circumstances of the procurement or that were the wrong section of the FAR and thus created ambiguity about whether circumstances warranted a noncompetitive award. In the first situation, at ICE, the justification to use a particular company’s online language learning services cited the wrong section of the FAR in two different ways. First, the order was placed under the firm’s GSA schedule contract (pursuant to FAR 8.4) and thus should have been justified under one of the exceptions in FAR 8.405-6, yet the FAR citation was to one of the exceptions to full and open competition under FAR 6.302. Orders placed under GSA schedule contracts are exempt from the requirements in FAR Part 6. Second, the justification itself was not even clear as to the circumstances warranting a noncompetitive order. Specifically, the justification incorrectly cited FAR 6.302-2 as “only one responsible source.” FAR 6.302-2 is used to justify sole source procurements that are urgent and compelling; FAR 6.302-1 is for procurements that have only one responsible source. Further, the justification should have been reviewed by the competition advocate and attorney based on the total estimated value of the procurement—the base year and 2 option years—but it was not. The justification described the features of the services provided, claiming that it was the best product available and that a pilot program testing this product had elicited a positive response. While planning the procurement, the contract specialist pointed out to the program staff that there were 31 GSA vendors that could offer these services and recommended that they try to obtain proposals from at least two other vendors. The contracting officers we spoke with explained that the program office was insistent on the use of this contractor for these services. The program office stated that the order was placed solely with this contractor primarily because DHS had undertaken a successful pilot program for these services with this contractor, and they were under time pressure to award the contract quickly. In the second example, at the Department of the Interior, the justification for a noncompetitive order on a GSA schedule contract to lease information technology licenses was similarly ambiguous because the citation used was FAR 8.405-6(b)(3), for urgent and compelling requirements, but the supporting narrative stated that this vendor was the only distributor that could offer all of the required products and services, i.e., a certain brand name of licenses. The justification also stated that this system was one of three that the government could use to meet its needs. In addition, the program office was pushing for this contractor because it was offering significant discounts if the award was made in a certain time frame. The Department of the Interior issued an order under an IDIQ contract for the Office of Historical Trust Accounting to provide assistance with historical accounting of trust funds for Indian Tribes. The order was placed for over $2.2 million noncompetitively through the 8(a) program to a tribally owned 8(a) firm. The contract specialist sent the firm’s proposal to the program official for price and technical review, and the program official responded in less than an hour that the contractor’s proposal “looked good,” with no documentation or description of what he had reviewed. The contracting officer at the time put a memo in the file stating that pricing for the labor categories was found to be in line with another order on the same contract and the base contract. We discussed this finding with agency officials, and a policy official at Interior’s Acquisition Services Directorate told us that she and other managers in her office have put a renewed emphasis on more detailed price analysis for orders under IDIQ contracts in their reviews of contract actions. Further, during the period of performance, the same program official from the Office of Historical Trust Accounting worked directly with the contractor— significantly overstepping his authority and circumventing the Interior contracting officer—to obtain services that were not included in an order by adding labor categories to the scope of work. The IDIQ contract under which this order had been issued was subsequently transferred to a new contracting officer, who noticed the unauthorized labor categories in the contractor’s invoices. This contracting officer modified the order to incorporate a new statement of work with the additional labor categories and a corresponding price increase of about $500,000. A more detailed price analysis was conducted, including development of an independent government cost estimate for these labor categories which was compared to the contractor’s proposed prices as a basis for the determination that the price was fair and reasonable. In another contract file at the Air Force, no price analysis had been documented for an order for integrated logistics support and engineering services in support of the Air Force’s Distributed Common Ground System. The pricing memorandum in the file for the order contained a brief note that the information was in the base contract file; however, when we looked in the base file the information was not there. After we raised this situation, the former contracting officer prepared a price negotiation memorandum after-the-fact explaining how the government arrived at a fair and reasonable price. To do so, however, the contracting officer had to rely on old e-mails as well as information supplied by the contractor. According to the current contracting officer, the value of this order grew from the initial $9.1 million to $18.8 million at the time of our audit. In another example, ICE contracting officers purchased communication equipment through an order under a Secret Service contract. ICE issued the order noncompetitively, using the justification that only one source was available. The justification stated that only the contractor could provide the equipment as the original manufacturer of a system in which the government had already invested significant resources in training and software. In the order file, the contracting officer noted that price analysis and legal review were not performed because the base contract at Secret Service was competed and prices were determined to be reasonable in part through competition. The base contract, however, was not competed. When we brought this to their attention, ICE contracting officials told us that they had misinterpreted the information in their internal acquisition planning database and from ICE program and senior management officials. Only after the order had been issued did they learn that the underlying contract had not been competed. They recognized that they should not have pointed to competition as a basis for the fair and reasonable pricing in the documentation for this order, but noted that they had compared the prices for this equipment to prices on the open market and GSA schedules contracts—which was noted in the contract file—and that this analysis, along with Secret Service’s determination of a fair price at the time of award of the base contract, would suffice as a determination that the price was fair and reasonable. Finally, a sole source contract at the Army for engineering and maintenance support for the Chinook helicopter program grew over a number of years from $34.7 million to about $477 million, but the acquisition plan was not revised in spite of this significant price increase. The FAR requires that whenever significant changes occur, and no less often than annually, the planner must review the acquisition plan and, if appropriate, revise it. The sole source justification prior to award of the contract, at $34.7 million, was correctly approved by the head of the contracting activity, which is all that is required for that dollar value. Although not reflected in the acquisition plan, a second phase of the requirement was identified and a second justification for this additional work, citing an estimated value of $134.6 million, was reviewed and approved by the Army’s senior procurement executive. When the contract was subsequently modified, however, the value was increased to $477 million with no further notification to the senior procurement executive about the significant price increase. The attorney reviewing the contract modification expressed serious concerns, including that the senior procurement executive was not being notified of the drastic increase in price. A senior DOD acquisition policy official told us that, given the significant increase in the contract’s value, additional notification should have occurred, such as in the form of an amended justification and approval or acquisition plan. Our sample also contained UCAs. In one case, it was unclear from the documentation to what extent the agency followed UCA policies; in another, the agency did not meet the DOD definitization requirements and key documentation was missing from the contract file. UCAs are binding commitments that can be entered into using different contract vehicles (i.e., letter contracts, orders under IDIQ contracts, or modifications to an existing contract). They are intended to be used only when the government needs the contractor to start work quickly and there is not enough time to negotiate all the terms and conditions for a contract. UCAs are required to be definitized within 180 days, or when the amount of the funds obligated under the contract action are 50 percent or more of the not-to-exceed price, to limit the risk to the government. In one case at the Air Force, an order was undefinitized for 17 months. This order, to provide F-15 engines to the Royal Saudi Air Force, was issued under a sole source IDIQ contract for development, production and other support of the F-15 weapon program. In addition to this lengthy undefinitized period, the contractor had begun work 7 months before the UCA was even issued, but neither the contracting officer nor program office official could locate any documentation showing that the government had authorized this work to begin. The Defense Federal Acquisition Regulation Supplement (DFARS) states that, while foreign military sales are not subject to the DFARS policy for UCAs, including the definitization requirements, contracting officers should apply the definitization requirements to the maximum extent practicable. The original contracting officer was no longer available, but we discussed this matter with the current contracting officer and an official from the program office, who were unable to explain the circumstances surrounding the initial authorization to start work or the lengthy undefinitized time frame. In another example, a UCA at the Navy remained undefinitized for 7 months, thereby not meeting DOD definitization requirements. This IDIQ contract was awarded noncompetitively to an ANC 8(a) firm after the previous contractor, another 8(a) firm, failed to meet certain contractual requirements and the government needed to quickly put another contract in place to avoid a break in service. Further, when the contract was definitized at an agreed to price of about $131 million, the only documentation included in the modification definitizing the contract was the award term plan and some contract line items, but no description of the scope of work required. The contracting officer explained that after a reorganization at the Navy, the contract file had been transferred to her from another contracting office with missing documentation and she could only make notes where documentation was missing. She said she would have expected more information to be included in the modification that definitized the contract. In one example, the government failed to follow sound procurement practices in several ways and did not adhere to certain sound management principles, or internal controls, in others. In the period following the terrorist attacks of September 11, 2001, Interior initially issued an order in our sample, on behalf of the Federal Bureau of Investigation (FBI), under a $50 million contract to an 8(a) ANC firm; the contract value subsequently grew to $100 million. This order was improperly transferred by the Interior contracting officer to a second contract—the ceiling price of which tripled, from $100 million to $300 million—with the same vendor. A modification in the file stated that “this task order is hereby transferred” from the first contract to the second contract after the ceiling on the first contract was reached. The contracting officer subsequently noted that some requirements under the order were not incorporated into the second contract, and took steps to modify the second contract. In addition, the work under the first contract included commercial-off-the-shelf information technology and telecommunication hardware and software and support services for civilian and DOD agencies; modifications to the contract added commercial and institutional building construction. The work under the second contract included information analysis and technical assistance support to provide a turnkey solution for operational support for services to the Foreign Terrorist Tracking Task Force, including quick reaction support for assessment planning and analysis and “other special requirements.” Additional labor categories—such as counter terrorism operations specialists—were later added. Interior contracting officials and FBI program officials acknowledged that the FBI’s requirements had changed significantly from a basic information technology support project to running a 24-hour program to track terrorist activities. The FBI officials noted that they were responding to a presidential directive to establish a terrorist watch list within a matter of months and that there was intense pressure to have this call center up and running quickly. Also, according to contract file documentation, the government accepted and paid for an invoice submitted by the contractor for a service—security guards—that was not within the scope of the first contract. The Interior contracting officer, in later scrutinizing the contractor’s invoices, noticed this out-of-scope issue and obtained reimbursement from the contractor. In another instance, the contracting officer noticed more improper billing, including an unauthorized expense for a hotel bar tab, and deducted the costs from the invoices. When the second contract was about to reach its ceiling price, Interior put in place a third contract with the same 8(a) ANC firm with an estimated ceiling price of $1 billion, with the contracting officer noting that after that award they would no “longer have to worry about contract ceilings” with this vendor. Finally, the FBI program official designated as the COTR did not have the necessary training to fulfill this position; the Interior contracting officer subsequently removed the official from the position. Apart from the issues with procurement practices discussed above, the administration of this contract lacked appropriate management controls— also referred to as internal controls. Examples follow. We found little evidence in the contract file that the contractor’s proposed prices had been analyzed for price reasonableness at various points during the life of the order. For example, the FBI program official’s analysis of the contractor’s proposed price for the order under the initial contract was a statement in an email that she had reviewed the price and found it reasonable, with no further documentation supporting the statement. Further, during preparations to award an option year on this order, an FBI program official again approved the contractor’s revised price proposal with a simple “yes” response when asked by the Interior contracting officer, and also explained to the contracting officer that the independent estimate—presumably for this option year—was different from the contractor’s proposed prices in part because the contractor included additional labor categories that the FBI did not require, but that they “agreed with.” Communications between the FBI and Interior were problematic. Specifically, the program office was communicating directly with the contractor about the growth in the requirements but not involving the contracting officers. For example, the contractor informed the Interior contracting officer that it had been directed by the program office, due to a change in requirements, to establish an increased level of continuity and retention within the terrorist screening center— particularly for the second and third shifts. The contractor proposed additional compensation for these shifts and told the contracting officer that approval to apply these shift differentials would be approved by the program office . The contracting officer told us the program office should have informed her first of the need for shift differential compensation and that the direction to the contractor should have come from her rather than from the program office. Interior contracting officials expressed dismay at the program office’s lack of communication with them and told us that sometimes they did not know what was going on with their own contract. Further, when Interior was preparing to award the second contract to the ANC firm, an FBI program official told the contracting officer that she understood delays in the award may be due to Interior’s legal review process; the FBI official, in an effort to expedite the process, then asked for the legal representative’s contact information to “move this along smoothly.” Interior officials told us that after several years of dealing with this contract, their office had undergone a culture change whereby they were starting to push back on customer demands instead of simply doing what the program offices wanted. Eventually, after a dispute about the interagency contracting fees the FBI was paying to Interior and Interior’s desire to not award another sole-source contract to the same vendor, the FBI officials told us that they pulled the requirement in-house and, under their own contract, awarded the requirement noncompetitively to another ANC 8(a) firm. FBI officials added that by this time, they had increased their own contracting staff so were able to handle this requirement themselves. Statute and regulation require that each executive agency establish an “advocate for competition,” commonly known as a competition advocate, at the agency level as well as at each procuring activity. In general, the agencies in our review have organized their competition advocates into the required department- and procuring activity levels. For example, the Department of the Interior has a department-level competition advocate and competition advocates at the bureaus, including the National Business Center, which houses the Acquisition Services Directorate. DOD has a competition advocate at the department level, and additional competition advocates are in place at the Army, Air Force, and Navy and local contracting activities. The advocates are to carry out a number of broad responsibilities, including promoting full and open competition and challenging barriers to competition, reporting to the agency’s senior procurement executive and chief acquisition officer on opportunities and actions taken to achieve competition, as well as conditions or actions that unnecessarily restrict it, such as unnecessarily detailed specifications or restrictive statements of work, and recommending to the senior procurement executive and chief acquisition officer a “system of personal and organizational accountability” for competition, which may include recognition and awards to program managers, contracting officers, or others. Competition advocates are also responsible for approving justifications for other than full and open competition within certain dollar limits, as depicted in table 3. Apart from these duties, agencies are left with much discretion regarding where in the organization the competition advocates should be placed, who should be appointed to this position, and how they should carry out their responsibilities. Some agency officials we spoke with said that, because the FAR is vague in this regard, and especially given the current emphasis on competition, more guidance related to this position could be helpful. For the agencies in our review, we found a range of approaches to the competition advocate position and placement, skills and expertise, and methods of implementing their responsibilities. Officials at the agencies we reviewed generally agreed that placement of the competition advocate at a high level within agencies is important, as the person in this role should have the clout to make difficult decisions regarding proposed sole-source awards and support procuring activity contracting officers who attempt to do so. A competition advocate who can exercise this level of leadership at a senior level can be more effective in the role and, by emphasizing the importance of competition to program office staff as well as contracting officers, has the potential to affect competition results. The Air Force competition advocate told us, for example, that he sees value in being highly placed within the organization; if people must come to him to explain why a particular procurement needs to be sole source, they are more likely to do things the right way and less likely to take short cuts. At the agencies in our review, competition advocates are in various positions and placements within the organization. Some are senior leaders in the acquisition arena. For example, the DOD competition advocate is the Director of Defense Procurement, Acquisition Policy, and Strategic Sourcing, and the DHS competition advocate is the Director of Oversight and Strategic Support within the Office of the Chief Procurement Officer. We found that the duties of competition advocate can be automatically tied to the person’s position within the agency. For example, Air Force, Navy, and Army acquisition regulations designate a specific deputy assistant secretary position to be the advocate. Agency officials told us that the ICE Deputy Assistant Secretary for Management is automatically designated as the competition advocate, and within the Secret Service, it is the Deputy Assistant Director. Also within DHS, the head of contracting activity was the competition advocate for the Coast Guard, but because this person was uncomfortable signing sole source justifications in two different capacities, she told us that she delegated her role as competition advocate to another senior official—the Deputy Assistant Commandant for Acquisition—within the acquisition directorate. Within Interior, the department competition advocate is a senior procurement analyst who reports to the senior procurement executive. The competition advocates at the bureaus, components, or procuring activities included in our review were, for example, acquisition policy chiefs or senior contracting officers. We found one situation, at ICE, where the competition advocate is higher in the organization than the head of contracting activity. Given the approval thresholds stipulated in the FAR, this means that the head of contracting approves justifications at a higher dollar threshold than the competition advocate, but in practice reports to the competition advocate. The competition advocate explained that ICE does not administer many high-dollar acquisitions, and therefore, having the competition advocate in a management position above the head of contracting brings greater visibility to noncompetitive requirements at lower-dollar thresholds. Expertise and Background Agencies can appoint any individual, other than the senior procurement executive, as a competition advocate as long as the individual’s broader functions and duties do not conflict with the responsibilities of a competition advocate as outlined in statute. Agency officials told us that a competition advocate should have the right skill set, which may include a background that enables them to recognize and question overly restrictive requirements which could lead to an unnecessary sole-source outcome, and the personality to ask tough questions. In general, we found that no specific qualifications were required for this role, but the Air Force’s Competition and Commercial Advocacy Program does call for its competition advocates to have “extensive qualifications and knowledge of the types of acquisitions” the procuring activity engages in. The department- and component-level competition advocates we spoke with had a variety of backgrounds. Several of them had been contracting officers and program managers, while one was involved in the operations side of the agency, and another served as an attorney to the agency. Some competition advocates offered examples of how their background and expertise influence how they approach their job of promoting competition. Interior’s competition advocate informed us that her experience as a contracting officer and bureau competition advocate influences how she reviews urgent and compelling requirements. During the Army’s deployment in Bosnia, she was a contingency contracting officer and processed wartime requirements. She said that because she worked with urgent and compelling requirements during the deployment, she is more likely to challenge sole source emergency requirements in her capacity as the competition advocate. In addition to her experience as a contingency contracting officer, she also served as the competition advocate for the U.S. Fish and Wildlife Service during Hurricane Katrina, service that required her to balance the needs of emergency response with her responsibility as a competition advocate to promote competition. She believes that her past experience leads her to have a different attitude toward her role as a competition advocate in comparison to other advocates who may be less likely to challenge emergency requirements. The Coast Guard’s competition advocate explained that his experience as a program manager allows him to ask questions that a contracting officer may not think to ask about a requirement. Over 25 years ago, he began working in structural design and subsequently served in several supervisory and program manager capacities in the Navy. His experience in program, fiscal, and technical management influences how he approaches his role as competition advocate. He reviews contract documentation strictly from the perspective of promoting competition. He stated that when a program office tries to solicit a requirement in a way that precludes certain vendors, his technical background of working with various platforms leads him to question these restrictive requirements. The competition advocate at one of the Air Force contracting activities we visited pointed to her experience as a program manager and her contracting background as an asset to her current position as competition advocate. She also noted that the support that she receives and relationships she has with the command-level and Air Force competition advocate has helped her challenge sole source requirements when program managers are pushing for a certain vendor. At one Navy location, the advocate pointed to her experience and familiarity with the program offices’ requirements as a plus in enabling her to question planned sole-source procurements. The Secret Service’s competition advocate has years of experience in operations but no experience as a contracting officer or a program manager. He told us he began working as a criminal investigator and worked in other positions, such as on the president’s protection detail and in congressional affairs. After over 20 years of working in operations for the agency, he told us he is learning about contracting operations and that he believes more agents who work in operations should hold the position of competition advocate. He works closely with the head of contracting activity, relying on her expertise and supporting her efforts to increase competition within the agency. The competition advocates at the agencies in our review are tackling their jobs in a variety of ways. Most department-level advocates told us that they review, and in some cases sign, justifications for sole source procurements that must also be approved by the senior procurement executive, that is, justifications for the highest dollar amounts, as another check point in the process to question planned sole-source procurements. The Army’s competition advocate said that one trigger for potentially rejecting a justification for a sole-source procurement is when the evaluation of the market is cursory. For example, a justification for a sole- source award to an 8(a) ANC firm was turned back for additional detail on market research and further review of the feedback from other potential offerors, although the same firm ended up with the follow-on contract. As noted above, competition advocates are required to recommend to the senior procurement executive and the chief acquisition officer a “system of personal and organizational accountability” for competition. The Navy competition advocate and his staff told us that, as part of their review of high dollar value justifications prior to the senior procurement executive’s review, they look to see whether program offices have made strides in improving competition in their programs if a prior justification had made this claim. Holding the program offices accountable is part of their overall plan for improving competition, an attitude that is also strongly held by the Navy’s senior procurement executive. Other competition advocates also pointed out that strong leadership, from the senior procurement executive and the competition advocate, can engender results. The DHS competition advocate, for example, noted that without this level of continued leadership and the will to enforce accountability, there could be a slip back to less competition. Other methods we found that competition advocates are using include: DOD officials told us that the DOD competition advocate holds quarterly meetings with competition advocates from the military services and other DOD agencies to review the progress toward meeting competition procurement goals and to challenge the barriers they identify as inhibiting competition. The Navy competition advocate told us that he holds quarterly contracting council meetings with senior contracting staff for each of the systems commands (some of whom are also competition advocates), discussing various topics including competition. Interior and DHS do not hold regular meetings with the competition advocates of bureaus or components, but, according to agency officials, they have other meetings with procurement staff where competition goals are discussed. The DHS competition advocate recently completed an investigation of noncompetitive contract awards at DHS and found that some contract files lacked required justifications for sole source procurements. Other contract files included justifications that did not adequately describe why only one source could perform the work. The competition advocate subsequently approved a memorandum signed by the Chief Procurement Officer that emphasized the need to include the justification in the contract file and provided examples of inadequate rationale for use of a noncompetitive contract. Competition advocates can also offer recognition and awards—including to program managers—for efforts to increase competition. Some advocates indicated that their agencies or procuring activities have competition awards programs that recognize the work of individuals or teams who increase competition. For example, since July 2007, DHS has instituted the DHS Competition and Acquisition Excellence Awards Program “as a means of renewing and increasing acquisition workforce interest in competition and related innovative procurement practices.” The Army competition advocate also said the Army was going to add competition as a metric to the Secretary of the Army’s award program, awarding contracting activities that raised and exceeded their competition goals. For their part, local competition advocates told us that they try to get involved in acquisitions as early as possible to have a greater impact on decisions related to competition. For example, the advocate at the Naval Air Warfare Center Weapons Division said that, for service contracts, she reviews acquisition strategies before she reviews justifications. If she identifies a potential issue related to competition in the acquisition strategy stage, she can have the program make appropriate changes before the justification is ever developed. The Warner Robins Air Logistics Center competition advocate and her staff also try to engage program offices early in the process and, if the procurement must be noncompetitive, the approval process is faster because they are already familiar with the support for the justification. The Air Force competition advocate noted that it is his responsibility to ensure that competition advocates in the field have a certain level of independence to allow them to push back, especially with regards to the requirements side, and to make sure there is a process that allows them to raise issues up the chain when warranted. Some degree of noncompetitive contracting is unavoidable, such as when only one responsible source can perform the work; and in some cases competition is impractical due to the government’s reliance on contractors stemming from decisions that were made long ago. Recent congressional actions to strengthen competition opportunities in major defense programs may take some time to demonstrate results. Further, OMB’s efforts to reduce agencies’ use of high risk contract types may help agencies refocus and reenergize efforts to improve competition. Despite these actions, other targets of opportunity still exist, but to take full advantage of them, it will be necessary to challenge conventional thinking to some extent. Key among these are establishing an effective, adequately trained team of contracting and program staff working together, starting early in the acquisition process. Competition opportunities should be considered when requirements are initially developed, and as complex programs mature and the government gains more knowledge about what it needs. Because program officials have an essential role in the acquisition process, as do contracting officers, it is just as important for them to advance competition whenever possible. Given the nation’s fiscal constraints, it is not acceptable to keep an incumbent contractor in place without competition simply because the contractor is doing a good job, or to resist legitimate suggestions that competition be imposed even though it may take longer. As discussed in this report, some agencies have implemented the leadership and accountability to make progress in this area, such as breaking out requirements to facilitate competition. However, there is no requirement to assess the circumstances under which competitive solicitations receive only one offer to potentially bring about a greater response from the market place. The competition advocates, in their unique role and in the context of OFPP’s call to reinvigorate their role, have the potential to implement changes to practice and to culture. However, to do so they need to be situated in the right organizational position and able to bring to bear the acquisition knowledge and leadership to engender change. We recommend that the Administrator of the Office of Federal Procurement Policy take the following three actions: Determine whether the FAR should be amended to require agencies to regularly review and critically evaluate the circumstances leading to only one offer being received for recurring or other requirements and to identify additional steps that can be taken to increase the likelihood that multiple offers will be submitted, with the results of the evaluation documented in the contract file. As part of efforts to reinvigorate the role of the competition advocate, issue guidance to federal agencies regarding appropriate considerations when appointing competition advocates, such as placement within the organization, skill set, and potential methods to effectively carry out their duties. Direct agencies to require their competition advocates to actively involve program offices in highlighting opportunities to increase competition. We requested comments on a draft of this report from OFPP, the departments of Defense, Homeland Security, and the Interior, and the SBA. In oral comments provided via email, the OFPP Administrator concurred with our recommendations, noting that they are consistent with the types of steps agencies have begun to take in response to the President’s direction to be more fiscally responsible in their contracting practices and to reduce use of high-risk contracting practices that can lead to taxpayers paying more than they should. The Administrator noted that there is still much work ahead and that OMB will periodically meet with agencies to review progress against their risk reduction goals. He stated that these efforts will include a push to achieve greater collaboration between contracting, program, finance, and other key stakeholder offices in the acquisition process. The Administrator also said that his office would continue in its efforts to build the capacity and capability of the acquisition workforce—including program and project managers and COTRs—to ensure that agencies are well-equipped to take the actions necessary to maximize the benefits of competition. The Administrator also stated that executive branch actions to draw agency attention to high risk contracting and establish goals for risk reduction provide a catalyst for change and that key among these are establishing an effective, adequately trained team of contracting and program staff working together, starting early in the acquisition process. Finally, the Administrator highlighted OFPP’s October 2009 memorandum, which stated that a spend analysis might be useful for identifying and analyzing competitions where only one offer is received, by comparing levels of competition achieved by different organizations within the agency or those similarly situated in other agencies to determine if more successful practices may exist for more competition for a given spending category. DOD also provided oral comments via email. The Director, Defense Procurement and Acquisition Policy, stated that, in general, he agrees with our findings and recommendations concerning opportunities to increase competition in cases where only one offer is received (a situation DOD terms “ineffective competition”). He stated that the department is taking a number of actions to increase the quality of competition in this regard. For example, the competition advocates will be required to measure and report on “ineffective competition,” contracting officers will be directed to perform cost analysis in all situations where only one offer is received, and the intent is to form a Contracting Integrity Panel subcommittee to specifically look at creating opportunities for more effective competition. In its written comments, Interior pointed out that its September 2008 internal policy, “Enhancing Competition,” is one step the department has taken to enhance competition. Interior also commented that we should clarify that noncompetitive orders justified as logical follow-ons are permitted. We believe our report makes this clear. In addition, Interior commented that one of the limited sources justifications discussed in our report, where the wrong FAR citation had been used, was a “minor typographical error.” We disagree. Because of this error, it is not clear what exception was being used and, therefore, the rationale is ambiguous, as we state in the report. Interior provided additional technical comments, which we incorporated as appropriate. Interior’s comments are included as appendix II. DHS had no comment on the draft report. We received technical comments from the SBA, which we incorporated where appropriate. The FBI requested a copy of the draft report and provided technical comments pertaining to one contract in our sample, which we also incorporated where appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution of it until 30 days from the date of the report. We will then send copies of this report to interested congressional committees and the Secretaries of Defense, Homeland Security, and the Interior; and to the Administrators of SBA and OFPP. This report is also available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-4841 or huttonj@gao.gov. Contact points for our offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to the report are listed in appendix III. The objectives of this review were to assess (1) the extent to which agencies are awarding noncompetitive contracts and contracts awarded competitively with only one offer received; (2) the exceptions to competition that agencies used when awarding noncompetitive contracts; (3) factors that affect competition in federal contracting; and (4) the extent to which the contracting approaches for the contracts in our sample reflected sound procurement or management practices. We also identified how agencies are instituting the roles of their competition advocates. To address these objectives, we identified through the Federal Procurement Data System-Next Generation (FPDS-NG) government-wide obligations to noncompetitive contracts in fiscal year 2008, the most recent available when we began our review. We included contracts and orders coded as “not competed,” “not available for competition,” “follow on to competed action” and “noncompetitive delivery order.” We found that a small percentage of obligations for orders under indefinite delivery / indefinite quantity (IDIQ) contracts were unlabeled in FPDS-NG as to extent of competition. We were able to match many of these unlabeled orders to their base contracts to obtain more complete information. In addition, we identified the fiscal year 2008 obligations under contracts where only one offer had been received. To select the agencies to include in our review, we identified the five agencies with the highest reported percentage of obligations under noncompetitive contracts in fiscal year 2008. These included the Air Force, Army, Navy, the National Aeronautics and Space Administration (NASA), and the Department of Homeland Security (DHS). For these agencies, the percentage of noncompetitive obligations ranged from 45.2 percent (Navy) to 25.2 percent (DHS). We performed additional analysis of NASA and DOD obligations, focusing on the types of services and products represented by the noncompetitive obligations, and found that the products were largely what is generally considered to be specialized equipment. To focus our review on services, for DOD and NASA we limited our analysis to the “R” codes in FPDS-NG, which reflect professional, administrative, and management support services. Each of the DOD agencies and NASA had a significant percentage of noncompetitive contracts for these services. After discussion with our congressional requesters, we eliminated NASA from our scope of work and focused on the DOD agencies, DHS, and the Department of the Interior. To identify the components, or specific procuring activities, within each of these agencies for our contract file reviews, we focused on those whose percentage of fiscal year 2008 obligations under noncompetitive contracts exceeded that of the agency (Army, Navy, Air Force, DHS, and Interior) as a whole. We also focused on those components with $100 million or more in contract obligations in 2008 and those with higher-dollar procurements, in order to avoid selecting low-dollar procurements. We then considered other factors, such as travel expenses, the locations’ percentage of obligations under noncompetitive contracts, and the locations’ percentage of obligations under contracts with only one offer received, into account in making our final selection. Following is more specific criteria applicable to the agencies in our review. For the Army, Air Force, and Navy, our analysis of fiscal year 2008 noncompetitive obligations was limited to obligations for professional, administrative, and management support services. For DHS: the Coast Guard, the Secret Service, Immigration and Customs Enforcement (ICE), Customs and Border Protection, the Federal Emergency Management Agency (FEMA), and the Transportation and Security Administration met our initial selection criteria. We removed the Coast Guard and FEMA from the scope of our review due to our continued audits of the Coast Guard’s Deepwater program, and FEMA because it had also been the subject of many recent audits. (However, we did speak with the Coast Guard’s competition advocate and head of contracting activity as part of our fifth objective, as discussed below.) Further, after discussion with the DHS Inspector General, we also removed the Customs and Border Protection and Transportation and Security Administration from our scope to avoid duplication of effort, as the Inspector General had completed or had current work underway on noncompetitive contracts at those locations. Therefore, we focused on the Secret Service and ICE. At Interior, two components—the Acquisition Services Directorate and Bureau of Indian Affairs—met our initial selection criteria. We included the Acquisition Services Directorate in our review due to its proximity to Washington, D.C. This fee-for-service contracting office (formerly GovWorks) awards and administers contracts on behalf of other federal agencies. Included in our contract sample from the Acquisition Services Directorate were contracts awarded on behalf of DOD agencies, the Federal Bureau of Investigation, and the National Institutes of Health. We then randomly selected a sample of 107 contracts and orders to review in depth. While our focus was on noncompetitive contracts, we also selected a small sample of competed contracts where only one offer had been received in order to gain an understanding of the circumstances leading to that situation. Table 4 shows the specific locations we visited, along with the number and type of contracts reviewed. To identify the extent to which agencies have reported obligations under noncompetitive contracts, we analyzed FPDS-NG data from fiscal years 2005 through 2009 using the fields “not competed,” “not available for competition,” and “follow on to competed action.” We determined that a contract or order was miscoded in FPDS-NG if it was coded as not competed, but our analysis of the contract file documentation showed that the contract or order was competed. Similarly, if a contract or order was coded as competed with one offer received, we determined that it was miscoded if the contract documentation showed that the requirement was not competed, or that it was competed and received more than one offer. We did not make a determination whether the “reason not competed” field in FPDS-NG was coded correctly. We also analyzed reported obligations under competed contracts where only one offer had been reported for the same time period. DOD’s total obligations in fiscal year 2009 reflect an approximately $13.9 billion downward adjustment made by DOD to correct an administrative error made in fiscal year 2008. As this adjustment significantly affected DOD’s reported obligations in fiscal years 2008 and 2009, the figures we report reflects what DOD’s total obligations would have been had the error not occurred. We also analyzed FPDS-NG data by quarter for fiscal years 2007 and 2008 to identify trends in obligations under existing and newly awarded noncompetitive contracts. In reviewing the contract files in our sample, we compared the reported competition data in FPDS-NG to the actual data in the contract file to determine if discrepancies existed in the way competition had been coded. As discussed in the first objective in this report, we found that about 18 percent of the contracts or orders had been miscoded as either competed or not competed. We found the FPDS-NG data to be adequately reliable for overall trend analysis on extent of competition and for selection of locations for our file reviews. To determine the exceptions to competition that were used, the factors affecting competition, and the extent to which certain contracting approaches reflected sound procurement or management practices, we reviewed documentation in the contract files such as the written justification, acquisition plan, statement of work, price negotiation memorandums, records of market research, and other key documents. Where our sample involved orders under IDIQ contracts, we reviewed the base contract file as well. We reviewed pertinent legislation, such as the Competition in Contracting Act, the Federal Acquisition Streamlining Act, National Defense Authorization Acts, and the Weapon Systems Acquisition Reform Act. We also reviewed relevant provisions in the Federal Acquisition Regulation (FAR), specifically Parts 6, 8, 16, and 19, Small Business Administration regulations, and pertinent agency guidance and supplements to the FAR. We also reviewed GAO and Inspector General audit reports dealing with competition. At the locations we visited, we interviewed the contracting officer and contract specialist responsible for the files we reviewed (when available) and, for many of the contracts, also interviewed cognizant program officials to obtain their views. In addition to discussing the specific issues related to the contracts in our sample, we also discussed general topics with these officials, such as their views on barriers to competition and how, if at all, they interact with the agency or local-level competition advocate. We also interviewed procurement policy officials at the department and local levels and a limited number of contractor representatives. To determine how agencies are instituting the role of the competition advocate, we reviewed statutory and FAR provisions, Office of Management and Budget and Office of Federal Procurement Policy memorandums (such as the May 2007 memorandum on reinvigorating the role of the competition advocate), pertinent agency regulations and guidance, and the annual competition reports for fiscal years 2008 and 2009 for the agencies in our review. We interviewed the competition advocates at DOD, the Army, Navy, and Air Force, the Department of the Interior, and DHS, as well as the advocates at the components included in our review. We discussed the competition advocates’ placement within their organizations, their backgrounds and areas of expertise, their strategies for promoting competition, and factors they identified as barriers to competition. We conducted this performance audit from October 2009 to July 2010, in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the contact named above, Michele Mackin, Assistant Director; Tatiana Winger; Julia Kennon; Anh Nguyen; Kenneth Patton; Jared Sippel; Sylvia Schatz; Kristin VanWychen; and Keo Vongvanith made key contributions to this report.

Competition is a critical tool for achieving the best return on the government's investment. While federal agencies are generally required to award contracts on the basis of full and open competition, they are permitted to award noncompetitive contracts in certain situations. Agencies are also required to establish competition advocates to promote competition. GAO assessed (1) trends in noncompetitive contracts and those receiving only one offer when competed; (2) exceptions to and factors affecting competition; (3) whether contracting approaches reflected sound procurement practices; and (4) how agencies are instituting the competition advocate role. GAO reviewed federal procurement data and 107 randomly selected contracts at the departments of Defense, Interior, and Homeland Security (which had among the highest noncompetitive obligations in fiscal year 2008) and interviewed contracting and program officials, competition advocates, and contractors. From fiscal years 2005 to 2009, reported obligations for noncompetitive contracts decreased from about 36 to 31 percent of total obligations, while obligations under contracts competed with only one offer received were steady, at about 13 percent of the total in each year. In comparing the data in the federal procurement data system to the information in contract files, we found that about 18 percent of the contracts sampled were coded incorrectly--as either not competed when they had been, or as competed with one offer received when they had not been competed at all. Agencies used a variety of exceptions to competition for the contracts and orders in our sample, with the two most common being "only one responsible source" and sole-source awards under the Small Business Administration's 8(a) business development program. For services supporting DOD weapons programs, the government's lack of access to proprietary technical data and decades-long reliance on specific contractors for expertise limit--or even preclude the possibility of--competition. In other cases, program offices may press for contracts to be awarded to the incumbent contractor without competition, largely due to their relationship and the contractor's understanding of program requirements. For competitive procurements where only one offer is received, factors include a strong incumbent, sometimes coupled with overly restrictive government requirements, or vendors forming large teams to submit one offer for broader government requirements, whereas previously several vendors may have competed. Contracting approaches for nine contracts reviewed did not reflect sound procurement practices and in some instances sound management practices, in some cases not leveraging the effectiveness of the market place. These approaches included ambiguously written justifications for noncompetitive contracts, very limited documentation of the reasonableness of contractors' proposed prices, instances where the contract's cost grew significantly or where labor categories were improperly authorized, and undefinitized contract actions that did not meet definitization requirements. Agencies have much discretion regarding where in the organization the competition advocates should be placed, who should be appointed to this position, and how they should carry out their responsibilities. As a result, agencies have taken a range of approaches regarding the placement of the competition advocates, their skills and expertise, and the methods they use to carry out their responsibilities. Some advocates cited their experience in program offices as helping them to question requirements that may be overly restrictive, while others had been contracting officers or procurement policy officials before assuming the position. Some agency officials said that regulations are vague regarding the role of the competition advocate, and that given the Office of Federal Procurement Policy's (OFPP) recent emphasis on competition, they would like to see more guidance on competition advocate roles and methods of implementing their duties. GAO recommends that OFPP take actions regarding assessment of the reasons only one offer is received and issue guidance on competition advocate roles, including their direct involvement with program offices to seek opportunities for competition. OFPP agreed with the recommendations, and DOD generally agreed with our findings and recommendations. Other agencies provided technical comments.

Many DOD organizations, collectively known as the missing persons accounting community, have a role in accounting for the missing, as discussed below. The Under Secretary of Defense for Policy (USD (PACOM) are the two top-level Policy) and U.S. Pacific Commandleadership organizations in the accounting community. USD Policy is responsible for developing, coordinating, and overseeing the implementation of DOD policy to account for personnel unaccounted for as a result of hostile acts. The Deputy Assistant Secretary of Defense for Prisoner of War/Missing Personnel Affairs, who reports to the Under Secretary of Defense for Policy, is responsible for, among other things, exercising policy, control, and oversight for the entire process of accounting for missing persons; monitoring and advocating for program funding requirements and resources for the mission; and leading and coordinating related communications efforts, such as the public outreach program. The Defense Prisoner of War/Missing Personnel Office (DPMO) is responsible for, among other things, overseeing archival research and standardizing procedures for methodology and prioritization; rendering final analytic judgments as to what constitutes fullest possible accounting for each case by identifying possibilities for future action, or determining when no further pursuit is possible; and defining, maintaining, and enumerating accounting lists. The DPMO Director is responsible for overseeing the execution of DPMO’s mission and duties. The Deputy Assistant Secretary of Defense for Prisoner of War/Missing Personnel Affairs serves as the DPMO Director and reports to USD Policy in that capacity as well. PACOM exercises authority over the Joint Prisoner of War/Missing in Action Accounting Command (JPAC), which is responsible for conducting operations in support of achieving the missing persons accounting mission. In 2003 JPAC was established as a Joint Command by the merger of the Joint Task Force-Full Accounting with the Central Identification Laboratory – Hawaii in order to achieve unity of command, permanence of operational elements, and efficiency and effectiveness in the use of DOD’s resources, as well as to strengthen the command and control of military forces in achieving the fullest possible accounting. JPAC’s functions include analysis, archival research, investigations, recoveries, repatriations, identifications, and reporting. The Central Identification Laboratory is the laboratory component of JPAC. The military services have a role, with their service casualty offices serving as the primary liaison for families concerning missing persons recovery and accounting. Officials from these offices also assist families and help explain the methods used to account for their missing loved ones. Additional activities include gathering family deoxyribobucleic acid (DNA) reference samples, coordinating responses to family inquiries and concerns, and maintaining family contact information. The past conflict accounting section of the Armed Forces DNA Identification Laboratory conducts DNA analyses of remains of missing persons from past military conflicts for JPAC and its laboratory component, the Central Identification Laboratory, and maintains the past conflict accounting family reference sample database, to include processing of all DNA references. The Armed Forces DNA Identification Laboratory is part of the Armed Forces Medical Examiner System, which reports to the Army Surgeon General. The Life Sciences Equipment Laboratory provides technical and analytical support to the accounting community, and is primarily tasked by JPAC’s Central Identification Laboratory to analyze and identify life science equipment-related artifacts that have been recovered and may potentially be related to missing persons cases. The Life Sciences Equipment Laboratory is part of the Air Force Materiel Command. In addition to these members of the missing persons accounting community, many other organizations play a role in the missing persons accounting process, including the Office of the Under Secretary of Defense for Personnel and Readiness, the Chairman of the Joint Chiefs of Staff, the Office of the Under Secretary of Defense for Intelligence, and the State Department. In addition, family and veterans organizations serve as constituency groups to the accounting community. The department’s response to the accounting-for goal established in the National Defense Authorization Act for Fiscal Year 2010 brought into sharp relief longstanding disputes that have not been addressed by top- level leaders, and have been exacerbated by the accounting community’s fragmented organizational structure. As I will describe in more detail later in this statement, leadership from the Under Secretary of Defense for Policy and Pacific Command have been unable to resolve disputes between community members in areas such as roles and responsibilities and developing a community-wide plan to meet the statutory accounting- for goal. Further, the accounting community is fragmented in that the community members belong to diverse parent organizations under several different chains of command. With accounting community organizations reporting under different lines of authority, no single entity has overarching responsibility for community-wide personnel and other resources. For example, although the Deputy Assistant Secretary of Defense for Prisoner of War/Missing Personnel Affairs has statutory responsibility for policy, control, and oversight of the entire accounting process, JPAC—which performs investigations, recoveries, identifications, and other key functions—falls under the authority of PACOM, rather than reporting to the Deputy Assistant Secretary of Defense for Prisoner of War/Missing Personnel Affairs. As a result, no single entity can implement or enforce decisions without obtaining widespread consensus. We have previously reported that having a single designated leader is often beneficial because it centralizes accountability for achieving outcomes and can accelerate decision-making. Concerns have arisen over the years, both within and outside of DOD, with regard to whether the current organizational structure of DOD’s missing persons accounting community enables the community to most effectively meet its mission. For example, a 2006 Institute for Defense Analyses study concluded that significant improvements could be made by increasing the lines of coordination in the accounting community and recommended that the community acknowledge DPMO as the leader in the accounting effort. The study also described some of the problems associated with the current organization; for example, that DPMO does not have tasking authority over the other organizations, and that while there are multiple lines of authority, no one organization has effective authority over execution of the entire mission. In our July 2013 report, we found that a majority of accounting community and DOD stakeholder organizations believe that an alternative organizational structure for the accounting community would be more effective. We administered a questionnaire asking representatives from each accounting community organization whether various options for reorganizing the missing persons accounting community could improve the community’s ability to meet its mission. One question asked respondents to rank five organizational options that would best enable the accounting community to meet its mission. We found that 12 of the 13 survey respondents who answered the question ranked an option with a more centralized chain of command as the most effective in enabling the accounting community to achieve its mission. Ten of these 12 respondents ranked the current organizational structure as the least effective or second least effective option for achieving the mission of the accounting community. Responses to our questionnaire also demonstrated a lack of confidence about the current organizational structure among many community and DOD stakeholder organizations. For example, 13 of the 14 survey respondents indicated that the current organizational structure did not enable or only somewhat enabled the community to develop the required capability and capacity to achieve the accounting-for goal. In addition, 12 respondents indicated that the current organizational structure did not enable or only somewhat enabled the community to collectively determine necessary resources. Furthermore, 9 respondents indicated that the current organizational structure did not at all enable the accounting community to define and agree on their respective roles and responsibilities. In contrast, not a single organization we surveyed ranked the current organizational structure as the most effective organizational option, and only three organizations—USD Policy, PACOM, and JPAC—ranked the current organizational structure as the second most effective organizational option. Illustrating a disconnect between leadership’s perspective and the rest of the community, only two organizations in our survey—USD Policy and PACOM, the two top-level leadership organizations in the accounting community—responded that the current structure greatly enables appropriate senior leadership involvement. USD Policy and PACOM stated that all of the organizational options, including the current organizational structure, offer access to DOD senior leadership. In addition, senior officials from these offices questioned whether the benefit of reorganization would result in real change and would be worth undergoing turmoil in the organization. While we recognize that a reorganization may pose challenges, such as creating the potential for short-term impacts on operations due to disruption, our findings in our July 2013 report show that the majority of accounting community members and other stakeholders lack confidence in the status quo, and we believe that the potential benefits of reorganizing and/or clarifying roles and responsibilities could outweigh those challenges. We recommended in our July 2013 report that the Secretary of Defense examine options for reorganizing the accounting community, to include considering organizational options that provide a more centralized chain of command over the accounting community’s mission. DOD concurred with this recommendation, stating that it will consider options for reorganizing the accounting community, ranging from maintaining the status quo to consolidation of DPMO and JPAC, as well as examining whether the Life Sciences Equipment Laboratory might also be included in this consolidation. DOD explained that the consolidated organization could be placed under the Office of the Secretary of Defense or a non- geographic combatant command to facilitate its worldwide mission and avoid competition for resources with a geographic combatant command’s war-fighting priorities. While DOD is working to revise its existing guidance and develop new guidance, the roles and responsibilities of the various members of the missing persons accounting community are not all clearly articulated in existing DOD directives or instructions. We have previously reported on the need for collaborating agencies to work together to define and agree on their roles and responsibilities. DOD has established several directives and instructions related to the missing persons accounting program. However, none of this guidance clearly delineates the specific roles and responsibilities of all the organizations comprising the missing persons accounting community in the four key areas that we examined for our July 2013 report: (1) equipment and artifact identification and analysis, (2) research and analysis, (3) investigations, and (4) family outreach and external communications. Disagreements over roles and responsibilities where the guidance is broad or vague enough to support different interpretations has led to discord, lack of collaboration, and friction among the community’s members, and particularly between DPMO and JPAC. For example, JPAC views itself as having the lead on operational activities, such as conducting investigation and recovery missions, and JPAC officials expressed concerns with DPMO’s plans to conduct some operational activities. Moreover, the lack of clarity in the guidance has given rise to overlapping and fragmented efforts among accounting community members. We have previously reported that overlap in efforts may be appropriate in some instances, especially if agencies can leverage each others’ efforts. In other instances, however, overlap may be unintended, may be unnecessary, or may represent an inefficient use of U.S. government resources. As described in table 1, in implementing the accounting mission, we found that overlapping and duplicative efforts have led to inconsistent practices and inefficiencies in four key areas. Today, I will highlight one of those areas: equipment and artifact identification and analysis. JPAC and the Life Sciences Equipment Laboratory disagree about the laboratory’s roles and responsibilities for equipment and artifact identification and analysis, and DOD guidance is vague regarding those responsibilities. As a result, the interactions between JPAC’s Central Identification Laboratory and the Life Sciences Equipment Laboratory have been inefficient and ineffective and have led to underutilizing government resources, as the following example demonstrates. JPAC and Life Sciences Equipment Laboratory officials disagree about roles and responsibilities in terms of which conflicts and types of equipment the Life Sciences Equipment Laboratory can analyze. JPAC officials told us it is unlikely that they would forward case work to the Life Sciences Equipment Laboratory for conflicts other than Vietnam, and that they do not send ground equipment remnants to the equipment laboratory, regardless of conflict. Conversely, Life Sciences Equipment Laboratory officials stated that their capabilities can support analysis of cases for conflict periods ranging from World War I through current military operations for all military services, and that their mission includes analyzing artifacts recovered at aircraft crash or ground action loss sites. Further, a 2004 memorandum of agreement between JPAC and the Life Sciences Equipment Laboratory states that the Life Sciences Equipment Laboratory has the capability to provide analysis for equipment from World War II, Korea, Vietnam, the Cold War, and current day conflicts. Life Sciences Equipment Laboratory officials expressed concern that JPAC and its Central Identification Laboratory are trying to exclude the Life Sciences Equipment Laboratory from the accounting process by downplaying its potential contributions. This example shows how the lack of clearly defined roles and responsibilities has led to disagreements and inefficient and ineffective interactions among community members. Since 2010, DPMO has attempted to address issues surrounding the accounting community organizations’ roles and responsibilities by developing new guidance or revising existing guidance, but these efforts have not been completed. DPMO has drafted a revision to DOD Directive 2310.07E and has also drafted a new DOD instruction to provide more clarity with regard to roles and responsibilities. As of May 2013, however, neither the draft instruction nor the revised directive had been finalized, because the drafts had been stymied by disagreements among community members regarding their respective roles and responsibilities as stated in the drafts. Both DPMO officials and JPAC officials said they have made progress in addressing these areas of disagreement, and DPMO officials stated that they hoped to have the draft directive finalized by September 2013 and the draft instruction published by March 2014. Because the drafts of these documents are still under revision, it is unclear whether the final guidance will clarify the roles and responsibilities sufficiently to address the four areas of overlap and disagreement summarized in table 1 above. Until DOD issues its revised directive and new instruction that more clearly define the roles and responsibilities of all the accounting community organizations, these areas of inefficient overlap may continue, and the disputing factions within the accounting community may continue to hinder future progress. Consequently, we recommended in our July 2013 report that the department revise and issue guidance to clarify roles and responsibilities of accounting community members and negotiate a new memorandum of agreement between the Life Sciences Equipment Laboratory and JPAC. DOD concurred with both of these recommendations. While DOD has made some progress in drafting a community-wide plan to increase its capability and capacity to meet the statutory accounting-for goal, as of June 2013 DOD had not completed a community-wide plan. We have previously reported that overarching plans can help agencies better align their activities, processes, and resources to collaborate effectively to accomplish a commonly defined outcome. However, our July 2013 report found that community-wide planning to meet the accounting-for goal established by Congress has been impeded by disputes and by a lack of coordination among members of the missing persons accounting community, with DPMO and JPAC developing two competing proposed plans, neither of which encompassed the entire accounting community. In response to a December 2009 memorandum from the Deputy Secretary of Defense directing the Deputy Assistant Secretary of Defense for Prisoner of War/Missing Personnel Affairs to begin planning to meet the accounting-for goal, USD Policy and PACOM allowed the development of these two competing proposed plans for obtaining additional funding and resources to meet the mandated capability and capacity. According to DPMO officials, neither the Joint Staff nor USD Policy provided oversight or intervention in the disagreement. These officials stated that such oversight and intervention could have helped JPAC and DPMO to resolve their impasse by improving communication, interaction, and cooperation. Both plans called for increased capability and capacity and for a new satellite remains identification laboratory located in the continental United States. However, the two plans differed as to which organization would have control over much of the increased capability and capacity, with each plan favoring the organization that authored it. The other accounting community members and their resource needs were not mentioned in either proposed plan. The dispute concerning the competing proposed plans was resolved through DOD’s Program Budget Review Process in January 2011, after being assessed by a DOD-wide team led by DOD’s Office of Cost Assessment and Program Evaluation. In a DOD resource management decision, DOD programmed more than $312 million in proposed additional resources over fiscal years 2012 through 2016 in support of JPAC’s plan, including an additional 253 personnel—reflecting a greater than 60 percent increase over JPAC’s 2011 level. However, key parts of JPAC’s plan are not being realized. For example, JPAC has been unable to conduct the number of investigation and recovery missions called for in the plan, in part due to an inability to hire the additional personnel who had been authorized and also in part due to the budget reductions and expected furloughs associated with sequestration. As of May 2013, the JPAC plan, which does not incorporate the larger accounting community, is DOD’s only plan to increase capability and capacity to account for missing persons. While the community has taken some recent steps to draft a community- wide plan as directed by the 2009 memo from the Deputy Secretary of Defense, we found that disagreements between JPAC and DPMO hindered progress in developing the community-wide plan. According to both DPMO and JPAC officials, the areas of disagreement included topics such as (1) the division of research and analysis responsibilities between DPMO and JPAC; (2) determination of the appropriate levels of effort for each of the various conflicts; and (3) agreement on a policy to address lower priority cases that have been on JPAC’s list of potential recovery sites for a long time. As of June 2013, DPMO and JPAC officials said that the areas of disagreement had been informally resolved and needed to be documented. DPMO had developed a draft of the community-wide plan, but DPMO officials explained that the draft would not be sufficiently comprehensive to share for review among the community members until it incorporated the informal agreements that have recently been resolved. The officials stated that they now plan to finalize the community-wide plan by the end of calendar year 2013. In the absence of a community-wide plan, the members of the accounting community have had varied success in independently identifying and obtaining funds and resources to help meet the accounting-for goal. Moreover, there is no community-wide process to provide resources for the missing persons accounting mission. Each member organization of the accounting community has its own processes for requesting resources, because they belong to diverse parent organizations, and these processes are not integrated or coordinated. Until DOD finalizes a community-wide plan that addresses the resource needs of community members as well as changes in planned operations, the accounting community will be challenged to justify the resources it needs to increase DOD’s capability and capacity to account for at least 200 missing persons a year by 2015, and DOD’s ability to achieve that required increase may be at risk. We recommended in our July 2013 report that the department finalize the community-wide plan to develop the increased capability and capacity required by statute, with the support and participation of all community members. DOD concurred with our recommendation. In total, our full report contains nine recommendations with which DOD generally concurred. The report also contains DOD’s comments, which state the steps the department plans to take to implement our recommendations. In conclusion, while we are encouraged that DOD generally concurred with all nine of the recommendations in our July 2013 report, we note that prompt action on the part of the department to address these recommendations is critical, because the 2015 timeframe for DOD to meet the accounting-for goal is rapidly approaching. Further, as time passes, the information needed for missing persons recoveries continues to deteriorate. Families have been waiting for decades to discover the fate of their loved ones, and the weaknesses that we identified in DOD’s capability and capacity to account for missing persons jeopardize the department’s ability to provide some measure of closure to those families whose loved ones are still missing as a result of their service to their country. Chairman Wilson, Ranking Member Davis, this concludes my prepared remarks. I would be pleased to respond to any questions that you or other Members of the Subcommittee may have. For future questions about this statement, please contact Brenda S. Farrell, Director, Defense Capabilities and Management, at (202) 512- 3604 or farrellb@gao.gov. In addition, contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Individuals who made key contributions to this statement include Margaret Best, Assistant Director; Renee Brown, Terry Richardson, Leigh Ann Sennette, Cheryl Weissman, Allen Westheimer, and Michael Willems. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

This testimony discusses GAO's findings and recommendations about DOD's missing persons accounting mission from our recently issued report, DOD's POW/MIA Mission: Top-Level Leadership Attention Needed to Resolve Longstanding Challenges in Accounting for Missing Persons from Past Conflicts. DOD reports that more than 83,000 persons are missing from past conflicts in Vietnam, Korea, the Cold War, the Persian Gulf, and World War II. Since the early 1970s, DOD has identified the remains of and accounted for approximately 1,910 persons. Several DOD components and organizations, collectively known as the missing persons accounting community, have a role in accounting for missing persons. Between 2002 and 2012, DOD accounted for an average of 72 persons each year. In 2009, Congress established an accounting-for goal in Section 541 of the National Defense Authorization Act for Fiscal Year 2010. This act required the Secretary of Defense to provide such funds, personnel, and resources as the Secretary considers appropriate to increase significantly the capability and capacity of DOD, the Armed Forces, and commanders of the combatant commands to account for missing persons, so that the accounting community has sufficient resources to ensure that at least 200 missing persons are accounted for annually, beginning in fiscal year 2015.The law also added all World War II losses to the list of conflicts for which DOD is responsible, thus increasing from about 10,000 to 83,000 the number of missing persons for whom DOD must account. In 2012, in a committee report to accompany a bill for the National Defense Authorization Act for Fiscal Year 2013, the House Armed Services Committee mandated that GAO review DOD’s efforts to increase its capability and capacity to account for missing persons. GAO will focus on three key issues identified in the report, specifically: (1) the accounting community’s organizational structure, (2) the lack of clarity regarding community members’ roles and responsibilities, and (3) DOD’s planning to meet the statutory accounting-for goal. The department's response to the accounting-for goal established in the National Defense Authorization Act for Fiscal Year 2010 brought into sharp relief longstanding disputes that have not been addressed by top-level leaders, and have been exacerbated by the accounting community's fragmented organizational structure. Leadership from the Under Secretary of Defense for Policy and Pacific Command have been unable to resolve disputes between community members in areas such as roles and responsibilities and developing a community-wide plan to meet the statutory accounting-for goal. Further, the accounting community is fragmented in that the community members belong to diverse parent organizations under several different chains of command. With accounting community organizations reporting under different lines of authority, no single entity has overarching responsibility for community-wide personnel and other resources. While the Department of Defense (DOD) is working to revise its existing guidance and develop new guidance, the roles and responsibilities of the various members of the missing persons accounting community are not all clearly articulated in existing DOD directives or instructions. GAO has previously reported on the need for collaborating agencies to work together to define and agree on their roles and responsibilities. DOD has established several directives and instructions related to the missing persons accounting program. However, none of this guidance clearly delineates the specific roles and responsibilities of all the organizations comprising the missing persons accounting community in the four key areas that GAO examined for the July 2013 report: (1) equipment and artifact identification and analysis, (2) research and analysis, (3) investigations, and (4) family outreach and external communications. Disagreements over roles and responsibilities where the guidance is broad or vague enough to support different interpretations has led to discord, lack of collaboration, and friction among the community's members, and particularly between the Defense Prisoner of War/Missing Personnel Office (DPMO) and Joint Prisoner of War/Missing in Action Accounting Command (JPAC). While DOD has made some progress in drafting a community-wide plan to increase its capability and capacity to meet the statutory accounting-for goal, as of June 2013 DOD had not completed a community-wide plan. GAO has previously reported that overarching plans can help agencies better align their activities, processes, and resources to collaborate effectively to accomplish a commonly defined outcome. However, GAO's July 2013 report found that community-wide planning to meet the accounting-for goal established by Congress has been impeded by disputes and by a lack of coordination among members of the missing persons accounting community, with the DPMO and JPAC developing two competing proposed plans, neither of which encompassed the entire accounting community.

While the term “data center” can be used to describe any room used for the purpose of processing or storing data, as defined by OMB in 2010, a data center was a room greater than 500 square feet, used for processing or storing data, and which met stringent availability requirements. Other facilities were classified as “server rooms,” which were typically less than 500 square feet and “server closets,” which were typically less than 200 square feet. Several factors led OMB to urge agencies to consolidate federal data centers. According to OMB, the federal government had 432 data centers in 1998; more than 1,100 in 2009; and 2,094 in July 2010. Operating such a large number of centers places costly demands on the government. While the total annual federal spending associated with data centers has not yet been determined, OMB has found that operating data centers is a significant cost to the federal government, including hardware, software, real estate, and cooling costs. For example, according to the Environmental Protection Agency (EPA), the electricity cost to operate federal servers and data centers across the government is about $450 million annually. According to the Department of Energy (Energy), data center spaces can consume 100 to 200 times as much electricity as standard office spaces. Reported server utilization rates as low as 5 percent and limited reuse of these data centers within or across agencies lends further credence to the need to restructure federal data center operations to improve efficiency and reduce costs. In 2010, the Federal CIO reported that operating and maintaining such redundant infrastructure investments was costly, inefficient, and unsustainable. Concerned about the size of the federal data center inventory and the potential to improve the efficiency, performance, and environmental footprint of federal data center activities, in February 2010 OMB, under the direction of the Federal CIO, announced FDCCI. This initiative’s four high-level goals are to promote the use of “green IT” by reducing the overall energy and real estate footprint of government data centers; reduce the cost of data center hardware, software, and operations; increase the overall IT security posture of the government; and shift IT investments to more efficient computing platforms and technologies. As part of FDCCI, OMB required 24 departments and agencies that participate on the Chief Information Officers Council (see table 1) to submit a series of documents that ultimately resulted in a data center consolidation plan. In addition to an initial data center inventory and preliminary consolidation plan, the departments and agencies were to provide the following: An asset inventory baseline, which was to contain more detailed information and serve as the foundation for developing the final data center consolidation plans. The final inventory was also to identify the consolidation approach to be taken for each data center. A data center consolidation plan, which was to be incorporated into the agency’s fiscal year 2012 budget and was to include a technical roadmap and approach for achieving the targets for infrastructure utilization, energy efficiency, and cost efficiency. In October 2010, OMB reported that all of the agencies had submitted their plans. OMB also announced plans to monitor agencies’ consolidation activities on an ongoing basis as part of the annual budget process. Further, starting in fiscal year 2011, agencies were required to provide an annual updated data center asset inventory at the end of every third quarter and an updated consolidation plan (including any missing elements) at the end of every fourth quarter. Agencies were further required to provide a consolidation progress report at the end of every quarter. To manage the initiative, OMB designated two agency CIOs as executive sponsors to lead the effort within the Chief Information Officers Council. Additionally, the General Services Administration (GSA) has established the FDCCI Program Management Office, whose role is to support OMB in the planning, execution, management, and communication for FDCCI. In this role, GSA collected the responses to OMB-mandated document deliveries and reviewed the submissions for completeness and reasonableness. GSA also sponsored three workshops on the initiative for agencies and facilitated a peer review of the initial and final data center consolidation plans. “…a data center is…a closet, room, floor or building for the storage, management, and dissemination of data and information and computer systems and associated components, such as database, application, and storage systems and data stores [excluding facilities exclusively devoted to communications and network equipment (e.g., telephone exchanges and telecommunications rooms)]. A data center generally includes redundant or backup power supplies, redundant data communications connections, environmental controls…and special security devices housed in leased,…owned, collocated, or stand-alone facilities.” OMB, Implementation Guidance for the Federal Data Center Consolidation Initiative (Washington, D.C.: Mar. 19, 2012). In December 2010, OMB published its 25-Point Implementation Plan to Reform Federal Information Technology Management as a means of implementing IT reform in the areas of operational efficiency and large scale IT program management. Among the 25 initiatives, OMB has included two goals that relate to data center consolidation: 1. By June 2011, complete detailed implementation plans to consolidate at least 800 data centers by 2015. 2. By June 2012, create a governmentwide marketplace for data center availability. To accomplish its first goal, OMB required each FDCCI agency to identify a senior, dedicated data center consolidation program manager. It also launched a Data Center Consolidation Task Force comprised of the data center consolidation program managers from each agency. OMB officials stated that this task force is critical to driving forward on individual agency consolidation goals and to meeting overall federal consolidation targets. OMB has also created a publicly available dashboard for observing agencies’ consolidation progress. To accomplish its second goal, OMB and GSA launched a governmentwide data center availability marketplace in June 2012. This online marketplace is intended to match agencies that have extra capacity with agencies with increasing demand, thereby improving the utilization of existing facilities. The marketplace will help agencies with available capacity promote their available data center space. Once agencies have a clear sense of the existing capacity landscape, they can make more informed consolidation decisions. We have previously reported on OMB’s efforts to consolidate federal data centers. In March 2011, we reported on the status of the FDCCI and noted that data center consolidation makes sense economically and is a way to achieve more efficient IT operations, but that challenges exist. For example, agencies reported facing challenges in ensuring the accuracy of their inventories and plans, providing upfront funding for the consolidation effort before any cost savings accrue, integrating consolidation plans into agency budget submissions (as required by OMB), establishing and implementing shared standards (for storage, systems, security, etc.), overcoming cultural resistance to such major organizational changes, and maintaining current operations during the transition to consolidated operations. We further reported that mitigating these and other challenges will require commitment from the agencies and continued oversight by OMB and the Federal CIO. In July 2011, we reported that agency consolidation plans indicate that agencies anticipated closing about 650 data centers by fiscal year 2015 and saving about $700 million in doing so. However, we also found that only one of the 24 agencies submitted a complete inventory and no agency submitted complete plans. Further, OMB did not require agencies to document the steps they took, if any, to verify the inventory data. We noted the importance of having assurance as to the accuracy of collected data and specifically, the need for agencies to provide OMB with complete and accurate data and the possible negative impact of that data being missing or incomplete. We concluded that until these inventories and plans are complete, agencies may not be able to implement their consolidation activities and realize expected cost savings. Moreover, without an understanding of the validity of agencies’ consolidation data, OMB could not be assured that agencies are providing a sound baseline for estimating consolidation savings and measuring progress against those goals. Accordingly, we made several recommendations to OMB, including that the Federal CIO require that agencies, when updating their data center inventories, state what actions have been taken to verify the inventories and to identify any associated limitations on the data. We also recommended that the Federal CIO require that agencies complete the missing elements in their consolidation plans and in doing so, consider consolidation challenges and lessons learned. We also made recommendations to the heads of agencies to complete the information missing from their inventories and plans. In response to our recommendations, OMB took several actions. Beginning in fiscal year 2011, in addition to the updated inventories due at the end of every third fiscal quarter, agencies are required to submit an updated consolidation plan by the end of every fourth fiscal quarter. Along with the updated plan, agencies are required to submit a signed letter from their CIOs, attesting to the completeness of the plan, stating what actions were taken to verify the inventory, and noting any limitations of inventory or plan data. The inclusion of this performance information will continue to be important to OMB as it makes decisions on how best to oversee the ongoing federal data center consolidations. By gathering this understanding of the validity and limitation on agencies’ data, OMB will be better assured that agencies are providing a sound baseline for estimating savings and accurately reporting progress against their goals. The extent to which agencies have completed information missing from their inventories and plans is discussed in the following section. More recently, in February 2012, we updated our March 2011 work and reported that although OMB had taken steps to ensure the completion of agencies’ consolidation plans, a preliminary analysis indicated that not all plans were complete. Also, in April 2012, we reported on the progress OMB and federal agencies made in implementing the IT Reform Plan, including one action item associated with data center consolidation. We reported that this goal was only partially completed, based on our conclusion that not all of the agencies’ updated data center consolidation plans included the required elements. As discussed earlier, OMB required agencies to submit an updated data center inventory that included information on each center and its assets by the end of June 2011, and an updated consolidation plan that included key information on the agencies’ consolidation approach by the end of September 2011. OMB subsequently issued revised guidance on the mandatory content of the data center inventories and consolidation plans, in May 2011 and July 2011, respectively. While the revised inventory guidance asked for different information from what was requested in 2010, it still required agencies to report on specific assets within individual data centers, as well as information about each specific data center. The revised guidance on consolidation plans was similar to the 2010 guidance, but included several additional requirements. Specifically, in addition to continuing to require information on key elements such as goals, approaches, schedules, cost-benefit calculations, and risk management plans, the revised guidance also required agencies to address the data verification steps, consolidation progress, and cost savings. Table 2 compares the original and revised requirements for key elements to be included in agency inventories and plans. While all agencies submitted updated inventories and plans in 2011, most of the agencies’ documents are still not complete. As required, all 24 agencies submitted their inventories in June 2011 and all but 2 submitted their updated consolidation plans in September 2011. The Social Security Administration (SSA) submitted its updated consolidation plan in October 2011 and the Department of Defense (Defense) submitted an updated consolidation plan in November 2011. However, of the 24 agencies’ submissions, only 3 of the inventories are complete and only 1 of the plans is complete. For example, while all 24 agencies report on their inventories to some extent, 8 agencies provide only partial information on the new category of physical servers and 17 provide only partial information on the new category of IT facilities and energy usage. Additionally, in their consolidation plans, 13 agencies do not provide a full master program schedule, 17 agencies do not provide full cost-benefit analysis results, and 21 agencies do not include all required cost savings information. In the absence of important information such as schedules and cost estimates, agencies are at risk of not realizing key FDCCI goals such as anticipated cost savings and improved infrastructure utilization. While agencies’ inventories and goals have changed since we last reported on FDCCI, agencies continue to report plans to significantly reduce the number of their centers and to achieve cost savings. Last year, we reported that as of April 2011, 23 agencies identified 1,590 centers (using the large data center definition) and established goals to reduce that number by 652. Our most recent analysis of 24 agencies’ documentation indicates that as of September 2011, agencies identified almost 2,900 total centers, and established plans to close over 1,185 of them by 2015. The new total number of data centers includes 648 large centers (500 square feet or more), 1,283 smaller centers (less than 500 square feet), and 966 centers of undetermined size.undetermined size are primarily comprised of 936 Defense facilities, a list of which was provided in a format that did not allow for an analysis of the The centers of size of the centers. An OMB official attributed the change in the number of large centers reported to agencies’ improvements in data quality. Table 3 contains a further breakdown of actual and planned closures by calendar year, for both large and smaller centers. The number of facilities in agencies’ inventories has changed over time, and will likely continue to evolve. For example, in July 2011, we reported that agencies reported having 1,590 large centers in their inventories, whereas they now report only 648. There are multiple reasons for these fluctuations. Some agencies have reported confusion over the evolving definition of “data center,” while officials from other agencies told us that some facilities have been reclassified or dropped from the inventory as more was learned about the facilities. Additionally, agencies have reported that their inventory totals are in a constant state of flux and changing on a regular basis as a result of their efforts to gather and refine information about data center inventories. Most agencies also continued to report expected savings from FDCCI. Specifically, Nineteen agencies reported anticipating more than $2.4 billion in cost savings and more than $820 million in cost avoidances, between 2011 and 2015. Additionally, as we also reported in 2011, actual savings may be even higher because 14 of these agencies’ projections were incomplete. One agency does not expect to accrue net savings until 2017. One agency does not expect to attain net savings from its consolidation efforts. Three agencies did not provide estimated cost savings. While we recognize that agencies’ planned savings of over $2.4 billion may grow as agencies complete their cost and savings assessments, the President’s budget for fiscal year 2013 states that FDCCI is expected to realize $3 billion in savings by 2015. This reflects a $600 million dollar disparity between what agencies are reporting and what OMB is expecting. Such a disparity highlights the need for agencies to continue to develop and refine their savings projections, in order to make clear an accurate picture of the goals to be realized by the governmentwide consolidation initiative. In our July 2011 report, we recommended that agencies complete the missing elements from their inventories. Further, as part of FDCCI, OMB required agencies to update their data center inventories at the end of the third quarter of every fiscal year. In guidance provided to the agencies, the 2011 updated inventories were to address five key elements for each data center: (1) physical servers, (2) virtualization, (3) IT facilities and energy, (4) network storage, and (5) data center information. One information category from 2010, IT software assets, was no longer required. Table 4 provides a detailed description of each of the five key elements. However, not all of the agencies used the revised format. Specifically, 21 of the 24 agencies submitted inventories in OMB’s updated format and 3 agencies (the Department of Agriculture (Agriculture), the Office of Personnel Management (OPM), and the Small Business Administration (SBA)) used the former format. Officials from all 3 agencies stated that they thought they were using the correct format at the time. Further, these officials said they plan to submit information consistent with OMB’s revised inventory template in the future. The confusion by selected agencies on which templates to use is due, in part, to a change in how OMB distributed its new guidance. While in prior years the Federal CIO wrote letters to agency CIOs and OMB posted its guidance on the FDCCI website, in conveying the direction to use a new template in spring 2011, the Federal CIO did not write letters to agency CIOs and OMB did not post its latest guidance online. Instead, the Federal CIO and OMB relied on more informal means, such as the FDCCI task force meetings, to disseminate the new guidance. Although the task force serves as an important communications conduit for FDCCI, the confusion we identified among agencies on which template to use demonstrates that the task force was not effective as the sole means of communication with the agencies. In providing guidance and direction, task force communications could be enhanced by leveraging other existing resources, such as sending letters from the Federal CIO to agency CIOs and posting the guidance on the initiative’s website. In assessing agencies’ inventories, we rated an element as complete if the agency provided all of the information required for the element, partial if the agency provided some, but not all, of the information for the element, and incomplete if the agency did not provide the information required for the element. A partial rating could result if an agency did not provide any information for selected facilities or if the agency did not fill in selected fields for its facilities. For example, both an agency providing data on two of five facilities and an agency providing incomplete data on energy usage across facilities would receive partial ratings. Of the 21 inventories in the new format, only 3 contain complete data for all five of the required elements. Additionally, while all agencies provide at least partial inventory data for all five elements, one agency provides complete information for four of the five eight agencies provide complete information for three of the five three agencies provide complete information for two of the five elements, two agencies provide complete information for one of the five elements, and four agencies do not have any complete elements in their inventories. Figure 1 provides an assessment of the completeness of agencies’ inventories, by key element, and a discussion of the analysis of each element follows the figure. In addition, a detailed summary of each agency’s completion of key elements is provided in appendix II. Physical servers. Thirteen agencies provide complete information on their physical servers and 8 agencies provide partial information. For example, the Department of Education (Education) provides complete information on its total rack count and counts of types of servers, while the Department of Health and Human Services (HHS) provides complete counts of individual servers, but partial information on total rack count. Additionally, the Department of Justice (Justice) provides partial information for both its total rack count and types of servers. Virtualization. Seventeen agencies provide complete information on their virtualization and 4 agencies provide partial information. For example, HUD, the Departments of State (State) and Veterans Affairs (VA), and the National Science Foundation (NSF), all provide complete information on their virtual host count and virtual operating system count. In contrast, the Departments of Defense, Homeland Security (DHS), Justice, and GSA provide partial information for both of those same elements. IT facilities and energy. Four agencies provide complete information on their IT facilities and energy, while 17 provide partial information. For example, the Nuclear Regulatory Commission (NRC) and SSA fully provide such information as total data center power capacity and average data center electricity usage. However, VA fully reports on total data center power capacity, but partially on average data center electricity usage and total IT data center power capacity. Further, the Department of Labor (Labor) partially reports on total data center IT power capacity and average data center electricity usage and does not report any information on total data center power capacity. Network storage. Fourteen agencies provide complete information on their network storage and 7 provide partial information. For example, the Departments of Commerce (Commerce) and Transportation (Transportation), EPA, the National Aeronautics and Space Administration (NASA), and the U.S. Agency for International Development (USAID) all fully report on their total and used network storage. Other agencies, such as Defense, HHS, and State partially report information in each of those two categories. Data center information. Three agencies provide complete information on their individual data centers, while 18 provide partial information. For example, HUD and SSA both fully report on data center-specific information such as data center type, gross floor area, and target date for closure. Other agencies, such as Energy and VA fully report on gross floor area and closure information, but partially report data center costs. Also, agencies such as Defense and DHS report partial information in all categories. Part of the reason the agencies’ inventories remain incomplete stems from challenges in gathering data center power information, a key component of the IT facilities and energy component, and more broadly, problems providing good quality asset inventories, as OMB requires. These challenges are discussed in more detail later in this report. Because the continued progress of FDCCI is largely dependent on accomplishing goals built on the information provided by agency inventories, it will be important for agencies to continue to work on completing their inventories, thus providing a sound basis for their savings and utilization forecasts. In addition to the agencies’ inventories, we previously recommended and OMB required agencies to update their consolidation plans to address any missing elements. OMB’s revised guidance on the contents of the consolidation plans retains key elements from its prior guidance and adds requirements to discuss steps taken to verify inventory and plan data, consolidation progress, and consolidation cost savings. OMB has previously reported on the importance of agencies’ consolidation plans in providing a technical road map and approach for achieving specified targets for infrastructure utilization, energy efficiency, and cost efficiency. Table 5 provides a detailed description of each of these elements. All 24 agencies submitted consolidation plans to OMB, but only 1 agency has a complete plan. For the remaining 23 agencies, selected elements are missing from each plan. For example, among the 24 agencies, all provide complete information on their qualitative impacts, but only 9 provide complete information on their quantitative goals. Further, 23 agencies specify their consolidation approach, but only 5 indicate that a full cost-benefit analysis was performed for the consolidation initiative. In many cases, agencies submitted some, but not all, of the required information. Figure 2 provides an assessment of the completeness of agencies’ consolidation plans, by key element, and a discussion of each element follows the figure. In addition, a detailed summary of each agency’s completion of key elements is provided in appendix II. Quantitative goals. Nine agencies provide complete savings and utilization forecasts, 13 agencies provide partial forecasts, 1 agency does not provide any information, and an official from 1 agency said that this element did not apply. For example, Agriculture and Labor were rated as providing partial forecasts because they provide complete savings forecasts, but incomplete utilization forecasts. State and NRC were rated as providing partial forecasts because they both provide incomplete savings and utilization forecasts. Some agencies identified reasons for not having completed these forecasts. Specifically, a Department of the Interior (Interior) official told us that it was not cost effective to gather the missing information, so it was not included. Officials from other agencies, such as Labor and NRC, told us of data quality problems or that their data centers lacked the ability to gather the required information. Further, a HUD official stated that the department did not have any quantitative goals because their consolidation effort was completed in 2005. Qualitative impacts. All 24 agencies fully describe the qualitative impacts of their consolidation initiatives. For example, Commerce’s plan describes goals such as controlling data center costs and shifting IT investments to more efficient computing platforms and technologies. Additionally, NASA reports that the consolidation effort will provide access to cost and power-efficient data centers that will meet all of the agency’s computing needs, as well as transform the data center environment, in part through virtualization and the use of cloud services. Further, SBA describes goals such as reducing the amount of physical resources consumed by technology systems and modernizing and updating agency systems. Summary of consolidation approach. Twenty-three agencies include a summary of the agencies’ consolidation approaches and an official from 1 agency said that this element did not apply. For example, Defense describes the department’s reference architecture for use in guiding the consolidation effort and also provides examples of how the Air Force and the Army are approaching aspects of their respective consolidations. Additionally, State’s plan details how the department will consolidate all domestic data centers into four enterprise data centers. Additionally, a HUD official stated that this element was not applicable because the department’s consolidation effort was completed in 2005. Scope of consolidation. Twenty-two agencies’ plans include a well- defined scope for data center consolidation, 1 provides partial information on the scope of their consolidation efforts, and 1 does not provide this information. Specifically, the agencies that provide this information list the data centers included in the consolidation effort and what consolidation approach will be taken for each center. For example, EPA lists the 25 facilities for which either the servers will be moved or the site will be decommissioned. Similarly, Justice lists the 36 centers that will be either consolidated or decommissioned. However, Labor only partially addresses consolidation scope because it only provides information on about half of its data centers. According to an agency official, the centers that have been addressed constitute the bulk of the agency’s computing power, but that the remaining facilities will be addressed in a later phase of the consolidation effort, the timing for which has not yet been determined. Additionally, Defense has not defined its consolidation scope. A Defense consolidation program official stated that the department was still working to better understand the full inventory for all departmental components. High-level timeline. Twenty-two agencies include a high-level timeline for consolidation efforts, 1 agency includes partial information on its timeline, and 1 does not provide a timeline. For example, Justice and EPA both provide the year for which action will be taken on their centers to be consolidated and NRC lists the years its three centers will be consolidated before they are replaced by NRC’s new data center. In contrast, Labor provides a timeline for about half of its data centers, and Defense does not provide a timeline because it has not fully defined the scope of its consolidation effort. Performance metrics. Eighteen agencies identify specific performance metrics for their consolidation programs, 1 agency provides partial information on its metrics, 4 agencies do not identify specific metrics, and an official from 1 agency said that this element did not apply. Specifically, Agriculture’s plan defines several key performance indicators such as the numbers of applications moved and physical servers eliminated. Additionally, several agencies, such as Commerce, Defense, and NSF, provide consolidation performance metrics based on quantitative savings and utilization goals. As an example of an agency with partial metrics, Education identifies metrics based on its savings goals, but is missing information on its progress in meeting utilization goals. Additionally, DHS and NRC do not identify any performance metrics. Officials from both DHS and NRC agreed that their agencies did not have such measures when their plans were published, but noted that the required metrics had since been developed or that they now have the resources to develop them. Further, a HUD official stated that this element was not applicable because the department’s consolidation effort was completed in 2005. Master program schedule. Nine agencies reference a completed master program schedule, 13 agencies do not reference such a schedule, and officials from 2 agencies said that this element did not apply. For example, HHS, VA, and GSA discuss their master program schedules, but other agencies, such as State and EPA do not reference schedules in their plans. State officials noted that the department has a schedule, but that it was not included in their consolidation plan due to a miscommunication. They stated that it would be included in their next plan update. Some agencies, such as Defense and Labor, are working to develop their schedules or will develop them in the future. A Defense official told us that the department has drafted a combined data center consolidation and cloud computing master schedule that is expected to be approved by the end of September 2012. Officials from Energy told us that their consolidation schedule existed, but that it was part of a larger departmental effort and did not provide detail down to the individual data center level. Officials from OPM questioned the utility of a master program schedule for relatively limited consolidation efforts. Two agencies reported that this requirement was not applicable to their situation. Specifically, officials from Education stated that this requirement was not applicable because of the small scale of their agency’s consolidation efforts. Additionally, a HUD official stated that this element was not applicable because the department’s consolidation effort was completed in 2005. Cost-benefit analysis. Five agencies provide results from a complete cost-benefit analysis that encompasses their entire consolidation initiative, 10 agencies provide only selected elements of a cost-benefit analysis, and 7 agencies do not provide a cost-benefit analysis. This element did not apply to 2 agencies. For example, Commerce details full annualized cost and savings estimates through fiscal year 2015, while other agencies, such as HHS and Interior provide only partial information. Specifically, HHS addresses projected savings, but not costs, and Interior acknowledges that an analysis has not yet been completed. Some agencies, such as Defense and Energy, plan to complete a cost-benefit analysis in the future. Officials from Transportation told us that the department was working on a new cost-benefit analysis, as the department no longer felt comfortable with their original savings projections. An Education official noted that the department’s consolidation did not cost anything and that although data will be moved out of the department’s one server room to be consolidated by the end of 2012, the facility would still operate as a network center. Additionally, a HUD official stated that this element was not applicable because the department’s consolidation effort was completed in 2005. Risk management plan. Eighteen agencies reference a consolidation risk management plan and require that risks be tracked, 4 agencies partially address risk management, and 2 agencies do not address risk management. For example, DHS describes its Data Center Services Project Risk Management Plan, including how risks are identified, assessed, and mitigated throughout the development life cycle. Additionally, Transportation addresses how its risk management plan identifies and tracks risks in three categories: people, process, and technology and administration. In contrast, agencies such as Energy and Interior are rated as partial because they are continuing to develop their risk management processes. An Interior official told us that the department’s plan is scheduled to be completed by June 2012. Officials from both OPM and SBA acknowledged that their consolidation plans did not address a risk management plan, but noted that risk was either being managed as part of individual projects or within a larger context within their respective organizations. Communications plan. Twenty-two agencies consider a communications plan for the agencies’ consolidation initiatives, 1 agency does so partially, and 1 agency does not. For example, HHS describes a series of organizational responsibilities for gathering and reporting project information, as well as communicating with other departmental stakeholders. Additionally, GSA describes how its communications approach ensures that stakeholders both within and outside of the agency are kept informed as to consolidation progress. The Department of the Treasury (Treasury) partially addresses its communications plan, noting that it is maintained as part of a larger departmental effort. OPM makes no such reference. Further, an official from OPM told us that a communications plan was not as critical for a small agency. Inventory and plan verification. Fifteen agencies fully describe the steps taken to ensure that inventories and plans were complete and accurate, and 9 agencies partially do so. For example, State describes how information was gathered and validated, addresses several limitations, and attests to the documents’ completeness. Additionally, EPA describes how information was validated, describes limitations on inventory data, and attests to the currency of the agency’s plans. However, other agencies, such as Agriculture, HUD, and SSA are rated as having partially completed this element because they note that information was validated, but do not address data limitations or the completeness of both the inventory and plan. A HUD official told us that the department was unaware of this requirement and agreed to consider what could be said in the next plan update. An SSA official acknowledged that this information was meant to be included, but was inadvertently omitted. Consolidation progress. Eleven agencies fully report on progress meeting consolidation goals, 11 agencies do so partially, and this element does not apply to 2 agencies. Specifically, Justice addresses progress against consolidation goals, discusses consolidation challenges, and references consolidation successes, such as integrating lessons learned from other organizations. VA similarly describes progress against goals and challenges, and also notes the department’s reliance on commercial and public best practices while updating its consolidation plan. However, both Education and NASA are rated as partially completing this element because they discuss progress against goals, but do not present specific successes or challenges. A NASA official agreed that this information was not included, but stated that the agency was aware of situations that addressed both categories of information. Additionally, a HUD official stated that this element was not applicable because the department’s consolidation effort was completed in 2005. OPM officials stated that the agency followed OMB’s original guidance when completing their updated consolidation plan, which did not include a requirement for reporting on consolidation progress. Cost savings. Only 1 agency fully reports on consolidation cost savings, while 13 agencies do so partially, and 8 do not. This element does not apply to two other agencies. Specifically, Commerce discusses net savings, future savings, budgetary impacts, and that the consolidation effort did not incur any unexpected costs. In contrast, HHS and Justice address net and future savings, but not budgetary impacts or unexpected costs. Additionally, other agencies do not include this information for various reasons. Notably, a Defense official told us that it was challenging to gather savings information from all the department’s components. An NSF official told us the information was not included because the agency had not yet realized any cost savings and so, had nothing to report. However, the agency expected to have more to report in the future. Additionally, a HUD official stated that this element was not applicable because the department’s consolidation effort was completed in 2005. Further, as with reporting on consolidation progress, OPM officials stated that they followed OMB’s original guidance, which did not include a requirement relating to cost savings. In the continued absence of completed consolidation plans, agencies are at risk of implementing their respective initiatives without a clear understanding of their current state and proposed end state. For example, OMB intends for agencies’ master program schedules to provide an agencywide plan drawn from detailed implementation schedules for each data center. However, only nine agencies have fully completed this activity. Further, OMB intends agencies’ cost-benefit analyses to assess planned investments and cost savings calculations on a year-by-year basis, thus capturing realistic estimates of funding needed or savings realized from the closing of facilities and associated reduction in energy use. Nonetheless, only five agencies have completed such a study. Without completing this information, agencies may not realize anticipated cost savings, improved infrastructure utilization, or energy efficiency. The importance of these two practices is further discussed in the following section. OMB requires both a master program schedule and a cost-benefit analysis as key elements of agencies’ consolidation plans, but none of the agencies we evaluated had complete schedules or cost estimates. A comprehensive schedule is an important foundational element for initiative planning and provides a road map for systematic project execution. A credible cost-benefit analysis, which is one type of cost estimate, is a key tool for management to use in making informed decisions and includes information such as relative benefits and the effect and value of cost trade-offs. However, of five agencies (Agriculture, DHS, Interior, Transportation, and VA) selected for further analysis, none had a schedule or cost estimate that was fully consistent with best practices. Of the five agencies, two did not have schedules at all and one agency had previously completed a cost estimate but no longer had confidence in those calculations and therefore, planned to do a new cost-benefit analysis. OMB is sponsoring the development of a standardized cost model that could help agencies provide future estimates based on a common set of assumptions, estimates, and calculations. The success of a program depends in part on having an integrated and reliable master schedule that defines when and how long work will occur and how one activity is related to another. A program schedule provides not only a road map for systematic project execution but also the means by which to gauge progress, identify and resolve potential problems, and promote accountability at all levels of the program. A schedule also provides a time sequence for the duration of a program’s activities and furthers an understanding of both the dates for major milestones and the activities that drive the schedule. Our research has identified four select attributes of properly sequenced schedule activities that are essential for a reliable schedule network. Table 6 provides a detailed description of these attributes. Of the five agencies selected, three agencies (Agriculture, VA, and DHS) provided their consolidation master program schedules and two agencies (Interior and Transportation) did not provide a master program schedule that we could evaluate. Of the three agencies that provided schedules, Agriculture and VA provided a single master schedule and DHS provided 4 schedules representing different aspects of the department’s future consolidation plans. However, none of these agencies’ schedules is fully compliant with the four attributes, although each agency was at least partially consistent with these practices. Table 7 provides an assessment of the agencies’ consistency with the four attributes of properly sequenced schedule activities. A discussion of the analysis of each characteristic follows the table. Identified dependencies. None of the three agencies’ schedules is fully consistent with this practice. Specifically, two of DHS’s schedules have activities missing predecessors, successors, or both. Additionally, almost half of Agriculture’s activities, and almost 40 percent of VA’s, have a similar condition. No dangling activities. Two of the three agencies are consistent with this practice and one agency is partially consistent. For example, neither schedule for Agriculture or VA has any dangling activities. In contrast, two of DHS’s four schedules do not have dangling activities, while the remaining two do have such activities. No start-to-finish links. Two of the three agencies’ schedules are consistent with this practice and one agency’s schedule is partially consistent. Both of Agriculture and VA’s schedules are consistent with this practice and have no start-to-finish links. However, while three of DHS’s schedules do not have start-to-finish links, one schedule does. No summary links. One of the three agencies was consistent with this practice, one agency was partially consistent, and one agency was not consistent. Specifically, Agriculture’s schedule does not have any summary links, while only one of DHS’s schedules meets the same condition. Three of DHS’s schedules include summary links, as does VA’s schedule. Department officials gave a variety of reasons why their respective department did not provide documentation of a completed master program schedule: An Agriculture official told us that the department had a detailed schedule for every individual closure, but that because those projects are not necessarily linked to one another, there was no need to link these activities in a master schedule. However, leading practices demonstrate that even a summary master schedule should be a roll- up of lower-level schedules and reflect milestones that are automatically calculated through the established network logic between planned activities. A schedule with proper logic can predict impacts on the project’s planned finish date of, among other things, misallocated resources, delayed activities, external events, scope changes, and unrealistic deadlines. DHS’s consolidation program manager stated that the department provided separate schedules because schedules are developed for individual facilities when placed under contract for closure. However, leading practices show that a program schedule should include the entire required scope of effort, including the effort necessary from all government, contractors, and other key parties for a program’s successful execution from start to finish. The DHS consolidation program manager acknowledged that the schedules in question were developed by contractors and that the department plans to incorporate the suggested best practices as appropriate. A VA official told us that because some of the tasks in the department’s schedule are expected to start on particular dates to ensure funding is available for the project task, they do not have predecessor tasks. While this can be a permissible step when the schedule constraints are clearly identified, the VA official was able to provide some, but not all, of those constraints. The VA official further told us that unnecessary tasks and constraints have since been removed from the department’s schedule. Interior officials stated that the department’s master program schedule was not yet complete. Transportation’s consolidation program manager stated that the department does not have a master program schedule dedicated to the FDCCI. Rather, the consolidation effort appears as a task on the department’s master IT projects schedule. In the absence of program schedules constructed in accordance with scheduling best practices, the agencies we evaluated are at risk of moving forward with their consolidation efforts despite having incomplete information that defines when and how long work will occur and how activities are related to each other. We have reported that the ability to generate a reliable cost estimate, such as a cost-benefit analysis, is a critical function necessary to support OMB’s capital programming process. Such estimates should also include information on the benefits of the project. Without such estimates, agencies are at risk of experiencing cost overruns, missed deadlines, and performance shortfalls. Our research has identified a number of best practices that are the basis of effective program cost estimating and should result in reliable and valid cost estimates that management can use for making informed decisions. Table 8 provides a detailed description of the four characteristics of a high-quality and reliable cost estimate. Of the five agencies selected, four (Agriculture, DHS, Interior, and VA) provided supporting documentation used to calculate the cost estimates found in the agencies’ consolidation plans and one (Transportation) did not. Transportation officials explained that they were no longer confident in their prior estimates and they planned to undertake a new cost-benefit analysis in 2012. None of the four agencies’ estimates was fully compliant with best practices, although all of the estimates were at least minimally consistent with these practices. Table 9 provides an assessment of the estimates’ consistency with the characteristics of a reliable cost estimate. A discussion of the analysis of each characteristic follows the table. Comprehensive. None of the estimates are fully consistent with this practice, although all four estimates satisfy about half of the criterion for this practice. For example, Agriculture includes most related costs and estimate assumptions, but does not include a work breakdown structure. Similarly, Interior includes most related costs and estimate assumptions, but also does not include a work breakdown structure. Well-documented. None of the estimates fully satisfy this practice. Specifically, one estimate satisfies most, but not all, of the practice, one estimate satisfies about half of the criterion for the practice, and two estimates satisfy a few of the criterion for the practice. For example, Interior documents its technical baseline but does not fully document how the estimate was developed. VA describes how its calculations were performed and discusses the estimate’s technical baseline, but satisfied only half of the criteria describing how the estimate was performed. Accurate. None of the estimates fully satisfied this practice. One estimate satisfied about half of the practice and three estimates satisfied some of the practice. For example, Agriculture’s estimate is partially based on historical estimates, but has not been updated since March 2009. Additionally, while DHS updated its estimate in July 2011, it did not adjust for inflation or document variations between planned and actual costs. Credible. None of the estimates fully satisfied this practice. One estimate satisfied about half the practice and three estimates satisfied some of the practice. For example, DHS addressed some aspects of a risk and uncertainty analysis, but did not conduct an estimate sensitivity analysis. Conversely, VA addressed some aspects of an estimate sensitivity analysis, but did not conduct a risk and uncertainty analysis. Neither agency conducted an independent cost estimate. Agency officials gave a variety of reasons for why their cost estimates were not complete. For example, the Agriculture CIO indicated that the department’s estimate was performed several years ago by a contractor and additional documentation was difficult to acquire. Additionally, Interior’s consolidation program manager stated that the department was in the process of revising its cost estimate using OMB’s cost model, but the effort was not yet complete. Further, Interior’s consolidation plan describes several efforts to estimate costs that the department ultimately did not include in their plan and indicates that the department will address this in a future deliverable. In May 2012, Interior officials stated that they recently provided this information to OMB. VA officials stated that they did not provide previous cost estimate documentation because the department expected to revise its cost estimate using new information regarding cost assumptions and that this information would affect life- cycle cost estimates. The DHS consolidation program manager noted that the department is now taking a different approach towards cost estimates through the use of enterprisewide contracts. Regarding Transportation, although it reported FDCCI-related estimated savings of over $26 million in its 2010 plan, the department’s updated consolidation plan states that the original cost estimates were no longer relevant and the department is in the process of conducting a new estimating effort that was not completed in time for the plan’s submission. Further, in March 2012, a department official confirmed that Transportation no longer felt comfortable with the original savings estimate and that planned cost savings were being reevaluated. The official further stated the department intends to complete a new cost-benefit analysis. Between the five agencies that we reviewed, there are plans to consolidate 375 data centers of all sizes. In the absence of reliable cost estimates, these five agencies are exposed to the types of risks that we have reported to be recurring problems in our program reviews—namely cost overruns, missed deadlines, and performance shortfalls. Because of the importance of a high-quality cost estimate to consolidation efforts as significant as these, it will be important for these agencies to work to improve their cost estimates, thus providing information on which management can make well-informed decisions as the agencies move towards their 2015 targets. To assist agencies in their data center consolidation efforts, the FDCCI Data Center Consolidation Task Force developed a standard Total Cost of Ownership (TCO) model in order to provide a comprehensive tool to help to inform consolidation decision making, model consolidation paths, and assist with the development of cost savings figures and funding needs. OMB provided the model to agencies for voluntary use starting in January 2012, noting that it is intended to provide a uniform and consistent method to derive agency cost savings figures and a modeling and simulation tool to inform consolidation decisions. At a high level, agencies load their raw agency inventory data into the spreadsheet-based model to develop three outputs: an “as is” view of current costs; a 5-year projection of costs based on maintaining current equipment and facilities at current growth rates; and a 5-year projection of costs, including equipment and facilities counts, based on the agency’s planned data center closure and efficiency targets. The model relies on a number of built-in assumptions—based on best practices in the public and private sectors and grouped into categories such as facilities, hardware, and software—to provide its outputs. The model also recognizes some limitations, such as an inability to capture costing data for individual facilities and an inability to recognize individual costs for hardware and software. To compensate, the model applies universal values for such information, while recognizing the inaccuracies this may cause in some costing elements. The model further allows agencies to adjust specific variables to input costs that are atypical or not already anticipated by the model. According to an official from the GSA program management office that maintains the cost model, while not intended to capture comprehensive program costs, the model does provide agencies with the ability to customize the input information to make it as comprehensive as they need it to be. As a result, agencies could use this tool to provide more consistent and reliable cost estimates. Moreover, the model provides standardized cost calculations, adjustment for inflation, and a scenario-analysis tool that agencies can use to analyze alternatives and develop plans. Thus, it can be used as a tool to help agencies improve their consolidation planning. Officials from several agencies told us that they plan to use the TCO model in future cost estimating efforts. For example, a Transportation official told us that the department intends to use the model as it recalculates its cost-benefit estimate. Additionally, the Interior consolidation program manager stated that the department planned to use the model to determine power estimates. Officials from other agencies, such as SSA and EPA, told us that the model was being considered for future use. Use of the TCO model could provide more consistent and reliable cost estimates, but using the model is currently voluntary. In light of the limitations identified above in our review of the five agencies’ cost estimates, the deployment of a standardized tool for planning consolidation efforts could help ensure that agencies develop consistent and uniform projections. Until OMB requires agencies to use the model, agencies will likely continue to use a variety of methodologies and assumptions in establishing consolidation estimates, and it will remain difficult to summarize projections across agencies. Agencies reported experiencing multiple areas of success in their consolidation efforts. Specifically, 20 agencies identified 34 areas of success, with the number of agencies reporting a particular success ranging from 9 to 1. However, only 3 successes were identified by multiple agencies and, of those, 2 represent over 45 percent of the total reported successes. Four agencies—Justice, Transportation, NSF, and SSA—did not report any successes. Table 10 details the reported successes as well as the number of agencies identifying that area of success; the two most common areas are further discussed after the table. Virtualization is a technology that allows multiple, software-based machines with different operating systems, to run in isolation, side-by- side, on the same physical machine. Cloud computing is an emerging form of computing that relies on Internet-based services and resources to provide computing services to customers, while freeing them from the burden and costs of maintaining the underlying infrastructure. OMB suggests both technologies as agency approaches, along with decommissioning and consolidation. Nine agencies reported that focusing on virtualization and cloud computing have proven successful for their consolidation efforts. The Interior consolidation program manager cited virtualization as the department’s greatest consolidation success, noting the efforts of the department’s Bureau of Indian Affairs as an example. Specifically, Interior has documented virtualization as a key enabler in the efforts of the bureau to close data centers. After closing 11 data centers in fiscal year 2011, the bureau turned its attention to remote sites with more than three servers. Through virtualization, the bureau was able to reduce all remote sites to either one or two physical servers. Additionally, on a site-by-site basis, other application and database servers were either virtualized or migrated to one of two primary bureau data centers. In doing so, the bureau’s virtualization effort reportedly produced over $114,000 in cost avoidance savings for fiscal year 2011, is expected to produce over $66,000 in savings for fiscal year 2012, and is planned to produce further savings of $66,000 annually. Table 11 details reductions and savings that the bureau has already realized and plans for the future. Other agencies also reported virtualization as a key factor in being able to realize resource reductions. For example, EPA officials told us that the agency was using virtualization to optimize their IT infrastructure. In 2011, the agency virtualized over 360 servers, increasing the agency’s virtualization by 6 percent. In 2012, the agency plans to consolidate and virtualize email hosting services, allowing the agency to decommission 14 percent (or over 300) of its physical servers, and migrate the agency’s email gateways to cloud services. In one EPA facility, the agency will migrate over 100 servers from eight server rooms to one primary data center. Additionally, NRC reports that it used virtualization to exceed its 2011 goals for Windows server reductions. Specifically, the agency was able to exceed its goal of reducing 13 servers and actually reduced 33 physical servers, a reduction of more than 10 percent from its baseline of 288 servers. Further, OPM officials reported that within 15 months, the agency was able to increase the virtualization of its Windows servers from 15 percent to 50 percent, resulting in cost savings for the agency. Other agencies reported on the less tangible, but still significant, importance of virtualization to their efforts. An Education official told us that the department’s biggest success has come from focusing on virtualization, rather than physical consolidation. DHS reported that the increased implementation of virtualization will reduce the overall costs of the department’s migration and postmigration operations. Further, officials from Labor told us that they expect virtualization to have an impact on the results of their consolidation, but that they had not yet documented any of those results. Officials from three agencies also shared with us the advantages of moving their organizations to cloud services. Specifically, a DHS official told us that the department’s cloud services technology was becoming operational and as a result, costs savings were becoming evident versus traditional consolidation. Whereas 2 years ago, the department had nothing in the cloud, a large percentage of services were now moving in that direction. The official specifically noted a DHS component that was originally only going to move to its own physical infrastructure, but was now joining with other components because of the benefits of moving services to the cloud. A HUD official stated that the department’s successes were related to higher efficiency and utilization of computing and storage resources. Essentially, HUD embarked on a cloud-like solution—by means of the department’s existing outsourcing contract— before cloud computing really existed as a service. As a result, the official noted that the department has been receiving a number of benefits such as green IT, regular technology refreshes, and high utilization of resources. A second HUD official noted that the department has been rated across the government as having the third-highest computing utilization and the second-most efficient use of storage capacity. Further, and as mentioned earlier, an EPA official noted that the migration of EPA’s email gateways to cloud services will enable the agency to decommission 14 percent (over 300) of the agency’s physical servers. Eight agencies reported consolidation successes that had been realized through agencies working together, both within and outside of their department, to identify consolidation opportunities. For example, several of the agencies that reported success with virtualization, as discussed earlier, also reported working with other departmental components as a key enabler of resulting savings. Specifically, Interior reported that as part of the department’s closing of 11 Bureau of Indian Affairs data centers in fiscal year 2011, two facilities were consolidated with other Interior bureaus, resulting in a reduction of 43 of the bureau’s 65 servers and producing immediate cost avoidance savings of over $114,000. Defense noted the willingness of its components to adopt the departmental strategy of first looking to the Defense Information Systems Agency for application and data hosting before pursuing any other options. Further, a SBA official told us that one success from the consolidation effort was that agencies have been looking for ways to work together. Specifically, the official cited the SBA’s effort to reach out to another agency in order to craft an interagency agreement to work together and move part of their operations into the hosting agency’s systems. A second agency official noted that because of this, the hosting agency has contacted SBA to make sure that it included SBA’s needs in its planning and requests. Additionally, a DHS official told us that departmental components were joining together to move services to the cloud. There were also reported successes in working with external agencies. For example, VA reported that the department was successful in working with the Defense Information Systems Agency on an agreement to consolidate mission critical enterprise IT systems into the agency’s Defense Enterprise Computing Centers. The department noted that considerable cost savings could be realized by entering into such an interagency agreement, as opposed to leasing from a commercial site, for mission critical health record systems. Additionally, a HHS official similarly reported that the department’s Indian Health Service, which has small data centers that cannot close because of communication difficulties in their locations, recognized that Interior’s Bureau of Indian Affairs had a data center in close proximity to an Indian Health Service facility. Consequently, the service was able to share space with the bureau and consolidate one of its data centers and the service is now looking for similar opportunities that will allow HHS to consolidate further. Further, Labor officials told us that the department was consolidating small server rooms in regional offices to co-located facilities and that this approach was expected to reduce costs. The consolidation successes experienced by agencies indicate that aspects of FDCCI are moving forward as planned. Further, almost half of these reported accomplishments directly relate to key tenets of OMB’s plans for the initiative, demonstrating that OMB has developed a consolidation road map that provides realistic means by which agencies can achieve their goals. In 2011, we reported on the challenges that agencies were facing during data center consolidations. These included challenges related to FDCCI as well as those that were cultural, funding related, operational, and technical. In 2012, agencies have continued to report many of the same challenges, have reported new challenges, and have stopped reporting challenges they previously identified. As we found in 2011, some challenges are more common than others. Specifically, the number of agencies reporting a particular challenge range from 15 to 1. Additionally, 25 challenges reported in 2011 were not reported in 2012. Two agencies, HUD and NSF, did not report any challenges. Table 12 details the reported challenges, the numbers of agencies experiencing that challenge, and identifies the challenges no longer being reported by agencies. The table is followed by a discussion of the most prevalent challenges. Agencies reported seven challenges that are specific to FDCCI, including obtaining power usage and providing good quality asset inventories, both as required by OMB. Specifically, 15 agencies reported that obtaining power usage information was a challenge, which is less than the 19 agencies that reported this challenge last year. For example, a NASA official told us that the agency only had one data center (out of 79) that was fully metered, but that the agency was working to establish metering capabilities at several more locations. An SBA official told us that the agency was still working to complete a power audit, but that it was questionable whether such an audit would be worth the amount of work required to install separate power meters in leased facilities. A USAID official reported that none of the agency’s facilities were metered and that the agency was not a landlord for any of its facilities, making power information difficult to obtain. Further, 10 agencies reported that providing good quality inventories was a challenge, which is more than the 4 agencies that reported this challenge last year. For example, an EPA official told us that the agency had trouble determining cost information for its server rooms because most were facilities within office spaces and which were part of larger federal leases or within GSA buildings. As a result, EPA focused their efforts on facilities greater than 500 square feet. Additionally, a Defense official reported that that gathering and verifying inventory information for an organization the size of the department was challenging. Agencies reported three cultural challenges to data center consolidation, including accepting cultural change that is part of consolidation and obtaining enterprise buy-in to the consolidation effort. One of the most prevalently reported cultural challenges, accepting cultural change, was cited by 5 agencies, which is 10 fewer agencies than last year. For example, Energy found that there was a perceived need for each facility or departmental organization to have “ownership” of their own data centers and server rooms in order to support their business or mission needs. Justice recognized that moving from the department’s current environment to a more unified, standardized, and cost-efficient approach for providing data center services requires change and consequently, efforts were underway to drive more significant consolidation. Another commonly reported cultural challenge was obtaining enterprise buy-in to the consolidation effort, which was reported by 5 agencies—an increase from the single agency that reported this last year. For example, DHS reported their consolidation effort to be a multistakeholder operation that required immense amounts of coordination and found that delays and issues arose when various stakeholders maintained differing visions, expectations, and commitment to the effort. Further, NRC reported that one of its main challenges was managing the level of coordination required by the number of internal and external entities involved in planning and the related activities that need to happen simultaneously. Agencies reported two funding challenges: acquiring the funding needed for consolidation and identifying cost savings to be realized by consolidation. Nine agencies reported challenges with acquiring funding, which is slightly fewer than the 11 agencies that reported this challenge last year. For example, Energy reported that the department had little or no funding available to invest in data center measurement systems, server utilizations assessments, or consolidation projects. Additionally, both Justice and Transportation reported challenges in providing upfront funding for consolidation efforts before cost savings accrue. Two agencies reported challenges with identifying cost savings, a decrease from the 9 agencies that reported this challenge last year. For example, an Interior official noted that the department would likely not be able to report on savings for 2011 because most bureaus absorbed the cost of consolidation within their budgets. Although site-specific plans were required by the department, most did not address costs. Additionally, an SSA official noted that the agency currently had too many uncertainties surrounding its consolidation effort to perform a cost-benefit analysis. Agencies reported eight operational challenges, including difficulties with procurement and technology and resource constraints, neither of which had been reported in 2011. Three agencies encountered challenges with procurement, including DHS, which had to create a team to streamline projects through the department’s procurement process. GSA reported encountering construction contracting challenges on all three of the agency’s calendar year 2011 data center consolidations. These contract challenges included: vendors that could not meet award schedules, nonresponsive vendors, and long lead times for some IT equipment. To counter such delays, GSA increased the time allotted for planned contracting efforts and vendor delivery schedules. Additionally, two agencies reported challenges with technology and resource constraints. Specifically, EPA reported encountering minor delays in consolidation plan execution due to such constraints and NRC reported another of its main challenges to consolidation being available resources and the impact on its critical path to consolidation, which is the timely completion of the agency’s new headquarters building. Agencies reported only one technical challenge to consolidation, planning a migration strategy. Specifically, this was reported by seven agencies, an increase from the two agencies that reported this in 2011. For example, Transportation’s consolidation plan notes that in the department’s organization, it is a long process to identify possible consolidations, present them to management, then to users, and then work the technical side of migrations. Transportation’s plan also noted that application mapping is a very difficult and time-consuming activity, but cannot be skipped in a successful completion of a migration. Further, an Education official told us that the department had to develop a two- step approach for migrating files after encountering technical issues with an earlier migration effort. Finally, Commerce reported in its consolidation plan that detailed consolidation planning was critical due to the number of moving parts and potential impact on applications and customers. As we have previously reported, one approach agencies can use to manage challenges such as the ones described above is through risk management processes. In 2011, we reported that less than half of the agencies included a discussion of risk management in their data center consolidation plans. As we stated earlier, 18 of the agencies, or 75 percent, now fully address risk management. By addressing consolidation risk, agencies have better positioned themselves to manage the challenges they have identified. In any significant IT initiative, it is important that both successes and challenges be highlighted. In the case of FDCCI, a success highlights approaches and strategies that are helping agencies to meet their consolidation targets and fulfill the intent of the initiative. Conversely, a challenge identifies an area where agencies are struggling to meet the requirements and intent of this governmentwide effort. The two most reported consolidation successes are both key tenets of OMB’s FDCCI: the use of virtualization and cloud services, and working with other agencies to find consolidation opportunities. Alternately, the three most reported challenges directly impact the ability of FDCCI to meet its goals: gathering power usage information, developing good quality data center inventories, and acquiring the funding needed for consolidation. In light of how closely the successes and challenges reported by agencies relate to FDCCI, it will be important for OMB to continue to provide leadership and guidance to the initiative. This includes, as we have previously recommended, utilizing the existing accountability infrastructure of the Data Center Consolidation Task Force to assess agency consolidation plans to ensure they are complete and to monitor the agencies’ implementation of their plans. With agencies reporting having closed 286 data centers by the end of 2011 and planning to close an additional 346 centers by the end of 2012, the data center consolidation initiative is expected to realize about $2.4 billion in cost savings through 2015. OMB now requires agencies to annually update both their data center inventories and their consolidation plans and has expanded the required content of both. However, agencies’ consolidation and savings goals continue to be built on incomplete inventories and plans. To better ensure that FDCCI improves governmental efficiency and achieves promised cost savings, we are reiterating our prior recommendation to the department secretaries and agency heads of the 24 departments and agencies participating in the federal data center consolidation initiative to fully complete their consolidation inventories and plans expeditiously. As OMB refines its approach to the data center consolidation initiative, it provides updated guidance to agencies. However, three agencies did not learn of the most recent changes in OMB’s required formats, in part because the guidance was provided in meetings and not in a formal letter from the Federal CIO to agency CIOs or disseminated on the website where all prior guidance had been disseminated. Until OMB uses more structured mechanisms to disseminate its guidance, it runs the risk that agencies will not learn of important changes in format or approach. Additionally, basic consolidation plan requirements, such as the need for schedules and cost estimates, are still unmet by almost 70 percent of the agencies. Among the five agencies selected for a detailed review, none of the agencies’ master schedules and estimates were completed in a manner consistent with best practices. For example, none of the agencies was able to demonstrate that its cost estimates were accurate, credible, or comprehensive. OMB’s cost of ownership model should help address a number of planning concerns. As more agencies use the model, OMB can use the model to ensure consistent planning and reporting on consolidation efforts across FDCCI. However, agencies’ use of the model is still voluntary. Until OMB requires agencies to use the model, it may miss opportunities to ensure consistency among agencies and it will remain difficult to summarize projections across agencies. As the federal consolidation effort matures, agencies are beginning to realize successes. These constructive experiences, which stem from OMB’s recommended consolidation strategies, indicate that FDCCI is moving in the right direction. However, as agencies work towards their consolidation goals, many continue to report challenges related to gathering the necessary technical information and funding the consolidation itself. While these challenges are consistent with those reported in the past, over 25 previous challenges were no longer reported by the agencies. Such a dynamic environment reinforces the need for agencies to remain in communication in order to facilitate knowledge sharing and transfer and for OMB to continue to provide leadership and guidance. In addition to reiterating our prior recommendation to agencies to complete the missing elements of their inventories and plans, we are making two recommendations to OMB. Specifically, we recommend that the Director of OMB direct the Federal CIO to ensure that all future revisions to the guidance on data center consolidation inventories and plans are defined in OMB memorandum and posted to the FDCCI public website in a manner consistent with the guidance published in 2010, and ensure agencies utilize OMB’s Total Cost of Ownership model as a standardized planning tool across the consolidation initiative. In addition, we recommend that the Secretaries of Agriculture, Homeland Security, Interior, Transportation, and Veterans Affairs direct their component agencies and their data center consolidation program managers to implement recognized best practices when completing required program schedules and cost-benefit analyses. We received comments on a draft of our report from OMB, the 5 agencies to which we made recommendations, and the other 19 agencies mentioned in the report. Specifically, OMB and the 5 agencies to which we made recommendations either agreed with, or had no comment on, the recommendations and the other 19 agencies had no specific comments on our recommendations. Multiple agencies also provided technical comments, which we incorporated as appropriate. Each agency’s comments are discussed in more detail below. In oral comments, OMB officials, including the Deputy Federal CIO and staff from the Office of E-government and Information Technology and the Office of the General Counsel, stated that they generally agreed with, and described planned actions to implement, our recommendations. These officials also provided technical comments, which we have incorporated as appropriate. In written comments, Agriculture’s Acting CIO stated that the department concurred with the content of the report and had no comments. The department offered no comments on our recommendations. The department also provided technical comments, which we have incorporated as appropriate. Agriculture’s written comments are provided in appendix III. In written comments, DHS’s Director of the Departmental GAO/OIG Liaison Office concurred with our recommendation, commented on the current and planned state of the department’s consolidation efforts, and outlined actions the department plans to take to implement our recommendation and update its data center inventory and consolidation plan. The department also provided technical comments, which we have incorporated as appropriate. DHS’s written comments are provided in appendix IV. In written comments, Interior’s Assistant Secretary for Policy, Management and Budget stated the department concurred with the report’s finding and recommendations, commented on the current status of the department’s consolidation efforts, and described the department’s plans to develop savings and cost avoidance projections. The department also provided technical comments, which we have incorporated as appropriate. Interior’s written comments are provided in appendix V. In comments provided via e-mail, Transportation’s Deputy Director of Audit Relations wrote that the department had no comments on the draft. The department offered no comments on the recommendations. In written comments, VA’s Chief of Staff stated that the department generally agreed with our conclusions, concurred with our recommendation, and described planned actions to address our recommendation. The department also provided technical comments, which we have incorporated as appropriate. VA’s written comments are provided in appendix VI. In written comments, Commerce’s Acting Secretary concurred with the report’s general findings as they applied to the department and with the specific reporting on the department’s consolidation plan. Commerce’s written comments are provided in appendix VII. In comments provided via e-mail, an audit liaison from Defense’s Office of the CIO wrote that the department had no comments on the report. In comments provided via e-mail, an official from Education’s Office of the Secretary wrote that the department had no comments on the report. In written comments, the Director of Energy’s Corporate IT Project Management Office stated that the department concurred with the findings reported for Energy and noted steps being taken by the department to address a consolidation challenge discussed in our report. The department also elaborated on facilities that we cited as not having been reported in Energy’s FDCCI inventory. Energy’s written comments are provided in appendix VIII. In written comments, HHS’ Assistant Secretary for Legislation stated our report was an accurate representation of the department’s 2011 data center inventory and consolidation plan and outlined actions the department plans to take to complete missing inventory and plan elements. HHS’ written comments are provided in appendix IX. In comments provided via e-mail, a HUD audit liaison wrote that the department had no comments or concerns regarding the report. In comments provided via e-mail, an official from Justice’s Office of the CIO wrote that the department had no comments on the report. In written comments, Labor’s Assistant Secretary for Administration and Management stated that the department did not have any comments on the draft to contribute. Labor’s written comments are provided in appendix X. In comments provided via e-mail, an official from State’s Office of the Chief Financial Officer wrote that the department had no comments on the report. In comments provided via e-mail, Treasury’s Deputy Assistant Secretary for Information Systems agreed with our report. The department also provided technical comments, which we have incorporated as appropriate. In written comments, the Director of EPA’s Office of Technology Operation and Planning provided technical comments, which we have incorporated as appropriate. The agency did not comment on the report’s findings. In comments provided via e-mail, an official from GSA’s GAO/IG Audit Response Division wrote that the agency had no comments on the report. In comments provided via e-mail, the team lead for NASA’s GAO/OIG Audit Liaison wrote that the agency was providing no comments on the report. In written comments, NSF’s CIO stated that the agency generally agreed with our characterization of their consolidation plan, but disagreed with our assessment of the agency’s master program schedule. The CIO asserted that we were provided with such a schedule, while also acknowledging that the schedule was inherently less detailed than those of agencies and departments with multiple components, but that it identified all NSF consolidation activities in the format and level of detail prescribed by OMB. However, OMB’s guidance on master program schedules states that such schedules are to be created from the detailed implementation schedules provided by data center managers, as well as driven by related federal government activities, such as OMB reporting and budgeting. While we acknowledge that NSF’s consolidation scope is less than that of some agencies, the high-level timeline presented as a master program schedule consists only of a single line item that states the fiscal year when NSF’s data center will be decommissioned. Further, this timeline does not include any of the detailed implementation activities or key baseline milestones required by OMB. As such, we believe our evaluation is reasonable and appropriate. NSF’s written comments are provided in appendix XI. In comments provided via e-mail, the NRC OIG and GAO Liaison wrote that the agency had reviewed the report and had no comments. The liaison also provided an update on NRC’s plans to move to a single data center. In comments provided via e-mail, an official from OPM’s Office of Internal Oversight and Compliance wrote that the agency had no comments on the report. In comments provided via e-mail, the program manager for SBA’s Office of Congressional and Legislative Affairs provided technical comments, which we have incorporated as appropriate. In comments provided via e-mail, a SSA audit liaison wrote that the agency had no comments on the report. In comments provided via e-mail, an official from USAID’s Office of the Chief Financial Officer wrote that the agency had no comments on the report. We are sending copies of this report to interested congressional committees; the secretaries and agency heads of the departments and agencies addressed in this report; and other interested parties. In addition, the report will be available at no charge on GAO’s website at http://www.gao.gov. If you or your staffs have any questions on the matters discussed in this report, please contact me at (202) 512-9286 or pownerd@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix XII. Our objectives were to (1) evaluate the extent to which agencies have updated and verified their data center inventories and data center consolidation plans, (2) evaluate the extent to which selected agencies have adequately completed key elements of their consolidation plans, and (3) identify agencies’ notable consolidation successes and challenges. For this governmentwide review, we assessed the 24 departments and agencies (agencies) that were identified by the Office of Management and Budget (OMB) and the Federal Chief Information Officer (CIO) to be included in the Federal Data Center Consolidation Initiative (FDCCI). Table 13 lists these agencies. To evaluate the agencies’ updated data center inventories and consolidation plans, we reviewed OMB’s guidance and identified key required elements for each type of document. We compared agency consolidation inventories and plans to OMB’s required elements, and identified gaps and missing elements. We rated each element as “Yes” if the agency provides complete information; “Partial” if the agency provides some, but not all, of the information; and “No” if the agency does not provide the information. We followed up with agencies to clarify our initial findings and to determine why parts of the inventories and plans were incomplete or missing, as applicable. We also compared our findings with those reported in 2011. To assess the reliability of the data agencies provided in their data center inventories and plans, we reviewed the letters agencies were required to submit attesting to the completeness and reliability of their inventories and plans, we interviewed agency officials about the actions taken to verify their data, and reviewed those results against our past reviews of agency inventories and plans. We concluded that the data were sufficiently reliable for our purposes, which was to report on the completeness of the inventories and plans. GAO-11-565. each practice as having been fully, partially, or not addressed. We discussed our findings with agency officials to determine why the schedules did not address all aspects of the best practices. To assess the agencies’ cost estimates, we compared documentation supporting the cost and savings estimates found in the agencies’ consolidation plans with relevant best practices. These practices include ensuring that each estimate is comprehensive, well-documented, accurate, and credible. In doing so, for each estimate, we rated each practice as having been met, substantially, partially, minimally, or not met. We also discussed our findings with agency officials to determine why the estimates did not address all aspects of the best practices. To assess the reliability of the data the five agencies provided in their master program schedules and cost estimates, we reviewed the schedules and estimates, compared them to our guidance on scheduling and estimating, and interviewed officials about how the schedules and estimates were constructed. We concluded that the schedules and estimates were generally unreliable and our report includes findings related to those assessments. The results of our evaluation at these five agencies cannot be generalized to other agencies. To identify the key successes and challenges encountered by agencies in consolidating data centers, we reviewed agency consolidation plans and interviewed agency officials. We then determined which successes and challenges were encountered most often. To assess the reliability of cost savings data reported by Interior, we confirmed that the information was included in the department’s updated consolidation plan, which the Interior CIO attested was assessed and determined to be accurate and complete. In doing so, we concluded that the quality of the information was sufficient for our purposes. We conducted this performance audit from September 2011 to July 2012, in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. As part of its data center consolidation initiative, OMB required 24 federal departments and agencies to submit an updated data center inventory and consolidation plan. Key elements of the inventory were to include, for each data center, information on physical servers, virtualization, IT facilities and energy, network storage, and data center information. However, 3 agencies reported their inventories based on 2010 guidance, in which case they included information for each data center on IT hardware, IT software, facilities/energy/storage, and geographic location. Key elements of the updated plan were to include information on quantitative goals, qualitative impacts, consolidation approach, consolidation scope, timeline, performance metrics, master schedule, cost-benefit analysis, risk management, consideration of a communications plan, inventory and plan verification, consolidation progress, and cost savings. For each of the agencies, the following sections provide a brief summary of the agencies’ goal for reducing the number of data centers, and an assessment of the completeness of their inventories and plans, as compared to what we reported in 2011. In the case of agencies that reported using the new inventory format, we have related the old key elements, where possible. Agencies that reported using the old format are directly compared to their previous results. The following information describes the key that we used in tables 14 through 37 to convey the results of our assessment of the agencies’ compliance with OMB’s requirements for the FDCCI. ● – the agency provides complete information for this element. ◐ – the agency provides some, but not all, aspects of the element. ○ – the agency does not provide information for this element. Agriculture plans to consolidate from 95 data centers (40 large and 55 small) to 27 centers (8 large and 19 small) by December 2015. However, the agency’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 2 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, Agriculture provides complete information for 9 of the 13 elements evaluated and provides partial information for the remaining 4 elements. An Agriculture official stated that the agency is dependent on component agencies to report complete inventory information. The official also stated the agency intended to provide the missing utilization plan information, as well as greater discussion of consolidation challenges in future consolidation plan updates. Table 14 provides our assessment of Agriculture’s compliance with OMB’s requirements in 2010 and 2011. The Department of Commerce (Commerce) plans to consolidate from 55 data centers (33 large and 22 small centers) to 30 data centers (21 large and 9 small centers) by December 2015. However, Commerce’s asset inventory remains incomplete, while its consolidation plan is now complete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, Commerce provides complete information for all 13 elements evaluated. A Commerce official stated that energy information is incomplete due to the lack of metering in its facilities and the inability of data center providers to supply agency-specific energy usage and cost information. Table 15 provides our assessment of Commerce’s compliance with OMB’s requirements in 2010 and 2011. The Department of Defense (Defense) plans to consolidate from 936 data centers to 392 by December 2015. However, Defense’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides partial information for all 5 key elements. Additionally, in its consolidation plan, Defense provides complete information for 5 of the 13 elements evaluated, provides partial information for 3 elements, and does not provide information for 5 elements. A Defense official stated that the agency’s next inventory update would include more complete information. In addition, the official stated that it was a challenge for Defense to collect all of the required information because of the scope and size of the agency’s consolidation effort. Table 16 provides our assessment of Defense’s compliance with OMB’s requirements in 2010 and 2011. The Department of Education (Education) plans to consolidate from five data centers (three large and two small centers) to four data centers (three large and one small center) by December 2012. However, Education’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, Education provides complete information for 8 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide information for 1 element. Education officials stated that 2 elements were not applicable because of the small scope of the agency’s effort. Table 17 provides our assessment of Education’s compliance with OMB’s requirements in 2010 and 2011. The Department of Energy (Energy) plans to consolidate from 56 data centers (26 large and 30 small centers) to 50 data centers (21 large and 29 small centers) by December 2015. However, Energy’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, Energy provides complete information for 8 of the 13 elements evaluated, provides partial information for 3 elements, and does not provide information for 2 elements. An Energy official stated that the agency’s next inventory update would include more complete information. In addition, the official stated that a risk management plan was under development and that the agency planned to work with OMB’s cost model to formulate better cost and savings information. Table 18 provides our assessment of Energy’s compliance with OMB’s requirements in 2010 and 2011. The Department of Health and Human Services (HHS) plans to consolidate from 181 data centers (43 large and 138 small centers) to 145 data centers (36 large and 109 small centers) by December 2015. However, HHS’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 1 key element and provides partial information for the remaining 4 elements. Additionally, in its consolidation plan, HHS provides complete information for 11 of the 13 elements evaluated and provides only partial information for the remaining 2 elements. An HHS official noted that it was difficult to gather every inventory element for all of its data centers. Table 19 provides our assessment of HHS’s compliance with OMB’s requirements in 2010 and 2011. DHS plans to consolidate from 101 data centers (40 large and 61 small data centers) to 37 data centers (3 large and 34 small centers) by December 2015. However, DHS’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides partial information for all 5 elements. Additionally, in its consolidation plan, DHS provides complete information for 10 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide information for 1 element. DHS officials stated that the completeness of inventory information has improved since 2011 and that they have developed performance metrics. They also noted that they do not expect to fully realize their cost savings until consolidation activities are complete. Table 20 provides our assessment of DHS’s compliance with OMB’s requirements in 2010 and 2011. The Department of Housing and Urban Development (HUD) has achieved its goal of consolidation prior to the start of the FDCCI and does not plan further consolidation of its existing base of contracts. Since 2005, the agency has operated in a fully outsourced infrastructure mode with two vendors providing consolidated departmental IT operations in hosting, storage, data transport, user environments, and systems integration, with off-site disaster recovery provided by one vendor. The agency’s asset inventory is complete, but its consolidation plan is not. Specifically, HUD provides complete information for 5 of the 13 elements evaluated and provides partial information for 1 element. A HUD official stated that 7 elements were not applicable because the agency has reached its consolidated end-state architecture. Table 21 provides our assessment of HUD’s compliance with OMB’s requirements in 2010 and 2011. Interior plans to consolidate from 232 data centers (158 large and 74 small data centers) to 135 data centers (90 large and 45 small centers) by December 2015. However, Interior’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, Interior provides complete information for 9 of the 13 elements evaluated, provides partial information for 3 elements, and does not provide information for 1 element. An Interior official stated that the agency expects to report more complete inventory information for the next inventory update and will report cost savings when it can more accurately estimate the agency’s expected savings. Table 22 provides our assessment of Interior’s compliance with OMB’s requirements in 2010 and 2011. The Department of Justice (Justice) plans to consolidate from 105 data centers (33 large and 42 small centers and 30 centers of unknown size) to 66 data centers (27 large and 39 small centers and no centers of unknown size) by December 2015. However, Justice’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides only partial information for all 5 key elements. Additionally, in its consolidation plan, Justice provides complete information for 10 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide any information for 1 element. A Justice official stated that the agency did not know it was required to report the missing inventory information, but that the agency had the information and would include it in the next inventory update. The official did not know when the agency’s savings and utilization goals would be updated. Table 23 provides our assessment of Justice’s compliance with OMB’s requirements in 2010 and 2011. The Department of Labor (Labor) plans to consolidate from 89 data centers (20 large and 69 small centers) to 54 data centers (20 large and 34 small centers) by December 2015. However, Labor’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 2 key elements and provides partial information for the remaining 3 elements. Additionally, in its consolidation plan, Labor provides complete information for 4 of the 13 elements evaluated, provides partial information for 6 elements, and does not provide information for 3 elements. A Labor official stated that the agency had difficulty obtaining energy information because of the lack of metering in its facilities. The official also noted that cost information would not be available until the end of fiscal year 2012 while savings information would not be available until fiscal year 2013. Table 24 provides our assessment of Labor’s compliance with OMB’s requirements in 2010 and 2011. The Department of State (State) plans to consolidate from 363 data centers (12 large and 351 small data centers) to 355 data centers (4 large and 351 small centers) by December 2015. According to agency officials, the 351 small data centers are located overseas and there are no current plans to consolidate these locations because of the resulting impact on information technology operations. However, State’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 1 key element and provides partial information for the remaining 4 elements. Additionally, in its consolidation plan, State provides complete information for 9 of the 13 elements evaluated, provides partial information for 3 elements, and does not provide information for 1 element. State officials stated that the agency focused on inventorying its larger domestic facilities and noted that it was difficult to capture inventory-related information, such as energy usage and costs, for its foreign posts. The officials added that State has since completed a cost-benefit analysis, the results of which would be included in the next update, and has developed detailed schedules for each year’s activities. Table 25 provides our assessment of State’s compliance with OMB’s requirements in 2010 and 2011. Transportation plans to consolidate from 328 data centers (33 large and 295 small centers) to 265 data centers (24 large and 241 small centers) by December 2015. However, Transportation’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, Transportation provides complete information for 8 of the 13 elements evaluated, provides partial information for 3 elements, and does not provide information for 2 elements. A Transportation official stated that the agency did not expect to see significant improvements for the energy-related information because not all facilities have meters. The official added that it was a challenge for the agency to collect inventory data for its small data centers. Table 26 provides our assessment of Transportation’s compliance with OMB’s requirements in 2010 and 2011. The Department of the Treasury (Treasury) plans to consolidate from 55 data centers (42 large and 13 small centers) to 40 data centers (29 large and 11 small centers) by December 2015. However, Treasury’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 2 key elements and provides partial information for the remaining 3 elements. Additionally, in its consolidation plan, Treasury provides complete information for 6 of the 13 elements evaluated, provides partial information for 4 elements, and does not provide information for 3 elements. A Treasury official stated that installing meters to gather all inventory power information would be cost prohibitive. In addition, the official stated that the agency is working to complete the missing plan elements, including the master program schedule, risk management plan, and communications plan. Table 27 provides our assessment of Treasury’s compliance with OMB’s requirements in 2010 and 2011. VA plans to consolidate from 97 data centers (51 large and 46 small centers) to 14 data centers (11 large and 3 small centers) by December 2015. However, VA’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 of the key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, VA provides complete information for 10 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide any information for the remaining 1 element. A VA official stated that installing equipment to gather all inventory power information would be cost prohibitive. Another official stated that the agency would more fully report on cost savings in future versions of their consolidation plan. Table 28 provides our assessment of VA’s compliance with OMB’s requirements in 2010 and 2011. The Environmental Protection Agency (EPA) plans to consolidate from 78 data centers (4 large and 74 small centers) to 53 data centers (4 large and 49 small centers) by December 2015. However, EPA’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 of the key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, EPA provides complete information for 10 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide any information for the remaining 1 element. An EPA official stated that the agency planned to develop energy estimates for the missing inventory information and to work with OMB’s cost model to develop better cost and savings information. Table 29 provides our assessment of EPA’s compliance with OMB’s requirements in 2010 and 2011. The General Services Administration (GSA) plans to consolidate from 21 data centers (21 large and no small centers) to 9 data centers (9 large and no small centers) by December 2015. However, GSA’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides partial information for all 5 key elements. Additionally, in its consolidation plan, GSA provides complete information for 10 of the 13 elements evaluated and provides partial information for the 3 remaining elements. A GSA official stated that the agency had now completed all missing IT facilities and energy information, but that there were continuing difficulties in calculating savings information due to changing schedules and lack of energy metering information for some GSA facilities. Table 30 provides our assessment of GSA’s compliance with OMB’s requirements in 2010 and 2011. The National Aeronautics and Space Administration (NASA) plans to consolidate from 79 data centers (75 large and 4 small data centers) to 22 large data centers by December 2015. However, NASA’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, NASA provides complete information for 10 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide information for the remaining element. A NASA official stated that currently only one facility has power metering and, as a result, it is difficult to determine costs. The official also noted that NASA expects to reach its 2012 consolidation targets. Table 31 provides our assessment of NASA’s compliance with OMB’s requirements in 2010 and 2011. The National Science Foundation (NSF) currently has only one onsite, centrally managed data center. Since 2007, the agency has been transitioning from owning and operating a data center to the use of commercial data center services and emerging cloud computing options. The agency’s plan is to complete transition of major legacy IT systems in a phased approach, with completion coinciding with the expiration of the NSF headquarters building lease, currently set for fiscal year 2014. The agency’s asset inventory is complete, but its consolidation plan is not. Specifically, NSF provides complete information for 10 of the 13 elements evaluated, provides partial information for 1 element, and does not provide information for 2 elements. An NSF official stated that the agency interpreted the guidance for consolidation progress and cost savings to apply only to ongoing or completed consolidations. However, the official noted that the agency would more fully report on these elements in future versions of its consolidation plan. Table 32 provides our assessment of NSF’s compliance with OMB’s requirements in 2010 and 2011. The Nuclear Regulatory Commission (NRC) plans to consolidate from three data centers (three large and no small centers) to one large data center by December 2015. However, NRC’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 4 of the key elements and partial information for the remaining 1 element. Additionally, in its consolidation plan, NRC provides complete information for 8 of the 13 elements evaluated, provides partial information for 3 elements, and does not provide information for the remaining 2 elements. An NRC official stated that the agency planned to gather missing data center information and that the agency’s planned single data center would be able to provide much of NRC’s missing energy information. The official also stated that both performance metrics and a master program schedule have now been developed. Table 33 provides our assessment of NRC’s compliance with OMB’s requirements in 2010 and 2011. The Office of Personnel Management (OPM) plans to consolidate from 4 data centers (one large and three small centers) to 3 centers (one large and two small centers) by December 2015. However, the agency’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 1 key element and provides partial information for the remaining 3 elements. Additionally, in its consolidation plan, OPM provides complete information for 6 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide information for 3 elements. Two elements were determined to be not applicable to the agency. An OPM official stated that several missing elements, such as more detailed and complete inventory information and a summary of the agency’s cost-benefit analysis would be provided in future updates. The official also stated that the agency was not aware that it had to include consolidation progress and cost savings information in its updated consolidation plan. Another OPM official indicated the agency intended to provide information required by OMB’s guidance in the future. Table 34 provides our assessment of OPM’s compliance with OMB’s requirements in 2010 and 2011. The Small Business Administration (SBA) plans to consolidate from four large data centers to two large centers by December 2015. However, the agency’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 2 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, SBA provides complete information for 6 of the 13 elements evaluated, provides partial information for 2 elements, and does not provide information for the remaining 5 elements. SBA officials stated that several missing elements, such as performance metrics, a schedule, and a risk management strategy, were either developed after the plan’s completion or would be developed in the future. Table 35 provides our assessment of SBA’s compliance with OMB’s requirements in 2010 and 2011. The Social Security Administration (SSA) has two large data centers and plans to replace one of them with a new facility. The agency expects the transition to begin in February 2015 and be complete in August 2016. However, SSA’s consolidation plan remains incomplete. In its asset inventory, the agency provides complete information for all 5 key elements. Additionally, in its consolidation plan, SSA provides complete information for 7 of the 13 elements evaluated, provides partial information for 4 elements, and does not provide information for the remaining 2 elements. An SSA official stated that the missing utilization plan elements and the plan verification information were unintentionally omitted and that those items would be included in the next update. Table 36 provides our assessment of SSA’s compliance with OMB’s requirements in 2010 and 2011. The U.S. Agency for International Development (USAID) plans to consolidate from six data centers (two large and four small data centers) to two small data centers by December 2012. However, USAID’s asset inventory and consolidation plan remain incomplete. In its asset inventory, the agency provides complete information for 3 key elements and provides partial information for the remaining 2 elements. Additionally, in its consolidation plan, USAID provides complete information for 7 of the 13 elements evaluated, provides partial information for 4 elements, and does not provide information for the remaining 2 elements. A USAID official stated that missing server information would be included in the next inventory update and that the agency has completed a new cost- benefit analysis and taken steps to verify its inventory data. The official also said that power-related information is difficult to obtain since the agency leases its data centers. Table 37 provides our assessment of USAID’s compliance with OMB’s requirements in 2010 and 2011. In addition to the contact named above, individuals making contributions to this report included Colleen Phillips (Assistant Director), Justin Booth, Kathleen Lovett Epperson, Rebecca Eyler, Dave Hinchman, Fatima Jahan, Jason Lee, John Ockay, Karen Richey, and Jessica Waselkow.

In 2010, as focal point for information technology management across the government, OMB’s Federal Chief Information Officer launched the Federal Data Center Consolidation Initiative—an effort to consolidate the growing number of federal data centers. In July 2011, GAO evaluated 24 agencies’ progress on this effort and reported that most agencies had not yet completed data center inventories or consolidation plans and recommended that they do so. In this subsequent review, GAO was asked to (1) evaluate the extent to which the 24 agencies updated and verified their data center inventories and plans, (2) evaluate the extent to which selected agencies have adequately completed key elements of their consolidation plans, and (3) identify agencies’ notable consolidation successes and challenges. To address these objectives, GAO assessed the completeness of agency inventories and plans, analyzed the schedule and cost estimates of 5 agencies previously reported to have completed one or both estimates, and interviewed officials from all 24 agencies about their consolidation successes and challenges. As of the most recent agency data submitted in September 2011, 24 agencies identified almost 2,900 total centers, established plans to close 1,186 of them by 2015, and estimated they would realize over $2.4 billion in cost savings in doing so. However, while the Office of Management and Budget (OMB) required agencies to complete missing elements in their data center inventories and plans by the end of September 2011, only 3 agencies submitted complete inventories and only 1 agency submitted a complete plan. For example, in their inventories, 17 agencies do not provide full information on their information technology facilities and energy usage, and 8 provide only partial information on their servers. Further, in their consolidation plans, 13 agencies do not provide a full master program schedule and 21 agencies do not fully report their expected cost savings. Officials from several agencies reported that some of this information was unavailable at certain facilities or that the information was still being developed. In a prior report, GAO recommended that agencies complete the missing elements from their inventories and plans. Until these inventories and plans are complete, agencies will continue to be at risk of not realizing anticipated savings, improved infrastructure utilization, or energy efficiency. OMB requires a master program schedule and a cost-benefit analysis (a type of cost estimate) as key requirements of agencies’ consolidation plans, but none of the five agencies GAO reviewed had a schedule or cost estimate that was fully consistent with the four selected attributes of a properly sequenced schedule (such as having identified dependencies), or the four characteristics that form the basis of a reliable cost estimate (such as being comprehensive and well-documented). For example, the Departments of Interior and Transportation did not have schedules and the Department of Agriculture’s schedule was consistent with three of four attributes. Additionally, cost estimates for the Departments of Homeland Security and Veterans Affairs were partially consistent with the four cost characteristics. In the absence of reliable schedules and estimates, these agencies are at risk of experiencing cost overruns, missed deadlines, and performance shortfalls. OMB has established a standardized cost model to aid agencies in their consolidation planning efforts, but use of the model is voluntary. Many federal agencies reported consolidation successes. Notably, 20 agencies identified 34 areas of success, although only 3 of those areas were reported by more than 1 agency. The two most-reported successes were focusing on the benefits of key technologies and the benefits of working with other agencies and components to identify consolidation opportunities. However, agencies have continued to report a number of the same challenges that GAO first described in 2011, while other challenges are evolving. For example, 15 agencies reported continued issues with obtaining power usage information and 9 agencies reported that their organization continued to struggle with acquiring the funding required for consolidation. However, other challenges appear to be less prevalent, including challenges in identifying consolidation cost savings and meeting OMB’s deadlines. Overall, 25 challenges that were reported in 2011 were no longer reported in 2012. In light of these successes and challenges, it is important for OMB to continue to provide leadership and guidance, such as—as GAO previously recommended—using the consolidation task force to monitor agencies’ consolidation efforts. OMB’s Federal Chief Information Officer should ensure that agencies use a standardized cost model to improve consolidation planning, and the 5 selected agencies should implement recognized best practices when establishing schedules and cost estimates for their consolidation efforts. OMB and 3 agencies agreed with, and 2 did not agree or disagree with, GAO’s recommendations.

Performance-based logistics is the DOD term for the process of (1) identifying a level of performance required by the warfighter and (2) negotiating a performance-based contract between the government and the product support integrator—that is generally the original equipment manufacturer of the total system—to provide long-term total system support for a weapon system at a fixed level of annual funding. Instead of buying spares, repairs, tools, and data in individual transactions, the method in a performance-based logistics arrangement is to buy a predetermined level of availability that meets the warfighter’s objectives. To implement performance-based logistics, DOD selects a product support integrator to serve as the single point of accountability, integrating support from all sources to achieve the performance outcome metrics specified in the performance-based support agreement. The metrics used include operational availability (a measure of the degree to which an item is in an operable state and can be committed at the start of a mission when the mission is called for at an unknown point in time); mission capability (the material condition, indicating that it can perform at least one and potentially all of its designated missions); and customer wait time (the total elapsed time between issuance of a customer order and fulfillment of that order). For example, the Navy now uses two metrics for its performance-based contract for the T-45 aircraft system—“ready for training,” which requires that the contractor have a minimum number of aircraft ready for training at 7:00 AM each business day in order to achieve a 57 percent aircraft availability; and “sortie completion,” which requires that the contractor meet 98 percent of the requirements for the scheduled training flights. As an incentive, the contract pays a performance bonus (maximum of $5 million annually) if the contractor exceeds the performance metrics. If the contractor only meets—or fails to meet—the minimum metrics, the contractor then receives none of the annual performance bonus. DOD Directive 5000.1, the Defense Acquisition System, highlights the department’s preference for using performance-based logistics at the platform level, stating, “Program Managers shall develop and implement performance-based logistics strategies that optimize total system availability while minimizing cost and logistics footprint.” As part of its implementation of this strategy, in 2003 DOD proposed that the Congress adopt legislative changes that would allow the services to increase the appropriations allocation flexibility within a weapon system program, allowing the program manager to use funds from different accounts (such as operation and maintenance; research, development, test, and evaluation; and procurement) to pay for system support costs. Although this proposal was not adopted, DOD continues to pursue various avenues that would support the overall objective of having greater flexibility by using a single line of support funding managed by the program office for total system operation and maintenance costs. Most recently, on February 4, 2004, the Deputy Secretary of Defense (1) directed the Under Secretary of Defense (Acquisition, Technology, and Logistics) in conjunction with the Under Secretary of Defense (Comptroller) to issue clear guidance on purchasing using performance criteria; and (2) directed each service to provide a plan to aggressively implement performance- based logistics, including transferring appropriate funding as needed, on current and planned weapon system platforms for fiscal years 2006–2009. While this directive does not preclude the services from using performance-based logistics contracts below the platform level, it does express DOD’s intent to apply the concept at the platform level as a preferred practice. As we discuss in the next section, DOD has established separate goals for implementing performance-based service contracts, and the services have identified many contracts as performance-based logistics arrangements that are, in fact, below the platform level. However, according to Office of Secretary of Defense officials, DOD would like to implement performance-based logistics at the platform level to move from contracting for material availability to weapon system availability. DOD considers that the platform level offers the metrics needed to implement a true performance-based logistics arrangement. The Office of Management and Budget indicates that performance-based service contracting, from which performance-based logistics has evolved, has been referenced in regulation, guidance, and policy for more than two decades, and federal agencies have used performance-based contracting to varying degrees for acquiring a range of services. In 1991 the Office of Management and Budget issued a policy letter establishing the use of a performance-based approach for service contracting, and in 1994 it initiated a governmentwide pilot project to encourage the use of performance-based service contracts in federal agencies, including DOD. The use of performance-based service contracts to acquire services offers a number of potential benefits, particularly when services are acquired by means of a fixed price agreement. Performance-based contracts can encourage contractors to be innovative and to find cost-effective ways of delivering services for a fixed level of funding. By shifting the focus from process to results, these contracts can potentially produce better outcomes and reduced costs. In view of the potential benefits, Congress has been encouraging greater use of performance-based service contracting. In an August 2003 memorandum to the military departments, the Under Secretary of Defense (Acquisition, Technology and Logistics) stated that DOD should continue to increase its use of performance-based service acquisitions. He noted that DOD has a goal to award 50 percent of contract actions and dollars using performance-based specifications by fiscal year 2005. The more specific concept of performance-based logistics as an approach for supporting military systems emerged from DOD’s 1999 study, Product Support for the 21st Century, which identified 30 pilot programs (10 in each military department) to test logistics support reengineering concepts that placed greater reliance on the private sector. Many of the pilots involved various types of contractor logistics support, prime vendor support, or performance-based type arrangements. Others focused on including reduced operation and support costs and improved readiness as performance requirements for new system development. The September 30, 2001, Quadrennial Defense Review advanced DOD’s move toward this concept by advocating the implementation of performance-based logistics with appropriate metrics that would be designed to compress the supply chain and improve the readiness of major weapon systems and commodities. A November 2001 Office of the Deputy Under Secretary of Defense document, Product Support for the 21st Century: A Program Manager’s Guide to Buying Performance, intended as a guide for program managers, stated that program managers will implement performance-based logistics on all new systems and on acquisition category I and II fielded systems selected on the basis of a sound business case. It is unclear how many performance-based logistics programs the services have implemented. In response to our inquiries, the Army identified 74 performance-based logistics programs, the Navy identified 106, the Air Force 4, and the Marine Corps 1. We noted a broad range of contract arrangements is identified under the performance-based logistics umbrella, with many of them initiated under a different name, such as contractor logistics support or total systems support responsibility and later identified as performance-based logistics arrangements. Most of the DOD performance-based logistics arrangements currently identified by the services are used for subsystems or components rather than for weapon system platforms. Fiscal years 2003 to 2007 Defense Planning Guidance required the services to submit plans that identified their implementation schedules for performance-based logistics to all new weapon systems and acquisition category I and II fielded systems. Similarly, a February 13, 2002, letter from the Under Secretary of Defense (Acquisition, Technology, and Logistics) to the services emphasized the need for the plans required by the Defense Planning Guidance and directed that the plans be issued by May 1, 2002. But although the services issued plans, they did not take an aggressive approach toward adopting this concept, according to Office of Secretary of Defense logistics officials. An October 2003 Defense Business Board report encouraged the department to move more quickly in adopting the performance-based logistics, stating, “Performance-based logistics is an industry best practice and a DOD best practice. DOD should consider using it for all its weapon systems, new and legacy, provided it is supported by a business-case analysis.” This task force was chartered by the Under Secretary of Defense (Comptroller) and Chief Financial Officer to describe private-sector best practices used in managing supply chain partnering arrangements and to propose how to apply such practices to the supply chain processes used by DOD. Citing this task force report, the aforementioned February 2004 Deputy Secretary of Defense memorandum to the military departments stated, “Delay in implementing this practice complicates our funding, limits industry flexibility, and increases DOD inventory. We must streamline our contracting and financing mechanisms aggressively to buy availability and readiness measured by performance criteria.” Because DOD proposes using performance-based logistics at the platform level as the predominant support strategy for its military systems, it may limit opportunities for savings from competition, volume discounts, and reduced administrative costs. Also, by often not contracting for long-term access to technical data, programs officials are further limiting their support options. In the private sector, performance-based contracting is a tool used according to the applicability of subsystem or component and circumstance, when it is cost-effective and reduces risk in a noncompetitive environment. DOD, by contrast, proposes using it as the predominant product support strategy for its military systems. Further, when private-sector companies use performance-based contracting, they use it at the subsystem or component level, retaining the program integration function themselves as a core business function essential to successful business operations. Conversely, DOD policy memoranda support using performance-based contracting at the platform level and using the contractor as the support integrator. Moreover, private sector companies emphasize the importance of having the rights to contracts and competition. DOD, in contrast, is frequently not acquiring the same level of technical data in its acquisition of new programs. While our review of private sector companies did find that half of those we interviewed are using performance-based contracting, the industry approach is much different from DOD’s preferred approach for performance-based logistics. As previously discussed, Office of the Secretary of Defense guidance has over the past several years encouraged the services to use performance-based logistics at the weapon system level as the preferred approach for life-cycle management of military systems. DOD officials have stated that this is an industry best practice and should be adopted more aggressively, but in 7 of 14 companies we interviewed that used some type of performance-based contracting, this agreement was used at the subsystem or component level—that is, for engines, auxiliary power units, wheels, or brakes— and it was generally used for older systems. The following chart characterizes the companies we interviewed by industry type, by the extent to which they outsource logistics support activities, by the predominant contracting practices used, and by the types of subsystems or components outsourced using performance-based contracting. Pseudonyms are used rather than the actual company names. These companies generate annual revenue generally exceeding $1 billion, and they use complex and expensive equipment for which they require high levels of availability and reliability as well as efficiency in managing lifecycle costs. The life-cycle management issues are comparable to those of DOD in managing its weapon system sustainment programs. As shown above, performance-based contracting in the companies we interviewed is most widely used in the air carrier industry, and it also has limited use in the energy exploration and mining industry. According to air carrier officials, time and material contracts are more prevalent than performance contracts, because industry prefers to use short-term (2 to 3 years) competitive contracts when possible. In a sole-source environment companies sometimes use longer-term (10 to12 years) performance-based contracts for supporting some subsystems such as engines, if they have sufficient historical data to establish an accurate baseline. For example, all but one of the air carrier industry companies had performance-based contracts for one or more engines. The amount of engine workload managed by performance-based contracts varied from company to company. For example, Company C, which outsourced 38 percent of its total maintenance workload, used performance-based contracts for one-fifth of its outsourced engine work; while Company A, which outsourced 65 percent of its maintenance workload, used performance-based contracts for all of its engine work. The air carrier companies did not use performance-based contracts for contracted work on airframes, work that generally comprises about 30 percent of the commercial aviation maintenance and repair market. Table 2 provides information regarding the percentage of dollars spent on the repair of each type of subsystem or component managed using performance-based contracts by the air carrier companies and the one non-air carrier company that used performance-based contracts. The subsystems or components for which the companies used performance-based contracts most widely were auxiliary power units and wheels and brakes. Company officials noted that performance-based contracts are a tool most often used selectively in a noncompetitive environment in an effort to control cost and reduce risk. Additionally, they said that performance- based contracting works better for subsystems and components where available cost and performance data are sufficient to establish a good business case analysis, noting that this is more difficult to accomplish for new systems, where performance data are uncertain. Performance-based contracts differ from traditional logistics contracts by focusing on the purchase of weapon system sustainment as an integrated package based on output measures—such as a predetermined level of system availability. In contrast, traditional transaction-based time and material contracts are used to purchase logistics inputs—such as quantities of spare parts, specific repair tasks, and engineering studies. Under transaction-based contracts, the government pays for each transaction as a separate deliverable; whereas under a performance-based contract, the contractor is being paid for achieving an outcome performance metric, regardless of what he does to achieve that performance. In concept, performance-based contracts encourage the contractor to achieve a high level of performance at a fixed cost. However, air carrier industry officials we interviewed said that entering into a performance- based contract without good baseline data introduces a higher level of risk that the arrangement may not be cost-effective. For example, officials from one company said they used a performance-based contract for the older of the two types of engines in the company’s inventory. Officials said they would wait to collect sufficient performance data on the newer engine before considering a performance-based contract. The officials noted that they had originally used a performance-based contract on the newer engine, but found that, because the reliability of the engine was greater than expected, the contract arrangement was not cost-effective. The company was able to change the contract to a time and material contract, to allow time to collect sufficient performance data to support a fact-based business case analysis to determine the company’s “should” cost amount for a performance-based contract. Performance-based contracting offers DOD opportunities to provide contractors incentives to achieve desired levels of operational performance at a fixed cost when the department has historical performance information. But in the absence of reliable and complete performance data as a baseline, the adoption of this approach as the preferred support strategy for new weapon systems could undermine DOD’s ability to negotiate cost-effective terms—particularly since the performance-based contracts at the weapon system level have cost-reimbursement elements, while the private-sector companies generally used fixed-price agreements. Private-sector officials noted that it is important to use fixed prices for materials, since the high price of materials is a key factor driving the companies to use performance-based contracts. DOD policy promotes using performance-based contracting differently from the way private-sector firms use it in supporting complex and expensive systems. The companies we reviewed generally used performance-based contracting at the subsystem level for engines and certain other components rather than at the platform level, as proposed by DOD. Furthermore, when using performance-based contracting, these companies do not contract out the program integration function, as the military services are doing. We found no performance-based contracts for maintenance of airframes or maintenance of any equipment platform among the private-sector companies we reviewed. Industry officials cited three reasons why they believe the use of performance contracts is more advantageous at the subsystem or component level. First, they prefer to take advantage of competition whenever it is available and to manage support contracts through the use of competitive procedures. For example, because airframe maintenance support is available from a competitive market, the companies generally use a combination of fixed price and time and material contracts for this category of service. Conversely, performance- based contracts are often used for engine repair because of the high cost of spare and repair parts that are available only from the original equipment manufacturer. Officials said that there are too few third-party repair vendors to foster competition. Second, company officials emphasized the importance of gaining purchasing power from volume discounts on subsystems or components across their entire fleet of systems as a reason for not implementing performance contracting at the platform level. Finally, by having contracts at the subsystem or component level, companies can avoid the administrative costs that would be charged by a prime integrator. Similar to the approach used by the companies we reviewed, we noted that the Navy has used performance-based contracts primarily at the subsystem or component level. Navy officials said that implementation at this level is easier because the service could implement this concept more readily under DOD’s current funding structure. The funding is handled through the working capital fund, with reimbursement to the fund coming from the sale of subsystems or components to the fleet. Navy officials also noted that by implementing performance-based logistics at this level, they can save money by competing subsystems or components where a competitive market exists, consolidating the requirements of multiple programs and leveraging their buying power to obtain a pricing advantage, and reducing administrative costs—advantages also recognized by the private sector. The Navy’s history of using a performance-based contract for logistics support of the T-45 trainer aircraft illustrates how savings may be achieved by implementing the concept at the subsystem level rather than the weapon system level. The program office originally had a performance- based contract for the entire weapon system with the original equipment manufacturer. The contract was a 5-year firm-fixed price with an option for a sixth year period. Program office officials said the sole metric used, ready-for-training aircraft, resulted in there being an insufficient number of aircraft available to fly scheduled training sorties. Additionally, because actual flying hours were fewer than forecasted, the Navy was paying for flying hours it was not flying. Concluding that benefits weren’t as expected, that the costs were too high, and that savings were achievable by negotiating separate contracts for the airframe and engine, the program office chose not to exercise the option. The new engine contract is a performance-based contract awarded on a sole-source basis to the engine manufacturer, and the airframe performance-based contract was awarded competitively. According to Navy program office officials, the revised approach resulted in a projected savings of $37 million in the first year and projected savings of $144 million at the end of a 5-year period. The savings are being achieved through elimination of the administrative costs charged by the prime contractor for the engine work and through competition for the aircraft system. Another potential adverse effect of awarding a performance-based contract at the weapon system level is the loss of management control and expertise over the system that private-sector firms said was essential to the success of their business operations. Industry officials said that managing their supplier base and ensuring the availability of their equipment to generate revenue is too critical to entrust to a second party. Further, they believe that contracting out support at a platform level by using a system integrator limits the potential to optimize savings through competition and volume discounts and adds administrative costs charged by the prime integrator for managing subcontractors. The spokespersons for every company we visited told us that when they purchase equipment they make sure to acquire the technical data necessary to support it, regardless of whether the company intends to support the equipment in-house or outsource some of its support operations. Company officials said that this data was essential to their own management and oversight functions. For example, officials from a company that outsources most of its repairs pointed out that its engineers use the data to perform such tasks as establishing reliability metrics, evaluating performance, and revising repair standards. Additionally, officials stated that owning the technical data afforded their companies the flexibility that enabled them either to perform the work in-house or to offer the work up for competition. Several company officials said that it is best to obtain the technical data at the time the equipment is purchased, when the buyer has the most leverage in its negotiations with the manufacturer. Trying to obtain the technical data at a later time is difficult to negotiate and more expensive. These companies do not price their technical data items separately. DOD program offices, however, negotiate a price for maintenance-and-repair technical data separately from the price of the military hardware systems. According to service competition- advocate officials, program managers faced with limited acquisition dollars often make trade-off decisions to buy increased weapon system capability in lieu of technical data. We reported in 2002 that DOD program offices have often failed to put adequate emphasis on obtaining needed technical data during the acquisition process. We recommended that DOD emphasize the importance of obtaining technical data and consider including a priced option for the purchase of technical data when considering proposals for new weapon systems or modifications to existing systems. DOD concurred with our recommendation, noting that there was a requirement in DOD 5000.2R for program offices to provide long-term access to data required for the competitive sourcing of systems support throughout the life cycle. Additionally, by implementing total life-cycle systems management, DOD would strengthen its emphasis on acquiring technical data when negotiating support agreements with logistics providers. Nonetheless, the DOD has further diminished the emphasis it places on the need to acquire rights to technical data. For example, in May 2003, DOD replaced its acquisition regulation with a streamlined instruction, which eliminated the prior regulation’s requirement for the program manager to provide for long-term access to data required for the competitive sourcing of weapon system support throughout the life cycle of the system. This language is now provided as guidance in the Interim Defense Acquisition Guidebook, but it is not mandatory that this guidance be followed. According to DOD and service logistics officials, program managers should develop strategies that provide the government with sufficient and affordable technical data rights to enable them to put the work out for competition or develop alternate public or private sources for weapon system support if performance-based logistics arrangements fail or become too expensive. Logistics officials recognize that program managers who implement performance-based logistics contracts on new weapon systems may wish to delay taking delivery of technical data early in the life of the system, because unlike the stable designs of commercial equipment purchased in the private sector, the data for cutting edge technology lacks maturity and is frequently changed. Alternatively, program managers sometimes pay the original equipment manufacturers both to maintain the technical and weapon system configuration data and to provide the program managers with sufficient access to enable them to manage and oversee the performance-based logistics contract. However, logistics officials agree that the product support strategy should clearly provide for the future delivery of the technical data when required to support competition or alternative source development. Service logistics and competition-advocate officials said that it is critical that this strategy be developed during the weapon system acquisition phase, when the program office has its greatest leverage in negotiating the price of the technical data and the conditions under which the manufacturer must deliver the data. For example, in the course of the acquisition of the V-22 aircraft engine, the Navy program office obtained a technical data license agreement, according to which the manufacturer agreed to deliver a complete data package if it failed to perform in compliance with the statement of work at the agreed-to price and schedule. Conversely, when the program office does not obtain the technical data at the time of purchase, the future costs for obtaining these data are not knowable and, without the leverage of the original package purchase, could be prohibitively expensive. In our review of data collected from DOD’s performance-based logistics program offices, we noted that DOD had not negotiated for the maintenance drawing packages for the Javelin missile, F-117 aircraft, and TOW missile improved target acquisition system, and DOD would have to purchase them at a later date at a price to be negotiated. In April 2004, the Logistics Management Institute reported in a review of performance-based logistics arrangements that it found no evidence to indicate either the quantity or the quality of logistics management data— including technical data—available to the government was compromised by the use of performance-based logistics arrangements. This report also noted, however, that the acquisition guidance published by the Office of the Secretary of Defense does not address strategies for terminating interim contractor support or performance-based logistics contracts. The Logistics Management Institute report recommended that the Defense Acquisition Working Group include performance-based logistics “exit strategy” guidance in the defense acquisition guidebook. Nonetheless, as we have previously noted, guidance in this handbook is not mandatory. The use of performance–based contracting for the support of complex and costly military systems offers opportunities for military program managers to incentivize contractors to achieve desired levels of weapon system performance. However, our review of the use of the practice in private-sector firms indicates that DOD’s proposed guidance to adopt performance-based logistics aggressively at the platform level could limit competition, and such guidance might not be the most cost-effective approach for using this concept. Additionally, although DOD based its rationale for using performance-based logistics at least partially on the perception that this is an industry best practice, it appears that perception is not the case. DOD’s approach toward implementing the concept appears inconsistent with the way private-sector companies we interviewed use performance-based contracting in acquiring support for their equipment, and DOD’s approach has risks that should be addressed as it develops its guidance for using performance-based logistics. Using performance-based logistics as the preferred approach for managing the support of major weapon system programs—even though private- sector company officials use performance-based contracting selectively, when appropriate and cost-effective—carries the risk of increasing life- cycle cost. Both private- and public-sector experiences with performance- based contracting illuminate the challenges involved in developing a meaningful baseline for establishing a performance-based arrangement for new systems, because not enough is known early in the program about performance characteristics and because there is risk to both the program office and the contractor that may translate into high cost. Additionally, the use of performance-based logistics can limit the competition that would be available for providing logistics support when support decisions are made at the subsystem or component level rather than at the platform level. Using performance-based logistics at the platform level also creates risk by contracting out the program integration function—a core function that private contractors consider essential for the cost-effective management of costly and complex systems over their life cycle. Finally, adopting performance-based logistics at the weapon system platform level may be influencing program offices to obtain access only to technical data necessary to manage the performance-based contract during the acquisition phase—and not to provide a strategy for the future delivery of technical data in case the performance-based arrangement fails. In such a case, the program manager would have limited flexibility in choosing whether to perform maintenance in-house, select an alternative vendor, or offer the work for competition. In order for the department to improve the implementation of performance-based logistics, we recommend that the Secretary of Defense direct that the Under Secretary of Defense (Acquisition, Technology and Logistics) and the Under Secretary of Defense (Comptroller) implement the following two recommendations: 1. Incorporate in DOD’s guidance to the services the private sector’s practice of using performance-based logistics as a tool to achieve economies at the subsystem or component level, rather than as a preferred practice at the platform level. Also, incorporate the private sector’s practice of using it when sufficient performance data are available to establish a meaningful cost baseline and 2. Consider requiring program offices, during weapon system acquisition, to develop acquisition strategies that provide for the future delivery of sufficient technical data to enable the program office to select an alternate source—public or private—or to offer the work out for competition if the performance-based arrangement fails or becomes prohibitively expensive. In commenting on a draft of this report, DOD concurred with our recommendations to enhance the implementation of performance-based logistics. Regarding our recommendation to incorporate in its performance-based logistics guidance to the services the private sector’s practice of using performance-based logistics as a tool to achieve economies at the subsystem or component level, DOD’s response stated that the department recognizes the need to re-emphasize the use of performance-based logistics for subsystems and components in its policy memorandum and guide books. Nonetheless, the response noted that the department believes that it is still prudent to pursue performance-based logistics strategies at the platform level where supported by a business case analysis. The private sector companies we interviewed noted that the more cost effective alternative is to use competitive procedures where practicable at the subsystem or component level supported by a cost analysis using reliable performance data. Regarding our comment that DOD also incorporate in its guidance the private sector practice of using performance-based logistics when sufficient performance data are available to establish a meaningful cost baseline, DOD stated that its policy is that a business case analysis should be performed to help make the determination to use performance-based logistics or traditional logistics support arrangement, and that the business case analysis incorporate the use of performance data, if available, in establishing a meaningful cost baseline. DOD stated that it will emphasize the use of performance data in a revised policy memorandum on performance-based logistics. However, based on information we obtained from the private sector companies we interviewed, developing reliable cost and performance data to support a valid cost analysis at the platform level for a new system will be a challenge and may not be reliable in identifying the most cost-effective support option over the life cycle of the system. As we noted in our report, one company tried a performance- based contract for a new engine but found that because the reliability of the engine was greater than expected, this contract management was not cost-effective. Company officials said they preferred to collect reliable performance data over a period of time to support negotiations for a performance-based contract. In response to our recommendation to consider requiring program offices to develop acquisition strategies that provide for the future delivery of sufficient technical data to select an alternate source—public or private— or to offer the work out for competition if the performance-based arrangement fails or becomes prohibitively expensive, DOD stated that it will take steps to address this issue in the next iteration of the DOD Directive 5000.1 and DOD Instruction 5000.2 acquisition regulation policy. According to the response, the new policy will require the program manager to establish a data management strategy that requires access to the minimum data necessary to sustain the fielded system, recompete or reconstitute sustainment if necessary, promote real time access vice delivery of the data, and provide for the availability of quality data at the point of need for the intended user. According to DOD, for performance- based logistics arrangements, these actions will include acquiring the appropriate technical data to support an exit strategy should the arrangement fail or become too expensive. The objectives of our review were to determine (1) what types of contractor support practices the private sector used to support complex and costly equipment that have life-cycle management issues similar to military weapons systems and (2) what potential lessons could be learned through a comparison of private sector contractor logistics support practices that DOD currently uses, or plans to use, under its implementation of performance-based logistics. To identify commercial industries that use complex and costly equipment with life-cycle management issues similar to military weapon systems, we interviewed DOD depot maintenance and logistics policy officials. We also conducted a literature search to identify appropriate industry groups and interviewed officials from the Industrial College of the Armed Forces, the Aerospace Industries Association, the American Association of Port Authorities, International Council of Cruise Lines, the Society for Mining Metallurgy and Exploration, the Construction Industry Institute, and the Council of Logistics Management to validate and refine the identified industries and to identify appropriate candidate companies within the industry groups. Within the air carrier, maritime shipping, energy exploration, mining, and entertainment industries, we identified over 250 companies and selected 67 companies based on sales/revenues, production rankings, and management awards that might be good candidates for our study. We eliminated three companies that did not outsource significant amounts of logistics support, and 50 companies either did not respond to our initial inquires or declined to participate in the study. Fourteen companies agreed to participate and completed our interviews and follow-up questions. Thirteen of the 14 companies we interviewed agreed to be identified and are listed below by industry group: Air carriers (Continental Airlines, Houston, Texas; Delta Air Lines, Atlanta, Georgia; FedEx Corp., Memphis, Tennessee; Southwest Airlines, Dallas, Texas; and United Airlines, San Francisco, California); Energy Exploration and Mining (British Petroleum, Houston, Texas; Diamond Offshore, Houston, Texas; Phelps Dodge, Phoenix, Arizona; and Vulcan Material, Birmingham, Alabama); Maritime (Carnival Cruises, Miami, Florida; Conoco Philips Polar Tanker, Long Beach, California; and Disney Cruise, Orlando, Florida); and Entertainment (Disney World, Orlando, Florida). To identify private sector support practices, including performance-based logistics, we conducted group discussions with respective company officials responsible for maintenance and support operations, budgeting, and contracting. To collect consistent information among the companies, we developed standard group discussion questions based on our literature search and discussions with industry experts. We also included questions to determine how the companies addressed logistics and contracting issues similar to those that DOD faced in implementing performance- based logistics. We analyzed the responses to identify the prevailing industry practices in supporting complex and costly equipment, especially focusing on the contracting approaches and practices used to outsource support functions and activities. We reviewed and discussed with Office of the Secretary of Defense and military department officials at the headquarters and major acquisition commands the department’s plans, policies, and procedures for using performance-based logistics. We also collected policy and guidance (published and under development) by the Office of the Secretary of Defense as well as the military departments’ policies and implementation plans. To assess what lessons could be drawn from the private sector companies’ experiences to guide DOD’s logistics support efforts, we interviewed DOD officials and reviewed ongoing logistics programs. We assessed the reliability of the projected cost and savings data we used in this report by reviewing supporting documentation and interviewing knowledgeable personnel; and we determined that it was sufficient for our purposes. We compared and contrasted the contract logistic approaches and practices used by private sector activities with those currently used by DOD and envisioned under its plans for implementing performance-based logistics. This comparison included such elements as the (1) use of performance- based contracting and the extent of its application, (2) assigning a single integrator for equipment or weapons system maintenance and logistics support on a platform level, (3) management and oversight including the importance of technical data, and (4) the degree of competitive sourcing and the importance of leveraging purchasing power. As part of our continuing review we are also conducting case studies on DOD performance-based logistics weapon systems to further compare the new DOD approach and practices with those of the private sector. This work is continuing and we expect to complete our final report early in 2005. We performed our work from September 2003 through June 2004 in accordance with generally accepted government auditing standards. We are sending copies of this report to the appropriate congressional committees, and it will be available at no charge on GAO’s Web site at http://www.gao.gov. We are continuing with our review of performance- based logistics in the private sector and in DOD and plan to report the results early in 2005. If you or your staff have any questions on the matters discussed in this letter, please contact me at (202) 512-8412 or solisw@gao.gov or my assistant director, Julia Denman, at (202) 512-4290 or denmanj@gao.gov. Larry Junek, Thom Barger, Pamela Valentine, Judith Collins, and Cheryl Weissman were major contributors to this report. 1. Only 7 of the 14 companies we interviewed use some type of performance-based contracting arrangements. None of the performance-based arrangements in the seven companies using them were at the platform level. The Government Accountability Office, the audit, evaluation and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is through GAO’s Web site (www.gao.gov). Each weekday, GAO posts newly released reports, testimony, and correspondence on its Web site. To have GAO e-mail you a list of newly posted products every afternoon, go to www.gao.gov and select “Subscribe to Updates.”

The Department of Defense (DOD) is pursuing a policy that promotes performance-based logistics at the platform level as the preferred product support strategy for its weapon systems, based in part on DOD's perception that this is an industry best practice. GAO was asked to compare industry practices for activities using complex and costly equipment with life-cycle management issues similar to those of military systems to identify lessons learned that can be useful to DOD. This is the first of two reports addressing DOD's implementation of performance-based logistics and is intended to facilitate DOD's development of new guidance on the use of this approach. DOD's current policy for implementing performance-based logistics as a preferred support approach at the weapon system platform level does not reflect the practices of private-sector companies that support expensive and complex equipment with life-cycle management issues. The companies GAO interviewed use performance-based contracting as a tool rather than as a preferred support concept at the weapon system platform level. While 7 of the 14 companies GAO interviewed use some type of performance-based contracting, they use it at the subsystem or component level--for commodities such as engines, wheels, and brakes--when it is cost-effective and reduces risk in a noncompetitive environment. DOD's proposed policy of pursuing performance-based logistics as the preferred support approach at the platform level results in contracting out the program-integration function--a core process the private-sector firms consider integral to successful business operations. Further, this proposed policy could limit opportunities to take advantage of competition when it is available for subsystems or components as well as limit opportunities to gain purchasing power from volume discounts on components across an entire fleet and avoid the administrative costs charged by a prime integrator. While DOD is proposing the aggressive use of performance-based logistics on both older and new weapon system platforms, the companies GAO interviewed use performance-based contracting at the subsystem or component level when it is cost-effective--often in a noncompetitive environment when the manufacturer controls expensive repair parts, such as engines. In general company officials said they rely more widely on other contracting vehicles, such as time and material contracts, particularly for new systems. Company officials noted that in the absence of accurate and reliable information on system performance to establish a baseline for evaluating the cost-effectiveness of a performance-based contract for new systems, the risk of the negotiated price's being excessive is increased. The companies GAO interviewed also emphasized the importance of having rights to the technical data--such as maintenance drawings, specifications, and tolerances--needed to support the management of all logistics contracts and, should the service provider arrangements fail, to support competition among alternate providers. In contrast, DOD program managers often opt to spend limited acquisition dollars on increased weapon system capability rather than on rights to the technical data--thus limiting their flexibility to perform work in-house or to support alternate source development should contractual arrangements fail.

The IG Act originally established IGs appointed by the President and confirmed by the Senate in 12 major departments and agencies of the government in 1978. Since then, additional IGs have been added through a series of amendments to the IG Act. The Inspector General Act Amendments of 1988 established IGs appointed by their respective entity heads in designated federal entities (DFE) identified by the act with duties and responsibilities similar to those of IGs appointed by the President. DFEs are generally smaller agencies established in various statutes as commissions, boards, authorities, corporations, endowments, foundations, institutions, agencies, and administrations. Prior to the 1988 amendments, both GAO and the President’s Council on Integrity and Efficiency, which preceded the Council of Inspectors General on Integrity and Efficiency (CIGIE), had found that the internal audit offices of small federal agencies lacked independence and provided inadequate coverage of important programs that could benefit from independent oversight by an IG. Additional criteria used by the Congress to determine where to establish these new IG offices included a budget threshold of at least $100 million for the DFEs. Specifically, those agencies with an annual budget of $100 million or greater were considered for inclusion in the 1988 amendments. However, other agencies below this budget threshold were also included for specific reasons. While the IGs in DFEs generally have the same authorities and responsibilities as those established by the 1978 IG Act, there is a clear distinction—they are appointed and removed by their agency heads rather than by the President and are not subject to Senate confirmation. The 1988 amendments established a new category of “federal entity,” which is defined to exclude departments and agencies and DFEs with statutory IGs under the IG Act, as well as judicial and legislative branch entities and others as specified. Further, the 1988 amendments require the Office of Management and Budget (OMB), in consultation with GAO, to annually publish a list of (1) DFEs and, for DFEs that are not boards or commissions, their DFE heads and (2) the federal entities, as that term is defined by the IG Act. OMB’s list of DFEs and federal entities is to be published annually in the Federal Register. The 1988 amendments also require that federal entities, which are defined to exclude entities with a statutory IG under the IG Act, report annually by October 31 to each House of the Congress and to OMB on, among other things, the audit and investigative activities in their respective organizations. GAO-11-770. GAO has long supported the creation of independent IG offices in appropriate federal departments, agencies, and entities, and we continue to believe that significant federal programs and entities should be subject to oversight by independent IGs. At the same time, we have reported some concerns about creating and maintaining small IG offices with limited resources, where an IG might not have the ability to obtain the technical skills and expertise needed to provide adequate and cost- effective oversight. In the final analysis, the determination of whether to place IGs in specific agencies is a policy decision to be decided by the Congress. As a result, we believe there are alternative approaches that the Congress may wish to consider to achieve IG oversight that is appropriate for federal agencies with relatively small budgets and resources. For example, we have recommended, on a case-by-case basis, that specific small agencies could benefit by obtaining IG oversight from another agency’s IG office where the missions of the two agencies are somewhat similar. The following provides examples from our previously issued reports on alternatives suggested for IG oversight of small agencies. Export-Import Bank. In 2001, we were asked to review the need for an IG at the Export-Import Bank, which was defined by OMB as a federal entity under the IG Act, and was not subject to IG oversight. We found that the Export-Import Bank obtained an annual financial audit from an independent public accountant and received additional audits of administrative operations from its internal audit group. We also found that the Export-Import Bank had the largest budget of all other federal entities on OMB’s list at the time, and that it was comparable in size to both departments and agencies with IGs appointed by the President and with DFEs with IGs appointed by the head of the DFE. The alternatives we provided for IG oversight of the Export-Import Bank included (1) establishing a new IG office through an amendment to the IG Act with an IG appointed by either the President or by the Export-Import Bank Chairman of the Board of Directors; (2) designating through legislation an existing IG office to provide oversight, such as the Agency for International Development IG; and (3) implementing a memorandum of understanding, which acts like a contract for outside IG services and would not require an amendment to the IG Act or other legislation. Subsequently, the Congress amended the IG Act in 2002 to establish a statutory IG for the Export-Import Bank, appointed by the President and confirmed by the Senate. Chemical Safety and Hazard Investigation Board (CSB). In 2008, we reported on the responsiveness of CSB to past IG recommendations. We concluded that after 10 years of operations, CSB continued to operate in noncompliance with its statutory mandates by not investigating all accidental chemical releases that involved a fatality, serious injury, or substantial property damage. Since fiscal year 2004, CSB had been obtaining IG oversight services from the Environmental Protection Agency (EPA) IG through a temporary statutory mandate included in its annual appropriation. However, because of the significant issues uncovered by our review, we provided for congressional consideration alternative oversight mechanisms that could be achieved either by amending CSB’s authorizing statute or by amending the IG Act to permanently give the EPA IG the authority to serve as the oversight body for CSB and to provide appropriations and staff allocations specifically for the audit function of CSB through a direct line in the EPA appropriation. Alternatives such as allowing CSB to contract for its own oversight or create an internal audit and investigative unit were not considered as options because of the potential limitations of contracting in terms of both audit independence and the potentially limited duration of the contracting relationship and due to the limited staffing that could reasonably be allocated to an internal oversight function at an agency of its size. The EPA IG has reported continuing oversight efforts at CSB in recent semiannual reports to the Congress. National Mediation Board (NMB). In a recent example, our mandated review of the programs and management practices at NMB concluded in a 2013 report that the board is a small agency, but with a vital role in facilitating labor relations in the nation’s railroads and airlines. We found that NMB’s strategic plan lacked assurance that its limited resources were effectively targeted toward the highest priorities. In addition, NMB lacked certain internal controls that could help achieve results and minimize operational problems. We also concluded that in addition to the periodic oversight by GAO and the annual audits of NMB’s financial statements by independent public accountants, an existing IG office assigned with the responsibility for providing ongoing audits and investigations of NMB and its operations would result in more effective oversight. We provided a matter for congressional consideration, which discussed the authorization of an appropriate federal agency’s IG office to provide independent audit and investigative oversight of NMB. Foreign Affairs Reform and Restructuring Act of 1998, Pub. L. No. 105-277, div. G, § 1314, 112 Stat. 2681-761, 2681-776-77 (Oct. 21, 1998), classified at 22 U.S.C § 6533. Overseas Private Investment Corporation. Finally, the Department of Transportation IG is authorized to provide oversight of the National Transportation Safety Board. Independence is the cornerstone of professional auditing and one of the most important elements of an effective IG function. The IG Act provides specific protections to IG independence that are unprecedented for an audit and investigative function located within the organization being reviewed. These protections are necessary in large part because of the unusual reporting requirements of the IGs, who are subject to the general supervision of their agency heads and are also expected to provide independent reports of their work externally to the Congress. The IG Act provides the IGs with independence by authorizing them to select and employ their own staffs, make such investigations and reports as they deem necessary, and report the results of their work directly to the Congress. In addition, the IG Act provides the IGs with a right of access to information, and prohibits interference with IG audits or investigations by agency personnel. The act further provides the IGs with the duty to inform the Attorney General of suspected violations of federal criminal law. With the growing complexity of the federal government, the severity of the problems it faces, and the fiscal constraints under which it operates, it is important that an independent, objective, and reliable IG structure be in place where appropriate in the federal government to ensure adequate audit and investigative coverage. The IG Act provides each IG with the ability to exercise judgment in the use of independence protections specified in the act; therefore, the ultimate success or failure of an IG office is largely determined by the individual IG placed in that office and that person’s ability to maintain independence both in fact and appearance. The Congress passed the IG Reform Act of 2008 (Reform Act) to further enhance IG independence and accountability. Among other provisions, the Reform Act requires the rate of basic pay of the IGs appointed by the President to be at a specified level, and for the DFE IGs, at or above that of a majority of other senior-level executives at their entities. The Reform Act also requires an IG to obtain legal advice from his or her own counsel or to obtain counsel from another IG office or from CIGIE. Additionally, the act provides a statutory process for handling allegations of wrongdoing by IGs so that such reviews are not done by the same management officials who are subject to IG oversight. The act also requires both the President and the DFE heads to give written reasons to the Congress for removing or transferring an IG at least 30 days prior to the action. The Reform Act also increased the visibility of the IGs’ budgetary resources through the annual budget process. Specifically, the act requires that IG budget requests include certain information and be separately identified in the President’s budget submission to the Congress. In addition, along with the separately identified IG budgets, an IG may include comments with respect to the budget if the amount of the IG budget submitted by the agency or the President would substantially inhibit the IG from performing the duties of the office. These budget provisions are intended to help ensure adequate funding and additional independence of IG budgets by providing the Congress with transparency into the funding of each agency’s IG while not interfering with the agency head’s or the President’s right to formulate and transmit their own budget amounts for the IG. In 2010, the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) amended the IG Act with provisions to enhance the independence of IGs in DFEs with boards or commissions. Specifically, the Dodd-Frank Act changed who would be considered the head of certain DFEs for purposes of IG appointment, general supervision, and reporting under the IG Act. If the DFE has a board or commission, the IG Act now requires each of these IGs to report organizationally to the entire board or commission as the head of the DFE rather than an individual chairman. In addition, the IG Act requires the written concurrence of a two-thirds majority of the board or commission to remove an IG. Prior to this protection, most DFE IGs reported to, and were subject to removal by, the individual serving as head of the DFE. In other past legislative reforms, the Congress has taken actions to convert IGs from appointment by the agency heads to appointment by the President with Senate confirmation as a way to enhance IG independence. For example, on the heels of the savings and loan and banking crisis over two decades ago, the role of the Federal Deposit Insurance Corporation’s (FDIC) IG became increasingly important in providing oversight. Because of the perceived limitation of the FDIC IG’s independence resulting from agency appointment, the Congress converted the IG from agency appointment to appointment by the President with Senate confirmation. In another example, the Congress took action to convert the Tennessee Valley Authority (TVA) IG to appointment by the President with Senate confirmation because of concerns about interference by TVA management. In both cases, Congress recognized that the IG’s independence would be enhanced by the presidential appointment. IGs play a critical role in federal oversight and we believe that all significant federal programs and entities should be subject to oversight by IGs. We have supported the creation of additional IG offices and the enhancements to their independence by past legislation. However, we continue to have some concerns about creating and maintaining IG offices in relatively small federal agencies where it may not be cost- effective to obtain the skills and expertise needed to provide adequate oversight. We believe there are alternatives to creating additional IG offices that can be both effective and less costly. These alternatives for oversight should be decided on a case-by-case basis depending on the critical nature of the small agencies’ missions and the risks identified that require increased oversight. Because the Congress relies on the IGs to provide current information about their respective agencies’ programs and activities, the determination of where and how to provide IG oversight in specific agencies is a policy decision addressed best by the Congress. This concludes my formal statement. Chairman McCaskill, Ranking Member Johnson, and Members of the Subcommittee, I would be pleased to answer any questions that you or the Subcommittee members may have at this time. If you or your staff have any questions about this testimony, please contact me at (202) 512-2623 or davisbh@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this testimony are Jackson Hufnagle (Assistant Director), Lauren S. Fassler, Gregory Marchand, Taya Tasse, and Clarence Whitt. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

The IGs play a key role in federal agency oversight by enhancing government accountability and protecting the government's resources. This includes a strong leadership role in making recommendations to improve the effectiveness and efficiency of government offices and programs at a time when they are needed most. This testimony focuses on (1) the creation of independent IG offices, (2) IG oversight of small agencies, and (3) IG independence and budgetary resources. This testimony provides updates of current IG responsibilities; provisions of the IG Act, as amended; and draws on prior GAO reports and testimonies conducted in accordance with GAO's Quality Assurance Framework. GAO has made numerous observations and provided matters for the Congress to consider in prior reports when addressing IG oversight at small federal agencies and IG independence. The Inspector General Act of 1978, as amended (IG Act), originally established inspectors general (IG) appointed by the President and confirmed by the Senate in 12 major departments and agencies of the government to conduct and supervise independent audits and investigations; recommend policies to promote economy, efficiency, and effectiveness; and prevent and detect fraud and abuse in their departments' and agencies' programs and operations. Based in part on GAO's findings that the internal audit offices of small federal agencies lacked independence and provided inadequate coverage of important programs, the Congress passed the IG Act Amendments of 1988 to establish IGs in designated federal entities (DFE), which are generally smaller agencies established in various statutes as commissions, boards, authorities, corporations, endowments, foundations, institutions, agencies, and administrations identified by the act. The DFE IGs are appointed by their respective entity heads with duties and responsibilities similar to those of IGs appointed by the President. The Congress used a budget threshold of $100 million to help determine which DFEs should have IGs. However, additional DFEs below this threshold were also included for specific reasons. Significant federal programs and agencies should be subject to oversight by independent IGs; however, small IG offices with limited resources might not have the ability to obtain the technical skills and expertise needed to provide adequate, cost-effective oversight. GAO has previously found that alternative approaches exist to achieve IG oversight that may be appropriate for federal agencies with small budgets and few resources. For example, GAO has recommended on a case-by-case basis that specific small agencies could benefit by obtaining IG oversight from another agency's IG office where the missions of the two agencies are somewhat similar. Independence is one of the most important elements of an effective IG function. The IG Act, as amended, provides specific protections to IG independence. The IG Reform Act of 2008 further enhanced the IGs' independence by providing specified pay levels, IG legal counsel, a process for handling allegations of IG wrong-doing, and required notification to the Congress before an IG is removed or transferred. The IG Reform Act also requires the IGs' budget requests to be visible in the budget of the U.S. government submitted by the President to the Congress. Additional provisions to enhance the independence of IGs in DFEs with boards or commissions were included in the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010. Specifically, these IGs are to report organizationally to the entire board or commission rather than a single chairperson. In addition, the IG Act requires a two-thirds majority of the board or commission to remove the IG.

The use of SSNs by government and the private sector has grown over time, in part because of federal requirements. In addition, the growth in computerized records has further increased reliance on SSNs. This growth in use and availability of the SSN is important because SSNs are often one of the “identifiers” of choice among identity thieves. Although no single federal law regulates the use and disclosure of SSNs by governments, when federal government agencies use them, several federal laws limit the use and disclosure of the number. Also, state laws may impose restrictions on SSN use and disclosure, and they vary from state to state. Moreover, some records that contain SSNs are considered part of the public record and, as such, are routinely made available to the public for review. Since the creation of the SSN, the number of federal agencies and others that rely on it has grown beyond the original intended purpose. In 1936, the Social Security Administration (SSA) created a numbering system designed to provide a unique identifier, the SSN, to each individual. The agency uses SSNs to track workers’ earnings and eligibility for Social Security benefits, and as of December 1998, SSA had issued 391 million SSNs. Since the creation of the SSN, other entities in both the public and private sectors have begun using SSNs, in part because of federal requirements. The number of federal agencies and others relying on the SSN as a primary identifier escalated dramatically, in part, because a number of federal laws were passed that authorized or required its use for specific activities. (See appendix I for examples of federal laws that authorize or mandate the collection and use of SSNs.) In addition, private businesses, such as financial institutions and health care service providers, also rely on individuals SSNs. In some cases, they require the SSN to comply with federal laws but, at other times, they routinely choose to use the SSNs to conduct business. In addition, the advent of computerized records further increased reliance on SSNs. Government entities are beginning to make their records electronically available over the Internet. Moreover, the Government Paperwork Elimination Act of 1998 requires that, where practicable, federal agencies provide by 2003 for the option of the electronic maintenance, submission, or disclosure of information. State government agencies have also initiated Web sites to address electronic government initiatives. Moreover, continuing advances in computer technology and the ready availability of computerized data have spurred the growth of new business activities that involve the compilation of vast amounts of personal information about members of the public, including SSNs, that businesses sell. The overall growth in the use of SSNs is important to individual SSN holders because these numbers, along with names and birth certificates, are among the three personal identifiers most often sought by identity thieves. Identity theft is a crime that can affect all Americans. It occurs when an individual steals another individual’s personal identifying information and uses it fraudulently. For example, SSNs and other personal information are used to fraudulently obtain credit cards, open utility accounts, access existing financial accounts, commit bank fraud, file false tax returns, and falsely obtain employment and government benefits. SSNs play an important role in identity theft because they are used as breeder information to create additional false identification documents, such as drivers licenses. Recent statistics collected by federal and consumer reporting agencies indicate that the incidence of identity theft appears to be growing. The Federal Trade Commission (FTC), the agency responsible for tracking identity theft, reports that complaint calls from possible victims of identity theft grew from about 445 calls per week in November 1999, when it began collecting this information, to about 3,000 calls per week by December 2001. However, FTC noted that this increase in calls might also, in part, reflect enhanced consumer awareness. In addition, SSA’s Office of the Inspector General, which operates a fraud hotline, reports that allegations of SSN misuse increased from about 11,000 in fiscal year 1998 to more than 65,200 in fiscal year 2001. However, some of the reported increase may be a result of a growth in the number of staff SSA assigned to field calls to the Fraud Hotline during this period. SSA staff increased from 11 to over 50 during this period, which allowed personnel to answer more calls. Also, officials from two of the three national consumer reporting agencies report an increase in the number of 7 year fraud alerts placed on consumer credit files, which they consider to be reliable indicators of the incidence of identity theft. Finally, it is difficult to determine how many individuals are prosecuted for identity theft because law enforcement entities report that identity theft is almost always a component of other crimes, such as bank fraud or credit card fraud, and may be prosecuted under the statutes covering those crimes. Most often, identity thieves use SSNs belonging to real people rather than making one up; however, on the basis of a review of identify theft reports, victims usually (75 percent of the time) did not know where or how the thieves got their personal information. In the 25 percent of the time when the source was known, the personal information, including SSNs, usually was obtained illegally. In these cases, identity thieves most often gained access to this personal information by taking advantage of an existing relationship with the victim. The next most common means of gaining access were by stealing information from purses, wallets, or the mail. In addition, individuals can also obtain SSNs from their workplace and use them themselves or sell them to others. Finally, SSNs and other identifying information can be obtained legally through Internet sites maintained by both the public and private sectors and from records routinely made available to the public by government entities and courts. Because the sources of identity theft cannot be more accurately pinpointed, it is not possible at this time to determine the extent to which the government’s use of SSNs contributes to this problem as compared to use of SSNs by the private sector. No single federal law regulates the overall use or restricts the disclosure of SSNs by governments; however, a number of laws limit SSN use in specific circumstances. Generally, the federal government’s overall use and disclosure of SSNs are restricted under the Freedom of Information Act and the Privacy Act. The Freedom of Information Act presumes federal government records are available upon formal request, but exempts certain personal information, such as SSNs. The purpose of the Privacy Act, broadly speaking, is to balance the government’s need to maintain information about individuals with the rights of individuals to be protected against unwarranted invasions of their privacy by federal agencies. Also, the Social Security Act Amendments of 1990 provide some limits on disclosure, and these limits apply to state and local governments as well. In addition, a number of federal statutes impose certain restrictions on SSN use and disclosure for specific programs or activities. At the state and county level, each state may have its own statutes addressing the public’s access to government records and privacy matters; therefore, states may vary in terms of the restrictions they impose on SSN use and disclosure. In addition, a number of laws provide protection for sensitive information, such as SSNs, when maintained in computer systems and other government records. Most recently, the Government Information Security Reform provisions of the Fiscal Year 2001 Defense Authorization Act require that federal agencies take specific measures to safeguard computer systems that may contain SSNs. For example, federal agencies must develop an agency-wide information security management program. These laws do not apply to state and local governments; however, in some cases state and local governments have developed their own statutes or put requirements in place to similarly safeguard sensitive information, including SSNs, kept in their computer systems. In addition to the SSNs used by program agencies to provide benefits or services, some records that contain SSNs are considered part of the public record and, as such, are routinely made available to the public for review. This is particularly true at the state and county level. Generally, state law governs whether and under what circumstances these records are made available to the public, and they vary from state to state. They may be made available for a number of reasons. These include the presumption that citizens need government information to assist in oversight and ensure that government is accountable to the people. Certain records maintained by federal, state, and county courts are also routinely made available to the public. In principle, these records are open to aid in preserving the integrity of the judicial process and to enhance the public trust and confidence in the judicial process. At the federal level, access to court documents generally has its grounding in common law and constitutional principles. In some cases, public access is also required by statute, as is the case for papers filed in a bankruptcy proceeding. As with federal courts, requirements regarding access to state and local court records may have a state common law or constitutional basis or may be based on state laws. When federal, state, and county government agencies administer programs that deliver services and benefits to the public, they rely extensively on the SSNs of those receiving the benefits and services. SSNs provide a quick and efficient means of managing records and are used to conduct research and program evaluation. In addition, they are particularly useful when agencies share information with others to verify the eligibility of benefit applicants or to collect outstanding debts. Using SSNs for these purposes can save the government and taxpayers hundreds of millions of dollars each year. As they make this wide use of SSNs, government agencies are taking some steps to safeguard the numbers; however, certain key measures that could help protect SSNs are not uniformly in place at any level of government. First, when requesting SSNs, government agencies are not consistently providing individuals with key information mandated by federal law, such as whether individuals are required to provide their SSNs. Second, although agencies that use SSNs to provide benefits and services are taking steps to safeguard them from improper disclosure, our survey identified potential weaknesses in the security of information systems at all levels of government. Similarly, sometimes government agencies display SSNs on documents not intended for the public, and we found numerous examples of actions taken to limit the presence of SSNs on documents. However, these changes are not systematic and many government agencies continue to display SSNs on a variety of documents. Most of the agencies we surveyed at all levels of government reported using SSNs extensively to administer their programs. As shown in table 1, more agencies reported using SSNs for internal administrative purposes, such as using SSNs to identify, retrieve, and update their records, than for any other purpose. SSNs are so widely used for this purpose, in part, because each number is unique to an individual and does not change, unlike some other personal identifying information, such as names and addresses. Many agencies also use SSNs to share information with other entities to bolster the integrity of the programs they administer. For example, the majority of agencies at all three levels of government reported sharing information containing SSNs for the purpose of verifying an applicant’s eligibility for services or benefits. Agencies use applicants’ SSNs to match the information they provide with information in other data bases, such as other federal benefit paying agencies, state unemployment agencies, the Internal Revenue Service, or employers. As unique identifiers, SSNs help ensure that the agency is matching information on the correct person. Also, some agencies at each level of government reported sharing data containing SSNs to collect debts owed them. Using SSNs for these purposes can save the government and taxpayers hundreds of millions of dollars, such as when SSA matched its data on Supplemental Security Income recipients with state and local correctional facilities to identify prisoners who were no longer eligible for benefits. Doing so helped identify more than $150 million in Supplemental Security Income overpayments and prevented improper payments of more than $170 million over an 8-month period. Finally, SSNs along with other program data, are sometimes used for statistical programs, research, and evaluation, in part because they provide government agencies and others with an effective mechanism for linking data on program participation with data from other sources. When government agencies that administer programs share records containing individuals’ SSNs with other entities, they are most likely to share them with other government agencies. After that, the largest percentage of federal and state program agencies report sharing SSNs with contractors (54 and 39 percent respectively), and a relatively large percentage of county program agencies report sharing with contractors as well (28 percent). Agencies across all levels of government use contractors to help them fulfill their program responsibilities, such as determining eligibility for services and conducting data processing activities. In addition to sharing SSNs with contractors, government agencies also share SSNs with private businesses, such as credit bureaus and insurance companies, as well as debt collection agencies, researchers, and, to a lesser extent, with private investigators. In addition, all government personnel departments we surveyed reported using their employees’ SSNs to fulfill at least some of their responsibilities as employers. Aside from requiring that employers report on their employees’ wages to SSA, federal law also requires that states maintain employers’ reports of newly hired employees identified by SSN. The national database is used by state child support agencies to locate parents who are delinquent in child support payments. In addition, employers responding to our survey said they use SSNs to help them maintain internal records and provide employee benefits. To provide these benefits, employers often share data on employees with other entities, such as health care providers or pension plan administrators. When a government agency requests an individual’s SSN, the individual needs certain information to make an informed decision about whether to provide their SSN to the government agency or not. Accordingly, section 7 of the Privacy Act requires that any federal, state, or local government agency, when requesting an SSN from an individual, provide that individual with three key pieces of information. Government entities must tell individuals whether disclosing their SSNs is mandatory or voluntary; cite the statutory or other authority under which the request is being state what uses government will make of the individual’s SSN. This information, which helps the individual make an informed decision, is the first line of defense against improper use. Although nearly all government entities we surveyed collect and use SSNs for a variety of reasons, many of these entities reported they do not provide individuals the information required under section 7 of the Privacy Act when requesting their SSNs. Federal agencies were more likely to report that they provided the required information to individuals when requesting their SSNs than were states or local government agencies. Even so, federal agencies did not consistently provide this required information; 32 percent did not inform individuals of the statutory authority for requesting the SSN and 21 percent of federal agencies reported that they did not inform individuals of how their SSNs would be used. At the state level, about half of the respondents reported providing individuals with the required information, and at the county level, about 40 percent of the respondents reported doing so. When government agencies collect and use SSNs as an essential component of their operations, they need to take steps to mitigate the risk of individuals gaining unauthorized access to SSNs or making improper disclosure or use of SSNs. Over 90 percent of our survey respondents reported using both hard copy and electronic records containing SSNs when conducting their program activities. When using electronic media, many employ personal computers linked to computer networks to store and process the information they collect. This extensive use of SSNs, as well as the various ways in which SSNs are stored and accessed or shared, increase the risks to individuals’ privacy and make it both important and challenging for agencies to take steps to safeguard these SSNs. No uniform guidelines specify what actions governments should take to safeguard personal information that includes SSNs. However, to gain a better understanding of whether agencies had measures in place to safeguard SSNs, we selected eight commonly used practices found in information security programs, and we surveyed the federal, state, and county programs and agencies on their use of these eight practices. Responses to our survey indicate that agencies that administer programs at all levels of government are taking some steps to safeguard SSNs; however, potential weaknesses exist at all levels. Many survey respondents reported adopting some of the practices; however, none of the eight practices were uniformly adopted at any level of government. In general, when compared to state and county government agencies, a higher percentage of federal agencies reported using most of the eight practices. However, despite the federal government’s self-reported more frequent use of these practices relative to the state and counties, it is important to note that since 1996 we have consistently identified significant information security weaknesses across the federal government. We are not aware of a comparable comprehensive assessments of information security for either state or county government. (For additional information on the eight practices we selected and how they fit into the federal framework for an information security program, see appendix II.) Further, when SSNs are passed from a government agency to another entity, agencies need to take additional steps to continue protections for sensitive personal information that includes SSNs, such as imposing restrictions on the entities to help ensure that the SSNs are safeguarded.Responses to our survey indicate that, when sharing such sensitive information, most agencies reported requiring those receiving personal data to restrict access to and disclosure of records containing SSNs to authorized persons and to keep records in secured locations. However, fewer agencies reported having provisions in place to oversee or enforce compliance with these requirements. In the course of delivering their services or benefits, many government agencies occasionally display SSNs on documents that may be viewed by others, some of whom may not have a need for this personal information. These documents include payroll checks, vouchers for tax credits for childcare, travel orders, and authorization for training outside of the agency. Also, some personnel departments reported displaying employees’ SSNs on their employee badges (27 percent of federal respondents, 5 percent of state, and 9 percent of county). Notably, the Department of Defense (DOD), which has over 2.9 million military and civilian personnel, displays SSNs on its military and civilian identification cards. On the state level, the Department of Criminal Justice in one state, which has about 40,000 employees, displays SSNs on all employee identification cards. According to department officials, some of their employees have taken actions such as taping over their SSNs so that prison inmates and others cannot view this personal information. SSNs are also displayed on documents that are not employee-related. For example, some benefit programs display the SSN on the benefit checks and eligibility cards, and over one-third of federal respondents reported including the SSN on official letters mailed to participants. Further, some state institutions of higher education display students’ SSNs on identification cards. Finally, SSNs are sometimes displayed on business permits that must be posted in public view at an individual’s place of business. In addition to these examples of SSN display, we also identified a number of instances where the Congress or governmental entities have taken or are considering action to reduce the presence of SSNs on documents that may be viewed by others. For example, the DOD commissary stopped requiring SSNs on checks written by members because of concerns about improper use of the SSNs and identity theft. Also, a state comptroller’s office changed its procedures so that it now offers vendors the option of not displaying SSNs on their business permits. Finally, some states have passed laws prohibiting the use of SSNs as a student identification number. These efforts to reduce display suggest a growing awareness that SSNs are private information, and the risk to the individual of placing an SSN on a document that others can see may be greater than the benefit to the agency of using the SSN in this manner. However, despite this growing awareness and the actions cited above, many government agencies continue to display SSNs on a variety of documents that can be seen by others. Regarding public records, many of the state and county agencies responding to our survey reported maintaining records that contain SSNs; however federal program agencies maintain public records less frequently. At the state and county levels, certain offices, such as state licensing agencies and county recorders’ offices, have traditionally been repositories for public records that may contain SSNs. In addition, courts at all three levels of government maintain public records that may contain SSNs. Officials who maintain these records told us their responsibility is to preserve the integrity of the record rather than protect the privacy of the individual SSN holder. However, we found examples of some government entities that are trying innovative approaches to protect the SSNs in such records from public display. Moreover, the general public has traditionally gained access to public records by visiting the office that maintains the records, an inconvenience that represents a practical limitation on the volume of SSNs any one person can collect. However, the growth of electronic record-keeping places new pressures on agencies to provide their data to the pubic on the Internet. Although few entities report currently making public records containing SSNs available on the Internet, several officials told us they are considering expanding the volume and type of such records available on their Web site. This would create new opportunities for gathering SSNs on a broader scale. Again, some entities are considering alternatives to making SSNs available on such a wide scale, while others are not. As shown in table 2, more than two-thirds of the courts, county recorders, and state licensing agencies that reported maintaining public records reported that these records contained SSNs. In addition, some program agencies also reported maintaining public records that contain SSNs. County clerks or recorders (hereinafter referred to as recorders) and certain state agencies often maintain records that contain SSNs because these offices have traditionally been the repository for key information that, among other things, chronicles various life events and other activities of individuals as they interact with government. SSNs appear in these public records for a number of reasons. They may already be a part of a document that is submitted to a recorder for official preservation. For example, military veterans are encouraged to file their discharge papers, which contain SSNs, with their local recorder’s office to establish a readily available record of their military service. Also, documents that record financial transactions, such as tax liens and property settlements, contain SSNs to help identify the correct individual. In other cases, government officials are required by law to collect SSNs. For example, to aid in locating non-custodial parents who are delinquent in their child support payments, the federal Personal Responsibility and Work Opportunity Reconciliation Act of 1996 requires that states have laws in effect to collect SSNs on applications for marriage, professional, and occupational licenses. Moreover, some state laws allow government entities to collect SSNs on voter registries to help avoid duplicate registrations. Although the law requires public entities to collect the SSN as part of these activities, this does not necessarily mean that the SSNs always must be placed on the document that becomes part of the public record. Courts at all three levels of government also collect and maintain records that are routinely made available to the public. Court records overall are presumed to be public; however, each court may have its own rules or practices governing the release of information. As with recorders, SSNs appear in court documents for a variety of reasons. In many cases, SSNs are already a part of documents that are submitted by attorneys or individuals. These documents could be submitted as part of the evidence for a proceeding or could be included in documents, such as a petition for an action, a judgment or a divorce decree. In other cases, courts include SSNs on documents they and other government officials create, such as criminal summonses, arrest warrants, and judgments, to increase the likelihood that the correct individual is affected (i.e. to avoid arresting the wrong John Smith). In some cases federal law requires that SSNs be placed in certain records that courts maintain, such as records pertaining to child support orders, divorce decrees, and paternity determinations. Again, this assists child support enforcement agencies in efforts to help parents collect money that is owed to them. These documents may also be maintained at county clerk or recorders’ offices. When federal, state, or county entities, including courts, maintain public records, they are generally prohibited from altering the formal documents. Officials told us that their primary and mandated interest is in preserving the integrity of the record rather than protecting the privacy of the individual named in the record. Officials told us they believe they have no choice but to accept the documents with the SSNs and fulfill the responsibility of their office by making them available to the general public. When creating public documents or records, such as marriage licenses, some government agencies are trying new innovative approaches that protect SSNs from public display. For example, some have developed alternative types of forms to keep SSNs and other personal information separate from the portion of a document that is accessible to the general public. Changing how the information is captured on the form itself can help solve the dilemma of many county recorders who, because they are the official record keepers of the county, are usually not allowed to alter an original document after it is officially filed in their office. For example, a county recorder told us that Virginia recently changed its marriage license application so that the form is now in triplicate, and the copy that is available to the general public does not contain the SSN. However, an official told us even this seemingly simple change in the format of a document can be challenging because, in some cases, the forms used for certain transactions are prescribed by the state. In addition to these efforts at recorders offices, some courts have made efforts to protect SSNs in documents that the general public can access through court clerk offices. For example, one state court offers the option of filing a separate form containing the SSN that is kept separate from the part of the record that is available for public inspection. These solutions, however, are most effective when the recorder’s office, state agencies, and courts prepare the documents themselves. In those many instances where others file the documents, such as individuals, attorneys, or financial institutions, the receiving agency has less control over what is contained in the document and, in many cases, must accept it as submitted. Officials told us that, in these cases, educating the individuals who submit the documents for the record may help to reduce the appearance of SSNs. This would include individuals, financial institutions, title companies, and attorneys, who could begin by considering whether SSNs are required on the documents they submit. It may be possible to limit the display of SSNs on some of these documents or, where SSNs are deemed necessary to help identify the subject of the documents, it may be possible to truncate the SSN to the last four digits. While the above options are available for public records created after an office institutes changes, fewer options exist to limit the availability of SSNs in records that have already been officially filed or created. One option is redacting or removing SSNs from documents before they are made available to the general public. In our fieldwork, we found instances where departments redact SSNs from copies of documents that are made available to the general public, but these tended to be situations where the volume of records and number of requests were minimal, such as in a small county. Most other officials told us redaction was not a practical alternative for public records their offices maintain. Although redaction would reduce the likelihood of SSNs being released to the general public, we were told it is time-consuming, labor intensive, difficult, and in some cases would require change in law. In documents filed by others outside of the office, SSNs do not appear in a uniform place and could appear many times throughout a document. In these cases, it is a particularly lengthy and labor-intensive process to find and redact SSNs. Moreover, redaction would be less effective in those offices where members of the general public can inspect and copy large numbers of documents without supervision from office staff. In these situations, officials told us that they could change their procedures for documents that they collect in the future, but it would be extremely difficult and expensive to redact SSNs on documents that have already been collected and filed. Traditionally, the public has been able to gain access to SSNs contained in public records by visiting the recorder’s office, state office, or court house; however, the requirement to visit a physical location and request or search for information on a case-by-case basis offers some measure of protection against the widespread collection and use of others’ SSNs from public records. Yet, this limited access to information in public records is not always the case. We found examples where members of the public can obtain easy access to larger volumes of documents containing SSNs. Some offices that maintain public records offer computer terminals on site where individuals can look up electronic files from a site-specific database. In one of the offices we visited, documents containing SSNs that were otherwise accessible to the public were also made available in bulk to certain groups. When asked about sharing information containing SSNs with other entities, a higher percentage of county recorders reported sharing information containing SSNs with marketing companies, collection agencies, credit bureaus, private investigators, and outside researchers. Finally, few agencies reported that they place records containing SSNs on their Internet sites; however, this practice may be growing. Of those agencies that reported having public records containing SSNs, only 3 percent of the state respondents and 9 percent of the county respondents reported that the public can access these documents on their Web site. In some cases, such as the federal courts, documents containing SSNs are available on the Internet only to paid subscribers. However, increasing numbers of departments are moving toward placing more information on the Internet. We spoke with several officials that described their goals for having records available electronically within the next few years. Providing this easy access of records potentially could increase the opportunity to obtain records that contain SSNs that otherwise would not have been obtained by visiting the government agency. While planning to place more information on the Internet, some courts and government agencies are examining their policies to decide whether SSNs should be made available on documents on their Web sites. In our fieldwork, we heard many discussions of this issue, which is particularly problematic for courts and recorders, who have a responsibility to make large volumes of documents accessible to the general public. On the one hand, officials told us placing their records on the Internet would simply facilitate the general public’s ability to access the information. On the other hand, officials expressed concern that placing documents on the Internet would remove the natural deterrent of having to travel to the courthouse or recorder’s office to obtain personal information on individuals. Again, we found examples where government entities are searching for ways to strike a balance. For example, the Judicial Conference of the United States recently released a statement on electronic case file availability and Internet use in federal courts. They recommended that documents in civil cases and bankruptcy cases should be made available electronically, but SSNs contained in the documents should be truncated to the last four digits. Also, we spoke to one county recorder’s office that had recently put many of its documents on their Web site, but had decided not to include categories of documents that were known to contain SSNs. In addition, some states are taking action to limit the display of SSNs on the Internet. Given the likely growth of public information on the Internet, the time is right for some kind of forethought about the inherent risk posed by making SSNs and other personal information available through this venue. SSNs are widely used in all levels of government and play a central role in how government entities conduct their business. As unique identifiers, SSNs are used to help make record-keeping more efficient and are most useful when government entities share information about individuals with others outside their organization. The various benefits from sharing data help ensure that government agencies fulfill their mission and meet their obligation to the taxpayer by, for example, making sure that the programs serve only those eligible for services. However, the gaps in safeguarding SSNs that we have identified create the potential for SSN misuse. Although the extent to which the government’s broad use of SSNs contributes to identity theft is not clear, measures to encourage governments to better secure and reduce the display of SSNs could at least help minimize the risk of SSN misuse. It is important to focus on ways to accomplish this. We will be reporting in more detail on these issues at the end of this month and look forward to exploring additional options to better protect SSNs with you as we complete our work. For further information regarding this testimony, please contact Barbara D. Bovbjerg, Director, or Kay E. Brown, Assistant Director, Education, Workforce, and Income Security at (202) 512-7215. Individuals making key contributions to this testimony include Lindsay Bach, Jeff Bernstein, Richard Burkard, Jacqueline Harpp, Daniel Hoy, Raun Lazier, Vernette Shaw, Jacquelyn Stewart, and Anne Welch.

The Social Security numbers (SSN), originally created in 1936 to track workers' earnings and eligibility for Social Security benefits is now used for many other purposes by both government and private sectors. The growth in electronic record keeping and the availability of information over the Internet, combined with the rise in identity theft, have heightened public concern about how their SSNs are being used. Federal agencies use SSNs to manage records, verify the eligibility of benefit applicants, collect outstanding debts, and do research and program evaluation. GAO found that federal laws designed to protect SSNs are not being followed consistently, Moreover, courts at all levels of government and offices at the state and county level maintain records that contain SSNs for the purpose of making these records available to the public. Recognizing that these SSNs may be misused, some government entities have taken steps to protect the SSNs from public display. At the same time, however, some government entities are considering making more public records available on the Intranet. Ease of access to electronically available files could encourage more information gathering from public records on a broader scale than possible previously.

Agencies communicate with the public for various reasons. In September 2016, we reported on categories of advertising and public relations activities that agencies may engage in, including: Public education and awareness – providing information on public health and safety issues, informing the public of its rights and entitlements, discouraging harmful or dangerous behavior; Customer service – providing information to users of agency General information – keeping the public informed of agency Recruitment – the process of attracting qualified applicants to apply Compliance with laws and policies – making information available in order to comply with statutes, executive orders, policies, and procedures. There are no single, commonly-accepted definitions of the terms “advertising,” “public relations,” or “public affairs.” Advertising. Although federal guidance does not define “advertising” for purposes of classifying contracts, the Office of Management and Budget (OMB) defines the term in relation to determining costs for grants and other federal awards. OMB Circular A-87 defines “advertising costs” as “the costs of advertising media and corollary administrative costs,” and “advertising media” as including magazines, newspapers, radio and television, direct mail, exhibits, and electronic or computer transmittals. Public relations. OMB Circular A-87 defines “public relations” as “community relations and those activities dedicated to maintaining the image of the governmental unit or maintaining or promoting understanding and favorable relations with the community or public at large or any segment of the public.” Public affairs. We have stated that “public affairs” involves “efforts to develop and disseminate information to the public to explain the activities of and the issues facing an organization.” While agencies have flexibility to use a variety of approaches to communicate with and reach out to the public, there are some prohibitions on public relations activities. For example, appropriations acts and other provisions of law prohibit the use of appropriated funds for certain types of communications. Since 1951, appropriations acts have included provisions prohibiting agencies from using appropriations for unauthorized “publicity or propaganda.” While Congress has not defined the meaning of such publicity or propaganda, we have recognized three types of activities that violate these prohibitions: (1) self-aggrandizement, (2) covert propaganda (defined as “communications such as editorials or other articles prepared by an agency or its contractors at the behest of the agency and circulated as the ostensible position of parties outside of the agency”), and (3) materials that are purely partisan in nature. Appropriations acts have also typically included provisions prohibiting the use of federal funding for certain grassroots lobbying. Typical language includes a prohibition against using appropriated funds for communications designed to support or defeat legislation pending before Congress, except in presentation to Congress itself. We have determined that in order for a violation to occur, there must be evidence of a clear appeal by the agency to the public to contact members of Congress in support of, or in opposition to, pending legislation. The four agencies we reviewed were selected, in part, for their different missions and interaction with the public. Table 1 summarizes each agency’s mission. We previously reported that government-wide annual obligations for advertising and public relations contracts averaged close to $1 billion over fiscal years 2006 through 2015, with the Department of Defense responsible for about 60 percent of obligations for these contracts over the 10-year period. The four selected agencies obligated on average a combined amount of $19 million annually over the past 10 years for advertising and public relations contracts, though annual obligations amounts vary substantially from year to year. As shown in table 2, these agencies obligated annual amounts ranging from -$0.3 million (reflecting deobligations) to nearly $18 million. The four agencies employed from 7 to 114 public affairs employees as of the end of fiscal year 2016, as shown in table 3. We previously reported that the government-wide number of full-time permanent public affairs employees was about 5,000 in fiscal year 2014, with about 42 percent of these employees at the Department of Defense. At each of the agencies we reviewed, obligations were concentrated in a small number of advertising and public relations contracts. The activities supported by these contracts ranged from promoting agency initiatives to providing communications support services. Appendix II provides details on each agency’s largest advertising and public relations contracts (in terms of obligations) over fiscal years 2012 through 2016. CFPB. CFPB’s four largest advertising and public relations contracts represented nearly 70 percent of obligations for these contracts over fiscal years 2012 through 2016 ($22.6 million out of $32.8 million). These four contracts all focused on increasing general awareness of the tools and resources CFPB offers to the public to promote better financial decision making. Contract documentation linked the need for increasing general awareness of CFPB’s tools and resources to CFPB’s statutory responsibility to develop and implement initiatives intended to educate and empower consumers to make better financial decisions. According to CFPB officials, promoting awareness of CFPB’s tools and the services it provides is intended to increase their use. In particular, research conducted by a CFPB contractor in 2015 showed that before CFPB’s campaign to increase awareness of its tools and resources, awareness and use of these tools and resources had remained level over the prior two and a half years. Figure 1 shows an example of CFPB outreach promoting its tools and resources. FEMA. The largest of FEMA’s advertising and public relations contracts represented just over 40 percent of all obligations for these contracts over fiscal years 2012 through 2016 ($8.7 million out of $20.7 million). This contract, and several others, supported the National Flood Insurance Program (NFIP). According to FEMA officials, part of its role in administering the NFIP involves public communications and advertising to encourage people to buy flood insurance to better prepare themselves for disaster. FEMA’s other high-value contracts included ones for services promoting emergency preparedness, including those supporting the Ready campaign. The Ready campaign promotes preparedness for many types of emergencies, including hurricanes, extreme heat, and tornadoes. Figure 2 illustrates Ready campaign outreach related to extreme weather alerts. NASA. NASA’s two largest advertising and public relations contracts represented just under half of all obligations for these contracts over fiscal years 2012 through 2016 ($7.6 million of $17.1 million). These two contracts, and other lesser-value ones, supported NASA’s Communications Support Services Center. This center provides support services across the agency, such as developing products that support the agency’s education and public outreach programs. According to NASA officials, the contracts are for services that are less efficient for NASA staff to perform or for which staff do not have the needed expertise. Figure 3 shows an example of NASA outreach illustrating a program to grow plants in space. USCIS. USCIS’ two largest advertising and public relations contracts represented just over 90 percent of obligations for these services over fiscal years 2012 through 2016 ($18.1 million out of $19.8 million).These two contracts supported planning and developing media and educational messaging tools for the E-Verify and Systematic Alien Verification for Entitlements (SAVE) programs. E-Verify is an Internet-based system that allows businesses to confirm the eligibility of their employees to work in the United States. SAVE is a verification service for benefit-granting agencies to verify applicants’ immigration or citizenship status. According to contract documentation, outreach on these programs aims to, among other things, increase understanding and use of these services. Figure 4 shows an example of USCIS outreach promoting a virtual assistant, which is one of the ways the agency interacts with those using or interested in its services. More broadly, most of the 68 contracts we reviewed focused on public education and awareness, general information, or both. Table 4 shows the frequency of the types of general activities supported by the contracts we reviewed at each agency. The contracts we reviewed also involved different types of tasks, including analysis, content creation, monitoring, distribution, planning, or technical or operational support. For example, planning activities include creating a media strategy with new or existing resources or conducting market research. Appendix II provides a description of these tasks and the frequencies with which they were included in contracts. Our review of position descriptions for public affairs employees across the selected agencies showed that these employees’ duties also primarily supported educating and providing information to the public. Officials at the agencies we reviewed supported this by most commonly citing public education and awareness, followed by general information, as public affairs staff’s primary activities. Of the 13 position descriptions we reviewed at various levels of responsibility, duties included: writing, editing, and analysis; communication planning and evaluation; disseminating information through various channels, including news media (wire services, radio, television, newspapers, and magazines), agency websites, and social media accounts; engaging with stakeholders, including those within and outside of managing relationships with media entities. In some cases, public affairs staff are involved in activities also supported by contracted services to provide information to the public. For example, FEMA’s Ready campaign involved contributions from both public affairs staff and a contractor’s staff. According to FEMA officials, public affairs staff responsibilities for the campaign included developing messages, responding to questions from the public, representing the campaign at outreach events, and tracking campaign efforts. The contractor’s staff, on the other hand, provided coding for the website, developed public service announcements, and created social media graphics and videos. Officials told us that the decision to contract a service instead of using public affairs staff, or the level of involvement of public affairs staff, depends on the type of work and who is able to provide the service more efficiently. For example, a USCIS public affairs official told us that the agency contracts for services related to outreach in situations where contractual services will save time and money or result in higher quality, such as providing a daily delivery of news clippings and media references to the agency, and maintaining a media contact database that agency staff use for outreach. Additionally, a NASA public affairs official told us that the agency contracts for services that employees do not have the expertise to do. For example, NASA headquarters has a photo office that is staffed by contractors, and contractors operate the cameras for NASA’s television station. In both cases, NASA officials told us they monitor the contractors. Advertising and public relations contracts do not necessarily capture all of the advertising and public relations activities carried out by the four case study agencies. As we reported previously, there are other product service codes aside from those related to advertising and public relations services that can encompass such activities. We did not focus on them because they may also include activities that are not related to public relations. Additionally, there is an element of subjectivity involved in selecting the appropriate product service code for a contract that may lead different people to appropriately select different codes for similar services. For example, three of the four agencies we reviewed had contracts for services to monitor media for references to them or their activities. Agencies coded these contracts as either advertising or public relations services. Alternatively, USCIS coded a contract for a digital broadcast monitoring service as “IT and Telecom - Other IT and Telecommunications.” USCIS officials said that this coding was because the monitoring related to digital broadcasting. Similarly, staff aside from those classified under the public affairs occupational series may also be involved in public affairs activities. For example, the agencies we reviewed used other staff to support public affairs activities by performing tasks such as translating material into other languages, editing materials to ensure that they meet standards for readability, and developing and maintaining information technologies that support outreach. Additionally, officials at both FEMA and NASA told us that public engagement is an integral part of their agencies’ activities, and that one could consider all of their employees to be involved in public affairs to some extent. The amounts the agencies we reviewed obligated to contracts coded as advertising or public relations services varied from year to year, though some agencies’ obligations were more stable. As shown in figure 5, CFPB and FEMA had the most significant changes over the last 10 years. According to CFPB officials, the increase in CFPB obligations over time is due to standing up its operations. Officials told us that during fiscal years 2013 through 2015, they piloted campaigns and conducted research on outreach. The results of these pilot campaigns and research informed later outreach activities. Increases in fiscal years 2015 and 2016 were due primarily to two contracts that focused on increasing awareness of CFPB and the tools it offers consumers. CFPB officials told us that fiscal year 2016 obligations reflect full operations, and they expect future years’ obligations to be more consistent. The changes in FEMA obligations over the past decade are due in large part to changes in three contracts related to supporting outreach and customer support services for the agency’s National Flood Insurance Program (NFIP). For example, in fiscal year 2010, most of the agency’s obligations for advertising and public relations services (about $12.5 million out of a total of $16.2 million) were for an integrated marketing, advertising, and public relations services contract for the NFIP. In fiscal year 2011, the amount obligated to that contract dropped to about $1 million. In fiscal year 2016, FEMA obligated about $8.7 million for another contract supporting the NFIP, which drove the sharp increase in obligations in that year. Officials said that the amounts obligated for outreach for the NFIP depend on the program’s other priorities, such as flood hazard mapping, which FEMA does to assess flood risks and uses to inform the development of NFIP regulations and flood insurance requirements. At NASA and USCIS, obligations were relatively stable over the past decade. NASA officials said they expected obligations to remain consistent or even decline in the future, in anticipation of declining resources. USCIS officials also said they expect obligations to remain relatively level in the future, though they may increase marketing activities if E-Verify is mandated nationwide. Employment of public affairs employees at the agencies we reviewed increased over the past decade, but was relatively stable at NASA. Table 5 shows the changes in public affairs employees at these agencies from 2007 through 2016. CFPB, FEMA, and USCIS officials identified several factors that caused the increase in the number of public affairs staff they use, including changes in operations and staffing structure. CFPB officials told us that the increase from five to seven public affairs employees between fiscal years 2011 and 2016 was due to standing up operations from fiscal year 2011. According to officials, the number of public affairs staff in fiscal year 2016 represents the steady state of operations for these employees at CFPB. FEMA officials cited organizational and structural changes, the rise of digital media, increasing stakeholder engagement, and an increase in work volume and duties as the reasons for its increase in public affairs employment. For example, in fiscal year 2007, the Protection and National Preparedness Directorate, which included public affairs staff, moved to FEMA. A USCIS public affairs official attributed the increase in the number of public affairs staff to a focus on having a greater number of lower- level public affairs staff versus a smaller number of higher level staff. They said they have found this staffing model to be effective. A NASA public affairs official said that the stability in the agency’s number of public affairs staff was due to the end of the space shuttle program in 2011, which reduced the public affairs workload, balanced by an increased focus on commercial cargo and crew flight, which in turn has increased the public affairs workload. The official told us he expects staffing to either remain stable or decline in future years as the agency manages resource constraints. Officials at three of the four agencies we reviewed identified increased use of digital media as a significant change over the past decade, and at all four agencies said that these platforms are their primary methods of outreach. While the increased popularity of these media represented a change for FEMA, NASA, and USCIS, officials at CFPB, which began operations in 2011, said that they have used digital media platforms since its beginning. All four agencies had websites and presences on multiple social media platforms, such as Facebook and Twitter. According to officials at the agencies we reviewed, the use of digital media has increased the reach of agency communications and changed the nature of these agencies’ interactions with the public. FEMA officials emphasized the usefulness of these media in obtaining information from the public by gaining awareness of conditions during disasters. For example, FEMA guidance described how during Hurricane Sandy public affairs employees were able to use social media to obtain information about power outages, volunteering and donations, and concerns about the response efforts. A USCIS public affairs official also said that digital media have improved and changed the agency’s reach. For example, USCIS has been working to provide customer service online by doing things like holding “office hours” on Twitter, during which Twitter users can post questions, which are answered by USCIS staff (see figure 6). The increased use of digital media has had mixed effects on resources at the agencies we reviewed. FEMA officials told us that they increased the number of public affairs staff in response to the increase and evolution of digital media and the public’s reliance on it during major disasters. For example, officials said during the Louisiana Floods in 2016, monthly engagement (shares, reactions, and comments) on the agency’s Facebook account increased to 49,000, compared to an average of 7,000 (see figure 7). NASA and USCIS officials said they have generally been able to adjust to the increased use of digital media using existing resources. For example, NASA officials told us that they focused on hiring new public affairs staff with expertise in these areas when hiring to fill open public affairs positions. They also trained public affairs staff in digital media skills and modified contracts to include more digital media tasks in place of more traditional ones. Although the agencies we reviewed used digital media to a large extent, they all continue to use more traditional media such as newspapers and radio. FEMA officials told us that some populations they are trying to reach do not have access to digital media, so they use a mix of media. For example, they have found that radio is an effective medium for reaching out to members of tribal nations. Officials also made the point that other media channels are important because, during a disaster, people may not be able to access digital media due to power loss or other connection failures. All of the agencies we reviewed identified performance indicators they use to assess the performance of outreach activities, including those supported by advertising and public relations contracts and public affairs staff. However, the type and extent of assessment depends on the types of outreach conducted. For example, all four agencies use web-based indicators when assessing digital outreach, including use of agency websites and social media accounts. Officials told us that digital media are well suited to performance measurement and offer richer analytical possibilities than more traditional media. Outreach types and related indicators are described below. Digital media: Agencies use web-based indicators when assessing digital outreach, including indicators of the number of people reached, such as number of visits to a website and click-through rates (a ratio showing how often people who see a digital advertisement end up clicking on it). Agencies also use indicators related to how engaged users are with the outreach materials, including bounce rates (the percentage of visitors to a particular website who navigate away from the site after viewing only one page), the length of time a user spends on a web page or watching a video, and the number of people who respond to or share a social media post. When assessing digital outreach, agency officials told us that they also consider data on the proportion of users who access websites from mobile versus desktop computers and the geographic location of users. Traditional media: Agencies use other indicators to evaluate the performance of outreach conducted through traditional media such as newspapers, radio, and television. For newspapers, these indicators included circulation and number of readers per copy. For radio and television, they included the number of listeners or viewers. Other types of outreach: In some cases, agencies designed indicators specific to more targeted outreach efforts. For example, CFPB has a program that involves working with libraries to provide websites, worksheets, guides, and other information to help with consumers’ financial decisions. CFPB directed a contractor supporting this program to assess this outreach through pilot testing of a guidebook and job aids with partner libraries, and to capture and document libraries’ feedback and consolidate recommendations to CFPB. CFPB also used focus groups, field input from stakeholders, and surveys to inform outreach efforts. Officials told us that obtaining qualitative information through such efforts requires more resources, but this information is extremely useful in assessing performance of CFPB outreach activities. In many cases, the agencies we reviewed identified quantitative and qualitative goals for their contract and employee activities. For example, one CFPB contract we reviewed included a goal of reaching 20 percent of the target audience (defined as adults ages 30–44 with household incomes between $35,000 and $125,000) at a frequency of five times per month. In another case, FEMA set a qualitative goal for a contract supporting the Ready campaign. The objectives of the contract included encouraging state and local governments to create localized efforts to encourage emergency preparedness. The agencies we reviewed all used performance information to assess services provided through advertising and public relations services contracts. The majority of the contracts we reviewed (43 out of 68) explicitly included an analysis component, which involved such actions as developing indicators, collecting quantitative data, or analyzing and reporting the effectiveness of outreach efforts. With the exception of NASA, each agency used contracts and public affairs staff to support a major outreach initiative. These agencies also assigned to contractors responsibility for supporting performance assessment of those initiatives. For example, USCIS contracts supporting education and outreach initiatives for the E-Verify and SAVE programs directed the contractors to develop performance indicators to analyze and assess the results of outreach efforts. Appendix II provides examples of how the agencies we reviewed assessed the performance of selected major initiatives. Agencies also assign responsibility to public affairs employees to assess the outreach activities they support. Position descriptions for public affairs staff at all four of the agencies we reviewed specifically included assessment of agency outreach activities. For example, a NASA public affairs official told us that public affairs staff are involved in reviewing data on how many people read certain features and analyzing the effects of releasing new information, such as photos taken from spacecraft. The agency’s Internet manager, who is classified under the public affairs occupational series, supervises the review of analytic information. Officials at the agencies we reviewed told us that they use performance information to inform decisions about outreach activities. For example: FEMA officials told us that they use the results of emergency preparedness surveys to inform the focus of future advertising for the Ready campaign. In response to people reporting lack of time as a barrier to discussing emergency preparedness with family members, FEMA developed messages for its Ready initiative focused on working towards emergency preparedness while doing other daily tasks, such as driving to school or eating dinner. USCIS used a report summarizing lessons learned from fiscal year 2016 outreach about the E-Verify program to inform fiscal year 2017 outreach efforts. The report included an assessment of which efforts in fiscal year 2016 most successfully supported objectives such as building awareness of the program. It included performance indicators, such as the number of times an online advertisement is shown on a search result page or other site (impressions), and the number of times a person clicks on an on-line advertisement (clicks). The summary report recommended approaches for fiscal year 2017 based on lessons learned from fiscal year 2016 performance. Despite the usefulness of web-based indicators and other tools used to measure outreach activities, officials at the agencies we reviewed acknowledged some challenges in using them. The following are examples of challenges officials at our case study agencies identified: Lack of qualitative feedback: Officials at USCIS and CFPB told us that while they have access to several indicators related to performance of outreach (for example, number of visitors to a site or downloads of materials), these indicators are not the same as understanding whether and how information is being used. A USCIS public affairs official said that the agency has this challenge with outreach through both digital and traditional media. For example, USCIS staff have information on the circulation of a newspaper in which they have placed an advertisement, but without additional assessment, it is difficult for staff to know how or if someone used information in the advertisement or had a favorable impression of it. Difficulty measuring long-term effects: NASA officials told us that digital media offers the ability to get virtually immediate feedback on indicators such as the number of people reached, and that such information helps inform decisions and ensure they make sound investments. However, officials noted that it is more difficult to determine the long-term effects of outreach activities. For example, it would be difficult for NASA to determine whether its outreach aimed at schoolchildren led them to eventually pursue careers in science, technology, engineering, and math fields. Difficulty identifying factors that influence performance: FEMA officials told us that there are many variables that affect performance of digital media outreach, including the time of day or week, the specific content of the message, and other news topics or marketing campaigns occurring at the same time. They said they use the data to make decisions, but in some instances it is difficult to determine why outreach did not perform well. In our prior work on advertising at the Department of Defense, we reported similar challenges in measuring the impact of advertising on recruitment. We stated that determining the precise impact of advertising on outcomes in this case is inherently challenging, in part due to concurrent effects of external factors, such as the influence of family support and the availability of other career or educational activities. Difficulty performing in-depth assessments: Tools that allow for more in-depth assessments of outreach activities may require more resources than collecting web-based or other indicators does. For example, under the Paperwork Reduction Act (PRA), agencies must receive approval from OMB for surveys or other efforts that involve collecting data from the public. CFPB officials told us the approval process for collecting information from the public can be challenging because the outreach initiative that is the subject of the information collection often changes or may evolve in the time it takes to receive approval, which can take up to a year. In our 2010 report on opportunities to strengthen agencies’ customer service efforts, we reported that in certain circumstances the PRA clearance process made obtaining customer input difficult because of the time it takes to obtain approval for surveys to collect customer input. In our 2014 report on customer service at selected agencies, we reported that use of the Fast Track process designed to speed OMB survey approval had varied among the agencies we reviewed and not led to significant improvements. Selected agencies have taken steps to address these challenges. For example, FEMA and CFPB have used focus groups or surveys to obtain richer information on whether and how people are using their outreach material. FEMA works with a contractor to administer surveys that provide data on how the public may be responding to the agency’s outreach, such as the extent to which people have taken action to prepare for emergencies. Both agencies have worked with OMB to obtain approval for their information collections, despite the resources involved. CFPB officials told us the information obtained through these collections makes it worth the time and resources involved in the approval process. We provided a draft of this report to CFPB, DHS, and NASA for comment. CFPB and DHS provided technical comments, which we incorporated as appropriate. NASA did not have comments. As agreed with your staff, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the Director of the Consumer Financial Protection Bureau, Acting Secretary of Homeland Security, Acting Administrator of the National Aeronautics and Space Administration, and interested congressional committees. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6806 or krauseh@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Our objectives were to review (1) the activities that selected agencies conducted using advertising and public relations contracts and agency public affairs employees, and their purposes; (2) how the level of resources selected agencies devoted to these activities changed over the past decade and the factors that officials identified as affecting these changes; and (3) how selected agencies measured the results of these activities. To address our objectives, we selected four agencies for case study based on several factors. First, we focused on agencies with high total obligations for advertising and public relations contracts or high total numbers of public affairs employees relative to other agencies with similar characteristics, and/ or large changes in these numbers in recent years relative to other agencies with similar organizational structures (e.g. comparing component agencies). Of those agencies, we selected ones with differing missions and types of interactions with the public. We also considered input on the suitability of agencies for case illustration purposes from our staff with expertise in agencies’ operations. The four selected agencies were the Consumer Financial Protection Bureau, Federal Emergency Management Agency, National Aeronautics and Space Administration, and U.S. Citizenship and Immigration Services. We previously reported that the Department of Defense (DOD) obligates more funding to advertising and public relations contracts, and employs more public affairs staff than any other agency. We did not include DOD in this review because we have recent work that examined advertising at that department. To identify the activities selected agencies conduct using advertising and public relations contracts and the purposes of these activities, we reviewed information on these contracts. Specifically, we used the Federal Procurement Data System-Next Generation (FPDS-NG) database to identify contracts at the selected agencies with values of $150,000 or more that were classified as advertising or public relations services over fiscal years 2012 through 2016. The FPDS-NG database captures information on the federal government’s contract awards and obligations. It includes data for most federal contracts that have an estimated value of $3,000 or more. The four agencies we reviewed had a combined total of 68 contracts with these characteristics. These 68 contracts represented just over a quarter (68 out of 253) and almost 95 percent ($90.7 million of the $97.4 million) of all advertising and public relations contracts and related obligations at these agencies over this time period. To illustrate the types of general activities the 68 contracts supported, we analyzed statements of work and other documentation using a data collection instrument (DCI).We used the DCI to categorize key characteristics of the services described in the contract documentation, such as the purpose of the activities, media channels for each of the activities, the intended audiences of the outreach, and any references to agency statutes or missions. We also interviewed agency contracting and public affairs officials about activities supported by contracts. We also classified the contracts according to five categories that we identified in our earlier work: (1) public education and awareness; (2) customer service; (3) general information; (4) recruitment; and (5) compliance with laws and policies. To conduct our analysis, one analyst used the DCI to review and code each contract. Another analyst verified the coding. Finally, we shared our classification according to the five categories with agency public affairs officials to ensure concurrence. We assessed the reliability of FPDS-NG data by considering known strengths and weaknesses of the data based on our past work that used the database, and by comparing FPDS-NG data to information in contracts provided by agencies. We also reviewed a nongeneralizable sample of 18 contracts not coded as advertising or public relations services, but that appeared to include some activities related to these services. We identified these other contracts by searching FPDS-NG for contracts with (1) vendors that had received other contracts for advertising and public relations services, and (2) contract descriptions that used the terms “advertising” or “public relations.” We determined that these data were sufficiently reliable for our purposes. To identify activities supported by public affairs staff, and their purposes, we analyzed employment data provided by agencies on employees classified under the Office of Personnel Management’s (OPM) public affairs occupational series. We also reviewed position descriptions and other documents provided by agencies and interviewed agency public affairs officials who manage public relations activities to describe the organization and role of public affairs employees in public relations and advertising activities, and the purposes of those activities. We assessed the reliability of agency employment data by comparing it to OPM’s Enterprise Human Resource Integration database—the primary government-wide source for information on federal employees—and determined that they were sufficiently reliable for our purposes. To review how the level of contract and staff resources at selected agencies has changed over the past decade and factors affecting these changes, we reviewed FPDS-NG data, including the total obligations of public relations and advertising contracts from fiscal years 2007 to 2016. We also analyzed employment data provided by agencies on the numbers of full-time permanent public affairs staff over fiscal years 2007 through 2016. We interviewed agency public affairs officials to discuss reasons for any changes, as well as descriptions of changes in the types of work performed. To review how selected agencies measure the results of activities supported by advertising and public relations contracts and public affairs staff, we reviewed agency performance information, including agency performance reports and reports describing specific outreach activities and campaigns. We also examined whether and how the contracts we reviewed involved performance measurement. We reviewed this information and interviewed agency public affairs officials to identify methods that agencies use to measure the effects of activities supported by these contracts and staff, any challenges that the agencies have with measuring their effects, and ways that agencies incorporate performance information into decision making. We conducted this performance audit from August 2016 to September 2017 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. At the agencies we reviewed, a relatively small number of advertising and public relations contracts represented a large portion of total obligations to these contracts. The following figures and notes illustrate and describe each agency’s largest contracts over fiscal years 2012 through 2016. Figure 8 shows Consumer Financial Protection Bureau (CFPB) obligations to its five largest and all other advertising and public relations contracts (“calls” or “task orders”) over fiscal years 2012 through 2016. CFPB had a total of 44 advertising and public relations contracts (“calls” or “task orders”) with combined obligations of $32.8 million over this time period. There may be multiple calls or task orders for a single contract. Figure 9 shows Federal Emergency Management Agency (FEMA) obligations to its five largest and all other advertising and public relations contracts over fiscal years 2012 through 2016. FEMA had a total of 148 of these contracts with a combined value of $20.7 million over this time period. National Aeronautics and Space Administration Figure 10 shows National Aeronautics and Space Administration (NASA) obligations to its five largest and all other advertising and public relations contracts over fiscal years 2012 through 2016. NASA had a total of 55 of these contracts with a combined value of $17.1 million over this time period. U.S. Citizenship and Immigration Services Figure 11 shows U.S. Citizenship and Immigration Services (USCIS) obligations to the four advertising and public relations contracts we reviewed. The agency had two other of these contracts over fiscal years 2012 through 2016, with a value of about $27,000 and -$100,000. Because the net value of these contracts is negative, we did not represent them in the figure. The contracts we reviewed involved different types of tasks. Table 6 lists and describes these tasks and provides examples from the agencies. Note that some contracts fell into two or more categories, as contracts may involve multiple types of services. Five of the six tasks (all except monitoring) were included in a majority of the contracts, and the three most frequent tasks were the three most closely related to distributing information (planning, content creation, and distribution). Planning activities often included conducting market research (of the 46 contracts that included planning, 31 included market research). Agencies asked contractors to create content for a variety of mediums. Of the 50 contracts that involved content creation, the most common tasks included were related to writing/editorial (37), graphic design (30), and video production (29), and the most common media were web/digital (38), and print (30). Agencies tasked contractors with distributing the information or buying media on a variety of platforms, but more often than not, did not include a specific medium in the contract. Of the 46 contracts that involved distribution, the most frequent method specified is through digital advertising (17). Table 7 shows examples of each agency’s major outreach initiatives, as indicated by the advertising and public relations services contracts with the highest obligations, along with related performance measurement activities. High-value contracts at CFPB, FEMA, and USCIS focused on a particular outreach initiative or initiatives, while NASA’s highest value contracts were for services that more generally supported outreach and other agency activities. Other GAO staff who made contributions to this report include Carol Henn (Assistant Director); Shari Brewster; Jenny Chanley; Cale Jones; Julia Kennon; Joshua Miller; Meredith Moles; Kathleen Padulchick; and Elise Vaughan Winfrey.

Agencies communicate with the public regarding their functions, policies, and activities. In September 2016, GAO reported that the federal government spends about $1.5 billion per year on public relations activities, carried out through advertising and public relations contracts and by public affairs employees. GAO was asked to describe the purposes and reported benefits of federal agencies' public relations investments. This report reviews (1) the activities selected agencies conducted using advertising and public relations contracts and public affairs employees, and their purposes; (2) how the level of resources the agencies devote to these activities has changed over the past decade and factors officials identified as affecting these changes; and (3) how the agencies measure results of these activities. GAO selected four case study agencies—CFPB, FEMA, NASA, and USCIS—based on factors including obligations for advertising and public relations contracts, numbers of public affairs employees, and agency missions and public interactions. GAO reviewed documentation for contracts valued at or over $150,000 from fiscal years 2012 through 2016; examined staff position descriptions, performance information, and employment data; and interviewed officials at these agencies. CFPB and the Department of Homeland Security provided technical comments on this report, which were incorporated as appropriate. NASA did not have comments. At the agencies GAO examined—the Consumer Financial Protection Bureau (CFPB), Federal Emergency Management Agency (FEMA), National Aeronautics and Space Administration (NASA), and U.S. Citizenship and Immigration Services (USCIS)—most of the advertising and public relations contracts GAO reviewed and public affairs staff responsibilities focused on informing and educating the public. These agencies' advertising and public relations contract obligations were concentrated in a small number of contracts that supported major agency initiatives and communications services. Specifically, CFPB: The four largest contracts ($22.6 million out of $32.8 million) focused on increasing public awareness of CFPB's tools and resources related to its statutory responsibility to educate and empower consumers to make better financial decisions. FEMA: The largest contract ($8.7 million out of $20.7 million) supported the National Flood Insurance Program (NFIP). The agency's role in administering the program includes advertising to encourage people to buy flood insurance. Other high-value contracts promoted emergency preparedness, including the Ready campaign. NASA: The two largest contracts ($7.6 million out of $17.1 million) supported NASA's Communications Support Services Center, which provides graphics and other services across the agency. USCIS: The two largest contracts ($18.1 million out of $19.8 million) were for outreach for two immigration-related eligibility verification systems—E-Verify and Systematic Alien Verification for Entitlements. Over the past decade, changes in advertising and public relations contract obligations and public affairs employees varied at the selected agencies due to changes in agency activities and increased use of digital media. For example, contract obligations were relatively stable at NASA and USCIS. CFPB, on the other hand, saw an increase in obligations due to ramping up operations from 2011, when the agency began operations. Variances in FEMA's obligations stemmed primarily from fluctuations in NFIP contracts due to changing priorities. The number of public affairs employees generally increased at the selected agencies, but was relatively stable at NASA. These increases were due to changes in operations and staffing structure. For example, USCIS changed its staffing model to add more lower-level public affairs staff. Officials at three of the four agencies we reviewed noted an increased use of digital media for public outreach, though the effects on contract and staff resources have been mixed. The agencies measured performance of their activities using web-based and other indicators, such as the number of website visits and length of time spent on a page, and reported using this information to inform decision making. However, agency officials identified challenges in measuring the performance of these activities, including a lack of qualitative data on whether and how information is being used. Selected agencies have taken some steps to address these challenges by, for example, administering surveys to obtain additional feedback.

DOJ’s Civil Rights Division (Division) was established in 1957 to enforce federal statutes prohibiting discrimination on the basis of race, sex, gender preference, disability, religion, and national origin. The Division’s enforcement responsibilities include enforcing antidiscrimination protections in education, employment, credit, housing, public accommodations and facilities, voting, and certain federally funded and conducted programs. The Division has three significant goals: (1) to fulfill the promise of basic civil-rights protections through effective and vigorous enforcement of the law; (2) to deter and remedy discriminatory and illegal conduct through the successful prosecution of these federal laws; and (3) to promote voluntary compliance and civil-rights protection through a variety of educational, technical-assistance, and outreach programs. To fulfill these goals, Division employees are required to travel to investigate potential discrimination, and pursue litigation in court where appropriate. Division employees also travel to engage in community outreach and education, and for training, conference, or administrative purposes. The Division comprises 11 sections, as shown in figure 1 below, all of which are stationed in Washington, D.C. Because of this, nearly all Division attorneys and, occasionally, some nonattorney personnel are required to travel since litigation activities occur in all parts of the United States. The Division spent a combined total of $6 million on travel in fiscal year 2012 and fiscal year 2013 as shown in table 1 below. Travel made up approximately 2 percent of the Division budget in fiscal year 2012 and fiscal year 2013. In 1991, the Division established a fair-housing testing program within its Housing and Civil Enforcement Section and commenced testing in 1992. Housing testing involves sending individuals who pose as prospective buyers or renters of real estate to gather information that can indicate whether a housing provider is complying with fair-housing laws. The primary focus of the section's fair-housing testing program has been to identify housing discrimination based on race, national origin, disability, or familial status. Under the Fair Housing Act, DOJ may initiate a lawsuit where it has reason to believe that a person or entity is engaged in a “pattern or practice” of discrimination or where a denial of rights to a group of persons raises an issue of general public importance. The section employs various means to accomplish testing in local communities, including contracts with private fair-housing organizations, contracts with individuals, and by using nonattorney DOJ employees throughout the country. The majority of fair-housing testing cases filed in court are based on testing evidence that involved allegations of agents misrepresenting the availability of rental units or offering different terms and conditions based on race, national origin, familial status, or disability. DOJ also uses the testing program to test for discrimination in lending and public accommodations. Most federal government travel is regulated by the Federal Travel Regulation (FTR) issued by GSA. The FTR implements statutory requirements and executive branch policies that applicable federal entities must follow, including the general principle that employees traveling on official government business exercise the same care in incurring expenses that a prudent person would exercise if traveling on personal business. In addition to the FTR, DOJ has issued its own travel policies, supplements, and bulletins to further describe and implement the specific travel requirements for the agency. Division employees are issued government travel charge cards through J.P. Morgan Chase and are to use them for all costs associated with government travel, including airfare, hotel rooms, food, and other miscellaneous expenses. The travel charge cards are issued directly to employees, and the cardholder holds all liability for any charges made to the travel card. Employees are expected to pay the card balance in full at the time it is due. There is no interest assessed on unpaid balances but cardholders may be charged a late fee. According to the travel authorization process the Division had in place at the time of our review (see fig. 2), prior to travel, generally the traveler or a travel arranger was to complete an authorization form. The authorization form would then be routed to the appropriate travel authorization-approving official within the section. According to Division officials, approving officials are generally the Section Chief, Deputy Section Chief, or Special Counsel. According to Division travel officials, the authorization must also be signed by a funds certifier, indicating that funds are available to cover travel expenses. Once the authorization form has been signed by the appropriate approving official, the forms are then either e-mailed or faxed to the Division’s finance group, which is responsible for entering the travel authorization information into the Financial Management Information System (FMIS). FMIS is the financial system of record that contains data on travel authorizations and vouchers for the Division. Upon completion of travel, Division employees are to submit a travel voucher document requesting reimbursement for travel expenses, per DOJ policy. In preparing and submitting voucher documents, Division employees are to follow a similar process to the authorization process. According to DOJ policy, the traveler or a travel arranger is to fill out the appropriate travel voucher forms, attach required receipts, and then submit the voucher to the travel voucher-approving official at the section level. The traveler must sign the voucher before the travel voucher- approving official signs the approval. According to Division officials, the same approving official generally signs off on both the authorization and the voucher unless the official is absent. After that, the voucher is to be sent to the Division’s finance group where an accounting technician enters the voucher information into FMIS. At the time of our review, all copies of travel authorization and voucher documents were maintained in a separate data system called Web Docs. The Division implemented the department’s new travel system called E2 in August 2013, shortly after the period covered by our review. E2 is a web-based, end-to-end travel-management application used for travel authorizations; booking of flights, rooms, and cars; and vouchering for reimbursement. E2 is intended to streamline travel management and enable real-time visibility into the buying choices of travelers, as well as assist in optimizing travel budgets while saving taxpayers’ money. According to DOJ financial management officials, this new travel system automates certain tasks compared to the process in place during the period of this review. For example, the E2 system allows travelers to submit and obtain approval for travel requests electronically. The new system directly interfaces with FMIS so that travel authorization and voucher information is automatically entered into the financial system. We did not analyze data from E2 because at the time of our review there was not a sufficient number of trips made and processed through E2 for valid analysis. Instead, after identifying weaknesses in the Division’s existing system, we consulted with GAO travel experts familiar with E2 and reviewed the Division’s documentation of E2 controls to determine whether these controls were designed in a manner that could address these weaknesses. We tested the outcomes of compliance with key travel policies and implementation of related internal controls in nine areas and found indications that controls functioned effectively in four areas but that weaknesses existed in five other areas. Two of these weaknesses should be addressed by the design of the new travel system. In three other areas our testing indicated that the control did not always function as intended and these weaknesses may remain unaddressed by the design of new or existing controls. Table 2 summarizes the travel policies and controls we tested and whether E2 is designed to address any weaknesses we identified. For four areas of travel that we tested, we found the outcomes indicated the Division complied with policy and that controls were effectively implemented and functioning to help ensure that travel complied with applicable rules. Because the outcomes of the related policy testing in three of these areas indicated effectively functioning systems, we did not conduct additional analysis of the underlying control mechanisms in those areas. Meals and Incidental Expenses (M&IE) Reimbursed Appropriately. Federal employees on official travel are reimbursed for lodging and for M&IE up to a set daily rate that varies by location and, for some locations, by time of year. These rates are published annually by the GSA. To examine the extent to which Division M&IE reimbursements complied with standard rates, we analyzed all Division travel for which we had location information from October 2011 through June 2013, by comparing the M&IE reimbursements from FMIS travel data to the standard rates. In almost all cases, the traveler’s reimbursement for M&IE was the amount it should have been or lower. Out of the 2,494 travel vouchers included in this analysis, we identified 13 instances where a higher amount was reimbursed, but the aggregate amount of potential overpayment was $159.50, and in all 13 cases the amount was $18.75 or less per case. Given the small number of cases identified and small dollar amounts potentially reimbursed above standard rates, we did not conduct further review of the associated travel documents for these trips. Required Receipts Provided to Justify Reimbursement. The FTR requires travelers to include receipts for lodging and any single expense over $75 when filing a travel voucher. DOJ policy states that travelers should submit all receipts for expenses over $75, and also specifies that the official approving the voucher should verify that all required receipts are included in the voucher documents. We analyzed travel documents for a generalizable sample of 105 Division trips covering travel from October 2011 through June 2013 to determine the extent to which Division travel complied with DOJ policy regarding the inclusion of required receipts to justify reimbursement. We estimate that approximately 95 percent of all travel vouchers included required receipts. Our sample data were not designed to estimate the dollar value of expenses without receipt for all Division travel, but through our review of trips in our sample we identified a total of $2,023 in travel expenses for four trips that lacked required receipts to justify reimbursement. For example, one voucher did not include a required receipt for $271.60 in airfare and another voucher included partial airfare receipts for $498.20 in airfare. Given the high level of compliance documenting expenses and that E2 currently offers a centralized way for travelers to capture receipts, we did not conduct tests of controls related to ensuring compliance with this policy. Prior Approval of Lodging Costs Higher Than Standard Per Diem. The FTR permits agencies to reimburse travelers above the per diem rate in specific, limited circumstances. The FTR states that travelers should request authorization for reimbursement above per diem rates in advance of travel. DOJ policy requires travelers to provide written justification for requesting lodging above the standard per diem rate and this justification should be on or attached to the travel authorization form for approval prior to traveling. We tested the extent to which this control was in place during the period of our review. Specifically, we analyzed FMIS travel data including all trips occurring from October 2011 through June 2013 where the traveler was reimbursed for lodging and we received location information for the trip. Through this analysis, we identified a small proportion of Division travel including reimbursement for hotel stays above the standard per diem rate, and found that most travelers obtained approval for hotel stays above per diem prior to travel. Specifically, we found that 111 of 2,099 trips with hotel stays, or about 5.3 percent of these trips, were potentially reimbursed above standard per diem rates, including about 1.6 percent of trips with hotel stays that cost $100 or more above the expected amount for the total stay. The total amount of lodging reimbursements above standard per diem rates is less than $22,000, or 3.4 percent of Division lodging reimbursements. Out of the 111 trips that we identified with lodging potentially greater than standard per diem, we reviewed supporting documentation for all trips where Division hotel reimbursement was $100 or more above per diem, which occurred on a total of 25 trips. The documents showed evidence of prior approval in 23 of 25 cases, and this approval was generally obtained through an e-mail from the traveler to the Division travel group that was included with the travel documents. According to Division officials, the traveler was expected to include the e-mail with the authorization document. The new E2 travel system is designed to automatically route any lodging over per diem for proper justification and approval at the time of travel authorization. Length of Travel for Fair-Housing Testing Was Consistent with Testing Activities. To determine whether travel for fair-housing testing was an appropriate length for the days in which testing occurred, we compared vouchers for travel associated with testing to testing data documenting activities. According to the FTR, only those travel expenses essential to the transaction of official business should be paid. DOJ policy states that travelers are only to be reimbursed for travel expenses related to official business. Based on our review of fair-housing tests that occurred from October 2011 through June 2013, travel for housing testing that we reviewed appeared consistent with fair-housing testing activities. All of the testing dates the Division provided matched to corresponding travel vouchers. In general, the length of travel conducted for fair-housing testing was consistent with the test dates provided. For example, a trip that included 3 consecutive days of testing should take no more than 5 days—the 3 days of testing and a travel day at the beginning and end of the trip. This is what we observed in 54 of the 56 testing trips in our data. The two other trips were 1 day longer than expected. On one trip, the traveler was unable to complete any tests on one day of the trip but did perform testing on the other days. According to officials in the Division’s Housing and Civil Enforcement section, they do not record dates in the testing database when testers are not able to complete a housing test. This can occur, for instance, when a tester arrives in town but is unable to reach a property manager. In regard to the second trip, the traveler was the testing coordinator and performed testing on one day of the trip. Officials explained that the traveler would not have conducted testing on all travel days as the testing coordinator’s main role is to oversee and coordinate the work of other testers during the trip. We found weaknesses in two control activities we tested but determined that the design of the controls in the new E2 travel system should address these weaknesses. Appropriate Authority of Officials Approving Travel Authorizations and Vouchers. DOJ policy generally requires that travel authorizations be approved by an individual at a higher level than the traveler, and that travel vouchers be approved by an official at a higher level than the traveler or a senior financial manager. To test the extent to which this control was functioning during the period of our review, we analyzed a generalizable sample of 105 travel vouchers for trips from October 2011 through June 2013 and estimated that 94 percent of travel authorization documents and 76 percent of travel voucher documents were approved by an official with the authority to do so. In some cases, it was unclear whether the control functioned as intended. Specifically, we could not match the signature we observed on authorizations and vouchers to the signature form of individuals with the authority to approve travel—we estimate this occurred in 5 percent of travel authorizations and 7 percent of travel vouchers. Also, we estimate that approximately 1 percent of travel authorizations and 17 percent of travel vouchers were approved by an official who did not have authority to approve travel vouchers according to documents provided by the Division. Division officials stated that this official, a financial management specialist, was authorized to approve travel but officials did not provide sufficient evidence to support this position. Despite these limitations, our review of documentation and consultation with GAO experts familiar with the E2 travel system showed that E2 is designed to enhance this control. Specifically, approvals are designed to be captured and recorded electronically in the system, and travel is to be automatically routed to the proper officials according to system rules that specify the officials who are authorized to approve a traveler’s authorization and voucher documents. Prior Approval of Premium Airfare. Federal regulations require that agencies must specifically authorize and approve other than coach- class airfare and set forth allowable circumstances in which noncoach airfare may be used. DOJ travel policy requires specific justifications for using travel above coach class. We found that Division travelers rarely used premium-class travel during the period of our review. To test the extent to which controls over premium-class travel were functioning from October 2011 through June 2013, we reviewed selected travel vouchers we identified as potentially including premium travel based on high transportation expenses or airfare tickets potentially above standard contract fares. Out of the 40 travel vouchers we reviewed with potential premium travel, we found two instances where the trip included airfare above coach class. In one case, the traveler flew to Seattle, Washington, from Washington, D.C., in February 2012 and took first-class flights on two of the three legs of the trip. The traveler was a manager in the Division and his travel voucher documents did not include evidence of prior approval for the premium class travel or explanation for the need for first class tickets. When asked about this trip, Division officials reported that this was a no-cost upgrade to first class using the traveler’s frequent flier miles, which is why the authorization document did not designate the trip as including premium travel. However, travel documents showed that the cost of the airfare was $1,043 whereas the contract rate for a roundtrip flight from Washington, D.C., to Seattle in fiscal year 2012 was $464–indicating that the government paid an additional $579 for the airfare. In the other instance of premium travel, the traveler tried to obtain approval for a first-class airfare from the Division travel group and reported that the tickets were obtained due to a mix up with travel dates that necessitated a last-minute change to the tickets. The traveler reported that the first-class tickets were the cheapest available at the time of the ticket change. However, the traveler did not obtain approval for the first-class tickets prior to the flight. In addition, for 3 of the 40 trips we reviewed, the travel voucher documents had missing or illegible receipts making it impractical to determine whether premium travel occurred. Division travel officials stated that the travel process in place at the time of our review relied on the traveler reporting any premium travel, and reported that their new travel system has the ability to identify premium travel by automatically retrieving airfare information from the reservation and entering it into the authorization. Travelers may also manually select premium travel in E2. Selection of premium travel in E2 is designed to result in special routing for authorization approval. Our review of documentation and consultation with GAO travel experts familiar with the E2 travel system corroborated Division officials’ observations. We identified weaknesses in two controls and one area of policy that, absent the attention of management or compensating controls, may remain even under the new travel system. Prior authorization of travel is a key control to ensure that travel is necessary and that travel funds are available. The FTR advises travelers to obtain prior authorization to travel except when it is not practical or possible. DOJ travel policy states that the approving official should sign the travel authorization form to document approval permitting travel and obligate estimated expenses for the trip. On the basis of our review of a generalizable random sample of 105 Division trips occurring from October 2011 through June 2013, we estimate that 16 percent of travel did not include documentation that travel was authorized in advance. Specifically, the travel authorization documents for these trips indicated that approving officials authorized the trips after travel had already started, based on the date the approving official signed the authorization form. According to Division officials, all travel is to be authorized in advance, but in some cases the approver may have given verbal approval for travel, and the travel authorization form may be completed after travel has already begun. Although travelers may, under unanticipated circumstances, need such flexibility, lack of formal, documented authorization could compromise the management of fiscal-year travel funds given that the signed travel authorization form also obligates funds for the trip and, without that form, the Division may not be aware that additional funds are needed. For example, we identified a trip where a Division official did not obtain prior authorization to cover travel costs for a speaker at a September 2012 training event and submitted a travel authorization 3 months after the travel, in the next fiscal year. Travel documentation related to the trip included an e-mail from the Division Comptroller expressing concerns about availability of funds to reimburse the traveler for the trip given the lack of prior authorization. While Division officials reported that there was ultimately no issue with availability of funds, this example illustrates potential difficulties imposed in managing travel funds when written approval for travel prior to a trip is not obtained. E2 provides an electronic authorization process that should ensure that the documentation of approvals is maintained, but E2 itself cannot ensure that travelers access and submit travel authorizations in advance. Thus, strengthening controls to promote prior authorization of travel in accordance with Division policies even when travel needs occur at the last minute—such as requiring an e-mail to be sent to one’s supervisor in advance of the travel rather than obtaining verbal approval—could help the Division better ensure that travel is necessary and that travel funds are available to cover the trip. We reviewed travel documents for eight selected travel vouchers to test whether there was documentation of prior approval for use of noncontract airfares (a control), as required by travel rules, and found that none of the Division travel vouchers we analyzed included such approvals. The FTR generally requires travelers to purchase airfare in coach class from a contract carrier unless an exception applies. For instance, noncontract tickets that are less expensive than the contract price can save tax dollars but can also carry additional risk as they are sometimes nonrefundable if travel plans change. DOJ policy encourages the use of noncontract airfares available to the general public when there is a high degree of certainty that travel will occur as planned and noncontract fares are available below the contract fare. In addition, noncontract fares can be appropriate if tickets on the contract carrier are unavailable or impractical based on the needs of the travel—for instance, if no flights are available at an appropriate time to correspond with mission needs. The FTR requires travelers to have an approved authorization for the use of a noncontract carrier before purchasing a noncontract ticket. DOJ policy also requires that authorizing officials approve noncontract airfares in advance. About 14 percent of Division airfares covering single-destination travel from October 2011 through June 2013 were for tickets on noncontract carriers. Of these fares, we found that in about 65 percent of cases (111 of 171 noncontract tickets), the cost of the ticket was below the corresponding contract fare. However, in the remaining 60 cases, the noncontract ticket cost more than the contract fare, including 41 cases where the noncontract ticket cost $100 or more above the contract fare. The total amount we identified that was spent on noncontract airfare above contract fares was approximately $12,000 from October 2011 through June 2013. We reviewed eight travel vouchers with noncontract airfares that were the greatest amount over the contract rate to review the extent to which controls over noncontract airfares were functioning during this same time frame. None of the selected vouchers contained documentation of approval for the airfares or the reason why a noncontract carrier was used. Officials reported that the travel system in place at the time of our review did not enable travel approvers and managers to easily identify whether a flight was a noncontract airfare, and they relied on either the traveler or travel preparer to self-report and manually document any use of noncontract airfares. Without effective controls on travelers using noncontract carriers, the Division does not have reasonable assurance that noncontract tickets were allowable and were selected for appropriate reasons and risks paying excessive amounts for airfare. E2 has the potential to improve management insight into use of noncontract airfares compared to the system in place at the time of our audit, but it is unclear whether the new system fully addresses the control weaknesses we identified. Specifically, E2 can be configured to document the use of noncontract fares automatically when the airfare is booked through its system, and officials provided documentation that the Divisions system had been configured in this fashion. However, officials reported that it is possible for a traveler to book an airfare outside the system, in which case the traveler would have to manually self-report and document that a noncontract fare was purchased—a process that is similar to the self-reporting mechanism in place at the time of our review. Officials believe that this type of airfare purchase would be uncommon, though they had not tested this assertion and therefore cannot be sure to what extent travelers will use this option and properly document noncontract airfare use. As a result, it is unclear to what extent the new system fully addresses control weaknesses identified by our review unless airfare data in E2 is evaluated for compliance with travel rules related to noncontract airfares. According to internal-control standards, evaluations of controls can be helpful to determine the effectiveness of a control when risks are identified, and these evaluations may include review of the control design and direct testing of the control. Given the challenges the Division faced under the previous system that relied on self-reporting by travelers, evaluating this aspect of the new travel system would confirm whether the new configuration and controls are functioning as intended or whether additional actions are needed. Travel rules generally require travelers to submit travel vouchers within 5 working days after travel ends, and this requirement is also reflected in DOJ travel policy. We reviewed the extent to which Division travel complied with this policy from October 2011 through June 2013 and found that Division travelers did not consistently submit vouchers within the required time frame. From our review of a generalizable sample of 105 travel vouchers for Division travel occurring during this time frame, we estimate that approximately 42 percent of all vouchers were not submitted within 5 working days, thus not complying with DOJ policy. Although most late vouchers were estimated to be submitted within a month after travel, we estimate approximately 5 percent of vouchers were not submitted until more than a month after travel ended. According to Division officials, oversight mechanisms included a monthly report of late vouchers provided to all Division sections, and vouchers over 30 days late would result in a notification to the Section Chief. In addition, officials said the Division conducted a quarterly review of open obligations that included reviewing travel authorizations that lack a corresponding voucher, and outstanding obligations would be reviewed with the appropriate section. However, these controls take place after a voucher is likely already late. When asked about controls to ensure timely submission of travel vouchers, Division officials did not identify any controls aimed at proactively improving the timeliness with which travelers submit vouchers. However, the Division reported that it provides policy training regarding timely voucher submission. Delayed submissions of vouchers can make managing travel funds difficult and may lead to travel card delinquencies that could affect an employee’s ability to travel and thus meet the Division’s mission needs. Specifically, while Division travel cardholders are expected to pay the balances on their cards whether they have been reimbursed for travel or not, we identified two cases in which travelers who did not submit timely vouchers became delinquent on their travel cards. For example, one traveler, a manager, did not submit a voucher for more than 6 months after an overseas trip. According to officials, control mechanisms related to late voucher monitoring have not changed since the start of our review. Controls aimed at preventing late voucher submission could help the Division reduce the amount of travel vouchers that are submitted late and better position it to manage travel funds. Most purchases made on Division travel cards aligned with evidence of official travel and appeared appropriate for travel, indicating that the Division complied with key policies we tested and that controls related to travel card use were effective, but we found one area of weakness that may remain (see table 3). Specifically, we found that the Division lacked documentation that a key control regarding the oversight of delinquent accounts was being implemented. FTR and DOJ travel policies state that travel cards can only be used while an individual is on official travel and can only be used for purchases associated with travel such as hotel costs, airline baggage fees, and rental car costs, among other items. Thus transactions made on dates outside of official travel or for items that would not be associated with travel needs can indicate inappropriate personal use of the travel card. To test the Division’s compliance with its policy regarding appropriate use of travel cards, we reviewed all purchases that appeared on Division employee travel charge card records from October 2011 through June 2013 and matched approximately 97 percent of travel card purchases to evidence of official travel, and almost all purchases appeared appropriate for government travel.We could not match 3 percent of purchase transactions corresponding to 1,488 transactions to evidence of official travel. Therefore we reviewed additional detailed documentation related to these transactions and found that the majority of the 1,488 transactions appeared appropriate for official travel but that 64 transactions totaling $2,956 contained evidence of improper use of the travel card, and we referred these cases to DOJ for additional review. These 64 transactions included the following: A Division cardholder who made a $600 purchase at a car dealership in Rockville, Maryland, near the cardholder’s duty station. This purchase did not align with evidence of official travel. A Division cardholder made a $93 purchase at a large retail store that did not align with evidence of official travel and the purchase occurred near the cardholder’s duty station. Lastly, we identified a Division employee who made dozens of purchases in the Washington, D.C., area (the employee’s duty station) that do not align with official travel dates, a suspicious pattern that could indicate personal use of the travel card. For example, in January 2012, the cardholder made a total of 14 purchases, all at restaurants, stores, or gas stations in Washington, D.C., or Arlington, Virginia (a suburb of Washington, D.C.), including four purchases at the same fast-food restaurant in Arlington. Also, this person’s account was suspended due to delinquency twice during the time frame included in our review from October 2011 through June 2013. Given that all Division offices are in Washington, D.C., local purchases that do not align with dates of official travel, such as those described above, may be unrelated to official travel. When we provided the Division with information on these purchases, they reported that the dealership purchase had been a mistake on the part of the cardholder and that shortly after the purchase the cardholder reported the issue to the Division and paid the bill. Also, officials stated that the Division had already identified the other cardholder who had numerous suspicious transactions in the Washington, D.C., area prior to our review, and taken appropriate disciplinary action. Division officials reported that they were continuing to look into the purchase at a large retail store described above based on the information we provided. In addition to analyzing purchase transactions, we also analyzed all cash advance transactions occurring on Division employee travel cards from October 2011 through June 2013, and found potential personal use of travel cards related to cash advance transactions. DOJ policy requires that cash advances align with official travel and do not exceed $40 per day unless justification for a higher amount and prior approval were provided. Overall, most travel cardholders did not make any cash advance transactions during the period of our review—only 19 percent of active cardholders made any cash advance transactions during this time. However, of those who made cash advance transactions, over 60 percent (61 of 98 cardholders) may not have followed DOJ travel policy, either making cash withdrawals that do not align with evidence of official travel or withdrawing cash above normally allowed amounts without documented approval. We identified 123 cash advance transactions totaling $21,424 associated with these cardholders that do not match to evidence of official travel in travel voucher data or travel card data, as shown in figure 3 below. There were 39 cardholders (of 61) who made cash advance transactions that did not match evidence of official travel. Three of these cardholders took out over $1,000 in cash advances with no evidence of travel, with one person withdrawing over $8,000 in cash during the period of our review with no evidence of official travel. Eleven of these cardholders also had accounts that were suspended due to delinquency during the period of our review. In addition, out of the 335 trips in which cardholders took cash advances, we identified 174 trips where cash advances exceeded the allowed amount per DOJ policy. As noted, according to DOJ policy in place from October 2011 through June 2013, travelers were normally allowed to withdraw up to $40 per day of travel. If employees needed more cash than $40 per day, they were required to obtain approval from a component executive officer or equivalent official and justify the need for the additional cash. We matched cash advance transactions to official travel dates from travel voucher data we received, and then compared the total cash withdrawn to the normally allowed amount for the length of the trip. Overall, 43 cardholders withdrew a total of $12,705 above the allowed cash advance amounts. We reviewed all 11 travel voucher documents for trips exceeding cash limits by $200 or more and did not find evidence of approvals for cash advances above the allowed amount. For example, one traveler took out cash advances totaling $1,316 over the course of three trips to Chicago, exceeding the normal cash limits by approximately $676. One of those trips was 3 days long, and the traveler took a cash advance of $345 for the trip. Travel documents for the three trips did not include an explanation for why the traveler required more cash than the allowed amounts, and there was no evidence of approval for the additional cash advances. Federal regulation and DOJ policy state that, while employees are required to use government travel cards during official travel, personal use of the card is prohibited and abuse or misuse of the card may result in disciplinary action. While responsibility for paying off any travel card transactions ultimately falls to the Division employee who was issued the card, lack of conformity with travel card rules may indicate increased risk of waste, fraud, or abuse. For example, personal use of travel charge cards could indicate that an employee is having financial problems and brings into question his or her suitability to hold a position of public trust. Also, according to the Association of Certified Fraud Examiners—a professional organization that sets fraud-examiner standards and provides antifraud training—a common travel card fraud scheme involves withdrawing cash for personal or inappropriate expenses and then adding seemingly appropriate but false cash expenses to a travel voucher for reimbursement. To help illustrate how this fraud scheme would work, a traveler could withdraw $200 cash from a travel card and spend it on personal items not related to official travel, and then include claims on a travel voucher for several fictitious taxi fares under $75 to obtain reimbursement for the cash withdrawal without having to provide receipts for those transactions. Our analysis identified three trips where a traveler’s transactions and voucher fit this risk profile. Specifically, one Division employee took out cash advances totaling $1,512 over the course of three trips, exceeding allowed cash limits by over $1,000. For these three trips, the traveler was reimbursed $893 for taxi fares. No receipts were provided for these taxi fares and they were all under the $75 limit requiring a receipt. Also, the travel documents did not include a justification for why cash over the allowed limits was needed or documentation of approval for such cash. While these expenses may have been legitimate, it illustrates how lack of conformance with policy— specifically the requirement for approval of excess cash advances—may put the Division at increased risk for waste or abuse of travel resources. We provided the Division with examples of potential travel card misuse we identified through our analysis of cash advances. Out of the four examples we provided, the Division had identified three of the cases prior to our review and pursued appropriate disciplinary action, according to an official speaking on behalf of the Division. As a result of our work, this Division official reported that the Division is currently looking into the fourth case to determine whether disciplinary action against the remaining employee is appropriate. In general, officials stated that they pursue disciplinary action against employees who misuse their travel charge card, and that employees found guilty of misuse or abuse are generally suspended without pay for up to 2 weeks. The employees are also required to repay any funds that could be owed to the agency. Division officials reported that at the time of our review, there was one employee—the Travel Card Coordinator—who was responsible for oversight of travel charge cards and associated transactions, and this individual’s time was split among a range of travel-related duties. Officials reported that the coordinator conducted oversight through reviews of various reports produced by the travel charge card contractor, such as a cash advance report, as well as an overall review of transactions. The Travel Card Coordinator was expected to flag any potential misuse of the travel charge card and report it to human resources and section officials, according to Division officials. However, given the range of duties of this individual and because there was only one person to oversee hundreds of travel cards, officials report that oversight may have been less frequent and thorough than would be desirable. The Division has increased oversight of travel charge cards since the start of our audit to better ensure appropriate use of the travel cards. The Division issued a new standard operating procedure on travel charge card oversight specifying that, in addition to reviews by Division accounting staff, employees at the section level are to review travel charge card transactions. The new procedure, issued in February 2014, also calls for monthly audits of all transactions to identify local transactions that may not be related to official travel, transactions in resort destinations, and cash withdrawals that are outside of policy. The new travel card procedure also specifies that 20 percent of all transactions should be randomly sampled each month and reviewed to determine whether the transaction occurred during official travel. Furthermore, the procedure calls for maintenance of records to document the results of the monthly reviews, and reporting of results to the Division Comptroller. If effectively implemented, these new controls may improve travel charge card oversight and address issues we identified related to potentially inappropriate travel card purchases and cash advance transactions. According to DOJ travel policy, the Travel Card Coordinator was responsible for closing the travel card account when an employee terminated employment with DOJ and was to review monthly reports from the travel card system to determine that cards are canceled for separated employees. Our review identified travel card accounts that were not closed in a timely manner upon employee separation from employment, but we did not find evidence of travel card use after employee separation. Our review of travel card accounts that were closed from October 2011 through June 2013 found that about 29 percent (25 of 87) were not closed in a timely manner upon employee separation from DOJ, contrary to its policy. In 6 cases, cards were closed over 100 days after employees had separated from the Division. The Division has taken steps to improve controls over travel charge card account closure. According to officials, the late closure of travel card accounts was due to an inefficient exit process as, in the past, the Travel Card Coordinator relied on notification from the human resources office to find out an employee had left the Division or agency, and the Travel Card Coordinator was not always notified of employee departures. Officials reported that they have implemented a new exit process that incorporates notifications to the Travel Card Coordinator. Also, a new standard operating procedure issued by the Division in February 2014 calls for monthly reviews of travel charge card accounts to ensure that all employees are current, and maintenance of records that the review occurred. In addition, in May 2015 DOJ implemented a policy in which finance staff review travel card accounts quarterly to identify any open accounts associated with employees that have separated from the Department. The new controls the Division introduced, if effectively implemented, could address the limitations we observed. Per DOJ policy, employees are required to pay their balances in full by the due date on their billing statement. Delinquency in payment of travel cards may result in disciplinary action and could affect the employee’s security clearance. According to DOJ policy, as a control to ensure that travel cardholders respond if they become delinquent, the travel card manager is to send delinquency notices to the supervisors of cardholders who are late paying their bill. In addition, per DOJ policy, it is expected that supervisors will talk to the employee about paying the balance on the travel card. According to Division officials, at the time of our audit the Travel Card Coordinator was responsible for sending out delinquency notifications and sent these notices to the section management team for the cardholder. Division officials could not provide evidence that these communications were consistently implemented for delinquent travel charge card accounts from October 2011 through June 2013. During the period of our review, Division travel card accounts were suspended due to delinquency 61 times. A total of 49 cardholders had accounts suspended due to delinquency at some point during the period of our review, with 11 cardholders facing account suspension multiple times. Furthermore, 4 of the 49 cardholders with suspended accounts were managers in the Division. We requested all delinquency notifications for delinquent accounts that were suspended from October 2011 through June 2013. Officials provided delinquent travel card account notices for about 38 percent of the suspensions due to delinquency that we identified. Also, the documentation provided included evidence that the delinquency was discussed with the cardholder in approximately 20 percent of the cases we identified. In one example, the Division cardholder who had the largest amount of cash advance transactions that could not be matched to evidence of official travel—over $8,000 during the period of our review— also had the account suspended twice due to delinquency. The Division could not provide evidence of delinquency notifications for this cardholder. Officials reported that the reason delinquent account notification documents were missing was because notifications were sent by e-mail and documentation was not maintained by the Division Travel Card Coordinator. Delinquency in payment can be an indicator of financial distress that could put the individual at greater risk of abusing his or her travel card— for instance, as noted earlier by padding a travel voucher for expenses never incurred—or other behaviors that could put Division resources, information, or reputation at risk. Also, DOJ receives a quarterly refund payment from the travel charge card contractor based on sales volume and speed of payment. If a travel charge card account becomes delinquent, the amount of refund payment that DOJ receives from the travel card contractor may be reduced. The Division has taken steps to strengthen oversight of delinquent travel card accounts since the period of our audit. Division officials report that they have implemented a new process requiring sections to review delinquent account information and report to the Division Comptroller and human resources office if any potential misuse is identified, and the Division pursues disciplinary action in cases of delinquent accounts that remain unpaid. Also, a new operating procedure issued in 2014 calls for monthly distribution of delinquency reports to each section, and historical maintenance of the distributed reports. However, the new procedure does not call for maintenance of documentation of communications with the cardholder, a key component of delinquent-account oversight. If the Division does not maintain evidence of communications with the cardholder, it will not have an institutional record to determine whether staff are implementing this key control, and therefore, if delinquencies persist, it will not be able to determine whether the underlying cause is lack of implementation of this control or whether additional adjustments are needed, such as strengthening controls or different processes. Waste, fraud, or abuse of official government travel can significantly affect the reputation of an agency and undermine public confidence in the integrity of federal employees, even where travel budgets are relatively small compared to overall federal spending, as is the case with the Division. While most Division travel we reviewed followed applicable rules and policies, we identified weaknesses in compliance with certain policies and related controls that could lead to travel waste or abuse. Deficiencies in documenting prior authorization of travel may compromise the management of fiscal-year travel funds and increase the risk that travel is not prudent. Also, late travel voucher submissions make managing travel funds difficult. Overall, travel rules surrounding authorization and reimbursement of trips are intended to help ensure that official travel is carried out in a responsible manner, and are a key component of the internal control system to help safeguard assets and prevent and detect errors and fraud. Furthermore, while the Division’s new travel system includes additional controls over proper justification and approval of noncontract airfares, the Division has not evaluated whether the new controls are effective, which would provide the Division with additional assurance that the controls are functioning as intended or indicate whether additional actions are needed. In addition, the Division has implemented new controls to improve oversight of travel charge cards, including improved oversight of delinquent travel card accounts. However, without documentation of communications with delinquent account holders, the Division does not have an institutional record to determine whether staff are implementing this key control, and therefore, if delinquencies continue, the Division will not be able to readily identify whether the underlying cause is lack of implementation of this control or whether there is a need to strengthen controls or implement different processes. To strengthen controls over Division travel, the Attorney General should direct the Assistant Attorney General for the Division to take the following three actions: 1. To better ensure that travel authorization and reimbursement comply with rules and regulations, strengthen existing or implement new internal controls over documenting that travel is authorized in advance and submitting travel vouchers in the required time frame. 2. To help ensure that noncontract airfare use is properly authorized and justified in accordance with travel rules, the Division should evaluate whether the configuration of its new travel system has implemented controls to address previous shortcomings in obtaining and documenting required approvals, and that travelers booking airfares outside of the system are documenting this decision. 3. To better document oversight of delinquent travel charge card accounts in accordance with DOJ policy, take steps to maintain documentation of communications with delinquent travel cardholders. We provided a draft of our report to the Attorney General for review and comment. We received written comments from the DOJ Civil Rights Division’s (Division) Acting Chief of Staff, which are reproduced in appendix II. The Acting Chief of Staff concurred with our recommendations, stating that the Division is committed to the effective and prudent stewardship of funds, and to continuous improvements in financial management in support of the Division’s mission. In response to our first recommendation to strengthen existing or implement new internal controls over documentation of travel authorization, the Division stated that it is implementing a new policy in December 2015 requiring written approval for all travel including cases of urgent travel where authorization through the travel system is not feasible. If implemented effectively, this new process should address our recommendation. In response to the second part of the first recommendation to strengthen existing or implement new controls over timely submission of travel vouchers, the Division reported that in late fiscal year 2013 it started distributing monthly reports to section management with all completed trips that had not been vouchered, and stated that the Division requires section management to take action to ensure travelers submit vouchers in a timely manner. In addition, the Division stated that leadership receives a monthly report with average number of days that a section’s vouchers are outstanding, for the purpose of identifying patterns of noncompliance. These new processes focus on identifying travel vouchers that are likely already late; however, it is possible that the new processes could improve timely voucher submission if problem areas are identified and action is taken to enforce more timely submission among groups that have issues with late vouchers. In response to our second recommendation regarding controls to ensure that noncontract airfares are properly authorized and justified, the Division responded that in cases where flight reservations are made outside the travel system and cannot be recorded in E2, it will require travelers to document in E2 whether a purchased flight was a contract or noncontract fare and to provide evidence of advance approval of any noncontract fares. If implemented effectively, this new process should address our recommendation. Lastly, in response to our third recommendation related to maintaining documentation of communication with employees with delinquent travel card accounts, the Division stated that it plans to provide instructions requiring management to document communications and actions taken with delinquent accountholders, and provide the documentation to the Division’s Comptroller. If implemented effectively, this new process should address our recommendation. The Division and DOJ’s Justice Management Division also provided technical comments, which we have incorporated as appropriate. We are sending copies of this report to relevant congressional committees, the Attorney General, the Assistant Attorney General for the Division, and other relevant parties. This report will also be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6722 or bagdoyans@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. This report examines the extent to which the Department of Justice (DOJ) Civil Rights Division (Division) implemented internal controls and complied with travel policy in key areas: (1) authorization and reimbursement of travel and (2) use of travel charge cards. To this end, we interviewed DOJ officials and reviewed travel regulations and policies. We also analyzed Division travel data from October 2011 through June 2013—the most recent data available when we made our request— including official authorization and voucher data from the Division’s Financial Management Information System (FMIS), travel card data from the J.P. Morgan Chase PaymentNet data system, and fair-housing testing data from the Fair Housing Testing System (FHTS). The data we analyzed included all trips associated with investigations resulting in a court case (which we refer to as “cases” in this report) and completed investigations not resulting in a court case (which we refer to as “closed investigations”). We estimate that the travel vouchers associated with cases and closed investigations represented approximately 61 percent of Division travel. We developed a methodology to test whether the Division effectively implemented internal controls and complied with policy in key areas. We excluded travel information related to ongoing investigations that could result in ongoing litigation. After identifying important travel controls and requirements as indicated in federal travel regulations and DOJ travel policies, we tested the implementation of controls and compliance with key travel policies by analyzing travel data, reviewing policy documents, interviewing officials, and comparing findings to criteria from GAO’s Standards for Internal Control in the Federal Government. We tested the implementation of controls by analyzing data that indicated whether the control had been implemented. In some cases, the data we analyzed directly indicated whether the control had been implemented, such as through compliance with policy, and we report on those outcomes. Where we found that Division travel complied with policy, we did not conduct further testing of related controls. In other cases, the data pointed to a weakness that required additional testing. If we identified a weakness in the implementation of controls that may have been addressed through a change in DOJ’s travel system subsequent to our analysis, we evaluated the design of any related new control as a way of assessing whether the control weakness had been mitigated. We performed data-reliability assessments on the FMIS authorization and voucher data, the PaymentNet travel charge card data, and the FHTS fair-housing test data. Data-reliability assessments included reviewing related documentation, interviewing knowledgeable DOJ and J.P. Morgan Chase officials on how the data are compiled, using control totals, and comparing records in the file against agency totals for travel vouchers and charge card transactions. We also performed electronic testing for completeness and accuracy of data. Overall, our assessment found the data to be sufficiently reliable for the purposes of this report. We conducted this performance audit from November 2014 through November 2015 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. To address the first objective of the report, we reviewed travel authorization and voucher data, as well as travel charge card data from October 2011 through June 2013—the most recent data available at the time of our request. DOJ implemented a new travel system in August 2013, and at the time of our request there were not a sufficient number of trips to evaluate the implementation of controls in the new system. Our review included travel conducted by Division employees as well as travel conducted by other DOJ employees who performed work for the Division and were authorized and reimbursed by the Division. We reviewed the implementation of internal controls and compliance with policy over the travel authorization and voucher process through a number of analyses. When we identified instances in which the implementation of controls was not effective, and if the data on implementation did not directly identify why the controls did not work as intended, we took additional steps as appropriate to make this determination. As part of our review, we drew a stratified, random probability sample of 105 trips generalizable to the 3,157 Division trips in the scope of this review. We used this sample to analyze the extent to which officials approving travel were authorized to do so by comparing the approving signature with a list of individuals authorized to provide such approvals. Using this sample we also determined the extent to which authorization was provided in advance of travel by examining the date of approval and comparing that to the dates of travel. We used this sample to review approval information for vouchers, including the position of the approver and whether the position was authorized to approve travel by comparing the signature to a list of individuals authorized to approve vouchers. We reviewed whether these vouchers were submitted timely in accordance with Division policy by looking at the submission date and determining whether it was within 5 days of completion of travel. We also determined whether vouchers included all required receipts needed to justify reimbursement by reviewing this documentation and comparing it to the requirements of travel policy. As stated above, our generalizable sample included a total of 105 trips from the 3,157 trips in the scope of this review (travel within the time frame of October 1, 2011, through June 30, 2013). The population of trips was divided into a certainty stratum with the 10 most expensive trips based on total voucher amount, and a noncertainty stratum with all remaining trips. The sample comprised all 10 trips in the certainty stratum and a generalizable random probability sample of 95 trips from the noncertainty stratum. Each sample element was subsequently weighted in the analysis to account statistically for all the trips in the population, including those that were not selected. Because we followed a probability procedure based on random selections, our sample is only one of a large number of samples that we might have drawn. Since each sample could have provided different estimates, we express our confidence in the precision of our particular sample’s results as a 95 percent confidence interval. All percentage estimates from this sample review have 95 percent confidence intervals of plus or minus 10 percentage points or less, unless otherwise noted. This is the interval that would be expected to contain the actual population value for about 95 percent of the samples we could have drawn. We also examined travel associated with the Fair Housing Testing program to determine the extent to which fair-housing testing activities aligned with official travel. We reviewed data from the Fair Housing Testing System (FHTS) to include fair-housing tests that occurred from October 2011 through June 2013, and were considered closed by the Division. As with the travel data, we excluded data related to ongoing investigations (in this case fair-housing tests) from our review. We matched fair-housing test dates from FHTS to official travel voucher dates, and reviewed the length of travel to identify any trips that appeared longer than necessary given that housing tests normally last 1 day. For trips that appeared longer than normal, we followed up with the Division to determine the reason for the longer trip. As part of our first reporting objective, we also reviewed the extent to which Division travel included premium airfare—business- or first-class airfare—and whether premium airfare was authorized and reimbursed in compliance with travel rules. We identified potential premium travel through several methods. First, we used the data field “VT_CLASS” in the FMIS travel voucher data to identify trips categorized as including premium travel. We reviewed travel documents for a selection of 11 trips identified in FMIS as including premium airfare to check the extent to which any premium airfare was properly authorized. The 11 trips were selected because the transportation costs associated with the trips were greater than the contract rate for airfare to the destination or there was no established contract rate. In addition, we used travel charge card transaction data to identify airfare purchases and compared the cost of the airfare to General Services Administration (GSA) contract rates for the travel destination. We identified cases where the airfare purchase appeared more expensive than the standard contract rate, and we selected the 10 trips where the difference between the contract rate and the actual airfare costs were greatest for further review. We reviewed these 10 travel vouchers and their associated travel documents to confirm whether the trip included premium airfare, whether the premium airfare was properly authorized, and what justification was provided for premium travel. In addition, we also selected the 10 travel vouchers with highest transportation costs, which would normally indicate cost of airfare, to review for any premium travel. We also used the travel charge card transaction data to identify airfare purchases that were on airlines other than the contract airline for a travel destination, also called noncontract airfares. We reviewed the 9 vouchers identified as including noncontract airfares that were the greatest amount above the contract rate. We also reviewed travel documents for these trips to determine whether the use of noncontract airfares was properly authorized and justified. We reviewed travel voucher data and compared reimbursed lodging rates and meals and incidental expenses (M&IE) rates to the standard GSA rates for a trip location. We identified all cases where M&IE was greater than the standard GSA rate using the travel data provided. We also identified cases where lodging potentially exceeded the standard GSA rate for the location, and requested and reviewed travel documents for trips that exceeded the standard lodging rate by $100 or more. We identified a total of 25 trips where lodging exceeded standard rates by $100 or more, and examined whether travelers obtained approval to exceed the standard rate, and the reasons given for the higher rate. To address the second objective of this report, we reviewed travel charge card information from the J.P. Morgan Chase PaymentNet system covering all Division travel cardholders from October 2011 through June 2013—the most recent data available at the time of our request. We received all transactions data for the period, and information on all Division cardholders, such as dates accounts were opened or closed. We reviewed the extent to which Division travel card transactions aligned with official travel by comparing PaymentNet charge card transactions to travel voucher travel dates. Because we did not have official travel vouchers for travel associated with ongoing investigations, we used hotel stays appearing in the charge card data to identify additional evidence of official travel. For those purchases that did not align with any signs of official travel, either in travel vouchers or hotel-stay data, we reviewed the data for any unusual transactions, such as large-dollar retail transactions, large-dollar transactions in the Washington, D.C., area, or a number of purchases in the area of the employee’s duty station in Washington, D.C. We also reviewed all transactions for any indications of potentially inappropriate purchases by reviewing the merchant categories of all purchases and reviewing any individual transactions for those merchants that appeared unusual for official travel. In addition to reviewing purchases, we reviewed whether cash advance transactions aligned with evidence of official travel, and whether cash advance transactions followed policy rules. We used PaymentNet transaction data to identify cash advance transactions, and then compared the timing of these transactions to travel voucher data from FMIS. For cash advances that could be matched to an official travel voucher, we compared the amount of cash withdrawn to normally allowed amounts under DOJ policy—$40 per travel day—to determine whether the cardholder took out more cash than is normally allowed. We requested and reviewed travel documents for 11 trips where cash advances exceeded the normally allowed amounts by $200 or more, and we reviewed whether the excess cash was authorized. We also identified cash advance transactions that did not align with evidence of official travel in FMIS travel voucher data or hotel-stay data in PaymentNet and reviewed the extent that certain cardholders took cash advances that do not align with evidence of official travel. We also took several steps to assess the extent to which travel charge cards were managed in line with DOJ policy. To review controls over closure of travel card accounts upon employee separation from the Division, we used PaymentNet travel charge card account data to determine all persons with accounts that closed from October 2011 through June 2013. We then provided the Division with a list of these account holders, and requested that they identify accounts that closed because of separation from the agency, and provide the date of separation. We compared the separation date to the account closure date to determine the extent to which accounts were closed promptly. Lastly, to review evidence of delinquent account oversight, we requested and reviewed delinquency notifications for travel cardholders who had accounts suspended due to delinquency during the period of our review, October 2011 to June 2013. In addition to the contact mentioned above, the following staff members made significant contributions to this report: Joah Iannotta and Phil Reiff, Assistant Directors; Celina Davidson, Analyst-in-Charge; Tracy Abdo; Melinda Cordero; Colin Fallon; Grant Mallie; Erin McLaughlin; Maria McMullen; James Murphy; Sonya Vartivarian; Shana Wallace; and Chris Zbrozek.

The sensitive nature of taxpayer-funded travel necessitates that federal agencies have strong internal controls in place to help ensure that travel complies with rules and regulations. GAO was asked to review the Division's travel controls. This report examines the extent to which the Division effectively implemented internal controls and complied with travel policy in key areas, including appropriate (1) authorization and reimbursement of travel and (2) use of travel charge cards. GAO analyzed Division travel data for all trips associated with investigations resulting in a court case and completed investigations not resulting in a court case (“closed investigations”), and travel charge card use from October 2011 through June 2013, the most recent data available at the time of GAO's data request. GAO did not analyze travel data associated with ongoing investigations. GAO also analyzed a generalizable sample of travel documents for 105 of the 3,157 trips within the period of its review. GAO also reviewed relevant federal and agency-specific travel rules and interviewed officials. GAO tested the implementation of internal controls and compliance with key policies in nine areas of the Department of Justice's (DOJ) Civil Rights Division (Division) travel authorization and reimbursement process for the period from October 2011 through June 2013 and found indications that controls functioned effectively in four of those areas. Weaknesses existed in the remaining five areas. Two of these weaknesses should be addressed by the design of a new travel system the Division began using in August 2013. Although the new travel system was in use during GAO's review, the Division had used the new system for too few trips for GAO to analyze its data. The four areas of travel in which the Division complied with policy and controls were generally functioning effectively included appropriate levels of per diem reimbursement, presence of required receipts, prior approval of higher-than-per-diem lodging, and appropriate length of trip for certain oversight activities (related to ensuring compliance with fair-housing laws). GAO identified weaknesses in three areas including the following: GAO found that 16 percent of travel authorizations did not include documentation of approval prior to travel, contrary to DOJ policy, which could hinder the effective management of travel funds. Although unanticipated travel may require flexibility for travelers, strengthening controls to promote prior written authorization could help the Division better ensure that travel is necessary and funds are available. GAO estimated that 14 percent of Division airfares were on noncontract carriers, and none of the eight highest cost vouchers in GAO’s sample had documentation of prior approvals as required by DOJ policy. The new DOJ travel system has controls designed to document approval of noncontract airfares but not if travelers book flights outside the system. The department has not yet evaluated whether this is occurring, yet doing so would confirm whether the new configuration of controls are functioning as intended. GAO estimated that 42 percent of Division travel vouchers were not submitted within required time frames per DOJ policy, which could result in difficulties in managing travel funds. Most Division travel charge card use appeared appropriate indicating effectively functioning controls and compliance with certain travel policies—for instance, more than 97 percent of purchases on Division travel cards aligned with evidence of official travel and appeared appropriate. GAO found three weaknesses related to travel card controls. First, cash advance transactions did not always comply with travel policies, as 19 percent of transactions did not align with evidence of official travel. Second, travel cards were not closed timely in 29 percent of cases reviewed. The Division has implemented new procedures since the start of GAO's audit, implementing new controls whose design should address these two issues. Third, the Division did not maintain documentation of communication with delinquent cardholders, a key component in addressing delinquent accounts. Without this documentation, the Division will not be able to determine whether staff are implementing this control, and thus, if delinquencies persist, it will be hindered in determining if the underlying cause is lack of implementation of this control or the need to strengthen controls or implement different processes. GAO recommends, among other things, that the Division strengthen controls related to prior authorization of travel and timely submission of vouchers, evaluate whether new controls for noncontract airfares are functioning properly, and improve how it documents oversight of delinquent travel card accounts. The agency concurred with GAO's recommendations.

Among other things, FSA is responsible for implementing USDA’s direct and guaranteed loan programs. FSA’s district office staff administer the direct loan program and have primary decision-making authority for approving loans. As of September 30, 2001, there were about 95,000 borrowers with direct loans outstanding, with an unpaid principal balance of about $8.5 billion. FSA farm loan managers are responsible for approving and servicing these loans. The factors FSA staff consider in approving or denying a loan include the applicant’s eligibility, (i.e., he or she must operate a family-size farm in the area), credit rating, cash flow, collateral, and farming experience. Once a farm loan application is complete, FSA officials have 60 days to approve or deny the application and notify the applicant in writing of the decision. Once FSA approves a direct loan, it helps borrowers develop financial plans; collects loan payments; and, when necessary, restructures delinquent debt. Direct loans are considered delinquent when a payment is 30 days past due. When a borrower’s account is 90 days past due, FSA county staff formally notify him or her of the delinquency and provide an application for restructuring the loan. To be considered for loan restructuring, borrowers must complete and return an application within 60 days. FSA staff process the completed application and notify borrowers whether they are eligible for loan restructuring. If a borrower does not apply or is not eligible for loan restructuring, and the loan continues to be delinquent, FSA notifies the borrower that it will take legal action to collect all the money owed on the loan (called loan acceleration). If the borrower does not take action to settle their account within a certain period of time, FSA may start foreclosure proceedings. When farmers believe that FSA has discriminated against them, they may file a discrimination complaint with USDA’s OCR. For the complaint to be accepted, it must be filed in writing and signed by the complainant; be filed within 180 days of the discriminatory event; and describe the discriminatory conduct of a USDA employee or the discriminatory effect of a departmental policy, procedure, or regulation. Farmers may also seek compensation for violations of their civil rights by filing individual or class action lawsuits. In 1997, African-American farmers filed a class action against USDA (Pigford v. Glickman). In 1999, this suit resulted in a multimillion-dollar settlement agreement for the farmers. Since then, women and other minority farmers have also filed class actions against USDA. As you know, to elevate the attention of civil rights matters at USDA, the Congress created the position of Assistant Secretary of Agriculture for Civil Rights in the 2002 Farm Bill. In addition, in September of this year, the Secretary of Agriculture announced the creation of a new office within FSA to work with minority and socially disadvantaged farmers who have questions and concerns about loan applications filed with local offices. During fiscal year 2000 and 2001, the national average processing time for direct loans for Hispanic farmers was 20 days—4 days longer than for non- Hispanic farmers—but well within FSA’s 60-day requirement. At the state level, loan processing time differences were more varied. For example, in the four states that account for over half of all Hispanic applications, processing times for Hispanic farmers were faster than for non-Hispanic farmers in three states and slower in the fourth state. However, all times fell well within FSA’s 60-day requirement. Table 1 shows the average processing times of non-Hispanic and Hispanic farmers’ applications nationwide and for the four states. The vast majority—91 percent—of all direct loan applications from Hispanic farmers were processed within FSA’s 60-day requirement. However, the loan approval rate for Hispanic farmers was lower than for non-Hispanic farmers during this 2-year period: 83 and 90 percent, respectively. FSA officials maintain that approval rate differences were not significant and attribute them to differences in the applicants’ ability to repay the loans they requested. Despite national differences, as shown in table 2, in three of the four states that received the largest number of Hispanic applications in fiscal year 2001, direct loan approval rates were similar. As part of FSA’s assessment of its civil rights performance, the agency monitors differences between minority and nonminority loan processing times and approval rates at both the national and state levels. In addition, FSA sends teams to state offices to conduct civil rights reviews. The teams review loan files to verify compliance with FSA policies and procedures and, if warranted, provide written recommendations to remedy problems they find. Through fiscal year 2001, each state office was reviewed once every 3 years; beginning in fiscal year 2002, the offices will be reviewed every other year. While FSA monitors variations in loan processing times and approval rates for minorities and nonminorities, it does not have established criteria for determining when observed variations are significant enough to warrant further inquiry. In addition, while FSA conducts periodic field reviews of state offices’ performance in civil rights matters and suggests improvements, it does not require the offices to implement the recommendations and does not monitor state office follow-up efforts. FSA is currently considering requiring state offices to provide information on how they have addressed weaknesses noted during reviews. USDA has a policy for issuing stays of foreclosure in cases when discrimination has been alleged in individual complaints filed with OCR, but not in response to individual or class action lawsuits with similar allegations. When an individual files an administrative discrimination complaint with OCR, FSA’s policy is to automatically issue a stay of adverse action—including foreclosure–until the complaint has been resolved. During fiscal years 2000 and 2001, this policy was followed in 24 of the 26 applicable cases involving Hispanic borrowers. The policy was not followed in the remaining two cases because of miscommunication between OCR and FSA in reconciling their respective lists of complainants. When FSA learned that complaints had been filed with OCR, it stayed its foreclosure actions, and, as of August 2002, no further collection actions had been taken against the two farmers. Although future data system improvements should alleviate this problem, OCR and FSA officials acknowledge that improvements could be made in the interim. USDA does not have a similar policy for issuing stays related to discrimination claims raised in an individual or class action lawsuit. Instead, FSA makes decisions on whether to issue stays on a case-by-case basis based on the advice of USDA’s General Counsel and the Department of Justice. Since 1997, USDA has issued stays of foreclosure related to African-American and Native American farmers’ class action discrimination lawsuits involving FSA loan programs. In contrast, USDA did not issue stays of foreclosure for other class action discrimination lawsuits involving FSA loan programs because the department believed that the circumstances did not warrant a stay. These class action lawsuits and how USDA handled stays of foreclosure are discussed in greater detail below. In October 1997, African-American farmers filed a class action lawsuit against the Secretary of Agriculture (Pigford v. Glickman) alleging racial discrimination by USDA in its administration of federal farm programs. On October 9, 1998, the court certified the class—issued the criteria for class eligibility. On January 5, 1999, USDA entered into a 5-year consent decree with the claimants of the suit to settle it. The federal district court approved the consent decree and a framework for the settlement of individual claims in April of the same year. As of August 29, 2002, about 21,800 claims have been accepted for processing. As part of the consent decree, USDA agreed to refrain from foreclosing on real property owned by a claimant or accelerating their loan account.In November 1999, Native American farmers filed a class action lawsuit against the Secretary of Agriculture (Keepseagle v. Glickman) alleging that USDA willfully discriminated against Native American farmers and ranchers when processing applications for farm credit and farm programs. Further, claimants alleged that some class members had previously filed discrimination complaints with USDA and that the department had failed to thoroughly investigate the complaints. In December 1999, USDA issued a notice to FSA offices directing them not to accelerate or foreclose on any direct loans held by Native American borrowers unless the national office, with the concurrence of the Office of General Counsel, specifically authorized such action against an individual. As scheduled, this directive expired at the end of 2000. In October 2000, Hispanic farmers (Garcia v. Glickman) and women farmers (Love v. Glickman) each filed class action lawsuits against USDA alleging similar claims that USDA willfully discriminated against them in processing applications for farm credit and farm programs. Specifically, they alleged that loans were denied, provided late, or provided with less money than needed to adequately farm. In addition, the plaintiffs alleged that when they filed discrimination complaints about the handling of their loan applications, USDA failed to investigate them. The department has not issued stays of foreclosure in either of these lawsuits. In June 2001, USDA’s Acting General Counsel wrote a memo that explained the department’s reasoning for issuing stays of foreclosure in response to some class action lawsuits, but not others. According to the memo, the stay of foreclosure agreement included in the Pigford consent decree was reached only in the context of litigation and only to settle a lawsuit in which a class action had already been certified by the district court. The memo went on to say that the stay of foreclosure policy issued in response to the Keepseagle lawsuit was implemented during the infancy of the lawsuit while USDA and the Department of Justice were evaluating how to proceed in defending it. In addition, the memo stated that USDA did not intend to continue a stay of foreclosure beyond the evaluation. Further, the Acting General Counsel wrote that in all three of the pending lawsuits—Keepseagle, Garcia, and Love—no adequate factual bases had been alleged to support the claims of discrimination made by most of the named plaintiffs. As a result, the department saw no reason to implement a policy to halt foreclosures and other similar actions affecting borrowers potentially involved in these lawsuits. As of September 2002, a class has been certified for the Keepseagle lawsuit, but not for the Garcia suit. USDA has not issued any further stays of adverse action for participants in any of these lawsuits. Although USDA has not issued stays of foreclosure for potential class members in Garcia, relatively few Hispanic farmers have been affected by this decision. According to our survey of state offices, FSA accelerated the direct loans of almost 1,500 borrowers during fiscal years 2000 and 2001; only 41 of these borrowers were Hispanic. FSA also foreclosed on the loans of 6 of these 41 farmers during this period. In addition to these 41 borrowers, 10 other Hispanic borrowers who had their loans accelerated in prior years were foreclosed on during fiscal years 2000 and 2001. To put these figures into context, during this period, FSA foreclosed on the loans of approximately 600 borrowers, 16 (or 3 percent) of whom were Hispanic. During this period, Hispanic farmers made up about 4 percent of the agency’s direct loan portfolio. FSA does not maintain historic information on accelerations or foreclosures in a manner for this information to be readily retrieved or analyzed. FSA officials acknowledged that such information is needed in light of the frequent charges of discrimination it faces. OCR has adopted many recommendations made in the past by USDA’s Inspector General and agency task forces. For example, in 2000, a USDA task force identified 54 tasks to help address problems with OCR’s organization and staffing, database management, and complaint processing. As of July 2002, the office had fully implemented 42, or nearly 80 percent, of these recommendations and plans to complete implementation of most of the others by October 2002. In addition, OCR has made some organizational modifications, such as creating separate employment and program directorates and adding three new divisions to the latter—Program Adjudication, Program Compliance, and Resource Management Staff. Further, from the beginning of fiscal year 2000 to the end of fiscal year 2001, OCR has made significant progress in reducing its inventory of complaints from 1,525 to 594. Despite these actions, however, OCR continues to fail to meet USDA’s requirement that program complaints be processed in a timely manner. Specifically, USDA directs OCR to complete its investigative reports within 180 days after accepting a discrimination complaint. However, during fiscal years 2000 and 2001, OCR took on average 365 days and 315 days, respectively, to complete its investigative reports. Furthermore, as shown in figure 1, the 180-day requirement covers only a portion of the three major stages of the entire processing cycle. Accordingly, even if the 180- day requirement were met, OCR still take 2 years or more to complete the processing of a complaint. In fact, when all phases of the complaint resolution are accounted for, OCR took an average of 772 and 676 days in fiscal years 2000 and 2001, respectively, to completely process complaints through the entire complaint cycle and issue the final agency decision. OCR has made only modest progress in improving its timely processing of complaints because it has yet to address severe, underlying human capital problems. According to USDA officials, OCR has long-standing problems in obtaining and retaining staff with the right mix of skills. The retention problem is evidenced by the fact that only about two-thirds of the staff engaged in complaint processing in fiscal year 2000 were still on board 2 years later. OCR officials also pointed out that this staffing problem has been exacerbated because management and staff have been intermittently diverted from their day-to-day activities by such tasks as responding to requests for information from the courts. Furthermore, severe morale problems have exacerbated staff retention problems and have adversely affected the productivity of the remaining staff. Management officials told us that they spend an inordinate amount of time and resources addressing internal staff complaints. In fact, during fiscal years 2000 and 2001, OCR had one of the highest rates of employee- filed administrative complaints in the department. This atmosphere has led to frequent reassignments or resignations of OCR managers and staff. According to senior OCR officials, the problem has reached the point where some staff have even threatened fellow employees or sabotaged their work. Although OCR’s Director believes that the situation has improved over the past few years, he acknowledges that some of the more serious morale problems have not been resolved. In conclusion, Mr. Chairman, USDA has continuously faced allegations of discrimination in its making direct loans to farmers over the past decade. To help guard against such charges, FSA needs to improve its monitoring and accountability mechanisms and make its systems and decision processes more consistent and transparent. Although FSA monitors variations in loan processing times and approval rates, it lacks criteria for determining when discrepancies warrant further inquiry. Similarly, while FSA conducts periodic reviews of its state offices’ civil rights conduct and makes suggestions for improvement, it cannot ensure that these suggestions have been effective—or even adopted– without a requirement that state offices implement its recommendations or, if not, explain their reasons for not doing so. In addition, USDA has also been criticized for its handling of the allegations themselves—whether they were handled through litigation or the agency’s complaint processes. In the case of class action lawsuits, USDA has been charged with treating different minority groups inequitably because it grants stays of foreclosures to some groups but not to others. Without a standard, transparent policy that lays out the factors USDA considers in deciding whether or not to issue stays, the department faces the continued problem of having its decisions viewed as unfair. Furthermore, if FSA and OCR do not improve their process for reconciling their respective lists of complainants, FSA runs the risk of violating its policy of not taking foreclosure actions against farmers with pending discrimination complaints. In addition, without maintaining historical information on foreclosures, USDA lacks an important tool to help it understand its equal opportunity performance.

For years, some minority and women farmers have alleged that the Department of Agriculture's (USDA) Farm Service Agency (FSA) discriminates against them, treating them differently from other farmers during the loan approval or foreclosure process. During fiscal years 2000 and 2001, FSA took, on average, four days longer to process loan applications from Hispanic farmers than it did for non-Hispanic farmers: 20 days versus 16 days. The FSA's direct loan approval rate was somewhat lower for Hispanic farmers than for non-Hispanic farmers nationwide: 83 and 90 percent, respectively. USDA's policies for staying foreclosures when discrimination has been alleged depend on the method used to lodge complaints. When an individual's discrimination complaint is accepted by the Office of Civil Rights (OCR), FSA's policy is to automatically issue a stay of adverse action, such as foreclosure, until the complaint has been resolved. OCR has made modest progress in the length of time it takes to process discrimination complaints. USDA requires OCR to complete the investigative phase of processing a complaint within 180 days of accepting it. In fiscal year 2000, OCR took an average of 365 days to complete just the investigative phase. Although OCR slightly improved this average to 315 days in fiscal year 2001, this continues to exceed the department's internal 180-day requirement.

The federal government established education provisions for American Indians through treaties dating back to the late 1700s. Since the early 1800s the federal government has funded schools to educate American Indians, and Interior’s BIE currently administers this school system. These federally funded schools were established in order to provide educational opportunities for American Indian children who largely live in remote areas. Today, an estimated 10 percent of American Indian children attend the 174 schools and 12 dormitories that receive funding from the Department of the Interior’s BIE. Although these schools are located in 23 states across the nation (see fig. 1), the majority of BIE students (83 percent) attend BIE schools in 6 states—Arizona, Mississippi, New Mexico, North Dakota, South Dakota, and Washington. According to BIE, in the 2006-07 school year, educational opportunities were provided to approximately 48,000 students in these schools located across 63 reservations. According to BIE, American Indian students enrolled in the BIE-funded schools represent 228 tribes, but the majority of students belong to a small number of tribes. The primary mission of BIE schools is to provide quality educational opportunities that are compatible with tribes’ cultural and economic well- being and their wide diversity as distinct cultural and government entities. To accomplish its mission, BIE’s elementary and secondary school system is multifaceted, with schools located in a variety of settings, including rural, town, suburban, and urban areas. However, the schools are located primarily in rural areas and small towns and serve American Indian students living on or near reservations. The BIE school system includes day schools, on-reservation boarding schools, and off-reservation boarding schools—which house and educate students from numerous tribes. BIE schools also vary in size, with an average enrollment of approximately 280 students in school year 2006-07. While the BIE helps fund 174 schools and 12 dormitories, it does not operate all of them; in the 2006-07 school year, 67 percent of BIE schools were tribally operated under federal contracts or grants (see table 1). Over the past 2 decades, these contracts and grants have transferred the operation of BIE-funded schools to tribes and tribal school boards, offering the potential for tribal groups to take greater ownership of their children’s education. The BIE is organized into two major divisions, with one division located in Albuquerque, New Mexico—called Central Office–West— and the other division located in Washington, D.C.—called Central Office–East. The Central Office–East division conducts research, policy analysis, and planning, and houses the Division of Post Secondary Education, which operates two post-secondary institutions and administers operating grants for 24 colleges operated by tribes and tribal organizations. The BIE performs some functions of a state education agency and receives grants from Education. Further, at the time of our review, BIE Central Office– West had oversight responsibilities for 21 BIE education line offices located in 10 states that provide assistance and/or oversight for the 186 schools and dormitories. Each education line office houses an ELO who functions similarly to a public school district superintendent in managing the schools and providing technical assistance to those schools that tribal groups operate through grants or contracts with the BIE. Throughout this report, we refer to officials from BIE’s Central Offices (East and West) as “BIE officials” and, while we recognize that the ELOs are also BIE officials, we refer to them as “ELOs.” Under NCLBA states are required to establish performance goals and hold their Title I schools accountable for students’ performance by determining whether or not schools have made AYP. The act requires states to set challenging academic content and achievement standards in reading or language arts, mathematics, and science, and determine whether school districts and schools make AYP toward meeting these standards. To make AYP, schools generally must: show that the percentage of students scoring at the proficient level or higher meets the state proficiency target for the school as a whole and for designated student groups, test 95 percent of all students and those in designated groups, and meet goals for an additional academic indicator, such as the state’s graduation rate. NCLBA requires states to establish these performance goals so that all students reach proficiency in reading/language arts, mathematics, and science by 2014. Schools that have not met their states’ performance goals for 2 or more consecutive years are identified for improvement and must implement certain remedial actions that are meant to improve student academic achievement. NCLBA required the Secretary of the Interior to develop a definition of AYP for BIE schools, through negotiated rulemaking. Interior established a No Child Left Behind Negotiated Rulemaking Committee (committee) to develop proposed rules to implement this requirement, among others. By law, the committee was to be comprised of representatives of the federal government and tribes served by BIE-funded schools. The committee held a series of meetings from June 2003 through October 2003 to develop its recommendations. After a public comment period, the final rule was published in April 2005. Under the rule, each BIE school must adopt the academic content standards, assessments, and definition of AYP of the state in which the school is located beginning with the 2005-06 school year. Moreover, if states do not give tribal groups access to their assessments, the tribal groups are obligated to develop alternative definitions of AYP. The regulations do not delineate how to determine AYP in the cases in which schools cannot access state assessments and have not developed an alternative. While NCLBA requires that states’ assessments be aligned with their standards, neither NCLBA nor BIE regulations require that schools’ curricula be aligned with state standards or assessments. Under the Secretary’s definition of AYP—i.e., that of the state in which the school is located—determining the AYP status of the 174 BIE schools requires that BIE officials apply 23 different definitions of AYP. The process is complex because of the many differences in assessments and criteria for AYP determination across the states. For example, some states assess students in additional areas, such as testing students in both reading and language arts. In addition, the complexity of state statistical formulas for calculating AYP also varies among states. Some states’ formulas include multiple confidence bands while other states use none; some states reference students’ improvement over their past performance while others use only current individual performance data on students. Similarly, annual measurable objectives, alternate AYP indicators, and formulas for calculating graduation rates also vary across states. Under NCLBA, tribal governments or school boards (tribal groups) must either adopt the Secretary’s definition of AYP—i.e., that of the state in which they are located—or waive all or part of the definition and propose an alternative. Specifically, tribal groups that waive all or part of the state’s definition of AYP must submit a proposal for an alternative definition of AYP within 60 days of the decision to waive. BIE regulations state that BIE will notify the tribal group within 60 days of receiving the proposed alternative definition whether the proposal is complete and, if complete, an estimated timetable for the final decision. All proposed alternatives are subject to the approval of the Secretaries of Interior and Education, with the tribal groups obligated to use the state’s definition, content standards, and assessments unless the alternative is approved. BIE is required to provide technical assistance upon request, to a tribal group that seeks to develop an alternative definition. Under BIE regulations, a tribal group that requires assistance in developing an alternative must submit a written request to BIE specifying the type of assistance it requires. BIE must acknowledge receipt of the request for technical assistance within 10 days of receiving the request. Within 30 days after receiving the original request for technical assistance, the BIE must identify a point of contact who will immediately begin working with the tribal group. In providing technical assistance to tribal groups in developing alternatives, the BIE can consult with Education. Under BIE regulations, in providing assistance, BIE may use funds provided by Education for assessment-related activities under section 6111 of the ESEA, as amended by NCLBA. According to BIE officials, BIE has used some of these funds on professional development training, development of a reporting system, and improvements to its student information management and tracking systems, which are appropriate uses of these funds. BIE officials stated they used most of these funds to develop BIE’s student information tracking system—the Native American Student Information System. In addition, BIE can use these funds to provide technical assistance to tribal groups in developing AYP alternatives. With respect to achievement under NCLBA, in the 2006-07 school year, BIE reported 51 of the 174 BIE schools made AYP as defined by the states in which the schools are located. Schools that fail to meet AYP for 2 consecutive years must implement remedial actions as required under NCLBA, although the requirements for BIE schools vary from those for public Title I schools (see table 2). For a BIE-operated school, implementation of required remedial actions is the responsibility of the BIE, whereas for schools that are tribally operated through contracts or grants, implementation of remedial actions is the responsibility of the tribal group. Unlike public schools, BIE schools that have an AYP status of school improvement, corrective action, or restructuring are exempt from offering public school choice and supplemental educational services. While the remedial actions applied to public schools and BIE schools under NCLBA may include change in governance, BIE officials told us that there was no provision to implement such a change with retrocession—reverting from grant or contract to BIE-operated status or from BIE-operated to another status—based on continued failure to meet AYP. Almost all of the BIE schools adopted the definition of AYP, content standards, and assessments of the state in which the school is located. While BIE had signed MOUs delineating the terms of accessing and scoring state assessments with 11 of the 23 states in which BIE schools are located, it had not completed MOUs with the other 12, as of April 2008. In addition, BIE experienced some challenges in applying the state definitions to determine whether the 174 schools had met AYP, and some schools, including about half of the schools we contacted, indicated they had not aligned their curricula with the state content standards. BIE officials told us that their schools generally use state definitions of AYP, content standards, annual proficiency goals, and assessments. Therefore, BIE makes AYP determinations for almost all 174 schools using the AYP definition of the state in which the school is located. Using the 23 state definitions of AYP, BIE reported that in 2006-07, 51 of the 174 schools had made AYP, 119 had not, and 4 did not have determinations. BIE officials told us that the AYP determinations were made by applying the criteria filed with Education by the relevant state, except in California and Florida, where BIE schools did not take the state assessment, and in Arizona and North Carolina where there was a data constraint. BIE officials told us that it was challenging to apply the various definitions of AYP and report their determinations to the schools prior to the beginning of the subsequent school year. As of December 2007, 93 of the 174 schools had been notified of their AYP status for school year 2006-07. By March 2008, the number of schools notified had increased to 146. BIE officials told us that, while they were aware that schools should have been notified of their AYP status prior to the beginning of the 2007-08 school year, the delay in notification was prolonged due to staffing issues, as well as schools and states missing deadlines to report assessment data. For example, BIE officials told us that there was a delay getting assessment results for the BIE schools in New Mexico due to a statewide scoring delay. In addition, BIE officials told us that it had been hard to collect attendance data and graduation data needed to make AYP determinations; however, they stated that these data will be more readily available in their new student information system—the Native American Student Information System. BIE officials told us that for the 2006-07 school year, they were unable to apply one feature of Arizona and North Carolina’s new definitions of AYP and made determinations for the 51 schools in Arizona and the 2 in North Carolina using those states’ respective AYP definitions without this new feature. In particular, BIE officials told us that Arizona and North Carolina had recently begun to use a growth model, which BIE was unable to use, as required by the states’ definition of AYP. Some growth models measure individual student progress across time and require a student data system that can link the individual students’ current test scores to those of prior years. BIE officials told us that their new Native American Student Information System has such capabilities, but had not been fully implemented. Officials expressed optimism they would be able to incorporate growth model-based components of AYP in the next round of AYP determinations (2007-08). In addition, BIE officials told us that four schools, two in California and two in Florida, were not administering the state exams. These schools were continuing to administer the standardized tests they had used in prior years. Officials from all four schools told us that their schools had adopted the academic content standards of their respective states, but had not administered the state assessments for different reasons. In these cases, BIE initially made AYP determinations for the 2005-06 school year but has recently suspended the AYP determinations for the four schools until issues regarding how to assess their students are resolved. In terms of content standards, BIE’s ELOs and some school officials told us that while the schools generally have access to state content standards and reported adopting them, some schools have not aligned their curricula to these standards. In particular, 10 of the 21 BIE ELOs stated that some schools in their purview had not aligned their curricula to the state standards for various reasons, including teacher turnover and resistance to change. For example, one ELO told us that some teachers who had been teaching the same material for over 40 years resisted changing the curriculum and preferred to continue to teach as they had been doing for years. Furthermore, officials from at least nine schools we contacted told us that their schools had not fully aligned their curriculum with the state content standards. For example, one school official told us that the school’s elementary reading curriculum was aligned with state content standards, but the elementary science curriculum was not. BIE uses MOUs with states to delineate the terms of BIE-funded schools’ access to the states’ assessment systems; however it had not completed MOUs with 12 of the 23 states, including 5 we visited—Arizona, California, Florida, Mississippi, and New Mexico. The 12 states without signed MOUs enroll about two-thirds of the students in BIE schools. BIE officials told us that in 2005, BIE asked the ELOs to work with state officials to establish MOUs with all 23 states in which BIE schools are located. By March 2006, 11 agreements had been completed, and no new agreements had been completed as of April 2008. The MOUs contain various aspects of administering and scoring the assessment, including delineating responsibilities for state and BIE officials (see table 3). For example, under the MOUs, the state’s responsibilities include inviting BIE school personnel to assessment-related training and informing the BIE of any changes to the state’s AYP definition and assessment system. The BIE’s responsibilities address, among other things, test security to ensure that the contents of the test are not improperly disclosed and proper test administration. BIE officials told us that they did not actively pursue MOU’s with the remaining states, in part because BIE’s leadership had not viewed the completion of the MOUs as a priority—most states were allowing BIE schools to access state assessments and scoring arrangements without such agreements. While BIE schools in 9 of the 12 states without signed MOUs were given access to the state assessments, BIE schools in California and, to a lesser degree in Mississippi, have encountered issues in accessing the state assessments. In particular, California state officials have not given the two BIE schools in California access to the state assessments. State officials in California told us that the state had invested millions of dollars on test development and that a breach in security could undermine the validity of the test. These officials also stated that several entities, including private schools, had requested permission to administer the test and that their approach was to restrict the test to public schools in California. State officials were willing to make an exception for BIE schools to administer the assessment, but requested a $1 million bond for security reasons. BIE and Education officials told us that they were trying to work with the state to resolve the issue. Education officials told us that they were hopeful that a solution, such as having BIE students assessed at public schools, could be worked out. Under BIE regulations, BIE schools without access to their state’s assessment must submit a waiver to develop an alternative definition of AYP. However, officials from the two BIE schools in California stated that developing an alternative definition was unreasonably burdensome and that they had no intention of submitting an alternative assessment in the foreseeable future. The eight BIE schools in Mississippi were able to administer the state assessment in both 2005-06 and 2006-07; however, they were not initially able to access a re-administration of the assessment in 2006-07 that some students needed in order to graduate. Tribal officials explained that they had to sign a special agreement personally guaranteeing the security of the test to administer the test in that instance. State officials and school officials told us that having a signed MOU in place could have expedited access to the test. In addition to concerns regarding test security, state officials we interviewed cited the lack of tribal input as a reason for delaying or rescinding an MOU (see table 4). For example, state officials in Washington told us that when they received the request to sign the MOU, they contacted tribal groups and realized that the tribal groups had been informed of the MOU, but not consulted regarding its details. After consulting with tribal groups, Washington state officials modified the proposed MOU and signed it. In addition, BIE does not currently have a valid MOU with New Mexico because the Governor of New Mexico suspended the state’s MOU with BIE shortly after signing it, in part because tribal groups indicated that they had not been consulted about the terms of the MOU. Officials from three tribal groups—the Navajo Nation, OSEC, and the Miccosukee Tribe of Indians—have informed BIE officials that they wish to pursue alternatives to state AYP definitions for a variety of reasons, including the desire to ensure that standards and assessments include components of native culture. However, the remaining tribal groups have not indicated that they will waive state definitions of AYP, in an effort to maintain compatibility with public schools or because of potential challenges to developing alternatives. According to ELOs and the school officials we interviewed, there are significant potential challenges involved in developing alternatives, as well as advantages to using the state assessments, including compatibility with public schools. As of March 2008, three tribal groups—Navajo Nation, OSEC, and Miccosukee—had formally notified the BIE of their intent to develop alternatives to state definitions of AYP. These tribal groups represent BIE- funded schools in five states and include about 44 percent of BIE students (see table 5). The tribal groups began the process of developing alternatives at different times, but all were still in the early stages of doing so. Officials from the Navajo Nation, with BIE schools in three states, have requested technical assistance for developing an alternative definition of AYP, citing the desire to include cultural components in the standards and assessments and to compare the progress of Navajo students across states. Navajo officials told us that they currently do not have a consistent method of measuring the academic progress of their students across the states in which they are enrolled. Navajo officials have recently (October 2007) requested technical assistance from BIE in their effort to develop an alternative to the relevant states’ definition of AYP. In their proposal to BIE, Navajo officials stated that while they are willing to work with existing assessment procedures as much as possible, they were seeking to develop a “Navajo specific” measure that would influence AYP determination, regardless of the state. OSEC, a consortium of tribal groups including representatives from 11 BIE-funded schools in South Dakota, has also requested technical assistance as it seeks to develop an alternative definition of AYP, primarily to improve student performance in its schools and to more accurately reflect the length of time it takes some students to graduate. First, it plans to define graduation rates differently from the state. In particular, South Dakota uses a 4-year window to determine graduation rates. OSEC officials told us that a definition of graduation rate that included those who successfully completed high school within 6 years would more accurately reflect the reality that many students take more than 4 years to graduate. In addition, OSEC officials told us that they wanted to replace the attendance component of the state’s definition of AYP with a language and culture component. Furthermore, OSEC would like to develop standards and assessments for its students in subject areas currently covered by the state assessment, such as reading, math, and science. To this end, the consortium has submitted a proposal to BIE officials that provides a framework for developing academic content standards for math, reading, and science, as well as developing an assessment. OSEC officials consulted with BIE officials regarding the proposal, and BIE has since forwarded the proposal to Education for review. Education officials met with officials from BIE and OSEC in November 2007 to evaluate OSEC’s needs and offer technical assistance. Education officials told us that they have a consultant who could help OSEC ensure that the new standards and assessments meet Education’s guidelines. Officials from the Miccosukee Tribe have informed BIE that, while they have aligned their curriculum to Florida’s academic content standards, they do not intend to administer the Florida state assessment system in their school. Miccosukee tribal officials explained that they did not want to implement the Florida assessment system because they thought it was flawed and inferior to the Terra Nova—the standardized test they were already using. They also told us that because attendance in the Miccosukee School was not compulsory, they rejected the use of attendance as an additional AYP indicator. After having met with Education officials and a consultant, the Miccosukee told us that they were considering various options in their development of an alternative assessment, including augmenting the Terra Nova or developing a new assessment based on a modified version of Florida’s academic content standards. Officials also told us that they were working on developing standards for Miccosukee culture and language to implement an assessment that would serve as the additional AYP indicator in lieu of attendance for their students in third through eighth grade. Officials representing BIE schools in California, Mississippi, and Washington told us that it was important that their schools be compatible with the local public schools. For example, officials from the BIE schools in Mississippi told us that they wanted their students to take the same tests as students attending Mississippi public schools, in part to ensure that they received the same diploma. In addition, officials from one California school explained that their students come from public schools and may return to public schools in high school. These officials told us that it made more sense for the students to take the state tests for continuity. In addition, BIE school officials in California, Mississippi, and Washington told us that because they followed the state curriculum, it would be logical to administer the state assessment. However, while the tribal groups representing the eight BIE schools in Washington have not waived the state definition of AYP, they have proposed a technical change that would affect how BIE officials determine AYP for these schools. In particular, BIE considers the 2002-03 school year as the baseline for its AYP determinations; however, officials representing the BIE schools in Washington told us that the 2005-06 school year is a more appropriate baseline, as it is the first year in which they administered the state assessment for AYP purposes. While the Washington state superintendent approved the schools’ request to change the baseline school year, BIE officials have not done so. As a result, officials representing one of the schools challenged BIE’s AYP determination for the 2005-06 school year. School officials and education line officers identified several potential challenges that tribal groups might encounter in their efforts to develop alternative standards or assessments, including a lack of expertise, funding, and time (see table 6). According to ELOs and school and Education officials, the specialized knowledge needed to develop an alternative definition of AYP is generally beyond the capacity of tribal groups. For example, ELOs and Education officials stated that the technical expertise needed to develop an assessment was not available among members of some tribes and would need to be obtained through consultant contracts. School officials from Mississippi and Washington agreed that developing such alternatives would require expertise beyond that available within their tribal groups. With regard to financing the development of alternatives, Education officials stated that developing standards and assessments could cost tens of millions of dollars—financial resources that are generally not available among many tribal groups for this purpose. Education officials and ELOs also agreed that developing alternatives requires an extensive time commitment that may not be sustainable given changes in leadership. In particular, Education officials told us that developing, piloting, and testing alternative content standards or assessments can take from 12 months to 3 years. Some of the education line officers we interviewed volunteered that the required time commitment could affect support for such a project. Five of the ELOs and two school officials specifically noted that the time needed to develop an alternative would be a challenge for their tribal groups or school boards, with one school official citing the time commitment needed to help teachers understand and incorporate alternative standards into their lesson plans. One school official stated that changes in BIE leadership had led to different interpretations of how to implement the NCLBA provision related to developing alternatives. Most tribal groups, school officials, and ELOs we spoke with said they had little guidance about the process BIE uses to help tribal groups develop alternatives. In addition, school officials and tribal groups we interviewed reported communication problems with BIE, including lengthy delays and a lack of response. Recently, however, BIE and Education officials have offered both technical assistance and funds to those tribal groups seeking to develop alternatives. Most tribal groups, ELOs, and school officials we spoke with said they had received little guidance about the process BIE uses to help tribal groups develop alternatives. Officials representing the two tribal groups and one consortium that have formally requested technical assistance stated they were uncertain about the BIE process for applying for an alternative. Likewise, we found school officials were also unsure of BIE’s process for applying for an alternative. For example, officials from the two BIE schools in California said they had no knowledge of the BIE process to assist tribal groups and school boards to develop alternatives. In addition, officials from one school said they hired legal counsel to assist them with their request because they were uncertain about BIE’s process for applying for an alternative. About half of the ELOs, despite being the first point of contact, told us they did not have enough information to accurately describe the process a tribal group would use to waive the Secretary of the Interior’s definition and pursue development of an alternative definition of AYP. This may be at least partly due to turnover among ELOs. Eight of the 21 ELOs said they had been in their current position for 12 months or less while 7 had been in their current position from 1 to 3 years. BIE officials told us that about 25 percent of the ELOs who attended training on the process to develop an alternative were no longer employed in that position. According to BIE officials, ELOs had received such training in 2005—although no requested documentation of this training and guidance was provided to us. During the course of our review, 19 of the 21 ELOs we interviewed also stated they had not received any training or written guidance on the BIE’s policy for approving a tribal groups’ request for an alternative, even though providing technical assistance to tribal groups developing an alternative is included in their job responsibilities. During our interviews, 11 of the 21 ELOs indicated they were knowledgeable about the NCLBA provision that allows tribal groups to waive the Secretary’s definition and develop an alternative and would be able to describe the provision to tribal groups. During our interviews, almost all of the ELOs (19 of 21) told us that they had not received any information from BIE officials on their role in providing technical assistance to tribes in developing content standards, assessments, or definitions of AYP. As a result, most tribal groups have not received any information from ELOs on the availability of technical assistance for developing alternatives. In particular, only 3 of the 21 ELOs stated they had provided any information on the availability of technical assistance for developing alternatives to tribal groups within their jurisdiction. BIE receives funds from Education that could be used to assist tribal groups with the development of alternatives, but BIE’s ELOs told us they had not been instructed that BIE funds were available for this purpose. All 21 ELOs told us they had not received any guidance from BIE on BIE funds that might be used to assist tribal groups seeking to develop alternatives. Some school officials and tribal groups we interviewed reported a lack of response from the BIE or lengthy delays in responding to requests for assistance related to development of alternative standards, assessments, or definitions of AYP. For example, OSEC’s written request for technical assistance in developing an alternative definition of AYP was not acted upon for 8 months. In another case, the Miccosukee’s written request to waive the state assessment and develop an alternative went unanswered by the BIE from October 2006 to June 2007. BIE officials, in acknowledging their slow response to the tribal groups’ requests for technical assistance, stated that in some cases tribal groups’ written requests were not always clear about what they wanted from the BIE or had not adhered to the regulation that requires the waiver request be submitted by either a tribal governing body or school board. Other tribal groups we interviewed reported frustration in communicating with BIE due to BIE’s failure to proactively initiate communication when necessary. For example, officials from one of the BIE schools in California stated that, although BIE officials were aware that the state had not given the schools access to the state assessment, BIE had not communicated with or offered any type of assistance to the schools. Further, OSEC submitted to BIE a written request for guidance and funds to pay for the development of assessment tools on developing an alternative definition of AYP. In its response, the BIE denied the consortium’s written request without further discussion or inquiry, noting that the request did not come from either a school board or a tribal governing body but rather a consortium of schools. BIE officials told us that their prior focus had been on ensuring that BIE schools were accessing and using the state standards and assessments and therefore did not devote resources to assist those tribal groups who sought to develop alternatives to the state systems. In addition, BIE officials told us that BIE had not initially been proactive in working with Education on issues related to alternative assessments. To address tribal groups’ requests for technical assistance, BIE assigned a staff person as the primary BIE contact for tribal groups that are requesting technical assistance or seeking to develop alternatives. However, this BIE staff person has several other key responsibilities including responsibility for applying 23 state AYP definitions to calculate the AYP status of BIE schools and responsibility for overseeing the special education program for all BIE schools. In addition, BIE officials informed Education officials in September 2007 of the OSEC and Miccosukee’s requests for technical assistance, and in November 2007 of the Navajo’s request for technical assistance. In response to the requests, BIE and Education officials have recently offered technical assistance to those tribal groups that are seeking to develop alternatives. For example, officials from BIE and Education met with the Miccosukee and OSEC in November 2007 to assess the type of technical assistance needed in order for the tribe to pursue development of its alternative. Likewise, officials from BIE and Education also met with the Navajo Nation in March 2008 to assess their technical assistance needs as they continue to pursue development of an alternative. In addition to identifying the types of technical assistance needed by those tribal groups that have formally submitted a request to waive state standards, assessments, or definitions of AYP, Education officials told us they have also sent a contractor to assist tribal groups as they pursue the development of alternative assessments. Specifically, in Florida, the Education contractor is charged with helping the Miccosukee to identify the steps needed to ensure its assessment complies with relevant regulations under NCLBA by reflecting Florida’s state standards—or any modified standards that the Miccosukee may adopt. Similarly, in South Dakota, the Education contractor is charged with working with the OSEC consortium to identify the actions needed to ensure that its alternative assessment will comply with NCLBA regulations. As of February 2008, according to BIE officials, none of the funds provided by Education to BIE under the NCLBA provision supporting assessment- related expenses had been spent to provide technical assistance to tribal groups seeking to develop alternatives. The BIE reported receiving from Education a total of $11.7 million for school years 2002-03 through 2007-08, that was targeted to assessment-related expenses. According to BIE, all of these funds had been obligated, primarily for improvements to BIE’s student information and tracking systems and other assessment-related uses, including professional development. In fact, some tribal groups told us they were not aware that BIE received funds that might be available to assist with development of alternatives. BIE officials stated that none of these funds had been spent on technical assistance, but said that they expected to spend some funds to provide technical assistance in the near future. In most cases, BIE schools that wish to adopt their state’s definition of AYP, standards, and assessments, have had no problems doing so, but the lack of MOUs between BIE and some states exposes the BIE schools in those states to the potential risk of losing access to state assessments. Under the existing MOUs, the state (or BIE) may terminate the agreement, but notice is required. Additionally, the MOU ensures that tribes’ access to tests is not dependent on decisions made by particular state officials or administrations, who could otherwise terminate or impose conditions on the sharing arrangements without notice. In part because BIE may have little leverage in negotiating with state education departments, BIE may encounter difficulty in reaching agreement on these MOUs, especially if a state imposes challenging conditions. In addition, a large burden is placed on tribal groups and schools that lack access to state assessments—in terms of developing an alternative assessment that meets federal guidelines. Without prompt assistance, such schools may lack appropriate measures of what children know and can do that could support plans for educational improvement. Similarly, lack of alignment between curricula and standards or inability to promptly produce determinations of performance can slow the pace of improvement for students and schools. Clearly, if tribal groups wish to propose an alternative, they must understand the process in place to pursue this option. Developing alternatives requires clear and timely communication between BIE and tribal groups, as well as between BIE and Education. To date, guidance from BIE on developing alternatives has been limited and BIE’s communication with tribal groups, BIE ELOs, and Education has been slow or lacking. Without improved guidance to tribal groups and ELOs, those tribal groups seeking to develop alternatives may lack information or receive inaccurate information about how to develop an acceptable alternative definition. Further, unless BIE establishes response time frames and processes, the communication between BIE and those tribal groups seeking alternatives will remain ineffective. As a result, these tribal groups could continue to view BIE as a hindrance rather than a partner in the process. While BIE and Education have recently begun offering technical assistance, clear guidance from BIE and timely communication between BIE and tribal groups could not only improve working relations, but also facilitate the use of the provision allowing the alternatives to address the unique cultural needs of the students. To improve support for tribal governments and school boards in their adoption of definitions of AYP, we are making the following four recommendations. We recommend that the Secretary of the Interior direct BIE to: Coordinate with relevant tribal groups in pursuing negotiation of MOUs with states that lack them, seeking facilitation from Education when necessary and appropriate. In close coordination with Education, provide prompt assistance to tribal groups in defining assessment options, especially in instances in which tribal groups are not accessing state assessments. Such assistance could include delineating options—such as using an already established assessment, augmenting an assessment, or incorporating cultural components as an additional academic indicator—and their associated costs. Provide guidelines and training on the process for seeking and approving alternatives to all tribal governments, tribal school boards, and education line offices. Establish internal response time frames and processes to ensure more timely responses to all correspondence with tribal groups as well as proactive communication with tribal groups and Education to resolve issues related to waivers, requests for technical assistance, and development of alternative definitions of AYP. We provided a draft of this report to Interior and Education for review and comment. Interior provided a written response to the report (see app. I); Education did not. Both agencies provided technical comments, which we incorporated in the report where appropriate. Interior agreed with all of our recommendations. In responding to our first recommendation, Interior explained that BIE is continuing to work jointly with Education to facilitate agreements with the states to ensure access to state assessments and to establish MOUs with those states where none currently exist. With respect to our second recommendation, Interior reported that the BIE has established a Scope of Work that addresses the full range of technical assistance needed to assist tribal groups that seek to waive all or part of the state’s definition of AYP, content standards, or assessments. In regard to our third recommendation, BIE stated that, in addition to continuing to provide guidance and training to tribal groups and tribal school boards, it has developed information on the process for seeking and approving alternatives that will be posted on its Web site as well as distributed to tribal groups and tribal school boards. In responding to our final recommendation, BIE stated it would continue to be more proactive in its communication with tribal groups and Education to resolve issues related to waivers of the state’s definition of AYP, requests for technical assistance, and development of alternative definitions of AYP. Moreover, as part of the project management with tribal entities that have sought technical assistance, a consultant will maintain a management document that identifies timelines, among other things. In addition to the steps BIE has mentioned, we continue to believe it is important for BIE to establish internal timelines to ensure more timely responses to all correspondence with tribal groups. We are sending copies of this report to the Secretaries of Education and the Interior; the Director of the Bureau of Indian Education; representatives of tribal groups identified in the report, relevant congressional committees, and other interested parties. In addition, the report will be available at no charge on GAO’s Web site at http://www.gao.gov. Please contact me on (202) 512-7215 if you or your staff have any questions about this report. Contact points for our offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix II. Betty Ward-Zukerman, Assistant Director, and Nagla’a El-Hodiri, Analyst- in-Charge, managed this assignment. Kris Trueblood and Tahra Nichols made significant contributions to all aspects of the work. Nora Boretti, Kimberly Granger, Angela Jacobs, Annamarie Lopata, and Sara Pelton assisted with data collection. Charlie Willson and Jessica Orr provided assistance in report preparation; Jeffery Malcolm provided expertise on Indian issues; James Rebbe and Doreen Feldman provided legal support; Jean McSween and John Mingus provided technical support; and Lise Levie verified our findings. No Child Left Behind Act: Education Actions Could Improve the Targeting of School Improvement Funds to Schools Most in Need of Assistance. GAO-08-380. Washington, D.C.: February 29, 2008. No Child Left Behind Act: States Face Challenges Measuring Academic Growth That Education’s Initiatives May Help Address. GAO-06-661. Washington, D.C.: July 17, 2006. Bureau of Indian Affairs Schools: Expenditures in Selected Schools Are Comparable to Similar Public Schools, but Data Are Insufficient to Judge Adequacy of Funding and Formulas. GAO-03-955. Washington, D.C.: September 4, 2003. Bureau of Indian Affairs Schools: New Facilities Management Information System Promising, but Improved Data Accuracy Needed. GAO-03-692. Washington, D.C.: July 31, 2003. Title I: Characteristics of Tests Will Influence Expenses; Information Sharing May Help States Realize Efficiencies. GAO-03-389. Washington, D.C.: May 8, 2003.

The No Child Left Behind Act (NCLBA) requires states and the Department of the Interior's Bureau of Indian Education (BIE) to define and determine whether schools are making adequate yearly progress (AYP) toward meeting the goal of 100 percent academic proficiency. To address tribes' needs for cultural preservation, NCLBA allows tribal groups to waive all or part of BIE's definition of AYP and propose an alternative, with technical assistance from BIE and the Department of Education, if requested. GAO is providing information on the extent of (1) BIE schools' adoption of BIE's definition of AYP; (2) tribal groups' pursuit of alternatives and their reasons as well as reasons other tribal groups have not done so; and (3) federal assistance to tribal groups developing alternatives. To obtain this information, GAO interviewed tribal groups, federal officials, and state education officials; conducted site visits to BIE schools; and reviewed laws, regulations, and other relevant documents. Although almost all of the 174 BIE schools have officially adopted BIE's definition of AYP--the definition of AYP of the state where the school is located--BIE had not yet completed memoranda of understanding (MOU) to delineate BIE and state responsibilities concerning BIE schools' access to the states' assessment systems for 12 of the 23 states with BIE schools. Without MOUs, states could change their policies regarding BIE schools' access to assessments and scoring services. Officials from the Navajo Nation, the Oceti Sakowin Education Consortium, and the Miccosukee Tribe have begun to develop alternatives to state AYP definitions, in part to make standards and assessments reflect their culture, while officials of other tribal groups have cited various reasons for not developing alternatives. The three tribal groups developing alternatives, representing about 44 percent of the 48,000 BIE students, have requested technical assistance in developing their alternatives. Other tribal officials cited a desire to maintain compatibility with public schools and/or cited challenges, such as a lack of expertise, as reasons not to pursue alternatives. The three tribal groups pursuing alternatives reported a lack of federal guidance and communication, although they have recently received some initial technical assistance from BIE and Education officials. These tribal groups reported receiving little guidance from BIE and difficulties in communicating with BIE because the Bureau did not always have internal response timelines or meet the ones it had. Moreover, BIE education line officers--the primary points of contact for information on the alternative provision--generally indicated that they had received no guidance or training on the provision. During the course of this review, BIE and Education officials began offering technical assistance to the tribal groups working to developalternatives

Because DOD is one of the largest and most complex organizations in the world, overhauling its business operations represents a huge management challenge. In fiscal year 2004, DOD reported that its operations involved $1.2 trillion in assets, $1.7 trillion in liabilities, over 3.3 million military and civilian personnel, and over $605 billion in net cost of operations. For fiscal year 2005, the department received appropriations of about $417 billion. Execution of DOD’s operations spans a wide range of defense organizations, including the military services and their respective major commands and functional activities, numerous large defense agencies and field activities, and various combatant and joint operational commands that are responsible for military operations for specific geographic regions or theaters of operation. To support DOD’s operations, the department performs an assortment of interrelated and interdependent business processes, including logistics management, procurement, health care management, and financial management. Transformation of DOD’s business systems and operations is critical to the department providing Congress and DOD management with accurate and timely information for use in the decision-making process. This effort is an essential part of the Secretary of Defense’s broad initiative to “transform the way the department works and what it works on.” Secretary Rumsfeld has estimated that successful improvements to DOD’s business operations could save the department 5 percent of its budget a year, which equates to over $20 billion a year in savings. For several years, we have reported that DOD faces a range of financial management and related business process challenges that are complex, long-standing, pervasive, and deeply rooted in virtually all business operations throughout the department. As the Comptroller General testified in November 2004, DOD’s financial management deficiencies, taken together, continue to represent the single largest obstacle to achieving an unqualified opinion on the U.S. government’s consolidated financial statements. To date, none of the military services has passed the test of an independent financial audit because of pervasive weaknesses in internal controls and processes and fundamentally flawed business systems. In identifying improved financial performance as one of its five governmentwide initiatives, the President’s Management Agenda recognized that obtaining a clean (unqualified) financial audit opinion is a basic prescription for any well-managed organization. At the same time, it recognized that without sound internal controls and accurate and timely financial and performance information, it is not possible to accomplish the President’s agenda and secure the best performance and highest measure of accountability for the American people. The Joint Financial Management Improvement Program (JFMIP) Principals have defined certain measures, in addition to receiving an unqualified financial statement audit opinion, for achieving financial management success. These additional measures include (1) being able to routinely provide timely, accurate, and useful financial and performance information; (2) having no material internal control weaknesses or material noncompliance with laws and regulations; and (3) meeting the requirements of the Federal Financial Management Improvement Act of 1996 (FFMIA). DOD does not meet any of these conditions. In September 2004, the DOD Comptroller identified 11 major deficiencies that would affect the department’s ability to prepare accurate and reliable financial statements for fiscal year 2004. Subsequently, the DOD Inspector General issued a disclaimer of opinion on DOD’s fiscal year 2004 financial statements, citing material weaknesses in internal controls and noncompliance with FFMIA requirements. Pervasive weaknesses in DOD’s financial management and related business processes and systems have (1) resulted in a lack of reliable information needed to make sound decisions and report on the status of DOD activities, including accountability of assets, through financial and other reports to Congress and DOD decision makers; (2) hindered its operational efficiency; (3) adversely affected mission performance; and (4) left the department vulnerable to fraud, waste, and abuse, as the following examples illustrate. Mobilized Army National Guard soldiers have experienced significant problems getting accurate, timely, and consistent reimbursement for out-of-pocket travel expenses. These weaknesses were more glaring in light of the sustained increase in mobilizations for Army National Guard soldiers over the last 3 years. Our case study units experienced a broad range of travel reimbursement problems, including disputed amounts for meals that remained unpaid at the end of our review and vouchers that were submitted five or more times before being paid. One of the primary causes for these problems is rooted in the paper-intensive process used by DOD to reimburse Army National Guard soldiers for their travel expenses. Manual processes and nonintegrated pay and personnel systems affect the Army’s ability to generate timely active duty medical extension orders and ensure that soldiers are paid correctly. The current stovepiped, nonintegrated systems are labor-intensive and require extensive error-prone manual entry and reentry. The inadequate control resulted in some soldiers being removed from active duty status in the automated systems that control pay and access to benefits, including medical care. In addition, because these soldiers no longer had valid active duty orders, they did not have access to the commissary and post exchange—which allows soldiers and their families to purchase groceries and other goods at a discount. In one case we reviewed, during a 12-month period, while attempting to obtain care for injuries sustained from a helicopter crash in Afghanistan, one Special Forces soldier fell out of active duty status four times. During the times he was not recorded in the system as being on active duty, he was not paid and he and his family experienced delays in receiving medical treatment. In all, he missed payments for 10 pay periods—totaling $11,924. The processes and automated systems relied on to provide active duty payments to mobilized Army Reserve soldiers are so error-prone, cumbersome, and complex that neither DOD nor, more importantly, the Army Reserve soldiers themselves could be reasonably assured of timely and accurate payments. Specifically, at eight Army Reserve units that we reviewed, 332 of 348 soldiers (95 percent) experienced at least one problem with the active duty pay and allowances they were entitled to receive. Many of the soldiers experienced multiple pay problems associated with their active duty mobilizations. Some of the pay problems soldiers experienced often lingered unresolved for considerable lengths of time, some for over a year. Of the $375,000 in active duty pay and allowance problems identified in our case studies, the majority were overpayments. We referred one individual for criminal investigation because he did not mobilize with his unit, but he erroneously received over $36,000 in active duty pay and did not report this overpayment. We also identified 294 soldiers who were underpaid a total of about $51,000 in active duty pay and allowances. Transformation of DOD’s business systems and operations is critical to the department having the ability to provide Congress and DOD management with accurate and timely information for use in the decision-making process. One of the key elements we have reported as necessary to successfully execute the transformation is establishing and implementing a business enterprise architecture (BEA). In this regard, the department has undertaken a daunting challenge to modernize its existing business systems environment through the development, maintenance, and implementation of the BEA, or modernization blueprint. As previously noted, the department has designated six domain owners to be responsible for implementing the BEA. The importance of developing, maintaining, and implementing an enterprise architecture is a basic tenet of both organizational transformation and IT management. Managed properly, an enterprise architecture can clarify and help optimize the interdependencies and relationships among an organization’s business operations and the underlying IT infrastructure and applications that support these operations. Employed in concert with other important management controls, such as portfolio-based capital planning and investment control practices, architectures can greatly increase the chances that organizations’ operational and IT environments will be configured to optimize mission performance. Our experience with federal agencies has shown that investing in IT without defining these investments in the context of an architecture often results in systems that are duplicative, not well integrated, and unnecessarily costly to maintain and interface. A key element of an enterprise architecture is the development and implementation of a transition plan. According to relevant guidance and best practices, the transition plan should provide a road map for moving from the “As Is” to the “To Be” environment. An important step in the development of a well-defined transition plan is an analysis that compares the “As Is” and “To Be” architectures to identify differences. Options are explored and decisions are made regarding which legacy systems to retain, modify, or retire, and which new systems either to introduce on a temporary basis or to pursue as strategic solutions. Accordingly, transition plans identify legacy, migration, and new systems, and sequence them to show, for example, the phasing out and termination of systems and capabilities and the timing of the introduction of new systems and capabilities. Furthermore, they do so in light of resource constraints, such as budget, people, acquisition/development process maturity, and associated time frames. To improve DOD’s control and accountability over business systems investments, Congress included provisions in the Ronald W. Reagan National Defense Authorization Act for Fiscal Year 2005. The act directs that a management structure be put in place that (1) makes domains responsible for the control and accountability over business systems investments, (2) requires domains to establish a hierarchy of investment review boards from across the department, and (3) directs the boards to use a standard set of investment review and decision-making criteria to ensure compliance and consistency with the BEA. The act also directs the establishment of a departmental review board called the Defense Business Systems Management Committee that is to be chaired by the Deputy Secretary of Defense. Until DOD has complete, reliable information on the costs and number of business systems operating within the department, its ability to effectively control the money it spends on these systems will be limited. DOD’s fiscal year 2005 budget request for its business systems was $13.3 billion, which on its face is about $6 billion, or 29 percent, less than its fiscal year 2004 budget request. We found that this decrease can be attributed to DOD’s reclassification of some business systems to national security systems, not to a reduction in spending on its systems. While some of the reclassifications appeared reasonable, our analysis showed that others were questionable or inconsistencies exist, which hinders DOD’s ability to develop a definitive business systems inventory. DOD CIO officials acknowledged that there were inconsistencies in the classifications of systems as either business systems or national security systems and stated that they were working to improve the classification criteria. DOD CIO officials also stated that they are working toward a single system repository for the department that will include information to be used in developing the budget. At the same time the amount of requested business system funding declined, the reported number of business systems increased by about 1,900—from 2,274 in April 2003 to 4,150 in February 2005. Given the ever-changing numbers—for both funding and systems— the department continues to lack reasonable assurance that all business systems are included within the reported inventory and that all business system funding is accurately detailed in the budget. For fiscal year 2005, DOD requested approximately $28.7 billion in IT funding to support a wide range of military operations as well as DOD business systems operations. Of the $28.7 billion, our analysis showed that about $13.3 billion was for business applications and related infrastructure. Of the $13.3 billion, our analysis of the budget request disclosed that about $8.4 billion was for infrastructure and related costs. Business applications include activities that support the business functions of the department, such as personnel, health, travel, acquisition, finance and accounting, and logistics. The remaining $15.4 billion was classified as being for national security systems. Of that amount, our analysis ascertained that about $7.5 billion was for infrastructure and related costs. Of the $13.3 billion, $10.7 billion was for the operation and maintenance of the existing systems and $2.6 billion was for the modernization of existing systems, the development of new systems, or both. The Office of Management and Budget requires that funds requested for IT projects be classified as either steady state (referred to by DOD as “current services”) or as development/modernization. Current services funds are to be for operating and maintaining systems at current levels (i.e., without major enhancements), while development/modernization funds are to be for developing new IT systems or making major enhancements to existing systems. Table 1 shows the distribution, by DOD component, of the reported $13.3 billion between current services and modernization funding. As an example of how a component applies funding for current services and modernization, the budget request for the Standard Procurement System, which is one of the department’s standard systems, will use a combination of current services and modernization funding. The fiscal year 2005 budget request for the Standard Procurement System identified about $56 million in total—$30 million for current services and $26 million for development and modernization. Incorrect system classification hinders the department’s efforts to improve its control and accountability over its business systems investments. For instance, the incorrect reclassification of business systems to national security systems precludes scrutiny by the business domains, including the process utilized to obtain the DOD Comptroller’s determination that authorizes the components to obligate amounts over $1 million for the improvement of financial management systems. Our comparison of the fiscal years 2004 and 2005 budget requests disclosed that DOD reclassified 56 systems in the fiscal year 2005 budget request from business systems to national security systems. The net effect of the reclassifications was a decrease of approximately $6 billion in the fiscal year 2005 budget request for business systems and related infrastructure. The reported amount declined from about $19 billion in fiscal year 2004 to over $13 billion in fiscal year 2005. In some cases, we found that the reclassification appeared reasonable. For example, in the fiscal year 2005 budget request, the Defense Message System was classified as a national security system. In our May 2004 report, we noted the inconsistent classification of this system among the military services. The Navy classified the Defense Message System as a business system but the Army and Air Force classified it as a national security system. Similarly, the reclassification of the Defense Information System Network initiative as a national security system appeared reasonable. For example, the Defense Information System Network is used to provide a secure telecommunication network—voice, data, and video— to the President, the Secretary of Defense, the Joint Chiefs of Staff, and military personnel in the field. These two systems account for over $2.6 billion of the $6 billion. However, our analysis of the 56 systems that were reclassified as national security systems also identified instances for which the reclassification was questionable. For example, Base Level Communication Infrastructure—initiative number 254—for several DOD entities was shown as a national security system in the fiscal year 2005 budget request. Our review of the fiscal year 2005 budget found that within the Air Force there were numerous other initiatives entitled Base Level Communication Infrastructure that were classified as business systems, not national security systems. The nomenclature describing these different initiatives was the same. Therefore, it was difficult to ascertain why certain initiatives were classified as national security systems while others, with the same name, were classified as business systems. The following are examples of the inconsistencies in the department’s classification of systems. The Joint Total Asset Visibility System had been classified as a business system since at least fiscal year 1999 and was identified as a business system in the department’s original list of 2,274 business systems in April 2003 and is still being reported as a business system in the department’s systems inventory. We found nothing in our review of the fiscal year 2005 budget request that warranted a change in classification. The Transportation Coordinators’ Automated Information for Movements System II was identified as a business system in the department’s original inventory of business systems in April 2003 and has been shown as a business system in the budget request since 1999. Furthermore, BMMP currently reports it as one of the department’s business systems. Yet it was reclassified as a national security system in the fiscal year 2005 budget request even though its functionality had not changed. According to the program management office, the DOD CIO classified the system as a national security system, but the program management office was informed by the DOD Comptroller that the system needed to be submitted to the DOD Comptroller for review in accordance with section 1004(d) of the fiscal year 2003 defense authorization act. However, the section of the act relating to the DOD Comptroller review does not apply to national security systems. In our May 2004 report, we noted that the Army reported that it obligated almost $22 million for modernization of the system in fiscal year 2003 without its having been subject to review by the DOD Comptroller for consistency with DOD’s BEA. The Army’s Global Combat Support System was reclassified as a national security system in the fiscal year 2005 budget, even though it is still being reported as a business system in the department’s current inventory. Furthermore, BMMP officials stated that reclassification was incorrect and the system was submitted to the DOD Comptroller for review in accordance with section 1004(d) of the fiscal year 2003 defense authorization act. The DOD Comptroller approved investment in the system in August 2003 and January 2004. In addition, this is the first year in which the Navy ERP was listed in the budget and incorrectly classified as a national security system. Its forerunners, four pilot ERP projects, have been classified as business systems since their inception. DOD officials were not able to provide a valid explanation why the program was classified as a national security program. For the fiscal year 2006 budget request, the Navy has requested that the DOD CIO reclassify the program from a national security system to a business system. The misclassification of systems in DOD’s budget hinders the department’s ability to develop a comprehensive list of its business systems, thereby affecting its ability to develop a well-defined “As Is” component of its BEA and a viable transition plan. In addition, improper classification diminishes Congress’s ability to effectively monitor and oversee the billions of dollars spent annually to maintain, operate, and modernize the department’s business systems environment. DOD CIO officials acknowledged that there were inconsistencies within the department in the classification of a system as either a business system or a national security system, which hinder its ability to develop a single comprehensive business systems inventory. DOD CIO officials also stated that with the implementation of the DOD Information Technology Portfolio Data Repository (DITPR)—the new systems inventory repository—they are working to improve the criteria governing the classification of systems, and these criteria are being incorporated into the new systems inventory repository. As discussed in our May 2004 report, DOD has several databases that contain information on its systems—business and national security—and these databases have not been reconciled and therefore inconsistencies exist. We recommended that the department establish a single database for its inventory of business systems. Subsequently, on July 13, 2004, the DOD CIO directed establishment of the DITPR. According to BMMP officials, the department is working toward one database that will contain information for use in developing the annual budget request and will be considered the single system repository for the department. Additionally, all existing databases will be eliminated. Furthermore, the DITPR will automatically classify each system based on selected information entered for each. For example, all intelligence, science and technology, and logistics-warfighting systems are to be classified as national security systems. In addition, to help identify and properly categorize its business systems, we recommended in May 2004 that DOD develop a standard definition for business systems. In July 2004, DOD established a definition of a system, but not specifically of a business system. DOD broadly defined a system as a set of information resources organized for the collection, storage, processing, maintenance, use, sharing, dissemination, disposition, display, or transmission of information. DOD’s definition also clarifies what should not be reported as a system. For example, it excludes such things as commercial office automation packages, information assurance initiatives, and architecture initiatives. Additionally, DOD’s definition is aimed at identifying all systems that would obligate $1 million or more for modernization in any year of the department’s 5-year defense plan. While DOD did not develop a specific definition of a business system, the fiscal year 2005 defense authorization act subsequently provided a definition. The act defines a defense business system as an information system, other than a national security system, operated by, for, or on behalf of the department that is used to support business activities, such as acquisition, financial management, logistics, strategic planning and budgeting, installations and environment, and human resources management. The act states that such systems are to include financial systems, mixed systems, financial data feeder systems, and IT and information assurance infrastructure. It is incumbent upon the department to ensure that the definition specified in the act is used consistently throughout the department and becomes the basis for entering information into the DITPR. As previously discussed, the accurate and complete identification of DOD’s business systems is crucial for developing a credible “As Is” component and transition plan for the BEA and ensuring that obligations for modernizations are reviewed and approved as required by the act. Furthermore, such information is needed in order to provide complete and accurate data to Congress for use in monitoring the department’s business systems investments. The department’s reported number of business systems continues to fluctuate, and DOD does not yet have reasonable assurance that the currently reported number of business systems is complete. As of February 2005, DOD reported that its business systems inventory consisted of 4,150 systems, which is an increase of approximately 1,900 reported business systems since April 2003. Table 2 presents a comparison of the April 2003 and February 2005 reported business systems inventories by domain. The largest increase is due to the logistics domain increasing its reported inventory of business systems from 565 in April 2003 to the current 2,005. We reported in May 2004 that the logistics domain had validated about 1,900 business systems but had not yet entered most of them into the BMMP systems inventory. Logistics domain officials informed us that they completed that process and this increase was the result. According to the logistics domain officials, in making this determination, they considered an initiative as a business system if it (1) is used by at least 50 people, (2) costs at least $50,000 annually to operate, and (3) runs on a network. The criteria used by the logistics domain are stricter than those developed by the DOD CIO. As previously noted, the department needs to ensure that the same criteria are used by all the domains in defining business systems in order to ensure that it develops a complete and accurate inventory of its business systems. As shown in table 2, the reported inventory of business systems for most of the other domains increased, except for the financial management domain, whose inventory declined. Domain officials attributed the increases to additional data calls and working closely with the components to identify systems. The financial management domain attributed the declines to eliminating nonsystems and duplicate entries in the inventory. For example, its analysis showed that previously spreadsheets, reports, or both were incorrectly reported as being systems. For the current inventory, some of the domains indicated that they used the definition of a system that was issued by DOD’s CIO in the July 2004 DITPR data call. Table 3 shows the distribution of the 4,150 business systems among the components and domains. The table shows the stovepiped, duplicative nature of DOD’s business systems. For example, there are 713 human resources systems across all components whose reported funding for fiscal year 2005 includes approximately $223 million for modernization and over $656 million for operation and maintenance. According to DOD officials, the Defense Integrated Military Human Resources System (DIMHRS) is intended to totally or partially replace 113 of these systems. We were informed that the remaining 600 human resources systems are to be reviewed in the context of the BEA as it is developed. Furthermore, a human resources domain official acknowledged that for two of the systems that are to be replaced by DIMHRS—the Army’s Electronic Military Personnel System and the Air Force Military Personnel Data System—continuing to spend money to modernize these systems was questionable. We also reported in June 2004 that the fiscal year 2005 IT budget request did not provide sufficient information to identify or justify the specific current services and modernizations for 97 of the 113 systems. Because, as noted in table 1, the funding is distributed to and controlled by the military services and DOD components, the domains have minimal influence over system funding. As a result, DOD continues to fund the modernization of systems that it intends to totally or partially replace. As discussed later, the new requirements and authorities included in the fiscal year 2005 defense authorization act are aimed at ensuring that the domains have a vital decision-making role in the control and accountability of the investments being made in the department’s business systems. While DOD has reported that its inventory of business systems has increased by about 1,900, the department continues to struggle with developing a comprehensive inventory. As detailed in table 3, there are 215 business systems with no component identified—although they have been assigned to a domain—and 140 business systems with no domain assigned. BMMP officials stated that they are reviewing each system and working with the domains to ascertain where each system should be placed. Without the component being identified, it would be difficult if not impossible to identify the DOD entity that is responsible for investment reviews of the systems. Furthermore, it is essential that a domain be identified for each system in order for the department to meet the requirements set forth in the fiscal year 2005 defense authorization act. In discussing the increase in the number of systems identified in DOD, some of the domains stated that many of the systems are not included in the IT budget request. They said that some of these systems were likely developed at the local level and financed by the operation and maintenance funds received at that location and therefore were not captured and reported as part of the department’s annual IT budget request. Financing business systems in this manner rather than within the IT budget results in Congress and DOD management not being aware of the total amount being spent to operate, maintain, and modernize the department’s business systems. As a result, Congress and DOD management do not receive complete and accurate data for use in monitoring the department’s business systems investments. In addition, according to Army officials, as part of its efforts to develop the Army’s enterprise architecture, the Army has identified about 3,000 systems, and it believes some of these systems should be categorized as business systems. At this time, the Army is uncertain how many should be classified as business systems. As shown in table 3, the BMMP inventory of Army business systems totaled 727 systems as of February 2005. Army officials did not specify the anticipated time frame for completing their analysis of these systems. Given that DOD does not know how many business systems it has or how much is spent on them, it is not surprising that the department continues to lack effective management oversight and control over business systems investments. Since February 2003, the domains have been given the responsibility to oversee the department’s business systems investments, yet the billions of dollars spent each year continue to be spread among the military services and defense agencies, enabling the numerous DOD components to continue to develop stovepiped, parochial solutions to the department’s long-standing financial management and business operation challenges. Furthermore, the department has testified that it does not know whether it was in compliance with the fiscal year 2003 defense authorization act, which provides that obligations in excess of $1 million for systems improvements may not be made unless the DOD Comptroller determines that the improvements are consistent with the criteria specified in the act. In this regard, based upon data reported to us by the military services and DOD components, obligations totaling about $243 million were made for systems modernizations in fiscal year 2004 that were not referred to the DOD Comptroller for the required review. Further, while the fiscal year 2005 national defense authorization act directs that the domains are to have increasing oversight of the department’s business systems investments, each of the military services has established its own management oversight structures. Thus, DOD does not yet have a comprehensive strategy in place that delineates the specific roles and responsibilities of the domains and military services. Absent an integrated strategy, the domains’ and military services’ efforts may be duplicative, resulting in the wasteful use of resources and hindering the overall transformation of the department’s business systems and related operations. DOD has not yet finalized guidance that clearly defines the roles and responsibilities of domains or assigns explicit authority for fulfilling these roles and responsibilities. It also has not established common investment criteria for system reviews or conducted a comprehensive review of its ongoing business systems investments. Over the past several years, we have made numerous recommendations aimed at improving the department’s control and accountability over its business systems investments. Many of the actions that DOD planned to take remained incomplete as of February 2005. DOD officials acknowledged that the following actions have not been completed: The March 2004 IT portfolio management policy memorandum signed by the Deputy Secretary of Defense provides the basic structure of the roles and responsibilities of the domains. In order for the guidance to be institutionalized within the department, DOD planned to issue a formal DOD directive to specify the roles and responsibilities of the domains and how they are to be involved in the overall business systems investment management process. As of February 2005, the DOD guidance had not been finalized, and DOD CIO officials did not have a time frame for its issuance. The domains are still working on developing standard and consistent criteria for performing system reviews. The BMMP program director recently acknowledged that the differing criteria are being used in the system review process. Although the domains have used draft guidance listing 27 critical questions since July 2004, this DOD guidance has not been finalized, and a time frame for approval has not been provided. The domains have not completed a comprehensive system review of their ongoing IT investments. As discussed previously, the reported business systems inventory has increased from 2,274 systems in April 2003 to over 4,000 systems in February 2005. A target date for completing these reviews has not been determined. We have previously reported that best practices recommend that an organization review ongoing investments periodically to ensure that they are consistent with its architectural development efforts. In a July 16, 2004, memorandum, the DOD Comptroller reiterated the importance of having all business systems modernizations with obligations exceeding $1 million approved. The memorandum expanded the congressional requirements for DOD Comptroller certification to all business systems and required business domains to submit a fiscal year 2005 system certification schedule to BMMP officials. Later, in November 2004, the DOD Comptroller testified that the department had started to take actions that would position it to meet the new, similar review requirements of the fiscal year 2005 act. The DOD Comptroller’s testimony noted that the department had already identified 132 business systems that represent 78 percent of fiscal year 2005 modernization funding. These systems are scheduled to be reviewed during the current fiscal year. However, as noted earlier, DOD has not yet established specific criteria for investment reviews. Such criteria would include elements to implement the definition of what constitutes a business system modernization that would be subject to the provisions in the fiscal year 2005 defense authorization act. While the act provides a general definition for a business system modernization, it is critical for DOD to issue specific implementation guidance and criteria to explicitly define business systems modernizations to ensure that the DOD components and the domains use clear, consistent guidance in performing the system reviews. We found that DOD is not in compliance with the fiscal year 2003 defense authorization act, which requires that all financial system improvements with obligations exceeding $1 million be reviewed by the DOD Comptroller. Based upon the reported obligational data provided to us by the military services and the defense agencies for fiscal year 2004, we identified 30 modernizations with obligations totaling about $243 million that were not submitted for the required review. As previously noted, DOD defines a modernization as an enhancement to existing systems or the development of new systems. For purposes of this report, we treat modernizations as defined by DOD the same as “system improvements” as that term is similarly defined in the fiscal year 2003 defense authorization act. Additionally, the 2003 act defines financial systems to include “budgetary, accounting, finance, enterprise resource planning or mixed information system.” We reviewed DOD’s representation of the functions performed by these systems made in the department’s fiscal years 2004 and 2005 budget requests. DOD has acknowledged that it does not have a mechanism to identify systems that should be reviewed in accordance with the statutory requirements. Because DOD lacks a systematic means to identify the systems that were subject to the requirements of the fiscal year 2003 defense authorization act, there is no certainty that the information provided to us accurately identified all systems improvements with obligations greater than $1 million during the fiscal year. BMMP officials stated that the domains were responsible for working with the components to make sure that business systems with obligations for modernizations greater than $1 million were submitted for review as required. There was also general agreement among the domains and BMMP officials that systems owners were responsible for initiating the $1 million review process. In essence, compliance was achieved via the “honor system,” which relied on systems owners coming forward and requesting approval. However, the approach did not work. During fiscal year 2004, the number of systems reviewed was small when compared to the potential number of systems that appeared to meet the obligation threshold identified in the fiscal year 2004 budget request. We analyzed the DOD IT budget request for fiscal year 2004 and identified over 200 systems in the budget that could involve modernizations with obligations of funds that exceed the $1 million threshold. However, BMMP officials confirmed that only 46 systems were reviewed, of which 38 were approved as of September 30, 2004. The remaining 8 systems were either withdrawn by the component/domain or were returned to the component/domain because the system package submitted for review lacked some of the required supporting documentation, such as the review by the Office of Program Analysis and Evaluation, if necessary. Moreover, although the modernizations of 38 business systems were reviewed and approved by the DOD Comptroller as required, this does not necessarily mean these were prudent resource investments or integrated solutions to DOD’s long-standing problems. Although the criteria for the DOD Comptroller review included compliance with the BEA, we have previously reported that the BEA did not include many of the elements of a well-defined architecture. For example, DOD does not have a comprehensive system inventory of its “As Is” environment and has not developed a transition plan to identify those systems that would not be part of the architecture. Further, the real value of a BEA is that it provides the necessary context for guiding and constraining systems investments in a way that promotes interoperability and minimizes overlap and duplication. Without it, expensive rework is likely to be needed to achieve these outcomes. In an attempt to substantiate that financial system improvements with over $1 million in obligations were reviewed by the DOD Comptroller, as provided for in the fiscal year 2003 act, we requested that DOD activities provide us with a list of obligations (by system) greater than $1 million for modernizations for fiscal year 2004. We compared the reported obligational data to the system approval data reported to us by BMMP officials. Based upon this comparison and as shown in table 4, 30 business systems with obligations totaling about $243 million in fiscal year 2004 for modernizations were not reviewed by the DOD Comptroller. Examples of DOD business systems modernizations with obligations in excess of $1 million included in table 4 that were not submitted to the DOD Comptroller include the following. The Navy obligated about $57 million for the Navy Tactical Command Support System in fiscal year 2004. We previously reported that for fiscal year 2003, the Navy obligated about $22 million for this system without submitting it to the DOD Comptroller for review. DFAS obligated about $3 million in fiscal year 2004 for the DFAS Corporate Database/DFAS Corporate Warehouse (DCD/DCW). In fiscal year 2003, DFAS obligated approximately $19 million for DCD/DCW without submitting it to the DOD Comptroller for review. Additionally, we reported in May 2004 that DFAS had yet to complete an economic analysis justifying that continued investment in DCD/DCW would result in tangible improvements in the department’s operations. The department has acknowledged that DCD/DCW will not result in tangible savings to DOD. Continued investment is being based upon intangible savings of man-hours reductions by DFAS. The Air Force obligated about $25 million for the Integrated Maintenance Data System in fiscal year 2004. We previously reported that for fiscal year 2003, the Air Force obligated over $9 million for this system without it being submitted to the DOD Comptroller for review. The Army obligated over $34 million for its Logistics Modernization Program in fiscal year 2004. In fiscal year 2003, the Army obligated over $52 million without the prerequisite review being performed by the DOD Comptroller. DLA obligated about $5 million for the Defense Medical Logistics Standard Support program in fiscal year 2004. We previously reported that DLA obligated about $5 million in fiscal year 2003 for this program. The 2003 act placed limitations on the legal authority of individual program and government contracting officials to obligate funds in support of the financial systems improvements for which they are responsible, but DOD did not proactively manage investments to avoid violations of the limitations and did not review investments in any meaningful way to enforce these statutory limitations. Appendix III provides a list of obligations exceeding $1 million for business systems modernizations for fiscal year 2004 that were reviewed and approved by the DOD Comptroller as required by the 2003 act. Appendix IV provides a list of the individual systems not submitted to the DOD Comptroller and the related amount of the reported obligations for fiscal year 2004, as required by the 2003 act. It should be noted that since passage of the fiscal year 2003 defense authorization act in December 2002 through the end of fiscal year 2004, based upon information reported to us, the military services and defense components obligated about $651 million for business systems modernizations without the required review by the DOD Comptroller. While this amount is significant, it is not complete or accurate because it does not include any fiscal year 2005 obligations that occurred prior to the enactment of the fiscal year 2005 defense authorization act on October 28, 2004. Additionally, our analysis also identified another 50 business systems with obligations totaling over $258 million that were not submitted for review as directed by the DOD Comptroller’s July 16, 2004, memorandum. The memorandum expanded the criteria set forth in the 2003 act to include the modernization of all nonfinancial business systems that support the operations of the business domains. Appendix V provides a list of the business systems not submitted for review in accordance with the July 2004 guidance and the related amount of obligation for each system. In an attempt to achieve compliance with the requirement of the 2003 act and the DOD Comptroller’s July 2004 memorandum, as of January 2005, the DOD Comptroller identified 48 business systems that had both fiscal year 2004 and 2005 budgets each greater than $1 million in modernization funding, but had not been reviewed and approved by the DOD Comptroller. For the 48 business systems, the DOD Comptroller withheld a funding amount equal to 50 percent of the systems’ fiscal year 2005 modernization funding, which amounts to over $192 million. We discussed the withheld amounts with BMMP officials, who told us they anticipated that virtually all of the systems would come off of the “withhold” list by the end of the fiscal year. According to BMMP officials, a system will be removed from the withhold list as soon as the system owner, in conjunction with the domains, has a business system certification package approved by the DOD Comptroller, thereby showing compliance with the DOD Comptroller’s July 2004 memorandum. DOD’s process of withholding funding is focused on meeting documentation requirements related to compliance with the BEA, rather than on the control of business systems investments. The “withhold” process ultimately will have very little impact on DOD’s control and accountability over its business systems investment. The department continues to perpetuate the proliferation of duplicative, nonintegrated, and stovepiped business systems by spending billions of dollars annually on the modernization of systems for which DOD lacks reasonable assurance that the investment will add value to DOD’s operations. To gain more control and accountability over such business systems funding, we have previously recommended that the funding be vested with the “owners” of the various functional areas or domains. We believe it is critical that funds for DOD business systems be appropriated to the domain owners in order to prevent the continued parochial approach to systems investment that exists today. While the department has stated that the domains would be involved in the fiscal year 2006 budget review process, as previously noted, we found this not to be the case. Unless the domains control the funding, it will be difficult for them to meet the requirements of the Ronald W. Reagan National Defense Authorization Act for Fiscal Year 2005, including control and accountability over business systems investments. The statutory requirements enacted as part of the Ronald W. Reagan National Defense Authorization Act for Fiscal Year 2005 are aimed at improving the department’s business systems management practices. The act directs DOD to put in place a definite management structure that is responsible for the control and accountability over business systems investments by establishing a hierarchy of investment review boards from across the department and directs that the boards use a standard set of investment review and decision-making criteria to ensure compliance and consistency with the BEA. More specifically, the act does the following: Directs DOD to establish specific management oversight and accountability with the “owners” of the various functional areas or domains. The legislation defined the scope of the various business areas (e.g., acquisition, logistics, and finance and accounting) and directed the establishment of functional approval authority and responsibility for management of the portfolio of business systems with the relevant under secretary of defense for the departmental domains and the Assistant Secretary of Defense for Networks and Information Integration and the CIO for the department. Stipulates that no later than March 15, 2005, the responsible approval authorities, or domains, establish a hierarchy of investment review boards with DOD-wide representation, including the military services and defense agencies. The boards are responsible for reviewing and approving investments to develop, operate, maintain, and modernize business systems for their respective business areas, including ensuring that investments are consistent with DOD’s BEA. Directs the Secretary of Defense to establish the Defense Business Systems Management Committee with representation including the Deputy Secretary of Defense, the designated approval authorities, and secretaries of the military services and heads of the defense agencies. The Deputy Secretary of Defense is the chairman of the committee with one of the designated approval authorities serving as the vice-chairman. Directs that effective October 1, 2005, funds may not be obligated for a defense business systems modernization that will have a total cost in excess of $1 million unless the conditions specified in the act are met. The Defense Business Systems Management Committee must agree with the designated approval authorities’ certification before funds can be obligated. More important, the obligation of funds without the requisite approval by the Defense Business Systems Management Committee is deemed a violation of the Anti-Deficiency Act. Requires that no later than March 15 of each year from 2005 through 2009, the Secretary of Defense shall submit a report to Congress that describes how DOD plans to comply with the requirements of the act. Stipulates that all budget requests, starting with the budget request for fiscal year 2006, include supporting information that (1) identifies each defense business system for which funding is proposed in that budget; (2) identifies all current services and modernization funds, by appropriation, for each business system; (3) identifies the designated approval authority for each business system; and (4) describes the required certification for each business system. While the success of BMMP and improved control and accountability of business systems investments are critical aspects of the department’s transformation efforts, equally important is the department’s ability to develop and implement business systems that provide the promised capabilities on time and within budget. As we have previously reported, the department has not demonstrated the ability to achieve these goals. Given that the domains have been designated as being responsible for reviewing and approving business systems investments, each of the domains’ investment review boards needs to provide effective management oversight of each system project’s performance and progress toward predefined cost and schedule expectations as well as each project’s anticipated benefits and risk exposure. Further, each investment review board should also employ early warning systems that enable it to take corrective action at the first sign of cost, schedule, or performance slippages. Effective project oversight requires having regular reviews of the project’s performance against stated expectations and ensuring that corrective actions for each underperforming project are documented, agreed to, implemented, and tracked until the desired outcome is achieved. We have previously recommended that until DOD assesses its current systems, investment should be limited to deployment of systems that have already been fully tested and involve no additional development or acquisition cost; stay-in-business maintenance needed to keep existing systems management controls needed to effectively invest in modernized new systems or existing system changes that are congressionally directed or are relatively small, cost-effective, and low risk and can be delivered in a relatively short time frame. DOD also has not acted upon this recommendation and continues to invest billions of dollars without effective oversight and control. With the fiscal year 2005 act placing more responsibility on the domains, the implementation of the above four limitations would be one means of obtaining improved control over business systems investments. We have previously reported that best practices recommend that to achieve successful transformation, an organization must change its culture so that it is more results-oriented, customer-focused, and collaborative in nature. To transform its culture, an effective performance management system can be a strategic tool to drive internal changes and achieve desired results by using specific key practices to create a clear link between individual performance and organizational success. An effective performance management system pertaining to DOD’s business systems would help the department determine the effectiveness of the domains and components in carrying out their responsibilities for the control and accountability of business systems investments. Although the requirements of the fiscal year 2005 defense authorization act establish a management structure, each of the military services has established its own business systems investment review process. At this point, it is uncertain how they will be integrated with the roles and responsibilities that are to be exercised by the domains. Given the size and complexity of the business systems and related operations transformation endeavor, it is critical that the military services and domains are fully integrated into one cohesive business system investment management strategy. However, it is not clear what specific role the military services will play. If the military services’ efforts are simply viewed as one more level of review, that would be counterproductive to the overall transformation goals and objectives. Absent guidance that clearly articulates the relationship and the related roles and responsibilities of the domains and military services, each will continue to have stovepiped approaches to business systems investment management that result in more duplicative efforts. As a result, the department will continue to lack an overall comprehensive corporate process for ensuring that the billions of dollars spent on business systems are being spent efficiently and economically. DOD’s difficulty in simply identifying all of its business systems and the money spent on them illustrates the enormity and complexity of transforming the department’s business operations. Since April 2003, the reported inventory has increased by about 1,900 systems. Also, the department is not aware of which DOD component controls all the systems nor have all the systems been assigned to a domain. Given these circumstances, the department has made limited progress in achieving effective management control and accountability over the billions of dollars invested annually in its business systems. The department continues to lack reasonable assurance that the billions of dollars spent annually on its business systems represent an efficient use of resources. Because DOD lacks a well-defined BEA and transition plan, billions of dollars continue to be at risk of being spent on systems that are duplicative, are not interoperable, cost more to maintain than necessary, and do not optimize mission performance and accountability. While the fiscal year 2005 defense authorization act provides a management structure to improve the control and accountability over the department’s business systems investments, the appropriate policies and procedures still must be developed, implemented, and institutionalized to allow the department to make informed systems investment decisions. An integrated, comprehensive strategy will be critical to help ensure the domains and military services do not proceed independently of one another. By doing less, DOD will continue to waste billions of dollars by perpetuating today’s legacy business systems environment. To improve the department’s control and accountability of business systems investments, we are making the following four recommendations. We recommend that the Secretary of Defense direct that the DOD CIO, in consultation with the domains, review the 56 systems reclassified from business systems to national security systems to determine how these should be properly reported in the fiscal year 2007 IT budget request; Defense Business Systems Management Committee work with the domain investment review boards to review the reported BMMP business systems inventory so systems are defined in accordance with the definition specified in the fiscal year 2005 defense authorization act; Defense Business Systems Management Committee develop a comprehensive plan that addresses implementation of our previous recommendations related to the BEA and the control and accountability over business systems investments (at a minimum, the plan should assign responsibility and estimated time frames for completion); and comprehensive plan we recommend above be incorporated into the department’s second annual report due March 15, 2006, to the defense congressional committees, as required by the fiscal year 2005 defense authorization act, to help facilitate congressional oversight. We received written comments on a draft of this report from the Under Secretary of Defense for Acquisition, Technology and Logistics, which are reprinted in appendix II. DOD concurred with our recommendations and identified actions it planned to take to improve the department’s control and accountability of business systems investments. For example, the Under Secretary of Defense for Acquisition, Technology and Logistics, stated that the department will conduct a review of the 56 systems reclassified from business to national security systems to determine the proper classification of these systems in the IT budget requests. In addition, the Under Secretary of Defense for Acquisition, Technology and Logistics, noted that on March 17, 2005, the department designated the DITPR as the database for all DOD business systems and that the DITPR has the capability to identify all reported BMMP business systems in accordance with the definition specified in the fiscal year 2005 defense authorization act. The Under Secretary of Defense for Acquisition, Technology and Logistics, also stated that the department is developing a plan and timeline to address our previous recommendations and that this plan will be included in the department’s report due March 15, 2006, to the defense congressional committees. In addition to the actions taken in response to our recommendations, DOD implemented several other key steps after we provided a draft of the report to the department for comment. Specifically, the department acted to address certain provisions and requirements of the fiscal year 2005 defense authorization act. On March 19, 2005, the Deputy Secretary of Defense delegated the authority for the review, approval, and oversight of the planning, design, acquisition, development, operation, maintenance, and modernization of defense business systems to the designated approval authority for each business area. Additionally on March 24, 2005, the Deputy Secretary of Defense directed the transfer of program management, oversight, and support responsibilities regarding DOD business transformation efforts from the Office of the Under Secretary of Defense, Comptroller, to the Office of the Under Secretary of Defense for Acquisition, Technology and Logistics. According to the directive, this transfer of functions and responsibilities will allow the Office of the Under Secretary of Defense for Acquisition, Technology and Logistics to establish the level of activity necessary to support and coordinate activities of the newly established Defense Business Systems Management Committee. As required by the act, the Defense Business Systems Management Committee, with representation including the Deputy Secretary of Defense, the designated approval authorities, and secretaries of the military services and heads of the defense agencies, is the highest ranking governance body responsible for overseeing DOD business systems modernization efforts. While these actions are important in establishing the administrative framework for implementing management reform, we continue to believe that a new executive position is needed to provide the strong and sustained leadership to guide these efforts. We have testified on the need for a chief management official (CMO) on numerous occasions, including our most recent testimony on April 13, 2005. The CMO would serve as the Deputy Secretary of Defense for Management and oversee the department’s business transformation efforts. The day-to-day demands placed on the Secretary of Defense, the Deputy Secretary, and others make it difficult for these leaders to maintain the oversight, focus, and momentum needed to resolve the weaknesses in DOD’s overall business operations. This is particularly evident given the demands that the Iraq and Afghanistan postwar reconstruction activities and the continuing war on terrorism have placed on current leaders. Furthermore, the breadth and complexity of the problems and the hierarchical nature of the department preclude the under secretaries from asserting the necessary authority to resolve these long- standing issues while continuing to fulfill their other responsibilities. A CMO could provide the sustained and focused leadership that these other top officials are unable to provide. On April 14, 2005, a bill was introduced in the Senate that requires the establishment of a CMO that would be appointed by the President and confirmed by the Senate, for a set term of 7 years. As agreed with your offices, unless you announce the contents of this report earlier, we will not distribute it until 30 days after its issuance date. At that time, we will send copies to the Chairmen and Ranking Minority Members, Senate Committee on Armed Services; Senate Committee on Homeland Security and Governmental Affairs; Subcommittee on Defense, Senate Committee on Appropriations; House Committee on Armed Services; Subcommittee on Defense, House Committee on Appropriations; and Ranking Minority Member, House Committee on Government Reform. We will also send copies to the Under Secretary of Defense (Comptroller); the Under Secretary of Defense (Acquisition, Technology and Logistics); the Under Secretary of Defense (Personnel and Readiness); the Assistant Secretary of Defense (Networks and Information Integration); and the Director, Office of Management and Budget. Copies of this report will be made available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions on matters discussed in this report, please contact Gregory D. Kutz at (202) 512-9095 or kutzg@gao.gov or Keith A. Rhodes at (202) 512-6412 or rhodesk@gao.gov. GAO contacts and key contributors to this report are listed in appendix VI. We reviewed the Department of Defense’s (DOD) approximately $28.7 billion fiscal year 2005 information technology (IT) budget request to determine what portion of the budget relates to DOD business systems. We reviewed the budget to determine, of the approximately $13.3 billion related to the department’s business systems, the amount allocated for operation, maintenance, and development/modernization. We also met with the domains to obtain an understanding of the process followed in determining the specific number of business systems applicable to each respective domain. In addition, we compared the fiscal year 2004 and 2005 IT budget requests to determine the systems that were reclassified from business systems to national security systems. We analyzed the 56 system reclassifications by using information in the budget requests, the Business Management Modernization Program (BMMP) systems inventory, and the list of business systems modernizations with obligations approved by the DOD Comptroller to determine if they were reasonable and consistent. For certain systems that had inconsistent information, we inquired of system program and BMMP officials about the appropriateness of the reclassifications. To determine the effectiveness of DOD’s control and accountability over its business systems investments, we met with DOD officials to obtain an update on the status of our prior recommendations. We also met with appropriate officials in the DOD Comptroller and DOD Chief Information Officer (CIO) offices to discuss the status of various draft policies and guidance that are aimed at improving the department’s control and accountability over business systems investments. We also reviewed and analyzed the DOD budget request for fiscal year 2004 to identify the business systems investments that could be subject to the requirements of the Bob Stump National Defense Authorization Act for Fiscal Year 2003, which requires the DOD Comptroller to review all financial system improvements with obligations exceeding $1 million and determine whether each improvement is in accordance with criteria specified in the act. To assess DOD’s compliance with the act, we obtained and reviewed obligational data on modernizations in excess of $1 million for business systems for fiscal year 2004. We compared the obligational data provided by the military services and defense agencies with information obtained from BMMP officials to determine if the modernizations were reviewed by the DOD Comptroller as stipulated in the fiscal year 2003 act. We did not review the accuracy and reliability of the obligational data reported by DOD. Given the department’s previously reported problems related to financial management, we have no assurance that the data provided were complete, but the obligational data reported by DOD are the only data available that can be used for determining the specific amount of modernization funding spent on each business system. To augment our document reviews and analyses, we interviewed officials from various DOD organizations, including the Office of the Under Secretary of Defense (Comptroller); the DOD CIO; and the Office of the Under Secretary of Defense (Acquisition, Technology and Logistics). We conducted our work from August 2004 through February 2005 in accordance with U.S. generally accepted government auditing standards. We requested comments on a draft of this report from the Secretary of Defense or his designee. We received written comments on a draft of the report from the Under Secretary of Defense for Acquisition, Technology and Logistics, which are reprinted in appendix II. Defense Civilian Personnel Data System-Sustainment Navy Tactical Command Support System Conventional Ammunition Integrated Information System SPAWAR Financial Management – ERP NAVSEA Regional Fleet Maintenance ERP Pilot Regional Maintenance Automated Information System Integrated Logistics System – Supply Depot Maintenance Accounting and Production System Financial Inventory Accounting & Billing System Regionalization of Civilian Personnel Support Job Order Production Master System Fuels Automated Management System Sustainment - Air Force Automated Budget Analysis/Centralized User System DFAS Corporate Database/DFAS Corporate Warehouse Defense Medical Logistics Standard Support Subsistence Total Order and Receipt Electronic System $9.8 (Continued From Previous Page) Personnel Electronic Records Management System US MEPCOM Integrated Resource System US Army Accessions Command Integrated Automation Architecture Target Location Design and Hand-Off System Electronic Military Personnel Record System Joint Engineer Data Management Information Control System MSC Afloat Personnel Management Center Mounted Cooperative Target ID System Surface Warfare Management Information Systems Reliability and Maintainability Information System Air Force Military Personnel Data System Education and Training Technology Applications Program Cadet Administrative Management Information System Air Force Recruiter Information Support System Programming Depot Maintenance Scheduling System 2.6 (Continued From Previous Page) Staff members who made key contributions to this report were Beatrice Alff, Francine DelVecchio, Francis Dymond, Lauren Fassler, Kristi Karls, Mai Nguyen, Philip Reiff, and Bernard Trescavage.

Despite its significant investment in business systems, the Department of Defense (DOD) continues to have long-standing financial and business management problems that preclude the department from producing reliable and timely information for making decisions and for accurately reporting on its billions of dollars of assets. GAO was asked to (1) identify DOD's fiscal year 2005 estimated funding for its business systems and (2) determine whether DOD has effective control and accountability over its business systems modernization investments. DOD's business and financial management weaknesses have resulted in billions of dollars wasted annually in a time of increasing fiscal constraint. These weaknesses continue despite DOD requesting over $13 billion in fiscal year 2005--about $6 billion less than in fiscal year 2004--to operate, maintain, and modernize its existing duplicative business systems. The difference is more a reclassification of systems rather than an actual spending reduction. Some of the reclassifications appeared reasonable and others were questionable due to inconsistent information. At the same time, DOD reported an increase in the number of business systems to 4,150 as of February 2005--an increase of about 1,900 systems since April 2003. The duplicative and stovepiped nature of DOD's systems environment is illustrated by the numerous systems in the same business area. For example, DOD reported that it has over 2,000 logistics systems--an increase of approximately 255 percent since April 2003. DOD still does not have an effective departmentwide management structure for controlling business systems investments. Furthermore, DOD is not in compliance with the National Defense Authorization Act for Fiscal Year 2003, which requires the DOD Comptroller to determine that system improvements with obligations exceeding $1 million meet the criteria specified in the act. Based on limited information provided by DOD, system improvements totaling about $243 million of obligations over $1 million were not reviewed by the DOD Comptroller in fiscal year 2004. Cumulatively, based upon DOD's reported data, system improvements totaling about $651 million of obligations over $1 million were not reviewed by the DOD Comptroller before obligations were made since passage of the 2003 act. The 2005 defense authorization act directed that DOD put in place a management structure to improve the control and accountability over business systems investments by placing more responsibility with the domains. At the same time, each military service has its own investment review process. Absent an integrated management structure that clearly defines the relationship of the domains and the military services, DOD will be at risk that the parochialism contributing to the current problems will continue.

The Census Bureau has discretion under the Constitution and federal statutes to decide whether to count Americans residing overseas. The federal decennial census is conducted pursuant to the requirement imposed by Article I, Section 2, Clause 3 of the Constitution, and Section 2 of the Fourteenth Amendment, that Congress enumerate “the whole number of persons in each State” as the basis for apportionment of seats in the United States House of Representatives. Under the Constitution, the census is to be conducted every 10 years “in such Manner as shall by Law direct.” Congress has exercised its authority under the Constitution by passing the Census Act, which assigns to the Secretary of Commerce the responsibility of “tak a decennial census of population as of the first day of April” of each census year. The Secretary does so with the assistance of the Census Bureau and its Director. The statutes governing the earliest censuses provided that enumerators should record all persons reported to them within their respective districts as having a usual place of abode there or as usually residing within that district, even though such persons might be “occasionally absent at the time of enumeration.” Statutes governing later censuses, including the current provisions in Title 13 of the United States Code, contain no similar provision, or any provision specifically governing the enumeration of inhabitants of the United States who are outside of its borders on the enumeration date. To determine who should be included in the census, the Bureau applies its “usual residence rule,” which has been defined as the place where a person lives and sleeps most of the time. People who are temporarily absent from that place are still counted as residing there. One’s usual residence is not necessarily the same as one’s voting residence or legal residence. Noncitizens living in the United States are counted in the census, regardless of their immigration status. Historically, the census has focused primarily on the domestic population and typically has not included any procedures designed to enumerate Americans residing outside of the United States. The first attempts to count Americans residing overseas were in the 1830 and 1840 censuses, which included procedures for counting the “crews of naval vessels at sea.” The naval personnel included in those censuses, however, were not allocated to any individual state, and thus were not included in the apportionment population. As shown in table 1, various overseas population groups were included in the census at different times. For example, while federally affiliated personnel were typically included in the enumerations that took place from 1900 through 2000, only the 1970, 1990, and 2000 censuses used the numbers for purposes of apportioning Congress. At the same time, private citizens living abroad were included only in the 1960 and 1970 censuses, but not for purposes of apportionment. In response to congressional direction and the concerns of various business, political, and other groups that represent overseas Americans, the Census Bureau embarked on a research and evaluation program aimed at determining the feasibility, quality, and cost of counting both federally affiliated and private citizens living abroad. The test enumeration began February 2004 and is to run through July 2004 at three sites: France, Kuwait, and Mexico. The Bureau selected these countries based on several criteria including their geographic diversity, the fact that large numbers of U.S. citizens reside there, and because of the existence of administrative records that can be used to compare to the test census counts for evaluation purposes. The Bureau estimated the implementation costs for the 2004 test at approximately $2.5 million in fiscal year 2004. Further, the Bureau estimates that by the end of fiscal year 2004, it will have spent an additional $3.5 million for planning and preparation during fiscal years 2003 and 2004. Americans can participate in the test census by completing a short-form paper questionnaire that is available at embassies, consulates, and other organizations that serve overseas Americans, or by completing the form on the Internet. The Bureau hired a public relations firm to develop a communications strategy to inform and motivate respondents living in the selected countries to answer the census. Responses from the paper and the Internet returns will be captured in order to analyze, among other things, the demographic characteristics of respondents and patterns of item nonresponse. The Bureau plans to conduct a 2006 overseas test if Congress appropriates requested funds in fiscal years 2005, 2006, and 2007. If Congress then indicates its desire that the Census Bureau conduct a general overseas enumeration in 2010, the Bureau will seek a supplementary appropriation in calendar year 2007 for that purpose and to conduct a 2008 overseas dress rehearsal beginning in 2007. A sound test is essential in order for the Bureau, Congress, and other stakeholders to resolve the numerous logistical, conceptual, policy, and other questions that surround the counting of overseas Americans. They include: Who should be counted? U.S. citizens only? Foreign-born spouses? Children born overseas? Dual citizens? American citizens who have no intention of ever returning to the United States? Naturalized citizens? How should overseas Americans be assigned to individual states? For certain purposes, such as apportioning Congress, the Bureau would need to assign overseas Americans to a particular state. Should one’s state be determined by the state claimed for income tax purposes? Where one is registered to vote? Last state of residence before going overseas? These and other options all have limitations that would need to be addressed. How should the population data be used? To apportion Congress? To redistrict Congress? To allocate federal funds? To provide a count of overseas Americans only for general informational purposes? The answers to these questions have significant implications for the level of precision needed for the data and ultimately, the enumeration methodology. How can the Bureau verify U.S. citizenship? Administrative records such as passports and Social Security data have limitations. For example, Americans can reside in Mexico and Canada without a passport and many Americans overseas do not have Social Security numbers, especially dependents. How can the Bureau ensure a complete count without a master address list? The foundation of the stateside decennial census is a master address list. Because the list is essentially the universe of all known living quarters in the United States, the Bureau uses it to deliver questionnaires, follow up with nonrespondents, determine vacancies, and determine individuals the Bureau may have missed or counted more than once. The Bureau lacks a complete and accurate address list of overseas Americans. Consequently, these operations would be impossible and the quality of the data would suffer as a result. Can administrative records be used to help locate and count overseas Americans? Administrative records such as passport and visa files, voter registration forms, as well as records held by private companies and organizations have the potential to help the Bureau enumerate Americans abroad. However, the accuracy of these records, the Bureau’s ability to access them, confidentiality issues, and the possibility of duplication all remain open questions. Do certain countries have requirements that could restrict the Bureau’s ability to conduct a count? According to the Bureau, in planning the overseas test, the Bureau was informed that French privacy laws prohibit asking about race and ethnicity, two questions that are included on the U.S. census questionnaire. Although the Bureau worked with French officials to address this problem, the extent to which the Bureau will encounter restrictions in other countries, or whether other countries will cooperate with the Bureau at all, is unknown. As agreed with your offices, our objectives for this report were to (1) assess the soundness of the Bureau’s test design and its suitability for addressing the Bureau’s specific research questions, and (2) examine what past court decisions have held about Americans’ rights and obligations abroad that could help inform whether and how they should be included in the census. To assess the soundness of the Bureau’s 2004 overseas enumeration test design, we interviewed knowledgeable Bureau officials and reviewed existing documents that described the Bureau’s test objectives, research questions, and test plans. We then systematically rated the Bureau’s approach using a checklist of over 30 design elements that, based on our review of program evaluation literature, are relevant to a sound study plan. For example, we reviewed the Bureau’s approach to determine, among other things, (1) how clearly the Bureau presented the research objectives, (2) whether the research questions matched the research objectives, (3) whether potential biases were recognized and addressed, and (4) the appropriateness of the data collection strategy for reaching the intended sample population. See appendix II for a complete list of the 30 design elements. We supplemented our ratings on the suitability of the test by gathering additional information through telephone and in-person interviews with representatives of several stakeholder organizations that represent various groups of Americans residing abroad. The organizations included Democrats Abroad, Republicans Abroad, Association of Americans Resident Overseas, and the American Business Council of the Gulf Countries. In addition, we interviewed representatives of the Mexican American Legal Defense and Educational Fund, National Coalition for an Accurate Count of Asians and Pacific Islanders, and California Rural Legal Assistance, Inc., to get their perspectives on the Bureau’s plans for counting American citizens living in Mexico, particularly migrant and seasonal farm workers, a group that the Bureau had trouble counting during the 2000 Census. These three organizations, while not actively involved in the planning of the overseas enumeration test, are members of the Secretary of Commerce’s Decennial Census Advisory Committee, a panel that advises the Bureau on various census-related issues. To examine what past court decisions have held about Americans’ rights and obligations living abroad, including their right to be counted in the census, we reviewed a judgmental selection of five federal laws and/or programs that cover large numbers of Americans stateside, in order to determine how those laws and programs treat U.S. citizens should they live overseas. We examined federal election law, federal income tax law, and federal laws relating to Social Security benefits, Supplemental Social Security Income, and Medicare. For each of these laws and programs, we reviewed relevant statutes, court decisions, and legal analyses. We requested comments on a draft of this report from the Secretary of Commerce, which were sent to us April 13, 2004 (see app. I). We address them in the Agency Comments and Our Evaluation section of this report. According to the Bureau, its objectives for the 2004 overseas test are “to determine the feasibility, quality, and cost of collecting data from U.S. citizens living overseas.” To meet those objectives, the Bureau developed eight research questions designed to gather data on such salient factors as participation levels, data quality, and the relative response from the two enumeration modes (Internet and paper questionnaire; see appendix III for a sample of the paper questionnaire). To assess the overseas test the Bureau is planning to complete a series of evaluations due in early 2005. The test objectives and related research questions are appropriate as written, but, as shown in table 2, because of various methodological limitations, the data that will result from the test will not fully answer key questions concerning feasibility, data quality, and cost. In short, the Bureau overstated the research test’s ability to answer its key research objectives and, as a result, congressional decision making on this issue will be that much more difficult. For the decennial census, the Bureau uses participation data as a key indicator of public cooperation with the census. The Bureau measures participation levels by what it calls the “return rate,” which it calculates as a percentage of all forms in the mailback universe (excluding vacant and nonexistent housing units) from which it receives a questionnaire. Stateside, the Bureau is able to perform this calculation because, as noted above, it has a master address list of all known housing units in the United States, an inventory that takes the Bureau several years and considerable resources to compile. However, no such address list exists for overseas Americans. Consequently, participation rates for overseas Americans cannot be calculated and the Bureau will only be able to tally the number of responses it receives (both overall for each site and within several demographic categories) and compare the results to counts obtained from administrative records. The sources for the records include a combination of tax, Medicare, and State Department data, as well as foreign census data if available. For a variety of reasons—some of which the Bureau has already acknowledged—the information generated from this exercise may not be relevant. First, the administrative records were developed for different purposes, and as a result, are not well suited as a base for comparing against overseas counts; thus, their relevance is uncertain. As the Bureau has already reported, each of the records it plans to use to compare to the census counts has coverage limitations. Further, as each of these records is associated with particular demographic groups, they could introduce systemic biases (we discuss potential problems with administrative records in greater detail below, under Research Question #5). Second, the census counts could be problematic because it is unclear who should participate in the overseas census, which in turn could confuse potential respondents. For the stateside enumeration, to determine where an individual should be counted, the Bureau uses the concept of “usual residence,” which it defines as “the place where a person lives and sleeps most of the time.” The Bureau has developed guidelines, which it prints on the stateside census questionnaire, to help people figure out who should and should not be included. However, the Bureau has not developed similar guidance for the overseas test. According to the Bureau, this was intentional; because this was an initial feasibility test, the Bureau did not want to restrict response, but rather to encourage the widest possible participation. Further, the guidance the Bureau has developed, which is available on its Web site and promotional literature, is vague and could confuse potential respondents. For example, the guidelines inform potential respondents that, “All U.S. citizens living in France, Mexico and Kuwait, regardless of shared citizenship, can and should participate in the test. U.S. citizens on vacation or on short business trips should not.” Unclear is what constitutes a short or long business trip. Is it 3 weeks or 3 months? Does it matter whether one stays in a hotel or an apartment? Also, should naturalized U.S. citizens, some of whom may not return to the United States, participate? What about children born in the United States to noncitizens, but who only lived in the United States a short time? Should students spending a semester abroad but who maintain a permanent residence stateside be included? Without clear residence rules and appropriate guidance indicating who should be counted, it is quite possible that some people might inappropriately opt in or out of the census, which would reduce the quality of the data. Bureau officials have told us that they are working to develop residence rules that it will apply if there is a second overseas enumeration test in 2006. Participation data might also be problematic because the Bureau’s enumeration methods strategy might not be as effective with certain groups compared to others. To the extent this occurs, it could introduce a systemic undercount. This is particularly true for dual nationals who, for cultural reasons, may not think of themselves as American citizens. For example, according to representatives of two advocacy groups we contacted, Mexican dual nationals include migrant farm workers, a group that often consists of poor, less-educated people living in rural areas. They noted that this population group has low literacy levels and thus might not understand the questionnaire, and is not likely to have Internet access. At the same time, they are not likely to pick up a copy of the questionnaire at an embassy. Further, the barriers that make it difficult to count migrant farm workers in the United States, such as a distrust of government and the fact that they may speak indigenous languages, also make it difficult to count this group in Mexico. The Bureau plans on measuring the quality of the data collected in the overseas test by tabulating item nonresponse, which refers to whether a respondent completed a particular question. The Bureau is to calculate this information by enumeration mode, test site, and various demographic categories. The Bureau also plans to compare this measure of nonresponse for key variables to those obtained in an earlier, stateside test held in 2003 by tabulating the rate respondents did not complete a particular question. According to the Bureau, patterns of item nonresponse are critical for improving question design, training, and procedures. However, as the Bureau acknowledges in its study plan for evaluating the quality of the overseas enumeration data, item nonresponse by itself does not address the quality of the data. Thus, at the end of the test, the Bureau will have, at best, only limited information on the quality of the overseas data. By comparison, the Bureau’s guidelines for measuring data quality in other surveys they conduct use measures such as coverage, unit response rates, imputation rates, and data collection errors. Because the Bureau lacks information on the universe of overseas Americans it will be unable to calculate these indicators. Therefore, it is misleading for the Bureau to state in its research objectives that it will determine the quality of the overseas data, when in fact it will deliver something far more limited. Comprehensive measures of data quality are critical because they could help Congress decide whether the data are sufficiently reliable to use for specific purposes. If the numbers were to be used to obtain a simple count of Americans abroad, absolute precision is not as critical. However, for other uses of the data, particularly congressional apportionment and redistricting, the quality of the data would need to be far higher. Counting people in their correct locations is essential for congressional apportionment, redistricting, and certain other uses of census data. With respect to Americans abroad, if the data are to be used for apportionment, the Bureau would need to assign respondents to a specific state. For purposes of redistricting and allocating federal funds, the Bureau would need to assign overseas respondents to specific neighborhoods and street addresses in the United States—a far more challenging task. Geocoding is the process of linking an address in the Bureau’s Master Address File (MAF) to a geographic location in the Bureau’s geographic database, known as the Topologically Integrated Geographic Encoding and Referencing (TIGER) mapping system. To obtain this information, the overseas enumeration form asks respondents to indicate their U.S. state of last residence, and their last street address within that state. Although the Bureau is assessing the level of geography to which it can geocode the overseas population, an important limitation is that the Bureau will not be able to make this assessment for people who live in certain noncity-style U.S. addresses; that is, U.S. addresses without a housing number and/or street name. Specifically, people whose U.S. addresses consist of a post office box will be excluded from the study. For the stateside enumeration, Bureau employees canvass the country, identify noncity-style addresses, and mark the locations of those residences on a map. During the 2000 Census, around 20 percent of U.S. households had noncity-style addresses. Indeed, the overseas enumeration questionnaire instructs respondents not to provide a post office box number for their last stateside address. However, if a respondent’s address includes a post office box or rural route number, it is unclear how they are supposed to complete this question. Also unclear is how migrant farm workers, who may not have had an address in the United States, would complete this question. To the extent they leave the question blank, the Bureau would be unable to distinguish between those people who did so because they have a noncity- style address, or left it blank for privacy or other reasons. This could affect the accuracy of the Bureau’s assessment. Moreover, the construction of the question could introduce a systemic bias because those states with large rural areas are more likely to have noncity-style addresses. The Bureau’s evaluation plan recognizes that respondents in the 2004 overseas enumeration test may provide noncity-style addresses which cannot be geocoded by the TIGER system. The Bureau intends to provide data on how many city-style and noncity-style addresses could be geocoded. Although the Bureau has taken a number of steps to publicize the overseas enumeration, evaluating the effectiveness of that effort will present a challenge. The Bureau awarded a $1.2 million contract to a public relations firm to develop a promotion strategy for the overseas enumeration test. As part of that effort, the public relations firm identified a number of stakeholder organizations that represent U.S. citizens living overseas in each of the three test countries. The organizations included advocacy groups, universities, church groups, and corporations. The Bureau anticipates that these stakeholders will help get the word out via e-mail, newsletters, and other media that a test census of Americans overseas is underway. In addition, the Bureau is to provide copies of the overseas questionnaire to stakeholders so that they can distribute them to their members and constituents. As noted earlier, questionnaires will also be available on the Internet, as well as at public places that Americans may visit, such as embassies and consulates. The Bureau has produced posters and pamphlets to promote the test (see fig. 1). The Bureau also plans to have articles about the census test placed in newspapers and magazines and stories run on local television and radio. Although paid advertising was not part of its original plans, the Bureau later decided to run a limited amount of paid advertising in Mexico and France. Fill out a U.S. Census questionnaire, available It’s easy, important — and confidential. The Bureau will attempt to gauge the effectiveness of the marketing program by measuring participation (as measured by the number of responses) and public awareness. According to the Bureau, respondents who submit a questionnaire via the Internet will be asked to complete a short survey eliciting information on how they learned about the census test and what motivated them to participate. No similar survey is planned for people who pick up their surveys at an embassy or other distribution site. Thus, the Bureau will not have a parallel set of data from a group of respondents that might be demographically or behaviorally different from Internet respondents. The Bureau does, however, expect to conduct focus group interviews and debriefings to obtain feedback from mail respondents and stakeholder organizations. Focus group interviews targeting nonrespondents are planned as well. Yet, as the Bureau acknowledges, participation, or the final count of U.S. citizens living in the selected countries, will only be an indicator of the number of people that heard about the test, completed the questionnaire, and submitted it to the Census Bureau. It will not be able to measure the Bureau’s success in getting Americans to respond because the universe of Americans overseas is unknown. Public awareness will also be difficult to measure because it includes an unknown number of people who were aware that a test census was being conducted but chose not to respond. Nevertheless, as noted above, the Bureau intends to interview both respondents and nonrespondents in an effort to determine their awareness and motivation for responding or not responding to the census test. To the extent the Bureau conducts these interviews, it will be important for it to include hard-to-count groups, such as dual nationals and migrant farm workers, that may have been outside the reach of the Bureau’s marketing campaign. The Bureau plans on using administrative records such as Medicare and passport data to provide comparison information to assess (1) participation, (2) an invalid return detection system, and (3) the records’ potential use for building an address list. Specifically, the Bureau plans to compare the number of people counted at each site to federal tax, Medicare, U.S. Department of State, and foreign census records. While it is important for the Bureau to assess the utility of administrative records, it is unclear from the Bureau’s study plans how it will make this determination given what Bureau officials have said is a large disparity between administrative record counts on the number of Americans living overseas at the three test sites. Further, because these administrative records were developed for different purposes and as a result are not well suited as a base for comparing against overseas counts, their relevance is uncertain. For example, not all American citizens who live abroad file tax returns; dependents are not always listed on tax returns; and dependents that are included in the tax form may not be U.S. citizens. After living abroad 30 days, Americans are no longer eligible for Medicare benefits; therefore, Medicare records may not be the most useful and only apply to U.S. citizens over 65 years old. U.S. State Department records are nonofficial and according to the Bureau inaccurate because the Department of State does not officially track either the number or location of U.S. citizens living in other countries. Finally, the type of administrative records kept by each country is unknown and earlier Bureau research found that census data from foreign countries do not contain the detailed information required for apportionment, redistricting, or other census uses. To determine the relative response from the two enumeration modes and their relative effectiveness, the Bureau will look at such data as the timing of returns by mode and site, and whether one mode provided more valid returns than the other. Respondents can either fill out a paper questionnaire or complete the form on the Internet. The Bureau will deploy an invalid return detection system to determine whether a form is valid and responses will be tallied by mode and by country. However, the Bureau recognizes that its analysis will contain several limitations. Key among them is that it will not be able to determine participation rates because the universe of Americans overseas is unknown. As a result, the analysis is limited to a count of the total number of forms returned at each site. In light of this and other limitations, it is unclear what conclusions the Bureau will be able to draw about the effectiveness of the two response modes. Indeed, just because one enumeration mode results in a larger number of returns from a particular demographic groupolder Americans, for examplemay not necessarily have anything to do with the mode itself. As a result, it would be inappropriate to say that one enumeration method was more effective than the other in counting senior citizens. This is because there are other factors that can influence the mode such as advertising or accessibility to the Internet or the paper questionnaire. Overall, while it will be important to collect information on the returns by method of enumeration, this information should not be construed as a measure of the effectiveness of that mode. Part of the processing of overseas returns involves validating that the respondents are within the scope of the enumeration; that is, that they in fact reside in one of the three overseas test areas. Thus, the Bureau needs a better method to detect invalid returns. To determine whether responses are within scope, the overseas enumeration questionnaire asks respondents to provide their Social Security and U.S. passport numbers. Although we agree with the importance of determining whether respondents reside overseas, the Bureau’s analysis will not assess this. Rather, the Bureau defines a valid return as one where at least one person in the household checked the U.S. citizen box on the questionnaire, provided a valid Social Security or U.S. passport number, and has been subject to and passes an algorithm that analyzes data from the questionnaire. All other returns are invalid. Thus, what the Bureau is really measuring is whether a questionnaire is eligible for further processing, and not whether the respondent lives abroad. Put another way, anyone who completes a questionnaire with valid data, including Social Security numbers, would be considered a valid return, regardless of whether the individual lived in the test areas. This is not an unlikely scenario given the way the Bureau set up its Internet site. Indeed, anyone—even if they live outside of the three test sites-can be included in the overseas count, so long as they provide the required information. The reason they can slip through the invalid return detection system is because the Bureau is unable to confirm the point of origin for questionnaires completed on the Internet. The Bureau is aware of this gap in the invalid return detection system, but has been unable to resolve this condition. Another potential problem is the Bureau’s sole reliance on Social Security numbers to validate returns for the 2004 test. The Bureau had also wanted to use passport numbers to validate returns. Although the Bureau has been negotiating with the Department of State for access to the passport database, the Bureau does not expect this to occur in time for use in the 2004 test. Bureau officials said they were aware that there was a strong possibility that they would not be able to gain access to the passport file because the Bureau had not worked with Department of State data and that a memorandum of understanding would first need to be in place before the data would be released. Nevertheless, the Bureau believes that the impact of this would be low if it could be shown that using Social Security numbers alone to validate returns was sufficient. A third potential problem is that some people may be reluctant to provide their Social Security and passport numbers for privacy reasons. Based on the Bureau’s research, requesting this information could reduce participation levels. The Bureau, based on its earlier research, has already identified a list of barriers to integrating overseas with stateside data. They include different questionnaire content for the overseas form, detecting and eliminating duplication within and between overseas and stateside enumerations, timely geocoding of addresses, and limited resources. The Bureau plans to document the lessons learned from the 2004 overseas test and how they might apply to a more integrated test in 2006. In particular, the Bureau is to focus on the issues encountered or associated with collecting, capturing, and processing overseas data. While it will be important for the Bureau to thoroughly document these issues and their implications for integrating the two data sets, the Bureau does not intend to actually integrate any data from the overseas test with data being collected from a parallel stateside test it’s conducting at three Georgia counties and in Queens, New York. Without an actual integration, the Bureau may miss problems that will not likely be detected until a next test in 2006. In addition to the limitations noted above, the overseas enumeration test has other limitations that will affect the Bureau’s ability to answer its key research objectives. Although one of the Bureau’s objectives for the overseas enumeration test is to determine the cost of collecting data from overseas citizens, the Bureau’s test design lacked a specific research question aimed at obtaining this information. More importantly, the cost information that the Bureau will collect will be of limited value because it will not be used to estimate the costs of future tests, nor model the costs of conducting a broader overseas enumeration in 2010. The Bureau developed a cost model for the 2000 Census that provided the agency with an automated means to estimate staffing and budget requirements. The Bureau used the cost model to support the budget process, as well as to answer questions from Congress, the Office of Management and Budget (OMB), and our office. The cost model could also estimate the budgetary impact of certain assumptions and alternative census-taking scenarios. The Bureau maintains that it would need more data than those that are now available to develop a cost model for an overseas enumeration in 2010. Nevertheless, while only at the beginning stages, the Bureau has some data points to begin developing a cost model for overseas enumeration or, at a minimum, for identifying major cost components. For example, Bureau officials told us that their agency will track and report the marketing, printing, postage, data capture, and processing costs for conducting the overseas enumeration at the three test sites. All of the costs provided by the Bureau are direct costs and could be specifically and uniquely attributed to a cost model for the overseas enumeration. Indirect costs, however—those that are not easily attributable to the overseas enumeration such as executive management or technical labor—would still need to be determined and captured. Cost will be an important factor to consider when making a decision on whether to enumerate Americans overseas. The cost of the 2010 Census, now estimated at more than $11.3 billion in current dollars, is the most expensive enumeration in the nation’s history. Consequently, it will be critical for the Bureau to have a mechanism for accurately and quickly estimating overseas enumeration costs so that Congress, other oversight groups, and the Bureau itself can have reliable information on which to base or advise decisions. Testing a questionnaire with a sample of intended respondents before it is distributed is a standard approach that survey organizations employ to ensure questions are clear and understandable, and that respondents will be able to provide accurate information. In short, testing is an important quality assurance procedure that increases the likelihood that respondents will provide the information needed and help reduce the likelihood of inaccurate responses. The Bureau is aware of the benefits of testing questionnaires. In preparation for the 2000 Census, Congress budgeted millions of dollars for the Bureau to develop and test questionnaires during the 1990s, which it did using focus groups and one-on-one interviews. Moreover, the Bureau’s policy requires that demographic survey questionnaires be tested. However, the Bureau did not test the overseas enumeration questionnaire. Instead, the Bureau gave stakeholders the opportunity to review and comment on the questionnaire before it went to OMB for final approval. Stakeholder feedback was generally positive. One change that was made in response to stakeholder feedback was that “United States” was added to “passport” in the question that asked for a respondent’s passport number. While sharing the questionnaire with stakeholders is important, it should not be seen as a replacement of questionnaire testing. According to Bureau officials, the Bureau developed the overseas questionnaire by modifying the Census 2000 short form to collect data needed for the overseas count. New questions asked respondents for their employment status, Social Security number, and passport number. The Bureau believed that the new questions did not require testing. However, as stated earlier, collecting Social Security numbers and passport numbers could be problematic. While some stakeholders believed this to be acceptable, other stakeholders believe it could reduce participation, especially in Mexico where dual nationals reside. In addition, the questionnaire requests data on everyone in the household even if a person is not a U.S. citizen. These questions could be seen as too intrusive and potentially could stop someone from completing the form, thus resulting in an undercount. Since there was no testing of the questionnaire, the Bureau cannot be certain of the impact of these questions. In a 2001 report to Congress, “Issues of Counting Americans Overseas in Future Censuses,” the Bureau indicated that it would provide Congress with data on the number of people in the military, federal, private business, nonprofit, and other categories. However, the Bureau is only collecting data on the number of military and federal workers; people working in other sectors will be grouped in an “other” category. The reason that the Bureau is unable to provide the additional breakdown is because the length of the questionnaire did not allow for additional check boxes. Therefore, the Bureau collapsed people working in the private sector and nonprofit organizations and others into one category. The soundness of the test design notwithstanding, the Bureau has already identified several country-specific challenges to counting American citizens at each of the test sites. Together, they suggest that an enumeration of Americans on a more global scale in 2010 would introduce a number of unforeseen obstacles that the Bureau would need to address on a country-by-country basis. For example, shortly before the test was to begin in France, the French government contacted the Census Bureau indicating that French law prohibited the collection of race and Hispanic origin data. Furthermore, they were also opposed to the U.S. government asking for information on persons who were not American citizens. The Bureau worked with the French government and it was agreed that an advisory would be posted on the Internet site explaining that under French law it was not mandatory to respond to the questionnaire. Problems have also surfaced in Mexico and Kuwait. Stakeholders and Bureau officials have told us that the mail system in Mexico is not always reliable. The concern is that the questionnaires may not make it to the Census Bureau, or arrive too late to be processed. In Kuwait, security concerns have prevented the Census Bureau from posting the location of sites where Americans can pick up the questionnaire. While the impact of these problems is difficult to quantify, it may prevent some Americans from completing the questionnaire and being counted. Americans residing abroad do not have the same rights and obligations under federal programs and activities as Americans living in the United States. In order to determine the rights and obligations of Americans residing abroad, one must examine the specific statutes governing each program. For this study, we examined whether overseas Americans can vote in federal elections; are subject to federal income tax; and can receive Social Security, Supplemental Security Income, and Medicare benefits (see table 3). Stateside, these programs cover millions of Americans; whether and how they extend to Americans living overseas could help inform the treatment of U.S. expatriates in the decennial census, to the extent there are any patterns. American citizens who reside outside of the United States have the right to vote in federal elections under the 1986 Uniformed and Overseas Citizens Absentee Voting Act (UOCAVA). Under this law, U.S. citizens residing on foreign soil can vote in federal elections as absentee voters of their last state of residence, even if they have no intention of ever returning to the United States. (American citizens residing in U.S. territories, however, cannot so vote. The territories include Puerto Rico, the Northern Mariana Islands, the U.S. Virgin Islands, Guam, and American Samoa.) UOCAVA repealed legislation enacted in 1955 that was designed to prevent members of the Armed Forces and their families from being denied their voting rights when absent from their home or in a far-off place. The goal was to make it easier for military personnel to cast votes in any federal primary, general, or special election through absentee balloting procedures. While the 1955 law was amended several times to, among other purposes, specifically include other Americans living overseas, in 1986 Congress acknowledged that there was a legitimate need for further legislation. UOCAVA’s main purpose was to facilitate absentee ballot voting, while also providing “for a write-in absentee ballot that may be used in Federal general elections by overseas voters who, through no fault of their own, fail to receive a regular absentee ballot in sufficient time to vote and return the ballot prior to the voting deadline in their State.” U.S. citizens are taxed on their worldwide income, subject to certain exclusions specified in the Internal Revenue Code. In 1913, the enactment of the Sixteenth Amendment to the U.S. Constitution gave Congress “the power to lay and collect taxes on income, from whatever source derived, without apportionment among the several States, and without regard to any census or enumeration.” After ratification of this constitutional provision, Congress imposed a tax on the net income of every U.S. citizen, wherever they lived. That the Constitution vests Congress with the power to tax Americans living overseas on their income earned outside of the United States was reinforced by the Supreme Court in 1924 when it ruled: “overnment, by its very nature, benefits the citizen and his property wherever found, and therefore has the power to make the benefit complete. Or, to express it another way, the basis of the power to tax was not and cannot be made dependent upon the situs of property in all cases, it being in or out of the United States, nor was not and cannot be made dependent upon the domicile of the citizen, that being in or out of the United States, but upon his relation as citizen to the United States and the relation of the latter to him as citizen.” There are statutory exclusions, however. Generally, a U.S. citizen or resident may exclude a portion (ascending each year to $80,000 during or after 2002) of his earned income if he is a resident of a foreign country for a full calendar year or is physically present in a foreign country for 330 days of any 12 consecutive months and otherwise meets certain requirements. This foreign earnings exclusion principally aims to encourage U.S. citizens and residents to work abroad without worrying about how it might increase their tax liability. Indeed, in view of the nation’s continuing trade deficits, Congress deemed it important to allow Americans working overseas to contribute to the effort to keep American business competitive. The Social Security Act provides individuals over the age of 62 who have worked for a minimum number of years with monthly benefit payments. Social Security checks generally follow recipients wherever they go around the world, subject to only a few exceptions. In fact, Social Security payments continue no matter how long a beneficiary stays outside of the United Stateseven if the individual retires overseas. Thus, if you are a U.S. citizen residing abroad you generally continue to receive your monthly benefits. Supplemental Security Income (SSI) is a “need-based” program designed to help individuals who are over the age of 65, blind, or disabled. SSI benefits cease once a recipient remains outside the United States for a period of at least 30 days. If the recipient returns to this country within 30 days, SSI benefits are to continue as before. Since Medicare became effective almost 35 years ago, the program has excluded coverage for health care services received outside of the United States, even if those services are medically necessary. There is a limited exception for services occurring near U.S. borders. The constitutionality of foreign exclusion was raised in a 1986 court decision. In that case, plaintiffs argued that by leaving Medicare beneficiaries without health care coverage once they leave the United States, the foreign exclusion deters these same individuals from traveling overseas and consequently infringes on their constitutional right to travel abroad. The district court, however, found the Medicare foreign exclusion to pass constitutional muster. The court reasoned that the difficulties of administering medical services abroad and the concern that Medicare funds be spent within the United States were not particularly compelling, but were rationally based. It thus concluded that the foreign exclusion satisfied the rational basis test and summarily dismissed the plaintiff’s claim. The Bureau has the discretion to decide whether to count American citizens abroad. Indeed, there is nothing in the Constitution, the Census Act, or court decisions that would require the Bureau to count Americans living overseas, or to not count such individuals. Consequently, if Congress wanted to require the Census Bureau to include this population in the 2010 Census, legislation would be needed. The issue of whether, and if so how, to count Americans living overseas was raised in federal court after both the 1990 and 2000 censuses. Massachusetts challenged the results of the 1990 Census claiming that it lost a seat in the House of Representatives as a result of the Secretary of Commerce’s decision to enumerate and apportion federal employees stationed abroad. Conversely, Utah challenged the results of the 2000 Census maintaining that it lost a congressional seat because no overseas Americans other than federally affiliated groups were included in the apportionment numbers. In both cases, the courts determined that the Census Bureau has discretion under the Constitution and the Census Act to decide whether to count Americans residing overseas. The design of the Bureau’s overseas enumeration test falls short in many respects, and the data that the Bureau will collect as a result of this exercise may only partially answer key questions relating to feasibility, cost, and data quality. The Bureau overstated the test’s ability to answer its key research objectives and, as a result, congressional decision making on this issue will be that much more difficult. The full results of the overseas enumeration test will not be available until early 2005, when the Bureau expects to issue the last of a series of evaluations. However, its experience thus far has made it clear that counting Americans abroad as an integral part of the 2010 Census would be an extraordinary challenge, one that would introduce new resource demands, risks, and uncertainties to an endeavor that is already costly, complex, and controversial. That said, to the extent that better data on the number and characteristics of Americans abroad might be useful for various policymaking and other nonapportionment purposes, such information does not necessarily need to come from the decennial census. This data could, in fact, be acquired through a separate survey or some other type of data collection effort, although it would still be a difficult undertaking. The Bureau is unlikely to decide—and in fact should not decide—on its own, whether or not to enumerate Americans abroad, and will need congressional guidance on how to proceed. Therefore, to give the Bureau as much planning time as possible, it will be important for Congress to soon decide whether the Bureau should be required to count this population group as part of the 2010 Census or as part of a separate data collection effort or whether there are so many obstacles to a successful overseas count regardless of the approach that the Bureau should shelve any plans for further research and testing. Should Congress desire an overseas count—as part of the decennial census or a separate effortit should consider providing the Bureau with input on how it expects to use the data on overseas Americans. The purposes of the data drive the design of the enumeration; therefore, once the Bureau has a clear idea on how the data will be used, it would be better positioned to plan a test that will more accurately assess the feasibility of the procedures, methodology, and resource requirements needed to accomplish the type of count that Congress desires. Moreover, if a second test of enumerating Americans abroad is needed in 2006, it will be important for the Bureau to address the shortcomings of the design of the 2004 overseas test. Conducting a second test without this information and a sound design to fulfill it would not be cost-effective. In order to give the Bureau as much planning time as possible, Congress may wish to consider coming to an early decision on whether the Bureau should be required to enumerate overseas Americans, and if so, whether they should be counted as part of the decennial census or by some other, separate data collection effort. Should Congress desire an overseas count—be it part of the decennial census or a separate data collection effortit should consider telling the Bureau how the data would be used (e.g., for purposes of apportionment, redistricting, allocating federal funds, or a tally of the U.S. overseas population). This information would enable the Bureau to more thoroughly evaluate procedures and resources needed to meet Congress’s specific requirements, and ultimately provide Congress with better information with which to gauge the feasibility of such an approach. To the extent that the Bureau proceeds with plans to conduct a second feasibility test in 2006, we recommend that the Secretary of Commerce direct the Bureau to take appropriate steps to improve the soundness of the test design and better address the objectives of an overseas enumeration. Specific steps should include the following 12 actions: Be more transparent with Congress and other stakeholders on what variables the Bureau is able to quantitatively measure, as well as what research questions the Bureau can and cannot answer. Develop and pretest clear residence rules and appropriate guidelines on who should be included in the count to prevent confusing prospective respondents. Ensure that its outreach and promotion strategy, data collection methods, and other aspects of the design are free from cultural and other biases that could introduce systemic errors. Explore the possibility of developing more comprehensive measures of data quality. Test the Bureau’s ability to geocode certain noncity-style addresses such as those with post office box numbers. Research how best to market the overseas census to hard-to-count groups and other, less visible, segments of a country’s overseas American population. Develop procedures to determine whether a return is within the scope of the enumeration—i.e., that it is truly from an overseas source. Actually integrate overseas data with stateside data to more thoroughly test this activity. Develop a cost model to provide the Bureau and Congress with better estimates of the costs of conducting an overseas census under different methodological and other scenarios. Thoroughly pretest the overseas questionnaire with various groups of potential respondents to ensure the questions are clear, appropriate, and can be accurately answered by the unique population groups that are found overseas. Add more response options to the questionnaire item concerning respondents’ primary activity. Specific information on whether an individual is retired, a student, or working for a private company, etc., could provide the Bureau with valuable data on the characteristics of overseas Americans that could be important for some of the purposes for which the data might be used, and just as important, could provide the Bureau with invaluable marketing data that the Bureau could use to develop a more effective outreach and promotion campaign. Work with Congress and other stakeholders to develop a broad consensus on what would be acceptable levels of accuracy, completeness, participation, and other key measures of performance. The Secretary of Commerce forwarded written comments from the U.S. Census Bureau on a draft of this report on April 13, 2004, which are reprinted in appendix I. The Bureau generally agreed with our key findings, conclusions, and recommendations, and suggested some additional context, technical corrections, and clarifications, which we have incorporated. The Bureau disagreed with our recommendation that it be more transparent with Congress and other stakeholders on what variables the Bureau is able to quantitatively measure, as well as what research questions the Bureau can and cannot answer. The Bureau believes that it has always been transparent about its plans and the likely limitations of this first test. Although the Bureau’s test plan describes the limitations associated with answering its various research questions, nowhere does it disclose that its fundamental research objective to “determine the feasibility, quality, and cost of collecting data from U.S. citizens living overseas” will only be partially answered. Indeed, none of the documentation we reviewed, including the test project plan or briefing slides provided to congressional staff in April 2003, indicated either explicitly or implicitly that this test was, as the Bureau maintains, “only the most basic assessment of feasibility,” or that the Bureau would not be able to fully answer its key questions regarding feasibility, data quality, and cost. For example, as noted in our report, even though one of the principal objectives of the test was to determine the cost of collecting data from U.S. citizens living overseas, the Bureau’s test design lacked a research question for this objective. The Bureau agreed, however, that as it completes its evaluations and documents its findings from the test, it will be “critical” to highlight the various qualitative and quantitative limitations that could affect congressional deliberations on this subject. As agreed with your offices, unless you release its contents earlier, we plan no further distribution of this report until 30 days from its issue date. At that time we will send copies to other interested congressional committees, the Secretary of Commerce, and the Director of the U.S. Census Bureau. Copies will be made available to others upon request. This report will also be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-6806 or by e-mail at daltonp@gao.gov or Robert Goldenkoff, Assistant Director, at (202) 512-2757 or goldenkoffr@gao.gov. Key contributors to this report were Lisa Pearson, Charlesetta Bailey, Betty Clark, Ellen Grady, Ronald La Due Lake, Andrea Levine, and Daniel Messler. 1. How clearly are the objectives of the research design presented? a. Are the test/research objectives and/or questions specified clearly in the design? b. Are concepts defined where necessary? 2. How sound is the research design? a. Do the research questions match the research objectives? b. Is the rationale for the determination of the size and type of sample explained? c. Are potential biases recognized and addressed (e.g., cultural bias, question item bias, or sample bias)? d. Does the sample selection reflect the full range and full cycle of entities or processes that should be considered? 3. How appropriate is the data collection strategy? a. Is the mode of data collection stated clearly? b. Is the timing and frequency of data collection considered? c. Is the data collection method appropriate for reaching the intended sample population? d. Is the data collection instrument appropriate for the sample population? e. Is a plan for administering and monitoring the data collection discussed in the design? f. How well does the design address factors that may interfere with data collection? g. How well are methods for assuring adequate response rates addressed? 4. How thoroughly did Census test the survey instrument(s)? a. Were any new survey items cognitively tested? b. Were field tests conducted to identify the best question wording and determine whether changes in questions are likely to achieve the change objectives? c. Were research studies used to address possible changes in the questionnaire? 5. How involved were relevant stakeholders in the research planning? a. Were relevant stakeholders for the research objectives identified? b. Were stakeholders involved in the planning or review of the methods of data collection? c. Were appropriate stakeholders participants in the review and testing of the questionnaire? 6. How sound is the design’s plan for reaching the target sample? a. Are the goals of the outreach strategy feasible? b. Are the methods of the outreach strategy viable? 7. How sound is the plan for implementing test site activities? a. Does the design consider possible mistakes and their consequences (including their seriousness)? b. Does the design assure that sufficient evidence will be gathered to answer the research questions? c. Does the design consider the level of difficulty in obtaining the data? 8. How good is the relationship between the research design and the time and resources allocated? a. Does the execution of the design appear feasible within the stated time frame? b. Do the proposed resources for the execution of the design appear feasible? c. Are the roles and responsibilities of key parties specified? 9. How well does the design develop a data analysis plan? a. Is the method of enumeration clearly presented? b. Are the proposed analytical techniques for analysis presented? c. Does the design provide a basis for comparing the results of the research? d. Does the design discuss the possible limitations of the findings/test results? The General Accounting Office, the audit, evaluation and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to e-mail alerts” under the “Order GAO Products” heading.

In the 1990 and 2000 Censuses, U.S. military and federal civilian employees overseas were included in the numbers used for apportioning Congress. Currently, the U.S. Census Bureau (Bureau) is assessing the practicality of counting all Americans abroad by holding a test census in France, Kuwait, and Mexico. GAO was asked to (1) assess the soundness of the test design, and (2) examine what past court decisions have held about Americans' rights and obligations abroad. Although the overseas enumeration test was designed to help determine the practicality of counting all Americans abroad, because of various methodological limitations, the test results will only partially answer the Bureau's key questions concerning feasibility, data quality, and cost. For example, one research questions asks, "How good is the quality of the data?" However, the Bureau will only measure item nonresponse, which indicates whether a person completed a particular question. As a measure of quality, it is far from complete. Similarly, although a key research objective was to determine the cost of counting Americans overseas, the Bureau's data will not inform the cost of conducting future tests or an overseas enumeration in 2010. Overall, the Bureau overstated the test's ability to answer its key research objectives. Overseas Americans have various rights and obligations to federal programs and activities. For example, Americans abroad are generally taxed on their worldwide income and can vote in federal elections, but are generally not entitled to Medicare benefits. There is nothing in the Constitution, federal law, or court decisions that would either require the Bureau to count overseas Americans, or not count this population group. As a result, Congress would need to enact legislation if it wanted to require the Bureau to include overseas Americans in the 2010 Census. Counting Americans abroad as part of the census would add new risks to an enterprise that already faces an array of challenges. Therefore, it will be important for Congress to decide whether overseas Americans should be counted as part of the census or counted as part of a separate survey or whether there are so many obstacles to a successful count regardless of the approach that the Bureau should shelve any plans for further research and testing. To the extent a second test is required, the Bureau will need to take steps to develop a more rigorous design.

To establish and manage its inventories, the Coast Guard must comply with the criteria contained in federal property management regulations and with the Department of Transportation’s (DOT) policy. Federal property management regulations state that each agency shall establish and maintain control of inventories to ensure that total costs will be kept to a minimum consistent with the needs of the agency’s programs. DOT’s policy states that inventories will be established and maintained only when it is more costly to purchase items on a case-by-case basis or when the items are so critical that a delay in delivery would negatively affect an agency’s mission. The policy also states that inventories must be managed in an effective manner to ensure that timely and adequate support is rendered and that optimum inventory levels are maintained. Within the Coast Guard, supply centers located at Curtis Bay and Baltimore, Maryland, stock about 18,000 different parts, including mechanical, electrical, radar, communication, computer, and hull items; these parts are valued at about $140 million. In addition, the Coast Guard authorizes each cutter to maintain a parts and supplies inventory that ranges in value from a few thousand dollars to over a million dollars, depending on the cutter’s size, missions, and operating area. However, individual purchases are restricted to established price thresholds that vary, depending on the cutter’s class; the thresholds range up to $5,000 for large cutters. (Table II.1 lists the 41 different classes of Coast Guard cutters.) For example, the cutters usually purchase such items as nuts and bolts, valves, seals, gaskets and other minor repair parts with their operating budgets. According to Coast Guard officials, about 55 percent of the cutters’ parts and supplies are purchased directly from commercial contractors and about 45 percent are purchased from the federal supply system. The cutters’ purchases from the federal supply system are divided between orders from the two Coast Guard supply centers and orders from other government agencies. Such other agencies as the Defense Logistics Agency; the Departments of the Army, Navy, and Air Force; and the General Services Administration fill about 90 percent of orders for parts from the federal supply system, while the Coast Guard’s two supply centers in Maryland fill the remaining 10 percent. Since storage space is limited on most vessels, cutters also store some of their parts at individual onshore storage facilities, including movable storage in tractor trailers, as well as at typical warehouse-type buildings. The Coast Guard does not have the organizational structure or computer systems necessary to effectively manage its parts and supplies inventory for its cutters. As a result, the Coast Guard does not know the value, type, quantity, and condition of many of the spare and repair parts in the overall inventory. In addition, the Coast Guard does not know whether cutters have a shortage or an excess of parts or whether the parts are available when needed. Management responsibilities for buying, storing, issuing, and tracking parts and supplies in the Coast Guard’s inventories are spread across various internal and external organizations. For example, the Maintenance and Logistics Commands at Alameda, California, and Governors Island, New York, manage the parts used during overhauls of cutters; the Coast Guard’s two supply centers in Maryland manage the unique parts needed by the cutters they support; and individual cutters manage parts they need to keep them operationally ready. However, no one organization or individual is responsible for consolidating inventory data and tracking the type, quantity, condition, or value of the Coast Guard’s total cutter inventory. The Coast Guard’s fragmented management structure also limits the agency’s ability to determine whether cutters have a shortage or an excess of parts and whether the parts are readily available when needed. We found, for example, that one cutter had 34 excess fuel injector nozzles (above its allowance of 16 nozzles), which cost about $580 each, and two excess starters (above its allowance of one starter), which cost approximately $6,600 each. During our visits to other cutters, we noted many other such excess items as drill presses, main engine cylinder heads, insulation, computer monitors, and galley equipment. Supply officials on the cutters told us that the excesses in their inventories were the result of several factors. For example, sometimes cutters procured larger quantities than needed to take advantage of volume discounts. However, if the Coast Guard centrally managed—at headquarters, a Maintenance and Logistics Command, or a supply center—the total cutter inventory, wherever located, it might be able to transfer excess items to cutters that have shortages of those items. For example, the Atlantic Coast Maintenance and Logistics Command purchased more than 15 new starters in 1 year, while one cutter had 2 excess starters in its inventory. Despite these excesses, officials responsible for the individual cutters’ inventories told us that they also had shortages of parts. For example, one cutter had parts shortages totaling $250,000 for such electronic items as circuit boards for radar and communication systems. According to these officials, these kinds of shortages occurred primarily because funding was not available to replenish the cutters’ parts or because the required parts were never issued when the cutters were first commissioned. If the Coast Guard centrally managed its total cutter inventory, some of these shortages might have been filled with excess items from other vessels’ inventories. The cutter officers noted, however, that although the shortages had not significantly affected their missions, they had resulted in costly emergency purchases that would not have been necessary if the parts had been in the cutters’ inventories, as required. During our visits, we found that many items in the cutter inventories were not available to the vessels when they were at sea. Each of the cutters we visited stored a portion of its inventory in its own individual onshore storage facility. Since no one was available to issue parts from these individual storage facilities when the cutters were at sea, these inventories were not fully utilized. We noted such useful items as valves, filters, engine and hydraulic oil, mooring lines, and damage control equipment—fire hoses and nozzles, submersible pumps, shoring, plugs, and oxygen canisters—in the individual onshore facilities. Unlike the Coast Guard, the Navy stores parts for its ships in centralized base supply centers and does not maintain individual onshore storage facilities for its ships. According to Navy officials, the centralized base supply centers provide more effective support than individual storage facilities because personnel are available at the centers to issue parts to the ships whenever the parts are needed. The centers can, for example, send the needed parts to ships at sea via another ship or an aircraft. When a Coast Guard cutter is at sea, no one is available to issue parts from the cutter’s individual onshore storage facility. According to Coast Guard officials, the agency is studying the use of regional support centers, but it does not expect to consolidate the individual onshore storage facilities before fiscal year 2002, when it expects to have total “visibility” of its cutters’ individual inventories. The Coast Guard uses several different systems to manage the cutters’ individual inventories. For example, during our visits to nine cutters, we observed a manual and three different computerized inventory control systems. However, the three automated systems that we observed could not exchange data with each other or with the systems at headquarters, the Maintenance and Logistics Commands, or the supply centers. To help ensure effective inventory management and to distribute parts more efficiently, the Coast Guard plans to implement a single automated system, CMplus, on its 101 largest cutters. The Coast Guard expects this system to integrate inventory and maintenance information into a larger fleet logistics system that will enable cutter crews to share data with each other, headquarters, the Maintenance and Logistics Commands, and the supply centers by the year 2002. According to Coast Guard officials, they expect to spend over $27 million to install the CMplus system on the cutters. Although we agree with the goals of the CMplus system, the Coast Guard can take interim actions to enhance the distribution of its inventory between now and the year 2002, when CMplus will be fully implemented. For example, the Coast Guard can utilize the most widely used, existing inventory system to enhance the distribution of parts by sending the inventory information to headquarters and to the Maintenance and Logistics Commands for analysis. In 1993, Coast Guard headquarters and one Maintenance and Logistics Command analyzed the inventories held by all of the 270-foot cutters. Because all 13 of these cutters used the same computer system, the Coast Guard was able to consolidate their inventory data. The data showed that the 13 cutters had more than $11 million worth of excess parts in their inventories and that $3 million of the excess could be redistributed among the cutters to offset their parts allowance shortfalls and reduce future acquisitions. The Coast Guard incurred minimal time and costs (less than 1 staff year) to perform the analysis because all of the 270-foot cutters had conducted full physical inventories of their parts and supplies and implemented a computerized inventory control program. Moreover, the payoff was significant and could be increased if the Coast Guard conducted similar analyses for most other classes of cutters because they have also already conducted full physical inventories when they implemented their computerized inventory control systems. Although such analyses could greatly improve the distribution of parts and supplies, they could not themselves ensure the optimal distribution of inventories for two reasons. First, because the cutters cannot directly transmit inventory data to the Maintenance and Logistics Commands conducting the analysis, the consolidated inventory information would not be current when the cutters began to redistribute their parts. Second, the Maintenance and Logistics Command conducted the redistribution study on a single class of cutters because conducting a fleetwide study would have taken much longer using current computer resources. If future analyses are conducted for only one class of cutter at a time, parts and supplies will not be redistributed between classes. When the Coast Guard’s fleet logistics system is implemented, it will deal with these two limitations of the current system. In 1993, the Coast Guard issued its Logistics Master Plan. The plan addresses numerous issues—for example, the Coast Guard’s lack of central management for the cutter inventories. The plan includes short-term actions that the Coast Guard expected to complete by fiscal year 1994, mid-term actions to be completed by fiscal year 1997, and 26 long-term actions that the Coast Guard expected to complete by the end of fiscal year 2002. Although we agree with the plan’s direction, we found that some initiatives are already behind schedule, increasing the potential for delays in the Coast Guard’s long-term efforts to centrally manage its inventories by fiscal year 2002. (App. I lists some of the initiatives that are in progress or planned.) Coast Guard officials told us that before the agency can centrally manage its inventory, they must complete the following long-term initiatives in the Logistics Master Plan: Develop a fully integrated, real-time computer system to track and consolidate inventory and maintenance information for the cutters. Create a single organization to integrate the maintenance guidance, technical, and supply functions now performed by headquarters and the two inventory supply centers. Designate an official to be responsible for all fleet logistics. We found that the completion of these and many other long-term initiatives in the plan are contingent upon the Coast Guard’s successfully completing numerous near- and mid-term initiatives. However, the Coast Guard has already experienced schedule slippages with some of the near- and mid-term initiatives begun in 1993 and 1994. For example, the Coast Guard had expected to implement the following new initiatives: A computerized inventory control system, CMplus, on the first cutter of the 378-foot class by the end of 1993. However, the system will not be operational until December 1994 (a 1-year delay) because of such operational commitments as transporting Haitian and Cuban refugees. According to officials, CMplus is a critical part of the Coast Guard’s long-term initiative to develop a fully integrated, real-time system to track and consolidate inventory and maintenance data for its cutters. A computer system at Curtis Bay by the fourth quarter of fiscal year 1994. However, the Coast Guard does not expect to have the system fully implemented until the third quarter of fiscal year 1996 (almost a 2-year delay) because of a 1-year delay in the award of the hardware contract and because of software development problems. According to officials, this system is needed to enable the Coast Guard to track and consolidate inventory and maintenance data for the cutters. Centralized shoreside support for its 110-foot cutters (49 vessels) by fiscal year 1996. The Coast Guard now expects to have this new management structure by fiscal year 1998 (a 2-year delay). Until that time, according to officials, the Coast Guard cannot centrally manage its cutter inventories because it does not have visibility of the inventories. The $140 million inventory held at the Coast Guard’s two supply centers does not reflect the agency’s total investment in spare and repair parts for its cutters. The Coast Guard does not know the type, quantity, condition, or total value of its total inventory of parts and supplies for its cutters. However, the Coast Guard estimates that it has additional inventory worth approximately $200 million stored onboard its cutters and in the cutters’ individual onshore storage facilities. Although Coast Guard officials contend that the agency’s lack of information on parts and supplies has not significantly affected the Coast Guard’s mission, it has resulted in inefficient management of resources. Consequently, the agency cannot minimize the cost of its total inventory as required by federal property management regulations and DOT’s policy. In addition, the Coast Guard does not expect to complete its integrated system to enhance the use and distribution of its inventory until the year 2002. Yet delays of as much as 2 years for some early initiatives raise concerns that the Coast Guard will not meet its targeted completion date. Since the Coast Guard may take many years to improve its inventory management system, some actions now could help alleviate shortages and excesses and help the Coast Guard better utilize its inventories. To enable the Coast Guard to manage its cutter inventories more effectively between now and when the Logistics Master Plan is fully implemented, we recommend that the Secretary of Transportation direct the Coast Guard Commandant to take the following interim actions: Make the use of the current automated inventory control program mandatory on all cutters that have sufficient computer hardware and have not implemented CMplus, consolidate and analyze inventory data for each class, and redistribute excess parts from additional cutter classes as warranted. Where economically feasible, consolidate at regional support centers those cutter inventories that are located at individual onshore storage facilities, particularly where several cutters from the same class are clustered or where the cutters’ individual onshore storage facilities are housed within a single building. Move up the implementation date for the Coast Guard’s initiative to establish a single source of accountability for all fleet logistics. This action will allow the Coast Guard to better coordinate interim actions to improve management of its cutter inventories while the fleetwide logistics system is being developed. We discussed this report with the Coast Guard’s Chief, Logistics Management Division, Office of Engineering, Logistics, and Development, and with other program officials, and we have incorporated their comments as appropriate. These officials generally agreed with our findings and recommendations. We conducted our work between November 1993 and December 1994 in accordance with generally accepted government auditing standards. Our objectives, scope, and methodology are discussed in appendix II. We are sending copies of this report today to the Secretary of Transportation; the Commandant, Coast Guard; and the Director, Office of Management and Budget. We will make copies available to others upon request. This work was performed under my direction. If you have any questions, I can be reached at (202) 512-2834. Major contributors to this report are listed in appendix III. The Coast Guard’s Logistics Master Plan sets out short-, mid-, and long-term objectives to improve the agency’s inventory controls by fiscal year 2002. This appendix provides (1) a brief description of the major initiatives related to central management of the Coast Guard’s inventories and (2) the status of the initiatives that were scheduled for completion in fiscal years 1993 and 1994. The Coast Guard relocated the Brooklyn, New York, supply center to Baltimore in 1993, as planned. This action was the first step toward the Coast Guard’s creating a single organization to integrate the maintenance, technical, and supply functions now performed by headquarters and the two inventory supply centers. The Coast Guard designated one Maintenance and Logistics Command to be responsible for an entire class of cutters, regardless of their home ports, including the development of a maintenance plan that lists the minimum information needed for a major overhaul or minor repairs at shipyards and bases. This initiative helped Curtis Bay to increase the availability of parts for 270-foot cutters from 65 percent in March 1992 to 94 percent in March 1994. In addition, the Coast Guard developed improved maintenance plans and long-range forecasts for its 210-, 180-, 157-, and 140-foot cutters in 1993 and 1994, as scheduled. The Coast Guard implemented a central supply department on its 378-foot cutters in 1992 and on its 399- and 270-foot cutters in 1993, as scheduled. Previously the Coast Guard maintained department-level inventories on these cutters that resulted in duplicate procurements, excess spare parts, reduced storage capacities, and longer casualty response times, according to Coast Guard officials. The new centralized supply departments have helped to alleviate many of these problems because all of the cutters’ parts information is located in one data base. Centralization of parts information also helps to save space on the cutters because duplicate parts that were previously stocked by more than one department are readily visible and can either be used, transferred, or scrapped. Finally, a central supply department increases operational readiness because procurements are coordinated across departments, making more effective use of available spare parts funding. The Coast Guard implemented an automated system, CMplus, to integrate shipboard supply and maintenance information on three of its cutters. According to Coast Guard officials, this system will be the cornerstone of its centralized fleet logistics system. The Coast Guard implemented the system on 1 of its 210-foot cutters (a class of 16 vessels), 1 of its 270-foot cutters (a class of 13 vessels), and 1 of its 378-foot cutters (a class of 12 vessels) in 1994. The Coast Guard had placed a prototype CMplus system on a 140-foot cutter in 1992 and had expected to implement the system on its eight remaining 140-foot cutters by the fourth quarter of 1994, but this date has slipped to fiscal year 1997. The Coast Guard had expected to purchase the hardware to replace its supply center computers in the fourth quarter of fiscal year 1993. Although the agency has purchased developmental hardware for the new system, procurement of the new production hardware is now scheduled for the third quarter of fiscal year 1995. This purchase is a key step in instituting the standardized fleet logistics system that the Coast Guard expects to have fully operational by the year 2002. The Coast Guard had expected to develop the software for its new supply center computer system by the end of 1994. Although the Coast Guard wants to get the new system on line as quickly as possible, the projected date for the initial software has slipped to the third quarter of 1996 because of technical difficulties and a delay in purchasing the needed hardware. Develop improved maintenance plans and long-range spare parts forecasts for the 110-foot cutters. Implement a central supply department on the 210-, 140-, and 110-foot cutters and study the feasibility of implementing it on smaller cutters and bases. Analyze the feasibility of transferring management of such consumable items as nuts, bolts, and bearings to the Defense Logistics Agency. Install the new automated inventory control system on the remaining 270-foot cutters and on the 399-foot cutters. Implement building block, software application groups for a standardized fleetwide logistic system. The application groups will include maintenance planning, scheduling, funds management, parts tracking, contract management, supply performance measures, and cost analysis. Install the new automated inventory control system on the remainder of the 378-foot and 210-foot cutters and on the 110-foot cutters. Integrate maintenance, technical, and supply functions, which are now performed by headquarters and the two supply centers, into a central engineering logistics center at Curtis Bay. Designate a single official responsible for all logistics. Complete procurement of both the hardware and software for the standardized fleet logistics system and implement the remaining software application groups, including customer service, technical information, and equipment management. The Coast Guard expects that this system will integrate shipboard logistics systems with shoreside systems so that the supply centers will have information about the cutters’ inventories, equipment usage, and costs. The former Chairman, Subcommittee on Oversight of Government Management, Senate Committee on Governmental Affairs, asked us to examine the Coast Guard’s inventory management system to identify any wasteful or inefficient practices that should be changed. As agreed with the former Chairman’s office, we focused our review on the Coast Guard’s inventory management system for its 240 cutters (vessels 65 to 399 feet in length) and developed the following specific questions to guide our work. First, does the Coast Guard have the systems needed to effectively manage its inventory of spare and repair parts and supplies? Second, if not, what initiatives does the Coast Guard have under way to improve its inventory management? In preparing this report, we reviewed federal property management regulations (41 C.F.R. 101); the Department of Transportation’s Order 4420.5, Management of Material Inventories; and the Coast Guard’s Supply Policy and Procedures Manual. To determine the cost effectiveness of the Coast Guard’s inventory management systems, we met with officials from the Coast Guard’s supply centers at Baltimore and Curtis Bay, Maryland, and reviewed their instructions, notices, and video tapes related to inventory management and supply support. We also reviewed Curtis Bay’s Supply Activity Reports for 1989 through 1993 and its list of inventory items for the 378-, 270-, and 210-foot cutters. We also met with officials from headquarters; the Maintenance and Logistic Command for the Atlantic Fleet; the Coast Guard District Five Office and the Naval Engineering Support Unit in Portsmouth, Virginia; the Coast Guard Group/Air Station in Cape May, New Jersey; and individual cutters. Using the Coast Guard’s register of cutters, we selected a judgmental sample of cutters to visit. Because of the large number of Coast Guard cutters (240), we defined our sample in three ways. First, we selected only cutters that were at least 82 feet long because larger cutters typically hold more inventory than smaller cutters. Second, we visited only cutters that had at least five ships in the class because we wanted the cutters to be typical of the largest number of cutters possible. Finally, when two or more classes existed for vessels of the same length and type (i.e., 210-foot, medium endurance cutters, 210A and 210B), we visited only one cutter from the combined classes because the cutters in the combined classes were still very similar to each other. Table II.1 lists the type, class, and number of Coast Guard cutters. Table II.2 lists the name, type, location, and size of each of the nine cutters we visited. The classes of the nine cutters account for 178 of the Coast Guard’s 240 cutters. (continued) Philadelphia, Pa. Governors Island, N.Y. Cape May, N.J. Cape May, N.J. Cape May, N.J. Cape May, N.J. Boston, Mass. Newport, R.I. Portsmouth, Va. To determine the Coast Guard’s initiatives related to its inventory controls, we reviewed the Coast Guard’s 1993 Logistics Master Plan. We also obtained information on the actions that the Coast Guard had undertaken that were not part of the Logistics Master Plan, such as the Supply Center Information Systems Plan and user manuals for the computerized inventory systems used on the Coast Guard’s larger cutters. We met with headquarters, supply center, and Maintenance and Logistics Command officials who were responsible for these initiatives to determine their status and obtain clarification on the benefits expected. (App. I describes some of the initiatives and their status.) M. Glenn Knoepfle, Adviser Michael J. Ferren, Evaluator-in-Charge The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (301) 258-4097 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed the Coast Guard's inventory management system for its fleet of 240 cutters to identify any wasteful or inefficient practices that should be changed, focusing on: (1) whether the Coast Guard has the systems it needs to effectively manage its inventory of spare and repair parts and supplies; and (2) initiatives the Coast Guard has under way to improve its inventory management. GAO found that: (1) the Coast Guard does not have the organizational structure or computer systems necessary to effectively manage its inventory for supporting cutters; (2) the Coast Guard does not know the value, type, quantity, and condition of many of the spare and repair parts in its inventory and cannot determine whether cutters have a shortage or an excess of parts, or whether the parts are readily available; (3) Coast Guard officials believe that the lack of inventory management information has not seriously affected the Coast Guard's ability to carry out its missions, although they acknowledge that it has resulted in costly purchases and excess inventory; (4) the Coast Guard plans to take actions to improve its inventory control problems by fiscal year 2002; and (5) some of the Coast Guard's initiatives are already behind schedule, delaying both its short-term actions and its long-term inventory management efforts.

We assessed the DEAMS schedule using the GAO Schedule Guide to determine whether it was comprehensive, well-constructed, credible, and controlled. To assess the schedule, we obtained and reviewed documentation, including the integrated master plan and work breakdown structure. In assessing the program’s cost estimate, we used the GAO Cost Guide to evaluate the DEAMS Program Management Office’s estimating methodologies, assumptions, and results to determine whether the cost estimate was comprehensive, well-documented, accurate, and credible. We obtained and reviewed documentation, including the program office estimate, software cost model, independent cost estimate, and risk and uncertainty analysis. We also interviewed key program officials, such as the Program Manager, lead schedulers, and cost estimators, to obtain information, such as explanations to resolve identified discrepancies. After we briefed DEAMS program officials on the results of our assessment, they provided an updated schedule dated October 2012. For this updated schedule, we determined the extent to which it met certain best practices for the comprehensive, well-constructed, and credible characteristics, because not implementing these best practices would affect the reliability of the entire schedule. In May 2013, program management officials provided another updated DEAMS schedule, which they acknowledged contained issues that prevented the schedule from meeting best practices. Although we did not independently assess the May 2013 schedule, we did confirm that it included certain information needed for long-term planning. We conducted this performance audit from May 2012 to February 2014 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. DEAMS was initiated in August 2003 and is intended to provide the Air Force with the entire spectrum of financial management capabilities, including collections, commitments and obligations, cost accounting, general ledger, funds control, receipt and acceptance, accounts payable and disbursement, billing, and financial reporting for the general fund. In February 2012, the DOD Deputy Chief Management Officer granted Milestone B approval for DEAMS to enter the Engineering Development Phase of the acquisition life cycle, which is considered the official start of the program. DEAMS program functionality is intended to be implemented across the Air Force in a series of releases in two increments—Increment 1 will include six releases and Increment 2 will include two releases. DOD has approved the funding for the Air Force to proceed with the acquisition of the functionality for the first increment of DEAMS. This funding is approximately $1.6 billion, with deployment scheduled to occur during the fourth quarter of fiscal year 2016. The Air Force reported that it had spent about $427.5 million as of September 30, 2013, on the program. As stated earlier, in October 2010, we reported that although the Air Force met best practices in developing a cost estimate, it did not meet best practices in developing the schedule estimate for implementing DEAMS. In particular, the Air Force had not developed a fully integrated master schedule that reflected all government and contractor activities. We recommended that the Air Force develop an integrated master schedule that fully incorporated best practices, such as capturing all activities, sequencing all activities, integrating activities horizontally and vertically, establishing the critical path for all activities, identifying float between activities, conducting a schedule risk analysis, and updating the schedule using logic and durations to determine dates. DOD concurred with our recommendation, and we discuss later in this report the status of DOD’s efforts to address this recommendation. In March 2009, we published the Cost Guide to address a gap in federal guidance about processes, procedures, and practices needed for ensuring reliable cost estimates. The Cost Guide provides a consistent methodology based on best practices that can be used across the federal government to develop, manage, and evaluate capital program cost estimates. The methodology is a compilation of characteristics and associated best practices that federal cost estimating organizations and industry use to develop and maintain reliable cost estimates throughout the life of an acquisition program. In May 2012, we issued the Schedule Guide as a companion to the Cost Guide. A consistent methodology for developing, managing, and evaluating capital program cost estimates includes the concept of scheduling the necessary work to a timeline, as discussed in the Cost Guide. Simply put, schedule variances are usually followed by cost variances. Because some program costs, such as labor, supervision, rented equipment, and facilities, cost more if the program takes longer, a reliable schedule can contribute to an understanding of the cost impact if the program does not finish on time. In addition, management tends to respond to schedule delays by adding more resources or authorizing overtime. Further, a schedule risk analysis allows for program management to account for the cost effects of schedule slippage when developing the life-cycle cost estimate. A cost estimate cannot be considered fully credible if it does not account for the cost effects of schedule slippage. A well-planned schedule is a fundamental management tool that can help government programs use public funds effectively by specifying when work will be performed in the future and measuring program performance against an approved plan. Moreover, as a model of time, an integrated and reliable schedule can show when major events are expected to occur as well as the completion dates for all activities leading up to them, which can help determine if the program’s parameters are realistic and achievable. A program’s success depends in part on the quality of its schedule. We found that the schedule for the DEAMS program did not meet best practices. The cost estimate did meet best practices, but the issues associated with the schedule could negatively affect the cost estimate. Specifically, the DEAMS schedule supporting the February 2012 Milestone B decision partially or minimally met the four characteristics for developing a high-quality and reliable schedule—it was not comprehensive, well-constructed, credible, or controlled. In addition, our assessment of the October 2012 updated schedule found that it was not comprehensive, well-constructed, and credible and thus was also not reliable. In contrast, the DEAMS cost estimate fully or substantially met the four characteristics of a high-quality and reliable cost estimate—it was comprehensive, well-documented, accurate, and credible. However, because the cost estimate is based on the schedule, the unreliability of the schedule could affect the cost estimate. For example, if there are schedule slippages, the costs for the program could be greater than currently estimated. Our analysis found that the DEAMS program partially met three and minimally met one of the characteristics of a reliable schedule estimate and therefore did not provide the information needed to support the February 2012 Milestone B decision (see table 1). Appendix I contains our detailed analysis of the DEAMS schedule estimate. The success of any program depends on having a reliable schedule of the program’s work activities that will occur, how long they will take, and how the activities are related to one another. As such, the schedule not only provides a roadmap for systematic execution of a program, but also provides the means by which to gauge progress, identify and address potential problems, and promote accountability. Comprehensive. A schedule should reflect all activities as defined in the program’s work breakdown structure, including activities to be performed by the government and the contractor; the resources (e.g., labor, materials, and overhead) needed to do the work; and how long each activity will take. We found that the schedule used to support the Milestone B decision included the activities to be performed by both the government and contractor for Releases 1 through 3 of Increment 1. However, the schedule did not reflect activities to be performed for Releases 4 through 6 of Increment 1 and for Releases 1 and 2 of Increment 2. The DEAMS Program Manager stated that a comprehensive schedule for Increment 1 that included the activities for all six releases would not be completed until mid-2014. The Program Manager also stated that Increment 2 had not been included because program officials did not know the detailed activities to be performed that far in advance. To address this issue, the DEAMS program office developed a roadmap depicting Releases 1 through 6 of Increment 1 and Releases 1 and 2 of Increment 2 with a full deployment date of fiscal year 2017. However, the program office did not provide a schedule that supported the estimated dates in the roadmap. A comprehensive schedule should reflect all of a program’s activities and recognize that uncertainties and unknown factors in schedule estimates can stem from, among other things, data limitations. As such, a schedule incorporates different levels of detail depending on the information available at any point in time. That is, near-term effort will be planned in greater detail than long-term effort. Effort beyond the near term that is less well defined is represented within the schedule as long-term planning packages. Planning packages are a summarization of the work to be performed in the distant future with less specificity. Planning packages are planned at higher levels such that a single activity may represent several months of effort, generic work to be accomplished, or even a future contract or phase. Planning packages can be used as long as they are defined and estimated as well as possible. By not including all work for all deliverables for both increments and all releases, the DEAMS program could incur difficulties resulting from an incomplete understanding of the plan and what constitutes a successful conclusion for the program. DEAMS program officials provided a draft of the Schedule Management Plan that documented their intent to use a planning package approach when updating the DEAMS schedule in the future. Resources were identified in the schedule; however, the resources were not assigned to specific activities in the schedule. Although our analysis determined that activity durations were manageable and reasonably estimated, resource availability affects estimates of work and its duration, as well as resources that will be available for subsequent activities. DEAMS program management officials told us that government resource allocations are determined by management as needed. These officials told us that management does not necessarily take into consideration the resource information captured in the schedule when determining resource allocations. However, DEAMS officials did not provide any documentation that specific resources were being mapped to the schedule. As mentioned above, the estimates of work required and duration for an activity are tied to the availability of resources; therefore, the lack of such information could hinder management’s ability to compute total labor and equipment hours, calculate total project and per-period cost, resolve resource conflicts, and establish the reasonableness of the plan. Well-constructed. A schedule should be planned so that critical project dates can be met. To meet this objective, all activities should be logically sequenced—that is, listed in the order in which they are to be carried out. In particular, activities that must finish prior to the start of other activities (i.e., predecessor activities), as well as activities that cannot begin until other activities are completed (i.e., successor activities), should be identified and their relationships established. The establishment of a critical path is necessary for examining the effects of any activity slipping along this path. The calculation of a critical path determines which activities drive the project’s earliest completion date. The schedule should also identify total float so that the schedule’s flexibility can be accurately determined. We found that the majority of logic used to sequence the activities within the schedule was generally error-free with a minimal use of lags, clearly indicating to program management the order of activities that must be accomplished. Although we found few missing logic relationships for Release 3 of Increment 1, approximately 25 percent of the remaining activities for Releases 1 and 2 of Increment 1 were missing logic relationships. Because interdependencies among activities were not identified, the DEAMS program management officials’ ability to properly calculate dates and predict changes in the future is impaired. We found a significant number of constraints for activities throughout the schedule. A schedule is intended to be a dynamic, proactive planning and risk mitigation tool that models the program and can be used to track progress toward important milestones. Schedules with constrained dates can portray an artificial or unrealistic view of the project. Constraints should be minimized because they can create false dates in a schedule. Further, the schedule did not have a valid critical path and identified critical activities more by their constraints than by logic. Rather than relying on constraints, the schedule should use logic and durations in order to reflect realistic start and completion dates for activities. Successfully identifying the critical path relies on several factors, such as capturing all activities, properly sequencing activities, and assigning resources, which, as noted earlier, had not been done. Without a valid critical path, management cannot focus on activities that will have detrimental effects on the key project milestones and deliveries if they slip. We found that total float was not reasonable, and that in some instances unreasonable float was a direct result of improper sequencing or missing logic. Releases 1 and 2 of Increment 1 showed that 25 percent of program activities had total float equal to or greater than 392 working days, meaning that those activities could slip almost 2 working years and not affect the end date of the program. Without knowledge of the reason float exists for a program activity, management cannot determine the flexibility of tasks and therefore cannot properly reallocate resources from tasks that can safely slip to tasks that cannot slip without adversely affecting the estimated program completion date. Credible. A schedule should be horizontally and vertically integrated. A horizontally integrated schedule links products and outcomes with other associated sequenced activities, which helps verify that activities are arranged in the right order to achieve aggregated products or outcomes. A vertically integrated schedule ensures that the start and completion dates for activities are aligned with such dates on subsidiary schedules supporting tasks and subtasks. Such mapping or alignment among subsidiary schedules enables different groups—such as government teams and contractors—to work to the same master schedule, and provides assurance that the representation of the schedule to different audiences is consistent and accurate. A schedule risk analysis should also be performed using statistical techniques to predict the level of confidence in meeting a program’s completion date. We found that Release 3 of Increment 1 exhibited horizontal integration, but Releases 1 and 2 of Increment 1 did not because date constraints prevented forecasted dates from being calculated realistically for future activities. If the schedule lacks horizontal integration, activities whose durations are greatly extended will have no effect on key milestones reflected in the schedule. We further found that Releases 1 and 2 of Increment 1 did not demonstrate vertical integration. For example, we found instances where the start dates for the same activities differed by 1 day, 1 week, and 1 month between the government and contractor schedules. Unless the schedule is vertically integrated, lower-level schedules will not be consistent with upper-level schedule milestones, affecting the integrity of the entire schedule and the ability of different teams to work to the same schedule expectations. DEAMS program management officials stated that a schedule risk analysis had not been conducted because the schedule had not been approved to be used as a baseline schedule—the target schedule against which program performance can be measured, monitored, and reported. These officials stated that although this analysis had not been conducted, they were collecting best-case and worst-case durations from the contractor with their periodic schedule delivery. These data can be used by program management to calculate more reliable estimates of durations for future activities. However, we found that the schedule did not contain best- or worst-case duration data for 600 of 605 detailed activities. For the five instances where duration data were contained in the schedule, we determined that four were questionable because two activities were already completed and two had already exceeded the worst-case estimate. If a schedule risk analysis is not conducted, program management cannot determine the likelihood of the project’s completion date, how much schedule risk contingency is needed to provide an acceptable level of certainty for completion by a specific date, risks most likely to delay the project, how much contingency reserve each risk requires, and the paths or activities that are most likely to delay the project. As discussed later, the lack of a schedule risk analysis can affect the credibility of the cost estimate. Controlled. A schedule should be continuously updated using logic, durations, and actual progress to realistically forecast dates for program activities. A schedule narrative should accompany the updated schedule to provide decision makers and auditors a log of changes and their effect, if any, on the schedule time frame. The schedule should be analyzed continuously for variances to determine when forecasted completion dates differ from planned dates. This analysis is especially important for those variations that affect activities identified as being in a program’s critical path and that can affect a scheduled completion date. A baseline schedule should be used to manage the program scope, the time period for accomplishing it, and the required resources. We found that DEAMS program management met weekly to discuss proposed schedule changes and updated the schedule’s progress. However, a schedule narrative was not prepared by DEAMS program management. In addition, we found a number of date anomalies throughout the schedule, including activities with planned start dates scheduled to occur in the past and activities with actual finish dates scheduled to occur in the future. We also found a number of out-of- sequence activities in the schedule—activities that started before their predecessors finished, in contradiction to the planned sequence. If the schedule is not continually monitored to determine when forecasted completion dates differ from planned dates, then it cannot be used to determine whether schedule variances will affect work needed to be accomplished at a future date. We also found that there was no baseline schedule that could be used to measure program performance. DEAMS program management officials did maintain a schedule narrative document that contained a list of custom fields and assumptions; however, the document did not explain ground rules and assumptions, justifications for logic, and other unique features of the schedule. These officials stated that other process documents were being developed. Without a formally established baseline schedule to measure performance against, management cannot identify or mitigate the effect of unfavorable performance. Our assessment of the updated schedule dated October 2012 found that it was not comprehensive, well-constructed and credible. Although the DEAMS Program Manager stated that the government and contractor activities for Releases 1 through 3 of Increment 1 had been integrated in the October 2012 schedule, this schedule was not comprehensive. Specifically, it excluded activities for both the government and contractor related to Releases 4 through 6 of Increment 1 and Releases 1 and 2 of Increment 2. If activities are missing from the schedule, then other best practices will not be met. The schedule was also missing relationships for a significant number of the remaining milestones and activities. In addition, the October 2012 schedule included a significant number of date constraints with little or no justification for their use in the schedule. Similar to the previous schedule, the updated schedule presented unreasonable float throughout and did not include a schedule risk analysis. As a result of these shortcomings, the updated schedule was not reliable. Further, program officials could not rely on this schedule as a baseline to effectively manage and monitor program performance. In May 2013, program management officials provided another updated DEAMS schedule that they stated included some improvements, but they acknowledged that it contained issues that prevented the schedule from meeting best practices. For example, these officials stated that the May 2013 schedule included long-term planning packages for activities related to Releases 4 through 6 of Increment 1 and Releases 1 and 2 of Increment 2, integrated government and contractor activities, and reduced the number of constraints and out-of-sequence activities in the schedule. However, the officials acknowledged that several outstanding issues remained related to, for example, vertical and horizontal integration, missing logic relationships, and the lack of a schedule risk analysis. Although we did not independently assess the May 2013 schedule to determine whether it met the four schedule characteristics, we did confirm that it included long-term planning packages, which are needed to create a complete picture of the program from start to finish and to allow the monitoring of a program’s critical path. The results of our analyses of the schedule that supported the February 2012 Milestone B decision and October 2012 DEAMS schedule reflect similar weaknesses to those we reported in October 2010. Therefore, given the findings of this review, our prior recommendation for improving the DEAMS schedule remains valid. We found that the DEAMS program fully or substantially met the four characteristics of a reliable cost estimate to support the Milestone B decision, as shown in table 2. However, because the cost estimate relies on dates derived from the schedule and we are questioning the reliability of the forecasted program dates, the credibility of the cost estimate can be affected. Appendix II contains our detailed analysis of the DEAMS cost estimate. A reliable cost estimate is critical to the success of any program. Such an estimate provides the basis for informed investment decision making, realistic budget formulation and program resourcing, meaningful progress measurement, proactive course correction when warranted, and accountability for results. Comprehensive. A cost estimate should include costs of the program over its full life cycle, provide a level of detail appropriate to ensure that cost elements are neither omitted nor double-counted, and document all cost- influencing ground rules and assumptions. The cost estimate should also completely define the program and be technically reasonable. We found that the cost estimate for DEAMS was comprehensive. The cost estimate included both government and contractor costs of the program over its life cycle—from the inception of the program through design, development, deployment, and operation and maintenance—as outlined in the roadmap prepared by program officials. As stated earlier, the roadmap provided an overall summary of the program’s key phases (increments and releases) and the expected milestones for completion. The cost estimate also included an appropriate level of detail, which provided assurance that cost elements were neither omitted nor duplicated, and included documentation of all cost-influencing ground rules and assumptions. The cost estimate documentation included the purpose of the cost estimate, a technical description of the program, and technical risks (e.g., the resolution for any identified deficiencies). Well-documented. A cost estimate should be supported by detailed documentation that describes how it was derived and how the expected funding will be spent in order to achieve a given objective. Therefore, the documentation should capture such things as the source data used, the calculations performed, the results of the calculations, the estimating methodology used to derive each work breakdown structure element’s cost, and evidence that the estimate was approved by management. We found that the cost estimate for DEAMS was well-documented. The cost estimate captured such things as the source data used, the calculations performed and the results of the calculations, and the rationale for choosing a particular estimating methodology. This information was captured in such a way that the data used to derive the estimate can be traced back to, and verified against, the sources so that the estimate can be easily replicated. However, there was no discussion of efforts taken, if any, to ensure the reliability of the data used. The DEAMS Program Management Office presented evidence of receiving approval of the estimate through briefings to management. Accurate. A cost estimate should be based on an assessment of most likely costs (adjusted properly for inflation), updated to reflect significant changes and grounded in a historical record of cost estimating and actual experiences on other comparable programs. We found that the cost estimate for DEAMS was accurate. The cost estimate provided results that were substantially unbiased, and the cost model detailed the calculations and inflation indexes underlying the estimate. Calculations within the model could be traced back to supporting documentation. The cost estimate was updated regularly to reflect significant changes in the program and updated annually to incorporate actual costs expended in prior fiscal years. Further, the cost estimate was based on historical data. However, the cost estimate did not discuss variances between planned and actual costs, which would enable estimators to assess how well they are estimating program costs and to identify lessons learned. Credible. A cost estimate should discuss any limitations of the analysis because of uncertainty or biases surrounding data or assumptions. In addition, the estimate’s results should be cross-checked and reconciled to an independent cost estimate to determine whether other estimating methods produce similar results. We found that the cost estimate was credible. The DEAMS Program Management Office conducted a risk and uncertainty analysis by identifying the cost elements with the greatest degree of uncertainty, determining the cost drivers for the program, and identifying the impact of changing major ground rules and cost driver assumptions. An independent cost estimate developed by the Air Force Cost Analysis Agency was reconciled to the program’s estimate. However, a sensitivity analysis was not completed for each of the major cost drivers. As a result, the cost estimator will not have a clear understanding of how each major cost driver is affected by a change in a single assumption and thus which scenario most affects the cost estimate. Further, as discussed previously, because a schedule risk analysis was not performed as required by best practices, the cost estimate does not include a contingency amount to account for any schedule slippage that could occur. To the extent that a schedule slippage does occur, there could ultimately be an impact on the cost estimate. The Air Force did not meet best practices in developing a schedule for the DEAMS program. As a result, this raises questions about the credibility of the deadline for acquiring and implementing DEAMS to provide needed functionality for financial improvement and audit readiness. Because of these questions, the cost estimate, while following best practices, may not fully capture all costs associated with the program, particularly if there is significant schedule slippage. Moreover, Air Force management did not have a reliable schedule estimate when making its decision to invest in the DEAMS program. It is critical to correct the deficiencies identified with the schedule estimate to help ensure that the projected spending for this program is being used in the most efficient and effective manner. To help provide for the successful implementation of DEAMS, we recommend that the Secretary of the Air Force direct the Under Secretary of the Air Force, in his capacity as the Chief Management Officer, to consider and make any necessary adjustments to the DEAMS cost estimate after addressing our prior recommendation to adopt scheduling best practices. We provided a draft of this report to DOD for review and comment. In its written comments, reprinted in appendix III, DOD concurred with our recommendation. DOD also provided a technical comment, which we incorporated. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees, the Secretary of Defense; the Secretary of the Air Force; the Assistant Secretary of Defense (Acquisition); the Deputy Chief Management Officer; the Under Secretary of Defense (Comptroller); the Under Secretary of the Air Force, in his capacity as the Chief Management Officer of the Air Force; and the Program Manager for DEAMS. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staffs have any questions about this report, please contact Asif A. Khan at (202) 512-9869 or khana@gao.gov or Nabajyoti Barkakati at (202) 512-4499 or barkakatin@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff members who made key contributions to this report are listed in appendix IV. This appendix provides the results of our analysis of the extent to which the Defense Enterprise Accounting and Management System (DEAMS) schedule estimate supporting the February 2012 Milestone B decision met the characteristics of a high-quality, reliable schedule. Table 3 provides the detailed results of our analysis. GAO’s methodology includes five levels of compliance with its best practices. “Not met” means the program provided no evidence that satisfies any of the criterion. “Minimally met” means the program provided evidence that satisfies a small portion of the criterion. “Partially met” means the program provided evidence that satisfies about half of the criterion. “Substantially met” means the program provided evidence that satisfies a large portion of the criterion. “Fully met” means the program provided evidence that completely satisfies the criterion. This appendix provides the results of our analysis of the extent to which the Defense Enterprise Accounting and Management System (DEAMS) cost estimate supporting the February 2012 Milestone B decision met the characteristics of a high-quality cost estimate. Table 4 provides the detailed results of our analysis. GAO’s methodology includes five levels of compliance with its best practices. “Not met” means the program provided no evidence that satisfies any of the criterion. “Minimally met” means the program provided evidence that satisfies a small portion of the criterion. “Partially met” means the program provided evidence that satisfies about half of the criterion. “Substantially met” means the program provided evidence that satisfies a large portion of the criterion. “Fully met” means the program provided evidence that completely satisfies the criterion. In addition to the contacts named above, Cynthia Jackson (Director), Karen Richey (Assistant Director), Beatrice Alff, Jennifer Echard, Patrick Frey, and Jason Lee made key contributions to this report.

The Department of Defense (DOD) has stated that the development and implementation of DEAMS is critical to the department's goal of producing auditable financial statements by September 2017. In October 2010, GAO reported that although the Air Force had developed a cost estimate that met best practices, it had not developed a schedule that met best practices for implementing DEAMS. GAO has published guides that identify the characteristics and associated best practices for developing reliable schedule and cost estimates. GAO was asked to review the schedule and cost estimates for selected DOD systems. This report addresses the extent to which the current schedule and cost estimates for DEAMS were prepared in accordance with GAO's Schedule and Cost Guides. Specifically, GAO's review focused on the schedule and cost estimates that supported DOD's February 2012 Milestone B decision, which determined that investment in DEAMS was justified. GAO assessed the schedule and cost estimates and supporting documentation. GAO also assessed an updated schedule dated October 2012. GAO interviewed DEAMS program officials, lead schedulers, and cost estimators. The Air Force's schedule that supported the February 2012 Milestone B decision for the Defense Enterprise Accounting and Management System (DEAMS) did not meet best practices. The cost estimate did meet best practices, but the issues associated with the schedule could negatively affect the cost estimate. GAO found that the schedule supporting the Air Force's decision to invest in DEAMS partially or minimally met the four characteristics for developing a high-quality and reliable schedule. For example, the schedule did not reflect all government and contractor activities, and resources were not assigned to specific activities. The schedule also lacked a valid critical path, preventing management from focusing on the activities most likely to cause critical program delays if they are not completed as planned. In addition, a schedule risk analysis was not conducted to predict a level of confidence in meeting the program's completion date. GAO found that the October 2012 updated schedule estimate was not comprehensive, well-constructed, and credible, and contained weaknesses similar to those found in the previous schedule. In May 2013, program officials provided a third schedule that they said contained some improvements but acknowledged that issues remained that prevented the schedule from meeting best practices. GAO found that the DEAMS cost estimate fully or substantially met the four characteristics of a high-quality and reliable cost estimate. For example, the cost estimate included both government and contractor costs for the program over its life cycle and provided for an independent assessment and reconciliation. Because the cost estimate relies on dates derived from the schedule and GAO is questioning the reliability of the schedule, the credibility of the cost estimate could be affected. GAO recommends that the Secretary of the Air Force update the cost estimate as necessary after implementing GAO's prior recommendation to adopt scheduling best practices. DOD concurred with the recommendation.

UAVs are pilotless aircraft, controlled remotely or by preprogrammed on-board equipment. The Outrider system consists of four air vehicles, ground control equipment, one remote video terminal, four modular mission payloads, communications devices, a means of launch and recovery, and one mobile maintenance facility for every three Outrider systems (see fig. 1). The Outrider ACTD grew out of the Joint Tactical UAV program. The original concept of the Joint Tactical UAV program was to acquire (1) a 50-kilometer UAV system, the Maneuver, to satisfy reconnaissance and surveillance needs of Army brigade and Marine Corps regimental commanders and (2) a 200-kilometer UAV system, the Hunter, to satisfy the reconnaissance and surveillance needs of Army corps and division commanders and Navy task force commanders. The Joint Tactical UAV program was restructured in fiscal year 1996. The Hunter portion was canceled and the Maneuver portion was reconstituted as the Outrider ACTD to evaluate one UAV system’s ability to perform both the Hunter and Maneuver missions. To streamline the acquisition process, DOD designated Outrider an ACTD in December 1995 and awarded a contract for a 2-year ACTD in May 1996. During this period, DOD will acquire 6 nondevelopmental Outrider systems with 24 air vehicles at a cost of approximately $57 million. DOD can procure more systems during the ACTD using low-rate production options built into the contract and, according to an Outrider program official, has requested $30 million for fiscal year 1998 to do so. According to DOD, the purpose of the Outrider ACTD is to evaluate the utility of the system through a series of operational demonstrations. The Army, the Navy, and the Marine Corps will prepare assessments of the system’s military utility based on the operational demonstrations. At the end of the ACTD, Defense Acquisition Board executives will review the service assessments and determine if the ACTD should become a formal acquisition program. If DOD approves transition to the formal acquisition process, program officials must prepare documentation identical to that required of traditional acquisition programs. Prior to beginning the Outrider ACTD, DOD acquired three other nondevelopmental tactical UAV systems: Pioneer, Hunter, and Predator. Each of these UAV programs provided DOD with important lessons about acquisition strategies, system integration, and logistic supportability. However, DOD is not applying these lessons to the Outrider ACTD. DOD’s acquisition strategy for the Outrider closely resembles the acquisition strategy used for the Hunter program. After a user demonstration, DOD awarded a low-rate production contract for 7 Hunter systems with 56 aircraft before demonstrating through operational testing that the system was potentially operationally effective and suitable.Testing of the low-rate production Hunter systems revealed numerous problems, and eventually DOD terminated the Hunter program. Similarly, according to an Outrider program official, DOD plans to exercise a contract option for low-rate production of three to six additional Outrider systems in April 1998 before conducting realistic operational testing. The program official stated that user demonstrations conducted prior to April 1998 as part of the ACTD will provide a sufficient basis for making a low-rate production decision. These user demonstrations, however, will not provide the same level of assurance for justifying a low-rate production commitment as would operational testing since such testing involves meeting minimally acceptable thresholds for key performance parameters. Outrider as an ACTD system has neither key parameters nor thresholds, and DOD is not required to establish them for the demonstrations. Lessons learned from prior UAV programs illustrate that nondevelopmental UAV systems should be operationally tested in realistic environments before beginning low-rate production. Our past work has shown that production of nondevelopmental UAV systems before operational testing can result in adverse consequences. DOD started producing two nondevelopmental UAVs—the Pioneer and, more recently, the Hunter—before subjecting either to any operational testing. The problems DOD has experienced with these systems clearly illustrate the adverse consequences of beginning production without having adequate assurance of satisfactory system performance. Specifically, in 1990, we reported that lack of Pioneer operational testing led the Navy to costly and time-consuming trial and error while trying to adapt the system for shipboard use. Ultimately, DOD spent about $50 million redesigning and modifying Pioneer systems initially acquired for $56 million. Undeterred by the experience with Pioneer, DOD then started production of the Hunter without subjecting it to operational testing. In 1992, we reported that DOD should not award a production contract for the Hunter based on limited testing in unrealistic environments. Nevertheless, DOD awarded a contract for seven Hunter systems. These systems were unable to meet requirements, and the program was terminated in 1995 after an investment of over $757 million. Integrating nondevelopmental components into a fieldable Outrider system is proving more challenging than DOD anticipated. According to program officials, integrating components necessary to satisfy the naval requirements, such as electromagnetic interference shielding and stronger landing gear, delayed Outrider’s first flight from November 1996 to March 1997. Because the Outrider ACTD has a 2-year time limit, schedule delays result in less time available for the users to assess the system’s military utility. These nondevelopmental UAV integration lessons are not new to DOD. The Hunter and Pioneer were both procured by DOD as nondevelopmental systems. Both systems required the expenditure of unexpected development time and money in retroactive attempts to solve integration problems. For example, we stated in our September 28, 1990, report, that the Pioneer system required substantial development to integrate the system into a shipboard environment. In addition, in 1995, DOD concurred with us that the complexity of the Hunter subsystem integration was significantly underestimated by both the government and the contractor. An independent DOD team that reviewed the Hunter UAV in 1995 reported that using nondevelopmental subsystems misled many into believing that integrating nondevelopmental subsystems would not require substantial development. The team recommended that the services should consider and reevaluate the advantage of attempting to procure nondevelopmental subsystems without allowing for some developmental effort needed to integrate them into the overall system. DOD plans to award a low-rate production contract for up to six Outrider systems without demonstrating a critical component of military utility—whether the system is user-supportable. The ACTD’s operational demonstrations will not realistically address the user-supportability of the Outrider system. According to an Outrider program official, the user will perform only basic maintenance during the operational demonstrations, while the contractor will perform all other maintenance. Furthermore, the Outrider ACTD will not include a logistics demonstration to show that the system is user-supportable without contractor assistance. UAV lessons learned show that procuring nondevelopmental systems without assurance that they are user-supportable results in cost growth and program delays. For example, a logistics demonstration conducted after DOD procured seven low-rate production Hunter systems revealed the system was not user sustainable. DOD analysts reported that the perception in the Hunter program was that logistics would be easy to add to a nondevelopmental system. In reality, adding military logistics to a nondevelopmental system proved a significant challenge. The analysts noted that an expensive, time-consuming developmental effort was needed to acquire the logistics support for Hunter. In addition, while ACTD unit cost may be low, militarizing capabilities and adding logistics support increases program costs. For example, while a Predator ACTD system cost about $15 million, a Predator combat-ready production system, with configuration changes, added subsystems, and full integrated logistics support provisions, costs about twice that amount. The Outrider system may not satisfy user needs unless problems associated with meeting joint requirements are resolved and interoperability with other DOD systems can be achieved. Design changes necessary to increase Outrider’s range to 200 kilometers have delayed the program and have increased the weight of the air vehicle to the point it may not be suitable for shipboard operations. Furthermore, developing an air vehicle engine suitable for naval use has proven problematic. In addition, the Outrider analog datalink is not compliant with DOD’s communications interoperability standards for reconnaissance assets and provides limited payload growth options. The Outrider system is encountering technical problems that must be resolved before the system can meet user needs. First flight of the Outrider system was delayed 4 months because of these problems. According to program officials, these problems arose from modifying the Outrider to satisfy joint requirements. The Outrider system was originally designed to satisfy the 50 kilometer, land-based, Army maneuver UAV requirement. Under the ACTD, Outrider’s joint range requirement is 200 kilometers and includes operation from amphibious ships. Modifications to satisfy joint requirements have necessitated several changes to the air vehicle design. These changes, such as adding electromagnetic interference shielding for shipboard operations and increasing air vehicle size to satisfy the range requirement, have added a large amount of weight to the air vehicle. Since DOD awarded the ACTD contract in May 1996, the weight of the fueled air vehicle has grown from the proposed 385 pounds to an actual of 578 pounds. The added weight increases the distance necessary to launch and recover the air vehicle. According to an Outrider oversight official, this could necessitate the use of arresting cables or barrier nets on the deck of a ship. According to Navy officials, the Navy is reluctant to use cables or nets to recover the Outrider because of the impact on other shipboard flight operations. The Navy has previously expressed concerns about the adverse impact of arresting cables and barrier nets on the normal flight operations of amphibious assault ships. In December 1995, we reported that Navy fleet officials opposed fielding the Hunter UAV on Navy ships because erecting barrier nets would adversely impact other flight operations from their amphibious assault ships. Additionally, Outrider’s joint requirements include a heavy fuel engine. Naval use requires a heavy fuel engine because the automotive gasoline currently used by the Outrider is considered too combustible for safe use on ships. DOD research officials estimate it may ultimately cost $100 million to develop a heavy fuel engine that is small enough to power the Outrider. Without a heavy fuel engine, the system will not satisfy naval users. A senior program official acknowledged the heavy fuel engine development is not proceeding as successfully as planned, and the current gasoline engine is not performing adequately. Consequently, 1 year into the ACTD, DOD now plans to acquire another gasoline engine. DOD is not capitalizing on opportunities to demonstrate that Outrider will be interoperable with other DOD systems during the ACTD period. DOD will not be demonstrating the Outrider with the Army and the Navy’s standardized computer workstations or with the software being designed to control all tactical UAVs, including the Predator UAV system, which is already in production. Nor will DOD be demonstrating the Outrider with a DOD-compliant Common Data Link (CDL) that would allow information from the Outrider to be more easily transferred to other DOD systems. DOD is developing a tactical control system that will control all tactical UAVs. The current Outrider and Predator control systems are incompatible and do not meet standards for communications compatibility with DOD’s other airborne reconnaissance systems. Although the Outrider will be required to work with the tactical control system, according to an Outrider program official, DOD will attempt to demonstrate interoperability on only one occasion during the ACTD. A potentially serious interoperability issue may arise if the Outrider development schedule is not aligned with the tactical control system program schedule. The tactical control system is primarily software designed to perform common mission planning and control for all tactical UAVs, including the Outrider, and it will be installed on computers already used by the services, such as the Navy’s TAC-4 and the Army’s Sunspark Systems. However, during the ACTD, DOD is allowing the Outrider contractor the option of using either (1) Outrider-specific hardware and software that is supposed to be interoperable with the tactical control system or (2) the tactical control system. According to the Outrider Demonstration Manager, the contractor has opted to use the Outrider-specific equipment, and only one demonstration of interoperability between the Outrider equipment and the tactical control system is planned for the ACTD. If the actual tactical control system and service computers are not used during the ACTD, the services’ overall assessments of military utility will not be based on actual system performance. DOD acknowledges the risk their plan creates of not achieving the required interoperability between the Outrider and the tactical control system. The Outrider datalink is not compliant with the CDL, DOD’s standard for communications interoperability for all airborne reconnaissance and surveillance missions, including those missions performed by the Outrider. The CDL requires a digital data link, whereas the Outrider employs an analog data link. According to officials from the Defense Airborne Reconnaissance Office, which is responsible for airborne reconnaissance and intelligence communications interoperability, the analog data link has no growth options and operates in the same widely used band of the microwave spectrum as European and Korean television. These officials noted that a CDL-compliant digital data link would offer the Outrider program several advantages over the current analog link. For example, a digital data link would (1) be less susceptible to distortion and interference, (2) minimize a system’s signature, (3) provide anti-jam capabilities, and (4) offer encrypted communications. The digital data link also provides for greater capability, including (1) a means to upgrade to all-weather payloads, such as the synthetic aperture and millimeter wave radars and (2) computer processing of gathered imagery. A Defense Airborne Reconnaissance Office study indicates that a short development effort could result in a CDL-compliant digital data link for the Outrider at an acceptable cost. However, Outrider officials maintain that a CDL-compliant digital data link would be too expensive given Outrider’s post-ACTD cost limit of $350,000 for the 33rd air vehicle and sensor. Because DOD’s strategy for acquiring the nondevelopmental Outrider system will not provide assurance of successful performance and interoperability before DOD’s planned low-rate production decision, and to avoid repeating the mistakes of prior UAV programs, we recommend that the Secretary of Defense delay low-rate production of the Outrider system until the results of operational testing of available systems demonstrate it is potentially operationally effective and operationally suitable for all intended users. DOD reviewed a draft of this report. DOD disagreed with most of our findings. It partially concurred with our recommendation. Specifically, DOD disagreed that it had not learned from problems in past programs and stated these problems in part led it to initiate the Outrider ACTD. DOD also disagreed that Outrider may not satisfy user needs unless it meets the Navy’s shipboard requirements and is interoperable with the tactical control system. It stated that the ACTD responds to an approved joint requirement and does not identify service unique requirements, but will address the effect of weight and engine type. DOD also noted that it has formed an integrated team between the Outrider and tactical control system programs and taken other measures to ensure interoperability. We recognize that DOD is aware of problems with past UAV programs and agree that an ACTD can provide useful insights. However, we remain concerned about DOD’s strategy for the Outrider because the planned demonstrations of military utility that will precede DOD’s low-rate production decision are (1) limited in scope; (2) will not be complete before the decision; and (3) may not identify and resolve serious system deficiencies, such as compatibility with joint requirements, and interoperability with the tactical control system. As detailed in this report, similar acquisition strategies for the Hunter and Pioneer programs resulted in the acquisition of additional systems that required costly modifications in order to meet user needs. DOD has the opportunity to operationally test the Outrider’s performance without risking commitment to additional unproven systems under low-rate production. DOD is acquiring 6 Outrider systems with 24 aircraft under the original contract. If the Outrider is assessed positively during the ACTD, DOD could modify the ACTD hardware to the production representative design for operational tests. If the required changes are so significant that the ACTD systems cannot be made production representative, DOD guidance on transitioning ACTDs to formal acquisition indicates that a new competition should be conducted. In responding to our recommendation, DOD concurred that Outrider should not enter production until the results of operational testing demonstrate its effectiveness and suitability. DOD noted that completing operational test and evaluation is a statutory requirement for formal acquisition programs entering production. DOD added, however, that this statute does not apply to ACTDs entering low-rate production. We recognize that full operational testing is not a statutory requirement for ACTDs entering low-rate production. However, our past work shows that awarding low-rate initial production contracts without any operational testing has resulted in the procurement of substantial inventories of unsatisfactory weapons requiring costly modifications to achieve satisfactory performance and, in some cases, deployment of substandard systems to combat forces. To determine whether DOD is applying lessons learned from prior UAV lessons learned to this program, and whether the Outrider would meet user needs, we reviewed program plans, test schedules, performance documents, and other records relating to the Outrider ACTD and examined DOD guidance related to systems acquisition, acquisition streamlining and reform, and ACTDs. We also interviewed and obtained information from knowledgeable officials of the Joint Chiefs of Staff; the Office of the Secretary of Defense; Defense Airborne Reconnaissance Office; Chief of Naval Operations; Department of the Navy, Program Executive Office for Cruise Missiles and UAV Joint Project; Department of the Army, Operational Test and Evaluation Command; and the Department of the Air Force, Deputy Chief of Staff Plans and Operations. All of these officials are located in the greater Washington, D.C., metropolitan area. Furthermore, we interviewed and obtained information from representatives of the Commander in Chief, U.S. Atlantic Fleet, Norfolk, Virginia; the Department of the Navy, Operational Test and Evaluation Forces Command, Norfolk, Virginia; the Joint Tactical UAV Project Office, Huntsville, Alabama; Defense Contract Audit Agency, Hopkins, Minnesota; Defense Contract Management Command, Hopkins, Minnesota; and the Outrider ACTD contractor, Alliant TechSystems, Hopkins, Minnesota. We performed our work from July 1996 to June 1997 in accordance with generally accepted government auditing standards. This report contains a recommendation to you. As you know, 31 U.S.C. 720 requires the head of a federal agency to submit a written statement on actions taken on our recommendations to the Senate Committee on Governmental Affairs and the House Committee on Government Reform and Oversight not later than 60 days after the date of the report. A written statement also must be submitted to the Senate and House Committees on Appropriations with an agency’s first request for appropriations made more than 60 days after the date of the report. We are sending copies of this report to appropriate congressional committees; the Secretaries of the Army and the Navy; and the Office of Management and Budget. We will make copies available to others on request. Please contact me at (202) 512-4841, if you or your staff have any questions concerning this report. Major contributors to this report were Tana Davis, John Warren, and Charles Ward. The following are GAO’s comments to the Department of Defense’s (DOD) letter, dated July 9, 1997. 1. We understand that the purpose of the Outrider Advanced Concept Technology Demonstration (ACTD) is to assess the utility of the Outrider system and note that DOD is acquiring 6 Outrider systems with 24 air vehicles under the original ACTD contract. If the Outrider is assessed positively, these could be used instead of building production representative systems under low-rate production. Specifically, DOD could modify the ACTD systems to create a production representative system that could be operationally tested prior to low-rate production. If required changes are so significant that the ACTD system cannot be successfully modified, DOD ACTD guidance indicates that a new competition should be conducted. 2. We agree that ACTDs should be based on mature technologies. However, DOD officials have acknowledged the Outrider system is not mature. We therefore continue to believe that DOD should resolve the integration challenges for Outrider before proceeding to a low-rate production decision. 3. Although DOD maintains that the development of Outrider is event rather than schedule driven, we note that DOD has not slipped the planned low-rate production decision or ACTD completion date in response to delays to the Outrider test schedule. 4. DOD states that it will demonstrate supportability prior to the full system acquisition. DOD ACTD guidance states that the full range of support areas must be considered if the plan for an ACTD is to transition to low-rate production. We believe that committing to further Outrider production without taking advantage of the opportunity to demonstrate supportability adds unnecessary risk to the planned acquisition program. 5. Our report specifically identifies the differences in the cost of a Predator ACTD system compared with a Predator production system. 6. We modified the text to clarify that the Outrider ACTD is based on joint requirements. 7. ACTD guidance points out that overall systems engineering efforts performed during the ACTD should include actions ensuring connectivity, compatibility, and synchronization of ACTD products with systems these products will operate with on the battlefield. Receipt of secondary imagery from the Outrider ground control station (level 1) does not provide any evidence that the tactical control system will be able to control or receive information directly from the Outrider air vehicle (levels 2 and 3). DOD’s plan to demonstrate Outrider’s compliance with tactical control system’s interoperability standards during the ACTD is not the same as demonstrating that levels 2 and 3 can be achieved in the field. 8. DOD’s response indicates a tactical Common Data Link (CDL) may be available for use in Outrider in less than 2 years. The ACTD is scheduled for completion in May 1998. If Outrider low-rate production were delayed until the CDL became available, DOD could avoid retrofit risks and expenses. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

GAO reviewed the Department of Defense's (DOD) acquisition of the Outrider, an unmanned aerial vehicle (UAV) system, through a streamlined acquisition process known as the Advanced Concept Technology Demonstration (ACTD), focusing on whether: (1) DOD is applying lessons learned from prior UAV programs to the Outrider; and (2) the Outrider is likely to meet user needs. GAO noted that: (1) DOD is not applying lessons learned from prior UAV programs to the Outrider ACTD; (2) for example, despite problems with the Pioneer and Hunter stemming from DOD's decision to award further production contracts without conducting operational testing or demonstrating that the system is user-supportable, DOD is pursuing the same strategy for the Outrider; (3) in addition, DOD has underestimated, as it did for the Pioneer and Hunter programs, the time and effort necessary to integrate nondevelopmental items into Outrider; (4) moreover, the Outrider system may not satisfy user needs unless problems associated with meeting joint requirements are resolved and interoperability with other DOD systems is ensured; and (5) consequently, DOD will not have assurance that the Outrider will meet user needs by the time of the planned fiscal year 1998 low-rate production decision.

As of December 31, 2010, there were 6,364 community banks (commercial banks with total assets of $1 billion or less). This represents about 92 percent of all commercial banks, although only about 10 percent of commercial bank assets nationwide. Banks in the United States are supervised by one of the following three federal regulators: FDIC supervises all FDIC-insured state-chartered banks that are not members of the Federal Reserve System. The Federal Reserve supervises commercial banks that are state-chartered and members of the Federal Reserve System. OCC supervises federally chartered national banks. Table 1 summarizes the regulators’ oversight responsibilities for community banks. The purpose of federal banking supervision is to help ensure that banks throughout the financial system are operating in a safe and sound manner, and are complying with banking laws and regulations in the provision of financial services. As we have identified in previous work, financial regulation more broadly has sought to achieve four goals: to (1) ensure adequate consumer protections, (2) ensure the integrity and fairness of markets, (3) monitor the safety and soundness of institutions, and (4) act to ensure the stability of the overall financial system. Federal banking regulators use a number of tools to achieve these goals. Capital requirements: Regulators require banks to maintain certain minimum capital requirements to help ensure the safety and soundness of the banking system, and generally expect banks to hold capital above these minimums—commensurate with their risks. Capital provides an important cushion against losses for banks, and represents the amount of money that can cover losses the bank may face related to nonpayment of loans and other losses on assets. Capital can be measured as total capital or tier 1 capital. Regulators oversee the capital adequacy of their regulated institutions through ongoing monitoring, including on-site examinations and off-site tools. When regulators require banks to hold capital above regulatory minimums, these requirements may result from an enforcement action through an agreement between the regulator and the bank. However, requiring banks to hold more capital may reduce the availability of bank credit and reduce returns on equity to shareholders. Examinations and ratings: Federal banking laws and regulatory guidance require on-site examinations, which serve to evaluate a bank’s overall risk exposure and its ability to identify and manage those risks— especially as they affect a bank’s financial health. At each full-scope examination, examiners review the bank’s risk exposure on a number of components using what is known as the CAMELS rating system (Capital adequacy, Asset quality, Management, Earnings, Liquidity, and Sensitivity to market risk). Evaluations of CAMELS components consider the institution’s size and sophistication, the nature and complexity of its activities, and its risk profile. In examinations, a bank is rated for each of the CAMELS components and given a composite rating, which generally bears a close relationship to the component ratings. However, the composite is not an average of the component ratings. The component rating and the composite ratings are scored on a scale of 1 (best) to 5 (worst). Regulatory actions typically correspond to the composite CAMELS ratings, with the actions generally increasing in severity as the ratings become worse. Table 2 describes the definitions of the composite scores under the Uniform Financial Institutions Rating System. Loan classifications: In examinations, examiners review a sample of banks’ internal ratings of loans to determine the adequacy of credit risk administration and identify loans that show undue risk and may be uncollectible. As part of this review, examiners determine which loans are considered “pass,” with no concerns noted, as well as those that are special mentioned or “classified”—that is, subject to criticism because they are not performing or may not perform in the future. There are three classification categories used by the federal banking regulators: substandard, doubtful, and loss (see table 3). These loan classifications, and the internal ratings that banks produce for all of their loans, are incorporated into how each bank calculates its allowance for loan and lease loss (ALLL), which is an estimate made according to accounting guidance of incurred losses on loans and leases. Therefore, if additional loans are classified substandard or doubtful, this information is included in a bank’s updated ALLL estimates. If loans are classified loss, they are charged off the bank’s balance sheet. The end result of an on-site examination is an ROE that includes the CAMELS ratings and other findings on the bank’s condition, which is provided to the bank’s management and board of directors. CRE encompasses many different property types that present different risks. Table 4 describes the key CRE property types. Regulators define CRE loans to include construction loans, loans to finance CRE that are not secured by CRE, loans secured by multifamily property, and loans secured by nonfarm, nonresidential property in which the primary source of repayment derives from the rental income associated with the property or the proceeds of the sale, refinancing, or permanent financing of the property. CRE loans in which the primary source of repayment is not the property itself are called owner-occupied loans and can include loans to businesses for working capital purposes that use real estate as collateral. For example, a line of credit for a business’s operating expenses might be secured in part by commercial property, such as an office. Owner-occupied properties generally are considered to carry less risk than non-owner-occupied properties because regulators consider them to be less sensitive to the condition of the CRE market. In ADC loans, also called construction and land development (CLD) loans, generally are considered to be the riskiest class of CRE, due to their long development times and because they can include properties (such as housing developments or retail space in a shopping mall) that are built before having firm commitments from buyers or lessees. In addition, by the time the construction phase is completed, market demand may have fallen, putting downward pressure on sales prices or rents—making this type of loan more volatile. In recent years, this type of loan also has tended to have a much higher loss volatility than loans secured by properties such as multifamily housing and other nonfarm, nonresidential commercial properties. Banks report on four broad categories of CRE in their quarterly call reports: CLD: loans secured by real estate to finance land development and construction. This includes new construction, as well as additions and alterations on existing properties. Multifamily: loans for residential properties with five or more dwelling units, such as apartment buildings. Nonfarm nonresidential: loans secured by real estate for business and industrial properties, as well as properties such as hotels, churches, hospitals, schools, and charitable organizations. This category includes offices, retail, and warehouse space. Loans to finance CRE, construction, and land development (not secured by CRE). Interagency guidance issued in 2006 on concentrations in CRE and sound risk-management practices define CRE loans within these categories to include those in which repayment is dependent on the cash flow generated from the real estate itself. When evaluating concentrations in CRE, examiners are instructed not to include owner-occupied properties in which the income or value of the property is not the primary source of repayment. CRE as an asset class—and especially properties in the ADC category—is prone to volatility and cyclical behavior, as illustrated in the current CRE market downturn. This volatility and cyclical behavior is attributed to characteristics such as information-gathering difficulties, infrequent transactions, high transaction costs, rigid and constrained supply, long construction times, and a two-fold reliance on external finance (shorter- term to cover construction, and longer-term for the occupancy period). Additionally, CRE is by nature diverse and localized. For example, shopping centers are one type of CRE, but even within this category properties can have significant differences depending on design, types of tenants, and other factors that can affect their value. Because of these factors, the supply of CRE in the marketplace is slow to respond to an increase in demand, which drives prices up when investor optimism rises. Conversely, the marketplace is slow to respond when the market supply of CRE catches up and new construction projects are delivered, resulting in oversupply and declining property values. According to regulatory officials, the weakness in the ADC sector during this crisis was primarily due to residential housing construction, and that weakness affected the performance of other areas of CRE (for example, failed residential housing developments will affect the ability of nearby strip malls to attract and generate rental income from tenants). The recent downturn in the CRE markets can be seen in the market values for commercial property and the condition of the commercial mortgage- backed securities (CMBS) market. Market values for all major commercial property types have declined significantly. As of December 2010, overall CRE market values were down more than 42 percent from their peak in 2007. This decline followed a rapid appreciation in CRE asset values during which CRE values increased by more than 85 percent from 2002 to the market’s peak in October 2007 (see fig. 1). Overall deterioration in CRE markets can be found in the condition of the CMBS market as well. By January 2011, the delinquency rate on loans included in CMBS was at a record high, above 9 percent. As we have reported, overall CMBS issuance slowed severely since the CRE downturn started. After peaking in 2007, the CMBS market came to a complete halt by the end of 2008. Although CRE market price deterioration appears to have leveled off recently, vacancies at many properties remain high and signs of a market recovery have been uneven in different areas of the country. High-value properties in markets such as New York, Washington, San Francisco, and Boston have performed well more recently. But low rental rates and high vacancies indicate that demand for office, retail, multifamily housing, and warehouse space remains relatively weak. Furthermore, although CRE markets nationally have experienced a downturn, some regions have experienced more distress than others. For example, the CRE markets in the South, Midwest, and West have experienced greater stress and deterioration than the East. These areas that have experienced greater CRE market stress also have experienced more bank failures, according to FDIC data. Community banks increasingly have moved toward providing CRE loans more than other kinds of loan products, in part because of competitive pressures. During the last decade, large banks and other financial institutions increased their market share for consumer loans, credit cards, and residential mortgages. As a result, community banks shifted their focus to CRE lending. Some market observers argue that community banks’ focus on CRE lending, and, therefore, the long-term trend of increased CRE concentrations, came about because community banks generally know their local CRE markets better than larger banks and are well-positioned to gather location-specific information for CRE properties. The increased exposure of community banks to CRE loans has been pronounced over the last 10–15 years. While CRE collateral backed about 30 percent of total loans and leases at community banks in 2000, a decade later that rate increased to more than 43 percent. Concentrations have been less pronounced at larger banks, which tend to rely less heavily on CRE lending. Since 2000, CRE as a percent of total loans and leases has ranged from about 15 percent to about 21 percent at commercial banks with more than $1 billion in total assets. Community banks also have come to hold large concentrations of CRE loans in comparison to their total capital. The average CRE concentration at community banks as a percentage of total risk-based capital increased from about 168 percent in 1996–2000 to about 289 percent in 2005–2010 (see fig. 2). Increased exposure to CRE has made community banks vulnerable to the decline of this market. According to FDIC, nearly 30 percent of community banks have concentrations of CRE to total capital above 300 percent, the concentration threshold established in the 2006 interagency guidance on CRE, above which regulators review banks’ risk-management controls more closely. This means that nearly a third of community banks are exposing three times their total capital to risks related to their CRE loans. As noted by a Bank for International Settlements report on bank lending and commercial property cycles, declining property prices increase the proportion of nonperforming loans, lead to a deterioration in banks’ balance sheets, and weaken banks’ capital bases. In particular, the decline in CRE values has contributed to more noncurrent CRE loans, charge-offs, and bank failures. Noncurrent CRE loans have increased. From the first quarter of 2008 through the fourth quarter of 2010, the percent of CRE loans at community banks that were noncurrent increased from 2.2 to 5.3 percent, well above the average rate of 0.9 percent from 2000 through 2007. The average from 2000 through 2010 is 1.94 percent. However, the data show that the volume of noncurrent CRE loans has begun to level off (see fig. 3). CRE loan charge-offs increased. From the end of 2007 through the end of 2010, the percent of CRE loans that had to be charged off at community banks rose from 0.19 to 1.34 percent. From 2000 through 2007, the average charge off rate of CRE loans at community banks was 0.09 percent, and from 2000 through 2010 it was 0.39 percent (see fig. 4). Charge-offs and expected losses that may arise from increases in noncurrent CRE loans have put stress on banks’ capital and ALLL. Increased bank failures linked to high CRE and ADC concentrations. Many bank failures are associated with high CRE and ADC concentrations. In 2009 and 2010, 102 of 106 of the MLRs issued by the IGs of the banking regulators cited high CRE concentrations, and 92 of 106 specifically cited ADC loans in particular as a contributing factor in bank failures. In the 106 MLRs, 76 of the banks reviewed were community banks. For example, the MLR for one failed bank—which contains findings similar to many other MLRs we reviewed—states that the rapid growth of its CRE portfolio “increased the institution’s exposure to a sustained downturn in the real estate market and reduced its ability to absorb losses due to unforeseen events.” CRE market value declines could reduce overall small business lending by community banks. As we previously reported, community banks tend to have a larger portion of small-business related loans compared with larger banks. Because CRE can be used as collateral for small business loans, diminished CRE values also can negatively affect credit availability to small businesses. Specifically, when the value of collateral declines, the amount of financing a bank is willing to lend against that collateral typically declines as well. Conversely, increases in collateral values lower the premium on external financing—improving credit availability for borrowers, boosting demand for real estate assets, and driving up prices. Therefore, falling property prices can generate a cycle of declining CRE values because they can accompany reduced credit. A report by the Bank for International Settlements notes that falling CRE prices decrease the value of collateral held by banks and, therefore, can give rise to significant losses by these banks and ultimately contract the supply of credit. The problem of CRE loan refinancing may exacerbate the negative CRE trends and limit lending to small businesses. CRE loans usually are written for 3–10 years, with a 20–30 year amortization schedule and a balloon payment at the end. Instead of making the large balloon payment, the borrower typically will sell the property or refinance the loan at the end of the term. Small businesses that are looking to refinance a loan against a property that has lost a significant amount of market value may have to put up more equity. Alternatively, such borrowers might default on the loan, or the bank could work with the borrower to restructure the loan and avoid default. Trepp LLC, a commercial mortgage analysis firm, estimates that about $1.7 trillion in CRE mortgages will mature between 2011 and 2015, with about half of that held at banks. Moreover, Trepp estimates that approximately 60 percent of CRE debt maturing in 2011 is “underwater”—meaning the value of the loan exceeds the value of the underlying collateral. Due to price declines and stricter bank underwriting standards compared to when these loans were originated, those mortgages will be difficult to refinance. Results from the Federal Reserve’s Senior Loan Officer Opinion Survey have suggested that new CRE borrowers have faced tighter credit conditions, due to banks tightening their underwriting standards in response to the downturn. The January 2011 survey found that 10.6 percent of respondents reported easing credit standards over the previous 3 months, compared to 10.5 percent who reported tightening them. However, this is a positive development from past results that showed severe tightening. The trend of tighter credit standards suggests that borrowers who previously were considered creditworthy might not meet banks’ higher standards. The regulators have been aware for some time of the risk-management challenges related to growing CRE concentrations at community banks and have taken steps to address these challenges. The regulators, for example, issued guidance to banks on managing CRE concentration risks, conducted training on CRE treatment, and conducted internal reviews to better ensure examiner compliance with CRE guidance. Even with the training and reviews, a number of bank officials we interviewed stated that regulators have applied guidance rigidly since the financial crisis and have been too harsh in classifying loans and improperly applying the 2006 CRE guidance, among other issues. Regulators have been incorporating lessons from the financial crisis in their supervisory practices, which in part may explain bank officials’ experience of increased stringency in supervision. Based on our review of a nonprobability sample of 55 bank examinations, examiners’ findings were consistent with CRE loan workout guidance, although in some instances examiners did not clearly support requirements for reduced CRE concentrations and did not calculate CRE concentrations according to concentration and risk-management guidance. Additionally, senior regulatory officials and examiners have differing views on the adequacy of the 2006 guidance, which may affect how consistently the guidance is applied. The regulators began to address CRE concerns before the financial crisis. Beginning in the early 2000s, the agencies reviewed CRE concentrations and risk-management systems across banks. For example, an OCC review found potential for improvement in banks’ risk-management processes for CRE concentrations, including the sufficiency of stress testing. The regulators issued draft guidance in January 2006 on CRE concentrations and risk management, based in part on the trends they had observed in CRE concentrations and risks. The draft guidance elicited considerable feedback from bank representatives, many of whom stated it would curtail their CRE lending, impose arbitrary limits on CRE concentrations, and require additional capital without explicitly stating how much would be required. The regulators revised the guidance based on the feedback received. Issued in final form in December 2006, the interagency guidance provides levels of CRE concentrations that will result in additional regulatory attention on risk-management systems: (1) 300 percent of CRE loans to total capital, (2) increases of 50 percent or more in CRE loans during the prior 36 months, and (3) 100 percent of CLD (or ADC) loans to total capital. In determining the concentration ratio, owner-occupied CRE is removed. The guidance also states that the concentration numbers are not limits, but rather indicate when banks will receive closer scrutiny of risk-management systems and capital adequacy. Such thresholds are not intended to be a “safe harbor”—that is, banks with lower concentrations may still receive scrutiny of their CRE loans in examinations—and banks that exhibit other risk factors can receive criticisms on their CRE concentrations. In October 2009, after the start of the financial crisis and the widespread deterioration in CRE loan performance, regulators issued interagency guidance on CRE loan workouts. According to the regulators, the guidance was issued to (1) help ensure consistent CRE loan and workout treatment among the regulators, (2) update and re-assert previous guidance, (3) inform banks of examiner expectations, and (4) ensure that supervisory practices do not inadvertently curtail the availability of credit to sound borrowers. Officials told us that the guidance was not intended as forbearance, but to encourage prudent loan workouts. While the 2009 guidance was similar to that issued in 1991 and 1993, regulatory officials noted that it includes updates for changes in accounting (for example, related to ALLL), information on risk management for CRE loan workouts, and numerous examples to assist banks and examiners in interpreting the guidance. The examples provide scenarios for different CRE loan classification outcomes and whether loans should be considered troubled debt restructurings (TDR), among other issues. Bankers with whom we spoke stated that the examples were helpful in understanding how to apply the guidance. Examiners told us that the examples have helped to resolve differences of opinion with bankers on how to treat certain CRE- related loans. In addition to the 2009 CRE loan workout guidance, the regulators issued statements on lending to creditworthy borrowers and creditworthy small business borrowers to encourage prudent lending among banks and balanced supervision among examiners. Federal banking regulators help ensure consistency among examinations, and appropriate application of guidance, primarily through the ROE review process—but also through training, internal reviews, quality assurance processes, and processes for obtaining input from banks. Such processes reflect federal internal control standards, which provide reasonable assurance that management directives are carried out. The federal government standards for internal control state that managers should have controls in place to compare actual performance to planned or expected results over time throughout the organization and analyze significant differences. All of the regulators’ processes were used in some form to implement CRE guidance and ensure consistent treatment of CRE loans. Some of the internal review processes have identified inconsistencies in the application of CRE guidance. The regulators’ examination drafting and review process is iterative, with multiple levels of internal review. According to regulatory staff, this design helps ensure consistent application of guidance and treatment of banks. Figure 5 illustrates the examination process of the federal banking regulators. The process and the staff involved vary slightly for each regulator and based on a bank’s CAMELS rating. For all examinations, a team of examiners will conduct on-site work, led by an examiner-in-charge (EIC), who drafts the ROE. Other staff and management also discuss findings with the examination team and review the examination report. The case manager in particular helps ensure consistency. Case managers review draft ROEs for consistency and adequate support of findings for a portfolio of banks. The analyst function at OCC serves a similar role: reviewing many ROEs, although not assigned to a specific portfolio of banks. While case managers and senior management generally do not review loan classification details, they review loan classification writeups in ROEs for accuracy and support for CAMELS ratings in a broad range of ROEs. According to Federal Reserve officials, case managers and senior management hold discussions with the examination teams as examination findings are being finalized, to ensure that they are appropriate. Therefore, these roles help ensure consistent application of guidance among various examiners. In certain instances, the regulators’ headquarters offices in Washington, D.C., may participate to help ensure consistent implementation of policy guidance. Regulators also provide training on new guidance to help ensure consistent application. All the agencies offered multiple training opportunities or conference calls on the 2006 CRE concentration and the 2009 CRE loan workout guidance. For the 2009 CRE loan workout guidance, Federal Reserve and FDIC headquarters staff visited field offices, and all regulators hosted conference calls with examiners to better ensure that the field examiners were implementing it as intended. Some field offices also included these CRE topics in their own training and team meetings. Regulators also have internal reviews and quality assurance processes to help promote consistency, including consistent application of the CRE guidance. Regulators’ internal reviews include comprehensive audits that recur periodically, and real-time and post-ROE quality processes. Some of these reviews identified inconsistencies or other needed improvements related to examiner treatment of CRE loans. For example: For certain FDIC regions, reviews found that examiners could have better documented findings on CRE concentrations or could have provided earlier or harsher criticism of CRE concentrations, and the regional office could have better monitored CRE concentrations among the banks it supervised. According to an FDIC official, the regions already have addressed some of these findings. Certain districts of the Federal Reserve System reviewed implementation of the 2006 CRE concentration guidance and found instances of inconsistency. The reports concluded that the guidance could have clarified expectations for how examiners should review banks’ compliance with CRE-related risk-management practices and noted that a common, mandatory process for reviewing banks’ compliance with the 2006 guidance would have resulted in better documentation and consistency. In particular, these reviews found that examination workpapers sometimes lacked sufficient documentation to determine if the examiner adequately assessed compliance with the guidance, particularly related to CRE portfolio-level stress testing. According to Federal Reserve officials, Washington staff also coordinated separate reviews related to implementation of the 2006 guidance, which resulted in internal clarifications of the appraisal review process, ALLL assessment guidance, examiner training sessions on the use of interest reserves and TDRs, and contributions to the 2009 loan workout guidance. According to a San Francisco Federal Reserve Bank official, the San Francisco Federal Reserve Bank implemented a “look back” process in June 2010 in which a senior official reviews the sufficiency of support for downgraded CRE loans for all issued ROEs—in response to criticism that examiners were being too harsh and curtailing credit. Of 11 CRE loan downgrades reviewed in the third and fourth quarter of 2010, this official stated that 2 should not have been downgraded. Specifically, according to this official, the loans initially were considered collateral dependent, but after closer review the official determined that the borrowers had some capacity to repay, so the loans should have been classified but not yet charged off. In these two cases, the loan downgrades were reviewed and changed before the ROEs were finalized and were consistent with the bank’s internal loan ratings. According to this official, the findings from this quality process will be used to improve the accuracy of CRE loan classification. Similarly, all OCC districts instituted a real-time review process in which experienced staff reviewed the accuracy of CRE loan classifications— before the ROE is finalized. For example, OCC’s central district in January 2009 reported that 95 percent of examiners’ ratings decisions on CRE loans, and 99 percent of their accrual decisions, were determined to be accurate. Among those classifications that were corrected, 3 percent were downgraded and 2 percent were upgraded. Based on these findings, the district concluded that the vast majority of examiner loan classifications were accurate but recommended more training to address the discrepancies. The western district’s quality assurance process concluded that 4 percent of risk ratings were incorrect. Regulators also use surveys to gather information on how to improve the examination process and understand the impact of policies on the banking industry. For example, the regulators issued an interagency survey for examinations conducted between May 31 and July 9, 2010. According to an FDIC analysis of the resulting data, more than 97 percent of respondents stated that the 2009 CRE loan workout guidance has been helpful. In addition, nearly 88 percent of banks stated that no specific guidance was inhibiting them from working with troubled borrowers. Among the remaining 12 percent, just over three-quarters of them raised concerns about accounting-related guidance and reporting of TDRs. Based on this information, regulatory officials from FDIC, the Federal Reserve, and OCC believe that the 2009 CRE loan workout guidance has been helpful. Comments provided to regulatory officials during the examination process and through the formal appeals process provide information about banks’ concerns on ratings and whether regulatory guidance has been applied consistently. According to bank and regulatory officials with whom we spoke, when bank officials have raised concerns, they tend to first approach the examination team. Bank officials also can contact the ombudsman offices at the regulators to seek confidential assistance. In addition to these avenues, banks formally can appeal regulatory decisions. Although bankers have multiple options for raising concerns about regulatory actions, our interviews with bank officials found that some were cautious about raising issues because of potential repercussions from regulatory officials. In contrast, other bank officials stated that they do contact regulatory officials when warranted, although a few raised concerns about the time and expense of pursuing formal appeals. Regulatory officials told us that many bank officials regularly reach out to them, and a few officials also noted that they provide information to bank officials about the ombudsman and the appeals process—should banks want to raise concerns. Interviews with officials from 43 banks in different parts of the country identified multiple concerns with examiner treatment of CRE loans and related issues. Many bank officials’ overarching concern was that examiners have been applying guidance more stringently than before the financial crisis—a shift that a few bankers commented was a difficult adjustment. Some bank officials we interviewed expressed concerns with examiner- required loan classifications. From the perspective of a few bank officials we interviewed, a loan is performing when the borrower continues to pay and should not be classified. However, according to regulatory guidance and statements of regulatory officials, such a loan may be classified if identified weaknesses suggest a reduction in the borrower’s capacity to repay that in turn could lead to future nonpayment. According to bankers and examiners with whom we spoke, “global cash flow” analysis always has been part of reviewing loans, but in the current economic downturn examiners have been focusing more closely on analysis and documentation of borrowers’ and guarantors’ global cash flows. The purpose of such analysis is to determine whether they have sufficient income and liquid assets to support loan payments. In some cases, such analysis shows that a borrower is unable to pay the loan, even if the borrower is currently paying. For example, the borrower’s income and other debt obligations, when reviewed as a whole, could show that the borrower’s debt obligations exceed income, which raises questions about whether the borrower will continue paying on the loan in the future. The renewed focus on global cash flow analysis, in part, led many bank officials to conclude that examiners were classifying loans based either on collateral value or because the bank lacked updated financial statements. Regulatory field staff acknowledged that global cash flow analysis can be difficult to conduct because borrowers sometimes do not provide banks with updated financial statements. However, regulatory officials told us that examiners do not classify loans solely because of a lack of financial statements, but a lack of such statements combined with other weaknesses could provide sufficient support for classifying a loan. If the borrower lacks the ability to repay, the loan then could be considered “collateral dependent.” Such loans are classified based on the value of the property, and tend to rely on values established in appraisals. According to some bank officials we interviewed, examiners have been more critical of recent appraisals. For example, these officials stated that examiners have been requiring appraisals on CRE collateral more often, criticizing the banks’ appraisal review process, and criticizing the appraisals themselves. Classifications are a major concern for banks primarily because they can result in reduced earnings. For example, an examiner may determine that a bank should place a classified loan on nonaccrual—which means the bank cannot accrue the interest income as earnings. While some bankers put loans on nonaccrual themselves, a few bank officials with whom we spoke stated that examiners have been asking bankers to place more loans on nonaccrual. Classifications also can reduce earnings because they factor into ALLL, which a bank estimates based on accounting guidance from the Financial Accounting Standards Board (FASB). As more loans are classified, more is generally reserved in ALLL to anticipate future losses from nonperforming loans. In our interviews with bank officials, some stated that examiners have been requiring additional ALLL and criticizing the ALLL methodology. In addition, a few bankers noted that some examiners want ALLL increased based on their peers’ ALLL rather than the bank’s individual situation. Many bank officials with whom we spoke were concerned that examiners have been misapplying the 2006 guidance. As we discuss above, the 2006 guidance states that the CRE and ADC concentration levels are not limits, but some bank officials told us that examiners have been interpreting them that way. A few bank officials also stated that examiners have not been calculating CRE concentrations according to the 2006 guidance: some examiners were including owner-occupied CRE and some were using the wrong type of capital for the calculation, which inflated the concentration levels (we also found this in our analysis, described in more detail below). A few bankers stated that examiners told them they exceeded the 300 percent concentration threshold for CRE, based on these faulty calculations, but according to the calculations in the guidance they actually were under the threshold. Furthermore, a few bank officials also stated that it was unclear to them what examiners expect in complying with requirements related to CRE and risk management—for example, on what is considered to be satisfactory stress testing. While more bank officials than not noted that examiners’ actions have been consistent with the letter of the 2009 CRE loan workout guidance, a few stated that examiners were not always complying with its spirit: to allow the banks time to work with borrowers until the current CRE downturn passes. Two bankers with whom we spoke stated that it was their experience that examiners were particularly stringent in 2008 when the financial crisis was escalating, but moderated their approach in late 2009, which is around the time that the 2009 CRE loan workout guidance was issued. An internal review by OCC’s Midsize and Community Banks Division came to a similar conclusion and noted that examiners have become more consistent in their CRE loan treatment through internal discussions, training, and policy communication efforts. A few bank officials with whom we spoke provided specific suggestions on how to improve policy guidance—such as clarifying what amount of debt service coverage is acceptable in a loan workout, how best to conduct global cash flow analysis, and when new appraisals are needed—but a few others felt that some form of general regulatory reprieve was needed for community banks to work through the downturn. Bank officials we interviewed also stated that their experience with increased supervisory scrutiny was being reflected in CAMELS ratings and additional capital requirements. For example, many bank officials thought the management component rating was more critically assessed now, and heavily driven by asset quality and other component ratings. A few bank officials added that they were being rated on deterioration in their portfolios that was due to the broader economic downturn and problems in their geographic market that were out of their control. Some bank officials with whom we spoke stated that examiners have been more aggressive about requiring additional capital—and two bank officials in particular stated they thought this was especially the case for banks with high CRE concentrations. However, a few bankers thought the additional scrutiny was appropriate, but should not necessarily be focused on all banks. For example, one banker stated that if examiners had been applying the guidance stringently before the crisis that perhaps the banks could have avoided some of the current problems. A few others also noted that stricter scrutiny was appropriate given the current CRE market situation and the severity of the economic downturn. Additionally, a few bank officials stated that the regulators should focus on the community banks that caused the problems, rather than subjecting all community banks to such strict scrutiny. Regulators have been incorporating into their regulatory processes a number of lessons learned from the financial crisis, including findings from the agencies’ IGs. Specifically, the IGs of the banking regulators completed 106 MLRs in 2009 and 2010 for banks that failed during the recent financial crisis. As noted in a December 2010 report by the FDIC IG that summarizes certain MLR findings, FDIC determined based on the MLRs that earlier supervisory action was needed to address banks with high risk profiles or weak risk-management practices. We also found in past work that regulators identified a number of weaknesses in institutions’ risk-management systems before the financial crisis began but did not always take forceful actions to address them. In addition, the Financial Crisis Inquiry Commission report notes areas in which regulators could have been more proactive in using regulatory tools to address certain risks in the financial system. A number of nonmanagement regulatory staff in the field offices we visited acknowledged they could have better followed up on outstanding issues. In response to lessons learned, regulatory officials stated that they have been following up more often on matters requiring attention (MRA) and refocusing their supervisory efforts. For example, OCC’s Midsize and Community Banks Division issued an MRA Reference Guide that provides examiners with OCC policy guidance on how to report, follow up on, and keep records related to MRAs. FDIC in June 2009 also implemented its “Forward Looking Supervision” program that re-emphasizes reviewing all aspects of a bank’s risks during the examination process. The program focuses on helping ensure that (1) CAMELS ratings reflect consideration of bank management practices without relying solely on a bank’s financial condition; (2) examiners review risks associated with concentrations and wholesale funding sources; (3) appropriate capital is maintained; and (4) examiners follow up on progress related to MRAs and enforcement actions. Based on our analysis of a nonprobability sample of 55 ROEs, examiners’ findings were generally consistent with policy guidance, with some exceptions. Examiners made statements in ROEs related to the quality of banks’ loan workouts that were consistent with the 2009 guidance on CRE loan workouts. ROEs in our sample that comment broadly on banks’ CRE loan workouts tend to cite areas for improvement (20 of 27). Examiner concerns include that the bank management or its board could have better identified, monitored, or tracked problem loan workouts, and therefore better identified loans that should have been reported as TDRs. Improvements to managing loan workouts that examiners cited include (1) reporting on current loan status and related developments (such as periodic analysis of the borrower’s financial situation); (2) creating and tracking benchmarks to measure workout progress; (3) providing the rationale for loan grades related to loan workouts; and (4) updating collateral values (as appropriate). Almost half of the ROEs in our sample that raise concerns related to CRE loan workouts (8 of 20) specifically note concerns about the nature of banks’ loan workouts. For example, in these cases, examiners most often stated that loans were restructured and extended on unsustainable terms that either resulted in further asset quality deterioration, did not adequately assess the borrower’s ability to repay, or did not follow the 2009 CRE loan workout guidance. In our sample, most ROEs (44 of 55 sampled) raise concerns about CRE concentrations and how they are managed. In more than half of those (24 of 44), the concerns are supported by risk-management deficiencies and explicitly cite the 2006 CRE concentration guidance. Specifically, the findings include concerns on the bank’s (1) need to update or enhance internal policies on CRE concentration limits and CRE underwriting acceptable to the bank; (2) monitoring of CRE concentrations; (3) management information systems used to track and report various types of CRE; (4) ability to inform the bank’s board of directors about trends in CRE concentrations; and (5) the adequacy of stress testing of CRE loans. However, in 7 instances (of 44), the ROE includes statements that the bank must reduce its CRE concentrations, but the basis for this requirement was unclear or appeared inconsistent with the 2006 CRE concentration guidance. For example: One ROE states that the bank needed to match its own internal policy on CRE concentrations to those in the 2006 guidance, and referred to the concentrations in the guidance as “limits.” Referring to the thresholds in the 2006 guidance as limits is inconsistent with that guidance. In two other instances the ROEs state that the bank must reduce its CRE concentrations—citing them as “excessive” for example—but do not focus on the bank’s risk-management systems. However, the enforcement actions for these banks did not require reduced CRE concentrations and emphasized improvements to risk management. One of these ROEs states repeatedly that the bank must reduce its CRE concentrations because the concentrations were too high “in the view of” the regulator. The related enforcement action did not require reduced CRE concentrations explicitly but required that the board “refine the concentration risk-management system” in part by establishing its own limits for certain CRE concentrations and adhering to them. Differences in message between an ROE and an enforcement action could be confusing to bank boards and management as they seek to comply with regulatory requirements for risk management related to CRE concentrations. Moreover, such confusion could lead bank officials to misunderstand whether they should focus on improving their risk-management systems or just reduce their total CRE concentration numbers. Another ROE acknowledges that the bank was below the CRE threshold indicated in the 2006 guidance but states the bank nonetheless should follow the guidance and assesses in detail the bank’s compliance with it. If banks are closely assessed on their compliance with the 2006 CRE guidance, although they have not reached the CRE and ADC thresholds, bank officials could be unclear on what the trigger is for compliance with CRE risk-management requirements. One case in particular provided ambiguous information to the bank about whether it had the appropriate risk-management systems in place to manage its CRE concentrations. The ROE requires a bank with a CRE concentration of about 600 percent of tier 1 capital to address a matter requiring immediate attention to improve stress testing for its CRE portfolio. The ROE also states that the bank needed to reduce its CRE concentrations. However, the ROE states that the bank was in compliance with the 2006 CRE guidance—which requires robust stress testing for CRE portfolios when banks reach certain levels of CRE. Requiring a bank to address a matter requiring immediate attention related to CRE risk management and noting that the bank must reduce its CRE concentrations—while also stating that it is complying with the 2006 guidance—could send an unclear message to bank officials about why they need to reduce CRE concentrations if overall they are complying with the guidance. Additionally, we found that out of 47 ROEs that include CRE concentration information, a number of them did not include CRE concentration numbers that were calculated according to the 2006 CRE guidance. For example, some ROEs appear to include owner-occupied CRE as part of the CRE concentration calculation (14 of 47), although the 2006 CRE guidance specifically excludes this type of CRE. We also found that 23 of these 47 ROEs include CRE concentrations calculated by using some combination of tier 1 capital (and sometimes also include a number simply referred to as a “concentration,” although how it was calculated was unclear). However, another 24 ROEs specify using either total risk- based capital or equity capital (or also include tier 1 capital), which is consistent with the 2006 guidance. The effect of calculating these concentrations differently can be to increase the total CRE concentration number, which increases the scrutiny placed on the banks’ risk- management systems. Sometimes the difference can be large: one ROE in our sample that includes both calculations shows a total CRE concentration of 432 percent when using tier 1 capital and 341 percent when using total risk-based capital. However, the difference also can be smaller: one ROE in our sample has a concentration calculated with tier 1 capital at 141 percent; with total risk-based capital it was 133 percent. While FDIC clarified to its examiners in April 2010 how to calculate CRE concentrations, other regulators have not done so. Moreover, two of the FDIC ROEs in our sample that use only tier 1 capital in the calculation were issued after the April 2010 clarification. Senior officials and field examiners have differing views on whether the 2006 CRE guidance is sufficient in addressing CRE concentration risks. We interviewed about 200 field staff associated with the bank examination review process at FDIC, the Federal Reserve, and OCC in Atlanta, Boston, Dallas, and San Francisco. A number of field examiners from all the agencies stated they did not have the tools to proactively address growing CRE concentrations when the economy was strong or that some banks ignored examiners’ concerns on CRE risk management. A number of examiners also admitted that during strong economic times they could have been more assertive in asking banks to implement risk-management changes related to CRE concentrations. When asked whether limits on CRE concentrations should be considered or whether specific amounts of capital should be required at certain concentration levels, some examiners did not think that limits were the answer, but others thought that there might be some level of concentration that was too high even when managed well, because a market downturn would expose banks to significant losses. In addition, some thought that a focus on additional capital would be sensible given the severe capital shortfalls some banks with CRE concentrations faced during the financial crisis. Senior regulatory officials with whom we spoke had differing views on whether the 2006 guidance was sufficient for examiners to address the buildup in CRE concentration risks. FDIC senior officials stated that the current guidance was sufficient for examiners to address risks related to CRE concentrations and did not think changes were needed. In contrast, OCC has been reviewing whether particular capital requirements should be set for banks that have higher CRE concentrations and stated that this could lead to changes in OCC or interagency guidance. At the Federal Reserve, senior officials believed that the existing 2006 guidance was largely sufficient, but noted that efforts to clarify expectations for stress testing and capital planning were ongoing. The examiners and senior officials with whom we spoke agreed that determining certain limits on CRE concentrations, or requiring specific amounts of additional capital for certain levels of CRE concentrations, would be difficult and require significant study. Federal Reserve officials also noted that limits or specific capital requirements would require studies on the potential effect on credit availability. Examiners exercise significant judgment during examinations, and different perceptions about the 2006 guidance among regulators could send mixed signals to examiners—which could affect how they apply the guidance. In addition, as noted earlier, the regulators’ processes to review and monitor application of examiner guidance and our review of ROEs identified some inconsistencies. Monitoring and revising existing controls, such as guidance, is a key component of a strong internal control system and reflects management’s efforts to implement findings from quality control processes. Regulators require banks to maintain certain minimum capital requirements to help ensure the safety and soundness of the banking system. However, increases in capital requirements can raise the cost of providing loans, which would lead to higher interest rates for borrowers, tighter credit terms, or reduced lending. While capital provides an important cushion against losses for banks, there is a trade-off between building up this cushion to provide greater protection against unexpected losses and increasing lending and returns to shareholders. Holding more capital against each loan means less equity is available to return to shareholders or back new loans. Empirical literature generally supports this basic understanding of the impact of bank capital requirements on lending. For example, researchers found in one study that bank capital requirements substantially affected bank loan growth during the last economic downturn. Specifically, in evaluating bank regulatory agreements in New England, researchers found that capital requirements significantly affected the lending behavior of banks. They noted that those banks with “low or no profits and an inability to obtain new capital at reasonable rates” decreased their assets and liabilities to meet the higher capital-to-asset ratios required by regulatory enforcement actions. Using the “capital crunch hypothesis,” the researchers found that institutions with lower capital ratios had slower loan growth (or loans shrank more rapidly) to try to satisfy capital requirements. Banks with capital bases that have been negatively affected by losses in their CRE portfolios—or those trying to improve their capital ratios or ALLL to guard against potential losses—may need to reduce lending to maintain an adequate capital-to-assets ratio on their balance sheets. Although limited research exists on the impact of CRE loan concentrations on a bank’s ability to lend, an existing study shows that high CRE concentrations can limit loan growth during economic downturns. For example, this study found that banks with high CRE concentrations before the recent financial crisis made loans to other sectors of the economy during this crisis at a “significantly slower rate” than banks that did not have high CRE exposure. In addition, the researchers found that the higher the bank’s CRE concentration prior to the crisis, the more its non-CRE lending slowed during the crisis. The reasons for the contraction of non-CRE lending during the crisis are being examined. The authors hypothesize that the rise in CRE lending and the substantially increased delinquencies on these loans “could have inhibited banks’ willingness or ability to lend to other segments of the economy— particularly when banks’ demand for liquidity and capital was high.” The authors also cite the possibility that high-CRE banks may have failed at a greater rate than banks without that level of CRE exposure. Our assessment of existing studies and interviews with bank officials found that market factors tend to drive CRE downturns and suggests that the regulatory effect of examiner actions on such downturns—as evidenced in studies of the downturn of the 1990s—was minimal. Bankers we interviewed attributed the CRE downturn to market factors such as problems in residential real estate that affected the CRE market and the severity of the broader financial and economic crisis. Although bankers did not state that regulatory policies were the cause of the CRE downturn, many noted that examiner actions were exacerbating it and a few stated banks needed to be given time to work through their troubled assets so they could continue to lend and support the economy. Other bankers stated that examiners have been impeding banks’ ability to make new loans—especially if the bank already had high CRE concentrations. That said, a few bankers noted that CRE loan demand from creditworthy borrowers was down significantly, and this was inhibiting loan growth. There is limited research on the effect of examiner actions on credit cycles; however, the studies that exist suggest the effects are minimal. Specifically: In an analysis of the credit crunch from 1989 to 1992, researchers found modest support for the hypothesis that increased regulatory “toughness” occurred and that this affected bank lending. In a comprehensive study spanning the last financial crisis, three hypotheses were tested regarding changes in regulatory toughness and its impact on bank lending. The data provided what the authors call modest support for all three hypotheses that: (1) toughness increased during the credit crunch from 1989 through 1992, (2) it declined during the boom from 1993 through 1998, and (3) differences in toughness affected bank lending. During the credit crunch they studied, the data show no more than 1 percent additional loans receiving classification or worsening of classification status. During the boom, the data show a similar change for a decrease in loan classifications. The authors also reviewed the economic significance of changes in regulatory “toughness” and found that growth in the number of classified assets by 1 percent would be predicted to decrease the ratio of real estate loans to gross total assets by less than 1 percentage point over the long term, and frequently this impact was projected to be significantly less than 1 percentage point. Changes in CAMEL ratings were not found to have a “consistent” impact on future lending behavior, and when there was an impact, the data show that it was minimal. Another article specifically focuses on how changes in CAMEL ratings affected loan growth from the period that spanned 1985–2004. During the 1985–1993 credit crunch, the authors found some evidence that changes in CAMEL ratings, both the composite and the components, had a significant impact on loan growth for commercial and industrial loans but a smaller effect on consumer loans and real estate loans. However, the authors found minimal evidence that adjustments in CAMEL ratings, including both composite and component ratings, had any systematic effect on loan growth during what they define as an economic recovery period (1994– 2004). The researchers also noted that, based on their research, banks may respond “asymmetrically” to changes in CAMEL ratings. Specifically, banks may decrease loan growth when their CAMEL ratings are downgraded, but they do not necessarily increase it when their CAMEL ratings are upgraded. These studies suggest examiners’ actions can affect lending, albeit minimally. However, the research suggests that regulators were more stringent during the past credit crunch, even holding financial conditions constant. Therefore, regulators should be aware of how their actions can affect broader credit markets, and help ensure that their actions do not contribute to unnecessary procyclical effects: that is, magnification of economic or financial fluctuations. In the aftermath of the recent financial crisis, regulators in the United States and abroad have been attempting to address the procyclical effects of their actions. The procyclical effects of regulation may not adequately discourage overly risky behavior during economic upswings or may inhibit bank lending during downturns, as banks may need to meet more stringent requirements during times when it is more difficult to do so. For instance, if regulators increase capital requirements at the same time that losses from an economic downturn decrease banks’ capital, banks may be less able to lend while they seek to rapidly raise additional capital. This can exacerbate downswings in credit cycles. Consistent and balanced application of policy guidance in strong and weak economic times is important to avoiding unnecessary procyclical effects. Regulatory efforts to better ensure that guidance is applied consistently during strong and weak economic periods helps to ensure that regulatory policies do not exacerbate economic upturns or downturns. While bankers with whom we spoke understood why examiners were looking more closely at CRE loans given the market’s downturn, a few said the shift to closer review and attention was a difficult adjustment. Bank examiners with whom we spoke in the field offices also noted that problems can be difficult to identify during strong economic times. Discussions about the procyclicality of regulation have been a central feature of recent deliberations on revised capital requirements among U.S. banking regulators and the Basel Committee for Banking Supervision. The discussions have been seeking to address the procyclical effects of capital requirements by having banks raise additional capital, commensurate with risks, during times of economic growth. In December 2010, the Basel Committee released the framework for Basel III, which includes increased and higher-quality capital requirements, enhanced risk coverage, the establishment of a leverage ratio as a “backstop” to the risk-based requirement, steps to encourage the “build up” of capital that can be used to absorb losses during “periods of stress,” and the introduction of two global liquidity standards. As part of the establishment of procedures to help ensure the consistent global application of this framework, the Basel Committee has developed standards that will be implemented gradually so that banks make the shift to higher capital and liquidity standards while “supporting lending to the economy.” Federal bank regulators also have acknowledged the importance of addressing procyclicality in regulatory requirements and have made efforts in this regard. For example, the Chairman of the Federal Reserve has stated the importance of regulators acknowledging the significance of procyclicality and he has noted that both the Basel Committee and the Financial Stability Forum have worked to address this as it relates to capital requirements. In addition, FASB has issued accounting guidance aimed at changes to mark-to-market accounting for assessing asset values in inactive markets. More recently, in May 2010, FASB released a proposed Accounting Standards Update, which encompassed proposals on the impairment of financial assets and suggested implementing a more forward-looking impairment model. Based on exposure draft comments, in January 2011 FASB proposed with the International Accounting Standards Board a common solution on how to account for the impairment of financial assets and presented the document for public comment. Efforts such as these seek to address procyclicality concerns related to capital and accounting requirements, while the potential procyclical effects of examiners’ application of policy guidance is something regulators consider in their supervision of banks, according to regulatory officials with whom we spoke. The recent financial crisis underscored how important it is for regulators to evenly apply guidance during strong economic periods, although this was not always the case, as previously noted. Consistent application of guidance is important to avoid unduly hampering credit provision throughout the economy. CRE still is working through a downturn sparked by the broader financial crisis, presenting an ongoing challenge to community banks and regulators. While community banks have been working through the challenges associated with many years of past growth in CRE concentrations, CRE portfolios continue to have a significant effect on community bank balance sheets as loan delinquencies and charge-offs remain historically high. Community banks continue to seek ways to work with their borrowers and shore up capital to remain solvent, but the current economic environment and the expectation that refinancing of CRE loans will bring a new round of market stress suggest many community banks may face these challenges for some time. And, because community banks tend to provide a greater proportion of their loans to small businesses, the availability of credit to small businesses could be constrained while community banks work through these difficulties. Prior to the financial crisis, regulators’ efforts included guidance on CRE concentrations and risk management, but these efforts were not as robust as they could have been in addressing CRE risks. Since the financial crisis, the shift in examination focus and differences in the application of the CRE concentration guidance has contributed to concerns among community banks about regulatory stringency. That is, a number of bankers remain concerned that regulators have become more stringent in reviewing CRE loans since the crisis, which could be explained partly by regulators re-emphasizing fundamentals and addressing lessons learned from the crisis. While we and the regulators have reviewed examiner application of the CRE guidance and generally found examiners’ actions were accurate or well-supported, some inaccuracies and inconsistencies were evident—particularly relating to the 2006 guidance. Regulators have mixed views about the adequacy of the 2006 guidance. Our findings from an analysis of ROEs were similar to those of the regulators’ internal reviews and assessments, which raised questions about the consistency of policy guidance application. Given these findings, and in light of lessons learned from the recent financial crisis and CRE market downturn, the regulators could reassess the adequacy of the guidance. Revising or supplementing the guidance to provide more details about risk- management practices and examples of when to reduce CRE concentrations would help both examiners and bankers better understand how to assess and manage such concentrations. Furthermore, incorporating CRE-specific analyses into the scope of internal and quality assurance reviews—at least while CRE issues remain a concern—will help ensure consistent application of the guidance and produce clearly articulated support in examination reports for requiring reduced concentrations. In this way, regulatory management can better ensure that examination practices related to CRE concentrations and risks will be carried out consistently from examiner to examiner and across regulators. Consistent treatment also will make clear to banks what regulatory expectations are for managing CRE concentration risks. Given the key role community banks play in business lending, bank regulators are aware of the potential effects of their actions in exacerbating the credit cycle. Many factors can affect a bank’s decision to lend, including regulatory requirements. But if such requirements are too stringent, they can unduly limit lending. Such limits on lending can have widespread effects on the economy, as acknowledged in Basel Committee for Banking Supervision discussions on capital requirements. Although isolating the impact of bank supervision is difficult, the recent severe cycle of credit upswings followed by the downturn serves as a useful reminder of the supervisory balance needed to help ensure the safety and soundness of the banking system and support economic recovery. To improve supervision of CRE-related risks, we recommend that the Chairman of the Federal Deposit Insurance Corporation, the Chairman of the Board of Governors of the Federal Reserve System, and the Comptroller of the Currency: Enhance and either re-issue or supplement interagency CRE concentration guidance—based on agreed-upon standards by FDIC, the Federal Reserve, and OCC—to provide greater clarity and more examples to help banks comply with CRE concentration and risk-management requirements and help examiners ensure consistency in their application of the guidance, especially related to reductions in CRE concentrations and calculation of CRE concentrations. After issuing revised or supplemental CRE concentration guidance, incorporate steps in existing review and quality assurance processes, as appropriate, to better ensure that the revised guidance is implemented consistently and that examiners clearly indicate within bank examination reports the basis for requiring a bank to reduce CRE loan concentrations. We provided a draft of this report to FDIC, the Federal Reserve, and OCC for review and comment. All of the agencies provided written comments that we have reprinted in appendixes III, IV, and V, respectively. The agencies also provided technical comments, which we considered and have incorporated as appropriate. In written comments, the Federal Reserve welcomed our recommendations to improve CRE concentrations guidance to help banks and examiners comply with it, and to ensure consistent implementation of such guidance. The Federal Reserve stated that it intends to work with its counterparts at the OCC and FDIC to develop and implement enhancements to CRE concentrations guidance. In its written comments, the OCC agreed with our conclusions and recommendations. OCC noted that community banks’ increased CRE concentrations exposes them to CRE market declines, which may require more explicit regulatory expectations for the robustness of risk- management systems, stress testing, capital planning, and capital levels when CRE concentrations increase. OCC also stated that it would discuss with the other federal banking regulators how to enhance the 2006 CRE concentrations guidance and would ensure the consistent implementation of any revised or supplemented guidance. In written comments, FDIC stated that it has already supplemented the 2006 CRE concentrations and risk-management guidance—citing a 2008 Financial Institution Letter, divisionwide training, and internal reviews. As noted in our report, the federal banking regulators all have numerous controls to help ensure examination consistency, such as training and internal reviews that FDIC highlights in its letter. Nonetheless, our review of ROEs from all three regulators, and their own internal reviews, uncovered instances of inconsistency in the treatment of CRE loans and application of the 2006 CRE guidance—demonstrating that no regulator was immune from consistency issues. We also noted in our report some concerns raised by bank examiners and bankers about applying the 2006 guidance, such as whether the current CRE concentration thresholds are hard limits, how the CRE concentrations should be calculated, and whether specific amounts of additional capital should be required for banks with elevated CRE concentrations. Though FDIC provided clarification on CRE risk management in March 2008 that could contribute to interagency action on CRE guidance, this guidance does not address all of the concerns raised more recently by examiners and bankers that we describe in the report, nor was it developed in conjunction with the other regulators. Therefore, we continue to believe that our recommendations on enhanced or clarified CRE concentration guidance—on an interagency basis that involves the FDIC—would help bankers and examiners better understand how to comply with CRE risk management requirements and help ensure the consistent application of such requirements. FDIC also cited a study by the Bank for International Settlements that concluded the social benefits of higher bank capital requirements outweigh the large reductions in economic activity from a banking collapse. As our report notes, additional capital provides an important cushion against losses, though this comes at a cost. From our review of the literature, there is disagreement on the magnitude of such costs on lending. We added information to the report to clarify this issue, which includes information on the study cited by FDIC. We are sending copies of this report to FDIC, the Federal Reserve, OCC, and other interested parties. The report also is available at no charge on the GAO Web site at http://www.gao.gov. If you or your staffs have any questions about this report, please contact A. Nicole Clowers at (202) 512-8678 or clowersa@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix VI. To assess the condition of the commercial real estate (CRE) market and the implications for community banks, we collected and analyzed the following data to provide information on overall trends in CRE: To determine trends in CRE prices, we analyzed monthly data from Moody’s/REAL Commercial Property Price Index from January 2002 through December 2010. We assessed their reliability for the purposes of providing a picture of the trends in CRE prices by reviewing documentation on how the data were collected and reviewed for accuracy. We determined that the data were sufficiently reliable for the purpose of determining price trends. To determine when CRE loans would be up for refinancing and what percentage of CRE loans were “underwater,” we analyzed data on CRE loan volume by maturity schedule and underlying collateral value provided by Trepp, a commercial mortgage analysis firm. To determine the accuracy of these data, we discussed the data with officials and reviewed documentation on their data quality processes. We determined that the data was sufficiently reliable for purposes of determining when CRE loans were maturing and how many were underwater. To understand how the condition of the CRE market has affected community banks, we analyzed call report data from the Federal Deposit Insurance Corporation (FDIC) for banks with assets of less than $1 billion to assess (1) CRE loan concentrations as a percent of total loans and leases, (2) trends in CRE loan concentrations as a percent of total risk- based capital, (3) trends in CRE loans that were noncurrent, and (4) trends in CRE loans charged off. We assessed changes in the rate of noncurrent CRE loans and CRE loan charge-offs over time. We also calculated average CRE concentration rates at community banks from 1996 through 2010 by taking the aggregate reported total of loans secured by CRE (the sum of construction and development, nonfarm nonresidential, and multifamily residential real estate) divided by the reported amount of total risk-based capital. We conducted similar analyses for commercial banks with assets more than $1 billion. We assessed the reliability of the call report data by reviewing the controls in place to help ensure its accuracy and by relying on past GAO reviews of these data. We determined that they were sufficiently reliable for the purpose of determining trends in CRE loan concentrations and performance at community banks. Reports by the inspectors general of the federal banking regulators provided additional information on the effect of the CRE market downturn on community banks. We reviewed all 106 of the material loss reviews (MLR) issued by the offices of the inspectors general of the Board of Governors of the Federal Reserve System (Federal Reserve), FDIC, and U.S. Department of the Treasury (for the Office of the Comptroller of the Currency, or OCC) from 2009 and 2010. In reviewing the MLRs, we noted which ones cited concentrations in either overall CRE or acquisition, development, and construction loans (ADC) as a factor in the bank’s failure. We also interviewed officials at the Conference of State Bank Regulators on the findings of their independent review of the MLRs. Additional reports provided background and information on the overall condition of CRE markets and their impact on community banks. The reports we reviewed were culled from research and literature reviews from academic journals, the Congressional Research Service, the Congressional Oversight Panel, FDIC, the Federal Reserve, and previous GAO reports. We also spoke with and reviewed speeches and public comments from officials at FDIC, the Federal Reserve, and OCC; bank officials; and academics and a think tank official with expertise on bank examinations or CRE. To determine how the banking regulators responded to trends in the CRE market, we interviewed regulatory and bank officials and reviewed reports of examination (ROE) from a sample of banks. Our sample selection for bank interviews and reviews of ROEs started with a data request to FDIC, the Federal Reserve, and OCC on bank examinations for banks with assets of less than $1 billion with data elements that included the value of their CRE loans, recent CAMELS ratings, total risk-based capital, recent enforcement actions, and other information. We determined that the data were sufficiently reliable for the purposes of selecting a sample, based on prior use of these data and interviews with agency officials. We further refined the sample so that we would have a breadth of banks for interviews and ROE reviews, based on the state in which they were located, their CAMELS composite rating, and their regulator. We also wanted to increase the likelihood that banks we sampled were relevant to our study; therefore, we selected banks with elevated CRE and ADC concentrations and those that had a recent bank examination. To achieve these goals, we selected the sample as follows. From the bank examination information we received from the regulators, we selected banks from California, Georgia, Massachusetts, and Texas. We selected these states because the banks in the first two states have had a relatively greater share of CRE-related nonperforming loans, and the latter two states a relatively smaller share of such loans (as measured by the percentage of CRE loans past due or in nonaccrual). We also wanted to include states from different regions of the country. To inform the selection, we conducted an analysis of call report data downloaded from SNL Financial and held discussions with the federal banking regulators. We analyzed the following data points for commercial banks with assets of less than or equal to $1 billion: (1) loans and leases for construction and land development, multifamily, owner-occupied real estate, other property, and nonfarm nonresidential; (2) loans 90 days past due for each of these categories; and (3) loans in nonaccrual status for each of these categories. We reviewed these commercial banks in each of the 50 states in relation to these measures and total assets in CRE loan categories, and arrived at our four states. We reviewed SNL data reliability and determined that it was sufficient for the purposes of selecting states for our sample. We further refined the sample to include only banks that had an examination conducted after October 2009, to capture those examinations for which the 2009 CRE loan workout guidance would have applied. From that subset, we placed banks into four “pools,” with the goal of speaking with officials from, and reviewing ROEs for, a breadth of community banks from a range of states in our sample, from all three of the banking regulators, and with a range of CAMELS ratings. Pool 1 consisted of banks with good CAMELS composite ratings and high CRE concentrations. In this group, we included banks with a CAMELS composite rating of 1 or 2 and a CRE concentration above 300 percent of CRE loans (including owner-occupied) to total capital. By including owner-occupied loans, we would ensure the largest possible pool of banks and also identify concerns that such banks might have about regulatory treatment of their CRE loans. Pool 2 consisted of banks that generally were in good condition but also had ADC concentrations. Pool 2 criteria were a CAMELS composite rating of 2 and both a 300 percent or greater concentration in overall CRE and a 100 percent or greater concentration in ADC loans to total capital. Pool 3 banks were intended to represent banks that were struggling and that had a higher number of loans classified as loss at the most recent examination. These had a CAMELS composite rating of 3, an overall CRE concentration of 300 percent or greater, and a proportion of loans classified loss to total assets in the 75th percentile and above, for banks within our sample. This “loss ratio” was calculated by taking the total assets classified loss for the most recent examination and dividing them by total assets. Pool 4 was the most distressed pool of banks and consisted of banks with CAMELS composite ratings of 4 or 5, a 300 percent or greater concentration in overall CRE in the 75th percentile or greater for the banks in our sample, and a loss ratio in the 75th percentile of our sample, calculated as described above. Once we selected these pools, we began contacting banks to interview. We simultaneously began requesting ROEs and some related workpapers from the federal banking regulators. In that process, we learned that FDIC and Federal Reserve data included ROEs led by the state banking regulator (because they share examination responsibilities with state examiners), and we excluded these from our sample. The resulting sample from this process is outlined below, broken down by regulator (table 5), state (table 6), and CAMELS ratings (table 7). OCC- supervised banks represent the greatest number of banks in our sample, in part because a number of the examinations in our sample for the Federal Reserve and FDIC were led by the state regulator and were out of the scope of our review. To avoid overweighting the sample with OCC- supervised banks, we randomly removed some OCC-led bank examinations for which we had a sufficient number of cases: banks rated 2 in Texas. To further understand how FDIC, the Federal Reserve, and OCC have responded to the CRE downturn and its effect on community banks, we collected and analyzed interagency policy guidance to examiners related to CRE loan treatment and interagency statements on lending, and also assessed regulatory efforts to address CRE-related concerns. As part of this work, we observed examiner training at FFIEC provided to commissioned examiners, to include a week-long training on CRE and a day-long training on regulatory updates related to policy guidance on CRE loans. To understand banks’ concerns about examiners’ treatment of their CRE loans and how examiners supported findings related to CRE loans, we interviewed bank and regulatory officials and analyzed ROEs. We spoke with community bank officials affiliated with 43 community banks in a number of states (see table 8). Most of the bank officials with whom we spoke were located in California, Georgia, and Texas based on our nonprobability sample. We initiated contact with 62 banks drawn from this sample and interviewed all of the 21 bank officials who responded to us. We gained additional bank views by supplementing this sample. Specifically, we interviewed officials from 22 additional banks that either testified before Congress on the issue of CRE, were identified to us through bank associations, or had learned of our work and wanted to talk to us. When we summarize statements from our interviews with bank officials throughout this report, we use the term “a few” to refer to 3–10 of 43 banks making the statement; “some” to refer to 11–25 of 43 banks making the statement; and “many” to refer to 26–43 banks making a statement. We do not provide specific numbers in the body of the report to avoid overstating the precision of the results from the nonprobability sample we used to select bank interviewees. We interviewed more than 230 regulatory field staff at FDIC, throughout the Federal Reserve System, and at OCC—along with additional staff at these regulators’ headquarters—to seek views and information about recent practices on CRE-specific guidance related to loan workouts and concentrations, training provided on the guidance, CAMELS ratings, capital requirements, and liquidity issues. The interviews were conducted at headquarters offices in Washington, D.C., and by telephone, video teleconference, and in person at district, regional, and field offices in California, Colorado, Georgia, Massachusetts, and Texas. We sought to determine if there were different views or practices related to CRE loan treatment among regulatory roles or between those in headquarters and the field offices, and therefore we interviewed staff in a range of roles and locations. The staff with whom we spoke in field locations included field examiners, their supervisors, case managers, analysts (at OCC), and senior management. Examiners were interviewed both in group settings and individually, and included those who served as examiners-in-charge (EIC) for ROEs in our sample. Examiners interviewed in group settings were gathered by the field offices, and based on our request, had a range of experience (either less than 5 years or more than 5 years, because examiners are usually commissioned within 5 years). Individual meetings with EICs were held based on our nonprobability sample. Of the field staff we interviewed, about 200 were directly involved in drafting or reviewing ROEs, and about 50 of the 200 were in senior management positions. We analyzed 55 ROEs based on our nonprobability sample, as described above, of banks in California, Georgia, Massachusetts, and Texas. We reviewed how examiners supported statements related to banks’ CRE loan workouts, concerns about CRE concentrations, and how examiners calculated CRE concentrations. To understand the controls the regulators have in place to help ensure consistent application of policy guidance, we reviewed the examination report review process at all of the regulators based on regulatory guidance and interviews with regulatory officials. We also collected, compared, and analyzed examination review reports and quality assurance processes for the three regulators to determine what processes were in place to identify and address inconsistencies among examiners, and what findings had resulted from these reviews. For FDIC, we reviewed reports conducted by FDIC’s Internal Control and Review Section for the Atlanta, Dallas, New York, and San Francisco regions, because these regions covered the states in our nonprobability sample. For OCC, we reviewed reports from all four of its districts related to the agency’s quality assurance process that focused on reviewing the classifications of CRE loans. For the Federal Reserve, we reviewed a special engagement report from the General Auditor at the Federal Reserve Bank of Atlanta and a Quality Assurance memorandum for the Atlanta, Kansas City, Philadelphia, and Richmond districts. We also discussed with Federal Reserve staff whether other similar studies existed. To determine and assess the factors that can affect banks’ lending decisions and their impact, we reviewed and summarized academic studies that included analysis of various factors on bank lending. With assistance from a research librarian, we conducted searches of research databases and report sources (Congressional Research Service, Congressional Budget Office, JSTOR, ProQuest, EconLit, Accounting and Tax Database, and Social Science Search). We also sought and reviewed studies cited by the American Bankers Association (ABA) and the Independent Community Bankers of America (ICBA). All studies included published papers released between 1991 and 2010. Based on our selection criteria, we determined that six studies were sufficient for our purposes. Specifically, with the assistance of a senior economist, we analyzed the methodologies underlying these studies and determined that they were the most relevant to our study and also had robust controls. Nonetheless, the research conducted in this area is not exhaustive and focuses primarily on what occurred during the previous economic downturn of the late 1980s and early 1990s. However, we did identify one study that examines the role of CRE and its effect on bank lending in the current financial crisis. To further demonstrate how loan losses, allowance for loan and lease losses, and capital requirements can affect lending, we developed an illustrative example of how loan losses affect capital ratios on a bank’s balance sheet, with assistance from certified public accountants. To obtain information on how examiner practices may affect lending, we interviewed bankers, examiners, and regulatory officials. We also interviewed officials from ABA, the Conference of State Bank Supervisors, and ICBA. Increases in capital requirements or the allowance for loan and lease loss (ALLL) can affect a bank’s ability to lend. Figure 6 illustrates relationships between loan losses, bank balance sheets, and capital requirements under certain assumptions. In this simplified example, the bank begins with $1,000 in assets ($1,010 in loans and $10 in ALLL), $900 in liabilities, and $100 in capital, which equates to a total capital ratio of 10 percent (calculated by dividing total capital by total assets). In this example the bank has analyzed the collectibility of its loans and decided to add $20 to its ALLL in anticipation of loan losses. Provisions for these losses increase the ALLL, which in turn are charged to the bank’s expenses, reduce income, and therefore reduce retained earnings that are included as part of total capital. As a result, the capital ratio falls to 8.2 percent. If the bank identifies as uncollectible and charges off a $20 loan that it holds as an asset, given the assumptions in the example, the bank’s total loans and ALLL would each decline by $20 with no change in capital. If the bank conducts another analysis of its loan collectibility and decides to add to its ALLL, then its capital would be further reduced. For this example, if the bank decides to—or is required to—meet a 10 percent capital ratio after adding to its ALLL to cover estimated future losses, it would need to raise capital by seeking it from the bank’s owners or from investors. If the bank cannot raise additional capital because it is in poor financial condition, or chooses not to because the cost of capital is prohibitive, then it could, among other options, reduce its assets (perhaps by selling assets such as other real estate owned) to decrease liabilities or by paying off borrowings to increase the capital ratio. Reducing assets is often referred to as “shrinking the balance sheet.” However, banks in better financial condition may be able to respond differently in the face of loan losses—they may be able to continue lending without raising additional capital or can rely on their stronger financial position to secure new capital. Kay Kuhlman, Assistant Director, and Robert Lee, Analyst-in-Charge, managed this review. Emily Biskup and Jason Wildhagen also led portions of the research and made significant contributions throughout the report. Anna Maria Ortiz and Rudy Chatlos provided methodological assistance related to our nonprobability sample. Michael Hoffman provided assistance reviewing the methodologies of our research studies and assistance on capital-related issues. JoAnna Berry provided assistance in identifying relevant research literature. Bill Cordrey, Jay Thomas, and Gary Chupka provided accounting assistance. Paul Thompson provided legal assistance. Marc Molino developed the report’s graphics. Barbara Roesmann provided editorial assistance. Jan Bauer, Kimberly Cutright, Andrea Dawson, Nathan Gottfried, Elizabeth Jimenez, Angela Messenger, Lauren Nunnally, Michael Pahr, Ellen Ramachandran, Maria Soriano, Winnie Tsen, Gavin Ugale, and Carrie Watkins held various roles in verifying our findings.

Since the onset of the financial crisis in 2008, commercial real estate (CRE) loan delinquencies have more than doubled. The federal banking regulators have issued statements and guidance encouraging banks to continue lending to creditworthy borrowers and explaining how banks can work with troubled borrowers. However, some banks have stated that examiners' treatment of CRE loans has hampered their ability to lend. This report examines, among other issues, (1) how the Federal Deposit Insurance Corporation (FDIC), Board of Governors of the Federal Reserve System (Federal Reserve), and the Office of the Comptroller of the Currency (OCC) responded to trends in CRE markets and the controls they have for helping ensure consistent application of guidance and (2) the relationships between bank supervision practices and lending. GAO reviewed agency guidance, examination review procedures, reports of examination, and relevant literature and interviewed agency officials, examiners, bank officials, and academics.. Aware of the potential risks of growing CRE concentrations at community banks, federal banking regulators issued guidance on loan concentrations and risk management in 2006 and augmented it with guidance and statements on meeting credit needs and conducting CRE loan workouts from 2008 to 2010. The regulators also conducted training on CRE treatment for examiners and internal reviews to help ensure compliance with CRE guidance. Nevertheless, a number of banks reported that examiners have been applying guidance more stringently since the financial crisis and believe that they have been too harsh in treatment of CRE loans. Regulators have incorporated lessons learned from the crisis into their supervision approach, which may help explain banks' experiences of increased scrutiny. GAO found that examiners generally provided support for exam findings on loan workouts, but identified some inconsistencies in applying the 2006 CRE concentration guidance-- which is similar to what some of the regulators uncovered in their internal reviews. Moreover, regulatory officials had varying views on the adequacy of the 2006 guidance, and some examiners and bankers noted that the guidance lacked clarity on how to comply with it. As a result, examiners and bankers may not have a common understanding about CRE concentration risks. Although many factors influence banks' lending decisions, research shows that the capital banks hold is a key factor. Capital provides an important cushion against losses, but if a bank needs to increase it, the cost of raising capital can raise the cost of providing loans. High CRE concentrations also can limit a bank's ability to lend because the bank may need to raise capital to mitigate the concentration risk during a downturn. Economic research on the effect of regulators' examination practices on banks' lending decisions is limited, but shows that examiners' increased scrutiny during credit downturns can have a small impact on overall lending. Although isolating these impacts is difficult, the recent severe cycle of credit upswings followed by the downturn provides a useful reminder of the balance needed in bank supervision to help ensure the banking system can support economic recovery. Federal banking regulators should enhance or supplement the 2006 CRE concentration guidance and take steps to better ensure that such guidance is consistently applied. The Federal Reserve and OCC agreed with the recommendations. FDIC said that it had implemented strategies to supplement the 2006 guidance.

The Civil Service Reform Act of 1978 (CSRA) provided the first whistleblower protections for disclosures of, among other things, violations of laws, mismanagement, or gross waste of funds for federal employees, former employees, and applicants for employment. The act established both MSPB and OSC and placed OSC within MSPB. OSC was tasked with investigating allegations of prohibited personnel practices (PPP), obtaining corrective actions for employees subjected to PPPs, and initiating disciplinary action against civilian government officials who commit PPPs, among other things. Thereafter, Congress passed the Whistleblower Protection Act of 1989 to strengthen protections for those who claim whistleblower retaliation. This act separated OSC from MSPB, making OSC an independent agency. The Whistleblower Protection Act also created the individual right of action (IRA), allowing whistleblowers to bring their appeals to MSPB after exhausting remedies at OSC. Congress further expanded whistleblower protections in 1994 when it made additional personnel actions subject to coverage and extending whistleblower protections to employees of government corporations and to employees in the Veterans Health Administration (VHA). On November 27, 2012, the President signed WPEA into law. WPEA clarified the scope of protected whistleblowing under the Whistleblower Protection Act; expanded the ability to bring IRA appeals for certain protected activity; provided the authority to award compensatory damages; afforded whistleblower protections to all Transportation Security Administration (TSA) employees; and mandated broader outreach to inform federal employees of their whistleblower rights. OSC and MSPB are the two primary agencies in the executive branch to which federal employees, former employees, or applicants for employment in the federal government go with whistleblower reprisal claims. OSC investigates PPP complaints with respect to a broad range of personnel actions, including appointments, promotions, details, transfers, reassignments, and decisions concerning pay, benefits, awards, and certain decisions concerning education or training. MSPB is an independent, quasi-judicial agency that serves the interests of prompt, procedurally simple dispute resolution. MSPB carries out its statutory responsibilities and authorities primarily by adjudicating individual employee appeals and by conducting merit systems studies. MSPB has jurisdiction over claims made by whistleblowers in two types of appeals—an Individual Right of Action (IRA) appeal and an Otherwise Appealable Action (OAA) appeal. A significant difference between the two types is in how they reach MSPB. In the first type of appeal—an IRA—the individual is subject to a personnel action, such as a reassignment with no reduction in pay or grade, and claims that the action was taken because of whistleblowing or other covered protected activity, but the action need not be one that is directly appealable to MSPB. Examples of the personnel actions that can be raised in an IRA appeal include an appointment, a reassignment, and a performance appraisal. In this kind of case, the individual can appeal to MSPB only if he or she files a complaint with OSC first and OSC does not seek corrective action on the individual’s behalf. In the second kind of case—an OAA—the individual is subject to a personnel action that is directly appealable to MSPB, such as a removal, demotion, or suspension of more than 14 days, and the individual claims that the action was taken because of whistleblowing. In this kind of case, the individual may file an appeal directly with MSPB after the action has been taken. In such an appeal, if the employee proves that MSPB has jurisdiction over the appeal, then both the appealable matter and the claim of reprisal for whistleblowing will be reviewed. IRAs and OAAs must be filed with one of MSPB’s regional or field offices. Once MSPB receives an appeal, it is then docketed in MSPB’s Case Management System and assigned to an administrative judge who issues an acknowledgment that the appeal has been received, orders the agency to file a response to the appeal, and determines whether or not MSPB has jurisdiction over the appeal. If the administrative judge determines that MSPB has jurisdiction, the administrative judge will then conduct a hearing (if requested by the employee), consider the evidence, and issue an initial decision. That decision may be appealed to the three- Member Board of MSPB on a petition for review. According to MSPB, the Board members review initial decisions in much the same way that appellate courts review the decisions of trial courts. Thus, petition for review outcomes are the decisions of the Board and it may issue a decision that affirms, reverses, modifies, or vacates, in whole or in part, the initial decision. The Board may also send back the appeal to the administrative judge for further review. An employee also may appeal either the initial decision (which becomes the final decision if neither party files a timely petition for review), or the petition for review decision, to a U.S. Court of Appeals. MSPB data show the agency received a total of 1,213 IRA appeals for whistleblower reprisal claims for fiscal years 2013 through 2015. As shown in figure 1, our analysis of MSPB data shows the agency received a higher number of IRA appeals in the 3 fiscal years after WPEA was enacted than in the 2 years prior to WPEA. MSPB data show that it received 461 IRA appeals in fiscal year 2013, 360 in fiscal year 2014, and 392 in fiscal year 2015. MSPB data show that it received 261 IRA appeals in fiscal year 2011 and 268 IRA appeals in fiscal year 2012. As discussed earlier, IRA cases often involve personnel actions that are not directly appealable to MSPB, thus employee whistleblowers can only bring an IRA appeal to MSPB if they first filed a complaint with OSC and OSC did not seek corrective action on their behalf. Conversely, MSPB’s data show that it received lower annual numbers of OAA appeals with whistleblower reprisal claims post WPEA. MSPB data show the agency received a total of 629 OAA appeals for fiscal years 2013 through 2015. While WPEA clarified the scope of protected disclosures and expanded the individual right of action appeal for certain other protected activities, the act did not alter MSPB’s jurisdiction over OAAs. As described earlier, for an OAA, the employee alleges whistleblowing as an affirmative defense to the adverse agency action, but the action was one that could be appealed to MSPB even without whistleblowing. Therefore, WPEA may not have necessarily impacted the number of OAAs filed with MSPB. MSPB’s data show it received 266 OAA appeals in fiscal year 2013, 173 in fiscal year 2014, and 190 in fiscal year 2015, compared to 292 in fiscal year 2011 and 294 in fiscal year 2012. As previously described, WPEA, among other things, clarified the scope of protected disclosures, afforded whistleblower protections to all Transportation Security Administration (TSA) employees, expanded IRA appeals rights, and mandated broader outreach to inform federal employees of their whistleblower rights. Each of these may have contributed to a higher number of appeals filed. For example, earlier court decisions had determined that the following disclosures were not protected: disclosures made to the alleged wrongdoer, disclosures made as part of an employee’s normal job duties, and disclosures of information already known. WPEA clarified that each of these disclosures does not lose whistleblower protection due to the nature of the disclosure. In addition, WPEA extended whistleblower protections to the approximately 60,000 TSA employees who were not previously afforded whistleblower protection. WPEA also expanded the IRA for reprisals for certain other protected activities, including filing a whistleblower appeal, cooperating with an inspector general or OSC investigation, or refusing to obey an order that would require the employee to violate a law. Another possible factor for the higher number of appeals may be that employees are more knowledgeable post WPEA about the whistleblower protection process because of the requirement under WPEA for each agency to designate a whistleblower protection ombudsman to educate agency employees about prohibitions on retaliation, and rights and remedies against retaliation, for protected disclosures. According to MSPB officials, MSPB is not in a position to know why the number of OAAs filed with MSPB decreased post WPEA; however they provided a number of factors that could have contributed to the lower numbers. They told us that agencies that furloughed large numbers of employees as a consequence of government sequestration in fiscal year 2013 were likely devoting significant resources to defending furlough appeals before MSPB during fiscal year 2014. MSPB officials explained that some agencies may have taken fewer adverse actions during fiscal year 2014 in light of the resources devoted to furlough appeals. MSPB officials indicated that the government-wide shutdown in October 2013, and the resulting disruption to agencies’ operations, may have also caused some agencies to take fewer adverse actions during fiscal year 2014. The officials said that another possible explanation for the reduction in OAAs is that employees with MSPB appeal rights might have elected to challenge appealable actions in other ways, such as by filing a grievance or an equal employment opportunity or OSC complaint. According to MSPB officials, such matters could still eventually come before MSPB. For example, if the employee files a complaint with OSC, the employee may file with MSPB once OSC has terminated its investigation or 120 days after the complaint was filed with OSC if OSC has not notified the employee that it will seek corrective action. Employees from 10 agencies accounted for more than one-third of the IRA and OAA appeals filed with MSPB during fiscal years 2013 through 2015. As shown in table 1, employees from these 10 agencies filed 636 of the total 1,842 IRA and OAA appeals received by MSPB during this period. Employees in the Department of Veterans Affairs’ (VA) Veterans Health Administration (VHA) filed the most whistleblower appeals with MSPB, a total of 290 OAA and IRA appeals combined, followed by employees at: VA (excluding VHA) (54 appeals); the Department of the Army’s Army Medical Command (50 appeals); DHS’s Bureau of Immigration and Customs Enforcement (40 appeals); DHS’s TSA (38 appeals); and the Social Security Administration (38 appeals). MSPB data show that the majority of IRA appeals closed from fiscal years 2013 through 2015 were dismissed, settled, or withdrawn and not adjudicated on the merits (see sidebars). As shown in figure 2, nearly half of these appeals—41 percent overall—were dismissed because MSPB determined either the appeal was untimely or the action alleged by the employee whistleblower was outside MSPB’s jurisdiction. For example, an administrative judge would dismiss the appeal if the employee whistleblower alleges reprisal because of a disclosure of a violation of agency policy, as this is not a protected disclosure under section 2302. Only a small percentage of IRA appeals—15 percent of the total appeals closed during the 3 fiscal years—advanced further in the appeals process where the cases were adjudicated on the merits, meaning the employee had the opportunity to present his or her case to the administrative judge. Of the 221 IRA appeals closed during fiscal year 2013, a total of 137 appeals (or 62 percent) were dismissed. Of the remaining 84 appeals, 26 (or 12 percent) were withdrawn by the employee, and 38 appeals (or 17 percent) were settled between the employee and the agency. The remaining 20 appeals (or 9 percent) were adjudicated by MSPB on the merits of the appeal. Settled According to MSPB officials, the agency encourages employees filing appeals to settle with their respective agencies before the case goes to a hearing. Parties may settle for many different reasons. Generally, officials cannot assess why certain appeals settle and others do not. MSPB officials also say that the agency has trained and certified mediators as part of its free Mediation Appeals Program (MAP). MAP facilitates a discussion between the employee and his or her respective agency to help identify issues to resolve disputes. Such steps may help settle the appeal quickly, economically, and to the benefit of both parties. However, both parties must agree to its use before the appeal will be accepted for MAP. Also, both must agree on its resolution before any settlement is concluded. Unlike the traditional appeal process, the parties control the result of the case under the guidance of the mediator. The mediator plays no role in deciding the appeal should an accord not be reached. More than half of all OAA appeals received by MSPB during fiscal years 2013 through 2015 were dismissed (see figure 3). According to MSPB officials, OAA appeals are dismissed without addressing the merits of the whistleblower claim for a variety of reasons including the appeal falling outside of MSPB’s jurisdiction, timeliness, or because the individual made a binding election to proceed with his or her claim in another manner, such as filing a grievance under a negotiated grievance procedure. MSPB data show that the number of IRA appeals increased that were timely, were within MSPB’s jurisdiction, and were neither withdrawn nor settled by the employee, and were therefore adjudicated on the merits. In fiscal year 2013, 20 IRA appeals were adjudicated on the merits (see figure 4). This number increased to 50 IRA appeals in fiscal year 2014 and 56 IRA appeals in fiscal year 2015. MSPB officials stated that, post WPEA, it is now more likely that an IRA appeal would be adjudicated on the merits because WPEA expanded through clarification the universe of disclosures that were considered protected. Additionally, according to MSPB officials, under the Board’s decision in Day v. DHS, 2013 M.S.P.B. 49, appeals pending on the effective date of the act that would have been dismissed prior to WPEA were ruled to be within MSPB’s jurisdiction. Corrective action was granted in 1 IRA appeal in fiscal year 2013. This number increased to 5 IRA appeals in fiscal year 2014, and 14 IRA appeals in fiscal year 2015. MSPB’s data do not identify what specific corrective action was granted for each appeal. Corrective action is ordered in any appeal if the employee has demonstrated that (1) he or she made a protected disclosure; (2) the agency has taken or threatened to take a personnel action against him or her; and (3) his or her protected disclosure was a contributing factor in the personnel action. All three elements must be present for MSPB to find for corrective action. MSPB will not order corrective action if, after a finding that a protected disclosure was a contributing factor, the agency demonstrated by clear and convincing evidence that it would have taken the same personnel action in the absence of such disclosure. Corrective action for a PPP violation may include reinstatement, back pay (lost wages), medical costs, compensatory damages, any other reasonable and foreseeable consequential damages, and attorneys’ fees and costs. MSPB granted corrective action to few OAA appeals that were adjudicated on the merits during fiscal years 2013 through 2015 (see figure 5). In approximately one-third of all OAA appeals that were adjudicated on the merits, MSPB determined that the agency would have taken the same action. We did not identify any cases where MSPB or any courts determined the allegations of reprisal in the appeals to be malicious or frivolous. Generally, MSPB, the Federal Circuit, and other courts do not make determinations on the intent, or the reason for, an employee filing a whistleblower claim. MSPB officials told us that because administrative judges do not use the term “malicious,” it would be unlikely that we would find the term used in appeals’ files or outcomes. Generally, to establish jurisdiction over an IRA appeal regarding whistleblowing, an employee whistleblower must prove that he or she exhausted his or her administrative remedies before OSC and make nonfrivolous allegations that (1) he or she engaged in whistleblowing activity by making a protected disclosure and (2) the disclosure was a contributing factor in the agency’s decision to take or fail to take a personnel action. If a whistleblower appeal were to be dismissed for lack of jurisdiction, one of the potential reasons for dismissal could be because the individual was unable to make a nonfrivolous allegation that he or she made a protected disclosure or that the disclosure was a contributing factor in a personnel action. However, MSPB officials stated that dismissing an appeal on this basis cannot be interpreted as an affirmative finding that the claim was determined frivolous. Moreover, according to MSPB officials, MSPB does not track or maintain such information on potentially frivolous appeals. As previously stated, WPEA required MSPB for the first time to include whistleblower appeals data in its annual performance reports. According to the act, each report should include, among other things, the number of appeals received in the regional and field offices and the outcomes decided. According to MSPB officials, the agency implemented a number of changes in reporting in response to WPEA. For example, MSPB created a standard form for appeals that could be used for data collection throughout its eight regional and field offices. Further, MSPB began collecting more specific information about the outcomes of whistleblower appeals post WPEA. Prior to WPEA, for example, MSPB’s Case Management System only captured that a whistleblower reprisal claim was not found in an IRA appeal. So, the appeal was adjudicated on the merits, but without any specific information related to the analysis of the reprisal issue. Post WPEA, MSPB changed its Case Management System and began capturing additional information about why corrective action was not granted in an IRA appeal adjudicated on the merits. This includes “no protected disclosure found,” “no contributing factor,” or “agency would have taken same action.” We found differences between the reported data in MSPB’s annual performance reports and the whistleblower data MSPB provided to us. In some cases, the differences between the two sets of data numbered a few appeals. However, other differences in the numbers were as much as 43 for IRAs received in fiscal year 2013 (MSPB reported 418 in its annual performance reports, compared to 461 in the data it provided to us) and 64 for IRAs dismissed for timeliness/res judicata/lack of jurisdiction in fiscal year 2013 (MSPB reported 161 in its annual performance reports, compared to 97 in the data it reported to us). According to MSPB officials, in some cases appeals were reported under more than one outcome due to human error and limitations in the data coding that resulted in over reporting of appeals closed. In addition, MSPB officials told us the agency did not save its original queries or datasets used to identify appeals data received and closed for fiscal year 2013. MSPB recreated these queries upon our request. MSPB officials stated that changes to its Case Management System and new data entry procedures are communicated routinely to appropriate staff, typically through e-mail and in-person and virtual meetings. MSPB officials told us that administrative judges were provided additional guidance on issue coding post WPEA; however, as of September 2016, MSPB had not updated its Case Management System data entry user guide in response to WPEA’s reporting requirements and changes in the required elements of data reporting that may affect data entry. The user guide, which outlines how appeals should be added and updated, was last issued in May 2004. Internal controls dictate that if there is a significant change in an entity’s process, management should review this process in a timely manner after the change to determine that the control activities are designed and implemented appropriately. Internal control standards also dictate that management should periodically review policies, procedures, and related control activities for continued relevance and effectiveness in achieving the entity’s objectives or addressing related risks. By reviewing and updating its data entry user guide for whistleblower appeals, MSPB may be able to more quickly identify potential coding errors and provide appropriate guidance. In reviewing the data provided by MSPB, we identified 156 separate whistleblower appeals that were reported under more than one outcome. After reviewing the duplicate entries identified by us, MSPB was able to identify the single correct outcome for 145 of the 156 cases, and those 145 cases were therefore included in the numbers we analyzed. MSPB officials indicated that the remaining duplicate entries were the result of either incorrect coding or unusual outcomes that did not fit comfortably within any of the enumerated categories. MSPB officials therefore recommended that those cases not be used for reporting purposes. We also identified a total of 364 cases that MSPB reported as closed but that were not included in any of the reported datasets for cases received. MSPB reviewed each of these cases and found that many of them were not reported as received because they had been received prior to fiscal year 2011, the first year for which we requested data. However, MSPB also found approximately 90 initial appeals that had not been reported. Those cases have now been added to the appropriate dataset. Internal controls state that management should process data into information and then evaluate the processed information so that it is quality information. By developing a process to identify data discrepancies or other anomalies in its data queries and the resulting datasets, MSPB may be able to quickly identify weaknesses overall in its data entry process allowing the agency to more accurately report the required appeals information. Congress relies on MSPB’s annual reports on the number of appeals received and the outcome of appeals alleging violations of whistleblower protection laws to help examine WPEA’s effectiveness and to identify unintended consequences of the legislation. MSPB, with improved reporting processes, has an opportunity to better assist Congress. Generally, the subject matter specialists who participated in our focus groups had mixed views as to whether MSPB’s authority should be expanded. Some strongly supported expanding MSPB’s authority, while others strongly opposed it. Focus group participants said that granting MSPB summary judgment may be advantageous for involved agencies and MSPB because greater efficiencies may be gained (see sidebar). However, in doing so, employee whistleblowers could lose their right to a hearing, which some participants said represents a disadvantage to employee whistleblowers. What is summary judgment? Summary judgment authority is a procedural device used when there is no dispute as to the material facts of the case, and a party is entitled to judgment as a matter of law and the responsibility of the court. Who has summary judgment authority? Other federal adjudicatory bodies, such as the Equal Employment Opportunity Commission (EEOC) and the Federal Labor Relations Authority, have summary judgment authority. According to EEOC officials we met with, summary judgment is frequently used at EEOC as a vetting process to determine whether or not a hearing will be held after the record has been developed. They also stated that summary judgment is used as a case management tool and its advantages include eliminating the need for a hearing. Focus group participants in favor of summary judgment for MSPB stated that it would be advantageous for involved agencies and MSPB because it would create greater efficiencies. They said that involved agencies would not have to engage in an exhaustive, extensive process when the facts do not warrant it if MSPB had summary judgment authority. One participant stated that having summary judgment would separate valid complaints from meritless complaints that may not have any facts in dispute, such as employees who are shielding themselves from misconduct they actually committed. Another participant said that involved agencies would delay making settlement decisions until summary judgment rulings were made instead of currently settling cases agencies deemed meritless to save agency resources. Participants also told us that MSPB could gain greater efficiencies from summary judgment because it could potentially decrease the number of appeals for which administrative judges would conduct hearings, thus allowing administrative judges to issue decisions on additional appeals, reducing potential backlogs, and resolving cases sooner. As a result, MSPB could conserve time and resources in the long term. One participant proposed a 5-year pilot to determine and measure the efficiency of summary judgment on affected parties. On the other hand, focus group participants opposing summary judgment for MSPB said that a motion for summary judgment may not resolve whistleblower cases any faster because it would require more discovery, depositions, and documents to establish the disputed facts thereby creating more prehearing litigation work. In addition, they said that while MSPB’s current caseload may decrease in the short term, MSPB may spend more time dealing with appeals of unfavorable summary judgment decisions. They explained that this could lead to prolonged litigation, thereby eliminating any potential efficiency gained by MSPB. Participants also discussed MSPB’s ability to dismiss cases under its jurisdictional test, noting that this process is like a summary judgment review. However, one participant distinguished MSPB’s current jurisdictional test from summary judgment because the jurisdictional test only involves the whistleblower, not the agency, and only reviews whether the whistleblower has exhausted his or her administrative remedies and can establish the jurisdictional requirements for MSPB review. Three participants pointed out the small number of appeals that are currently adjudicated on the merits as an example of MSPB’s current efficiency in using its jurisdictional test. Adding summary judgment authority to MSPB, according to another participant, would not only be duplicative but would also create two barriers for a whistleblower’s case to move forward. Focus group participants not in favor of granting summary judgment to MSPB also stated that doing so could unfairly erode employee whistleblowers’ right to a hearing. They explained that the procedural nature of responding to a summary judgment motion may include legal technicalities that employee whistleblowers, who choose to represent themselves without legal counsel, may not understand. Specifically, they stated that summary judgment may be too complicated a legal tactic to master for the average employee whistleblower, who may be confused about the burden needed to overcome a summary judgment motion. For example, employee whistleblowers may lack legal expertise to properly complete required paperwork to address the motion for summary judgment. Conversely, agency attorneys who represent the involved agencies’ position in whistleblower appeals may be better positioned to draft sophisticated briefs and well-prepared affidavits to which employee whistleblowers would be unable to respond—a scenario that focus group participants believe favors involved agencies. One participant said that involved agencies already win a majority of the appeals at MSPB, and if summary judgment were granted, the odds of employee whistleblowers prevailing against an agency’s motion would be nonexistent. The participant stated that the current process is already an uneven playing field. Another participant explained that proving retaliatory intent by the agency for whistleblowing may be too difficult to achieve where the employee whistleblower must rely on submitting a brief and documents to support the employee’s allegations rather than a hearing. In addition, focus group participants cited the additional costs of conducting discovery as another potential disadvantage for employee whistleblowers. Conducting discovery includes gathering documentation and conducting depositions to establish disputed facts in order to draft motions required for summary judgment. Focus group participants generally agreed that it would be beneficial for employee whistleblowers if U.S. District Courts were granted jurisdiction for whistleblower cases. They said that this would give whistleblowers access to a jury trial similar to nonfederal employee whistleblowers. One participant pointed out that a double standard already exists because corporate whistleblowers can go to district court while federal employee whistleblowers are unable to do so. Another participant stated that it would be a good idea to get federal employee whistleblower cases into district court because the only option typically available to get into federal court is on appeal to the U.S. Court of Appeals for the Federal Circuit, which this participant said overwhelmingly upholds the Board’s decisions. This participant also said that it would be more feasible for federal employees to go to district court for a full review of their claims rather than to appeal to the U.S. Court of Appeals for the Federal Circuit, which has a limited scope of review. Another participant explained that adding additional procedural options for employee whistleblowers, such as district court, could help strengthen the law. While it may be advantageous for employee whistleblowers, participants also pointed out that allowing U.S. District Courts this jurisdiction could be a disadvantage to those courts because of the increased overall workload that would result from having jurisdiction for whistleblower cases. However, views on the extent to which the court’s workload might increase were mixed. Participants in favor of district court jurisdiction stated that there would be a negligible impact because not all employee whistleblowers would exercise their right to court for a number of reasons. For instance, they said the relatively high cost of filing and any kick out provisions that may require employee whistleblowers to wait before filing in district court may keep the burden on courts low. Two participants said that the district court would most likely only be used for high-profile or complex, in-depth proceedings involving scientific questions or technical cases, and not for all types of whistleblower reprisals. Another participant said that while district court is slower in resolving cases, employee whistleblowers could get injunctive relief. Participants against district court jurisdiction generally concurred that the current MSPB process is sufficient and efficient and provides employee whistleblowers with an opportunity for a hearing. They stated that adding the district court as an option would undermine efficiency because it would create more backlogs for an already overburdened court system. Another potential disadvantage cited by focus group participants against granting jurisdiction to U.S. District Courts was that involved agencies would have to relinquish control over their cases. Participants told us that the responsibility for appeals cases would shift from the involved agencies to the Department of Justice. As a result, the Assistant U.S. Attorney would be responsible for litigating all whistleblower reprisal appeals that went to district court while agency counsel would assist. This could be a potential advantage or disadvantage because involved agencies would no longer have responsibility for litigating such cases. One participant said that Assistant U.S. Attorneys would push to settle whistleblower cases because many of them are already overburdened. Another participant echoed that Assistant U.S. Attorneys are so busy that they may be unwilling to devote appropriate attention to whistleblower cases, in contrast to counsel for involved agencies. While focus group participants generally agreed that the U.S. District Courts should be granted jurisdiction for some whistleblower appeals, they stated that not all cases should be eligible for U.S. District Courts. However, they differed on which whistleblower appeal cases should be eligible and at what stage of the process. For instance, participants were divided on whether all or a subset of whistleblower appeals, such as only terminations or demotions, should be eligible to be heard in U.S. District Court. We asked focus group participants for their input on how the process could work and what types of appeals should be considered if employee whistleblowers had the right to file their whistleblower case at the U.S. District Court. The current process requires employee whistleblowers to file a claim with either OSC or MSPB, depending on the type of agency action alleged. If whistleblowers are dissatisfied with the initial decision at MSPB, they can appeal it to the U.S. Court of Appeals or to MSPB’s Board through a petition for review. In fiscal year 2015, employee whistleblowers filed a total of 582 whistleblower reprisal appeals with MSPB and 95 petitions for review with the Board. For scope of jurisdiction, the focus group participants primarily proposed two options as shown in figures 6 and 7. As shown in figure 6, the first option proposed by focus group participants would require employee whistleblowers to exhaust all administrative remedies before permitting them to proceed to district court. In discussing this option, participants also discussed whether to allow employee whistleblowers to go to district court after a specified amount of time if no action is taken by MSPB. One participant highlighted whistleblower statutes that applied to his agency which included employees’ right of action to go to district court if a decision was not rendered after a certain period. However, participants expressed concern that this process would allow too many “bites of the apple,” or opportunities to appeal, resulting in inefficiencies because employee whistleblowers could prolong the entire process by filing claims at district court for a full, de novo review after receiving adverse MSPB decisions. Using discrimination complaints as an example, participants explained that complainants can go to district court at many different points in the process if they choose to do so. Specifically, complainants have a right to file a civil action in district court after final agency action or after 180 days of filing a complaint if the agency has not taken final action, or, after the complainant appeals to EEOC, after a final EEOC decision or after 180 days from the date of filing an appeal with EEOC if no final decision has been issued. In response to the EEOC example, one participant found that there was a negligible increase of between 0.01 and 0.015 percent of EEOC cases that went to district courts based on a study conducted by his organization. Another participant stated that the option for employee whistleblowers to elect to go to district court was previously proposed in prior legislation, but with a specified time limit if no action was taken by MSPB. As shown in figure 7, the second option proposed by focus group participants would allow employee whistleblowers to have “one bite of the apple”—in other words, selecting either MSPB or district court to have their appeal heard. Participants generally agreed that this would be the preferred method if employee whistleblowers were permitted to go to district court. Two participants proposed that the employee whistleblower should choose between MSPB and district court following OSC’s initial review. Participants also said that the election of remedies may be more suitable because employee whistleblowers may choose to go to MSPB because it does not charge a filing fee and therefore is more affordable than the court route. Meanwhile, attorney-represented employee whistleblowers or those with high-profile and complex cases may choose to bypass MSPB and go to district court. Finally, participants discussed what types of whistleblower retaliation should be eligible for district court review. The two proposals discussed were (1) all whistleblower reprisal appeals, including minor and major offenses; and (2) only a subset of whistleblower reprisal appeals, such as the more serious adverse actions. However, participants disagreed on which whistleblower appeal types should be eligible. One participant stated that it should not be limited to a subset and that all whistleblower reprisal appeals should go to district court. On the other hand, those in favor of having a subset of whistleblower reprisal appeals eligible for court review stated that only the more severe adverse actions, such as terminations or demotions, should be allowed. If whistleblower appeals could go to district court, the statute would need to clearly define what is eligible for court review. One participant pointed out that none of the corporate whistleblower statutes have a breakdown of which serious adverse actions can go to district court versus administrative adjudication. We solicited input from MSPB and the Administrative Office of the U.S. Courts to obtain their opinions on how their respective organizations might be impacted if Congress granted U.S. District Court jurisdiction for whistleblower cases. MSPB officials declined to comment. Officials at the Administrative Office of the U.S. Courts told us that forecasting specific impacts would be speculative. They also said that legislation that imposes new workload requirements on the federal judiciary may necessitate additional resources. They added that, should specific legislation be proposed by Congress, the Judicial Conference and its committee(s) and staff would then carefully review the legislation, and consider its effect to determine if a Judicial Conference position on the legislation was warranted. Lawmakers have recognized that whistleblowers are crucial in helping to expose waste, fraud, abuse, mismanagement, and threats to public health and safety across the federal government; however, those who come forward may face reprisal from their employers. Courts, over the years, have narrowly interpreted the type and recipient of the disclosure that qualifies for whistleblower protection. The Whistleblower Protection Enhancement Act of 2012 was enacted to further ensure that employees come forward and report violations of law, or certain agency mismanagement or ethical violations, primarily by expanding protected disclosures. As required by WPEA, MSPB has collected and reported information on the number of whistleblower appeals filed and on the outcomes of cases decided by MSPB. However, we found discrepancies between the data MSPB publicly reported and the data MSPB provided to us. Some of these discrepancies may have been caused by the lack of updated data entry user guides and the lack of a quality check in its data analysis and reporting process. To help ensure the accuracy of MSPB’s reporting on whistleblower appeals received and closed, the Chairman of MSPB should take the following two actions: Update MSPB’s data entry user guide to include additional guidance and procedures to help improve the identification of appropriate whistleblower appeal closing codes to use. Add a quality check in MSPB’s data analysis and reporting process to better identify discrepancies or other anomalies in data queries and the resulting datasets. We provided a draft of this report to the Chairman of MSPB for review and comment. In its written comments, reproduced in appendix III, MSPB agreed with our recommendations and stated that it is currently incorporating them. MSPB pointed out that it already provides an “impartial adjudication” of all appeals filed there, and takes no policy position on the issue of providing appellants the option for a jury trial in a U.S. District Court. As stated in our report, the documented views for and against granting jurisdiction to U.S. District Courts for whistleblower cases were obtained from focus groups consisting of whistleblower protection subject matter specialists. MSPB also provided technical comments on the draft report, which we incorporated in the report as appropriate. We are sending copies of this report to the Chairman of the Merit Systems Protection Board, as well as to the appropriate congressional committees and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2717 or jonesy@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. The objectives of this engagement were to describe the Merit Systems Protection Board’s (MSPB) implementation of the Whistleblower Protection Enhancement Act of 2012 (WPEA). Specifically, this report: (1) describes changes in the number of whistleblower reprisal appeals filed with MSPB, as well as the outcome of appeals, since WPEA’s effective date, including whether or not MSPB, the United States Court of Appeals for the Federal Circuit, or any other court determined the allegations to be frivolous or malicious; and provides (2) subject matter specialists’ views on granting MSPB summary judgment authority for whistleblower cases, and (3) subject matter specialists’ views on granting jurisdiction for some subset of whistleblower cases to be decided by a district court of the United States, and the potential impact on MSPB and the federal court system. To address the first objective, we reviewed the whistleblower reprisal appeal data published by MSPB for fiscal years 2013 through 2015. We requested and obtained from MSPB data on all whistleblower reprisal appeals filed in fiscal years 2013 through 2015, after WPEA’s effective date, as well as all whistleblower reprisal appeals closed during the same time frame. We also requested and reviewed MSPB whistleblower reprisal appeals data for 2 fiscal years prior to WPEA’s effective date, specifically for fiscal years 2011 and 2012, in order to have 5 years of consecutive data to identify changes in the number of whistleblower reprisal appeals filed after WPEA’s effective date. We reviewed relevant MSPB documents, including data entry guidelines and annual performance reports. To supplement the documentary evidence obtained, we interviewed MSPB officials on whistleblower reprisal appeals and WPEA’s enactment. To assess the reliability of the data, we compared data on whistleblower reprisal appeals reported in MSPB’s annual performance reports with data that MSPB provided us and we identified discrepancies. We provided MSPB with the results of our data analysis and coding language. We discussed the discrepancies with MSPB staff as discussed in the report. Based on our analysis and our review of related documentation and interviews with MSPB officials, we corrected the data and the resulting data set that we used for analysis was sufficiently reliable to provide a general indication of a change in the numbers of appeals received and closed post WPEA. Although MSPB has jurisdiction for other appeals, we evaluated only information related to the whistleblower reprisal appeal process. We did a keyword search using Lexis Nexis to determine if the U.S. Court of Appeals for the Federal Circuit or any other appeals court had concluded that a whistleblower reprisal appeal from MSPB was “malicious” or “frivolous” post WPEA. To address the second and third objectives, we conducted six focus groups in July 2016 with subject matter specialists knowledgeable about WPEA. Of the six focus groups, four included a mix of whistleblower advocacy and agency representatives (the mixed sessions) while the remaining two were conducted with agency representatives from the Chief Human Capital Officers Council. To select the 28 participants for the six focus groups, we solicited suggestions from MSPB and the Office of Special Counsel based on their direct knowledge of whistleblower cases, as well as the Equal Employment Opportunity Commission, that has summary judgment authority. We also conducted our own research to identify additional participants in the whistleblower community and reached out to the federal inspector general and Chief Human Capital Officers communities. Although these focus group discussions provided rich information, they are nongeneralizable and do not reflect opinions of all in the whistleblower community. The selected participants represent nonprofit organizations, a national union, private law firms, and federal agencies. For a list of organizations represented by the focus group participants, see appendix II. We asked each focus group to discuss the following issues: (1) WPEA’s impact on various entities, such as the alleged employee whistleblower and the involved agency; (2) reasons for or against granting MSPB summary judgment authority for whistleblower cases; and (3) reasons for or against granting U.S. District Court jurisdiction for whistleblower cases. All six focus groups were recorded and transcribed. We analyzed the results to identify common themes and patterns among the mixed and Chief Human Capital Officers Council focus groups. Additionally, we solicited input from MSPB and the Administrative Office of the U.S. Courts to obtain their opinions on how they might be impacted if Congress granted U.S. District Court jurisdiction for whistleblower cases. We conducted this performance audit from January 2016 to November 2016, in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the contact named above, Clifton G. Douglas Jr., Assistant Director; Dewi Djunaidy, Analyst-in-Charge; Amy Bowser; Sara Daleski; Karin Fangman; Joshua Garties; Susan Sato; and Cynthia Saunders made major contributions to this report. Also contributing to this report were Ellen Grady; Robert Robinson; Mark Ryan; and Stewart Small.

Federal employee whistleblowers—individuals who report violations of law or certain agency mismanagement— may risk reprisals from their agencies. WPEA seeks to strengthen the rights of and protections for federal whistleblowers. WPEA includes a provision for GAO to report on the law's implementation. This report (1) describes changes in the number of whistleblower reprisal appeals filed with MSPB, as well as the outcome of appeals, since WPEA's effective date; (2) provides subject matter specialists' views about granting MSPB summary judgment authority for whistleblower cases; and (3) provides subject matter specialists' views about granting jurisdiction for a subset of whistleblower appeals to be decided by a district court of the United States. GAO obtained data from MSPB on whistleblower appeals received and closed from fiscal years 2011 through 2015. This period begins approximately 2 years prior to WPEA's effective date of December 27, 2012, and continues through the end of fiscal year 2015. GAO also conducted six focus groups with subject matter specialists that include whistleblower advocates and agency labor relations officials. Of the two types of whistleblower appeals—individual right of action (IRA) and otherwise appealable action (OAA)—Merit Systems Protection Board (MSPB) data show higher numbers of IRA appeals received by MSPB after enactment of the Whistleblower Protection Enhancement Act of 2012 (WPEA). In an IRA appeal, an individual has been subject to a personnel action, such as a reassignment, and claims the action was reprisal for whistleblowing. In contrast, the number of OAA appeals decreased after WPEA. In an OAA appeal, an individual has been subject to an action directly appealable to MSPB, such as a demotion, and claims that the action was taken because of whistleblowing. WPEA, among other things, clarified the scope of protected disclosures which may have contributed to a higher number of IRA appeals. WPEA did not alter MSPB's jurisdiction over OAAs. MSPB has taken steps to collect and report whistleblower appeals data. GAO identified a number of weaknesses in MSPB's data collection. Some were due to shortcomings in data coding that resulted in over reporting appeals closed. Further, MSPB has not updated its data entry user guides to reflect new reporting requirements nor has it instituted checks to ensure data accuracy, which are inconsistent with internal control standards. Subject matter specialists varied in their opinions about the benefits of granting summary judgment authority—a procedural device used when there is no dispute as to the material facts of the case—to MSPB. Benefits cited included a more efficient process; but negatives included the loss of a whistleblower's right to a hearing. Similarly, the specialists provided a mix of opinions about granting jurisdiction for cases to U.S. District Courts. A benefit cited was access to a jury trial, while a negative cited was the increased workload of the courts. GAO recommends that MSPB help ensure the accuracy of its reporting on whistleblower appeals received and closed by (1) updating its data entry user guide to include additional guidance and procedures, and (2) adding a quality check in its data analysis and reporting process to better identify discrepancies. MSPB agreed with these recommendations.

As a result of coverage limits, restrictions, and exclusions in NFIP policies, insurance payments for flood damage may not pay all of the costs of repairing or replacing flood-damaged property. Certain NFIP limitations are embedded in statute; others have been promulgated by FEMA pursuant to its statutory authority. FEMA officials said that the coverage limitations are necessary to keep the NFIP self-supporting and actuarially sound. Thus, the program is designed to strike a balance between premium prices and coverage. The following are several examples of NFIP coverage limitations, restrictions, and exclusions that affect the premium and amount a claimant could expect to receive for flood damage: Homeowners are required to insure their homes for the amount of their federally backed mortgages. If a home is insured for less than 80 percent of its full replacement cost or the maximum coverage amount of $250,000, or it is not a primary residence, NFIP will pay the actual cash value for the damage. The actual cash value represents the original cost of the structure less depreciation and, in most cases, will not cover the full cost to repair damage to or replace the dwelling. The value of physical depreciation is based on the age and condition of the item. If a home is insured for 80 percent or more of its full replacement cost or the maximum coverage amount of $250,000 and is a primary residence, NFIP will pay replacement costs for damage to the dwelling. The policy defines replacement cost as coverage to replace the damaged part of the dwelling with materials of like kind and quality to what was damaged. The policy will pay the amount actually spent for this repair or replacement up to its limit. A homeowner may choose not to insure personal property under the program. A deductible amount is applied against claims for dwellings and personal property. Basements, which are defined as building areas below grade level on all sides, have limited coverage that does not include payment to repair or replace finished walls, floors, furniture, and other personal property. The personal property limit paid for jewelry, artwork, and home business equipment is $2,500 for all items combined. No coverage is provided on these items if they are located in a basement. Actual cash value, not replacement value, is paid on all covered furniture and other personal property. Thus, personal property is also depreciated. A detached garage is covered by the dwelling policy only if it is used solely for vehicles and storage. If the garage is improved (e.g., a sink is installed), flood damage to the structure is not covered under the NFIP. In a hypothetical property adjustment we developed with the assistance of FEMA’s director of claims, a poorly maintained 30-year-old home located in a designated flood zone had flood damage when a nearby river overflowed. The property was valued at $60,000. It was insured under the NFIP for $30,000. Although a contractor estimated it would cost $40,000 to repair damages to the structure and personal property losses totaled another $10,000, a NFIP adjuster determined that payment on the claim was $8,000. The following circumstances reduced the amount of coverage: The homeowner had chosen not to insure his personal property. Because the homeowner did not insure the structure for at least 80 percent of its value, actual cash value will be paid for repairs or replacement of damage to the dwelling. Since the condition is poor, the actual cash value will be low. A $1,000 deductible will be applied. The adjuster determined that some problems that needed to be addressed had not been caused by the flood (e.g., leaking pipes in the bathroom and preexisting mold in the basement). Only limited coverage is allowed in the basement of the home, where the largest amount of damage occurred. The work of selling, servicing, and adjusting claims on NFIP policies is carried out by thousands of private sector insurance agents and adjusters who work independently or are employed by insurance companies or vendors under subcontract to insurance companies to handle their flood business. In contrast, according to a FEMA official, about 40 FEMA employees are responsible for regulating, managing, and overseeing the program, which is expected to grow to about 4.7 million policies in 2005. FEMA is assisted in this effort by about 170 contractor employees. According to FEMA, about 95 percent of the NFIP policies in force are written by agents who work for or represent 94 private insurance companies that issue policies and adjust flood claims in their own names. The companies, called write-your-own companies, receive an expense allowance from FEMA of about one-third of the premium amounts for their services and are required to remit premium income in excess of this allowance to the National Flood Insurance Fund. The insurance companies share the FEMA expense allowance with the agent selling and servicing the policy and a vendor, if the company has subcontracted with one to handle all or part of its flood insurance business. It is clear that some agents do not understand the program. It is very complex and different from the other lines of insurance. Flood insurance is much more complex than automobile and homeowners insurance. Some items of specific concern are definitions of elevated buildings and basements. Officials at FEMA, the four insurance companies, and the vendor said that they offered support to the insurance agents who sell and service NFIP policies. Reported support included training, help from telephone hotline customer service representatives, development of rate quotes, and Web sites with NFIP information. However, other than requiring that agents meet basic state insurance licensing requirements, neither FEMA nor the four insurance companies have historically required that agents complete training or demonstrate a basic level of knowledge of the NFIP to sell flood policies. When losses occur, flood adjusters employed by insurance companies or independent contractors become the eyes and ears of the NFIP. Claims adjusters are assigned to policyholders by their insurance companies after the policyholders have notified their agents of a flood loss and the agents have written loss reports. Adjusters are responsible for assessing damage; estimating losses; and submitting required reports, work sheets, and photographs to the claimants’ insurance company, where the claim is reviewed and, if approved, processed for payment. They work under the coordination of a general adjuster assigned to manage claims adjustments for the flood event. Unlike agents who sell flood insurance policies, adjusters must be certified by FEMA to work on NFIP claims. To be approved to adjust residential flood losses, an adjuster must have at least 4 consecutive years of full-time property loss adjusting experience and have attended an adjuster workshop, among other requirements. To keep their certifications current, adjusters are required to take a 1-day refresher workshop each year and pass a written examination testing their knowledge each year. FEMA’s program contractor maintains a database of independent adjusters who are qualified to adjust flood claims. A FEMA official said that 4,844 flood-certified adjusters are registered in the database, as of April 2005. A NFIP official noted that the adjuster community is stretched thin when a major flood event occurs. Adjusters and insurance companies are paid for claims settlements from the National Flood Insurance Fund based on the size of the losses they settle. The write-your-own company receives about 3.3 percent of the incurred loss, according to FEMA’s NFIP claims director. Adjusters are to be paid at the time claims are settled based on a standard fee schedule. For example, an adjuster receives $1,000 for a claim of between $25,000 and $35,000. FEMA’s primary method of overseeing the work of write-your-own companies is to conduct an operational review of every participating company at least every 3 years. In addition, FEMA relies on about 10 general adjusters employed by its program contractor to check the work of claims adjusters in reinspections of a sample of adjustments done after every flood event. According to the FEMA director of NFIP claims, one or two employees from FEMA’s NFIP Claims and Underwriting sections go on-site to review the operations of write-your-own companies at least every 3 years. They do reviews more frequently, if necessary, to follow up on any findings from a previous visit. The auditors are to request that a random sample of 100 files be pulled for them to review. Files that are closed without payment and those with particularly large settlements are to be included in the sample of files reviewed. Auditors are to check the files for completeness and accuracy. For example, they must make sure that there are photographs to document damage. Auditors are also to look at internal controls in place at the company. If a write-your-own company does not pass an operational review, FEMA requires that it develop an action plan to correct the problems and schedules a follow-up review in 6 months to determine whether progress has been made, according to the NFIP director of claims. If the company continues to have problems and fails to implement an action plan, it can ultimately be withdrawn from the NFIP. According to FEMA officials, a company has been asked to withdraw from the NFIP once in the program’s history. About 3 years ago, a write-your-own company was withdrawn from the NFIP in part because of issues raised in operational reviews and in part to other financial problems. Three of the four flood program managers for write-your-own companies whom we interviewed thought operational reviews were an effective way for FEMA to ensure that the NFIP is run according to established legislation and regulation. The fourth manager said that he had no opinion one way or the other. Interviewees noted that the reviews caught problems, and while FEMA had a small audit staff, the auditors were knowledgeable and provided about the right level of review. Two of the four flood program managers said that recent operational reviews had identified problems on policies they had recently purchased from other companies and that they were working to rewrite some policies and address other oversight issues. General adjusters are to do reinspections of open claims. FEMA chooses a random sample of about 4 percent of the claims for every flood event to reinspect, according to the NFIP claims director. If the general adjuster determines that a company paid for an expense that should not have been covered, FEMA is to be reimbursed by the write-your-own company. If a general adjuster finds that an adjuster missed an item in the original inspection, the general adjuster is to add it back into the claims report so that the policyholder will be compensated for it. The instructors at an adjuster refresher training session we attended noted the following as common errors identified in reinspections of claims: improper measurement of room dimensions; improper allocation of damage between wind and flood (homeowners’ policies cover wind damage, while the NFIP covers flood losses); poor communication with homeowners on the process they are following to inspect the property and settle the claim. “I am in the flood insurance business because I believe in the program. It does a lot of good. Floods are horrible occurrences. A homeowner sees the water coming but can do nothing to stop it. The smell is horrible. Whole communities are affected, and the emotional toll is tremendous. I have seen the NFIP do great good for many people.” Each of the interviewees, when asked how the NFIP could be improved, said that FEMA should look for ways to make the program less complex and more similar to other property insurance programs. For example, a vendor manager noted, “if the customers, the agents, and the adjusters all have difficulty understanding the program, it is too complicated.” “As FEMA has tried to make the flood program more actuarily sound, it has made it more complex. FEMA has required of us more information, more forms, and more photos to be scanned into files. Those requirements cost money to implement. As an industry, we are looking at how the flood line might be more compatible with other lines of insurance business to be more cost-efficient. Now the flood business is so unique that it requires special handling.” FEMA officials said that some documentation (i.e., elevation certificates) is required because the NFIP is part of FEMA’s broader flood plain management strategy that combines insurance protection with hazard mitigation to reduce future flood damage to homes. The officials noted that, while the NFIP has different requirements than homeowners insurance, it is not necessarily more complex and that the more familiar agents become with the requirements of the NFIP, the easier it becomes for them to routinely handle documentation requirements. Congress mandated that within 6 months of the enactment of the Flood Insurance Reform Act, FEMA establish (1) insurance agent education and training requirements, (2) new processes for explaining coverage to policyholders when they purchase and renew policies, and (3) an appeals process for claimants who are dissatisfied with the settlement of their claims. The 6-month mandated deadline elapsed on December 30, 2004, but FEMA is still working to complete these mandated efforts. According to FEMA officials, in order to address the requirements to establish insurance agent education and training and for explaining coverage to policyholders, the agency must go through the rule-making process. FEMA officials also said to address the requirement for explaining policy coverage, they are waiting for DHS approval before finalizing the draft materials that will accompany the flood insurance policy. When DHS approves the draft materials, they will be published in the Federal Register as part of the rule-making process. Regarding the requirement for an appeals process, the agency must initiate and complete formal rule making. FEMA officials said that this process takes more than 6 months and could not be completed within the mandated time frame. To address the requirement in the Flood Insurance Reform Act of 2004 to establish insurance agent education and training requirements, FEMA is working with state insurance commissions. An official said FEMA is still in the planning stages of meeting the requirement and is discussing options with state insurance commissions, but has not yet developed an action plan. When a customer purchases a flood insurance policy, the main document he or she is to receive from the insurance agent is the policy. A congressional report accompanying the Flood Insurance Reform Act stated that the NFIP did not provide “simple” forms or claims guidelines for flood victims to follow, making access to information about flood insurance policies difficult to obtain. To address this concern, the act requires FEMA to provide simplified forms and a flood insurance claims handbook to policyholders at the time of purchase or renewal and at the time of flood loss. FEMA has drafted new materials that would be provided to the policyholder at the time of purchase or renewal of the flood insurance policy. The draft material includes: a supplemental form that would explain the policy, such as the amount of deductibles, the exact coverage being purchased, exclusions from coverage, and an explanation of how lost items and damages will be valued under the policy at the time of loss; a flood insurance handbook to describe procedures to be followed to file a claim and provide detailed information on an appeals process that FEMA is to develop; and an acknowledgment form that the policyholder has received the flood insurance policy and that the policy only covers building property for the dwelling and does not provide coverage for contents or personal property. Before the materials are finalized, FEMA must go through rule making and publish them in the Federal Register. FEMA expects to have these forms and handbook finalized by October 2005. If a policyholder has a grievance about a flood insurance claim, proof of loss, or loss estimate, he or she may informally appeal to the insurance agent, to the insurance adjustor’s supervisor, or to a hotline where a customer representative is to provide assistance. There is currently no official recourse for the policyholder. To provide official recourse to policyholders, section 205 of the Flood Insurance Reform Act requires that FEMA establish a formal appeals process through which policyholders may appeal decisions on their claims. FEMA is developing a formal appeals process for a policyholder to follow if he or she has a grievance. The proposed new appeals process must go through the rule-making process with publication of a draft and a final set of procedures in the Federal Register. A FEMA official was uncertain when the process would be completed, but said that it would be after October 2005. Mr. Chairman and Members of the Committee, this concludes my prepared statement. I would be pleased to answer any questions you and the Committee members may have. For further information about this statement, please contact William O. Jenkins, Jr. Director, Homeland Security and Justice Issues, on (202) 512-8777 or jenkinswo@gao.gov or Christopher Keisling, Assistant Director, Homeland Security and Justice, on (404) 679-1917 or at keislingc@gao.gov. Major contributors to this testimony included Christine Davis, Pawnee Davis, and Deborah Knorr.

According to the National Flood Insurance Program (NFIP), 90 percent of all natural disasters in the United States involve flooding. Because of the catastrophic and unpredictable nature of floods, private insurance companies do not typically cover flood losses. Congress established the NFIP in 1968 to provide an insurance alternative to disaster assistance in response to the escalating costs of repairing flood damage. During congressional hearings on provisions of the Flood Insurance Reform Act of 2004, several legislators testified on NFIP shortcomings, as reported by constituents whose properties had been flooded by Hurricane Isabel in September 2003. The act required GAO to study coverage provided under the NFIP. It also required the Federal Emergency Management Agency (FEMA), the administrator of the NFIP, to take steps to address concerns about coverage and claims procedures. Today's testimony is based on work in progress to address this mandate. It provides preliminary information on (1) the types of coverage limits, restrictions, and exclusions under the NFIP; (2) how FEMA, in partnership with private insurers, manages and oversees the NFIP and the views of selected private sector program managers on how the program is working; and (3) the status of FEMA's efforts to comply with provisions of the Flood Insurance Reform Act. As a result of policy limits, restrictions, and exclusions, insurance payments to claimants for flood damage may not cover all of the costs of repairing or replacing damaged property. Some limitations are embedded in statute and others have been promulgated by FEMA pursuant to its statutory authority. FEMA officials said that the coverage limitations are necessary to keep the NFIP self-supporting and actuarially sound. Thus, the program is designed to strike a balance between premium prices and coverage. For example, homeowners may choose not to insure personal property under the program. If they do elect to have this coverage, the value of personal property is depreciated. Basement coverage does not include payment to repair or replace finished walls and floors. The work of selling, servicing, and adjusting claims on NFIP policies is carried out by thousands of private sector insurance agents and adjusters under the regulation, management, and oversight of about 40 FEMA employees assisted by about 170 contractor employees. Agents are the main point of contact for policyholders. Four private sector NFIP managers we interviewed said that the agents have varying levels of NFIP knowledge. While training and support are available, historically neither FEMA nor the insurance companies have required completion of training or demonstration of basic program knowledge. Flood-certified adjusters are responsible for assessing damage and estimating losses when flooding occurs. Unlike agents, adjusters have mandatory training requirements. FEMA has oversight mechanisms in place to review the operations of the insurance companies and the work of adjusters. The private sector NFIP managers GAO interviewed were generally supportive of the program. However, they said that FEMA should find ways to make it less complex than and more similar to other property insurance programs. FEMA has taken steps to address its mandates in the Flood Insurance Reform Act, but it did not meet the 6-month timeframe specified. For example, to establish an insurance agent training requirement, an official said FEMA is discussing options but has not developed an action plan. To meet the requirement to provide "simple and complete information" to NFIP policyholders, FEMA has drafted materials explaining coverage, deductibles, and claim- and appeals-related procedures that it expects to have finalized by October 2005.

The HEA, as amended, specifies a formula, known as the federal need analysis methodology, to determine students’ eligibility for federal need- based student aid. A student’s need for financial aid is calculated using a formula that subtracts a student’s expected family contribution (EFC) from the student’s cost of attendance (COA). The EFC represents the applicant’s household financial resources that are considered available to help pay for postsecondary education expenses and is calculated by reducing the financial resources reported by applicants by certain expenses and allowances, including state and other tax allowances. The factors used to calculate the EFC differ based on whether students are classified as financially dependent on their parents or are independent. For dependent students, the EFC is based on such factors as the student’s and parents’ income and assets, as well as family size and whether the family has other children enrolled in college. For independent students, the EFC is based on such factors as the student’s and, if married, spouse’s income and assets and whether the student has any dependents other than a spouse, as well as the number of family members enrolled in college. The COA at a postsecondary institution includes tuition, fees, books, and living expenses. If the COA is greater than the EFC, the difference between the two represents the student’s financial need. For example, if a postsecondary institution has a COA of $10,000 and a student has an EFC of $4,000, the student is eligible for up to $6,000 of federal need-based aid. If the EFC is greater than the COA, the student is not eligible for federal need-based aid but may qualify for aid that is not need-based. In the 2007 to 2008 academic year, more than 12 million prospective students applied for federal student aid. Education requires student aid applicants to complete the FAFSA to collect students’ data for the federal need analysis formula. Although the primary purpose of the FAFSA is to help Education distribute federal student aid, the form also accommodates the needs of state and institutional aid programs that rely on the FAFSA data for their own eligibility calculations. Prior to the creation of the FAFSA, separate application forms were required to apply for various types of federal, state, and institutional aid. As required by law, in 1992, Education streamlined the student aid application process by consolidating many of these forms into a unified FAFSA. Since then, Education has undertaken periodic efforts to modify the form’s design and instructions and reduce data elements required of applicants. In addition, several amendments to the HEA have also modified the FAFSA by adding, for example, some new questions to the application. The 2009 to 2010 FAFSA consists of more than 100 questions that collect information ranging from basic contact information to the current value of assets. While less than half of the questions ask for financial information, many of these questions require applicants and the parents of dependent applicants to search for information located on tax returns as well as bank, business, and investment records. While both online and paper versions of the FAFSA are available, Education recommends that applicants file online to take advantage of features that are not available on the paper form, such as skip-logic, which allows applicants to skip questions that do not pertain to them. For example, independent students are not asked for their parents’ financial information. The online FAFSA can also detect many errors prior to applicants’ submission and allow the applicant to make corrections. If such errors are made on the paper form, they may take weeks to resolve, delaying financial aid eligibility notification from Education. According to Education, 98 percent of FAFSA applications are submitted online. Education’s student aid application processing cycle covers an 18-month period. For example, applicants seeking federal aid for the 2009 to 2010 award year can submit a FAFSA from January 1, 2009, through June 30, 2010; however, most states and institutional aid programs have earlier FAFSA deadlines. After Education processes an applicant’s FAFSA, a report is sent to the applicant or made available online. This report includes the applicant’s EFC, the types of federal aid for which the applicant qualifies, and information about any errors—such as questions the applicant did not complete—that Education identified during FAFSA processing. Colleges send applicants award letters after admission, providing students with types and amounts of federal, state, and institutional aid, should the student decide to enroll (see fig. 1). Title IV of the HEA, as amended, authorizes the following federal aid programs. Grants. Generally, grants do not need to be repaid unless the recipient withdraws from school and owes a refund. They include the following types: Pell Grant. Grants to low-income undergraduate and certain postbaccalaureate students who are enrolled in a degree or certificate program and have federally defined financial need. For the 2009 to 2010 award year, the maximum award is $5,350. Supplemental Educational Opportunity Grant. Grants to undergraduate students with federally defined financial need. Priority for this award is given to Pell Grant recipients. In general, an annual award may not be less than $200 and may not exceed $4,000. Academic Competitiveness Grant. Grants to Pell-eligible students enrolled at least half-time in their first or second year of study who completed a rigorous secondary school program of study. First year students may receive up to $750, and second year students who have at least a 3.0 cumulative GPA at the end of the first year of study may receive up to $1,300. National Science and Mathematics Access to Retain Talent (SMART) Grant. Grants up to $4,000 per year to Pell-eligible students in their third or fourth year of study (or fifth year of a 5-year program) majoring in certain subject areas with at least a 3.0 cumulative GPA. Teacher Education Assistance for College and Higher Education (TEACH) Grant. Grants to undergraduate, postbaccalaureate, and graduate students who are taking or will be taking course work necessary to begin a career in teaching. TEACH provides up to $4,000 per year to recipients who agree to teach full-time in a designated teacher shortage area for 4 years, or the grant will be converted to a loan that must be repaid with interest. Work-study. Work-study is employment in on-campus or certain off- campus jobs for which students who have federally defined need earn at least the current federal minimum wage. The college or off-campus employer pays a portion of their wages, while the federal government pays the remainder. Work-study is awarded based on a student’s need minus other aid awarded. Colleges participating in the program administer the funds and make award decisions based on the student’s financial need. Loans. These are funds that are borrowed and must be repaid, with interest. Perkins Loan. Low interest—5 percent—loans made through participating schools to undergraduate and graduate students. Interest does not accrue while the student is enrolled at least half-time in an eligible program. Priority is given to students who have exceptional federally defined need. Undergraduate students can borrow up to $5,500 annually, and graduate students can borrow up to $8,000 annually. Stafford and Plus loans. Loans made by private lenders and guaranteed by the federal government (Federal Family Education Loan Program) or made directly by the federal government through a student’s school (Direct Loan Program). Subsidized Stafford Loan. A loan made to students enrolled at least half-time in an eligible program of study and have federally defined financial need. The federal government pays the interest costs on the loan while the student is in school. The amount students can borrow is based on their year in school and whether they are classified as financially dependent on their parents or independent. Unsubsidized Stafford Loan. A nonneed-based loan made to students enrolled at least half-time in an eligible program of study. Although the terms and conditions of the loan (i.e., interest rates, etc.) are the same as those for subsidized loans, students are responsible for paying all interest costs on the loan. PLUS Loan. A nonneed-based loan made to credit-worthy parents of dependent undergraduate students enrolled at least-half-time in an eligible program of study, and credit-worthy graduate and professional degree students. Borrowers are responsible for paying all interest on the loan, and can borrow up to the cost of attendance minus any financial aid the student receives. Currently, dependent students may borrow combined subsidized and unsubsidized Stafford loans up to $5,500 in their first year of college, $6,500 in their second year, and $7,500 in their third year and beyond. Independent students can borrow combined subsidized and unsubsidized Stafford loans up to $9,500 in their first year, $10,500 in their second year, and $12,500 in their third year and beyond. There are aggregate limits for an entire undergraduate education of $31,000 for dependent students and $57,500 for independent students. Graduate and professional degree students can borrow combined subsidized and unsubsidized Stafford loans up to $20,500 per year, and their aggregate limit for undergraduate and graduate education generally cannot exceed $138,500. Many study group participants said using federal income tax data the government already collects on annual income tax forms could shorten the application process, making it easier on students and their families. Specifically, these participants proposed that relevant federal income tax data be directly transferred to the appropriate answer fields on each applicant’s online FAFSA. With answers to as many as 20 FAFSA questions already collected on federal tax forms, such a change could decrease the quantity and complexity of the financial questions for the majority of applicants who complete the FAFSA with information from tax returns. Several participants said the FAFSA questions that take the longest to complete tend to be those that require applicants to search their tax forms for answers, such as questions on combined income and untaxed portions of retirement accounts. One participant noted that directly populating the FAFSA with tax data could particularly ease the burden on many first- generation college students and their parents, who may have less familiarity with the application process. Several participants also suggested that the use of federal income tax data could increase the number of applications completed, because fewer applicants would be discouraged by the number of questions they had to answer. One participant referred to her research showing that, by electronically populating an applicant’s FAFSA with IRS data, an independent applicant could complete the online FAFSA in less than 10 minutes, on average. Another participant noted that financial questions are the source of most errors on the FAFSA, resulting in students and colleges spending additional time making corrections and verifying information. Currently, Education requires colleges to verify that up to 30 percent of their federal aid recipients provided accurate financial information. This process involves the school’s financial aid office comparing an applicant’s or his family’s information on the FAFSA to supporting documentation, including tax returns that the student must provide to the school. Although 98 percent of applicants submit the FAFSA electronically, a few participants noted that some low-income applicants may not have reliable internet access in their homes. These participants said that applicants without such access would be more likely to complete a paper FAFSA and would not benefit from an electronic transfer of IRS data. Many study group participants proposed changes to the design and contents of the FAFSA that could help streamline the form and make the application process less daunting for prospective students. Instructions. Although Education has worked to clarify the online and paper FAFSA instructions in recent years, some participants said the length and complexity of the instructions continue to confuse applicants and should be further reduced and clarified. Beginning in January 2010, Education plans to improve instructions for the online FAFSA by customizing the directions for each question based on information the applicant has already provided. For example, if applicants enter their marital status as single, the directions for each question will only provide information pertinent to single FAFSA applicants. Tone. A few participants raised concerns about the tone of some questions on the FAFSA—saying they conveyed the wrong message to applicants—with one participant likening the application to a “beware of dog” sign instead of a welcome mat. For example, two participants recommended rewording a question that asks if applicants will attend college full-time or part-time, saying the question erroneously gives applicants the impression that they must commit to one of these options in order to apply for aid. However, this question is not used to determine federal aid eligibility, and students do not have to make this decision until they decide to enroll in college. Education has recently announced plans to make changes to the online FAFSA that are designed to encourage applicants to complete the application process. For example, in January 2010, Education plans to begin providing status indicators throughout the application that will inform students of their progress in completing the FAFSA. Skip-logic design. Several participants praised the online FAFSA feature—known as skip-logic—which allows applicants to bypass some questions that are not relevant to their student aid eligibility, based on their answers to previous questions. Education’s recent expansion of skip-logic now allows applicants to bypass a selective service registration question unless they are male and younger most dependency questions if they are at least 24 years of age or married, three homeless determination questions unless they are 21 years of age or younger and answered yes to a question asking if they are homeless or at risk of being homeless, and all parental data for dependent applicants who only wish to apply for an unsubsidized loan if their parents refuse to provide their data on the FAFSA and refuse to provide financial support to the applicant. In addition, upcoming enhancements planned for January 2010 will allow applicants to skip asset information if they have low incomes and assets are not required to drug conviction questions if they are first-time college students, as federal aid eligibility is not affected by drug convictions that occur prior to college enrollment; and the state of legal residence and date of residency question if they confirm that, for at least the previous 5 years, their state of legal residence is the same as the state on their mailing address. Two participants recommended improving the skip-logic for financial questions by grouping together all questions requiring applicants to reference their tax forms and reordering the FAFSA questions to match the order in which data are collected on tax forms. Another participant noted that while skip-logic may be helpful for online applicants, it does not benefit the approximately 2 percent of individuals who complete the paper FAFSA. Content. Many participants offered recommendations to streamline the FAFSA’s contents. A few participants suggested it would be helpful to know the extent to which each question is used in determining eligibility for federal, state, and institutional aid, since the value of information gained from particular questions may be outweighed by the potential loss of applicants due to the form’s length. In addition, a few participants recommended significantly shortening the FAFSA by removing all questions not used to determine federal eligibility or financial need. For example, some states consider the highest level of education an applicant’s parents have completed in targeting aid. However, some participants expressed concern that eliminating such questions from the FAFSA may cause states and colleges to develop additional forms in order to get the data they need, which could in turn increase the overall burden on applicants. One participant added that it might be difficult for colleges to get a comparable response rate if they tried to collect nonfinancial data. However, another participant suggested that colleges could collect this information on the acceptance form students submit after receiving letters of admission. Two participants also suggested eliminating questions currently asked on the FAFSA to determine aid eligibility—such as those regarding selective service registration and drug convictions—that are not used to calculate financial need. Many study group participants supported changing the need analysis formula to require less financial information from federal student aid applicants. Because the formula is specified by federal statute, any modifications would require legislative change. In discussing the need for a simpler formula, several participants noted both the sheer number of questions required to compute aid eligibility and the relative difficulty of answering the financial questions. For example, one participant stated that applicants have a far easier time answering questions about their marital status than they do complicated questions about their assets. In particular, participants discussed the merits of relying solely on a family’s income— as measured by adjusted gross income (AGI) on federal income tax forms—and the number of tax exemptions to determine aid eligibility. Similar proposals have been suggested previously. Such a shift would greatly reduce the number of financial questions asked on the FAFSA— from more than 45 items to only 2—which several participants said could decrease the burden applicants face in completing the form. Nevertheless, a few state aid administrators we interviewed said they saw no need to change the current formula, and one added that the online form’s skip- logic keeps the formula from being too burdensome for most applicants. Several participants also noted that a simpler formula could increase applicants’ awareness of their potential financial aid eligibility, and perhaps increase the probability that they will go to college. For example, if eligibility were determined solely by AGI and number of tax exemptions, Education could publish a reference table that would allow students to estimate their aid eligibility far earlier in the aid application process and plan accordingly. Supplying applicants with earlier, more precise information on eligibility could ultimately render the EFC unnecessary, replacing the estimate of an applicant’s or family’s contribution to the cost of education with a direct calculation of federal aid eligibility. Several participants said that providing this type of early information could lead to an increase in the number of financial aid applications submitted and could encourage prospective students to apply for aid earlier in the cycle. In addition, participants said that reducing the financial information required by the formula could in turn simplify the verification process for financial aid administrators and applicants. If fewer financial items were included in the formula, financial aid administrators would have to collect and verify less information, and students selected for verification would similarly be relieved of the burden of providing large amounts of documentation. Many participants said that although it would make the application process easier on prospective students, reducing the amount of financial information collected would likely result in some change in the distribution of federal, state, and institutional aid, and would create new winners and losers among the pool of aid applicants. Participants differed in their assessment of whether the benefit of simplifying the formula outweighed the potential cost in how federal aid is distributed among applicants. For example, several participants were concerned that eliminating asset information from the federal formula could result in some applicants with high-value assets, such as large bank accounts or trust funds, receiving more need-based aid—such as Pell Grants—than they would under the current system. By potentially increasing the pool of applicants who qualify for need-based aid, eliminating assets could result in a smaller award amount for each Pell Grant recipient, as the maximum amount of the grant depends on program funding and can change each year. Several other participants, however, asserted that simplifying the formula would be beneficial to applicants—particularly those with the greatest need and those who do not currently apply—and is therefore worth the potential cost of a shift in who receives federal aid. One participant’s research suggests that redistribution at the federal level would be relatively small if the formula included only AGI and number of tax exemptions. Specifically, she said she has found that approximately 85 to 90 percent of the variation in how the Pell Grant is awarded can be explained by those two factors. According to participants, formula changes could also affect the distribution of state and institutional aid to varying degrees, as many states and institutions use the eligibility determinations from the FAFSA to allocate their awards. Consequently, some participants were concerned that—much like with federal aid—the removal of asset data from the formula could increase the overall pool of eligible award recipients, and in turn reduce the size of state and institutional financial aid awards available to the neediest applicants. Some participants asserted that, while a change in the formula may not greatly affect Pell-eligible students, state and institutional need-based aid reaches into middle income ranges where the implications may be far greater. One of these participants added that when her state modeled what would happen to its aid program if it eliminated assets from the eligibility formula, it found that expenditures would increase by 12 percent. Because her state, like many others, has a program in which all eligible applicants are entitled to receive aid, she explained that such a change would result in the state either having to cover additional costs or providing less money to each eligible student. For a few participants, concerns over how formula changes might affect state aid extended to the way in which family size is calculated. These participants said that the number of tax exemptions is a poor measure of the household size of an applicant or applicant’s family. For example, some children or other family members may live with an applicant but not be listed as dependents on tax forms. However, other participants countered that household size is already difficult to measure accurately under the current formula. Several participants recommended further analysis on how and to what extent applicants for both federal and state aid would be affected by possible changes to the federal formula. Many participants stressed that, as federal aid does not cover the entire cost of education for most students, the information needs of states and institutions must be addressed in any plan to simplify the federal formula. Two participants, however, maintained that it was not reasonable to expect a single application to serve the needs of both the federal aid program and programs from all states and institutions. Education’s recent proposal to limit the federal formula to financial information available through federal income tax forms would eliminate 26 financial questions—including those on assets—while retaining up to 20 financial questions that could all be answered with federal income tax data. Although such a formula would decrease the burden on applicants, one participant noted that it would not be concise enough to allow for a simple reference table that prospective students could use to estimate aid eligibility, as it would if it were limited to AGI and tax exemptions. Proposed legislation passed by the House of Representatives and under consideration in the Senate would simplify the student need analysis formula by setting an asset cap for some aid programs and eliminating assets from the need analysis of students whose families do not equal or exceed the cap. While many study group participants noted the potential benefits of using IRS data to populate the online FAFSA, some raised questions about the feasibility and limitations of this approach. Applicants currently complete the FAFSA with income information from the tax year prior to the beginning of the school year for which they are applying for aid. For example, an applicant who completed the FAFSA with the intent of beginning college in fall 2009 is required to use 2008 income information. However, Education officials said that because the tax calendar permits most tax filers to file their income taxes for the prior calendar year as late as April 15, the IRS could not make tax data electronically available to student aid applicants until July. Some participants said that besides not giving students sufficient time to plan for college costs, completing the FAFSA this late would cause many students who plan to enroll in the fall to be ineligible for aid from states and colleges. Acknowledging these limitations, Education officials said making the electronic transfer of IRS tax data to the FAFSA feasible for fall college applicants would likely require the use of income data that would be one year older than the information Education currently uses to determine financial aid eligibility. For example, under this scenario, an online applicant who completed the FAFSA in March 2009 with the intent of beginning college in August 2009 would be required to use income tax data from 2007. As of July 1, 2010, the Higher Education Opportunity Act authorizes the Secretary of Education to allow such older data to be used in calculating applicants’ aid eligibility. However, some participants expressed concern that by using older data— often referred to as prior-prior year data—there is an increased risk that the data may no longer reflect an applicant’s current economic need. For example, a college applicant could have a higher or lower income than they did two years prior to attending college. Currently, school financial aid officials can use professional judgment to change an applicant’s eligibility for aid upon an applicant’s request, if they determine that there are special circumstances. For example, applicants may request professional judgment if they think their financial aid award does not match their current economic need. Participants said that while professional judgment may be used to increase or decrease an applicant’s financial aid, it is unlikely that applicants with an improved economic status will ask their colleges to use professional judgment to decrease their student aid award. Therefore, some participants said they think that this will lead to increases in the numbers of applicants eligible for federal and state aid. However, one participant noted that although the implementation of using prior-prior year tax data would likely cause an initial increase in applicants qualifying for aid, the cost might level off in subsequent years. In addition, some participants expressed concern about the possible effects of using prior-prior year data on applicants who are not required to file income taxes. According to Education, about 6 percent of dependent applicants’ parents and about 13 percent of independent applicants who completed the FAFSA in the 2008 to 2009 academic year did not file taxes. One participant said that although it may be challenging for tax-filing applicants who submit paper forms to find tax forms from up to 2 years earlier, it would likely be particularly difficult for applicants who do not file taxes to provide information on their income from 2 years earlier. A few participants also expressed concern that Education had not offered a plan to simplify the FAFSA for applicants not required to file income taxes. Education’s recent proposals do not address how changes would affect these applicants. Education plans to begin providing applicants who both complete the FAFSA and enroll in college between January 1 and June 30, 2010, with the option of electronically transferring IRS data into the online FAFSA. Education officials said this pilot is feasible since spring 2010 applicants are required to use income tax information from 2008, which the IRS can make available electronically. Education officials stated that during the spring 2010 online FAFSA sessions, a question on the screen will ask if applicants would like to electronically retrieve their IRS tax data—or their parents’ data if they are dependents—to answer financial questions. If the applicants agree, they will be taken to an IRS Web site to confirm their identity and obtain tax information that they can electronically transfer into the appropriate FAFSA fields with a single push of a button. After piloting this electronic transfer of tax data with spring applicants, Education plans to make it available to all students who apply online for aid between July and December 2010. Some study group participants said Education’s plan to pilot the electronic transfer of IRS tax data to the FAFSA is a good initial step. Eventually, should Education elect to use its new statutory authority to allow the use of prior-prior year data, the option of electronically transferring IRS data into the online FAFSA could be made available to all applicants year-round. One participant said that additional piloting will be needed if Education ultimately decides to base student aid eligibility on prior-prior year tax data. Many participants noted ways in which other technology could facilitate additional improvements to the application process. One such possibility would be in better linking the electronic applications for federal and state aid. In 2001, Education began piloting a link between the online FAFSA and New York’s online state aid application. Currently, New York residents who submit the online FAFSA are immediately provided a link to the New York state aid Web site. Once New York applicants register and receive a personal identification number from the state student aid office, they may begin the online application process for state aid. The New York online application is automatically populated with FAFSA data that are sent electronically from Education. Applicants are asked to verify that the populated information is correct, and may be asked for additional information not collected by the FAFSA, such as their spouse’s social security number. Many participants said linking the FAFSA to other state application sites could prevent the possible negative effects on state aid programs of changing the federal need analysis formula by allowing states to ask additional questions that are not available on the FAFSA. One participant further noted that providing states with this option could make it feasible for Education to eliminate all FAFSA questions not needed to determine federal student aid eligibility without affecting the needs of states. However, a few participants expressed concern that such a change could lead to states adding a large number of additional questions on their applications, jeopardizing Education’s efforts to streamline the overall application process for students. Other participants said that the cost of setting up state online applications could create a barrier that would prevent some states from linking to the FAFSA. Beginning in January 2010, Education plans to offer this type of connection to all states, but the costs to states—and whether Education will provide financial assistance to states to facilitate this change—are not yet known. In addition to technological improvements, participants suggested that efforts to simplify the application should be accompanied by a strategy to increase public outreach efforts. For example, one study group participant suggested that Education should reach out to students from middle school through high school to help raise awareness about the affordability of college and the process of applying for financial aid. Some participants also suggested that Education provide monthly updates to states and colleges by zip code about how many students have completed the FAFSA. They said that this could help officials better target certain geographical areas with low FAFSA completion rates to raise awareness about student aid eligibility. Other participants suggested that Education send parents with children in middle school through high school annual estimates of their child’s current eligibility for student aid. Education has announced its intent to launch public outreach efforts that are designed to inform high school students about the availability of federal aid for college beginning in fall 2009. In creating an application process to distribute student aid, the federal government has had to consider multiple competing demands to promote college access and affordability: developing a formula precise enough to ensure that resources reach the target population, collecting enough information to assist states and institutions in administering their own aid programs, and making the application process easy to use and transparent for applicants. This last issue has proven particularly challenging, and the complexity of the application form and underlying formula has become a pain point for students and their families. The prospect that the application itself may discourage students from applying for aid—and perhaps to college—is especially troubling in light of current economic conditions, as postsecondary access and affordability become more challenging for some students. Our study group participants proposed various options for mitigating some of the complexity in the application process, and Education has proceeded with the early phases of its new plan for simplification, which includes a public outreach component. While most of these changes will result in streamlining the application process and will not affect eligibility for federal, state, and institutional aid, some— as is often true of policy and process changes—come with trade-offs. In particular, any changes to the formula used to compute eligibility may result in new winners and losers among aid applicants. Because the formula is complicated and no means of calculating eligibility—including the current method—is a perfect prediction of financial need, the effects of potential modifications on the pool of eligible applicants must be weighed against the goals of federal student aid. The issue to be considered is whether the benefit of simplifying the formula outweighs the potential loss in the precision of how aid is targeted, and depends not only on how great the overall change in the distribution of federal aid is, but on how much various types of applicants gain or lose. We provided a draft of this report to the Department of Education, Department of Treasury, and the Internal Revenue Service for review and comment. These agencies had no comments on the draft report. We are sending copies of this report to relevant congressional committees, the Secretaries of Education and Treasury, the Commissioner of Internal Revenue, and other interested parties. In addition, this report will also be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staffs have any additional questions about this report, please contact me at (202) 512-7215 or scottg@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in Appendix III. Simplifying the Federal Student Aid Application Process A Government Accountability Office Expert Panel Breakfast available in the room Approaches to shortening the form and changing the application process What are some approaches to shorten the Free Application for Federal Student Aid or otherwise make it less time-consuming to complete? How do states and institutions use the data collected on the current application form? What are the possible risks of simplifying the application form? Break Approaches to changing the statutory need analysis formula How could the statutory need analysis formula be changed to reduce the amount of financial information collected, without causing significant redistribution of federal grants and subsidized loans? What are the possible risks of simplifying the need analysis formula? Following any modifications to the need analysis formula, what are the best means of addressing the needs of states and institutions that rely on the federal application to administer their own aid program? Working lunch with presentations from other GAO engagements related to student financial aid issues Operationalizing changes to the application form and underlying formula What is the feasibility of the IRS providing individuals’ financial data directly to the Department of Education for the purposes of determining aid eligibility? How can changes to the application form and underlying formula be operationalized? Debra Prescott, Assistant Director; Rebecca Woiwode, Analyst-in-Charge; and James E. Lloyd III made significant contributions to this report in all facets of the work. In addition, Jean McSween and Luann Moy assisted in design; Sheila R. McCoy and Doreen Feldman provided legal support; Mike Brostek, Dave Lewis, and Ron Fecso lent subject matter expertise; Susannah Compton provided writing assistance; and James Bennett provided help with graphics.

Federal student aid is intended to play an integral part in fulfilling the promise of greater academic access and success for less affluent students. However, many experts have expressed concern about the length and complexity of the Free Application for Federal Student Aid (FAFSA) and the statutory need analysis formula used to determine aid eligibility. The Higher Education Opportunity Act required GAO to form a study group to examine options and implications in simplifying the financial aid process. The study group focused on (1) identifying ways to shorten the FAFSA and make it less burdensome to complete, (2) identifying changes to the statutory need analysis formula that would reduce the amount of financial information required by the FAFSA without causing significant redistribution of federal and state student aid, and (3) determining how any changes to the FAFSA and the statutory need analysis formula could be implemented. To address these questions we convened an expert panel on May 7, 2009, and conducted additional interviews with experts. This summary captures the ideas and themes that emerged at the panel and during interviews. It does not necessarily represent the views of GAO or of the organizations whose representatives participated in the study group. Study group participants said using federal income tax data that the government already collects and revising the form could shorten the application process, making it easier on students and their families. Many participants proposed that relevant federal income tax data be directly transferred to the appropriate answer fields on each applicant's online FAFSA, an approach that the Department of Education (Education) plans to pilot for some applicants in January 2010. Such a change could decrease the amount and complexity of some of the financial questions on the application. In addition, many participants proposed changes to the design and contents of the form to clarify and streamline the application. Education has recently taken steps to shorten and reorganize the online form and has plans for further improvements. Participants said changing the federal formula to reduce required financial information would ease applicants' burden, but such a shift would likely result in some change in the distribution of aid. Many study group participants supported changing the need analysis formula to rely solely on a family's income and number of tax exemptions to determine aid eligibility. These changes would greatly reduce the number of complicated financial questions on the FAFSA. However, reducing the amount of financial information collected could change the distribution of federal, state, and institutional aid, prompting some concern about this approach. Education's recent legislative proposal to limit the formula to financial information available through tax forms would eliminate 26 financial questions, including those on assets. Participants said technology and public outreach efforts could improve the federal student aid application process, but successful implementation of changes hinges on the ability of federal and state agencies to address several challenges. While it is feasible to electronically transfer tax data directly from the Internal Revenue Service (IRS) to the FAFSA by using income data one year older than what is currently required, participants expressed some concern about the potential implications of such a change. Specifically, using older tax data might result in increased aid eligibility for some applicants whose data may not reflect their current economic needs. In addition, it may be more difficult for applicants who do not file taxes to provide sufficient documentation of their income from two years earlier. Education and the IRS have begun developing a plan to allow some applicants to electronically access their tax data when they apply for aid online. However, because taxpayers can submit their data as late as April 15, these data will not be available in time to accommodate most aid applicants. Many participants also called for linking state aid Web sites to the online federal application to mitigate the potential effects of federal formula changes on state aid. Education plans to offer this option to states in January 2010. In addition, participants said that efforts to simplify the application process should be accompanied by a public outreach strategy aimed at increasing knowledge of the availability of federal student aid. Education plans to undertake a public outreach campaign beginning in fall 2009.

The Department of Defense understands that a comprehensive transformation of its logistics structures, processes, and supporting information systems is necessary to improve its customer services and reduce the cost of support. To lead this transformation, the Department established the Logistics Reform Senior Steering Group—comprised of senior officials from the Joint Staff, the military services, the Defense Logistics Agency, and the U.S. Transportation Command—to develop its Fiscal Year 2000 Logistics Strategic Plan. This plan was designed during a series of workshops that focused the collective attention and resources that the Department believed were necessary for achieving its key objectives. This top-leadership effort concentrated on developing a common mission, vision, and associated performance goals. To meet the Department of Defense goal to have a successful logistics transformation, the Logistics Reform Senior Steering Group prescribed the following specific “success criteria.” Optimize cycle times—acquisition, supply, maintenance, transportation, and distribution. Manage the total life-cycle through integration of acquisition and logistics processes. Meet deployment and sustainment requirements across the full spectrum of military operations. Guarantee joint total asset visibility through fully integrated, secure information systems, with asset visibility being the capability for users to view information on the identity and status of Defense material assets and, in some cases, complete a business transaction using that information. Meet or exceed the Department of Defense logistics metrics and cost- reduction goals. To address the above criteria, the Logistics Strategic Plan includes six broad objectives developed by the Logistics Reform Senior Steering Group to assist logistics managers in planning and executing the Department’s priority initiatives for transforming logistics. The objectives are as follows: Optimize support to the warfighter. Improve strategic mobility to meet warfighter requirements. Implement customer wait time as the Department-wide logistics metric. Fully implement total asset visibility across the Department of Defense. Reengineer and modernize applicable logistics processes and systems. Minimize logistics costs while meeting warfighter requirements. The Department’s Fiscal Year 2000 Logistics Strategic Plan directed the services and Defense commands to develop implementing plans that reflect the vision, objectives, and metrics of the Department-wide plan. The plan also specified that the implementing strategies and related plans are governed in content and format by the provisions of the Government Performance and Results Act of 1993 (P.L. 103-62), commonly referred to as the Results Act. As such, these plans should include detailed actions, performance measures, completion dates, and resource requirements. To support the implementation of the Logistics Strategic Plan, the Deputy Secretary of Defense directed the military services, the Defense Logistics Agency, and the U.S. Transportation Command in a March 23, 2000, directive entitled the Department of Defense Reform Initiative #54, to establish logistics transformation plans for submission to the Office of the Secretary of Defense. This directive also provided guidance for the preparation of these plans, which were intended to serve as vehicles for aligning the military component initiatives, documenting their approach for achieving the objectives in the Department’s Logistics Strategic Plan, and obtaining resources. Appendix I provides an overview of the logistics reengineering initiatives that each of the components is individually pursuing as part of its transformation plan. We reported in December 1996 that the Department’s 1994 version of its Logistics Strategic Plan contained similar weaknesses to those outlined in this report. Specifically, we reported that the 1994 version of the plan did not adequately (1) link the action plans to resource requirements, (2) link the services’ and the Defense Logistics Agency’s plans to the strategic plan, or (3) identify interim approaches that could be developed and implemented when milestones of a priority strategy had been extended. Other related GAO reports and testimonies are listed on the last page of this report. While the Department-wide Logistics Strategic Plan outlines six key objectives and a general time frame for implementation, it is not sufficiently comprehensive and does not provide an adequate framework for guiding the military services’, the Defense Logistics Agency’s, and the U.S. Transportation Command’s logistics reengineering initiatives. Specifically, the plan does not address the logistics lifecycle process from acquisition through support and system disposal. In addition, the Department’s plan does not specify how the Department will be organized in the future to fulfill the logistics requirements that will be needed to support the National Military Strategy or how the Department will eventually attain a new logistics structure. The Department’s plan identifies that the logistics vision is that, by fiscal year 2006, the joint logistics process will be a highly efficient, integrated system that ensures required support to the warfighter. This vision, however, does not identify the logistics requirements the Department will have to fulfill, how it will be organized to fulfill these requirements, or who will be responsible for providing specific types of logistics support. Furthermore, while the Department will likely face further changes in size and structure in the near future, its strategic plan has not identified the logistics facilities or personnel it will need to support future logistics requirements and has not specified a process for determining what resources it will need. Without addressing all logistics functions, as well as these facilities and personnel requirements through its strategic planning process, the Department will be unable to identify the resources (i.e., maintenance facilities, storage capacity, transportation assets, and depot personnel) it will need to support the Department’s future logistics concepts. In assessing the Logistics Strategic Plan, we also reviewed the relationship between the Department’s logistics architecture initiative and the strategic plan. The architecture initiative is intended to provide a strategic framework to synchronize logistics improvements for the years 2010 and beyond and to define a Department-wide logistics organization. In a previous report on the Department of Defense’s reengineering initiatives, we identified that in October 1999, the Department established the Office of the Assistant Deputy Under Secretary of Defense for Logistics Architecture to oversee defense logistics reengineering efforts. This office was given responsibility for designing a logistics system for business processes, physical infrastructure, and information technologies, as well as for defining the services’ responsibilities within the new logistics architecture. With the assistance of a contractor, the Office began work on an integrated logistics architecture that was intended to facilitate the implementation of reengineered logistics support processes and procedures. To some extent, this logistics concept was intended to address the logistics facilities and personnel requirements. This project has progressed slowly and it is not clear how it will fit into the Department’s Logistics Strategic Plan. A planned report on the contractor study has not yet been completed. Further, according to defense officials, this planning effort may be refocused to incorporate the vision of the Department’s new leadership regarding how future logistics planning should proceed. The title of the office overseeing this effort has been changed to the Assistant Deputy Undersecretary of Defense for Logistics Plans and Programs and its stated mission has been expanded to cover the entire future logistics environment. The Office is pursuing an approved long-term logistics planning strategy and plans to identify specific operational efforts to undertake to move the Department along toward achieving that strategy. Nonetheless, service planning officials expressed concerns regarding the realism of this study effort and the extent to which the proposed concept would work in wartime. They also questioned how this effort fits in with other planning initiatives, such as the Logistics Strategic Plan. Therefore, it is too early for us to determine whether this new concept for logistics support planning is an effective one, whether it will be accepted throughout the Department, and whether it will be feasible for future implementation. While each of the military services, the Defense Logistics Agency, and the U.S. Transportation Command prepared implementation plans in the form of transformation plans and other documents to support the Logistics Strategic Plan, these plans will not likely result in overall improvements to the economy and efficiency of logistics activities. We found that these plans (1) are not always consistent with the actions in the Defense-wide plan, (2) are not directly related to the Defense-wide plan or to each other, and (3) do not contain some key management elements as outlined in the Results Act, such as performance measures and specific milestones. According to officials in the Office of the Secretary of Defense, the components’ plans were generally a summation of ongoing initiatives reformatted to support the Department-wide generic objectives and did not include any new initiatives. According to Air Force officials, one contributing factor to this lack of any new initiatives was that the approximately 3 months the components were given to prepare their implementation plans was insufficient given the scope of the task. The military services, the Defense Logistics Agency, and the U.S. Transportation Command have initiated implementation actions to address the Department’s six objectives set forth in the Department-wide strategic plan. Tables 2 through 7 outline the six strategic objectives and the stated performance measures and provide a summary of the weaknesses we identified in the corresponding parts of the components’ implementation plans. A more detailed discussion of our findings related to each objective follows the individual tables. While we found that each component, with the exception of the U.S. Transportation Command, addressed this objective, mission capable rates are not defined and/or measured the same among the components. For example, to measure mission capable rates, the Army plans to use the percent of equipment that is fully mission capable and the revised readiness report for tracking mission capable rates. These readiness reports are embedded in the development of the Army’s Global Combat Support System. However, this system has not been fielded Army-wide. The Air Force, on the other hand, uses historical and forecasted aggregate aircraft mission capable rates to measure readiness and has implemented four initiatives to improve logistical support to the warfighter—aircraft spare parts availability, informational advancements, reengineered logistical support concepts, and supply chain management. Many of the Marine Corps initiatives to support this objective are still under development. The Navy included a sample of 12 ongoing initiatives to support this objective. These initiatives, however, are not directly linked to defining or measuring mission capable rates. For example, the Navy’s One- Touch Support initiative allows a customer to use the Internet to access the supply system, identify the location and status of stock, and input requisitions. The Navy has predicted that the upgraded version will be an electronic single point of entry that will link Navy users via the Internet to over 20,000 suppliers and manufacturers. The military components have not fully defined specific performance measures that can be used to monitor the implementation of actions that will meet the Department’s objective. The measures are generally not specified or are under development. For example, the Army plans to utilize its revised readiness reports when available. However, these reports will be developed at a later date based on an ongoing Army War College study being conducted at the request of the Army’s Deputy Chief of Staff for Logistics. The Air Force states that its current best measure is logistics response-time days. However, the Department is transitioning from the logistics response-time measurement to customer wait time. The Marine Corps developed draft metrics during a working group session that was completed in May 2000. The operational performance of the Navy’s One Touch Support initiative is measured by the number of “logins” by the customer and the number of “hits” experienced by the program. The military components have not established the necessary metrics to measure progress toward a successful implementation of this objective or defined a strategy that recognizes the interrelationship of the components’ initiatives. With regard to the components’ initiatives, the Army and Air Force, for example, are relying on the future outcome of the Mobility Requirements Study. The Department believes this study is the absolute key to determining its strategic mobility requirements. However, the expected completion date is not until fiscal year 2005. The Army is also relying on the future deployment of its Transportation Coordinator Automated Information for Movement System II. This system is intended to provide automation support to warfighters during deployment, sustainment, and redeployment operations and to provide source data to strategic command and control systems to increase the visibility of in- transit personnel and items during those operations. However, the Army has encountered delays in the development of this system and fielding will not be completed until September 2007. We reported in February 2000that a critical $22.7-million software requirement was unfunded and could further delay the project’s completion. The Army’s Transformation Plan indicates that some of these requirements are still unfunded. The Defense Logistics Agency states in its transformation plan that objective two does not apply to them. We agree with their position since this agency’s mission does not directly involve mobility issues. While the U.S. Transportation Command has initiated many actions, the successful implementation of these actions has been hindered by inadequacies in the Department’s transportation systems. For example, in order for the U.S. Transportation Command to execute its global mission in support of a National Military Strategy, a healthy and robust Defense transportation system infrastructure, including all mobility assets and critical nodes (i.e. installations, depots, rail/highway networks, air bases including en route bases, and seaports) is required worldwide. The continued decline of the U.S. flag merchant marine fleet and the maritime industry may affect the Command’s ability to meet peacetime and wartime Defense requirements. Therefore, the success of the Command’s initiatives depends largely on the services’ actions. We found that in most cases the military services, the Defense Logistics Agency, and the U.S. Transportation Command have not identified performance measures to indicate a successful implementation of objective two. The proposed measures that were outlined simply define a goal and do not include baseline data or interim steps for implementing the initiatives necessary to meet this objective. The Department considers customer wait time, in general, to be the total elapsed time between issuance of a customer order and satisfaction of that order. However, the capability to capture and to report customer wait time is still under development; and the Department is relying on the individual components to develop a more specific definition of, and a process for measuring, customer wait time. Further, the Department has not identified how it plans to integrate the military components’ efforts. The Logistics Strategic Plan does not define how customer wait time will be calculated or provide information on expected outcomes, such as to what extent customer wait time will be reduced. The capability to capture and report customer wait time is still under development. The Marine Corps, for example, uses retail and wholesale order ship time for all classes of supply. According to the Marine Corps, its order ship time measurement is identical to the Department’s logistics metric of customer wait time. The Marine Corps believes that implementing customer wait time will not require significant changes in its business process or the way response time is measured. According to the Army, its Single Stock Fund initiative will provide the Army with its initial capability to collect customer wait time data. Initiatives included in the plans submitted by the Defense Logistics Agency and the U.S. Transportation Command did not relate to establishing customer wait time. In addition, the performance measures specified in the other component’s plans are primarily stated objectives and are too broad to measure whether a successful implementation will occur. The military components are focusing on achieving 100-percent asset visibility by fiscal year 2006 through the use of automatic identification technology and automated information systems. However, some of the systems that the Department is relying on to achieve that 100-percent visibility may not be fully developed in time to support that schedule. For example, the Army’s Global Combat Support System is expected to substantially improve asset visibility for the warfighter and the logistics support community. This system is being developed in three tiers concurrently—retail, wholesale, and joint integration—but the expected completion date is also fiscal year 2006. The Air Force has tasked the Air Force Audit Agency to perform a series of reviews to determine the continued need for Air Force manual accounts that are targeted for total asset visibility by fiscal year 2004. We noted in our November 2000 reportthat the Air Force plans to complete these reviews by December 2001. Air Force audit officials stated that the scope of this work has recently changed and that they will likely not complete the work by December 2001. In many cases, the performance measures will be based on the progress made in achieving the components’ individual initiatives to have 100- percent visibility by fiscal year 2006. For example, the Army has outlined several initiatives to meet this goal; but in many cases, no stated performance measures were given. The Defense Logistics Agency states that its sample inventory accuracy for all products in each fiscal year will be greater than or equal to 95 percent and that the depot location accuracy for all products will be greater than or equal to 99 percent. The U.S. Transportation Command did not define a performance measure. With the exception of the U.S. Transportation Command, we found that the components have ongoing initiatives to address objective five. However, the components’ individual initiatives are generally not directly linked to the Department’s Logistics Strategic Plan or to each other. It appears that the components’ initiatives will have limited, if any, interface. For example, the Navy’s Transformation Plan outlined 16 objectives to represent a sample of the initiatives that are ongoing throughout the Navy, but it did not indicate how these systems will interface with the other military components. The Air Force has established a Logistics Transformation Team specifically chartered to reengineer overarching Air Force logistics system processes and identify opportunities to increase performance and optimize costs. This team will also develop plans and schedules that will outline the approach for identifying process enhancements for the logistics system and describe the key activities required. However, the plan does not state how the different Air Force systems will be linked with similar systems in the rest of the Department. The Marine Corps has initiated a program called Integrated Logistics Capability to reengineer and modernize its logistics processes and system. The Army’s stated solution to logistics software modernization includes both its Global Combat Support System, which is scheduled to be completed in fiscal year 2006, and its web-based logistics system. The web- based logistics system will transform existing serial, sequenced, batch processing into a real-time set of logistics management systems that will connect enterprises with customers, suppliers, and transportation providers worldwide. The Defense Logistics Agency plans to implement commercial business-based systems and practices. For objective five, we also found that the military components have not defined adequate performance measures. In most cases, the performance measures will be determined once the initiatives have matured. The Marine Corps states that its detailed schedule for key activities and milestones will identify performance measures and will be finalized by September 2001. For the Navy’s 16 initiatives, either no firm decision has been reached for the performance measures; or it will be based on future progress in achieving completion of the Navy’s stated objectives. Under objective six, the Department is relying on the individual components to reduce the overall costs of logistics support for selected weapon systems. With the exception of the U.S. Transportation Command, which did not address this objective, the components have each outlined initiatives to meet this objective. However, the components did not identify which weapon systems they are focusing on to reduce logistics costs. For example, the Marine Corps has established a Total Ownership Cost Integrated Product Team to identify its total ownership cost methodology and to obtain the necessary data to compute the costs and minimize the logistics costs. The Air Force has a cost savings modernization initiative to identify and highlight cost reduction opportunities and to ensure that the opportunities receive due consideration for funding. The Army initiatives are ongoing, and it plans to reduce the total ownership costs for its weapons. In most cases, the components’ performance measures were not defined or were under development. The goal is to reduce costs by fiscal year 2006. While the Department of Defense has taken a number of steps toward improving the economy and efficiency of its logistics support systems, its Logistics Strategic plan lacks a realistic and practicable overarching logistics framework to effectively guide the Defense components’ logistics planning and implementation efforts. In the absence of a clearly defined, Department-wide logistics strategy, the Department is unable to develop a Defense-wide logistics structure that is both economical and efficient and supports the needs of the warfighter. Furthermore, because the Department’s plan and the components’ implementation plans lack a comprehensive strategy that addresses the logistics life-cycle process from acquisition through disposal, the Department cannot evaluate and prioritize the initiatives on a Department-wide basis. The Department also faces an increased risk that the Defense components will continue to develop individual initiatives that may not be compatible with each other or may have differing objectives. Furthermore, without detailed performance measures and milestones, the Department will be unable to monitor the progress of its logistics initiatives or ensure that the components’ individual initiatives are contributing to meeting the Department’s overall objectives. To provide the military services, the Defense Logistics Agency, and the U.S. Transportation Command with a framework for developing a Department-wide approach to logistics reengineering, we recommend that the Secretary of Defense direct the Under Secretary of Defense for Acquisition, Technology, and Logistics to revise the Department-wide Logistics Strategic Plan to provide for an overarching logistics strategy that will guide the components’ logistics planning efforts. Among other things, this logistics strategy should specify a comprehensive approach that addresses the logistics life-cycle process from acquisition through support and system disposal, including the manner in which logistics is to be considered in the system and equipment acquisition process and how key support activities such as procurement, transportation, storage, maintenance and disposal will be accomplished; identify the logistics requirements the Department will have to fulfill, how it will be organized to fulfill these requirements, and who will be responsible for providing specific types of logistics support; and identify the numbers and types of logistics facilities and personnel the Department will need to support future logistics requirements. We also recommend that the Under Secretary of Defense for Acquisition, Technology and Logistics establish a mechanism for monitoring the extent to which the components are implementing the Department’s Logistics Strategic Plan. Specifically, the Under Secretary of Defense for Acquisition, Technology, and Logistics should monitor the extent to which the components’ implementation plans are (1) consistent with the Department-wide plan, (2) directly related to the Department-wide plan and to each other, and (3) contain appropriate key management elements, such as performance measures and specific milestones. We received written comments from the Department of Defense, which are reprinted in appendix III. The Department generally agreed with the report and our recommendations. The Department agreed with our recommendation that the Under Secretary of Defense for Acquisition, Technology, and Logistics should revise the Department-wide Logistics Strategic Plan to provide for an overarching logistics strategy that will guide the components’ logistics planning efforts. The Department specifically stated in its comments that the Deputy Under Secretary of Defense (Logistics and Materiel Readiness) will lead the Department in developing a comprehensive Logistics Strategic Plan that incorporates our recommendation, the results of the recent Quadrennial Defense Review, and the logistics requirements of the new National Defense Strategy. The Department also agreed with our recommendation that the Under Secretary of Defense for Acquisition, Technology, and Logistics should establish a mechanism for monitoring the extent to which the components are implementing the Department’s Logistics Strategic Plan. Specifically, the Department stated in its comments that it will establish metrics to measure performance of the components’ implementation of the Logistics Strategic Plan, within the context of the normal program and budget review cycles. We are sending copies of this report to the appropriate congressional committees; the Secretary of Defense; the Secretaries of the Army, the Navy, and the Air Force and the Commandant of the Marine Corps; the Director of the Defense Logistics Agency; the Commander-in-Chief of the U.S. Transportation Command; and the Director of the Office of Management and Budget. We will also make it available at www.gao.gov and to others. If you or your staff have any questions regarding this report, please contact me at (202) 512-8412. Key contributors to this report were Julia Denman, David Schmitt, Patricia Albritton, and Marjorie J. Hunt. This appendix provides our analysis of the logistics transformation plans and other implementation plans prepared by the military services, the Defense Logistics Agency, and the U.S. Transportation Command as they relate to the Department of Defense’s Logistics Fiscal Year 2000 Strategic Plan. The overall objective of the Department-wide plan was to focus the collective attention and resources necessary for achieving the key objectives required to improve the Department’s logistics support to the warfighter. The Logistics Strategic Plan includes the following six basic objectives: (1) optimize support to the warfighter, (2) improve strategic mobility to meet warfighter requirements, (3) implement customer wait time as the Department-wide logistics metric, (4) fully implement total asset visibility across the Department of Defense, (5) reengineer and modernize applicable logistics processes and systems, and (6) minimize logistics costs while meeting warfighter requirements. To support the implementation of the Department’s Logistics Strategic Plan, the Department of Defense Reform Initiative 54, dated March 23, 2000, requires the military services, the Defense Logistics Agency, and the U.S. Transportation Command to develop logistics transformation plans to relate the 400 different service-sponsored logistics reengineering initiatives to the Department-wide Logistics Strategic Plan. These plans serve as the primary vehicles for aligning the military component initiatives, obtaining resources, and documenting the approach for achieving the Logistics Strategic Plan goals and objectives. With the exception of the U.S. Transportation Command, each component prepared a transformation plan based on high-level guidance provided by the Department to address the six strategic objectives. The U.S. Transportation Command’s plan primarily focused on objective 2–improve strategic mobility to meet warfighter requirements–and, to a lesser extent, objective 4–fully implement joint total asset visibility. The Defense Logistics Agency submitted an annotated agency performance contract that specifically addressed its implementation plan for the Department’s six objectives. The following is both a brief overview of the logistics reengineering initiatives that each of the military services, the U.S. Transportation Command, and the Defense Logistics Agency are individually pursuing and our analyses of how these initiatives related to the Office of the Secretary of Defense’s Fiscal Year 2000 Logistics Strategic Plan. The components’ implementation plans are embodied in their transformation plans and other related documents. The Navy’s focus on High Yield Logistics began several years ago and includes 46 ongoing initiatives in its logistics transformation plan. According to Navy officials, the Navy’s vision of logistics transformation is captured in its High Yield Logistics Strategy. However, the Navy is using its logistics transformation plan as the tool to meet the challenge and address all aspects of logistics transformation required under the Department of Defense’s Reform Initiative #54. These initiatives, however, do not directly link with the other components’ initiatives. For example, the Navy’s four ongoing initiatives to support how it plans to implement customer wait time, objective three, do not relate to the other components’ initiatives regarding customer wait time. While the Navy was scheduled to begin collecting data to define customer wait time after the Logistics Reform Senior Steering Group met in June 2000, this initiative is still in process. Additionally, the Navy has implemented a “Response to Failure”1 metric, which the Navy is synonymous with customer wait time. The Response to Failure metric has been prototyped and developed for use by cognizant fleet and Headquarters staff. However, the specific goals for Response to Failure man-hours by fiscal year are under development. The Navy also plans to utilize the Logistics E-Business Concept of operations to integrate e-business interfaces, applications and data environments. The Navy has not developed, however, any performance measures to monitor the success of implementing these initiatives. The Navy defines “Response to Failure” as the total time that maintenance personnel wait for material. fiscal year 2002. However, we reported in November 20002 that the Army has encountered delays in developing this system and does not expect to complete fielding it until September 2007, which may be further delayed due to a reported critical $22.7-million unfunded software development requirement. The Army’s Transformation Plan indicates that some of these requirements are still unfunded. The Transportation Coordinator Automated Information for Movement System II system will also have a direct link to the Global Transportation Network that will be relying on more than 20 Defense automated logistics systems to provide data that our prior reports and Department of Defense audit reports have found inaccurate. With regard to reengineering and modernizing applicable logistics processes/systems, the Navy outlined 16 objectives to represent a sample of ongoing initiatives, such as Enterprise Resource Planning, Technical Publications Lifecycle processes, and update and revise Military Sealift Command Logistics Systems Procedures. However, it is not clear how these initiatives will correlate with the initiatives of other components. GAO-01-30, Nov. 15, 2000. its retail and wholesale order ship time measurement is identical to the Department-wide customer wait time logistics metric and will not require significant changes in its process or the way response time is measured. However, it will require new tools to capture customer wait time rather than order ship time data. The Navy is establishing baseline data based on its current performance. To fully implement objective four, total asset visibility, the Marine Corps plans to develop and field total asset visibility/in-transit visibility systems along with the automatic identification technology to support the identification and processing of materiel within the supply and distribution pipeline. During 1998, the Marine corps initiated a program—Integrated Logistics Capability—to reengineer and modernize its logistics processes and systems. The Army plans to synchronize its logistics transformation with the overall Army Force transformation, which is expected to be completed by about 2010. The Army Force transformation is a three-phased approach to develop a consistent, Army-wide force structure. According to Army officials, while there are many Army-wide supporting plans, the plans most critical to the success of logistics transformation are its Strategic Logistics Plan and Combat Support/Combat Service Support Transformation Campaign Plan. The Army’s Strategic Plan includes initiatives from all elements of the Army’s logistics community. In order to define customer wait time, the Army is relying on an ongoing effort—the Single Stock Fund initiative—that is intended to provide the Army with the initial capability to collect customer wait time data. According to the Army, the Global Combat Support System will be the platform for the customer wait time data collection and dissemination. However, this system is being developed in three tiers and the expected completion date for fielding this system is not until fiscal year 2006. To fully implement Total Asset Visibility, the Army plans to achieve this objective through the use of its automatic identification technology/automated information systems and transformed business practices. The Army is also relying on the Global Combat Support System that is being used to define customer wait time as the solution to the Army’s logistics software modernization but as mentioned earlier, the fielding of this system is not scheduled for completion until fiscal year 2006. The Air Force integrated several of its ongoing initiatives into its Transformation Plan to meet the Department of Defense’s objectives. However, it is uncertain as to how the actions the Air Force has taken will achieve a successful overarching framework. For example, the Air Force plans to measure customer wait time starting with the logistics pipeline documentation of a requisition by a customer to receipt of the asset by the customer to include retail transactions. However, it does not indicate how this action will link with the other components or how this objective will be assessed, since the metrics have not been defined. To implement the Joint Total Asset Visibility concept, the Air Force plans to rely on a series of ongoing studies being conducted by the Air Force Audit Agency, in response to section 349 of the National Defense Authorization Act for Fiscal Year 1999, to assess its policies, procedures, and business practices regarding controls over assets. Additionally, the Air Force has established a Logistics Transformation Team specifically designed to create an overarching logistics system process and identify opportunities to increase performance and optimize costs. The Air Force is relying on this team to develop a set of plans and schedules that will outline the approach for identifying logistics system process enhancements. However, none of the Air Force actions indicates a successful coordination with the other components. To meet the Department of Defense’s goals, the Defense Logistics Agency annotated its Fiscal Years 2001 through 2005 Performance Contract sorted by the objectives outlined in Department’s Logistics Strategic Plan. According to officials, the agency’s basic mission, operations, authority, or reporting chain was not altered in any way. However, according to these officials, only two of the six objectives were in line with the Defense Logistics Agency’s operations and easy to address. These were objective three–implementing customer wait time–and objective four–total asset visibility. Officials stated that these two objectives were easy to support due to the way the Defense Logistics Agency does its business. Additionally, the Defense Logistics Agency plans to develop a “balanced scorecard” approach to measure its strategic performance. According to officials, this approach will allow the agency to more closely align its performance indicators used to measure its strategic goals and objectives that support logistics transformation. However, the agency’s actions defined do not show a correlation to the Department’s overall objective. For example, to implement customer wait time, the Defense Logistics Agency states that it will consistently provide responsive, best-value supplies and services to its customers. Specifically, the logistics response time reliability for “supply–non-energy materials” will improve over the program period to reflect greater percentages of requisitions processed within shorter timeframes. To fully implement objective four, total asset visibility, the Defense Logistics Agency plans to sample inventory accuracy for all products, as determined by a statistical sampling in each fiscal year. The sampling accuracy is expected to be greater than or equal to 95 percent. The agency plans to shift to commercial practices for its hardware, energy, and troop support items as a way to reengineer/modernize its applicable logistics processes/systems. The U.S. Transportation Command used its 1999 Strategic Plan to implement the Department of Defense’s Logistics Strategic Plan and develop its Logistics Transformation Plan. The Command’s plan consists of two parts: (1) the Strategic Guidance, which identifies the Command’s mission, vision and long-term goals for executing each of its five core processes, and the Strategic Objectives, which must be accomplished to ensure the vision is met, and (2) the Corporate Resource Plan, which is intended to link the long-term goals and objectives, strategies and various resources needed to accomplish the goals and objectives, and an avenue to evaluate, establish, and revise strategic goals and objectives. The Command incorporated a table in its transformation plan to depict the Command’s strategic objectives and how they support the Department of Defense’s Logistics Strategic Plan objectives. Of the six objectives, the Command primarily focused on two–objective two–improving strategic mobility to meet warfighter requirements–and to a lesser extent, how it plans to fully implement objective four–total asset visibility. However, the Command has not identified in its plan how the actions it plans to implement will be coordinated with those of the other components. To assess the Department of Defense’s logistics strategic planning process, we reviewed the Department’s August 1999 Logistics Strategic Plan and the various logistics transformation plans prepared by the military services, the Defense Logistics Agency, and the U.S. Transportation Command. We met with officials in the Office of the Secretary of Defense, each of the military services, the Defense Logistics Agency, and the U.S. Transportation Command to discuss these various planning documents and the Department’s planning approach. We also relied on our prior work regarding logistics planning and reengineering. Our review of the logistics strategic planning process focused on the Department’s Logistics Strategic Plan because the purpose of this plan was to provide an overall Defense-wide corporate direction for accomplishing the Department’s logistics mission. To determine whether this plan provides an adequate overarching logistics strategy to guide logistics reengineering initiatives, we analyzed the contents of the plan and the extent to which it contained the elements we believe are necessary for focusing current and future initiatives. Specifically, we assessed whether the plan included an appropriate definition of (1) the future role of logistics in supporting the operational forces; (2) how the Department should be organized and staffed to fulfill its logistics mission; and (3) the types of capabilities, facilities, and systems that will likely be needed to meet future logistics requirements. We also reviewed the status of the Department’s long-range logistics architecture initiative because the intent of this initiative is to identify the framework for logistics support for the years 2010 and beyond. Specifically, we met with officials in the Office of the Deputy Under Secretary of Defense for Logistics Architecture and of the Science Applications International Corporation to discuss the objectives, status, and future plans for this project. These officials provided us with the basic principles and long-range vision of the logistics architecture initiative and discussed with us the current status and proposed milestones for this project. To determine whether the components’ related implementation plans are likely to result in overall improvements to the economy and efficiency of logistics activities, we reviewed the plans, comparing the objectives and planned actions outlined in each of the various transformation plans and other logistics reengineering initiatives. Our review concentrated on determining whether there was a direct linkage among the various plans and initiatives and whether the objectives and actions outlined in these documents represented a coordinated approach to logistics reengineering on a department-wide basis. Specifically, we reviewed the components’ implementation plans to determine the extent to which these plans are (1) consistent with the Defense-wide plan, (2) directly related to the Defense- wide plan and to each other, and (3) contain appropriate key management elements. We also reviewed the various plans to determine whether they contained an appropriate management framework for implementation. For these analyses, we used the requirements of the Government Performance and Results Act, commonly referred to as the Results Act, as a model for the types of information the plans should contain. We compared the contents of the plans and the requirements of the Results Act. Additionally, we reviewed the plans in terms of outcome-oriented Results Act principles and identified areas in which they could be improved to achieve successful implementation. Congressional reports and administrative guidance regarding the Results Act indicate that activities such as strategic planning should be subject to the outcome-oriented principles of the Results Act. We did not assess the merits of the Department’s proposed actions or the likelihood of success for these actions. We conducted our review from January to May 2001 in accordance with generally accepted government auditing standards. Actions Needed to Overcome Capability Gaps in the Public Depot System (GAO-01-612, Oct. 2001). Defense Maintenance: Sustaining Readiness Support Capabilities Requires a Comprehensive Plan (GAO-01-533T, Mar. 23, 2001). Major Management Challenges and Program Risks: Department of Defense (GAO-01-244, Jan. 2001). Defense Inventory: Implementation Plans to Enhance Controls Over Shipped Items Can Be Improved (GAO-01-30, Nov. 15, 2000). Defense Logistics: Actions Needed to Enhance Success of Reengineering Initiatives (GAO/NSIAD-00-89, June 23, 2000). Defense Inventory: Plan to Improve Management of Shipped Inventory Should Be Strengthened (GAO/NSIAD-00-39, Feb. 22, 2000). Defense Inventory: DOD Could Improve Total Asset Visibility Initiative With Results Act Framework (GAO/NSIAD-99-40, Apr. 12, 1999). Performance and Accountability Series: Major Management Challenges and Program Risks – Department of Defense (GAO/OCG-99-4, Jan. 1999). High-Risk Series: An Update (GAO/HR-99-1, Jan. 1999). Defense Inventory Management: Problems, Progress, and Additional Actions Needed (GAO/T-NSIAD-97-109, Mar. 20, 1997). High-Risk Series: Defense Inventory Management (GAO/HR-97-5, Feb. 1997).

The Department of Defense's (DOD) Logistics Strategic Plan is not comprehensive enough and does not provide an adequate overall logistics strategy to effectively guide the defense components' logistics plans. The military services, the Defense Logistics Agency, and the U.S. Transportation Command each developed separate logistics transformation and other implementation plans to support the Department-wide Logistics Strategic Plan. However, these plans also have weaknesses and are not likely to improve the overall economy, efficiency, and effectiveness of logistics activities.

Under the Stafford Act, when a major natural catastrophe, fire, flood, or explosion occurs that is beyond the capabilities of a state and local government response, the President may declare that a major disaster exists. This declaration activates the federal response plan for the delivery of federal disaster assistance. The response plan is an agreement signed by 27 federal departments and agencies, including the American Red Cross. Under the Stafford Act, FEMA is responsible for coordinating both the federal and private response efforts. President Jimmy Carter established FEMA in 1978 to consolidate and coordinate emergency management functions in one location, addressing concerns about the lack of a coordinated federal approach to disaster relief. FEMA most recently redesigned its public assistance program in 1998. The federal assistance coordinated by FEMA is designed to supplement the efforts and available resources of state and local governments and voluntary relief organizations. While FEMA had the lead in coordinating the federal response to the attacks on NYC, other federal agencies, including the Department of Transportation (DOT), the Small Business Administration (SBA), the U.S. Army Corps of Engineers, and the Environmental Protection Agency (EPA) also provided significant assistance. The disaster declaration from the President triggers not only a role for FEMA as coordinator of the federal emergency response plan, but also a role in delivering assistance through several programs it administers. These programs include individual assistance to victims affected by a disaster and hazard mitigation funds to state and local governments to take steps to prevent future disasters. However, FEMA’s public assistance program is typically its largest disaster assistance effort. It is designed to provide grants to eligible state and local government agencies and specific types of private nonprofit organizations that provide services of a governmental nature, such as utilities, fire departments, emergency and medical facilities, and educational institutions, to help cover costs of emergency response efforts and work associated with recovering from the disaster. According to FEMA regulations, work eligible for public assistance must be to repair damage that occurred as a result of a declared event, located within an area declared by the President as a disaster area, and the legal responsibility of an eligible applicant. The Stafford Act sets the federal share for the public assistance program at no less than 75 percent of eligible costs of a disaster with state and local governments paying for the remaining portion. The assistance is to be provided to repair, restore, reconstruct, or replace eligible facilities. The amount of public assistance provided is reduced by, among other considerations, insurance proceeds and salvage value. Because the assistance provided by the program is limited by these factors, as well as certain eligibility criteria, the amount of public assistance funds FEMA provides in a disaster does not equal the total financial impact of a disaster on an affected community or area. The Stafford Act has been amended several times since its enactment in 1974, and FEMA has taken steps over the years to redesign its public assistance program with internal policy changes to make eligibility criteria for public assistance clearer, and more consistent and accurate. The Senate report on the Disaster Mitigation Act of 1999 noted that the congressional interest in reducing the federal cost of disaster assistance would be achieved by, among other things, reducing the types of facilities and activities that may receive assistance in the event of a disaster. In August 2001, we reported that in a period of about 2 years since FEMA had completed a 1998 redesign of the public assistance program, it had developed or revised public assistance program policies in 35 areas or topics in part to make clearer eligibility criteria and improve the consistency and accuracy of eligibility determinations for individual projects. FEMA’s public assistance program is the largest portion of the federal assistance provided to New York in the aftermath of the World Trade Center attacks. Of a total of over $20 billion in federal assistance approved for this disaster, either in the form of direct assistance or in the form of tax benefits, about $7.4 billion was funded through FEMA’s public assistance program or through public assistance-related spending authorized by Congress through appropriations to FEMA. Figure 2 shows that FEMA’s public assistance program is providing the largest single portion of the federal contribution to the NYC area’s disaster recovery effort. Housing and Urban Development (HUD) funding ($3.5 billion) Liberty zone tax package ($5.0 billion) FEMA may assign work or enter into agreements with other federal agencies and the American Red Cross to handle aspects of public assistance within their areas of expertise. These agreements are called mission assignments and interagency agreements. Mission assignments were widely used in the first few months after the World Trade Center disaster to provide assistance for short-term projects. Interagency agreements—used for long-term projects—are similar to mission assignments in that they are funding agreements between agencies to provide goods and services on a reimbursable basis. In March 2003, FEMA and its responsibilities were placed entirely into DHS in the largest reorganization of the federal government since the formation of the Department of Defense. The Emergency Preparedness and Response Directorate within DHS has responsibility for the public assistance program and continues to be referred to as FEMA, which we do in this report. The approximately $7.4 billion of public assistance and public assistance- related work funded through FEMA is providing a broad range of aid to the NYC area. For example, public assistance-related funding was, or will be, provided to reimburse NYC authorities for immediate response and recovery actions—such as debris removal operations and emergency efforts by the NYC Departments of Design and Construction, Sanitation, Fire, and Police—and for long-term actions to repair and upgrade damaged facilities and transportation systems. Because of the unique nature of the NYC disaster, existing FEMA data system categories for tracking and reporting public assistance do not provide for some of the large public assistance-related efforts. Based on our analysis, we categorize the public assistance and related funding for NYC into six general areas: debris removal operations and insurance; reconstruction of the Lower Manhattan transportation infrastructure under an interagency agreement with DOT; reimbursement of police and fire department costs; reimbursement of expenses incurred by NYC agencies other than the Departments of Design and Construction, Sanitation, Police and Fire for such activities as DNA and forensic testing to identify victims and exterior building cleaning; reimbursement of expenses to agencies that are not part of the NYC government (i.e., New York state agencies, the Port Authority, and private non profits) for disaster-related costs such as transportation work not covered under the interagency agreement discussed above; and reimbursement of public assistance-related expenses authorized by Congress that would not otherwise have been eligible for assistance (i.e. heightened security after the terrorist attacks) from funds made available after the June 30, 2003, close out of the traditional public assistance program. Refer to figure 1 on page 4 for a graphic illustration of how public assistance funding to the NYC area was or will be distributed within these six categories. Each category of public assistance funding and some of the major efforts funded in each of them, are described in the following sections. FEMA funded about $1.7 billion in work related to debris removal operations and to reimburse the NYC Departments of Design and Construction and Sanitation for debris removal expenses. The most significant and costly activities in this category were removing and disposing of the destroyed World Trade Center buildings, screening debris for victims’ remains and personal effects, and establishing an insurance company for possible claims resulting from debris removal operations. Workers spent an estimated 3.1 million hours over 9 months to remove about 1.6 million tons of debris from the World Trade Center site. Debris from the collapse of the World Trade Center towers extended 7 stories into the earth and more than 11 stories high at Ground Zero. Thick dust covered streets, buildings, and vehicles for blocks around the site. FEMA provided $620.9 million for removing the debris from the World Trade Center site and barging it to a landfill in Staten Island, N.Y., for screening, sorting, and disposal. Original estimates projected that the recovery effort and cleanup would take 2 years and $7 billion. Figure 3 shows debris removal and barging operations. The need to sort and screen the debris to recover the remains and personal effects of victims and criminal evidence made the debris removal operation even more difficult. FEMA provided $72 million to the U.S. Army Corps of Engineers to manage the debris inspection at the landfill. The sorting activities were an intense, meticulous effort to recover remains and personal belongings of victims to return them to their families and to gather criminal evidence related to the terrorist attacks. The Corps of Engineers provided labor, heavy equipment, conveyer belts, and screening equipment. The Corps also provided temporary buildings for storage and to shelter workers, worker decontamination facilities, and food service facilities. Figure 4 shows debris screening and inspection operations at the landfill. In addition to the costs of debris removal and disposal, FEMA set aside $1 billion to establish a debris removal insurance company to cover contractors and NYC for liability claims resulting from debris removal operations. According to city officials, private contractors came to Ground Zero to do search and rescue, recovery, and debris removal work in the immediate aftermath of the terrorist attacks before entering into formal contract agreements with NYC. The outstanding issue that kept the contractors and NYC from reaching a final agreement on compensation for the work done was liability insurance coverage. City officials said that liability insurance could not be obtained from a private insurance company because of the unknown risks and potentially large number of liability claims. Based on input from insurance experts, city officials and FEMA determined that the best solution was to establish an insurance company with $1 billion in federal capital to provide $1 billion in coverage for a payout period of up to 25 years. The insurance fund will cover NYC workers and contractor employees. As of June 2003, the details of the insurance coverage had not been finalized. Additional perspectives on how aspects of FEMA’s establishment of the insurance fund differed from a traditional public assistance activity can be found on page 30 of this report. FEMA provided $2.8 billion to help fund an interagency agreement with the DOT to reconstruct the Lower Manhattan transportation system. The terrorist attack at the World Trade Center severely damaged the intermodal public transportation system that was used by about 80 percent of the 350,000 daily commuters to Lower Manhattan—the highest percentage of people commuting to work by public transit of any commercial district in the nation. The Port Authority of New York and New Jersey (Port Authority) commuter station underneath the World Trade Center was destroyed, and subway stations servicing the area were sufficiently damaged to prevent trains from stopping at them. In addition, some tunnels were temporarily closed, preventing commuter buses from entering Lower Manhattan. Access to and mobility within Lower Manhattan was severely diminished. Many streets were closed due to debris from the collapsed buildings and the subsequent debris removal operations. Large rescue vehicles and heavy debris removal equipment also damaged the area streets, making them more difficult to navigate. Plans are underway to rebuild and improve the Lower Manhattan transportation system with funding from FEMA and DOT. These agencies, under an interagency agreement, will contribute $4.6 billion to these transportation system projects, with FEMA providing $2.8 billion and DOT providing an additional $1.8 billion. The agreement will result in not only rebuilding a system that was damaged, but also improving the overall Lower Manhattan transportation system. The agreement designated DOT’s Federal Transit Administration (FTA) as the lead agency in charge of administering the federal assistance and coordinating with state and local implementing agencies. In February 2003, the Governor of New York submitted funding requests to FEMA and DOT for three priority projects estimated to cost between $2.55 billion and $2.85 billion—the World Trade Center Transportation Hub, Fulton Street Transit Center, and South Ferry Subway Station to improve the overall flow of commuter traffic in lower Manhattan. Although the uses for the remaining $1.7 billion to $2.0 billion of the $4.6 billion in FEMA/DOT funds had not been determined as of June 2003, uses for the remaining funds being evaluated included improvements in access to JFK Airport and Long Island, improvements to West Street Route 9A, a tour bus facility, the World Trade Center sub grade infrastructure, and commuter ferries and street configuration work. Figure 5 shows the extensive damage to the PATH commuter station beneath the World Trade Center Towers after the terrorist attacks and a model of the permanent station planned to be constructed in its place with FEMA/DOT interagency agreement funds. FEMA is also funding transportation-related work for the Port Authority outside of the scope of this interagency agreement. This work is discussed on page 21 of this report. We provide additional perspective on how aspects of this interagency agreement differ from FEMA’s traditional public assistance response to major disasters on page 28 of this report. FEMA provided about $643 million in assistance to the NYC Police and Fire Departments to pay benefits and wages to emergency workers during response and recovery efforts and to replace vehicles and equipment. As first responders, these departments suffered heavy casualties and damages in the collapse of the twin towers of the World Trade Center: 343 NYC fire department employees, 23 active city police officers, and 5 retired city police officers died in the line of duty, and 238 emergency vehicles, as well as radios and other equipment were lost or destroyed. In the months after the attack, nearly 100 firefighters per shift worked at the disaster site around the clock standing over contractor-operated steel-ripping machines looking for victims’ remains. Similarly, police officers were stationed 24 hours a day, 7 days a week to provide security at the disaster site. Figure 6 includes photographs of police and firefighters during the search and rescue phase of work immediately after the terrorist attacks and 2 of the emergency vehicles that were destroyed in the World Trade Center collapse. Public assistance grants to these two city agencies included $341 million for police overtime and death benefits and $223 million for firefighter overtime, death benefits, and funeral costs. Grants also reimbursed emergency service departments $44 million to replace 98 firefighter vehicles, radios, and other equipment; and $26 million to replace 140 police emergency vehicles and emergency equipment that were destroyed in the terrorist attacks. Although the NYC Departments of Design and Construction, Sanitation, Fire, and Police were the city agencies that received the largest amounts of FEMA public assistance funding for debris removal and insurance and for emergency response losses and expenses related to the terrorist attacks, FEMA also provided direct public assistance to a number of other NYC agencies for a wide range of work totaling almost $300 million. Projects included: $46.7 million to the Office of the Chief Medical Examiner for DNA testing, forensic analysis and equipment to help identify victims of the terrorist attacks; $8 million to the Department of Elections to reimburse the expenses it incurred to reschedule elections that were being held on September 11, 2001, and to replace damaged voting equipment; $19.3 million to the NYC Department of Education to pay for instructional time for students who missed school due to closures, delayed openings, and school relocations; and $8.6 million to the NYC Department of Environmental Protection for exterior building cleaning. Other examples of funding that went to city agencies are $12.9 million to the NYC Department of Citywide Administrative Services for emergency supplies, equipment and services, and $10.6 million to set up the facilities and provide equipment and furniture for the NYC Family Center and reimburse city and state personnel for overtime at the Family Center who provided services for NYC residents in the aftermath of the terrorist attacks. Figure 7 shows the cloud of dust that covered buildings for blocks around the World Trade Center. FEMA provided over $700 million in public assistance-related funding to agencies that were not part of the NYC government, including the Port Authority, state agencies, counties, and private nonprofit organizations. Among the agencies receiving some of the largest amounts was the Port Authority, which sustained substantial losses of lives and property as a result of the terrorist attacks. The funding for the Port Authority was in addition to the FEMA transportation funding provided in its interagency agreement with DOT to rebuild and improve the Lower Manhattan transportation system, as discussed on page 16. FEMA reimbursed the Port Authority for a wide range of work including $285.0 million to relocate offices that were located in the World Trade Center, repair commuter train tunnels that were damaged in the terrorist attacks, implement emergency ferry services, open a temporary PATH station, and pay overtime to the Port Authority police. The damage to the Port Authority’s PATH train system was extensive; tunnels leading from the station to New Jersey were flooded and the Exchange Place station in New Jersey had to be closed because the station could not operate as a terminal. All tunnel components (i.e., fiber optics, conduits, pipes, lighting, ductbanks, track, contact rail, and ballast) needed to be replaced. The Port Authority also received public assistance funds to replace equipment it lost when its World Trade Center facilities were destroyed, including its voice telephone network, desktop computers, and fax and photocopy machines, and to pay overtime labor costs for the emergency response. Figure 8 shows PATH tunnel repair and construction efforts. FEMA also provided public assistance funds to many other non-NYC government agencies to reimburse them for emergency and repair costs. For example, the New York State Police received $45 million for security operations, and New York University received $5.9 million for air monitoring, environmental cleaning, and emergency supplies and services. Other examples include the NYC Office of Emergency Management, which received $11.8 million from FEMA to replace destroyed equipment and leased office space that was located in the World Trade Center; Pace University, which was provided $4.4 million for damaged buildings; and the Battery Park City Authority, which received $3.9 million to repair damaged facilities. Lastly, $1.2 billion was made available in June 2003 as a result of FEMA’s early close out of its traditional public assistance program to NYC and state for congressionally authorized costs associated with the terrorist attacks. Most of these costs would not have been eligible for reimbursement under FEMA’s traditional public assistance program. The close out freed funds for discretionary public assistance-related uses by NYC and state and ensured that FEMA would spend the entirety of the appropriated assistance to the NYC area. Funds obligated for all of FEMA’s programs, including individual assistance and hazard mitigation, were reconciled, and funds that had not been expended for approved projects as of April 2003 were de-obligated to be used for discretionary public assistance-related expenditures. To receive the $1.2 billion reimbursement for public assistance-related costs, FEMA officials reported that NYC and state officials must prepare traditional grant applications to document that disaster-related costs have been incurred; however, Congress authorized wide discretion on the type of costs that could be reimbursed. As we concluded our review, the list of projects to be funded had not been determined, but NYC and state had requested reimbursements for heightened security in the aftermath of the terrorist attacks and cost-of- living adjustments to pensions of the survivors of firefighters and police officers killed in the line of duty in the terrorist attacks. A $19 million reimbursement has been made for a public awareness campaign called “I Love New York,” which was designed to attract visitors back to the city after the terrorist attacks. We discuss the heightened security reimbursements in more detail on page 32 of this report as an example of funding that was different in scope than a typical public assistance project and that would not have been eligible for FEMA funding unless it was specifically authorized by Congress. Each disaster to which FEMA responds has aspects that make it unique from other disasters, resulting in some differences in forms of assistance provided to affected communities within the parameters of the Stafford Act eligibility requirements, according to the head of FEMA’s public assistance program. While FEMA followed traditional processes for considering most applications, the public assistance response in the NYC area after the terrorist attacks differed significantly from the traditional approach FEMA has used in providing assistance under the Stafford Act after major natural disasters. The three significant differences were: the elimination of any local sharing of disaster response and recovery costs, capped amounts of funding that resulted in significant modifications to the project selection and close out processes, and the size and type of projects funded. Many of these differences are based on presidential and congressional direction; however, some are the result of FEMA’s interpretations of the Stafford Act to allow the approval of funding for certain assistance to New York. The Stafford Act sets the federal share for the public assistance program at no less than 75 percent of eligible costs. The President can increase the federal share for the public assistance program if it is determined that the disaster costs greatly exceed a state’s financial capabilities. In practice, the federal share has reached 100 percent for emergency work, for limited periods of time, if determined that it was necessary to prevent further damage, protect human lives, or both. In 1992, for example, after Florida and Louisiana suffered large disaster expenses as a result of Hurricane Andrew, FEMA funded 100 percent of all public assistance costs above $10 per capita. According to a FEMA official, the 1994 Northridge, California earthquake, which cost almost $7.0 billion, was FEMA’s most costly disaster funding effort until the World Trade Center attacks occurred; FEMA provided for 90 percent of all public assistance costs. In discussing the question of state and local sharing of public assistance costs, FEMA officials stated that they are reluctant to recommend a 100 percent federal share for projects unless there are compelling reasons to do so because the traditional process with a matching share creates incentives for state and local officials to control costs and closely evaluate projects. In the days immediately following the terrorist attacks, the President determined that the magnitude and nature of the disaster justified the federal government funding the total cost of public assistance projects, and he directed that FEMA fund 100 percent of the eligible costs with no state or local matching funds. This increased FEMA’s costs and significantly reduced costs to NYC and other recipients. For example, on the transportation repair and improvements efforts, NYC area recipients did not have to make a financial contribution that could have totaled nearly $680 million—25 percent of the $2.75 billion that FEMA is providing. Although New York received the benefits of 100 percent FEMA funding of public assistance projects, the President reduced the amount of related Hazard Mitigation Grant Program funds provided to New York. Created in 1988 by the Stafford Act, this grant program provides funds to communities affected by major disasters to undertake mitigation measures following a major disaster. At the time of the terrorist attacks, grants funds up to 15 percent of the total amount of FEMA assistance provided are available to states following a disaster. However, in this case, the President limited the mitigation grant funds to 5 percent of the amount spent. Had the hazard mitigation funding percentage not been reduced, more than $1.2 billion in mitigation funds would have been required using the customary 15 percent of total cost criteria. In a typical major disaster, FEMA’s consideration of whether work is eligible for public assistance is not constrained by a limit on the total amount of public assistance funding that can be spent, and disasters remain “open” with FEMA until public assistance work is substantially completed. Generally, FEMA officials approve all public assistance applications that meet eligibility criteria under the Stafford Act, and they fund the work from FEMA’s disaster relief fund. Also, according to a FEMA public assistance official, direct congressional appropriations are not typically made for a specific disaster. The official explained that damaged facilities are identified within 60 days following a kick-off meeting to begin federal disaster assistance between FEMA officials and state and local officials of the area impacted by the disaster. Proposed work is then considered for eligibility and funded through “project worksheets”—applications for specific funding amounts to complete discrete work segments. Project worksheets document the scope of work, cost estimates, locations, damage descriptions and dimensions, and special considerations of each work segment. No limit is set on the dollar amount of eligible work that can be approved. As the response and recovery progresses, states reimburse applicants for all costs that meet the Stafford Act’s public assistance eligibility criteria and FEMA reimburses the states for the federal share. A public assistance official noted that disasters remain open with FEMA long after public assistance funds have been obligated. For example, as of June 2003, the Northridge, California, earthquake was still an open FEMA disaster 9 years after it occurred due to large and long-term reconstruction efforts. Disasters are “closed” when the project is complete, the final costs are known, and all appeals of funding decisions have been resolved. Following the terrorist attacks, however, the process of selecting projects that were eligible for funding and closing out the public assistance for the NYC area did not follow FEMA’s customary process because FEMA had a set amount of funds available for public assistance efforts. Congress provided FEMA with specific appropriations for the terrorist attacks that resulted in a capped funding amount of $8.8 billion for its efforts to aid the NYC area from the President’s pledge of at least $20 billion in federal assistance. In consideration of funding required for its other programs (assistance for individuals impacted by the disaster and hazard mitigation grants), $7.4 billion remained available for public assistance and public assistance-related projects. To help ensure that the amount of public assistance did not exceed this amount, FEMA asked that city and state officials prioritize their funding needs. As a result, about $400 million in funding initially budgeted for the Port Authority was eventually reallocated to other projects. FEMA also delayed a decision on funding for city and state pension actuarial losses resulting from line of duty deaths of police and fire fighters at the World Trade Center site so that officials could be certain that the costs of the project would not cause FEMA to exceed its total appropriation for the disaster. A second major difference from how FEMA typically manages a disaster occurred when it established a June 30, 2003, deadline for closing out the regular public assistance program and the disaster before work was completed. According to FEMA officials, they established this deadline for closing out public assistance projects eligible for funding under the Stafford Act so that any remaining funds could be used for work identified as high priorities by city and state officials in New York and authorized by Congress. They said that deadlines for closing out public assistance had not been set in any prior disaster until work was completed, but that they believed it was necessary for the NYC area to manage the available funds to ensure that its priorities are best met as quickly as possible. The response to the NYC terrorist attacks was the largest public assistance effort in FEMA’s history and by far its largest response to a terrorist event. Prior to the World Trade Center attacks, FEMA’s most costly disaster assistance—almost $7 billion—was provided to aid in the recovery from the Northridge, California, earthquake in 1994. FEMA spent more than $1 billion for five other disasters in its history. Further, FEMA’s experience with terrorism was limited to two occasions prior to the World Trade Center attacks. In April 1993, a major disaster was declared in the aftermath of an explosion caused by terrorism at the World Trade Center. FEMA spent about $4.2 million on that disaster recovery. In April 1995, an emergency and then a disaster were declared in Oklahoma City, Oklahoma, in the aftermath of the bombing of the Murrah federal building—FEMA spent about $530 million on that recovery effort. In its response to terrorism in the NYC area, FEMA provided public assistance funds for the same types of projects that are funded after a natural disaster (e.g., removing debris, repairing roads, and replacing emergency vehicles that were destroyed). However, other work funded was quite different because of the magnitude and nature of the disaster. FEMA officials said that they determined that some non traditional work was eligible for its public assistance program using flexible interpretations of the Stafford Act. Examples of public assistance projects approved by FEMA that we identified as being different from traditional public assistance work due to their size and/or type of work done included improvements to the Lower Manhattan transportation system and air quality testing. Some of these projects are discussed as follows. Improving Lower Manhattan’s Transportation System ($2.75 billion). Public assistance has traditionally been limited to repair of disaster-related losses and damages to existing infrastructure. Assistance has not generally been provided to enhance or modernize the infrastructure beyond its pre disaster conditions. In recognizing the interdependence of Lower Manhattan’s transportation system, however, FEMA officials said that they broadly interpreted their guidelines to enter into an interagency agreement with DOT to rebuild physical facilities that were damaged from the attacks and construct new facilities that may improve the overall Lower Manhattan transportation system. FEMA attorneys said that they determined that the Stafford Act would permit funding for the restructuring of the Lower Manhattan transportation system because they concluded that repairing and replacing individual elements would not completely restore the system’s functionality. Testing air quality and cleaning buildings ($36.9 million). FEMA officials said that air quality testing and removing dust from buildings had not been an issue in prior major disasters, however, it was important to the physical and psychological well being of NYC citizens in the aftermath of this disaster. FEMA determined that the testing of air quality and cleaning were eligible for public assistance funding where the collapse of the World Trade Center buildings, resulting fires, and subsequent debris removal caused potential health issues related to air quality. To meet this need, FEMA entered into interagency agreements with EPA to sample and test air quality in the NYC area, as well as to test ways to clean potentially hazardous dust in building interiors. FEMA also provided funding to the New York Department of Environmental Protection for the exterior cleaning of 244 buildings and the interior cleaning of residences. EPA provided oversight over the interior cleaning program as part of the interagency agreement with FEMA. Reimbursing costs for rescheduling New York elections ($11 million). According to a FEMA official, this disaster was the first during which elections were being held on the day of a federally declared disaster event. FEMA officials said that they considered whether the costs of canceling the elections statewide and rescheduling them at a later date were eligible for public assistance or were increased operating expenses for the state and local governments that are not considered to be eligible for assistance under the Stafford Act. After initially denying the public assistance application for reimbursement, FEMA officials reconsidered and determined that the costs were eligible for reimbursement as disaster related expenses. NYC was also reimbursed for costs of damaged and destroyed computers, voting machines, and ballots as Stafford Act eligible public assistance. Aiding WNET Public Television (covered completely by private insurance) and the Legal Aid Society of New York for Public Assistance ($1.6 million). According to FEMA officials, WNET, a nonprofit television station, requested reimbursement from the public assistance program for expenses for a communications antenna that was damaged in the World Trade Center attacks. The New York Legal Aid Society asked for reimbursement of disaster-related costs including repair of damages to its building and reconstruction of its data hub that was destroyed in the attacks. Although public television stations are not among the specific types of non profit organizations that are normally considered to be eligible applicants for public assistance because they provide essential government services (i.e. educational, medical, water, and sewer treatment facilities), FEMA determined that WNET was eligible as a public facility because it provided health and safety information to the general public during the crisis. Later, WNET received full coverage for its claims from a private insurance company, so FEMA funds were not awarded. Similarly, FEMA officials said that although legal aid societies are not generally eligible for public assistance, the Legal Aid Society of New York was eligible because it provided government services as the public defender for NYC. These projects were not traditional because they required flexibility in FEMA’s interpretation of Stafford Act definitions of private nonprofit and public facilities that are eligible for public assistance. Notwithstanding its efforts to be flexible in defining public assistance activities eligible under the Stafford Act, FEMA officials denied some applications because they determined they were not eligible for public assistance under the Stafford Act, but the Congress directed FEMA to reimburse the NYC area for some public assistance-related costs that would not otherwise have been eligible for funding. An estimated total of $2.2 billion of FEMA’s public assistance funds—about 28 percent—will go to these costs. This public assistance-related funding was different from work FEMA funds under the Stafford Act. The projects included authorizing a debris removal insurance fund for workers at the World Trade Center site and reimbursing NYC agencies for the costs of providing heightened security after the terrorist attacks. In addition, as discussed on page 24 of this report, as we concluded our review, FEMA and NYC and state officials were considering projects to be funded with $1.2 billion that became available after the close out of traditional work in June 2003 for congressionally authorized purposes. None of these reimbursements were eligible for funding under FEMA’s public assistance program. Reimbursements being considered included payment of increased costs of the Medicaid program to meet health needs of recipients after the attacks, a public awareness campaign called “I Love New York,” which was designed to attract visitors back to the city after the terrorist attacks, and cost of living adjustments made to the pensions of survivors of firefighters and police officers killed in the line of duty in the terrorist attacks. Debris removal insurance for workers at the World Trade Center site ($1 billion). As discussed on page 15, this project establishes an insurance company to insure NYC and its contractors for claims arising from debris removal at the World Trade Center, including claims filed by workers who suffer ill health effects as a result of working on debris removal operations. FEMA officials said that the project is unprecedented in its size and complexity and because it involves long-term health and environmental issues of a scope FEMA had not considered in prior major disasters. Although officials said that FEMA has never established an insurance fund to manage claims from other major disasters, FEMA Office of General Counsel officials noted that FEMA does frequently pay for contractors’ insurance because it is built into the contract between the public assistance applicant and the contractor. In this instance, workers rushed to the disaster site before any contracts were approved, and no private insurance company would carry the insurance because of unknown liabilities. FEMA officials said that the portion of the project pertaining to contractors qualified for public assistance under the Stafford Act and is a disaster-related cost that FEMA has traditionally assumed in major natural disasters. Expanding the coverage to include liability for claims filed against NYC or by city workers was an eligibility issue that was under consideration within FEMA when Congress authorized the funding in the Consolidated Appropriations Resolution for fiscal year 2003. Reimbursement for heightened security costs in the aftermath of the terrorist attacks (amount of funding not determined). FEMA denied applications for public assistance to reimburse city agencies, including the Departments of Environmental Protection, Corrections, Fire, and Transportation to cover costs for increased security (e.g., the Department of Environmental Protection took increased security measures to protect the city water supply). A FEMA official said that the applications were not eligible for public assistance because the work was of the sort that was being done nationwide after the terrorist attacks and were intended to prevent future attacks rather than respond to the disaster that had occurred. However, NYC Office of Management and Budget (OMB) officials said that some of the heightened security costs would be reimbursed as a result of the enactment of the Consolidated Appropriations Resolution for fiscal year 2003, which allowed NYC flexibility in covering disaster-related costs not otherwise reimbursable under the Stafford Act. At the time of our review, the amount of funding to be provided for heightened security costs had not been determined, but it was anticipated by FEMA officials to be over $100 million. Reimbursement for instructional time for students to make up for days missed after the terrorist attacks ($19.3 million). FEMA initially denied a public assistance request to pay for additional hours of instructional time for students who missed school due to closures, delayed openings, and school relocations in the aftermath of the terrorist attacks. FEMA officials said that the application was denied because the after-school program designed by the NYC Board of Education to make up for the lost instructional time was predicated on direct FEMA funding, but it did not meet the standards of emergency work for which applicants must perform work immediately after a disaster, regardless of who will pay, to eliminate an immediate threat to health, life, and safety. However, FEMA was specifically directed by the congressional conference committee making supplemental appropriations for fiscal year 2002 to provide funds for the additional instructional time. The conference report also directed FEMA to provide compensation to the NYC school system for costs stemming from the terrorist attacks for services and supplies, including mental health and trauma counseling, guidance and grief counseling, and replacement of lost textbooks and perishable food. NYC Board of Education had spent $19.3 million of a total $77.6 million approved for this work as of April 30, 2003. The remainder of the funding was de-obligated to be used for public assistance related spending authorized by Congress. Because the public assistance response to the NYC area after the terrorist attacks was unique and expanded in terms of the level and types of assistance provided, it creates uncertainty about how public assistance will be delivered if another catastrophic terrorist attack occurs. Both NYC and FEMA officials, including managers of the World Trade Center Federal Recovery Office and top officials of the NYC Offices of Emergency Management and OMB, agreed that they were uncertain regarding the level and type of future FEMA assistance. These officials stated that if another major terrorist disaster occurs, other communities might seek similar types of assistance as was received in the federal public assistance response to New York. In this regard, an official of the NYC OMB anticipated that one of the first calls by a mayor of a city that experienced a major terrorist event would be to NYC to discuss the decisions made in the aftermath of the World Trade Center attacks. FEMA Recovery Office officials agreed that the decisions made in New York would be on the table at discussions of federal assistance for any future terrorist event. They noted that it would remain to be seen whether an approach similar to the one that evolved in NYC, including a 100 percent federal share for public assistance funding, a capped funding amount, and flexibility in addressing needs, would be used following any future event. The Congressional Research Service noted similar concerns in a June 2002 report about the implications that the response and assistance provided to the NYC area may have on future federal response to catastrophic terrorist events. The agency’s report pointed out that one of the long-standing principles of federal disaster assistance policies has been that federal aid should supplement—not supplant—nonfederal efforts and that the actions taken in the aftermath of the terrorist attacks might have established precedent for an expanded federal role in consequence management after terrorist attacks. The report noted that traditionally, the types and amounts of assistance provided after one disaster have been sought following succeeding catastrophes. The report also states that the overriding question is whether the range of existing federal policies for responding to disasters is appropriate if a terrorist attack more devastating than that of September 11 were to occur. This is a question to which NYC and FEMA officials have differing positions. With respect to the effectiveness of the Stafford Act in dealing with a major terrorist event of an impact equal to or greater than the World Trade Center attacks, the officials from NYC involved in the response and recovery efforts whom we interviewed did not believe that the act fully addressed the needs of the city and did not think it should be used to respond to major terrorist events unless it had significant amendments to address the unique challenges related to terrorist events. According to top officials of both the NYC Office of Emergency Management and OMB, the public assistance program authorized by the Stafford Act is not a good fit for the needs of a large municipal government that is coping with the effects of a terrorist event. They pointed out that the impacts of the terrorist attacks in NYC were different than impacts from the natural disasters that the act was created to address. For example, the Stafford Act does not address concerns such as the federal government’s responsibility for addressing long-term environmental liabilities. Additionally, a NYC emergency management official noted that the Stafford Act lacked provisions for cities and states to be eligible for reimbursement of money spent to provide security in the immediate aftermath of terrorist attacks. The city officials noted that funding to help alleviate these impacts was eventually approved, but not without considerable discussion with FEMA officials and specific direction from Congress. A key NYC OMB official also said that the Stafford Act is too restrictive for responding to a major terrorist event because it does not allow the reimbursement to affected communities for budget shortfalls resulting from lost tax revenues. The official said that NYC lost tax revenues, both from real estate taxes from the destroyed buildings and corporate, sales, and income taxes from displaced businesses and individuals that were eligible for reimbursement under the Stafford Act. He said that NYC requested $650 million in reimbursement for revenue shortfalls in fiscal years 2002 and 2003 that were directly related to the terrorist attacks. While FEMA officials agreed that the estimate seemed reasonable, the amount was not eligible for reimbursement under the Stafford Act. Congress recognized the problem and provided the city some flexibility to cover expenses in these areas. However, the New York OMB official said that a federal block grant would have allowed the city to spend the money in ways that were most needed without specific congressional authorization to do so; he viewed a block grant approach to providing disaster relief as preferable to trying to obtain the funding under the Stafford Act. In contrast, FEMA officials said that the Stafford Act worked appropriately for the NYC area. FEMA attorneys said that the Stafford Act contains enough flexibility to allow funding for non traditional activities. They added that every disaster has unique aspects, which continually challenge FEMA officials to exercise their discretion under the act to provide needed assistance. Furthermore, they point out that it is always the prerogative of the Congress to provide additional assistance to disaster-affected areas to address specific and unique needs. If Congress saw a need to fund public assistance-related work not covered under the Stafford Act in the event of another major act of terrorism, it could appropriate funds specifically for the disaster, as it did in NYC. Consequently, the FEMA officials are generally satisfied that they are able to apply provisions of the Stafford Act to respond to the terrorist attacks and, as of June 2003, did not believe significant changes to the legislation were necessary in the aftermath of September 11, 2001. Nevertheless, FEMA recently initiated an effort to develop a concept for redesigning the public assistance program. A working group of the Public Assistance Program Redesign Project, formed at the request of the director of FEMA’s Recovery Division, held its first meeting in May 2003. Members included FEMA public assistance and research and evaluation staff and state program managers to provide a broader perspective on the issues and concerns. The project was established to suggest proposals to improve the public assistance program and make it more efficient and capable of meeting community needs for all types and sizes of disasters, including those resulting from terrorism. Among other things, the project seeks to transform the program to one that: is flexible enough to meet the demands of disasters of all types and sizes, reduces overall resource requirements, offers incentive for timely close outs, places operational control principally with states and applicants, and eliminates redundancies in decision making and processes. The working group will examine potential options for redesigning the program that include an annual block grant program managed by the states, a disaster-based state-managed program, and a capped funding amount. The project is currently scheduled to hold a listening session for local officials and representatives of other organizations in August 2003, and develop a basic concept design for revising the program by September 30, 2003. The public assistance program FEMA delivered in the NYC area after the terrorist attacks was substantially different in several ways from a “typical” FEMA public assistance response. For example, in the NYC area there was a lack of cost sharing with state and local governments; a smaller than usual federal share of hazard mitigation funding; a different process for project review, selection, and close out; and, most significantly, the size and scope greatly exceeds the traditional public assistance response after a major natural disaster. The reasons for these differences are many and include the President’s early commitment to providing a specified amount of funding to New York, congressional direction on activities to fund, and FEMA’s discretion under the Stafford Act. Irrespective of the reasons for the differences in the way public assistance was delivered after the terrorist attacks, these differences raise questions about FEMA’s response to any future major terrorist event in this country. The key issue is whether the differences in the ways the public assistance program in the NYC area was delivered will serve a baseline for the federal approach in the event of another major terrorist event. Should such a terrorist event occur, it is not unrealistic to assume that affected communities will expect to receive public assistance comparable to that provided for the NYC area to meet their needs. DHS, within its Emergency Preparedness and Response Directorate, has an opportunity to assess the questions raised as a result of these differences and, if necessary, revise the public assistance program or provide Congress with suggestions for legislative changes that are needed so that it will be positioned to address new expectations for disaster assistance. The newly formed Public Assistance Redesign Project, established as we were concluding our audit work at the request of the Director of FEMA’s Recovery Division, plans to address many of the issues raised in this report, including whether the approach used in NYC is the appropriate way to provide federal assistance for recovery from terrorist acts. It is too early for us to assess the impact the project will have on the public assistance program in the future; however, it is a promising first step toward addressing these issues and better ensures that DHS will have a process in place to deliver public assistance that eliminates uncertainties and questions about the ways in which the needs of affected communities will be met in the event of another major terrorist attack. In commenting on a draft of this report, the Acting Director of FEMA’s Recovery Division said that FEMA officials are proud of the agency’s response in delivering public assistance programs to NYC and state, and that they are satisfied that FEMA’s authority was adequate and flexible enough in most circumstances to meet the response and recovery needs of New York. The Acting Director did not take exception to any of the information provided in our report. FEMA’s comments are reprinted in appendix II. FEMA also provided technical comments on our draft, which we incorporated into the report where appropriate. As we agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution of it until 30 days from the date of this letter. We will then send copies of this report to the Secretary of Homeland Security and interested congressional committees. We will make copies available to others upon request. In addition, this report will be available at no charge on our Web site at http://www.gao.gov. If you or your staff have any questions about this report, please contact me on (202) 512-2834 or at heckerj@gao.gov. Individuals making key contributions to this report are listed in appendix III. To determine what activities the Federal Emergency Management (FEMA) supported through its public assistance program, we analyzed published FEMA reports and FEMA’s National Emergency Management Information System (NEMIS) data. NEMIS is FEMA’s primary information system that manages disaster grant funding, and we analyzed NEMIS data on public assistance funding for this disaster. Though we were not able to completely assess the reliability of the published FEMA program data, we did perform logic tests of the data and found no obvious errors of completeness or accuracy. Also, according to FEMA officials, the published reports are the most reliable information available. The officials said that published FEMA reports were compiled based on NEMIS data, as well as the knowledge of public assistance program managers of funding for specific projects. We also updated spending amounts for some projects to reflect changes made after FEMA’s June 30, 2003,closeout of the traditional public assistance program, based on technical comments to our draft report. We interviewed FEMA headquarters, regional, and recovery office officials in New York City, N.Y., and Washington, D.C. We analyzed FEMA, Office of Management and Budget (OMB), and Congressional Research Service reports on federal assistance to the New York City (NYC) area to recover from the terrorist attacks. We reviewed the Stafford Act and FEMA regulations for ensuring that public assistance program funds are spent appropriately on eligible work and discussed oversight processes with FEMA headquarters, regional, and recovery office officials. We also discussed the agreements that FEMA used to coordinate responses of other federal agencies. We selected and examined the FEMA agreements with the U.S. Army Corps of Engineers, Department of Transportation (DOT), and Environmental Protection Agency (EPA) and reviewed support documents. We met with officials of the FEMA Inspector General to discuss planning for full audits of selected projects within 3 years of their completion. To determine how the federal government’s response to the terrorist event differed from FEMA’s traditional approach to funding public assistance in other disasters, we selected 10 projects for detailed review from an issue matrix created by the public assistance officer at the World Trade Center Federal Recovery Office. The issue matrix tracked 32 public assistance funding issues and other types of concerns that required higher than normal levels of review. In making our selection of projects, we consulted with officials of the FEMA Office of General Counsel in Washington, D.C., and FEMA officials at the World Trade Center Recovery Office in New York City, N.Y. For each project selected, we reviewed available written documentation such as project worksheets, case management files, letters, and memoranda. We reviewed the legislation that directed FEMA to fund selected projects. Using structured interview instruments, we interviewed FEMA project managers and representatives of agencies that applied for public assistance to discuss how the challenging issues were considered and resolved. Table 1 lists the 10 projects we reviewed and the applicant organizations that participated in interviews on each of them. We also discussed FEMA’s staffing processes with human resources officials at FEMA Headquarters in Washington, D.C., World Trade Center Federal Recovery Office managers, and representatives of each of FEMA’s three technical assistance contractors who sent staff to NYC. To identify some of the implications these different approaches may have on the delivery of public assistance should terrorist attacks causing similarly catastrophic damage occur in the future, we interviewed FEMA officials in NYC, and FEMA and Congressional Research Service officials in Washington, D.C. We also analyzed our report and Congressional Research Service reports on federal emergency response and recovery policies, and we reviewed the Stafford Act and FEMA regulations. We conducted this review from August 2002 to July 2003. We performed our audit work in accordance with generally accepted government auditing standards.

The terrorist attacks on New York City created the most costly disaster in U.S. history. In response, the President pledged at least $20 billion in aid to the city. Approximately $7.4 billion of this aid is being provided through the Federal Emergency Management Agency's (FEMA) public assistance program, which provides grants to state and local governments to respond to and recover from disasters. The Senate Committee on the Environment and Public Works requested that GAO determine (1) what activities FEMA supported in the New York City area through its public assistance program after the terrorist attacks; (2) how the federal government's response to this terrorist event differed from FEMA's traditional approach to providing public assistance in past disasters; and (3) what implications FEMA's public assistance approach in the New York City area may have on the delivery of public assistance should other major terrorist attacks occur in the future. FEMA has supported many activities through its $7.4 billion in public assistance-related funding to the New York City area. Activities funded include grants to state and local governments for emergency response, such as debris removal, and permanent work, such as the repair of disaster-damaged public facilities. FEMA also provided public assistance-related funding specifically directed by Congress that would not otherwise have been eligible for assistance (e.g. reimbursing costs of instructional time for students who lost school time after the terrorist attacks). The major uses for this funding are as follows: $1.7 billion for debris removal operations and insurance; $2.8 billion to repair and upgrade the transportation infrastructure of Lower Manhattan; $0.6 billion to the New York City Police and Fire Departments for such purposes as emergency efforts and replacing destroyed vehicles; $0.3 billion to miscellaneous city agencies for a wide range of activities (e.g., instructional time for students and building cleaning); $0.7 billion for non-New York City agencies for many purposes (e.g. office relocations and repair of damaged buildings); and $1.2 billion available on June 30, 2003, for public assistance-related reimbursements to New York City and state (work to be decided). The provision of public assistance to the New York City area differed in three significant ways from FEMA's traditional approach. FEMA and New York City officials agreed that FEMA's public assistance approach in the New York City area creates uncertainties regarding the delivery of public assistance in the event of another major terrorist event. They differed on the effectiveness of using the public assistance program as currently authorized as the vehicle for federal disaster response to a future major terrorist event. Key New York City officials said that the program needed major revisions, while FEMA officials said it worked well along with the congressional prerogative to provide additional assistance. Nevertheless, FEMA has begun to consider ways to redesign the program to make it better able to address all types and sizes of disasters, including terrorist attacks.

When the WTC buildings collapsed on September 11, 2001, an estimated 250,000 to 400,000 people in the vicinity were immediately exposed to a noxious mixture of dust, debris, smoke, and potentially toxic contaminants in the air and on the ground, such as pulverized concrete, fibrous glass, particulate matter, and asbestos. Those affected included people residing, working, or attending school in the vicinity of the WTC and thousands of emergency responders. Subsequently, an estimated 40,000 responders who were involved in some capacity in the days, weeks, and months that followed, including personnel from many government agencies and private organizations as well as other workers and volunteers, were also exposed. A wide variety of physical and mental health effects have been observed and reported among people who were involved in rescue, recovery, and cleanup operations and among those who lived and worked in the vicinity of the WTC. Many health effects have persisted or worsened over time. Physical health effects included injuries and respiratory conditions, such as sinusitis, asthma, and a new syndrome called WTC cough, which consists of persistent coughing accompanied by severe respiratory symptoms. Almost all firefighters who responded to the attack experienced respiratory effects, including WTC cough. A recent study suggested that exposed firefighters on average experienced a decline in lung function equivalent to that which would be produced by 12 years of aging. Commonly reported mental health effects among responders and other affected individuals included symptoms associated with posttraumatic stress disorder—an often debilitating disorder that can develop after a person experiences or witnesses a traumatic event, and which may not develop for months or years after the event. Behavioral effects such as alcohol and tobacco use and difficulty coping with daily responsibilities have also been reported. The five programs that were created for monitoring the health of WTC responders vary in aspects such as the implementing agency (i.e., federal, state, or local governments or private organizations) and eligibility requirements. (See table 1.) Each program received federal funding, the majority of which was provided by the Department of Homeland Security’s Federal Emergency Management Agency (FEMA), as part of the approximately $8.8 billion in federal assistance that the Congress appropriated to FEMA for response and recovery activities after the WTC disaster. FEMA is authorized to use a portion of its WTC-related funding for screening and long-term monitoring of responders. With regard to treatment, however, FEMA may generally fund only short-term care after a disaster, such as emergency medical services, and not ongoing clinical treatment. FEMA entered into interagency agreements with HHS to fund most of the health monitoring programs. OPHEP, which coordinates and directs HHS’s emergency preparedness and response program, entered into separate interagency agreements with FOH to implement the federal responder screening program for current federal workers and with NIOSH to implement the screening program for former federal workers. We reported in February 2006 that four of the five monitoring programs had made progress in screening and monitoring affected individuals and gathering data. (See table 1.) These four programs—the FDNY WTC Medical Monitoring Program, the worker and volunteer WTC program, the New York State responder screening program, and the WTC Health Registry—had collected information that monitoring officials said could be used by researchers to help better understand the health consequences of the attack and improve treatment, such as by identifying which types of treatment are effective for specific conditions. In contrast to the progress made by the other programs, the HHS WTC Federal Responder Screening Program had lagged behind and accomplished little. The program was established to provide free voluntary medical screening examinations for federal workers whom their agencies sent to respond to the WTC disaster from September 11, 2001, through September 10, 2002, and who were not eligible for any other WTC health monitoring program. Through March 2004, the program—which started about a year later than the other WTC monitoring programs—completed screenings of 394 federal workers. HHS put the program on hold in January 2004, when it stopped scheduling new examinations, because it wanted to resolve several operational issues, including HHS’s determination that FOH did not have the authority to provide examinations to people who are no longer in federal service. Under an agreement between OPHEP and FOH that was established in July 2005, the program resumed providing examinations for current federal workers in December 2005, and in February 2006, OPHEP executed an agreement with NIOSH calling for NIOSH to arrange for the worker and volunteer WTC program to provide examinations to former federal workers. Many participants in the monitoring programs required additional testing or needed treatment for health problems that were identified during screening examinations. The FDNY WTC Medical Monitoring Program referred participants to the FDNY Bureau of Health Services, but the other programs primarily referred participants to their primary care physician or to privately funded programs available to responders, such as treatment services provided by the Mount Sinai clinical center that are funded by the American Red Cross. We previously reported that officials told us that finding treatment services for such participants was an important, but challenging, part of the programs’ responsibility. For example, officials from the worker and volunteer WTC program stated that identifying providers available to treat participants became a major part of their operations, and was especially difficult when participants lacked health insurance. In December 2005, the Congress appropriated $75 million to CDC to fund programs providing baseline screening, long-term monitoring, and health care treatment for emergency services and recovery personnel who responded to the WTC disaster. The law required CDC to give first priority to programs coordinated by the FDNY-BHS, Mount Sinai-Irving J. Selikoff Center for Occupational and Environmental Medicine, and New York City Department of Health and Mental Hygiene, which have existing monitoring programs, and to programs coordinated by the POPPA program and Project COPE. The mission of the POPPA program, which offers peer-to- peer mental health counseling to New York City Police Department (NYPD) officers, is to reduce unresolved emotional trauma that can result in problems ranging from poor performance to suicide. The POPPA program counseled over 5,000 NYPD officers in the 10 months following the WTC attack. Project COPE, a collaboration of the New York City Police Foundation and Columbia University Medical Center, uses a hotline and outreach efforts to encourage NYPD uniformed and civilian employees to obtain mental health services, which are provided by Columbia University Medical Center and private providers. As of August 2006, over 18,000 employees had attended educational sessions held at police facilities, and over 5,000 had received individual counseling or therapy consultations. Since February 2006, an additional 1,385 federal responders have registered for screening examinations, bringing the total number registered on the WTC Federal Responder Screening Program Web site to 1,762 as of late August 2006, including 283 former federal workers. Because the total number of federal responders involved in the WTC disaster is uncertain, it is not possible to determine what proportion of the total number of federal responders have registered. HHS’s efforts to conduct outreach to federal agencies resulted in the identification of 2,200 federal responders. As of late August 2006, FOH had completed screening examinations for a total of 907 federal workers, 380 of whom were screened since February 2006. Through OPHEP’s agreement with NIOSH, the worker and volunteer WTC program has provided screening examinations to 13 former federal workers and scheduled 11 more. Most of the former federal workers reside outside the New York metropolitan area, where the worker and volunteer WTC program is located, and NIOSH is working to establish a national network of providers to screen these workers. HHS reported that as of late August 2006, a total of 1,762 federal responders had registered for screening examinations on the WTC Federal Responder Screening Program Web site, including 1,479 current federal workers and 283 former federal workers. Of the 1,762 federal responders who registered, 1,385 had registered since February 2006, including 1,134 current federal workers and 251 former federal workers. It is not possible to determine what proportion of the total number of federal responders involved in the WTC disaster have registered because the total number involved is uncertain. In determining the total number of individuals eligible for its program, the WTC Health Registry developed an estimate of 8,621 federal responders, based on information from 31 federal agencies in the New York area and information from FEMA on 22 Urban Search and Rescue teams that were deployed to the WTC area. This estimate does not account for all federal responders from other geographic areas. As we reported previously, in the aftermath of the WTC disaster, HHS did not have a comprehensive list of all federal agencies and federal responders who were involved. In an effort to develop such a list, OPHEP and ATSDR entered into an agreement in April 2005 for ATSDR—which had developed the WTC Health Registry—to identify and register federal responders. Under the agreement, ATSDR, through a contractor, contacted federal agencies, developed a list of WTC federal responders, and conducted outreach to encourage the responders to register on the new Web site that the contractor established. As a result of this effort, 46 federal agencies were identified and provided contact information for 2,200 federal responders. The agreement between OPHEP and ATSDR expired on April 30, 2006, ending the outreach efforts to federal agencies. Under an agreement with OPHEP, NIOSH assumed responsibility for maintaining the WTC Federal Responder Screening Program Web site through December 31, 2006. As of late August 2006, FOH had completed screening examinations for a total of 907 of the federal workers who had registered; 380 of the 907 were screened since February 2006. Under its agreement with OPHEP, FOH is responsible for regularly retrieving from the registration Web site requests for screening examinations for current federal workers and for assigning individuals to a provider for screening. FOH officials told us that they contact the individual and the provider to inform them of the need to arrange an appointment for screening. The program relies on individuals to call the designated provider and schedule their appointment. FOH officials told us that individuals who have registered do not always contact the provider to schedule an appointment or may not keep an appointment or call to reschedule it. FOH officials said that they have attempted to contact such individuals but often received no response. We reported in our February 2006 testimony that under the July 2005 agreement FOH clinicians can refer current federal workers for follow-up care if the screening examination—which includes a medical questionnaire, clinical tests such as a chest X-ray, and a full physical examination—reveals significant physical or mental health symptoms. On July 31, 2006, FOH told us that it had referred 39 current federal workers with mental health symptoms to an FOH employee assistance program (EAP) for counseling; 24 to ear, nose, and throat specialists; 19 to pulmonary medicine specialists; and 1 to a cardiology specialist. As of late August 2006, 283 former federal workers had registered to receive screening examinations, which under OPHEP’s agreement with NIOSH are to be provided by the worker and volunteer WTC program. Under the agreement, former federal workers receive a one-time examination comparable to the examination that FOH is providing to current federal workers. As of July 31, 2006, 13 screening examinations had been completed and 11 were scheduled. These completed and scheduled examinations are in addition to the 139 former federal workers that FOH screened after the WTC Federal Responder Screening Program resumed because FOH thought they were current federal workers. A key challenge in providing screening examinations to former federal workers has been that a large number do not reside in the New York metropolitan area, where the worker and volunteer WTC program is based. The 283 former federal workers who have registered for screening examinations reside in 40 states, and about 240 of them reside outside the New York metropolitan area. NIOSH officials said that making arrangements to screen these widely dispersed responders has presented challenges, such as ensuring that the arrangements comply with federal privacy protections. NIOSH is negotiating with the Association of Occupational and Environmental Clinics (AOEC) in an effort to establish a national network of providers to screen these federal workers. CDC plans to award the $75 million appropriated for screening, monitoring, and treatment to the five organizations that the law identified as having priority for funding. CDC officials expect to make awards to the WTC Health Registry, Project COPE, and the POPPA program over a 3-year period and to award funds to the FDNY WTC and worker and volunteer WTC programs in response to their treatment costs. CDC officials have a proposed spending plan but told us that because they are uncertain about how quickly treatment costs could deplete the available funds, they may need to make adjustments. Officials from the FDNY WTC and worker and volunteer WTC programs told us that they expected that their estimated portion of the appropriated funds would be depleted well before the end of 3 years. As of August 2006, CDC awarded about $4.5 million of the $75 million—about $1.9 million to the WTC Health Registry, $1.5 million to the FDNY WTC program, and almost $1.1 million to the worker and volunteer WTC program. In addition, CDC expects to award $1.5 million to the POPPA program and $3 million to Project COPE in September 2006. CDC is waiting to make further awards until agency officials have reached certain decisions about the coverage of treatment services, such as which prescription drugs would be covered in the FDNY WTC and worker and volunteer WTC programs. CDC expects to begin making further awards around February 2007. CDC has decided to award the $75 million for screening, monitoring, and treatment that was appropriated to the agency in December 2005 to the five organizations identified as having first priority for funding. The organizations to which CDC plans to provide funds are the FDNY WTC program, for monitoring and treatment; the worker and volunteer WTC program, for monitoring and treatment; the WTC Health Registry, for monitoring; Project COPE, for treatment; and the POPPA program, for treatment. CDC plans to make awards through cooperative agreements with the programs. In general, it plans to send letters to the organizations inviting them to submit applications for funding; the applications would then undergo a two-stage peer review process. At the first stage a panel of outside experts would assess the merit of the application, and at the second stage CDC officials would determine the amount of funding the applicant would receive. CDC has made preliminary decisions about how to allocate the $75 million among the five organizations. As of September 1, 2006, CDC’s proposed spending plan indicated that awards would be made in the following way: $53.5 million for treatment and $8 million for monitoring, to be divided between the FDNY WTC and worker and volunteer WTC programs; $9 million for the WTC Health Registry; $3 million for Project COPE; and $1.5 million for the POPPA program. CDC officials expect to make awards to the WTC Health Registry, Project COPE, and the POPPA program over a 3-year period, but are not sure over what period they will make awards to the FDNY WTC and worker and volunteer WTC programs. A CDC official told us that the agency would award funds to the latter two programs in response to the treatment costs they incur. He said that agency officials are uncertain about how quickly treatment costs could deplete the available funds, because CDC does not know how many additional people will seek monitoring and what the extent of their treatment needs will be. For example, previous media reports about illnesses diagnosed in responders have resulted in increases in responders seeking examinations. Officials from the FDNY WTC and worker and volunteer WTC programs told us that they expected that their estimated portion of the appropriated funds would be depleted well before the end of 3 years. CDC has developed a proposed spending plan that indicates that about 36 percent of the funds would be awarded by the end of fiscal year 2007 and about 63 percent would be awarded during fiscal year 2008, although a CDC official told us that, depending on the extent of treatment needs, the funds could be used more quickly. The current plan is based in part on an agreement CDC made with the American Red Cross in April 2006. According to a CDC official, under this agreement, American Red Cross funds would be used for the treatment services that are eligible for American Red Cross support—such as basic clinical examinations and certain tests—for as long as such funds are available and the CDC funds would be used to cover other program expenses—such as infrastructure costs, more sophisticated diagnostic tests, and the conversion of medical records into an electronic format. As of August 2006, CDC had awarded a total of about $4.5 million of the $75 million to the WTC Health Registry, FDNY WTC program, and worker and volunteer WTC program. According to CDC officials, the WTC Health Registry applied for about $1.9 million in April 2006 for continuation of its collection of health data, and CDC awarded the registry $1.9 million in May 2006 and about $56,000 in July 2006. On August 10 and 11, 2006, respectively, the worker and volunteer WTC and FDNY WTC programs submitted applications to CDC for funds related to treatment services. In response to these applications, CDC made what an agency official termed emergency awards to the FDNY WTC and worker and volunteer WTC programs on August 11, 2006. CDC provided $1.5 million to the FDNY WTC program for leasing treatment space that previously had been provided by New York City at no cost. CDC provided almost $1.1 million to the worker and volunteer WTC program to hire an additional physician to help reduce the 3- to 4-month waiting time for treatment appointments that recently developed at the Mount Sinai clinical center, as well as to hire three administrators and a medical assistant. Officials from the clinical center told us that this waiting time had developed because additional people were seeking monitoring due to media reports about illnesses diagnosed in responders and because the proportion of responders who needed to be referred for treatment had increased. In addition to having awarded about $4.5 million, CDC plans to award an additional $4.5 million in September 2006. In spring 2006, CDC invited Project COPE and the POPPA program, two programs that provide mental health services to members of the NYPD, to apply for funding through a peer review process. In their applications, the POPPA program requested $1.5 million over 3 years, and Project COPE requested funding of $3 million over 3 years. CDC received their applications in June and July, respectively, and plans to implement the application review process in time to be in a position to make awards in September 2006. CDC does not plan to award additional funds from the $75 million to the FDNY WTC and worker and volunteer WTC programs until it makes certain decisions about the coverage of treatment services. These decisions include determining which medical conditions will be covered; developing a prescription drug formulary, that is, the list of drugs that will be covered; and determining the extent to which inpatient care will be covered. CDC officials said that they expected to make the coverage decisions in late 2006 and that they would obtain input from the American Red Cross and the programs. A CDC official told us that making decisions about which prescription drugs to cover could be the greatest challenge CDC and the programs face, because of the potentially high cost of drugs needed to treat responders. An FDNY WTC program official said that prescription drug costs are a looming financial problem for the FDNY WTC program. The CDC official told us that the most common diagnoses of WTC responders— gastroesophageal reflux disease, obstructive pulmonary disease, and mental health conditions—frequently are treated with prolonged and expensive drug therapy. For example, medications for respiratory therapy can cost $1,000 a month and may continue for a year. The FDNY WTC program official estimated that 100 percent coverage of prescriptions for firefighters and emergency medical technicians could cost $10 million to $18 million per year and potentially consume all of the funding that CDC would provide to the program. Clinicians at the worker and volunteer WTC clinical center at Mount Sinai stated that spending on prescription drugs at their center was increasing by $5,000 to $10,000 each month and amounted to $60,000 in July 2006. Another coverage decision that CDC faces is to determine the extent to which inpatient care will be covered. Currently, the FDNY WTC and worker and volunteer WTC programs provide only outpatient care, but officials involved with these programs believe that the treatment funds from the $75 million should cover some inpatient care, such as when a responder’s WTC-linked asthma becomes exacerbated to an extent that requires hospitalization. CDC officials told us that they plan to reach decisions about treatment coverage in fall 2006. They also plan to invite the FDNY WTC and worker and volunteer WTC programs to submit applications for treatment funding in the fall. If the applications are submitted by December 2006, CDC officials expect to be able to review them in time to provide funding to the programs by February 2007. CDC is also in the process of resolving issues related to providing access to screening, monitoring, and treatment services for WTC responders, including former federal workers, who reside outside the New York metropolitan area. CDC is negotiating with AOEC about possibly using AOEC clinics around the country to provide these services. CDC officials told us they intend that monitoring and treatment services available to responders around the country would be comparable to services provided by the worker and volunteer WTC program. Mr. Chairman, this completes my prepared remarks. I would be happy to respond to any questions you or other members of the subcommittee may have at this time. For further information about this testimony, please contact Cynthia A. Bascetta at (202) 512-7101 or bascettac@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Helene F. Toiv, Assistant Director; Fred Caison; Anne Dievler; Keyla Lee; and Roseanne Price made key contributions to this statement. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Responders to the World Trade Center (WTC) attack--individuals involved in rescue, recovery, or cleanup--included New York City Fire Department (FDNY) personnel, federal government workers, and others from New York and elsewhere. They were exposed to numerous hazards, and concerns remain about the long-term effects on their physical and mental health. In February 2006, GAO testified that four of the five key federally funded programs that were monitoring health effects in responders had made progress but that the Department of Health and Human Services' (HHS) WTC Federal Responder Screening Program, implemented by the Office of Public Health Emergency Preparedness (OPHEP), lagged behind (GAO-06-481T). GAO also reported that the Congress appropriated $75 million in December 2005 to HHS's Centers for Disease Control and Prevention (CDC) for monitoring and treatment for responders and that CDC was deciding how to allocate the funds. This statement updates GAO's February 2006 testimony. GAO examined (1) progress made by HHS's WTC federal responder program and (2) actions CDC has taken to award the $75 million appropriated. GAO reviewed program documents and interviewed HHS officials and others involved in WTC monitoring and treatment programs. The WTC federal responder program has registered and screened additional federal responders since February 2006, and arrangements are being developed to screen responders who are former federal workers residing outside the New York area. An additional 1,385 federal responders have registered for screening, including 1,134 current federal workers and 251 former federal workers, bringing the total number registered as of late August 2006 to 1,762, including 283 former federal workers. Because the total number of federal responders is uncertain, the proportion of the total who have registered is unknown. As of late August 2006, Federal Occupational Health Services (FOH) had completed screening of 907 federal workers, 380 of whom were screened since February 2006. Under an OPHEP agreement with CDC's National Institute for Occupational Safety and Health (NIOSH), former federal workers are being screened through the worker and volunteer WTC program, one of the five key federally funded programs. As of July 31, 2006, the worker and volunteer WTC program provided screenings to 13 former federal workers and scheduled 11 more, and 139 former workers had been screened by FOH as part of the 907 workers. Most of the former federal workers reside outside the New York area, where the worker and volunteer WTC program is located, and NIOSH is working to establish a national network of providers to screen these workers. CDC has awarded a small portion of the $75 million appropriated for screening, monitoring, and treatment and plans to make decisions about treatment coverage before awarding most of the funds. The agency plans to award the $75 million to the five organizations that the law identified as having priority for funding. CDC officials expect to make awards to the WTC Health Registry, the Police Organization Providing Peer Assistance (the POPPA program), and the New York City Police Foundation's Project COPE over a 3-year period and to award funds to the FDNY WTC and worker and volunteer WTC programs in response to the treatment costs they incur. CDC officials have a proposed spending plan that allocates about $53.5 million for the latter two programs' treatment costs, but the officials told GAO that because they are uncertain about how quickly treatment costs could deplete the available funds, they may need to make adjustments. Officials from the FDNY WTC and worker and volunteer WTC programs told GAO that they anticipated that their estimated portion of the funds would be depleted well before the end of 3 years. As of August 2006, CDC awarded about $4.5 million of the $75 million: about $1.9 million to the WTC Health Registry, $1.5 million to the FDNY WTC program, and almost $1.1 million to the worker and volunteer WTC program. In addition, CDC expects to award $1.5 million to the POPPA program and $3 million to Project COPE in September 2006. CDC is waiting to make further awards until it has reached certain decisions about the coverage of treatment services, such as which prescription drugs would be covered. CDC expects to begin making further awards around February 2007.

In its role as the nation’s tax collector, IRS is responsible for collecting taxes, processing tax returns, and enforcing the nation’s tax laws. Since 1990, we have designated IRS’s enforcement of tax laws as a governmentwide high-risk area. In attempting to ensure that taxpayers fulfill their obligations, IRS is challenged on virtually every front. While IRS’s enforcement workload—measured by the number of tax returns filed—has continually increased, only recently have the resources IRS has been able to dedicate to enforcing the tax laws begun to increase. IRS estimates that the annual gross tax gap, that is, the difference between what taxpayers should pay on a timely basis and what they actually pay, is about $345 billion. IRS has reported that its enforcement activities, coupled with late payments, recover about $55 billion of that amount, leaving an annual net tax gap of almost $300 billion. IRS has a statutory limitation on the length of time it can pursue unpaid taxes, generally 10 years from the date of the assessment. The amount of cumulative outstanding tax debt that IRS has identified either through taxpayer reporting or through its various compliance programs is also substantial. As of September 30, 2005, IRS’s master file database of taxpayer accounts reflected about $250 billion in cumulative outstanding taxes owed by businesses and individuals. The amount of unpaid taxes ranges from small amounts owed by individuals for a single tax period to millions of dollars owed by businesses. The taxes owed include individual income, corporate income, payroll, and other types of taxes, as shown in figure 1. As a part of its tax administration, IRS maintains over 24 million separate tax debt account records in its master file database for businesses and individuals. Within the master file database, IRS records collection actions and the current status of tax debts through a series of codes. The codes, referred to as status or transaction codes, display a host of information, including the stage of the collection process the tax debt is in; the capacity of a tax debtor to pay, such as whether a tax debtor is considered to be experiencing financial hardship; or other data such as whether the tax debtor is under an arrangement with the IRS to pay the tax debt in installments. IRS uses these codes to monitor and manage its inventory of outstanding tax debt and its tax collection efforts. To improve the collection of unpaid taxes, the Congress, in the Taxpayer Relief Act of 1997, authorized IRS to collect delinquent tax debt by continuously levying (offsetting) up to 15 percent of certain federal payments made to tax debtors. The payments include federal employee retirement payments, certain Social Security payments, selected federal salaries, and contractor and other vendor payments. Subsequent legislation increased the maximum allowable levy amount to 100 percent for payments to federal contractors and other vendors for goods or services sold or leased to the federal government. The continuous levy program, now referred to as the Federal Payment Levy Program (FPLP), was implemented in 2000. Under the FPLP, each week IRS sends the Department of the Treasury’s Financial Management Service (FMS) an extract of its tax debt files. These files are uploaded into the Treasury Offset Program. FMS sends payment data to this offset program to be matched against unpaid federal taxes. The program electronically compares the names and taxpayer identification numbers on the payment files to the control names (first four characters of the names) and taxpayer identification numbers of the debtors listed in the offset program. If there is a match and IRS has updated the weekly data sent to the offset program to reflect that it has completed all statutory notifications, the federal payment owed to the debtor is reduced (levied) to help satisfy the unpaid federal taxes. In creating the weekly extracts of tax debt to forward to FMS for inclusion in the offset program, IRS uses the status and transaction codes in the master file database to determine which tax debts are to be included in or excluded from the FPLP. For example, IRS cannot levy the assets of individuals and businesses to recover tax debts while the tax debtor is involved in a bankruptcy proceeding. In such cases, IRS uses the bankruptcy status code in the master file to block the tax debt from being submitted to the FPLP. Under other circumstances, IRS collection personnel can enter a transaction code into the tax debtor’s tax account to block the debt from being levied through the FPLP. Consequently, the accuracy and appropriateness of status and transaction codes is vital to the effective operation of the FPLP. We reported in 2004 that incorrect or out-of-date IRS status and transaction codes in IRS’s records had inappropriately blocked delinquent tax debt from being referred to the FPLP. IRS currently excludes 62 percent of all tax debt from the FPLP because of either statutory or policy reasons. As shown in figure 2, at September 30, 2005, IRS excluded over $73 billion (29 percent) from the FPLP for statutory reasons and about $82 billion (33 percent) for policy reasons. Cases excluded from the FPLP for statutory reasons include tax debt that had not completed IRS’s notification process, or tax debtors who filed for bankruptcy protection or other litigation, who agreed to pay their tax debt through monthly installment payments, or who requested to pay less than the full amount owed through an offer in compromise. Cases excluded from the FPLP for policy reasons include those tax debtors whom IRS has determined to be in financial hardship, those filing an amended return, certain cases under criminal investigation, and those cases in which IRS has determined that the specific circumstances of the cases warrant excluding it from the FPLP. Since the inception of the FPLP, we have identified numerous issues that have impeded the levy program from achieving its full potential. In response to many of the issues we raised, IRS and other agencies have made numerous improvements to the levy program that have contributed to increased tax collections. IRS and FMS officials, along with Department of Defense, General Services Administration, Office of Management and Budget, and Department of Justice officials, created a multiagency task force—referred to as the Federal Contractor Tax Compliance Task Force—in 2004, primarily to address the issues raised in our 2004 report related to defense contractors and the FPLP. The multiagency nature of the task force reflected that the involvement of several agencies was required for the FPLP to reach its full potential. The task force, which has now become a semipermanent body, has worked toward its stated goals and, along with the efforts of the individual agencies, has been instrumental in making significant improvements in the program. For example, the task force has achieved its goal of adding most of the Department of Defense’s payment systems to the FPLP. IRS, in conjunction with the task force, has made several policy changes directed toward increasing the amount of unpaid tax debt that it is submitting to the FPLP. For example, IRS altered its policies to include the following tax debt in the levy program that had previously been excluded: cases waiting in a “queue” to be actively worked by an IRS collections official—formerly IRS blocked such cases from the FPLP for a year each time a case entered the queue; nearly half of the cases assigned to its Automated Collection System most cases in the field that are being worked by an IRS revenue officer; cases that have low dollar balances and cases for which the IRS has been unable to locate or contact the tax debtor. IRS has also worked with FMS to improve the process of matching tax debtor names between FMS’s payment files and IRS’s tax debt files to increase the number of payments and debts that are matched. This work was important because the FPLP relies on matching both the tax identification number and the control name in the payment to those in the tax files to identify a federal payment for levy. The FPLP has proved to be a cost-effective means of collecting outstanding tax debt from tax debtors who receive payments from the federal government, and the improvements IRS and other agencies have made in the program have significantly increased tax collections since 2003, as shown in figure 3. Although the FPLP collected almost $300 billion dollars in previously unpaid taxes during fiscal year 2006, the program has an even greater effect on total tax collections. In previous reports, we have estimated that IRS collects three times the amount of the direct levy collections through voluntary revenues received as a result of taxpayers responding to IRS’s notice that their federal payments would be levied. To maximize the effectiveness of the FPLP as a tool to collect outstanding federal taxes, it is crucial that IRS record and maintain accurate status codes for all tax debt within its systems. To test the accuracy of the codes, we selected statistical samples of tax debt excluded from the FPLP for both statutory and policy reasons to determine if these status codes appropriately reflected the current condition of the tax debt. Our testing of IRS’s exclusion codes consisted of samples of 100 tax debts excluded for statutory reasons, and 100 tax debts excluded for policy reasons as of September 30, 2005. While our review of the sample of tax debts excluded for policy reasons did not identify a significant number of coding errors that would affect the FPLP, our review of the sample of tax debts excluded for statutory reasons did. On the basis of our samples, we estimate that over a half-million tax records with over $2.4 billion in tax debt were erroneously excluded from the FPLP. At September 30, 2005, IRS had about $73 billion of outstanding tax debt associated with about 9 million tax records that were excluded for statutory reasons. As shown in figure 4, these tax records were almost exclusively in four statutory exclusion categories: notice, bankruptcy, offers in compromise, and installment agreements. In reviewing the 100 tax records coded as statutorily excluded tax debt, we identified six instances in which the records were incorrectly coded. Table 1 presents the number of errors we found by exclusion category. As indicated in table 1, four of the errors we identified involved tax debtors erroneously coded as paying on an installment agreement and thus excluded from the FPLP. In each of the four cases, IRS had not terminated the installment agreement within 5 months after the tax debtor stopped making agreed-to payments. Although IRS’s guidance on the installment agreement termination process does not contain a specific time frame, 5 months is the minimum amount of time that would elapse if IRS’s Internal Revenue Manual requirements on terminating installment agreements were laid out in a timeline. In one of the cases, IRS took 23 months to terminate the agreement after the tax debtor had stopped making payments. One error involved tax debt that had been erroneously kept in IRS’s notice phase. The notice phase is IRS’s first phase in the tax debt collection process and consists of a series of letters IRS sends to tax debtors informing them of the tax debt and requesting payment. Each letter is represented by a specific status code. The one error we identified in this exclusion category resulted when an IRS computer programming change in 2005 inadvertently blocked certain status codes from being updated and thus prevented the related tax debt from exiting the notice phase. IRS personnel took action to correct this systemic error after we informed them of the issue. We also found one bankruptcy-related case erroneously excluded from the FPLP due to IRS failing to reverse a bankruptcy transaction code after the bankruptcy had ended. According to IRS officials, IRS’s time frame for initiating action to reverse a bankruptcy code is 30 days after bankruptcy actions have been completed. However, in this case, the bankruptcy had ended almost a year before the time of our review, yet IRS had not updated the status code in the tax debtor’s account. IRS reversed the bankruptcy code after we informed IRS personnel of the issue. However, as a result of the error, the tax debt had been erroneously excluded from the FPLP and all other collection action for almost a year. We found no errors in the status codes for the five offer in compromise cases we reviewed. In total, the errors we found in the sample of tax records excluded for statutory reasons equate to a 6 percent projected error rate. As a result of these errors, we estimate that over a half-million tax records containing about $2.4 billion in uncollected tax debt were erroneously excluded from the FPLP. At September 30, 2005, IRS had about $82 billion of outstanding tax debt associated with about 7 million tax-period records that were excluded for policy reasons. As shown in figure 5, tax records were excluded primarily for three reasons: cases designated as financial hardship, cases currently in or awaiting assignment to IRS’s collection function, and cases designated as currently not collectible for reasons other than financial hardship, including low-dollar cases. IRS is authorized to exclude tax debt from the FPLP based on a policy determination of financial hardship. Tax debt in the other two categories is excluded on a case-by-case basis. In other words, the categories, themselves, are not explicitly excluded from the FPLP, but individual cases in those categories may be excluded by IRS personnel based on the circumstances of the particular case. For example, cases that are in IRS’s field collection status being worked by a revenue officer are generally eligible for the FPLP; however, the revenue officer can block the tax debt from inclusion in the FPLP when the officer determines that pursuing other collection actions may be more effective. In reviewing the 100 tax records coded as excluded for policy reasons, we identified one instance in which the records were incorrectly coded. Table 2 presents the results of our review of the sampled cases. The one coding error we found involved a tax debtor who defaulted on an offer in compromise, but IRS did not put the tax debt into the FPLP. Although IRS correctly coded the tax debtor as having defaulted on the agreed-to payment terms of the offer, IRS’s system had not been programmed to reverse the original “pending” code that IRS personnel placed in the tax debt record while IRS was considering the tax debtor’s offer. Even though the tax debtor had defaulted on the offer, the unreversed pending offer code continued to exclude the case from the FPLP. On the basis of our finding, IRS implemented a computer programming change to reverse existing pending codes for defaulted offer in compromise cases. We found no errors in the cases that IRS had designated as currently not collectible for reasons other than financial hardship. Although IRS is not going to actively seek collection from them, these cases are generally included in the FPLP. However, IRS tax collections personnel can exclude these cases from the FPLP on a case-by-case basis. Our review of the limited data IRS retains related to financial hardship cases and our own review of the tax debtor’s financial condition using available IRS information and outside data sources did not identify any cases in which we believe IRS had erroneously coded a tax debtor as being in financial hardship. However, as discussed later, we do believe that IRS’s existing processes increase the risk that outdated status codes related to financial hardship cases could occur and not be detected. In total, the errors we found in the sample of tax records excluded for policy reasons equate to a 1 percent projected error rate. IRS’s current monitoring of the ongoing status of accounts did not identify and correct the errors in our sample. In addition, although we found no errors in the coding of financial hardship cases, our analysis revealed that the design of IRS’s policies for monitoring the status of such cases is not sufficient to ensure the ongoing accuracy of hardship designations. The coding errors we identified in our samples of tax debts excluded from the levy program for statutory and policy reasons could have been avoided if IRS had more effectively monitored the ongoing status of accounts to detect and prevent delays in putting tax debt into the FPLP. In the one case from our sample of statutory exclusions involving a bankruptcy-related coding error, the transaction code blocking the case from inclusion in the FPLP was not reversed within IRS’s stated time frame. IRS policy is that bankruptcy codes should be reversed within 30 days after a bankruptcy judge has dismissed the case. In such cases, the tax debtor again becomes liable for repaying the tax debt. IRS did not reverse the bankruptcy code in a timely manner because the case was repeatedly transferred to different IRS personnel without anyone taking action to reverse the code. As a result of confusion caused by the repeated transfer of the case within IRS and no one person having responsibility for monitoring the disposition of the case, IRS did not recognize that the bankruptcy code had not been reversed until we notified IRS officials during our review of the case. In the four coding errors we identified involving installment agreement cases, the errors were caused by a computer programming problem— corrected in January 2006—that prevented the installment agreement codes from automatically reversing within IRS’s systems. Generally, IRS’s computer systems automatically begin the process to reverse an active installment agreement code after a tax debtor fails to make two scheduled monthly payments, but that did not happen in these cases. IRS officials were unable to determine specifically why this occurred, and stated that they do not monitor whether installment agreement transaction codes are reversed within the 5-month time frame indicated by IRS’s Internal Revenue Manual for terminating installment agreements. Until the installment agreement code is reversed in the system, the tax debt remains excluded from the FPLP. Had IRS been monitoring the timely termination of installment agreements, these cases would have come to IRS’s attention and afforded it an opportunity to investigate the cause. Two coding errors—one statutory exclusion case and the other a policy exclusion case—were also caused by deficiencies in IRS’s computer programs. In the statutory exclusion case, the tax debt did not automatically move through the notice process because IRS did not include one of its several notice status codes in a computer programming change. As a result, when the programming change was implemented, the existing cases in that notice status were prevented from automatically continuing their movement through the notice phase and into collection. As a result, these cases remained excluded from the FPLP and other collection actions. After we brought this case to IRS’s attention, it took corrective action to address this programming deficiency. In the second case, involving a policy exclusion related to a defaulted offer in compromise, IRS continued to exclude the tax debt from the FPLP because, although IRS personnel properly entered a code indicating a default on an offer in compromise, IRS’s systems did not reverse the code indicating the case had an initial pending offer. IRS had recently implemented a programming change to the way it processes offer in compromise-related transaction codes so that the code that reverses an active offer in compromise transaction code also reverses any pending offer in compromise codes related to the same tax case. However, the programming change only affected offer in compromise cases occurring subsequent to the date the change was implemented; it did not affect pending offer codes that existed prior to the programming change. After we notified IRS personnel of the error we identified, they took corrective action to reverse the status code in this and similar cases. As discussed earlier, in our sample of tax debt cases excluded from the FPLP for policy reasons, we found that all 31 cases that were excluded due to the tax debtor being designated by IRS as being in financial hardship were correctly coded based on IRS’s existing policy and our review of the limited documentation IRS maintained regarding the tax debtor’s income. However, we found deficiencies in IRS’s procedures for monitoring the ongoing status of financial hardship cases, which hinders IRS’s ability to ensure the ongoing accuracy of the financial hardship designation. This, in turn, could result in additional tax debt that should be eligible for levy not being forwarded to the FPLP. To make the determination of whether a tax debtor is facing financial hardship and thus does not have the means to pay the tax debt, IRS analyzes the tax debtor’s financial condition using guidelines for allowable costs. On the basis of these guidelines, IRS officials place individuals in one of nine income categories, or income thresholds. These thresholds represent income level ceilings above which the tax debtor again becomes subject to IRS’s collection actions, including forwarding of the tax debt to the FPLP. Once IRS designates a tax debtor as being in financial hardship, it performs an automated evaluation of the debtor’s income based upon their annual tax return filings. Specifically, IRS compares the income reported on the tax debtor’s tax return to the threshold level assigned to the tax debtor. If the reported income exceeds the threshold, the financial hardship designation is terminated and the tax debt becomes subject to collection and can be put into the FPLP. IRS policy discourages any other monitoring or follow-up of financial hardship cases beyond the automated match. IRS does not routinely update the tax debtor’s allowable expenses or perform a periodic review— such as once every 3 years—of the tax debtor’s overall financial condition. In fact, the Internal Revenue Manual directs IRS personnel working financial hardship cases to not request future follow-up reviews to check on compliance with future income tax filing requirements or to update a tax debtor’s financial condition. Consequently, IRS relies only on the accuracy of the information reported in the tax return filed by the tax debtor, with no review of income information reported to IRS by third parties, such as Form W-2 and Form 1099 information reports, to assess the ongoing accuracy of hardship designations. IRS’s procedures do not require it to remove a tax debtor from the financial hardship status if the tax debtor fails to file a tax return, and failing to file does not flag the case for IRS personnel to perform a review of the financial hardship designation. Because of its monitoring policy, when a tax debtor with a financial hardship designation does not subsequently file an annual income tax return, IRS has no means of determining whether the tax debtor’s financial condition has improved and the hardship designation should be terminated. Since individuals designated as being in financial hardship are excluded from the FPLP—as well as all other tax collection actions—not knowing whether the hardship designation remains valid can result in IRS inappropriately excluding the tax debt from the FPLP. Generally, individuals with a financial hardship designation who do not file a tax return are treated like other nonfilers: they can be eventually subject to review by IRS’s automated substitute-for-return process. In that review, IRS examines other available data on the taxpayer, assesses whether a tax return should have been filed, and estimates the amount of tax due. However, that process generally does not occur for more than a year after the failure to file, and only individuals who meet certain criteria are reviewed. A financial hardship designation is not one of the criteria and, therefore, these cases do not have a high priority. On the basis of our review of the sample cases, ceasing to file is not an uncommon occurrence for tax debtors with hardship designations. Twenty-four of the 31 tax debtors designated as financial hardship in our sample cases had ceased filing tax returns after IRS had determined the tax debtor was in financial hardship. IRS’s current practices also do not prevent tax debtors with a financial hardship designation from accumulating additional tax debt by not paying their current taxes. A financial hardship designation puts a tax debtor’s past debt in abeyance, but the hardship designation does not automatically exempt the tax debtor from paying current taxes. However, we found that IRS does little to prevent the further accumulation of tax debt by these tax debtors. Of the 31 tax debtors with financial hardship designations in our sample cases, we found that 4 filed but had not paid income taxes subsequent to being identified as a financial hardship case. As with not filing a tax return, accumulating new tax debt does not cause IRS to automatically terminate the financial hardship designation, and IRS’s procedures allow IRS personnel to include the newly acquired tax debt into the hardship designation, sometimes without any additional analysis of the tax debtor’s financial condition. Thus, a tax debtor’s ever-increasing tax debt can remain excluded from the FPLP as well as other collection actions. The effect of IRS’s collection policy regarding financial hardship tax debtors who accumulate new debt is essentially to both cease collection of old debt and not require tax debtors to pay the current taxes they owe. Allowing such tax debtors to continually not pay current taxes without consequence appears to be giving tax debtors with financial hardship designations an additional exemption from paying current taxes as well as old tax debt and may contribute to the noncompliance of other taxpayers. In fiscal year 2006, IRS initiated a withholding compliance program that has potential to help prevent wage-earning tax debtors from accumulating more unpaid tax debt. The program is designed to identify individuals who incur tax debt because they did not have their employer withhold sufficient wages to cover their taxes due for the current year. The program identifies those debtors and requires their employers to increase the withholdings. However, due to resource constraints, IRS actively pursues only a small portion of the tax debtors who underwithhold. Additionally, while the program prioritizes cases for review, a financial hardship designation is not a prioritization criterion. IRS has significantly improved the effectiveness of the FPLP by making an additional $28 billion in unpaid tax debt eligible for the program since 2004. However, certain changes in IRS’s policies could result in additional billions of dollars in tax debt entering the levy program for potential collection or entering the program earlier. Under current IRS policy, all tax debt for which the debtor is designated as being in financial hardship, including those debts associated with tax debtors earning relatively high income levels, are excluded from the levy program. In addition, half of the cases in IRS’s ACS are excluded from the program, as are all cases throughout IRS’s notification process. IRS has established policies that allow it to designate tax debtors earning up to $84,000—nearly twice the national median income of about $44,000—as being in financial hardship. IRS is authorized to grant tax debtors a designation of financial hardship when collection of the tax debt would cause the tax debtor to be unable to pay his or her reasonable basic living expenses. IRS’s Internal Revenue Manual provides examples of financial hardship cases, such as a disabled taxpayer who lives in a modest house that has been equipped to accommodate the disability and whose fixed income is not sufficient to both meet his or her living expenses and pay the tax debt. IRS has the authority to determine allowable living expenses according to the unique circumstances of individual tax debtors; however, unique circumstances do not include the maintenance of an affluent or luxurious standard of living. Once designated as being in financial hardship, the tax debtors are excluded from the FPLP and are also exempt from any other IRS collection action until their self-reported income rises above one of nine designated income thresholds. Since 1992, IRS has almost tripled the income it allows tax debtors in financial hardship to earn without pursuing collection, but IRS does not have documentation of any data analysis that justified the large increases. Consequently, as of September 30, 2005, about 65 percent of the tax debt in the financial hardship category was owed by tax debtors who were allowed to earn more than the national median income before being subject to collection actions. Of the $247 billion total tax debt in IRS’s records, IRS is not pursuing collection of almost 10 percent of that amount—$22.6 billion—as a result of its financial hardship determinations. As discussed previously, IRS makes a determination as to whether a tax debtor qualifies as a financial hardship case based on an analysis of the amount of income earned and the allowable expenses owed by the tax debtor. If IRS determines that a tax debtor is unable to pay the outstanding tax liability due to financial hardship, it places a financial hardship transaction code in the tax debtor’s account. The transaction code is assigned one of nine subcodes (called closing codes) indicating the income threshold level ceilings at which IRS has determined that the tax debtor should be able to begin repaying the tax debt. Tax debtors will not face any IRS collection action until their total positive income—roughly equivalent to adjusted gross income—exceeds the designated income threshold ceiling. IRS’s financial hardship income thresholds range in $8,000 increments from $20,000 to $84,000, as depicted in table 3. Five of the nine income thresholds included in IRS’s financial hardship designation have upper range ceilings above the 2004 national median household income of $44,389. Of the approximately 1.8 million tax debt records designated as financial hardship in IRS’s unpaid assessments database at September 30, 2005, approximately half—about 900,000—were debts owed by tax debtors in one of the five income threshold categories above the national median. Over 286,000 tax records—with associated tax debt of about $6.4 billion—were for tax debtors in the top income level threshold for financial hardship of up to $84,000. The exclusion of tax debt from collection actions may be appropriate in many circumstances to provide relief for those experiencing financial difficulty. However, as shown in figure 6, $14.8 billion in tax debt (65 percent of the tax debt) in financial hardship is owed by tax debtors whom IRS will allow to earn more than the national median household income before they have to begin repaying their tax debt. As shown in table 4, IRS’s income thresholds used to determine whether tax debtors are experiencing financial hardship and therefore cannot currently pay their outstanding tax debt have not always been this high. IRS significantly increased each of the nine income thresholds in 1997 and again in 2004. IRS had previously set rates in 1992. The 2004 increases averaged 77 percent but the individual threshold increases ranged from 100 percent for the lowest threshold to 68 percent for the highest. In justifying the large increases from previous threshold ceilings, IRS stated that the new 2004 thresholds more accurately reflected current economic conditions and that the new values were supported by Bureau of Labor Statistics data and were consistent with the allowable expenses in IRS guidance. IRS also stated that the revised income thresholds were based on an analysis of allowable expense standards for high-cost geographic areas considered in conjunction with current Bureau of Labor Statistics poverty levels. Though it raised the top threshold to $84,000, IRS had considered raising its top threshold for financial hardship to $100,000. Other than the above statements, IRS could not provide documentation of any data analysis that supported its reasons for the large increases since 1992. However, measures of median income raise questions about the size of the increases to the income thresholds for financial hardship determinations. As table 5 depicts, IRS’s increases in the financial hardship income thresholds has had the effect of raising the maximum income threshold from about equivalent to the national median income in 1992 to almost twice the median income in 2004. With respect to high-cost areas, New Jersey’s $61,359, the highest state median income in 2004, was well below IRS’s 2004 top three income threshold levels. The lowest state median income in 2004 was $31,500. As a result of these large increases, almost two-thirds of all tax debtors with IRS financial hardship designations are allowed to earn more than the national median household income in 2004 before their unpaid tax debt again becomes subject to IRS collection action. In contrast, in 1992, no tax debtor with a financial hardship designation was allowed to earn more than the median income without becoming subject to collection action. Measures of inflation also raise questions about the size of IRS’s increases. Bureau of Labor Statistics national inflation rate data indicate that the effects of inflation would have justified lower increases. For example, using inflation data from 1997 to 2004, the top 2004 threshold would have been about $60,000, far below IRS’s $84,000 level, and would have kept the top threshold at roughly 35 percent above the national median income as it was in 1997. Exacerbating the effect of IRS’s increases in its hardship thresholds was the policy it used to implement the increases. IRS did not change the subcodes indicating the income threshold ceilings or reexamine the financial condition of tax debtors when it raised the income thresholds ceilings. For example, the IRS subcodes indicating that tax debtors were in the highest income threshold of $50,000 prior to the threshold increases were not updated to reflect the new thresholds. Thus the tax debtors remained in the highest income threshold and were allowed to earn up to $84,000 before IRS would begin taking collection action. Therefore, after the 2004 increases, the tax debtors in the top income threshold category were allowed to earn $34,000 more annually before IRS would remove the tax debtor from the financial hardship exclusion category and begin pursuing collection of the outstanding tax debt. IRS neither reassessed the financial condition of tax debtors with existing financial hardship designations nor changed their existing designation to one that closely matched their original income threshold amount. Allowing relatively high income tax debtors, such as those earning $84,000, to avoid tax collection action calls into question the fair application of the tax system and may contribute to noncompliance by other taxpayers. In addition, dramatically increasing the financial hardship income threshold ceilings has effectively resulted in increasing the number of tax debtors IRS classifies as being in financial hardship. This, in turn, reduces the portion of IRS’s inventory of tax debt under active collection and reduces the portion eligible for inclusion in the FPLP. Although IRS made policy changes in 2004 to allow about 40 percent of the tax debt in ACS to enter the FPLP, IRS continues to exclude the other 60 percent. The ACS is an automated call system designed to schedule and follow up on IRS’s outgoing calls to, and incoming calls from, tax debtors. ACS personnel’s primary activity is to contact tax debtors by phone to attempt to collect outstanding tax debt. At September 30, 2005, the ACS contained an inventory of about $8 billion of unpaid tax debt. To manage the large inventory of tax debt in ACS, IRS has divided the ACS inventory into 40 subcategories. In general, those subcategories describe the status of IRS collection actions within ACS, such as indicating that an installment agreement is pending or specifying a collection action that is awaiting approval by a supervisor. Prior to 2005, all the tax debt in ACS was excluded from the FPLP. In 2005, in response to issues raised in our 2004 review of Department of Defense contractors with outstanding tax debt, IRS changed its policies to allow some of the tax debt assigned to ACS to enter the FPLP. However, IRS has continued to exclude tax debt in 19 of the 40 ACS subcategories from the FPLP. Those 19 subcategories contain 60 percent, or about $5 billion, of the total tax debt in the ACS inventory. Two of the excluded subcategories, which IRS calls R-5 and I-6, contain approximately $3.9 billion of tax debt, and involve cases in which IRS is placing a levy against a tax debtor’s assets. These “paper” levies, as IRS refers to them to distinguish them from automated FPLP levies, are generally one-time levies placed against a tax debtor’s bank accounts or other financial assets, although they can also be an ongoing garnishment of wages. FPLP levies, in contrast, are continuous levies of all federal payments, including federal salaries, pensions, social security, and contractor-related payments. IRS has the authority to levy a tax debtor’s assets to collect outstanding tax debt. Therefore, simultaneously levying through both the paper levy process and the FPLP would seem to be appropriate, especially since many paper levies are one-time levies of a tax debtor’s assets. Additionally, the FPLP is a cost-effective means of collecting from tax debtors. By excluding tax debt from the FPLP while IRS personnel are working on a paper levy, IRS is relegating the FPLP to a secondary role in the tax collection process. Because of its potential, we have previously recommended that IRS use the FPLP as one of the first steps in the IRS collection process and keep the debt in the levy program until the taxes are fully paid. At September 30, 2005, IRS had excluded $25.1 billion from the FPLP because it was in the process of notifying the tax debtor of the taxes owed. The Internal Revenue Code prohibits IRS from levying a tax debtor’s assets, including doing so through the FPLP, until the tax debtor has been given time to respond to a notice from IRS that a tax debt exists. IRS’s process of issuing a series of notice letters and waiting for the tax debtor to respond can take 6 months for individuals. IRS excludes tax debt from the FPLP during the entire notice phase. For individuals, the notification process consists of sending a series of three or four computer-generated letters with increasingly urgent language notifying the tax debtor of the debt and requesting payment. Per the Internal Revenue Manual, IRS waits 5 weeks between letters and up to 10 weeks after the last letter before moving the tax debt into one of IRS’s active collection statuses such as ACS. Consequently, the notification process can take up to 6 months or longer to complete, during which time IRS excludes the tax debt from other collection activity, including the FPLP. Although IRS excludes all tax debt from the FPLP during the entire notice process, legally, tax debt could be included in the FPLP in about 3 months—about half way through the general notice process for individuals. IRS must allow the tax debtor 90 days after notification of a potential tax debt liability to respond. If IRS does not receive a response within that period, it can issue a notice of tax deficiency and demand for payment. If the tax debt is not paid within 10 days after notice and demand for payment, IRS can initiate the procedures for levy, including forwarding the tax debt to the FPLP. Under this scenario, IRS could forward tax debt to the FPLP about 14 to 15 weeks after the first notice is sent to the tax debtor, and IRS could fulfill its statutory requirement with only two notices before initiating the levy process. For business tax debt, IRS essentially follows this sequence. The notice process for business tax debt consists of only two notices and is generally completed in about 15 weeks, at which time the tax debt can be included in the FPLP. In addition to putting tax debt into the FPLP sooner in the overall tax collection process, IRS could potentially enhance the tax collection potential of notices by informing the tax debtor early in the process of sending notice letters that unpaid tax debt can be subject to levy. As shown in figure 7, about 70 percent of tax collections resulting from notice letters are received as a result of IRS’s first notice letter. Very little is collected from subsequent notices until the last notice letter, which includes specific language of IRS’s authority to levy or place a lien on the tax debtor’s property. The FPLP is a powerful tool for encouraging collection that goes beyond the direct taxes collected through federal payment levies. We have previously estimated that the threat of a levy brings in over three times more collections than the levy itself. IRS could take advantage of this potential during the notice phase if it were to inform tax debtors early in the notice process that their tax debt could be included in the FPLP. Although the collection of taxes is always important, it takes on added prominence in times of severe budgetary uncertainty. As the nation’s tax collector, IRS must seek out and utilize the most cost-effective means of collection at its disposal and apply those means to the broadest application of tax debt. The FPLP is a cost-effective program that has enabled IRS to greatly increase collection. The program’s full success is dependent on the accuracy of IRS’s status and transaction codes as well as the appropriateness of IRS’s policies, procedures, and practices regarding the exclusion of tax debt from the FPLP. Improvements are needed in both arenas. The errors we identified in the status and transaction codes of tax debt cases highlight potential problem areas that have led to tax debt being inappropriately excluded from levy action and therefore require IRS’s attention. With regard to its current policies, IRS continues to exclude over 60 percent of all tax debt from the FPLP and does not appear to have fully adopted our previous recommendation to use the FPLP as one of the first steps in the tax collection process. Viewing the FPLP as a primary and efficient collection tool could lead IRS to reevaluate its FPLP exclusion policies and to reduce the extent and length of time tax debt is excluded. Such changes hold the potential to subject billions of dollars in additional tax debt to the FPLP, thus increasing the government’s chances of collecting some of this tax debt. IRS faces tough challenges in balancing its tax collection activities against its available resources. In times of tough budgetary constraints, this can provide an incentive to close cases quickly or otherwise reduce the active tax collection inventory, possibly at the expense of maximizing tax collections. While reducing the number of active cases does, in fact, reduce the resources required, it can also have the effect of reducing collections, diminishing compliance, and eroding the public’s confidence in the fairness of the tax system. For instance, in financial hardship cases, beyond those tax debtors granted relief from paying tax debt due to unexpected financial difficulty, each tax debtor who is allowed to avoid filing required tax returns or paying current taxes, or who is perceived to live well while facing little tax collection consequence, represents not only less money for vital federal programs but one more advertisement for others to do the same. Therefore, in setting financial hardship or other tax collection policies, it is incumbent upon IRS to be particularly judicious in setting income threshold levels and monitoring tax debt to ensure that it is acting fairly toward all taxpayers. To increase the amount of tax debt eligible for, and to expedite the entry of tax debt into, the FPLP, we recommend that the Commissioner of Internal Revenue take the following actions: monitor the timely termination of defaulted installment agreements to help ensure tax debt is made available to the FPLP as soon as possible; place tax debt in the notice phase into the FPLP as soon as legally consider adding language to IRS’s second communication in the notice process informing the tax debtor that IRS has the authority to collect the debt by levying the tax debtor’s income and assets if the tax debt is not paid voluntarily; and modify FPLP exclusion policy to allow tax debt in ACS subcategories R-5 and I-6 that is being considered for a levy on financial assets through paper levies to be concurrently included in the FPLP. To help ensure that IRS’s financial hardship FPLP exclusions are appropriate, we recommend that the Commissioner of Internal Revenue take the following actions: reevaluate whether the dollar ranges for existing financial hardship income thresholds, especially those that exceed the national median income, are appropriate and reasonable; consider changing the financial hardship closing codes for tax debtors designated as being in financial hardship prior to the 2004 income threshold increases to a closing code that most closely corresponds to the originally designated income threshold—for example, tax debtors who were in a threshold of $50,000 prior to the change would be given a different subcode (closing code) so that the tax debtor’s income ceiling stays as close to the original $50,000 ceiling as possible under the new income thresholds; establish a policy so that in implementing future financial hardship income threshold changes, tax debtors’ financial hardship subcodes (closing codes) are changed to ones that maintain the tax debtor’s income ceiling as close as possible to the ceiling prior to the change; establish a policy to review tax debtors’ financial condition periodically to verify the continued validity of the financial hardship designation; evaluate the ongoing validity of the financial hardship designations whenever tax debtors fail to file their annual tax returns by comparing third-party income information to the tax debtors’ designated financial hardship income threshold ceilings; and refer tax debtors with a financial hardship designation to IRS’s withholding compliance program for special attention if those tax debtors do not pay their current income tax obligations. In commenting on a draft of this report, IRS noted improvements made to the FPLP and the extent to which such improvements have resulted in increased collections while at the same time ensuring that taxpayer rights have been protected. IRS also described several initiatives it had undertaken to improve its program for taxpayer accounts classified as currently not collectible, including a study to determine whether changes to IRS’s allowable living expense tables, used in the determination of financial hardship status, would be appropriate given the availability of additional economic data. We made 10 recommendations: IRS agreed with 5, partially agreed with 2, and disagreed with 3. We modified the 2 recommendations with which IRS partially agreed in order to address issues raised by IRS while retaining the intent of our recommendations. With respect to the four recommendations we made to either increase the amount of tax debt eligible for the FPLP or expedite the entry of tax debt into the FPLP, IRS agreed with one recommendation, partially agreed with another, and disagreed with the remaining two recommendations. IRS agreed with our recommendation that it monitor the timely termination of defaulted installment agreements, and noted it would identify those taxpayer accounts in installment agreement status but which show no payment activity within the last 60 days and determine if it needs to change the way it monitors installment agreements. While IRS disagreed with our recommendation that it add language about its legal authority to levy income and assets to its first notice letter, it stated that it would consider adding stronger language regarding possible enforcement activity in subsequent collection notices. As an explanation for its reluctance to include this course of action in the event of nonpayment, IRS noted that it had received criticism in the past for early aggressiveness and not affording taxpayers an opportunity to voluntarily satisfy their liability. While IRS is not legally precluded from providing language concerning its enforcement powers in the initial taxpayer notice, we understand IRS’s desire to attempt to provide sufficient opportunity for taxpayers to voluntarily comply with their tax obligations without threat of enforcement action. Accordingly, we have modified our recommendation to add informative language about IRS’s levy starting with the second taxpayer notice rather than the first. The important point to us is that IRS inform the tax debtor of its levy authority earlier in the process. IRS disagreed with our recommendation that it place tax debt in the notice phase into the FPLP as soon as legally possible, stating that it believed its current notification process was the most cost effective. In stating its position, IRS noted that over three-fourths of tax debtors pay their tax debt after receiving the first notice, and that it believed the action recommended is not appropriate for individual taxpayers who have a high payment rate during the notice process. We do not believe that our recommendation would diminish the effectiveness of IRS’s notice process, especially the voluntary tax collections resulting from the first notice. In fact, those tax debtors who would typically pay their debt upon receipt of the initial notice would be unaffected by the action we are recommending. Although implementing our recommendation would allow IRS to begin levy procedures during the notice phase, in practice, the tax debt generally would not be levied before IRS completes the notice phase. As we discuss in our report, tax debt could be included in the FPLP about three months after IRS notifies the tax debtor of the tax liability, giving the tax debtor sufficient time to respond to both the first and second notices before the levy process would actually commence. Additionally, once the levy process begins, IRS must still send the tax debtor a Collection Due Process notice and wait about two and one-half months before levying a payment through the FPLP. Consequently, tax debt would not generally be levied before the notice phase is completed. However, by starting the levy process during the notice phase, IRS would be able to begin levying payments earlier than would otherwise be the case if the tax debtor does not voluntarily fully pay or otherwise resolve the tax debt during the notice phase because IRS would not have to continue to delay levy action while it issues the Collection Due Process notice and waits for the tax debtor to respond. IRS also disagreed with our recommendation that it modify its FPLP exclusion policy to allow tax debt in two subcategories of its ACS to be eligible for the FPLP, citing concern that this could result in duplicate levies and thereby create unanticipated hardships for taxpayers. IRS also noted that it attempts to issue levies on cases within these ACS subcategories that could generate more in collections than would be collected through the FPLP due to the program’s limit of 15 percent of each federal payment made to the tax debtor. We do not believe these concerns have merit. We believe that IRS’s current process for manually blocking tax debt from the FPLP would provide a sufficient safeguard against duplicate levies while at the same time preserving adequate flexibility for other collection actions. As we discuss in our report, IRS has the ability to block, and, on a case by case basis, does block individual tax debt accounts from levy through the FPLP. IRS could apply this same approach to these two ACS subcategories. To manually block tax debt from the FPLP, IRS can place a transaction code in the tax debtor’s account that blocks the FPLP from automatically levying the tax debt. The same transaction code can be placed in the tax record if IRS wants to levy more than the 15 percent allowable under the FPLP. This process would allow IRS to increase the effectiveness of the FPLP while preserving its ability to use paper levies when appropriate and minimizing the risk of duplicate levies. With respect to the six recommendations we made to help ensure that IRS’s financial hardship FPLP exclusions are appropriate, IRS agreed with four recommendations, partially agreed with one recommendation, and disagreed with the remaining recommendation. Specifically, IRS agreed with our recommendations to (1) reevaluate whether the dollar ranges for existing financial hardship income thresholds are appropriate and reasonable; (2) establish a policy that when future financial hardship thresholds are changed, tax debtors’ hardship closing codes are changed to ones that maintain the tax debtor’s original income ceiling; (3) evaluate the ongoing validity of financial hardship designations whenever tax debtors fail to file their annual tax returns; and (4) refer tax debtors with a financial hardship designation to IRS’s withholding compliance program for special attention if those tax debtors do not pay their current income tax obligations. Although IRS agreed to reevaluate whether the dollar ranges for existing financial hardship income thresholds are appropriate and reasonable, it raised concerns that imposing a rigid national median amount would disregard circumstances such as family size, medical needs, and geographic variations in average income. It is important to note that our recommendation does not advocate imposing the national median amount as a rigid maximum threshold limit. We recognize that IRS attempts to accommodate the needs of tax debtors with varying family sizes, geographical locations, and various other circumstances. However, as our report discusses, between 1992 and 2004, IRS raised its top financial hardship income threshold ceiling from an amount equal to the median national household income to an amount almost twice the median income without any detailed analysis supporting either the large increases or the deviation in the relationship of these thresholds from the national median income. While IRS disagreed with our recommendation that it establish a policy to review tax debtors’ financial condition every 3 years to verify the continued validity of the financial hardship designation, IRS did agree to consider including a time factor. Specifically, IRS noted that as part of its initiatives to improve its program for taxpayer accounts classified as currently not collectible, of which financial hardship is a significant aspect, it will consider including a time factor. Accordingly, we have modified our recommendation, replacing “every 3 years” with “periodically” to reflect IRS’s willingness to consider including a time factor for reviewing a tax debtor’s financial condition. However, in deciding upon the time factor to use, we believe that IRS should take into account that tax debt is typically only legally available for collection for 10 years. Thus, implementing a time period of greater than 3 years could result in IRS affording itself only one opportunity to reconsider the validity of the financial hardship designation. Finally, IRS stated that it could not agree with our recommendation that it consider changing the financial hardship closing codes for tax debtors designated as being in financial hardship prior to the 2004 increases it implemented in the income thresholds until it has determined how many tax debt accounts would be affected by the recommendation. IRS said that implementing the recommendation to change existing closing codes would require significant computer programming and system changes that it may not be able to implement, and which may not be cost effective. Our recommendation is for IRS to consider changing the hardship closing codes for the affected accounts; we are not recommending that IRS must do so. Implicit in our use of the word “consider” in our recommendation is a cost-benefit determination. In considering whether to change the hardship closing codes, IRS should take into account the work and cost involved in making this change as well as the potential for increased collections in determining the cost effectiveness of any modifications. However, we do believe that IRS erred in not changing the financial hardship closing codes for existing cases when it implemented the 2004 increases in the income thresholds. As discussed in our report, by not changing the closing codes, IRS allowed tax debtors who it previously believed could begin paying off their tax debt at a certain income threshold to immediately begin earning up to, on average, 77 percent more before IRS would hold them liable for their tax obligations. This created the potential for inequitable treatment between taxpayers in these same income brackets who pay their taxes and tax debtors who do not, especially when some of those tax debtors, on the day IRS changed the thresholds, were thereafter allowed to earn up to $34,000 more income without IRS considering whether they continued to warrant the hardship designation. Consequently, in considering whether or not it is cost effective to implement a change in the closing codes of the effected accounts, IRS should also consider the issue of fairness with respect to the taxpayer population as a whole. We are sending copies of this report to the congressional committees with jurisdiction over IRS and its activities, the Secretary of the Treasury, the Commissioner of Internal Revenue, and interested congressional committees and members. We will also make copies available to others upon request. In addition, this report will be available at no charge on the GAO Web site at http://www.gao.gov. If you have any questions about this report, please contact me at (202) 512- 3406 or sebastians@gao.gov. Key contributors to this report are listed in appendix III. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. To determine whether and to what extent Internal Revenue Service (IRS) tax records contain inaccurate or out-of-date status or transaction codes that exclude them from the Federal Payment Levy Program (FPLP), we used IRS’s unpaid assessments database as of September 30, 2005, to select two statistical samples. We used IRS’s criteria for the statutory and policy exclusions from the FPLP to segment the tax records in the unpaid assessments database into the two categories. The population of statutory exclusions consisted of 9,068,508 tax records that contained tax debt of $72,167,432,455. The population of policy exclusions consisted of 7,183,880 tax records that contained tax debt of $81,492,531,369. We selected a statistical sample of 100 tax debts that were excluded from the levy program based on IRS’s designation of their tax record as being excluded because of a legal—statutory—requirement of the Internal Revenue Code. We also selected a statistical sample of 100 tax debts that were excluded from the levy program based on IRS’s policy determinations. We randomly selected probability samples from the populations of tax debt accounts excluded from the FPLP for statutory reasons and policy reasons. With these probability samples, each tax account in each of the populations had a nonzero probability of being included and that probability could be computed for any account. Each sample tax account selected was subsequently weighted in the analysis to account statistically for all the tax accounts of its respective population. The sample we selected from each population was only one of a large number of samples that we might have drawn because for each sample we followed a probability procedure based on random selections. Since each sample could have provided different estimates, we express our confidence in the precision of each sample’s result as 95 percent confidence intervals, which are intervals that would contain the actual population value for 95 percent of the samples we could have drawn. As a result, we are 95 percent confident that the confidence intervals presented in this report will include the true values in the respective study populations. For the statistical error rate projection, we used a point estimate with a 95 percent confidence interval. The projected point estimate combined with the confidence interval surrounding the point estimate means that although we estimate the error rate to be at the point, we are 95 percent confident that the true error rate is somewhere between the interval’s lower and upper limits. For each sampled tax period, we obtained and reviewed IRS records on the status and history of tax collection action with particular emphasis on actions that affected the FPLP status of the tax debt. We performed additional searches of criminal, financial, and public records. We compared each sampled tax debt to IRS’s FPLP exclusion and inclusion criteria and determined the accuracy of the status or transaction code that excluded the tax debt from the FPLP. We categorized a sample tax debt as an error if the tax period did not meet at least one exclusion criterion or had exceeded IRS’s time frame for ending an exclusion, such as the time frame for terminating an installment agreement. In some cases, the time frame for terminating an installment agreement was exceeded at the time IRS provided us records to review rather than at the September 30, 2005, date of the unpaid assessments database. To determine whether IRS’s policies, procedures, and practices could be strengthened to ensure the accuracy and timeliness of its status and transaction codes, we reviewed IRS’s Internal Revenue Manual and interviewed IRS officials to obtain criteria, guidance, and internal controls on (1) coding cases for inclusion and exclusion from the FPLP (2) processing cases in the notice phase; and (3) processing and terminating cases in installment agreements, offers in compromise, and financial hardship. To determine whether opportunities exist to increase the amount of tax debt included in the FPLP, we analyzed the effect of IRS’s exclusion criteria as well as the potential effect of changes in the exclusion criteria on the amount of tax debt included in the FPLP. In assessing the effect of potential changes in the statutory exclusions, we compared the potential for modifying IRS’s existing FPLP exclusion criteria within the exclusion’s legal framework, and we discussed the potential changes with cognizant IRS officials. For the IRS database we used, we relied on the work we perform during our annual audit of IRS’s financial statements. While our financial statement audits have identified some data reliability problems associated with the coding of some of the fields in IRS’s tax records, including errors and delays in recording taxpayer information and payments, we determined that the data were sufficiently reliable to address this report’s objectives. Our financial audit procedures, including the reconciliation of the value of unpaid taxes recorded in IRS’s master file to IRS’s general ledger, identified no material differences. The following individuals made major contributions to this report: William J. Cordrey, Amy Bowser, Ray Bush, Kenneth Hill, Inna Livits, Dave Shoemaker, Sidney Schwartz, and Mark Yoder.

GAO previously testified that federal contractors abused the tax system with little consequence. While performing those audits, GAO noted that the Internal Revenue Service (IRS) records sometimes contained inaccurate or outdated tax information that prevented IRS from taking appropriate collection actions against those contractors, including submitting their tax debt to the Federal Payment Levy Program (FPLP) for collection. As a result, GAO was asked to review IRS's coding of tax debt excluded from the FPLP to determine whether (1) IRS tax records contain inaccurate status or transaction codes that exclude tax debt from the FPLP, (2) IRS's monitoring could be strengthened to ensure the accuracy of its status and transaction codes, and (3) other opportunities exist to increase the amount of tax debt included in the FPLP. IRS tax records had inaccurate information that resulted in it erroneously excluding cases from the FPLP and other tax collection actions. The FPLP is a cost-effective automated system used to collect unpaid taxes from certain federal payments. GAO estimates that as of September 30, 2005, over 500,000 tax records' equating to about $2.4 billion in tax debt--contained inaccurate codes that IRS systems used to exclude tax debts from the FPLP. Inaccuracies included tax debts coded as having active installment agreements even though the tax debtor had stopped making payments. IRS's monitoring of cases was insufficient to identify and correct the coding errors GAO identified. Additionally, IRS's monitoring of financial hardship cases is not sufficient to ensure their ongoing accuracy. IRS grants tax debtors experiencing financial difficulty a hardship designation that excludes them from the FPLP and other tax collection activities until their income increases. To measure this, IRS solely uses the income reported on the tax debtor's annual tax returns. However, IRS does not monitor those tax debtors to ensure they are filing and paying current taxes. For 31 financial hardship cases GAO examined, 24 had ceased to file tax returns. Although IRS has increased the amount of tax debt it submits to the FPLP, additional policy changes could further improve the program's effectiveness. Since 1992, IRS has almost tripled the maximum income it allows tax debtors in financial hardship to earn; raising it to $84,000 in 2004--almost double the national median income. As a result, whereas in 1992 no one earning above the median income was considered to be in financial hardship (and therefore excluded from the FPLP), in 2005 almost two-thirds of the tax debt in financial hardship was owed by individuals earning over the median income. Although a financial hardship designation may be appropriate in many situations, allowing relatively high-income tax debtors to avoid tax collection action, including the FPLP, calls into question the fair application of the tax system and may contribute to noncompliance. IRS policy also limits the amount of tax debt in the FPLP by excluding $5 billion in tax debt from the program while IRS is pursuing levies from other assets or income sources. Additionally, during notification IRS excludes individuals' tax debt from the FPLP about twice as long as legally necessary.

IRS’s Business Systems Modernization program, which began in 1999, is a multibillion-dollar, high-risk, highly complex effort that involves the development and delivery of a number of modernized tax administration and internal management systems, as well as core infrastructure projects, which are intended to replace the agency’s aging business and tax processing systems and provide improved and expanded service to taxpayers and internal business efficiencies for IRS. CADE, one of the core systems of the Business Systems Modernization program, was intended to provide a modernized system of taxpayer accounts, with the ultimate goal of eventually replacing the Individual Master File (IMF), a 1960s legacy tax master file that maintains all taxpayer records for individual taxpayers. The IMF is the authoritative data source for individual tax account data. Tax data and related information pertaining to individual taxpayers are posted to the IMF, and the IMF is updated annually to incorporate new tax law procedures and changes. The IMF is a critical component of IRS’s ability to process tax returns, as all the other IRS information system applications use data from this source. According to IRS, the IMF is a complex computer system that maintains all individual taxpayer records but has inherent limitations that significantly constrain IRS’s ability to achieve its mission. The delays in weekly updates of taxpayer accounts, the lack of synchronization of taxpayer data across dozens of systems, the complexity of the data, the large volumes of data, and the ongoing need to embed new tax laws and business rules contribute to IRS systems storing, sharing, and processing data that is often incomplete, untimely, or inaccurate. With CADE, IRS started establishing a database and related functionality for posting, settlement, maintenance, refund processing, and issue detection for taxpayer account and return data for relatively simple tax returns for individual taxpayers. Since 2005, CADE has processed and recorded tax return and tax account information for increasing numbers of individual taxpayers. IRS reported that, in 2010, CADE processed about 41 million relatively simple returns (about 30 percent of the total individual income tax returns received); and issued about 36 million refunds that totaled in excess of $66 billion (approximately 14 percent of total refunds), and processed over 7 million taxpayer payments amounting to approximately $90 billion—with zero defects in balancing. In addition, IRS reported that it issues refunds with CADE approximately 5 days faster than refunds processed by the IMF. Although IRS made progress in processing individual taxpayer returns with CADE, IRS found that each successive release of the system was far more difficult as more complex accounts were to be supported, raising concerns about the effort required to address these increasing complexities. Also, since the CADE system operates concurrently with IMF processing, work must be done across the two environments, increasing the complexity of filing season operations and resulting in a greater risk for errors in processing taxpayer account information. IRS was concerned that, without a significant change in approach, CADE would not be completed until at least 2020. These challenges led the IRS Commissioner to form a team of senior IRS technologists and external advisers to review CADE and IRS’s modernization strategy for individual taxpayer accounts. The review identified six risks with the CADE strategy. As a result of the review, in August 2008, IRS began defining a new strategy to address the challenges confronting CADE and deliver improved individual tax processing capabilities sooner. Specifically, the new strategy called for accelerating completion of a modernized database and converting to a single processing system sooner than the current approach would allow. The new strategy is referred to as CADE 2. Table 1 describes the CADE risks reported by IRS and how CADE 2 intends to address them. As noted in the table, IRS used prototypes to test and prove complex technical and performance concepts early and provide guidance and information needed for detailed design. These prototypes were completed in 2010, although some remaining tasks in one of them—the Database Performance Test Prototype—will be completed later. CADE 2 is expected to deliver its functionality incrementally through three phases known as transition states. Table 2 shows the target completion dates and key characteristics for transition state 1 (TS1), transition state 2 (TS2), and the target state. TS1 consists of the following two projects: Daily Processing: According to IRS, this project is to enable IRS to process and post all individual taxpayer returns filed and other transactions by updating and settling individual taxpayer accounts in 24-48 hours with current, complete, and authoritative data, and provide employees with timely access. More specifically, through the Daily Processing project, IRS is expected to modify the IMF to run daily, i.e., modify current transaction processing to occur daily versus weekly; make changes to the Integrated Data Retrieval System (IDRS) and the notice generation process to support the daily vs. weekly processing cycle; and prepare daily “loads” to update the CADE 2 database for IDRS and other downstream systems as they are able to support daily processing. Database Implementation: According to IRS, this project is to establish the CADE 2 database, a relational database that will house data on individual taxpayers and their accounts; develop a capability to transfer data from IMF to the database; and provide for the transfer of data from the database to downstream IRS financial, customer service and compliance systems. Specifically, the project is expected to enable IRS to perform a one-time initialization of the database when TS1 begins (this will involve loading all taxpayer account data from IMF and the CADE database into the CADE 2 database and then validating it); perform daily updates to the database after TS1 begins by extracting, transforming, and loading daily processing changes to IMF data; and utilize the database to provide individual taxpayer account information to selected downstream systems to assist IRS service and compliance personnel. For example, the CADE 2 database is intended to provide daily updates to IDRS, as well as weekly updates to the Integrated Production Model (IPM)—a downstream data repository designed to support IRS compliance functions that house IMF and Business Master File data, information returns, and other data. IPM is expected to include current individual taxpayer data from the CADE 2 database. Regarding the status of these two projects, IRS completed efforts to define the TS1 projects’ logical design in December 2010 and began the detailed design phase, which includes activities such as documenting the physical design of the solution. IRS expects to complete this phase by April 2011. IRS has also defined overall objectives for TS2 and the target state, but detailed planning of TS2 is in the early stages, with the formation of a planning team that expects to define a high-level approach by early May 2011. In December 2009, the Treasury Inspector General for Tax Administration (TIGTA) identified several challenges that IRS needed to address to effectively manage identified CADE 2 risks, including, among other things, implementing a governance structure for the PMO to provide oversight and direction for the implementation of CADE 2 and developing contingency plans in the event that CADE 2 cannot be fully implemented. IRS officials completed a contingency strategy in November 2010. In May 2010, we reported that, while much had been done to define CADE 2’s transition states, IRS had not identified time frames for completing key planning activities for TS2. Consequently, we recommended IRS establish these time frames. In response, IRS stated that it would develop a plan for launching TS2 activities that will outline the approach and a high-level plan for the transition state. IRS also noted that business requirements would be completed in parallel with the development of the plan. In December 2010, IRS initiated TS2 planning activities and set a target of May 2011 for agreeing on the scope and plan for the transition state. In November 2010, TIGTA reported on the prototypes undertaken by IRS to gain confidence in the CADE 2 solution. TIGTA found that the five prototype teams that were established generally managed their objectives effectively, and that the teams also identified risks that faced the successful execution of the prototype plans and took steps to overcome the barriers. TIGTA recommended that IRS take several actions to reemphasize compliance with the elements of the CADE 2 Prototype Process. In its response, IRS agreed with TIGTA’s recommendations and described steps to address them. While the MITS organization has primary responsibility for developing, managing, and delivering the CADE 2 program, IRS has established a new governance approach for the program, including the CADE 2 PMO that is to manage the program and the relationship with the program’s stakeholders, define the solution specifications to meet the requirements and the delivery of the program scope, and specify the roles and responsibilities of all parties. The new governance approach is intended to foster rapid decision-making and proactive risk management and issue resolution, as well as ensure that the appropriate stakeholders are involved, engaged, and collaborating as a unified team in making CADE 2 decisions. The CADE 2 PMO is headed by an Associate Chief Information Officer and made up of several groups, each responsible for a different aspect of the program. These groups include Delivery Management. Responsible for overseeing and coordinating activities across the projects that are implementing the solution for each of the transition states. Program Management and Control. Responsible for executing the program management processes, CADE 2 solution planning, and governance and control activities. Chief Architect. Responsible for setting the overall technical direction in alignment with IRS’s business and technical architecture, as defined in the enterprise architecture. Chief Engineer. Responsible for reviewing detailed design and interfaces to ensure the CADE 2 solution will work, is secure, and integrates with other systems and infrastructure. Business Operations. Responsible for providing the administrative support for all branches of the PMO. These groups are to work with (1) delivery partners who are other MITS organizations (e.g., Applications Development and Enterprise Operations) that work to deliver the program scope, (2) business partners (e.g., W&I and the Chief Financial Officer) who are organizations who either use the system or define the business requirements and, as such, play a key role in overseeing CADE 2, and (3) other stakeholder groups who support, influence, or oversee the program. To oversee and guide the PMO, IRS also established several executive- level boards and advisory councils composed of business and IT officials. The executive-level boards include the following: The CADE 2 Executive Steering Committee, which consists of senior executives from MITS, W&I, and Department of the Treasury, and serves as an oversight group that ensures the program stays aligned with the IRS strategic plan and approves decisions with significant organizational or external impact; and The CADE 2 Governance Board, which consists of Associate Chief Information Officers from CADE 2 and Applications Development and the business modernization executive from W&I, and ensures that objectives are met; decisions and issues are resolved in a timely manner; risks are managed appropriately; and the expenditure of resources allocated is fiscally sound. The CADE 2 Governance Board also approves program risk response plans, milestone exits, and resolves escalated issues. The advisory councils include the Executive Oversight Team, which consists of executives from the CADE 2 PMO, delivery partners, and senior-level business executives, and serves as the oversight for the day-to-day execution of the CADE 2 program; Associate Chief Information Officer Advisory Council representing leaders of each of the MITS technical units; Program Leadership Advisory Council, a broad group of executives from W&I, Agency Wide Shared Services, and across MITS that consult on key issues, risks, analyses, and recommendations on an as needed basis; and Architecture/Engineering Review Council recently created to provide architectural and engineering leadership across the program to ensure design and development efforts are complete and adhere to solution architecture and engineering best practices. In addition, while the projects chartered under CADE 2 are expected to follow the IRS enterprise life cycle methodology that other projects are required to follow, IRS also defined a program management framework for CADE 2 that defines the phases, activities, deliverables, milestones, and reviews necessary to manage both the program and each of its component projects in a coordinated and integrated manner. The PMO and supporting boards are to play key roles in reviewing deliverables and approving milestone exits for the program and assuring that the projects are properly aligned and integrated with the program. Finally, IRS has also defined and started implementing several processes to support its management of the CADE 2 program. They include processes for managing risks, the integrated master schedule, the prototype process, and other activities. While several of these processes were already established for projects, IRS enhanced them for CADE 2 to take into account the complexities associated with integrating projects into a program. For TS1, IRS has identified 20 benefits. These benefits span three categories: service; compliance; and other benefits, including reduced system costs and improved security benefits. Our past work at IRS established that quantitative targets can be useful for tracking program performance. While it may not always be possible to quantify targets, doing so helps to objectively measure the extent to which expected benefits have been realized. As shown in table 3, IRS has set quantitative targets for 15 of the 20 benefits for TS1. (Table 7 in app. II describes all the benefits IRS has identified and the related quantitative targets where they exist.) An example of a benefit with a quantitative target is the percentage of refunds IRS will process on a daily basis. IRS expects CADE 2 to increase this from 30 to 80 percent, resulting in faster refunds to many taxpayers. Another example is the security benefit of reducing the number of IRS databases that contain extracts of data from CADE 2. IRS expects this to be reduced by 20 percent from the 2011 baseline. IRS officials have stated that they have not set quantitative targets for five TS1 benefits for two reasons. First, officials told us that two benefits related to improved security and updated account information cannot be quantified and therefore do not have numeric targets. (See app. II for more details.) Second, IRS officials stated that competing priorities have impeded their ability to define the remaining targets. While they stated they plan to define these targets, they have not committed to specific time frames for doing so. Until IRS has set quantitative targets where feasible in TS1, it and other stakeholders including Congress will not have complete information on the specific benefits that can be expected from funds spent on this phase. In addition, it will be more difficult to objectively measure whether expected benefits have been delivered. In contrast with TS1, IRS has not finalized the benefits expected in TS2 or defined related quantitative targets. Specifically, IRS has identified a list of 22 service, compliance, and other benefits including addressing financial material weaknesses it expects to achieve beyond TS1. These benefits are summarized in table 4 and described in greater detail in table 7 in appendix II. However, IRS has not yet determined whether these benefits are to be delivered in TS2 or the target state. According to IRS, committing to what can be delivered in TS2 is contingent upon a number of decisions yet to be made. Specifically, officials stated that IRS must first finalize its approach for TS2 (as noted earlier, the agency expects to do this by May 2011). Officials also stated that funding decisions for the upcoming fiscal year could impact IRS’s plans for TS2. In addition, IRS has not finished defining the full range of possible benefits it expects beyond TS1. For example, one possible compliance benefit that IRS is exploring is requiring providers of information returns to send those returns to IRS and taxpayers at the same time, which would allow IRS to use this data earlier to conduct some compliance checks. Conducting earlier analyses of tax return information using CADE 2 and IPM could allow IRS to stop an erroneous refund before it goes out. IRS currently waits until well after the filing season to conduct some compliance checks, such as computerized matching of information returns and tax returns and examinations of taxpayers’ books and records. Such compliance checks can lead to costly collection efforts from noncompliant taxpayers. Finally, as summarized in table 4 and shown in greater detail in appendix II, IRS has not yet set quantitative targets for any of the benefits defined or committed to any time frames for doing so. As noted for TS1, until IRS has set quantitative targets where feasible in TS2, it and other stakeholders will not have complete information on the specific benefits than can be expected from funds spent, and it will be difficult to objectively measure whether expected benefits have been delivered. We acknowledge that thinking through the expected TS2 benefits (including those from the table in app. II and other potential benefits) and related targets as the approach and design are being considered may result in having to make adjustments later. However, the benefits and targets generated at this stage can influence design decisions and allow IRS to identify early how associated systems and business processes might be affected. In July 2009, IRS reported preliminary life cycle cost estimates for TS1 and TS2 of about $1.3 billion through 2024, including about $377 million for development and $922 million for operations and maintenance. The estimates include costs for eight categories of work needed to achieve the goals defined for TS1 and TS2. The majority of the costs are for the Operational Framework category ($663 million or 51 percent of total costs), which is to provide for the infrastructure on which to develop, test, and deploy the new and enhanced applications under the CADE 2 program. The CADE 2 preliminary cost estimates are summarized in table 5. The estimated annual costs from 2009 through 2024 are shown in figure 1. At the time they were developed, IRS estimated CADE 2 development costs to end in 2015 and total costs to peak during 2011 at about $162 million. Estimated costs between 2015 and 2024 are all recurring operations and maintenance costs. IRS’s estimates did not include expected costs for the following items: operations and maintenance of the IMF system to include enhancements needed to achieve the CADE 2 target state, prototype development and assessments, technical training (estimated at $1 million to $3 million), work required to reach the final target state, and process reengineering costs/business costs associated with implementing the current weekly batch processing of accounts to the daily processing required at the completion of TS1. The process IRS used to develop the CADE 2 preliminary cost estimates was generally consistent with the best practices outlined in the GAO cost estimating guide. Specifically, IRS followed a well-defined process for developing a comprehensive, documented, accurate, and mostly credible life cycle cost estimate for the first two phases of CADE 2. For example, the estimation process was performed in accordance with a well-defined plan by an independent Estimation Program Office. Parametric estimation models were used that contained catalogs based on historical data, along with cost estimating relationships, designed to produce estimates analogous to similar programs. In addition, IRS used simulation techniques to incorporate risks and determine confidence bounds on the estimates. Further, IRS plans to update its estimates, consistent with the best practice to update estimates to reflect changes in technical or program assumptions. Specifically, since the estimates were prepared, IRS has developed a better understanding of the work required to meet CADE 2 objectives for TS1 and the current work breakdown structure (i.e., definition of work to be performed) no longer matches the one used to develop the preliminary estimates. IRS officials told us that, consistent with IRS’s Estimation Program Office Estimator’s Guide, they plan to revise their estimates after detailed design information for the database implementation and daily processing projects is completed and expect them to be available by the completion of our audit. CADE 2 PMO officials stated they did not expect the revised estimates to differ significantly from the current estimate. It is important to note that, even when all TS1 cost elements have been updated to accurately reflect the current understanding of the work to be performed, the total cost of ownership through TS2 could still significantly differ from the new estimate, given that, as previously noted, the approach for TS2 has yet to be defined. We identified the following three practices IRS did not follow that could strengthen its methods for estimating CADE 2 costs to ensure they are more credible: First, according to best practices, items excluded from estimates should be documented and explained. While IRS documented four of the five items that were excluded from the estimates, it did not provide an explanation for excluding these costs. In addition, IRS did not document that it excluded the business costs associated with moving to daily processing. IRS officials stated that, at the time the estimates were developed, these business costs were not well understood by the estimators and now believe that these costs would be relatively small, representing mostly costs for staff already being paid out of existing budgets. Nevertheless, consistently documenting excluded costs and providing a rationale for excluding them would improve credibility. Second, IRS’s estimates did not include inflation. According to best practices, adjusting costs for inflation correctly is necessary if the cost estimate is to be credible. Inflation reflects the fact that the cost of an item usually continues to rise over time. If a mistake is made in applying inflation, or if inflation is not included, cost overruns can result. Third, IRS did not conduct a sensitivity analysis to examine the effects of changing assumptions and ground rules on its estimates. While IRS developed risk-adjusted estimates, it did not perform the formal sensitivity analysis necessary to understand which variables most affect the cost estimate. IRS officials stated they intend to include a sensitivity analysis in the revised cost estimates expected to be available by the completion of our audit. Performing this analysis will not only improve the credibility of the estimate, but it will provide management a better understanding of the relationship of key factors to cost, which is important in evaluating different options as part of subsequent risk and issue mitigation strategies and contingency plans. Given the size and significance of the CADE 2 program, it is important that IRS have an effective risk management process. According to best practices, an effective risk management process identifies potential problems before they occur, so that risk-mitigating activities may be planned and invoked as needed across the life of the product and project in order to mitigate adverse impacts on achieving objectives. It includes assigning resources, identifying and analyzing risks, and developing risk mitigation plans and milestones for key mitigation deliverables. IRS’s risk management process for CADE 2 is generally consistent with best practices. For example, regarding resources assigned to risk management, the CADE 2 PMO’s Director for Program Management and Control is responsible for “executing” the risk management process. Biweekly meetings are held where the Director for Delivery Management and project officials discuss project status and risks and determine whether risks need to be escalated as candidate risks to the program level. Monthly risk and issue meetings are held to discuss the program-level risks. Finally, those risks that require additional oversight are then presented to the Chief Technology Officer for review. Program-level and project-level risks are scored and tracked in a consistent manner using a risk repository tool. Risk mitigation plans and associated milestones are defined for each risk. Officials from the PMO have also told us that the management structure and processes that have been established for the program are intended to help mitigate the risks associated with developing such a large and complex program. While IRS is generally effectively carrying out the activities associated with the risk management process, the two key documents that support this process are in some cases inconsistent and do not fully reflect actions being taken. For example, the Risk and Issue Management Process document calls for automatically escalating those projects with a high to very high risk exposure score to the next level of management while the Risk and Issue Management Plan calls for escalating risks based primarily upon the judgment of the risk owner and responsible manager. The Risk and Issue Management Process identifies a specific set of actions to mitigate risks depending on the risk exposure score while the Risk and Issue Management Plan provides high-level guidance describing various alternatives that can be taken to mitigate risks. As another example, both documents identify the risk management roles but do not specify when in the process these roles are assigned or what the required qualifications are. IRS recognizes its risk management procedures could be strengthened, and the Director for Program Management and Control intends to review these procedures by the completion of our audit, to ensure they are consistent and clearly reflect the process in place. This action will help IRS achieve optimal effectiveness in identifying, prioritizing, or managing risks that could affect the cost and schedule of the program. Using its risk management process, IRS has identified several significant program risks for CADE 2. As of November 2010, six of these were considered “top risks.” These “top risks,” which include TS1 contingency planning and resource contention among IRS priorities are listed in table 6 along with the mitigation strategies that have been defined for them. (For a complete list of program risks, see app. III.) While IRS has taken many steps to reduce the risks associated with CADE 2, the agency will nevertheless be challenged in meeting its January 2012 schedule for delivering the benefits associated with daily processing and database implementation given the amount of work needed to complete TS1 over the next year. This includes completing logical design and detailed design activities, testing and integration of core daily applications and their development, and testing and population of the database with current CADE and IMF data. Further, as already noted, waiting to finish (1) defining targets for the TS1 benefits and (2) finalizing benefits and related targets for TS2 creates a risk of costly or time-consuming late design changes. Recognizing the challenges it faces, IRS has taken additional steps to increase the likelihood of meeting the schedule. These include moving certain activities up, performing others concurrently and adding checkpoints to monitor the program’s status. In addition, IRS has recently completed a TS1 contingency strategy. While these actions may increase the likelihood of meeting the schedule, some of them, such as performing activities concurrently, could potentially introduce more risk to CADE 2’s successful development and implementation. Defining benefits and related quantitative targets where feasible is important to understanding what to expect and how to measure performance against expectations. For TS1, while IRS has identified numerous expected benefits, it has not yet defined quantitative targets for all of them. Given that IRS is finalizing the design and preparing to implement TS1, it is important for the agency to complete defining the targets associated with TS1 benefits. Delaying the setting of targets could result in costly and time-consuming changes to the design and associated requirements. Similarly, addressing TS2 benefits and related targets as the design is being considered could influence design decisions and help identify early on how systems and processes might be affected. In developing cost estimates for the first two phases of CADE 2, IRS did not disclose all excluded costs, provide a rationale for documented excluded costs, adjust for inflation, or perform a sensitivity analysis to examine the effects of changing assumptions and ground rules. IRS stated it would perform a sensitivity analysis in the revised estimates it expects to release by the completion of our audit. To its credit, IRS has established a risk management process that is generally consistent with best practices and has identified mitigation strategies to address each identified risk. Several of the risks identified are significant and will require continued attention from IRS. We recommend that the Commissioner of Internal Revenue direct the appropriate officials to take the following five actions: to provide a better basis for measuring CADE 2 performance in achieving TS1 benefits, set the remaining quantitative targets where feasible as soon as possible in 2011; in conjunction with developing the approach for TS2 expected in May 2011, finalize the phase’s expected benefits, set quantitative targets where feasible, and identify how related systems and business processes might be affected; and to ensure the revised estimates for CADE 2 are credible, include inflation when calculating costs, include the costs IRS explicitly excluded or provide a rationale for excluding them, and include any business costs associated with moving to daily processing or document that these costs were excluded and provide a rationale for excluding them. IRS’s Commissioner provided written comments on a draft of this report (reprinted in app. IV.) In its comments, IRS stated that it appreciated that the report recognized the progress the agency had made and that it is following best practices on important disciplines like risk management and cost estimation. IRS also agreed with our recommendations and stated it would provide a detailed corrective action plan addressing each one along with its response to our final report. IRS also provided revised cost estimates for CADE 2 that address our finding related to the use of a sensitivity analysis. We are sending copies of this report to the appropriate congressional committees, and to the Chairmen and Ranking Members of other Senate and House committees and subcommittees that have appropriation, authorization, and oversight responsibilities for IRS. We will also send copies to the Commissioner of Internal Revenue, the Secretary of the Treasury, the Chairman of the IRS Oversight Board, and the Director of the Office of Management and Budget. The report also is available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff members have any questions or wish to discuss the material in this report further, please contact us at (202) 512-9286 or pownerd@gao.gov or at (202) 512-9110 or whitej@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix V. As agreed, our objectives were to 1. determine whether the Internal Revenue Service (IRS) has identified the expected benefits of Customer Account Data Engine (CADE) 2 and set targets for measuring success, 2. examine the estimated costs of CADE 2 and assess IRS’s process for 3. assess IRS’s process for managing the risks associated with CADE 2 and describe the risks IRS has identified using this process. For our first objective, we reviewed relevant documents, including the CADE 2 Program Business Case and Benefits Management Plan, and interviewed officials from IRS offices: Wage and Investment and Modernization and Information Technology Services (MITS). We also used criteria from our prior work to identify attributes of successful performance measures. We reviewed documents including the Program Roadmap, Business Case and Benefits Management Plan, and Benefits Roadmap, and the Office of Management and Budget Exhibit 300. For our second objective, we summarized the estimated costs reported in the July 2009 CADE 2 Estimate Summary/ Basis of Estimate and totaled costs by year for each of the major categories for which estimates were developed. We reviewed several documents, including the July 2009 CADE 2 Estimate Summary/ Basis of Estimate and the Estimation Breakdown Summary reports and capabilities and solution concept documents (e.g., the Integrated Master File Enhancements: Capabilities Definition and Solution Concept document and the Financial Settlement: Capabilities Definition and Solution Concept document), which support the Basis of Estimate. We also reviewed IRS’s Estimators’ Reference Guide and aspects of the SEER-SEM methodology that was used to develop the estimates. We also interviewed staff from the Estimation Program Office, as well as members of the Program Management and Control function of the CADE 2 Program Management Office (PMO) responsible for tracking the program’s cost. We compared what we learned with the 12 steps for developing well-documented, comprehensive, accurate, and credible cost estimates identified in the GAO Cost Estimating and Assessment Guide. For our third objective, we reviewed documented procedures supporting the risk management process, including the Risk and Issue Management Process, the Risk and Issue Management Plan, the user guide for the Item Tracking Reporting and Control system that is used to track risks and the MITS Risk Identification Procedure. We also reviewed artifacts of the risk management process, including program risk logs and detailed risk reports. We participated in meetings in which risks were discussed and reviewed agendas and meeting minutes for others. We also interviewed several officials, including the CADE 2 Program Director, the Director of the CADE 2 PMO Program Management and Control office, and the Director of the CADE 2 PMO Delivery Management office. Through document reviews and interviews with appropriate officials, we also examined the following areas in greater depth: (1) schedule, (2) contractor oversight, (3) requirements management, (4) prototypes, and (5) human capital. We selected these areas because, based on our experience, they represent areas of risk or have been problematic for IRS in the past. In addition, we examined IRS’s prototyping efforts to understand how the agency was using them to mitigate risks. We evaluated what we learned about IRS’s risk management process against relevant criteria for Software Engineering Institute’s Capability Maturity Model Integration (CMMI) to determine the effectiveness of the process. To obtain background information on the program, we reviewed a host of documents developed by the CADE 2 PMO, including the CADE 2 Program Charter, Solution Architecture, Program Roadmap, Program Management Plan, the charters of the two CADE 2 transition state 1 projects (Daily Processing and Database Implementation), as well as governance team meeting minutes. We also observed CADE 2 meetings, conducted interviews with and received briefings from CADE 2 officials, including the IRS Chief Technology Officer, the Deputy Chief Information Officer for Strategy and Modernization, the Associate Chief Information Officer who serves as executive-in-charge of the CADE 2 program, and the Chief Architect of the CADE 2 program. We conducted this performance audit from January 2010 to March 2011 primarily at the New Carrollton Federal Building in New Carrollton, Maryland, where the CADE 2 staff are located, and Atlanta, Georgia, where the Wage and Investment division staff are located, in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Table 7 shows a list of benefits the IRS has defined to date for transition state 1 (TS1) and beyond, a related quantitative target for a benefit if applicable, and the transition state in which IRS expects to realize the benefit. According to IRS, defining which benefits will be delivered in transition state 2 (TS2) is contingent upon a number of decisions yet to be made and actual benefits may differ from those currently expected depending on the outcome of these decisions. The following (see table 8) are the IRS’s reported CADE 2 program risks and their planned mitigation strategies as of November 2010. In addition to the individuals named above, Sabine R. Paul, Assistant Director; Joanna Stamatiades, Assistant Director; Rebecca Eyler; Mary D. Fike; Nancy Glover; Sairah R. Ijaz; Robert Kershaw; Paul B. Middleton; John Ockay; and Jennifer Wong made key contributions to this report.

In August 2008, the Internal Revenue Service (IRS) began defining a new strategy for modernizing the way it manages individual taxpayer accounts. The strategy, known as Customer Account Data Engine (CADE) 2, is expected to provide service, compliance, and other benefits to IRS and to taxpayers beginning in 2012. IRS expects to implement CADE 2 in three phases. The first phase is expected to be delivered in 2012, the second in 2014, and the third at a later yet to be determined date. GAO was asked to (1) determine whether IRS has identified the expected benefits of CADE 2 and set targets for measuring success, (2) examine the estimated costs and assess IRS's process for developing them, and (3) assess IRS's process for managing the risks associated with CADE 2 and describe the risks IRS has identified using this process. To do so, GAO reviewed relevant documentation, attended program review meetings, and interviewed IRS officials. IRS has identified 20 service, compliance, and other benefits for the first phase of CADE 2, including increasing the percentage of refunds processed daily and reducing the number of erroneous notices due to better account information, and has set quantitative targets for most of these benefits. GAO has previously reported that quantitative targets can be useful for tracking program performance. While it may not always be possible to quantify targets, doing so helps to objectively measure the extent to which expected benefits have been realized. However, IRS has not yet finalized expected benefits for the second phase or set related quantitative targets, because, according to officials, these are contingent upon yet to be made design and funding decisions. Nevertheless, addressing the second phase's benefits and related targets as the design is being considered could influence design decisions and help identify early on how systems and processes might be affected. IRS reported preliminary life cycle cost estimates for the first two phases of the CADE 2 program of about $1.3 billion through 2024. This includes about $377 million for development and $922 million for operations and maintenance. IRS's process for developing the preliminary estimates was generally consistent with best practices. However, the agency did not follow three practices intended to improve the credibility of cost estimates. Specifically, IRS did not (1) consistently document excluded costs or provide a rationale for excluding them; (2) use inflation in calculating costs; and (3) perform an analysis to examine the effects of changing ground rules and assumptions. While IRS stated it would perform the analysis of changing ground rules and assumptions in revised estimates to be available by the completion of our audit, until the agency implements all these practices its estimates may not be credible. IRS's process for managing the risks associated with CADE 2 is generally consistent with best practices. Through its process, IRS identified significant risks facing CADE 2, including that filing season and other top information technology investment priorities may result in contention for key resources, the delivery of the first phase of CADE 2 may be delayed if deficiencies identified in requirements are not corrected in a timely manner, and the risk that technical challenges and other risks to implementing the database identified as a result of prototyping efforts may not be addressed. To its credit, IRS has developed mitigation strategies for each identified risk. While IRS is working to ensure CADE 2 is successfully managed, the schedule for delivering the initial phase is nevertheless ambitious. IRS officials have acknowledged this and are taking actions to increase their chances of meeting it, including moving certain activities up, performing others concurrently, and adding checkpoints to monitor the program's status. While these actions may increase the likelihood of meeting the schedule, some of them, such as performing activities concurrently, could potentially introduce more risk to CADE 2's successful development and implementation. GAO's recommendations include (1) identifying all of the second phase benefits, setting the related targets, and identifying how systems and business processes might be affected; and (2) improving the credibility of revised cost estimates by including all costs or providing a rationale for excluded costs, and adjusting costs for inflation. In its comments on a draft of this report, IRS agreed with GAO's recommendations.

Not since the creation of the Department of Defense in 1947 has the federal government undertaken an organizational merger of the magnitude of DHS. In 2003, we designated the implementation and transformation of DHS as one of the high-risk areas across the federal government because it represented an enormous undertaking that would require time to achieve in an effective and efficient manner. Moreover, the components that became part of the department already faced a wide array of existing challenges, and any failure to effectively carry out their missions would expose the nation to potentially serious consequences. The department has remained on our high-risk list since 2003. Most recently, in our January 2009 high-risk update, we reported that, although DHS had made progress in transforming into a fully functioning department, its transformation remained high risk because it had not yet developed a comprehensive plan to address the transformation, integration, management, and mission challenges we identified in 2003. In designating the implementation and transformation of DHS as high risk, we noted that building an effective department would require consistent and sustained leadership from top management to ensure the needed nto an transformation of disparate agencies, programs, and missions i integrated organization. Our prior work on mergers and orga transformations, undertaken before the creation of DHS, found that successful transformations of large organizations can take at least 5 to 7 years to achieve. GAO-03-669. and people—in areas such as information technology, financial management, procurement, and human capital—as well as in its security and administrative services, for greater efficiency and effectiveness. On one level, management integration refers to integration of the elements mentioned above—processes, systems, and people—within management functions (sometimes referred to as vertical integration), from the department level down through each of the corresponding management functions in the component agencies. An example of this is the use of consistent human capital management policies at the DHS CHCO level and for each of the corresponding component agency human capital management functions. On another level, management integration refers to integration of the elements mentioned across management functions (sometimes referred to as horizontal integration), such as the integration of human capital management and financial management activities in areas related to payroll. In February 2009, DHS’s Management Directorate provided us with a definition of its approach and responsibilities for implementing management integration in the department. According to the Management Directorate, DHS defines management integration as including three different levels of activities: (1) strategic integration, (2) operational coordination, and (3) functional integration. The directorate further stated that the first level, strategic integration, consists of efforts to ensure that all component activities and acquisitions align with DHS mission goals through appropriate leadership oversight and policies and procedures. The second level, operational coordination, consists of the delivery of management services in order to increase cross-component collaboration and reduce costs by achieving efficiencies for managing assets such as real property, for procuring volume discounts of supplies and services, and acquiring common technology platforms through shared information technology infrastructure. The third level, functional integration, consists of, among other things, management oversight of component-level internal controls and standard operating policies to ensure departmentwide compliance with presidential directives, congressional mandates, and other legal requirements and DHS policies; and consistent business practices that support financial reporting and operational assurance statements. The Management Directorate includes the CFO, the CSO, the CHCO, the CAO, the CPO, and the CIO. They are referred to as the departmental management chiefs. In addition to the department’s Management Directorate, each of the seven DHS component agencies has its own component management chief for the procurement, financial, human capital, information technology, administrative, and security management areas. Figures 1 and 2 show the DHS and DHS Management Directorate’s organizational structures. The Homeland Security Act of 2002 gave DHS’s USM responsibility for the management and administration of the department, including the transition and reorganization process, among other things. The Implementing Recommendations of the 9/11 Commission Act of 2007 (9/11 Commission Act) enhanced the USM position by designating the USM as the Chief Management Officer (CMO) of DHS and principal advisor to the Secretary on matters related to the management of the department, including management integration and transformation. DHS also defined the USM responsibilities for integration in department management directives following the creation of the department. For example, a DHS management directive assigns the USM responsibility and accountability for designing departmentwide integrated systems to improve mission support. Within the Management Directorate, the management chiefs’ roles and responsibilities for the integration of the department are established in DHS management directives. For example, DHS management directives give the departmental management chiefs responsibility to ensure the integration of their management function and to review their programs in order to recommend program improvements and corrective actions where appropriate. DHS management directives also require the departmental management chiefs to annually establish milestones for the integration of their management function’s activities. Component management chiefs are to implement initiatives within their respective functional areas that relate to management integration. In 2004, the Secretary of Homeland Security issued a departmentwide memo on DHS efforts to integrate management functions. In order to ensure that both department and component personnel took responsibility for supporting performance of management functions, the memo describes the concept of dual accountability in which both the heads of component agencies, such as TSA or CBP, and the DHS management chiefs, such as CPO or CIO, share responsibility for implementing management functions. For example, the TSA Administrator and DHS’s CPO are both responsible to the DHS Secretary, through their respective chains, for procurement performance at TSA. An accompanying memo from the Deputy Secretary of Homeland Security in 2004 noted that component agency heads would be responsible for accomplishing the mission of their component agencies, and management chiefs would be responsible for providing the support systems to help components accomplish their mission. The memos also set out the “dotted line” reporting relationship of the component management chiefs, such as TSA’s CPO or CBP’s CIO, to the department management chiefs, DHS’s CPO or DHS’s CIO. Resulting management directives were developed for each DHS management function in 2004 as principal documents for leading, governing, integrating, and managing the management functions throughout DHS. These management directives require DHS management chiefs to collaborate with component agency heads on the recruiting and selection of key component management officials, and provide input into component management chiefs’ performance agreement and evaluation, among other things. The 9/11 Commission Act requires DHS to develop a strategy for management integration as part of the department’s integration and transformation to create a more efficient and orderly consolidation of functions and personnel in the department. In our 2005 report, we recommended that DHS develop an overarching management integration strategy for the department that would, at a minimum, have the following characteristics: look across the initiatives within each of the management functional units; clearly identify the critical links that must occur among these initiatives; identify trade-offs and set priorities; set implementation goals and a time line to monitor the progress of these initiatives to ensure the necessary links occur when needed; and identify potential efficiencies, and ensure that they are achieved. We pointed out that a comprehensive management integration strategy would, among other things, help the department look across initiatives within each of the functional units to clearly identify the links that must occur among initiatives and develop specific departmentwide goals and milestones that would allow DHS to track critical phases and essential activities. By including these characteristics in DHS’s management integration strategy, we said that Congress, DHS employees, and other key stakeholders would have access to more transparent information regarding departmental integration goals, needed resources, critical links, cost savings, and status documentation, thereby providing a means by which DHS could be held accountable for its management integration efforts. In commenting on our 2005 report, DHS discussed actions it was taking to address our recommendation that it develop a management integration strategy, stating that it was establishing an integrated project plan / integration strategy that would define roles and responsibilities and identify key deliverables and milestones. DHS has not yet developed a comprehensive strategy for management integration that is consistent with statute and that contains all of the characteristics we identified in 2005. According to DHS’s USM, the department has not yet developed a comprehensive management integration strategy because, in part, the Management Directorate has focused on building the management operations capacity within the functional areas, such as financial management and information technology. As a result, the Management Directorate has not yet focused on integration across the functional areas and has not clearly or systematically identified trade-offs and linkages among initiatives in different functional areas. In the absence of a comprehensive management integration strategy, DHS’s USM, Chief of Staff, and department and component management chiefs stated that various departmental documents collectively contribute to the department’s strategy for implementing and achieving management integration. In particular, DHS officials identified (1) departmentwide documents that provide guidance that relate to management integration across the department; and (2) documents for management of functional areas. With regard to the departmentwide documents, DHS officials included the following as particularly relevant to aspects of management integration: DHS Integrated Strategy for High Risk Management. This document is intended to be a corrective action plan outlining the department’s framework for its transformation efforts and methods by which the department will seek to improve performance in high-risk areas we have identified since 2003. For the high-risk area of DHS implementation and transformation, the document discusses five areas of focus for the department—utilizing a management framework to unify 22 disparate organizations, creating joint requirements planning and risk assessment processes, instituting an Investment Review Board, implementing a corrective action plan, and consolidating and integrating a financial management system. Management Directorate Strategic Plan Fiscal Years 2009 through 2014. This plan sets out the Management Directorate’s vision, core values, guiding principles, goals and objectives, as well as the organizational structure and responsibilities of the Management Directorate and department management chiefs. The plan provides the following four objectives for the Management Directorate: (1) provide structure (strengthen unified organizational governance to enhance departmentwide communication, decision making, and oversight); (2) optimize processes and systems (integrate functional operations to facilitate cross-component synergies and streamline coordination ensuring reliable and efficient support of mission objectives); (3) foster leadership (adhere to core values and guiding principles of DHS in performing duties, effecting progress, and leading with commitment for the mission); and (4) leverage culture (leverage the benefits of commonalities and differences across components to promote cooperative intra- and inter-agency networks and implement best practices). The plan also discusses four methods that the Management Directorate will use to achieve the plan’s objectives—provide guidance, offer representation, deliver tools, and manage services. Integrated Planning Guidance Fiscal Years 2011 through 2015. This document describes the DHS Secretary’s policy and planning priorities for the 5-year budget time frames, such as for fiscal years 2011 through 2015. The Integrated Planning Guidance is part of the DHS strategic planning process and, among other things, provides general risk management guidance for prioritizing programming and budget proposals within the department. Future Years Homeland Security Program (FYHSP) Fiscal Years 2009 through 2013. This document provides a summary and breakdown of DHS program resources over a 5-year period; including resource alignment by goals, component appropriations, and component programs, as well as program descriptions, milestones, performance measures, and targets. In the fiscal year 2009 through 2013 FYHSP, DHS projected funding for 65 priority programs within 13 components in support of the five goals of the DHS Strategic Plan. Internal Control Playbook Fiscal Year 2009. This document comprises DHS’s plan to design and implement departmentwide internal controls with respect to three areas: (1) internal controls over financial reporting (which provides an overview of efforts to establish reliable financial reporting); (2) internal controls over operations (which outlines plans to maximize the effectiveness and efficiency of operations); and (3) conformity with financial management system requirements (which summarizes efforts to strengthen the internal controls over the department’s financial systems). Specific examples of these internal control areas include: integrating internal control assessments across component lines of business, integrating financial system security assessments through tests of operating effectiveness, and incorporating results into plans of action and milestones. With regard to functional area documents, DHS officials indicated that both management directives and functional area strategic plans contain elements of the department’s strategy for achieving management integration. DHS issued management directives for each of the six department management chiefs—the CAO, CFO, CHCO, CIO, and CPO management directives were issued in 2004 (with updates for the CIO and CPO in 2007 and 2008, respectively); the management directive for CSO was issued in 2006. These directives communicate standard definitions of the management chiefs’ respective roles and responsibilities; define the concept of dual accountability for both mission accomplishment and functional integration as the shared responsibility of the heads of DHS’s individual agencies or components and the department management chiefs; and establish the need for the department management chiefs, along with the heads of agencies, to annually recommend and establish integration milestones for the consolidation of the chiefs’ functions. Functional area strategic plans generally discuss, among other things, the missions and goals of the department management chiefs and the link between the goals and objectives in each functional area strategic plan and the goals and objectives in DHS’s Strategic Plan. Among the six department chiefs, four have issued strategic plans for their functional areas—the CAO, CIO, CHCO, and CSO. While some of the documents DHS officials identified as contributing to the department’s strategy for implementing and achieving management integration address some of the characteristics we have previously identified for such a strategy, these documents, either individually or taken together, do not include all of the characteristics we have identified. These documents described by DHS officials as contributing to the department’s strategy for achieving management integration can provide high-level guidance for integration efforts and can help the department to manage those efforts. For example, two of the functional area strategic plans set goals, objectives, and milestones for implementing certain initiatives within functional areas. Moreover, the Management Directorate Strategic Plan and other departmentwide documents, for example, set performance goals, measures, and targets for achieving certain management initiatives. Such elements as goals, objectives, milestones, performance targets, and priorities documented in these plans and strategies can help the department to manage, implement, and monitor the specific initiatives to which these elements apply. They can also help to guide efforts to consolidate policies, processes, and systems within each management functional area. However, among the documents cited by DHS officials as being part of the department’s management integration strategy, DHS has not yet looked across the management initiatives within management functional areas to identify the critical links that must occur among these initiatives to integrate the department’s management functions both within and across functional areas. Furthermore, the documents generally do not identify the priorities, trade-offs, and potential efficiencies among management initiatives, nor do they set implementation goals and a time line for monitoring the progress of initiatives to ensure the critical links occur when needed. Thus, when considered either individually or together these documents do not constitute a management integration strategy containing all of the characteristics we have identified. See table 1 for more detailed information on the plans. In addition to these functional area and departmentwide documents, DHS officials identified three other documents that are related to management integration: (1) the DHS Strategic Plan; (2) the Quadrennial Homeland Security Review (QHSR); and (3) the Business Operations Manual. The DHS Strategic Plan includes, among other things, the department’s vision, mission, core values, and guiding principles, as well as the goals and objectives by which the department will continually assess performance. The department’s latest strategic plan for fiscal years 2008 through 2013 includes five strategic-level goals related to the department’s mission and management functions. Goal 5 of this plan—“Strengthen and Unify DHS Operations and Management”—sets out the department’s goal for its management functions; information and intelligence sharing; and policy, planning, and coordination functions. Under this goal, the first objective describes what the department plans to achieve for its management functions, through the Management Directorate, and includes a reference to achieving management integration. Specifically, Objective 5.1— “Improve Department Governance and Performance”—states that the department will lead efforts that provide structure to enhance departmentwide governance, decision making, and oversight, including internal controls and performance management tracking, and optimize processes and systems to facilitate integration and coordination. DHS’s Strategic Plan sets out strategic-level goals and objectives for the department’s overall mission and management functions but is not intended to constitute a management integration strategy. As we have previously reported, a management integration strategy goes beyond what is contained in an agency strategic plan, as it provides the more specific operational and tactical information to manage the integration effort. A strategic plan contains the high-level goals and mission for an agency, while a management integration strategy would provide the activities and time lines needed for accomplishing the goals of the integration effort. As required by the 9/11 Commission Act, DHS is developing its first QHSR to conduct a comprehensive examination of the homeland security strategy for the nation, including recommendations regarding the long- term strategy, priorities for homeland security, and guidance on the programs, assets, capabilities, budget, policies, and authorities of the department. The QHSR includes five principal study areas, based on the Secretary’s priorities for homeland security, one of which is maturing and unifying DHS and the homeland security enterprise. DHS is also developing a Business Operations Manual that, according to DHS officials, will provide an overview of the key DHS processes including strategic requirements planning, risk assessment, programming, budgeting, acquisition, and performance assessment, and will also show how these processes link together to ensure coordinated decision making. As the QHSR and Business Operations Manual are still under development, it is too soon to tell whether or how these documents will contribute to DHS’s management integration efforts. DHS has developed some performance goals and measures to measure management activities, but has not yet established measures for assessing management integration across the department. For example, DHS has increased the number of departmentwide performance measures for the Management Directorate in support of Goal 5 of its strategic plan. Specifically, since fiscal year 2008, DHS has added 13 new measures and retired 3 others for the Management Directorate in support of Objective 5.1 of the strategic plan, going from 5 performance measures for the Management Directorate in fiscal year 2008 to 15 measures in fiscal year 2009, as shown in table 2. These measures relate to activities in functional areas but do not help to measure management integration. The Government Performance and Results Act of 1993 (GPRA) provides a framework for strategic planning and reporting intended to improve federal agencies’ performance and hold them accountable for achieving results. Effective implementation of this framework requires agencies to, among other things, clearly establish performance goals for which they will be held accountable and measure progress toward those goals. Although DHS has added measures for the Management Directorate since fiscal year 2008, DHS has not yet clearly communicated what the linkages, if any, are between these measures and the management integration of the department. DHS officials told us that the department’s current measures do not allow the department to gauge the status of management integration and that the department has focused on the development of measures for departmental components, offices, and directorates—such as a measure for the attrition rate for career Senior Executive Service (SES) personnel and a measure for the percentage of improper payments collected. However, these performance measures do not allow the department to assess its progress in achieving departmental goals for management integration within and across functional areas. DHS officials stated that the department’s goal is to develop a set of measures that will help the department assess its management integration. Without such a set of measures, DHS cannot assess its progress in implementing and achieving management integration both within and across its functional areas. A comprehensive strategy for management integration that clearly sets implementation goals and time lines could help the department establish measures for assessing its management integration. Through various management councils, the Management Directorate shares information related to the implementation of management initiatives, solicits feedback from the components, and provides a forum for coordination between component management offices. The Management Directorate has several councils that it uses to communicate through the department, as shown in table 3. Each management chief chairs a functional council to address issues pertaining to that management function. For example, the DHS CFO leads a council that includes component or agency CFOs across DHS and addresses and coordinates departmentwide financial management issues, such as financial management internal controls. Likewise, the USM chairs a Management Council made up of the DHS management chiefs and a representative from each component that discusses issues of departmentwide importance, such as training and development programs. DHS management directives give five of six functional councils responsibility for developing and executing formal communications programs for internal and external stakeholders. The functional councils also have charters that generally define the role of the councils to communicate information and provide input on goals or priorities within their management function. The Management Council does not have a formal charter. We found the Management Directorate uses the councils to share information related to management initiatives with their counterparts from the components and solicit their input on departmentwide issues. For example, when we observed the Human Capital Leadership Council (HCLC) in May 2009, the DHS CHCO updated the council members on his office’s efforts to establish an automated performance management tool. Members of the HCLC’s Human Capital Subcommittee on Performance Management also solicited feedback from the HCLC on whether changing the dates of the department’s performance management cycle could be explored by the subcommittee because it currently falls at a challenging time during the fiscal year. The HCLC discussed the issue and raised points, such as the relationship with the SES performance cycle and the impact of potential continuing resolutions. Ultimately, the HCLC agreed the subcommittee should pursue the issue. Similarly, when we observed a Management Council meeting in April 2009, the council members shared information on issues that affect multiple management functions, such as the Transformation and Systems Consolidation (TASC) initiative to consolidate and integrate its financial management, acquisition, and asset management systems. The council meetings also provide a forum for the component chiefs to provide input into departmentwide plans, such as the functional area strategic plans. For example, component officials from the information technology and human capital management functions collaborated with their corresponding DHS management chief on the development of their functions’ strategic plans at council meetings or council-sponsored off-site meetings. The councils also provide a forum for component management chiefs to raise concerns and suggestions about departmentwide management initiatives. For example, when we observed the CIO Council meeting in April 2009, a component official expressed concern about the component’s outdated financial management system, which they have not replaced because they are waiting for the departmentwide TASC solution. The official said that the component is repeatedly receiving negative results on their financial systems audit while they wait. The DHS CIO responded that if the TASC initiative experiences further delays, she will work with the DHS CFO’s office to jointly determine a solution to allow the components to make progress and identify areas of possible audit mitigation while waiting for TASC to be implemented. Finally, we found that the six functional councils provide the component management chiefs with an opportunity to collaborate with their peers in other components and share best practices. The FEMA Assistant Administrator for Management explained that FEMA is in a better position today because of its management chiefs’ participation in the councils. He said the management chiefs have been able to better handle issues because they are able to learn best practices from their counterparts in other components who are dealing with the same issues and would not have the same access to the other components without the functional councils. While DHS does not have a comprehensive management strategy, its Management Directorate is working to consolidate management policies, processes with associated governance boards, and systems. The Management Directorate has developed and implemented departmentwide policies to replace policies from each of the legacy agencies that make up DHS in all six management functions. For example, the DHS CAO’s office completed a comprehensive review of directives that govern departmentwide activities. According to the DHS Internal Control Bluebook for fiscal year 2008, results of this review reduced the number of directives by over 56 percent. The DHS CAO’s office also implemented a new initiative to develop uniform policies and programs for radiation safety across the department. The DHS CFO’s office launched an online Financial Management Policy Manual tool, which serves as the single authoritative guide on financial management and the foundation for departmentwide financial management knowledge sharing and standardization. According to officials from the DHS CFO’s office, the Financial Management Policy Manual is part of its approach to integrate within the financial management function and is critical to enable financial management employees to carry out their duties and responsibilities effectively and efficiently. The Management Directorate has also taken steps toward consolidating some management processes and established governance boards to manage the processes in the areas of acquisition, information technology, financial management, and resource allocation, as shown in table 4. As we previously reported, the Management Directorate recognized historical shortcomings in its acquisition review process and released an interim acquisition management directive in November 2008. The interim directive established a revised acquisition review process, including roles and responsibilities of DHS approving authorities, threshold levels for acquisitions, and acquisition decision events and the corresponding documentation required. Specifically, it established the Acquisition Review Board (ARB) as the department’s highest review body and charged it with reviewing and approving all programs at key milestone decision points that are above $300 million in life cycle costs. In September 2009, we testified that DHS has also reinstated regular ARB meetings and acquisition decision memorandums. Specifically, as of September 15, 2009, DHS’s ARB reports that it completed 14 acquisition reviews in 2008, and has thus far completed 18 reviews in 2009, including reviews of major acquisitions, such as SBInet, US-VISIT, and Secure Flight. DHS also reports that 7 additional reviews are scheduled to occur by the end of the fiscal year. We previously reported that while recent actions establishing the ARB and an acquisition process represent progress, the department’s previous acquisition review process was not able to effectively carry out its oversight responsibilities and keep pace with investments since 2004. It is too soon to tell whether DHS’s latest efforts will be sustained to ensure investments are consistently reviewed as needed. The DHS IG has also reported that DHS faces challenges in implementing corrective actions for acquisition oversight. In addition, DHS established an Enterprise Architecture Board (EAB) to guide and approve new information technology investments. Enterprise architecture provides systematic structural descriptions of how a given organization operates today and how it plans to operate in the future, and it includes a plan to transition from the current state to the future state. The EAB reviews and approves information technology investments to ensure they align with DHS’s enterprise architecture and transition plan. Based on these reviews, the EAB makes recommendations to the ARB, mentioned above, which the ARB includes in its review of information technology acquisitions. In September 2009, we testified that since 2003, DHS has issued annual updates to its enterprise architecture that have improved prior versions by adding previously missing content. However, DHS has yet to adequately address how it determines and ensures that an investment is aligned with its enterprise architecture. Specifically, while the Management Directorate has recently chartered its EAB and assigned it responsibility for ensuring that each investment is architecturally aligned throughout its life cycle, it has yet to define a methodology, including explicit criteria, for making a risk-based alignment determination. The Management Directorate established a mission action plan process and Senior Management Council (SMC) for Internal Controls to assist the department in identifying, assessing, prioritizing, and monitoring the progress of efforts to remediate material weaknesses. A mission action plan presents an overall plan for correcting a control deficiency that includes milestones with specific dates and remediation actions and is published annually in the department’s Internal Control Playbook. In November 2008, the Management Directorate created its first Internal Control Bluebook, which provides the status of the department’s efforts to design and implement departmentwide internal controls. The SMC oversees the mission action plan process and determines when sufficient action has been taken to correct material weakness. The Management Directorate has faced challenges in implementing the mission action plan process at the components. For example, the DHS IG reported that while FEMA prepared mission action plans for the fiscal year 2009 Internal Control Playbook that address known deficiencies, its financial reporting mission action plan did not adequately emphasize the primary root cause of control weaknesses. Similarly, the DHS IG reported that the TSA’s financial reporting mission action plan in the fiscal year 2009 Internal Control Playbook lacked specific milestones related to some root causes and lacked clear linkage from the root cause to actions and milestones to address the deficiencies. The Management Directorate also participates in the Program Review Board (PRB), which governs the department’s programming efforts as part of the broader Planning, Programming, Budget, and Execution process. This process is the department’s effort to ensure goals and priorities are translated into actionable requirements, programmed and budgeted for appropriately, and realized through execution. Specifically, the USM is a member of the PRB. The PRB considers major multi year programmatic issues across the department and recommends resource allocation decisions to the deputy secretary based on priorities. These decisions provide department approved 5-year resource profiles by component, and provide the foundation for the next DHS budget sent to the Office of Management and Budget. The DHS Program Analysis and Evaluation Director told us that the USM has been instrumental in helping components prioritize activities. The PRB gives the USM a forum for providing input into the resource decisions from a management perspective. However, without a management integration strategy for the department with clear priorities, it is unclear how the management initiatives related to integration that are considered are being prioritized, and whether resources are being used in the most efficient and effective manner. Additionally, the Management Directorate has taken steps in an effort to consolidate the department’s systems. For example, the TASC initiative is the department’s current effort to consolidate its financial management, acquisition, and asset management systems. DHS has been working to consolidate its financial management systems since the department was first created. A prior effort focused on financial management systems integration began in January 2004, known as the Electronically Managing Enterprise Resources for Government Effectiveness and Efficiency (eMerge) project. This project was expected to integrate financial management systems departmentwide and address existing financial management weaknesses. However, DHS officially ended the eMergeprocess for a contractor to implement TASC, but DHS expects to award the contract in early 2010. While DHS officials told us they believe communications between the department’s CFO and the component CFOs for TASC seem to be working well, in October 2009, we testified that the department has not yet completely defined its financial management strategy and plan to move forward with financial management integration efforts. The Management Directorate also has an initiative under way to consolidate its information technology data centers, which are facilities that contain electronic equipment used for data processing, data storage, and communications networking. The Data Center consolidation initiative is an effort to move from DHS’s 17 legacy data centers to two large-scale enterprise data centers. According to DHS’s fiscal year 2010 Budget in Brief report, DHS expects the reduced number of data centers to help streamline the department’s maintenance and support contracts as well as enhance security and improve information sharing with stakeholders. While the Management Directorate intends to complete the relocation of legacy data centers to the new data centers by fiscal year 2011, it is facing challenges in the implementation of the Data Center consolidation. For example, the DHS IG reported that the department has not established necessary connectivity between the two data centers so they are able to provide backup capabilities for each other because necessary telecommunications equipment and circuits are not in place to transmit data between the two centers. During a transformation, strategic goals must be clear and enable stakeholders and employees to understand what they need to do differently to help the organization achieve success. The organization’s performance management system can help to show how individual performance can contribute to overall organizational results, and can help manage and direct the transformation process. Specifically, we have reported that several practices are critical to ensuring that the performance management system supports change. To be successful, transformation efforts must align individual performance expectations with organizational goals. These practices support efforts to create a “line of sight” showing how unit and individual performance can contribute to overall organizational results, and in the case of transforming DHS and integrating the department, can enable the USM and department management chiefs to align activities of subordinate management officials in support of the management integration strategy. A line of sight that connects management integration goals should show how the USM, department management chiefs, and management chiefs of DHS components all contribute to and support DHS management integration goals. Figure 3 provides an example of how individual goals for the USM and department and component management chiefs support the Management Directorate and department goals and objectives for management integration activities related to a particular management integration initiative—in this case, DHS Data Center consolidation. The figure also depicts how management officials at each level provide performance input to align the activities of subordinate levels. As the designated CMO of the department, the USM is specifically tasked with leading management integration at DHS, and is the first link in the line of sight that connects organizational and individual goals and objectives, necessary to ensure that once a management integration strategy is developed, management leaders at each level support its implementation. The 9/11 Commission Act requires that the USM enter into an annual performance agreement with the Secretary, including measurable individual and organizational goals, and be subject to an annual performance evaluation by the Secretary, with a determination of progress made toward achieving those goals and measures. Similarly, we have reported that top leadership should drive the transformation, and have previously stated that the organization’s CMO should have a clearly defined, realistic performance agreement. To support departmentwide goals, the USM’s performance plan should reflect the DHS Strategic Plan and Management Directorate Strategic Plan, and when developed, the management integration strategy. In reviewing performance management linkages at the USM’s level, we found that the Deputy Secretary provided input into the USM’s performance plan in October 2007, and conducted a performance evaluation in 2008 based on this agreement. According to DHS officials, the Deputy Secretary conducted the performance agreement and evaluation—rather than the Secretary—based on delegated responsibilities for the performance of management reform as the department’s chief operating officer. Performance objectives in the USM’s agreement and evaluation show linkages to strategic plans, and include references to several efforts related to management integration. Specifically, the USM’s performance objectives included clear linkages to the fiscal year 2009 through 2014 Management Directorate Strategic Plan, and to the fiscal year 2008 through 2013 DHS Strategic Plan Goal 5—“Strengthen and Unify DHS Operations and Management.” In terms of the content of individual objectives, three of the performance objectives refer to projects or initiatives specifically: (1) Designing a new acquisition review system, (2) finalizing and implementing a plan to improve management controls, and (3) establishing a certified SES performance system. One objective refers generally to improving management programs and initiatives for DHS’s headquarters. The efforts to implement a new acquisition review system and management controls and centralize management of SES positions involves increasing integration of management functions. The Integrated Strategy for High Risk Management referenced in the performance plan was mentioned by DHS as providing guidance for management integration efforts. Other initiatives that contribute to integration, including the consolidation of DHS data centers, are described as accomplishments in the USM’s evaluation related to the implementing programs and initiatives for DHS’s headquarters objective. The second link in the line of sight involves the USM’s relationship with the department management chiefs. Five department chiefs report directly to the USM, and the CFO has a dual reporting relationship to the Secretary and the USM. Based on performance management practices mentioned above, department management chiefs’ performance plans should support organizational goals included in the Management Directorate Strategic Plan. We would expect these performance plans should also support a department management integration strategy, when one is developed. In reviewing department management chiefs’ performance agreements, we found that they supported higher level Management Directorate goals and objectives, and included references to management integration-related activities. For example, performance agreements for the six department management chiefs consistently include a reference to the Management Directorate’s Strategic Plan. We also learned from DHS officials that fiscal year 2009 was the first year that the USM provided a common objective to department management chiefs. Specifically, the fiscal year 2009 management chiefs’ performance plans included a joint performance objective related to management support for the expansion of NPPD. In addition, the agreements consistently include objectives related to management integration. The following are examples of management integration-related objectives in performance agreements: The CSO’s agreement included an objective to “Integrate security services department wide through development and implementation of security policies and practices for the department.” The Acting CIO’s agreement included objectives for consolidating legacy networks and consolidating component data centers (as depicted in figure 3). The CHCO’s agreement included an objective for providing DHS-wide policy and guidance on all major human resources matters. The CAO’s agreement included an objective to establish a consolidated headquarters for DHS. Performance agreements also showed evidence of common goals. For example, as discussed previously, each chief’s agreement includes support for the expansion of NPPD. The performance agreements also refer to specific actions for support from that management function, such as providing space for the new employees, in the CAO’s case, and providing information technology hardware and software to support the new employees, in the CIO’s case. The third link in the line of sight involves the department management chiefs’ relationships with the management chiefs in DHS’s component agencies. The component management chiefs directly report to their component agency heads, while also having a “dotted line,” or indirect, reporting relationship to their respective department management chief. The arrangement of component heads and department chiefs both supporting integration of management functions is referred to as “dual accountability.” When we reviewed DHS’s management integration progress in 2005, the department had recently established the dual accountability structure of reporting relationships. Management directives define department and component management chiefs’ responsibilities, including specific ways that department management chiefs should provide direction to component management chiefs. For example, management directives require the department chief to establish annual milestones for integrating the management function. The directives also require the management chiefs to provide written performance objectives to the component management chief at the start of each performance cycle, feedback to the component rating official on the component chief’s accomplishment of objectives, and input on bonus or award recommendations, pay, and other forms of commendation. Also, in accordance with performance management practices mentioned previously, to ensure accountability for change, component management chiefs’ performance agreements should reflect the department management chiefs’ specific performance objectives for the component chiefs’ management functions, and should also reflect the Management Directorate and departmentwide strategic plans. All department management chiefs except for the CSO said that they specifically established annual priorities of some sort—either goals, objectives, milestones, and / or expected results—for their function. In addition, we reviewed documentation of goals, objectives, milestones, or expected results for each of these management functions. Four management chiefs—the CFO, CPO, CHCO, and CAO—said that the priorities were determined through annual planning processes for the function, either at an offsite meeting or through the management function’s council. The CSO indicated that the management function’s strategic plan served in place of annual milestones, although the plan provided did not identify its applicability to any given year or distinguish any priorities or target for implementation within a particular year. At an individual level, however, the department chiefs did not consistently provide individual input at the beginning of the component management chiefs’ performance cycle—either through written goals and objectives or via direct input into the performance agreement. Management directives for each management function include a requirement that department management chiefs provide the component management chief with written objectives at the start of the annual performance cycle. In our review we found that only two department chiefs—the CAO and CPO— said that they provided individual input with regard to component chiefs’ performance. The CAO said that he provided written objectives attached to a memo that was sent to each component CAO, and the CPO said that individual input and goals were provided annually in the form of a letter. We also reviewed these objectives and goals provided to the components. The other four management chiefs said that they did not provide individual input, and instead pointed to collective goals or objectives developed through planning processes and contained in strategic or operational plans. While these collective processes and overall plans provide general guidance for the management function, they do not meet the standard established by the management directives of annual, individualized performance input. The USM told us that the functional councils have improved in their development of common management goals for their functions, but she agreed that they have not yet consistently followed through by putting those goals into individual performance plans. She added that the department’s management chiefs would be including this information in component chiefs’ performance plans for 2010. Despite the lack of department input in the four component management chiefs’ performance agreements mentioned above, in reviewing the agreements we found that some of them included a link to the Management Directorate Strategic Plan, a management function strategy or plan, or the DHS Strategic Plan Goal 5 for management. Some performance agreements also referred to supporting department-level efforts, with references to activities such as supporting department-level strategic plans and council activities and implementing departmentwide management initiatives. In addition to input into component chief performance agreements, management directives require department chief feedback to the component rating official regarding the component chiefs’ accomplishment of annual objectives. The CFO, CSO, and CAO told us that they provided input into component chiefs’ performance appraisals, while the CIO and CPO did not provide input. The CPO stated that he would be providing input beginning with the fiscal year 2010 performance appraisals. The CHCO said that, due to his limited tenure in the position, he could not state whether input had occurred. In addition to individual input, department chiefs have the opportunity to review the component chiefs’ performance ratings and bonuses and / or pay adjustments at the conclusion of the department Performance Review Boards prior to their approval by the deputy secretary. While this assessment provides an additional opportunity for department oversight, it does not satisfy the management directives’ requirement for input by the department chief to the component rating official. The USM said that departmental chiefs’ input into component chiefs’ performance appraisals would be a priority in the future. In summary, performance management practices to help ensure accountability for management integration between the department and component management chiefs are not consistently in place. While linkages are being most clearly defined at the department chief level within their individual management functions, department chiefs are not consistently providing the guidance and input required by department management directives and in accordance with performance management leading practices. The inconsistent application of such guidance and practices presents challenges to institutionalizing individual accountability and enabling the effective exercise of authority at the department. Without ensuring that the management chiefs provide input into component chiefs’ performance plans and evaluations as required, the Management Directorate cannot be sure that component chiefs are fully implementing management integration. In the more than 6 years since its establishment, DHS has taken actions that could help it transform organizationally and integrate its management functions to establish a unified department. In particular, the department has taken actions to vertically integrate the component agencies by developing common policies, procedures, and systems within individual management functions, such as human capital and information technology However, DHS has placed less emphasis on integrating horizontally, and bringing together these multiple management functions across the department. In addition, key characteristics that are necessary to guide and ensure successful management integration are not yet in place, such as identification of trade-offs, priorities, and implementation goals, and the implementation and transformation of the department remains on our high-risk list. Current plans are a step in the right direction, but in the absence of a comprehensive strategy for management integration as required by the 9/11 Commission Act and meeting all of the previously identified characteristics for such a strategy, it is unclear how management integration will be more fully achieved across the department. We therefore reiterate our prior recommendation, not yet fully implemented, that DHS develop a comprehensive management integration strategy. We continue to believe that a comprehensive strategy for management integration is warranted, and would help the department to ensure that its management initiatives are implemented in a coherent way. It would also help DHS to communicate its approach for management integration and measures for evaluating progress made. Moreover, while DHS has been implementing management initiatives and processes across the department, in the absence of a comprehensive management integration strategy, it is unclear how these efforts are being prioritized and sequenced, and trade-offs between them are being recognized. In addition, a comprehensive strategy for management integration would help the department establish performance measures to better gauge its progress in integrating its various management policies, processes, and systems across DHS. Although the department has developed certain management measures, these measures do not allow the department to assess the extent to which it is making progress in implementing and achieving management integration both within and across functional areas. The “dotted line” reporting relationships between the department chiefs will be particularly important once DHS develops a management integration strategy that would involve decisions and trade-offs that are dependent on component compliance to succeed. Implementation of existing performance management mechanisms—such as the departmental management chiefs’ input into component chiefs’ performance plans and evaluations, and linkages between department goals and objectives and individual performance plans for component management chiefs—is necessary to ensure that the Management Directorate can exercise its authority and leadership to implement a management integration strategy. To strengthen its management integration efforts, we recommend that the Secretary of Homeland Security direct the Under Secretary for Management, working with others, to take the following four actions: Once a comprehensive management integration strategy is developed, consistent with statute and as we previously recommended, establish performance measures to assess progress made in achieving departmentwide management integration; Ensure that department management chiefs provide written objectives for component management chiefs’ performance plans at the beginning of each performance cycle, and that the objectives are representative of determined priorities and milestones for the management functions during that period; Ensure that department management chiefs provide input into component management chiefs’ annual performance evaluations; and Ensure that component management chiefs’ individual performance plans are reflective of and include linkages to the goals and objectives for the Management Directorate and relevant department management function. We provided a draft of this report to the Secretary of the Department of Homeland Security for comment. In written comments on a draft of this report, the DHS Under Secretary for Management provided information on steps the department was taking or planning to take to develop a strategy for management integration, as we had recommended in our 2005 report, and to link this strategy to SES performance appraisals for the management chiefs. Specifically, the Under Secretary for Management said that she is leading the process for developing a detailed, measurable plan that will include the actions and milestones necessary to accomplish management integration at the department. Additionally, the Under Secretary for Management stated that the integration plan will be tied to the SES performance appraisals for each management chief for the fiscal year 2010 performance cycle, and that the plan will also serve as the required annual performance agreement between the Secretary and the Under Secretary for Management. While DHS’s letter did not directly comment on our recommendations in this report related to the need for performance measures for management integration and additional steps needed to strengthen accountability for the management chiefs, the Director of DHS’s Internal Control Program Management Office noted in a subsequent e-mail that DHS concurred with our report and its written comments were intended to discuss steps to implement the recommendations in our report. DHS’s written comments are contained in appendix I. We incorporated technical comments provided by DHS as appropriate. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Department of Homeland Security and other interested parties. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any further questions about this report, please contact me at (202) 512-6543 or steinhardtb@gao.gov, or David Maurer, Director, at (202) 512-9627 or maurerd@gao.gov. Points of contact for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix II. In addition to the contact named above Sarah Veale, Assistant Director; Rebecca Gambler, Assistant Director; S. Mike Davis; Barbara Lancaster; Jared Hermalin; Bion Bliss; Jyoti Gupta; Tom Beall; and Karin Fangman made significant contributions to this report. Financial Management Systems: DHS Faces Challenges to Successfully Consolidate its Existing Disparate Systems. GAO-10-210T. Washington, D.C.: October 29, 2009. Homeland Security: Despite Progress, DHS Continues to Be Challenged in Managing Multi-Billion Dollar Annual Investment in Large-Scale Systems. GAO-09-1002T. Washington, D.C.: September 15, 2009. High-Risk Series: An Update. GAO-09-271. Washington, D.C.: January 2009. Department of Homeland Security: A Strategic Approach Is Needed to Better Ensure the Acquisition Workforce Can Meet Mission Needs. GAO-09-30. Washington, D.C.: November 19, 2008. Department of Homeland Security: Billions Invested in Major Programs Lack Appropriate Oversight. GAO-09-29. Washington, D.C.: November 18, 2008. Department of Homeland Security: Progress Made in Implementation of Management Functions, but More Work Remains. GAO-08-646T. Washington, D.C.: April 9, 2008. Organizational Transformation: Implementing Chief Operating Officer / Chief Management Officer Positions in Federal Agencies. GAO-08-322T. Washington, D.C.: December 13, 2007. Department of Homeland Security: Better Planning and Assessment Needed to Improve Outcomes for Complex Service Acquisitions. GAO-08-263. Washington, D.C.: April 22, 2008. Organizational Transformation: Implementing Chief Operating Officer / Chief Management Officer Positions in Federal Agencies. GAO-08-34. Washington, D.C.: November 1, 2007. Homeland Security: DHS Enterprise Architecture Continues to Evolve but Improvements Needed. GAO-07-564. Washington, D.C.: May 9, 2007. Federal Real Property: DHS Has Made Progress, but Additional Actions Are Needed to Address Real Property Management and Security Challenges. GAO-07-658. Washington, D.C.: June 22, 2007. Homeland Security: Departmentwide Integrated Financial Management Systems Remain a Challenge. GAO-07-536. Washington, D.C.: June 21, 2007. Department of Homeland Security: Progress Report on Implementation of Mission and Management Functions. GAO-07-454. Washington, D.C.: August 17, 2007. Information Technology: DHS Needs to Fully Define and Implement Policies and Procedures for Effectively Managing Investments. GAO-07-424. Washington, D.C.: April 27, 2007. Department of Homeland Security: A Comprehensive and Sustained Approach Needed to Achieve Management Integration. GAO-05-139. Washington, D.C.: March 16, 2005. Results-Oriented Cultures: Implementation Steps to Assist Mergers and Organizational Transformations. GAO-03-669. Washington, D.C.: July 2, 2003.

Significant management challenges exist for the Department of Homeland Security (DHS) as it continues to integrate its varied management processes, policies, and systems in areas such as financial management and information technology. These activities are primarily led by the Under Secretary for Management (USM), department management chiefs, and management chiefs in DHS's seven components. The Government Accountability Office (GAO) was asked to examine: (1) the extent to which DHS has developed a comprehensive strategy for management integration that includes the characteristics recommended in GAO's 2005 report; (2) how DHS is implementing management integration; and (3) the extent to which the USM is holding the department and component management chiefs accountable for implementing management integration through reporting relationships. GAO reviewed DHS plans and interviewed management officials in DHS's headquarters and in all components. DHS has not yet developed a comprehensive strategy for management integration as required by the 9/11 Commission Act of 2007 and with the characteristics GAO recommended in a 2005 report. Although DHS stated in response to the 2005 report that it was developing an integration strategy, it has not yet done so, in part because it has focused on building operations capacity within functional management areas. In the absence of a comprehensive management integration strategy, DHS officials stated that documents such as management directives and strategic plans address aspects of a management integration strategy and can help the department to manage its integration efforts. However, they do not generally include all of the strategy characteristics GAO identified, such as identifying the critical links that must occur among management initiatives and time lines for monitoring the progress of these initiatives. In addition, DHS has increased the number of performance measures for the Management Directorate, but has not yet established measures for assessing management integration across the department, although DHS officials stated that the department intends to do so. Without these measures DHS cannot assess its progress in implementing and achieving management integration. In the absence of a comprehensive strategy, DHS's Management Directorate has implemented management integration through certain initiatives and mechanisms to communicate and consolidate management policies, processes, and systems. The directorate uses councils to communicate information related to the implementation of management initiatives, among other things. The directorate has also established governance boards and processes to manage specific activities. Further, the directorate is in the process of consolidating certain management systems. However, without a documented management integration strategy, it is difficult for DHS, Congress, and other key stakeholders to understand and monitor the critical linkages and prioritization among these various efforts. The USM and department and component management chiefs are held accountable for implementing management integration through reporting relationships at three levels--between the Secretary and the USM, the USM and department chiefs, and the department and component chiefs--in which, among other things, the Secretary of Homeland Security, USM, and department chiefs are required to provide input into performance plans and evaluations. The Deputy Secretary--through delegation from the Secretary--and the USM have provided input into the USM's and department chiefs' plans and evaluations, respectively. Although department chiefs are required by management directives to provide component chiefs with written objectives at the start of the annual performance cycle, in fiscal year 2009 only two out of six department chiefs provided such input to component chiefs. Without ensuring that the management chiefs provide input into component chiefs' performance plans and evaluations as required, the directorate cannot be sure that component chiefs are fully implementing management integration.

DOD faces a number of long-standing and systemic challenges that have hindered its ability to achieve more successful acquisition outcomes, such as ensuring that DOD personnel use sound contracting approaches and maintaining a workforce with the skills and capabilities needed to properly manage the acquisitions and oversee contractors. While the issues encountered in Iraq and Afghanistan are emblematic of these systemic challenges, their significance and effect are heightened in a contingency environment. For example, in 2004, we raised concerns about DOD’s ability to effectively administer and oversee contracts in Iraq, in part because of the continued expansion of reconstruction efforts, staffing constraints, and the need to operate in an unsecure and threatening environment. Similarly, we reported in July 2007 that DOD had not completed negotiations on certain task orders in Iraq until more than 6 months after the work began and after most of the costs had been incurred, contributing to its decision to pay the contractor nearly all of the $221 million questioned by auditors. In 2008, we reported that not having qualified personnel hindered oversight of contracts to maintain military equipment in Kuwait and provide linguistic services in Iraq and questioned whether DOD could sustain increased oversight of its private security contractors. The contract closeout process includes verifying that the goods or services were provided and that all final administrative steps are completed, including an audit of the costs billed to the government and adjusting for any over- or underpayments on the final invoice. To close a contract, DOD must complete a number of tasks, including making final payment to the contractor, receiving a release of claims from the contractor, and deobligating excess funds, among other tasks (see fig. 1). A contract is eligible to be closed once the contract is physically complete, which is generally when all option provisions have expired and the contractor has completed performance and the government has accepted the final delivery of goods or services in the form of a receiving report or the government has provided the contractor a notice of complete contract termination. From this point, contracts should be closed within time frames set by the FAR—6 months for firm-fixed price contracts and 36 months for cost-type contracts and time and materials contracts. Additional time is allowed for the closeout of these latter contract types as the contracting officer and DCAA may need to ensure any incurred costs are allowable, allocable, and reasonable. Additional time is also needed to set the final indirect overhead rates, which determine, in part the contractor’s final payment on cost-type contracts. When the contract completion statement, also known as the DD 1594, is signed by the contracting officer, the contract is considered closed and contract documents can be stored and retained. A contract not closed within the FAR time frames is considered to be over age for closeout and increases an organization’s exposure to a number of financial issues. If contract closeout does not take place in a timely manner and funds are not deobligated when currently available, the agency loses the use of those funds for new obligations. Even if funds are expired when they are deobligated, the agency can still use them for up to 5 years after they expire to pay for authorized increases to existing obligations made from the same appropriation. Any funds remaining after the 5-year period are considered canceled and must be returned to Treasury. If closeout does not take place until after they are canceled, and the agency identifies a need for the government to pay the contractor for an unanticipated cost, the government must use other funds that are currently available. Additionally, the risk of late payments to contractors increases when contracts are not closed within required time frames and in turn may result in the government paying interest. Further, the longer an organization waits to close a contract the more difficult it becomes to identify and recover improper payments to contractors. In addition, closing a contract years after the performance is complete can be more time consuming because key documentation, such as invoices and receiving reports, and contracting personnel with first-hand knowledge of the contract may no longer be available. DOD does not have visibility into the total number of its Iraq contracts eligible for closeout, but our analysis of available data indicates that relatively few of these contracts will be closed within the time frames prescribed by the FAR. C3, which awarded the majority of the Iraq contracts, did not have sufficient internal controls to ensure its contracting data were accurate and complete, and was further affected by limitations of its contracting systems, turnover in contracting personnel, and other competing demands. In 2009, to help reduce the backlog of contracts to be closed, C3 transferred 66,760 Iraq contracts and 14,336 contracts in which a place of performance was not specified to the Task Force. As it was unclear how many of these contracts were closed before being shipped, Task Force personnel are in the process of reviewing each contract and, as appropriate, closing any open contracts. As of April 2011, however, over 54,000 of these contracts still needed to be reviewed. DOD officials noted that record keeping generally improved for C3’s firm-fixed price contracts awarded after fiscal year 2008. C3 also improved visibility of its large, cost-type contracts awarded between fiscal year 2003 and 2010 after delegating contract administration, including closeout responsibilities, to DCMA Southern Europe in 2008. Based on available data provided by C3 and the other DOD contracting organizations we reviewed, there are at least an additional 4,298 Iraq contracts—90 percent of which are already over age—that need to be closed. C3 and its predecessor organizations awarded the majority of DOD’s contracts to support reconstruction and stabilization efforts, yet weak internal controls, turnover in contracting personnel, and competing demands contributed to incomplete or inaccurate information that hindered management oversight of its contracting activities, including whether it was meeting FAR closeout requirements. DOD officials noted C3 did not have a contract writing and management information system in Iraq between 2003 and 2008, which contributed to the use of multiple manual databases. Each regional contracting center awarded manually written contracts and documented contract actions on independent spreadsheets. C3 and Army officials noted some of the challenges with manually written contracts included duplicate or inaccurate contract numbers and inaccurate period of performance dates. They also noted that each regional contracting center maintained and managed its contract data on spreadsheets differently as there was not an Iraq-wide standard for how to maintain contract data and that data input was often unverified. These contract documentation challenges were exacerbated by the constant turnover of contracting personnel and the command’s emphasis on awarding contracts to support the warfighter. Additionally, C3 and Army officials said that an unknown number of contracts were never input into C3’s database and could not be accounted for because contract files were lost, damaged, or destroyed. Our analysis of C3’s data on its Iraq contracts found at least 55,000 contracts were recorded as being awarded between fiscal years 2003 and 2008, but we determined that the data had numerous discrepancies. These discrepancies, which included missing or invalid period of performance and physical completion dates as well as invalid or duplicative contract numbers, affect the data needed to maintain visibility on the contracts eligible to be closed. Army officials acknowledged that the contract information reflected in C3’s database through fiscal year 2008 was unreliable for determining the actual number of contracts it awarded or which contracts were eligible to be closed. Consequently, the Army underestimated the total number of contracts that the Task Force needed to close. In 2008, the Army estimated that the Task Force would need to close approximately 24,000 contracts awarded by C3 in Iraq and Afghanistan from 2003 to 2008, but the Task Force recorded that C3 sent it 103,693 contracts (see table 1). Our analysis of the Task Force’s data indicates that C3 transferred at least 66,760 Iraq contracts, including approximately 8,500 more contracts awarded between fiscal years 2003 and 2008 than what was reflected in C3’s database. Additionally, the Task Force inventoried another 14,336 contracts for which the place of performance was not specified. Army officials stated that C3 had closed some of these contracts before sending the files to the Task Force, but acknowledged that the C3 data did not accurately reflect which contracts were closed. Therefore, the Army required Task Force personnel to review each contract and close those that remain open. Army officials stated, however, that there have been no attempts to reconcile the C3 contracting data with the Task Force’s findings. The extent to which the contracts that have not yet been reviewed by Task Force personnel and will need to be closed is uncertain, in part, because some that were reportedly closed by C3 still required contract administration. For example, Task Force personnel stated that contracts sometimes included a signed DD 1594 even though the contracts still required administrative actions. To improve the management of its contracts, C3 began using the Standard Procurement System in fiscal year 2009. Both Army and C3 officials stated that the Standard Procurement System had better quality control checks to generate valid contract numbers with automated prompts requiring contracting personnel to insert required data fields, such as period of performance, at the time of award. These officials also said that the quality control checks improved the completeness and quality of C3’s data and provided better insight needed to manage the contract closeout process. Army officials said that once the Standard Procurement System was deployed in Iraq, the regional contracting centers were able to transmit data back to Army locations in the United States which could be used to run automated reports on contracts closed, eligible for closeout, and over age for closeout. Army and C3 officials acknowledged that while the data improved, C3 continued to identify problems with the data input by contracting personnel. In a July 2010 memorandum, C3 directed its personnel to take actions to improve the overall quality, accuracy, and timelines of C3’s contracting actions. For example, it identified specific data fields, including those that help to determine a contract’s eligibility for closeout, that personnel are required to capture in C3’s data systems. C3 obtained better visibility of its firm-fixed price contracts awarded in fiscal years 2009 and later as well as their large, cost-type contracts. C3’s data on these firm-fixed price contracts indicates that C3 closed over 9,600 of its Iraq contracts awarded between fiscal years 2009 and 2010. Similarly, DOD officials indicated that C3 had better visibility of its large, cost-type contracts awarded between fiscal years 2003 and 2010, in part because it generally delegated contract administration for these contracts, including closeout responsibilities, to DCMA Southern Europe in 2008. DCMA officials reported that when it accepted C3’s cost-type contracts, the files were in generally poor condition and missing documents. DCMA officials reported, however, that they devoted the resources necessary to collect missing information for these contracts and developed their own data to manage the closeout of these contracts and task orders. Our analysis of these firm-fixed price and cost-type contracts indicates that 97 percent were over age as of May 2011 (see table 2). ACC-RI, AFCEE, and USACE officials indicated that the use of existing contracting systems at the onset of military operations in Iraq provided them better visibility into the number of contracts they had awarded to support efforts in Iraq. Agency officials acknowledged, however, that they sometimes encountered challenges with using their existing systems. For example, USACE officials noted that the standard reports used to determine which Iraq contracts needed to be closed were initially inaccurate because period of performance or physical completion dates were not correctly entered into their contracting systems. As a result, USACE officials found in March 2011 that USACE’s closeout reports underestimated the number of contracts eligible and over age for closeout due to inaccurate period of performance dates. USACE revised its reports using period of performance dates from other data sources, which identified that 639 contracts were eligible to be closed, more than 300 contracts than its initial report reflected. Similarly, AFCEE’s data indicate that the period of performance ended for 154 of its Iraq contracts but the data did not reflect whether final goods and services had been delivered and whether the contract was physically complete. Our analysis indicates that the period of performance ended at least 3 years ago for 37 of these contracts, but AFCEE personnel stated that they cannot close these contracts until they receive final documentation that the goods and services have been delivered. Overall, we estimate that about 66 percent of these organizations’ 907 eligible contracts are over age (see table 3). Our analysis of data provided by these contracting organizations reflects a higher percent of eligible firm-fixed price contracts that are over age compared to eligible cost-type contracts, in part due to the longer period of time allowed by the FAR to close out cost-type contracts. For example, our analysis indicates that about 81 percent of the firm-fixed price contracts eligible to be closed were over age compared to approximately 40 percent of eligible cost-type contracts. Nevertheless, these organizations have closed few of their cost-type Iraq contracts. For example, USACE data indicate that it had closed 7 of its 77 Iraq cost-type contracts and AFCEE had closed just 10 of its 239 Iraq cost-type contracts awarded since 2003. DOD’s ability to close the contracts it awarded to support efforts in Iraq is hindered by several factors, including the failure to plan for or emphasize the need to close these contracts until reconstruction efforts were well underway, staffing shortfalls, and contractor accounting issues. DOD did not plan for or focus on closing its Iraq contracts until 2008, in part because DOD’s contingency contracting policy and guidance do not emphasize the need to plan for contract closeouts during the early stages of a contingency operation. DOD has taken steps to reduce the number of firm-fixed price contracts it needs to close, but ACC-RI has not been able to hire enough personnel to replace Task Force personnel during the transition of closeout responsibilities, which has slowed these efforts. Similarly, efforts to close its large, cost-type contracts is hindered by staffing shortages at DCAA and unresolved issues with contractors’ cost accounting practices that preclude completing the necessary audits of the contractors’ incurred costs. As a result, DOD is unlikely to close 226 cost- type contracts with over $19.1 billion in obligations in the near future. DOD contingency contracting doctrine and policy do not specifically include closeout as part of the advanced planning for a contingency operation. Since 2006, a contract support integration plan annex termed Annex W—which provide details on the contractor support required during a contingency, including the military’s organizational requirements needed to acquire and oversee such support—has been required to be in DOD’s most detailed operation plans. In October 2008, DOD established its first doctrine to standardize guidance for planning, conducting, and assessing operational contract support integration, contractor management functions, and contracting command and control in support of joint operations in its Joint Publication 4-10, Operational Contract Support. In part, this doctrine provides guidance for contingency contracting requirements that should be planned for within the Annex W. While it states that an Annex W should outline all activities necessary to execute contract support integration requirements in an operational area, it does not specifically direct DOD commands to determine an approach for closing contracts in advance or even during the initial stages of a contingency operation. Joint Publication 4-10 advises that contracts be closed as performance is completed, consistent with the requirements established in the FAR, but makes no reference for the need to plan for the resources needed to close contracts within required time frames. Instead, contract closeout is described as part of the redeployment and contract termination phase, the fourth and final operational phase of a contingency. In 2009, DOD issued a template for planners to use when developing Annex Ws and plans to incorporate the template into planning policy. The template does not, however, specifically call attention to the need to plan for the closeout of contracts. Furthermore, in March 2010, we reported that few of the operation plans approved by the Secretary of Defense or his designee even included an Annex W and when they did, those annexes restated broad language from DOD’s high-level guidance on operational contract support. The contracting organizations included in our review generally did not conduct any planning to close the contracts they awarded to support operations in Iraq until several years after the contracts were initially awarded. DOD officials noted that the department initially assumed that post-conflict stability and reconstruction efforts would not last for an extended period and as such, any organization that awarded contracts to support these efforts would close contracts under the organization’s standard processes. Officials acknowledged that as these efforts continued and the level of contracting activity increased, C3’s predecessors attempted to close contracts as time and resources permitted, but did not develop a plan needed to do so. For example,  The Army did not develop a plan to close its Iraq contracts until 2008, long after reconstruction efforts were underway in Iraq. According to the Army, the 2007 Gansler Commission report’s finding that only 5 percent of eligible Iraq contracts were closed prompted the Army to begin planning for and taking steps to address the backlog of over- age Iraq contracts. To do so, in October 2008, the Army established the Task Force and delegated responsibility to DCMA Southern Europe to close a number of C3’s cost-type contracts.  According to USACE personnel, they began focusing on contract closeouts after the Army identified that the Army had more than 660,000 over-age contracts as of January 2009 and established a goal to close all of its over-age contracts by the end of fiscal year 2011. In January 2011, USACE established a contract closeout cell in Winchester, Virginia.  AFCEE personnel, with 96 over-age Iraq contracts, stated they have not developed an Iraq contract closeout plan and continue to close these contracts as part of their routine contracting activities. AFCEE personnel stated, however, only two contracting personnel are assigned to closing the Iraq contracts and do so only when time and other responsibilities permit. DOD officials also noted that the need to focus limited staff resources on fulfilling urgent requirements in support of the war effort, and other contingency-related challenges, contributed to the backlog of contracts to be closed. One senior Army official noted that as there were not enough contracting officers in theater to handle both awards and closeouts, the command focused its attention on awarding contracts. Similarly, C3 and USACE contracting personnel we spoke with stated that they were responsible for awarding, administering, and closing contracts, but to meet urgent requirements, they prioritized contract awards over other activities. In addition, an Army official noted that contracting personnel have little incentive to close contracts, as their success is often measured by contracts awarded. Contracting personnel who are responsible for closing contracts stated, however, that emphasis on timely contract closeout is especially important in a contingency environment because the longer the time from when the contractor completes its work and when the contract is closed, the more difficult it becomes to determine the status of contracts, resolve documentation and administration issues, obtain a release of claims, and negotiate final payments. For example,  To close a $16.8 million guard services contract, contracting personnel in Iraq described the process of determining how payments were made as “putting together pieces of a puzzle.” Personnel stated that they spent several weeks identifying what the contractor billed and was paid by reviewing invoices, contract modifications, and e- mails.  Similarly, contracting personnel in Iraq stated that resolving an overpayment of over $500,000 has delayed the closeout of another $17 million guard services contract. The contracting officer who awarded and administered the contract was no longer in Iraq when the contracting personnel began closing the contract. These personnel stated that they relied on e-mails in the contract file and obtained payment information from DFAS to determine the extent to which the contractor was overpaid and are awaiting further guidance from DFAS on what steps are needed to recover funds from the contractor.  Task Force personnel noted that while closing a $1.3 million contract for life support services, they found that there was no documentation in the contract file to explain why services were not performed at three camp sites listed in the contract. The contractor told Task Force personnel that he was instructed not to perform the services but was never provided anything in writing. Task Force personnel noted that the contractor then refused to sign the release of claims, so personnel unilaterally deobligated the remaining funds on the contract to close the contract.  According to one senior C3 official, contracting officers sometimes relied on documents provided by the contractor to resolve claims because they were not maintained in the contract files. In one instance, while closing a vehicle lease contract, C3 personnel stated that they found 149 damage claims for vehicles, but oversight personnel often did not keep records or pictures of the condition of the vehicles when they were picked up and dropped off by the contractor. The contracting personnel stated that they are coordinating with the payment office and resource managers but said that it may not be possible to locate someone who can verify or dispute the claims. Task Force personnel stated that they often needed to perform routine contract administration tasks on contracts, including reconciling payments and obligations, acquiring receiving reports, contacting contractors in theater to obtain invoices and release of claims, and piecing together incomplete contract files to provide reasonable assurance that the government received what it paid for and the contract could be closed. Task Force personnel illustrated some of the challenges they often encounter in the following two examples: In one case involving the closeout of a $55 million contract for shotguns, goggles, and radios, Task Force personnel stated that they had to reconcile payments against nine different task orders because payments were not made to the correct task orders, including one lump-sum payment for $8 million that did not correspond to any task order, and the contract was missing receiving reports and payment documents. Task Force personnel contacted DFAS to determine how much should have been paid on the task order and verified payments through a data system. Task Force personnel eventually closed all of the task orders between March and December 2010.  During the closeout of another contract for $101,000 to lease buses from an Iraqi contractor, Task Force personnel found that the contractor was not paid for 1 month of service and not compensated for damages to two of the buses. After contacting DFAS and determining that there were enough funds on the contract to cover the missing payment and repair costs, Task Force personnel notified the payment office to make a final payment to the contractor. Task Force personnel were able to close the contract after the contractor was paid and a release of claims was received. C3 has taken steps to reduce the number of firm-fixed price contracts it needs to close, but difficulties with hiring ACC-RI personnel have slowed these efforts. The Army and C3 initially established the Task Force to address the backlog of C3’s firm-fixed price contracts awarded before fiscal year 2009 and planned at that time to close any contracts awarded in fiscal year 2009 and later in theater. To ensure the contracts remaining in theater were closed, a senior C3 official established closeout goals in October 2010 and required each regional contracting center to appoint personnel responsible for completing contract closeout. While Army data indicate that progress was made in closing contracts in Iraq, C3 officials told us that closeout goals were tracked informally and acknowledged that some regional contracting centers were unable to meet these goals. By February 2011, the Army changed its strategy and decided that when the Task Force is shut down in September 2011, all C3 contracts, including those awarded after fiscal year 2009, would eventually be transferred to ACC-RI for closeout. According to C3’s commanding general, this decision was made because ACC-RI has a workforce that can handle complex contract actions and has expertise in southwest Asia contracting. By June 2011, the Army had transferred about 15,000 Iraq and Afghanistan contracts awarded between fiscal years 2008 and 2010 from the Task Force to ACC-RI. According to the Army, ACC-RI personnel are in the process of inventorying these contracts and identifying which are closed or require additional administration. Army officials stated that they are reviewing ACC-RI closeout procedures and data collection efforts to ensure Army data are accurate and complete. During this transition period, ACC-RI has not been able to hire the number of individuals it estimated it needed to manage the anticipated workload and the number of contracts reviewed and closed by the Task Force has fallen considerably. According to Army officials, ACC-RI will need to hire 25 individuals by the time it fully assumes the Task Force’s responsibilities. Army officials stated that ACC-RI has experienced challenges hiring contracting personnel in part due to potential applicants’ hesitation to accept these positions, which are term positions that expire by October 2012. Army officials stated as of June 2011, ACC-RI had only hired 4 staff but efforts are underway to hire additional personnel. Until these positions can be filled, other ACC-RI personnel are temporarily supporting the closeout efforts. In addition, in July 2011, the ACC-RI issued a task order for contract closeout support to AbilityOne, which provides job opportunities on federal contracts for individuals who are blind or have other disabilities. According to one ACC-RI official, ACC-RI plans to hire nine AbilityOne employees under this contract. It remains uncertain, however, when the Army will be able to review and, as necessary, close the contracts that remain at the Task Force. Similarly, Army officials stated that the Task Force’s capacity to close contracts has decreased, as 10 of its 25 staff have resigned in advance of the Task Force’s planned closure. During the week of September 3, 2010, the Task Force closed 439 contracts but by the week of June 9, 2011, the Task Force only closed 267 contracts. DOD’s efforts to close its large, cost-type contracts are hindered by staffing shortages at DCAA and unresolved issues with contractors’ cost accounting practices. DOD reported that it had 226 over-age, cost-type Iraq contracts with approximately $19.1 billion in obligations (see table 4). A critical step to closing these contracts is to determine how to allocate a contractor’s general administrative and overhead costs to each of its contracts. To do so, DCAA performs annual incurred cost audits on a contractor-by-contractor basis—versus a contract-by-contract basis—by reviewing incurred cost proposals from the contractor for each year of performance. DCAA auditors test direct and indirect costs to determine whether they are allowable, allocable, and reasonable. The direct and indirect costs form the basis for DCAA’s recommended indirect cost rate, which is usually used by the contracting officer to negotiate a final rate with the contractor. When the indirect cost rate for the final year of contract performance is settled and the final price of the contract is determined, contract closeout may proceed. DOD’s cost-type contracts related to Iraq often spanned multiple years and as such DCAA must complete incurred cost audits for each year of performance. For example, on one contract with performance from 2004 through 2008 and 5 divisions of the contractor claiming costs, DCAA is required to complete 25 audits of costs incurred, one for each year of performance per division. DCAA, however, is still completing audits for this contractor for costs incurred in 2004 and 2005, with audits of the remaining years scheduled for 2011 and after. DCAA officials told us that this condition is due in part to a DCAA-wide shortage of auditors. DCAA data indicates that from fiscal years 2000 to 2011, its workforce grew by 16 percent while DOD research and procurement spending, an indicator of DCAA’s workload, increased by 87 percent. In addition, DCAA officials stated that in response to GAO’s finding in 2009 on problems with DCAA’s audit quality, including insufficient testing of contractors’ support for claimed costs, DCAA now requires more testing and stricter compliance with government auditing standards, which adds to the amount of staff time required to complete each audit. DCAA officials stated that as their workload increased and resources remained relatively constant, auditors prioritized time-sensitive activities, such as audits to support new awards, and incurred cost audits were not completed, creating a backlog. In planning for its fiscal year 2011 workload requirements, DCAA determined that it had the resources to complete only about half of its entire portfolio of required audits and activities, including both Iraq and non-Iraq work. As a result, DCAA prioritized its high-risk audits, which included the backlog of incurred audits for C3’s 106 over-age, cost-type contracts. As of July 2011, DCAA reported that of the 116 incurred cost audits needed to close these C3 contracts, it had completed 27 audits and estimated another 19 audits will be completed by the end of fiscal year 2011. The remaining 70 audits are planned to be completed after fiscal year 2011. DCMA contracting officials responsible for closing C3’s cost-type contracts stated that regardless of whether DCAA completes the 19 audits as planned, none of the C3 contracts can be closed by the end of fiscal year 2011 because most of the contractors claimed costs through 2008 or 2009, and the audits will only be completed for costs incurred mostly through 2004 and 2005. Further, there are an additional 31 AFCEE over-age cost-type contracts that will not have final incurred cost audits completed before the end of fiscal year 2011. To address its resource challenges, DCAA officials reported that it hired over 500 new employees in the past 2 years. DCAA has also requested authority to hire 200 auditors per year over each of the next 5 years. DCAA officials noted, however, that it often takes several years before auditors are properly trained to conduct an incurred cost audit. In addition, in January 2011, DOD issued a memorandum that shifted some audit responsibilities, such as lower dollar price proposal audits and purchasing system reviews, to DCMA to allow DCAA to devote more resources to high-risk work, like the incurred cost audits needed to support the closeout of Iraq contracts. DCAA officials also stated that they plan to dedicate additional auditors to solely focus on conducting incurred cost audits in fiscal year 2012. DCAA has identified a number of deficiencies at major defense contractors, which provided support in Iraq, that need to be resolved before the incurred cost audits can be completed. These deficiencies include    accounting practices that are not compliant with cost accounting inadequate incurred cost proposals and cost documentation; inadequate contractor business systems; standards, leading to misallocation of costs;  delays in providing DCAA access to needed records;  disputes with contractors over unallowable costs; and  other challenges, such as those due to ongoing litigation. The following examples illustrate the challenges that DCAA reported for several contractors.  Due to inadequate incurred cost proposals, DCAA has completed incurred costs audits only through 2003 for one major Iraq contractor that incurred costs through 2010. In total, DOD has $15.3 billion in obligations on over-age, cost-type Iraq contracts awarded to this contractor. DCAA reported that it issued the 2003 incurred cost audit 5 years after the costs were incurred, in part because the contractor repeatedly submitted inadequate incurred cost proposals and did not provide adequate support for costs (see fig. 2). Further, DCAA officials stated that the incurred cost proposals submitted by the contractor for 2004 through 2009 are inadequate but will continue its audits of the 2004 and 2005 proposals.  Additionally, DCAA reported that this contractor had deficient accounting systems, unresolved issues associated with unallowable costs, noncompliant accounting practices, and legal investigations that further delayed incurred cost audits. In 2006, DCAA reported that the contractor had significant deficiencies in its accounting system that resulted in the contractor charging over $370 million to incorrect task orders from 2002 to 2004, requiring reclassification of costs to the proper task orders. The reclassifications were completed in January 2005. Then, in 2009 and 2010, DCAA found over $185 million in unallowable costs that are pending negotiations with DCMA and settlement of contractor claims. In 2010, DCAA auditors found the contractor did not comply with the cost accounting standard associated with insurance costs, which resulted in an estimated $1.6 million in costs that were misallocated. DCAA reported that the contractor did not respond to DCAA’s finding because it had not completed its management review of the allocated costs. Further, according to the auditors, DCAA’s incurred cost audit reports could be delayed as the auditors coordinate the issuance of audit reports with various investigative agencies. DCAA auditors do not expect to complete the 2004 and 2005 incurred cost audits for this contractor before the end of fiscal year 2011. In May 2011, the contractor withdrew its 2006 through 2009 incurred cost proposals and stated that it plans to delay its submission of the 2010 incurred cost proposal until November 2011.  For another major contractor, DCAA identified that the contractor’s accounting practices were not compliant with cost accounting standards. DOD has $316 million in obligations on over-age, cost-type Iraq contracts awarded to this contractor with performance between 2004 and 2009. DCAA reported in 2006 that the contractor’s accounting practices did not sufficiently remove unallowable costs from a cost proposal, which DCAA auditors stated put additional onus on them to test whether the costs were allowable. In one case, DCAA auditors found the contractor had included over $500,000 in bonuses to senior executives in the incurred cost proposal, even though these costs are expressly unallowable under law. The contractor disagreed with DCAA’s findings but agreed to remove these costs from its proposal. As of July 2011, DCAA has completed 5 of the 18 incurred cost audits required to close the contracts.  DCAA identified deficient subcontract management systems, disputes over unallowable costs, and challenges with access to records as contributing to delays in completing incurred cost audits for another contractor. DOD has $212 million in obligations on over-age, cost-type Iraq contracts awarded to this contractor. In 2005, 2006, and 2009, DCAA auditors reported significant deficiencies in the contractor’s subcontract management system that resulted in potential unreasonable and unallowable costs being billed to the government, subcontracts being awarded noncompetitively, and inadequate price analysis. As a result, DCAA auditors had to audit the subcontractors’ costs, even though doing so is generally the prime contractor’s responsibility. The contractor generally disagreed with DCAA’s findings but stated it would evaluate and revise its procedures where necessary to comply with DCAA’s recommendations. In addition, in 2010 and 2011, DCAA auditors reported the contractor had over $22.5 million in unallowable subcontract costs, some of which have been appealed by the contractor and some of which are being settled by DCMA. Finally, in 2010, DCAA auditors repeatedly requested but were denied access to support for the 2006 incurred cost proposal, including a $2.3 million procurement file. DCAA reported that its auditors requested the data over a period of 5 months and stated that when the contractor provided the data, they were still inadequate in supporting the claimed costs. DCAA auditors stated as a result, it deemed those costs as unallowable for reimbursement. DOD has taken steps to address the challenges with auditing contractors’ incurred costs. For example, effective June 2011, the FAR was revised to list the minimum information that contractors must include for proposals to be adequate to address the delays resulting from inadequate incurred cost proposals. Also, to improve its oversight of contractor business systems, DOD revised the Defense Federal Acquisition Regulation Supplement in May 2011 to more clearly define contractor business systems, including accounting, estimating, and purchasing, and to allow payments to be withheld from contractors if their business systems contain significant deficiencies. DOD has taken steps to identify unspent contract funds and recover improper payments, but limited visibility into its contracts has hindered such efforts. For example, DOD has deobligated some funds to make them available to meet other DOD needs, but there remains at least $135 million that will potentially not be available for use by DOD at the end of fiscal year 2011. DOD generally cannot identify to which contracts these funds are associated. Additionally, instances of improper payments and potential fraud were sometimes found years after final deliveries were made, but contracting personnel may not be able to recover funds owed to the government. DOD prioritizes deobligating funds that may potentially be returned to Treasury at the end of each fiscal year so these funds would be available for other DOD uses. DOD contracting organizations, however, have varying degrees of visibility into the amount of funds remaining on their Iraq contracts. Contracting organizations we met with generally could not identify the total and unliquidated obligations associated with their Iraq contracts, in part because the systems used to track contracting information were not linked with systems used to track financial and payment data. Similarly, DOD resource managers, who are responsible for maintaining information on the availability of funding, tracked unspent funds at the appropriation level but did not always have such information on a contract-by-contract basis. DOD estimates that at least $135 million in contract funding could return to Treasury by the end of fiscal year 2011 if not deobligated but there may be additional funds not yet identified (see table 5). C3, AFCEE, and USACE contracting organizations generally do not track unspent funds that could be returned to Treasury on a contract-by- contract basis. As a result, resource management personnel stated they are responsible for notifying contracting personnel of these funds. Resource management personnel, however, reported that identifying the appropriate contracting personnel can be time-consuming and labor- intensive, in part because of the rapid turnover of contracting personnel, which often caused the contact information listed in the data systems to be invalid. Contracting personnel stated that once they were aware that funds may be potentially returned to Treasury, they took steps to prioritize deobligating these funds, including checking whether there were pending invoices or claims requiring payments. For example: C3 did not maintain visibility of unspent funds at the contract level, in part due to limitations in its contracting and financial management systems, but available data indicate that DOD may lose $18.6 million for its use and which will be returned to Treasury at the end of fiscal year 2011. While C3 officials noted that some contracting officers may have tracked unspent funds for contracts for which they were responsible, we found that C3’s contracting data systems did not maintain such financial data. After being delegated closeout responsibility for C3’s large, cost-type contracts, DCMA Southern Europe undertook efforts to manually track unspent funds on a contract-by-contract basis. DCMA personnel reported that $15.0 million of funds that could be returned to Treasury remained on C3’s cost-type contracts as of May 2011, but anticipated having most of these funds deobligated by the end of July 2011. Similarly, without visibility into which firm-fixed price contracts had unspent funds, Task Force personnel focused their efforts on reviewing C3 contracts awarded in fiscal year 2006 to deobligate funds but told us they do not believe they will be able to close all of these contracts before these funds are returned to Treasury. Resource managers at U.S. Army Central—which manages the funds associated with C3’s contracts—stated they believe that, as of June 2011, $3.6 million on these contracts will potentially be returned to Treasury. AFCEE contracting personnel stated that they generally do not maintain visibility into AFCEE’s unspent funds at the contract level. For AFCEE’s own contracts, contracting personnel generally deobligate funds down to 10 percent of the total obligated amount, or $100,000, whichever is less, to pay for any additional costs that may be identified during DCAA’s incurred cost audits. AFCEE contracting personnel reported that for these contracts, they do not believe any funds will be returned to Treasury at the end of fiscal year 2011. AFCEE contracting personnel stated that for the contracts awarded on behalf of other organizations, they are notified by the customers of unspent funds on an ad-hoc basis. AFCEE contracting personnel stated that they prioritize the deobligation of these funds when they are made aware of them, but do not track the total amount of funds that may be returned to Treasury. USACE contracting personnel stated that they do not maintain information on unspent funds on a contract level, but rather USACE resource managers tracked funds at the account level. For these accounts, USACE resource managers notify contracting personnel, who attempt to identify which contracts are associated with these funds and, as appropriate, take steps to deobligate these funds. USACE reported, however, that $104.9 million have not been deobligated as of March 2011. USACE personnel stated that a majority of these funds are on contracts awaiting DCAA audits. Conversely, ACC-RI’s LOGCAP office tracked funds that could be returned to Treasury on a contract-by-contract basis. ACC-RI contracting personnel stated that they hold weekly meetings with the contractor and resource managers to reconcile financial records and identify funds that could be deobligated. ACC-RI personnel told us that $12.3 million of funds that could be returned to Treasury have not been deobligated as of June 2011, but anticipated having most of these funds deobligated by the end of July 2011. In some instances, DOD discovered improper payments during the contract closeout process years after the contractors delivered the final good or service, but some attempts to recover overpayments were unsuccessful and, at times, late payments to contractors resulted in interest fees. According to DFAS personnel responsible for recovering overpayments made on some Iraq contracts, if contracts were closed immediately after final payments are made, overpayments could be discovered earlier, which increases the likelihood of recovering payments. For example, when the contractor is still conducting business with the government, DFAS can reduce payments on one contract to offset overpayments made on another contract. Task Force personnel noted that for a 2005 vehicle lease contract, contracting personnel in theater found the contractor was overpaid by over $41,000 on several invoices and subsequently DFAS withheld payments on several of the contractor’s other contracts to completely offset the overpayment. DFAS personnel, however, stated that the more time that has passed from when the contractor was mistakenly paid, the more difficult it becomes to recover those payments because the contractor may no longer be in business with the U.S. government or may have changed address or name. In several instances, overpayments on contracts for goods or services delivered in 2007 or earlier were not referred to DFAS until 2010 (see table 6). DFAS personnel stated that in these cases, despite numerous attempts to contact the contractor, they have yet to recover the overpayments. As of June 2011, two of the contracts have been referred to Treasury and one contract has been referred to another DFAS office for further debt collection efforts. In a few instances, Task Force personnel did not refer overpayments to DFAS because they determined the excess payments were relatively small in value or unlikely to be recovered. For example, Task Force personnel found that the U.S. government overpaid a contractor by $8,100 for trash services provided in 2006 and 2007. After unsuccessful attempts to contact the contractor, Task Force personnel closed the contract in 2009, noting that so much time had passed since the final payment that it was unreasonable to expect that the overpayment could be recovered. C3 is unable to mitigate the amount of interest payments that may be associated with late payments because the contracting and financial management systems cannot identify which contracts still require payment, especially for contracts awarded between 2003 and 2007. Task Force personnel stated that given the limitations of these systems, they must review the contract file to determine whether a contract requires additional payment. For example, while closing a $94,500 contract for vehicle lease services in Iraq, Task Force personnel discovered the contractor may not have been paid for 2 months’ worth of vehicle lease services, so the Task Force is attempting to contact personnel in theater to confirm whether services were rendered. Additionally, some contracts requiring final payments were not paid until years after the final delivery, which resulted in interest payments. DFAS personnel reported that DFAS has paid $2.8 million in interest payments on Iraq contracts as of June 2011, though it is not possible to determine the amount of interest payments associated with over-age contracts. DOD took steps to improve its payment processes in Iraq, but some challenges with timely payments remain. According to DFAS officials, in 2008, DFAS became responsible for making payments for contracts awarded in theater with obligations of $25,000 or more and in 2010 DFAS and C3 agreed to lower this threshold to $3,000. DFAS officials stated this decision was made to improve internal controls by ensuring that adequate documentation was available before payments are made in theater. DFAS officials noted, however, there were some payment delays because payment documentation requirements were not always met. One C3 official noted that these payment processing delays led to some Iraqi vendors being unwilling to do business with the U.S. government and walking off job sites. C3’s commanding general stated that when contracts are not closed out and vendors have not been paid for goods and services that they provided to the U.S. government, this contributes to negative perceptions about Americans. Finally, late contract closeouts may hinder efforts to identify and address potential fraud found on the C3 contracts because they were reported to investigators years after the potential fraudulent activities took place and the contract files were poorly maintained. As Task Force personnel reviewed and closed C3 contracts, they identified 151 contracts with potential fraudulent activities and referred these contracts to the Army’s Criminal Investigations Division. For example, in one contract for a cable fiber network, Task Force personnel stated that they found evidence that the contracting officer had made a payment of $84,000 in cash, but the contractor’s invoice was only for $64,000. There was no documentation in the file to account for the $20,000 difference between the disbursement and invoice, so Task Force personnel referred this case to the Army’s criminal investigators. According to an Army investigator, it was difficult to determine whether this case and other cases were due to fraudulent activity or contracting errors, in part because the contracts did not have enough documentation to build a case. Furthermore, the Army investigator stated that many of the referred contracts had been awarded many years ago so following up on these cases has been challenging, as many of the contracting personnel and contractors involved are no longer available. DOD reported that actions are underway to address the lessons learned in Iraq, including developing deployable contract management systems and explicitly requiring that contract closeout requirements be incorporated into contingency contracting planning documents. DOD officials acknowledge, however, they are likely to face similar problems with closing contracts awarded to support efforts in Afghanistan. For example, the backlog of C3’s Afghanistan contracts that need to be closed is growing steadily, but the Army’s capacity to close these contracts in the United States remains in question due to challenges with transitioning closeout responsibilities from the Task Force to ACC-RI. In October 2010, as part of the Army’s Operational Contract Support Lessons Learned Program, C3 identified lessons learned from contracting in Iraq between 2005 and 2010. As part of this effort, C3 identified the need to improve and consolidate data management, improve contract oversight, and increase emphasis on contract administration and closeout. DOD officials told us they had already implemented or planned new practices, as the following examples illustrate.  C3 officials noted that they had implemented the Standard Procurement System in both Iraq and Afghanistan to better document information on contracts awarded during and after fiscal year 2009 and have worked to improve the data input into the system. Defense Procurement and Acquisition Policy officials and a representative from the Joint Chiefs of Staff told us they are also identifying and developing deployable contract writing and management systems with the intent that one day contingency contracting personnel will use the same contract management tools in theater that are used in the United States.  C3 also identified that contract oversight was a historic problem and noted the need to ensure contracting officer’s representatives fulfilled their oversight responsibilities. In March 2010, the Under Secretary of Defense for Acquisition, Technology and Logistics issued new certification requirements for contracting officer’s representatives to ensure they are experienced and trained before they are appointed to oversee contractor performance. In June 2011, we reported, however, that DOD personnel in Afghanistan were not always fully prepared for their roles and responsibilities to provide adequate oversight there.  Defense Procurement and Acquisition Policy has also issued and since updated the Defense Contingency Contracting Handbook, which includes reference material to ensure contingency contracting officers maintain proper contract documentation and complete closeout duties. For example, the handbook includes guidance on the essential documents that should be in a contract file, identifies steps to ensure contracts are properly enumerated to avoid duplicate contract numbers, and recognizes the need to close contracts as soon as possible. Finally, DOD is in the process of determining how it will address the problems C3 attributed to a lack of planning for the contracting requirements in Iraq. A senior C3 official recommended that operational campaign plans include a contracting annex, such as an Annex W. In such cases when an Annex W would be required, we found that Joint Publication 4-10 and DOD’s Annex W guidance do not fully address the need to plan for contract closeout requirements—including identifying responsibilities, either in or outside of theater, for closing contracts. United States Forces-Iraq issued an Annex W in 2011, which included directions for personnel to take steps to close contracts in Iraq, well after C3’s backlog of contracts was identified. Representatives from the Joint Chiefs of Staff responsible for revising Joint Publication 4-10 and the Annex W guidance recognize the need to incorporate more specific language on the need to plan for contract closeout during the contingency contracting planning process. These officials stated that they plan to issue new Annex W guidance by the end of 2011 and intend to add more specific language regarding contract closeout. As was the case in Iraq, C3 officials stated that prior to the build-up of forces in Afghanistan, contract closeout was a challenge because there were not enough contracting personnel in theater to meet competing contracting demands. To address its backlog of contracts awarded before fiscal year 2009, C3 delegated responsibility for closing at least 22,597 Afghanistan inactive contracts to the Task Force. Task Force data indicate that 3,510, or about 16 percent, of these contracts have been reviewed as of April 2011. Task Force personnel stated that they faced the same challenges with closing the Afghanistan contracts as those associated with the Iraq contracts, such as poor contract documentation and improper payments. According to C3’s commanding general and senior contracting officials, these challenges were exacerbated during the build-up of U.S. military personnel in Afghanistan, and the focus remains on meeting the warfighter’s needs. C3 officials told us the number of contracting officers in Afghanistan increased from about 60 in 2008 to about 200 in April 2011. In part, this increase in personnel enabled C3 to close over 18,600 contracts awarded between fiscal years 2009 and 2011. Despite these efforts, however, the number of contracts eligible to be closed continues to grow. For example, as of April 2008, C3 data indicated that 1,471 Afghanistan contracts remained in theater that were eligible but over age for closeout. As of May 2011, the number of contracts eligible but over age for closeout has increased to over 16,900 contracts. Additionally, C3 will have to close over 7,000 other contracts awarded during this period that are eligible but not yet over age for closeout. C3 officials told us they expect that more Afghanistan contracts will be transferred out of theater to be closed by ACC-RI, likely after much of the remaining Iraq contracts are closed. As previously noted, however, the Army’s ability to close contracts remains in question due to challenges with transitioning closeout responsibilities to ACC-RI. Contract closeout is a key step to ensure the government receives the goods and services it purchases at the agreed upon price and, if done in a timely manner, provides opportunities to utilize unspent funds for other DOD needs. In Iraq, however, contract closeout was often an afterthought or was done as time permitted. The complications DOD has faced with closing its Iraq contracts underscore the importance of advanced planning to close contracts awarded in a contingency environment, encouraging a greater command emphasis on completing and overseeing administrative requirements, establishing a process to provide better management visibility and insight into contracting efforts, and ensuring that DOD’s contracting workforce has the capacity to provide appropriate contract administration and contractor oversight. Meeting warfighter needs is paramount, but doing so does not lessen the need to ensure that contracts are properly administered and executed. DOD’s recognition in 2008 that it needed to address the backlog of contracts that are over age for closeout and its establishment of the Task Force came too late in the operation to make a significant difference in closing contracts within the required time frames. By not fully understanding the scope of the backlog and waiting to address it, DOD underestimated the efforts required to close these contracts. Further, the limited visibility provided by the contracting and financial management systems hindered DOD’s ability to identify and address improper payments. Challenges with transitioning closeout responsibilities to ACC- RI appear to have hindered the progress the Army had made in closing its Iraq contracts. With over 100,000 C3 Iraq and Afghanistan contracts that need to be reviewed and closed, as appropriate, further delays in closing these contracts can be expected. Finally, closing the large cost-type contracts is further hindered by DCAA’s shortage of auditors and problems with contractor accounting practices. DOD has recognized the need to increase DCAA’s staffing and address contractor business systems, but fully implementing these initiatives will take several years. To help address the current backlog of contracts supporting the efforts in Iraq and Afghanistan that need to be closed out, we recommend that the Secretary of Defense direct the Secretary of the Army to take steps to ensure ACC-RI’s planned resources are adequate to meet forecasted closeout demands. To help improve DOD’s ability to manage the closeout of contracts awarded in support of future contingencies, we recommend that the Secretary of Defense, in coordination with the Chairman of the Joint Chiefs of Staff, take the following two actions: revise DOD’s contingency contracting doctrine and guidance to reflect the need for advanced planning for contract closeout; and require senior contracting officials to monitor and assess the progress of contract closeout activities throughout the contingency operation so steps may be taken if a backlog emerges. DOD provided written comments on a draft of this report. DOD concurred with the three recommendations and identified a number of ongoing and planned actions to address them. For example, DOD noted that Army Contracting Command-Rock Island will utilize contractors and explore additional options, such as the Wounded Warrior program, to assist in closing contracts. DOD also noted that it recently amended the Defense Federal Acquisition Regulation Supplement and provided additional guidance to DOD personnel to underscore the need to understand the unique requirements and considerations associated with planning and executing contingency contract administration services in contingency operations. DOD also plans to further revise its guidance to address the need for contracting officers to do advance planning for closeout of contracts performed in contingency areas. DOD also indicated it intends to issue a revised Joint Publication 4-10, its contingency contracting planning doctrine, in June 2012 to reflect the need for such planning. DOD also provided technical comments, which were incorporated as appropriate. DOD’s comments are reprinted in appendix II. We are sending copies of this report to the Secretary of Defense, the Secretaries of the Army and Air Force; the Under Secretary of Defense (Acquisition, Technology, and Logistics); the Director, Defense Procurement and Acquisition Policy; the Under Secretary of Defense (Comptroller) and Chief Financial Officer; the Chairman, Joint Chiefs of Staff; the Commander, U.S. Central Command; the Director, Defense Contract Audit Agency; the Director, Defense Finance and Accounting Service; and interested congressional committees. In addition, the report will be made available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-4841. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix III. To assess the Department of Defense’s (DOD) efforts to close its Iraq contracts, under the authority of the Comptroller General to conduct evaluations on his own initiative, we examined the (1) total number of its contracts with performance in Iraq that are eligible for closeout and the extent to which DOD closed these contracts within required time frames, (2) factors that contributed to contracts not being closed within required time frames, (3) steps DOD took to manage the financial risks associated with not closing contracts within required time frames, and (4) how DOD captured and implemented lessons learned from closing its Iraq contracts. To determine the number and value of DOD’s Iraq contracts eligible for closeout and the extent to which DOD will close these contracts within required time frames, we reviewed the Federal Acquisition Regulation (FAR) and the Defense Federal Acquisition Regulation Supplement which provide the time frames and the procedures for closing contracts. For the purpose of our review the term contracts refers to all base contracts, task orders, and blanket purchase agreement call orders. We obtained contract data from four DOD organizations which our prior work indicated had been responsible for awarding the majority of contracts with performance in Iraq: CENTCOM Contracting Command (C3), Army Contracting Command-Rock Island (ACC-RI), US Army Corps of Engineers (USACE), and Air Force Center for Engineering and the Environment. These organizations may retain responsibility for administering and closing the contracts they awarded, or may they may delegate such responsibilities to another organization. In those instances, we obtained contract data from that organization, which includes Defense Contract Management Agency, ACC-RI, and C3’s Contract Closeout Task Force Office (Task Force). From each organization, we requested the following data for contracts for which they are responsible: contract and order numbers, period of performance, contract type, contract status, total obligations, total unliquidated obligations, and physical completion dates. We identified contracts that were eligible for closeout and over age for closeout based on the time frames established in the FAR. We also identified contracts that did not have complete data to determine eligibility for closeout, but we determined these contracts to be eligible and over age according to data available. We assessed the reliability of these data reported by the contracting organizations through interviews with knowledgeable officials and electronic data testing for missing data, outliers, and obvious errors within each database. While we found that C3’s contract data from fiscal years 2003 through 2008 were generally unreliable for determining the closeout status of contracts, they were sufficiently reliable for determining the minimum number of contracts awarded during this time period. We did not evaluate or assess the reliability of the financial management systems used to provide financial data for the purpose of our review. We also did not independently evaluate whether DOD closed individual contracts in accordance with the procedures outlined in the FAR or other DOD guidance. To identify the factors that contributed to contracts not being closed within FAR-required time frames, we analyzed data provided by and interviewed officials at each of the contracting organizations and the Defense Finance and Accounting Service (DFAS), which is responsible for making payments on some of the Iraq contracts. To understand any challenges faced by DOD contracting personnel in closing individual contracts, we reviewed contract documents for 25 firm-fixed price contracts purposefully selected to obtain a variety of closeout organizations and a range of closeout difficulty and interviewed contracting personnel on their experiences with closing them. We also reviewed Task Force and ACC- RI closeout data to assess the Army’s ability to close C3’s contracts. In addition, to identify the factors that affected the closeout of cost-type contracts, we interviewed personnel at each of the contracting organizations. In addition, we purposefully selected eight contractors with varying amounts of over-age cost-type contracts, obligations on contracts, and remaining unliquidated obligations and reviewed DCAA’s incurred cost and other audit reports for these contracts, and interviewed DCAA officials at headquarters and eight field offices to determine the factors affected their ability to complete the audits. We also reviewed Joint Publication 4-10; the Defense Contingency Contracting Handbook; and the Defense Contract Management Agency’s contract closeout guidance and handbook to assess the guidance provided to DOD contracting personnel regarding the need to plan the contract closeout process. To determine the steps DOD has taken to manage the financial risks associated with not closing contracts within FAR time frames, we reviewed the DOD Financial Management Regulation and each contracting office’s closeout guidance. We also interviewed contracting and financial management personnel at the Office of the Under Secretary of Defense, Comptroller; Office of the Assistant Secretary of the Army, Financial Management & Comptroller; U.S. Forces – Iraq, Force Structure Resources and Assessment (J-8); U.S. Army Central Command; and USACE. In addition, we analyzed unliquidated obligation data provided by both the contracting personnel and financial management personnel to determine how these funds were managed. To determine the steps DOD has taken to manage other risks of not closing contracts timely, we reviewed data and interviewed officials from C3; the Task Force; DFAS, which is responsible for collecting overpayments and tracking interest payments; and the Army’s Criminal Investigations Division, which is responsible for investigating instances of fraudulent activity found in contracts. To assess the extent to which DOD captured and implemented lessons learned from closing contracts in contingency operations, we interviewed contracting officials at each of the organizations we visited to identify any lessons learned and reviewed documentation when available. We also interviewed senior contracting officials in Iraq and Afghanistan to identify any changes made in response to the lessons learned from closing the C3 contracts. We reviewed DOD’s current contingency contracting doctrine and guidance, and interviewed officials from the Joint Chiefs of Staff who are responsible for revising the doctrine and guidance. We also interviewed officials from the Office of Under Secretary of Defense for Acquisition, Technology, and Logistics’ Office of Defense Procurement and Acquisition Policy and the Office of the Deputy Assistant Secretary for the Army (Procurement) to identify any policy changes that may result from the lessons learned in Iraq. We obtained and reviewed C3 data on the total number of its Afghanistan contracts eligible and over age for closeout to assess its ability to close these contracts. We conducted this performance audit from July 2010 through September 2011 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the individual named above key contributors to this report were Timothy DiNapoli, Assistant Director; Johana Ayers; Noah Bleicher; Seth Carlson; Morgan Delaney-Ramaker; Justin Jaynes; Julia Kennon; John Krump; Claire Li; Anne McDonough-Hughes; and Roxanna Sun.

Since 2002, DOD obligated at least $166.6 billion on contracts supporting reconstruction and stabilization efforts in Iraq and Afghanistan. Many of these contingency contracts, in particular those awarded in Iraq, need to be closed. Contract closeout is a key step to ensure the government receives the goods and services it purchased at the agreed upon price and, if done timely, provides opportunities to use unspent funds for other needs and reduces exposure to other financial risks. To assess DOD's efforts to close its Iraq contracts, GAO examined the (1) number of contracts that are eligible for closeout and the extent to which they will be closed within required time frames, (2) factors contributing to contracts not being closed within required time frames, (3) steps DOD took to manage the financial risks associated with not closing contracts within required time frames, and (4) extent to which DOD captured and implemented lessons learned from closing its Iraq contracts. GAO reviewed contingency contracting guidance, analyzed contract and closeout data for contracts awarded between fiscal years 2003 and 2010, and interviewed DOD officials from six organizations responsible for awarding or closing out these contracts. DOD does not have visibility into the number of its Iraq contracts eligible for closeout, but available data indicate that DOD must still review and potentially close at least 58,000 contracts awarded between fiscal years 2003 and 2010. GAO's analysis indicates that relatively few of its contracts will be closed within required time frames. For example, about 90 percent of the limited number of contracts for which DOD could provide closeout data are already over age for closeout. The U.S. Central Command's Contracting Command (C3) and its predecessors, which awarded many of DOD's Iraq contracts, did not have sufficient internal controls to ensure that contracting data were accurate and complete. C3's management visibility was further affected by limitations of its information systems, staff turnover, and poor contract administration. DOD's ability to close its contracts has been hindered by the lack of advance planning, workforce shortfalls, and contractor accounting challenges. For example, DOD's contingency contracting doctrine and guidance do not specifically require advanced planning for contract closeouts. DOD took steps in 2008 to address its backlog of contracts needing to be closed but such actions came too late to make significant difference in closing contracts within required time frames. DOD is now transitioning responsibility for closing out C3's contracts to the Army Contracting Command. Staffing challenges, however, during this transition have hindered efforts to close these contracts. Efforts to close large, cost-type contracts have been further hindered by Defense Contract Audit Agency staffing shortages and unresolved issues with contractors' accounting practices, which have delayed audits of the contractors' incurred costs. DOD's efforts to identify unspent contract funds and improper payments--two examples of financial risks that timely closeout of contracts may help identify--are hindered by limited visibility into its Iraq contracts. DOD identified at least $135 million in unspent funds that could potentially not be available to meet other DOD needs. If not used, these funds will be returned to the U.S. Treasury at the end of fiscal year 2011. Should DOD identify a need to pay for an unanticipated cost on these contracts, it will need to use other funds that are currently available. Additionally, instances of improper payments and potential fraud were sometimes found years after final contract deliveries were made, making it harder for DOD to recover funds owed to it and increasing the risk that it may need to pay contractors interest fees on late payments. DOD has identified and addressed some of the problems related to the closeout of Iraq contracts, but the growing backlog of over 42,000 Afghanistan contracts that need to be closed suggests the underlying causes have not been resolved. DOD officials noted that the lessons learned in Iraq highlight the need to improve contract data, increase the emphasis on contract administration and closeout, and improve contingency contracting doctrine and guidance. DOD officials reported that actions are underway to correct these deficiencies in future contingencies, but fully implementing these initiatives may take several years. GAO is making three recommendations to ensure DOD has sufficient resources to close its Iraq and Afghanistan contracts and to better plan for and improve visibility of closeout efforts in future contingencies. DOD concurred with each of the recommendations.

FAA engages in three primary activities: aviation safety oversight, ATC, and airport infrastructure development (see fig. 1). The costs associated with each of these activities generally depend on the nature and usage of the specific service FAA provides. FAA safety activities include the licensing of pilots and mechanics, as well as the inspection of various aspects of the aviation system, such as aircraft and airline operations. According to FAA, the costs associated with these safety activities are primarily driven by the volume of each (e.g., the number of licenses and inspections). ATC includes a variety of complex activities that guide and control the flow of aircraft through the NAS. Generally, commercial aircraft fly under instrument flight rules (IFR) that require ATC services throughout a flight. Such flights rely on FAA staff in control towers to guide them from the terminal to the runway, and through takeoff. Once in the air and beyond the immediate vicinity of the airport, they rely on terminal radar approach control centers (TRACONs) to guide them out of the airspace in a broader area surrounding the airport. Services provided by control towers and TRACONs are referred to as terminal services. The TRACONs then pass flights off to air route traffic control centers (ARTCC), which provide en- route control until the flights near their destinations; services provided by ARTCCs are referred to as en-route services. When a flight nears its destination, control is passed back to a TRACON, and then to tower guidance, to land and proceed to an airport gate. General aviation’s (GA) use of these services varies greatly. Nearly all business jet flights file flight plans for IFR services, as do roughly half of GA piston flights. Many GA flights operate entirely under visual flight rules (VFR) and may not require any ATC services at all if they do not fly to airports that have towers. These other GA flights may require ground control, or rely on beacons or flight service stations en route. FAA states that the costs imposed by each flight are influenced by the amount and nature of the specific services it uses, and by whether the flight operates at peak periods. FAA funds airport infrastructure development through the Airport Improvement Program (AIP). AIP is a multibillion-dollar grant program that provides funding for the airports included in FAA’s National Plan of Integrated Airport Systems, which includes airports that range from the largest commercial service airports in the United States to small GA airports. Unlike safety and ATC services, AIP expenditures are not the direct result of costs imposed by users of the NAS. FAA distributes AIP funding based on congressional priorities established in authorizing and appropriation legislation. Accordingly, apart from some relatively small administrative expenses, FAA’s spending for AIP does not represent a “cost” of providing services to users. Therefore, it is not possible to establish a direct link between AIP expenditures and taxes or charges paid by system users based on their use of FAA services. The Trust Fund was established by the Airport and Airway Revenue Act of 1970 (P.L. 91-258) to help fund the development of a nationwide airport and airway system. The Trust Fund provides funding for FAA’s two capital accounts, AIP and the Facilities and Equipment account, which funds technological improvements to the ATC system. The Trust Fund also provides funding for the Research, Engineering, and Development account, which funds continued research on aviation safety, mobility, and environmental issues. In addition, the Trust Fund supports part of FAA’s operations. To fund these accounts, the Trust Fund is credited with revenues collected from system users through the following dedicated excise taxes: 7.5 percent ticket tax on domestic airline tickets $3.30 domestic passenger segment tax (excluding flights to or from rural airports) 6.25 percent tax on the price paid for transportation of domestic cargo or $0.043/gallon tax on domestic commercial aviation fuel $0.193/gallon tax on domestic GA gasoline $0.218/gallon tax on domestic GA jet fuel $14.50/person tax on international arrivals and departures, indexed to 7.5 percent tax on mileage awards (frequent flyer awards tax) $7.30 per passenger tax on flights between the continental United States and Alaska or Hawaii (or between Alaska and Hawaii), indexed to inflation Trust Fund revenues totaled $10.7 billion in fiscal year 2005. The ticket tax was the largest single source of Trust Fund revenue in fiscal year 2005, totaling about $5.2 billion, or about 48 percent of all Trust Fund receipts. The passenger ticket tax was followed by the passenger segment tax and the international departure/arrival taxes, which each totaled about $1.9 billion; fuel taxes, which totaled $870 million; the cargo/mail tax, which totaled $461 million; and interest income, which totaled $430 million. Figure 2 shows the shares received from each source in fiscal year 2005. Since the Trust Fund’s creation in 1970, revenues have, in aggregate, exceeded spending commitments, resulting in a surplus or an uncommitted balance, although expenditures from the Trust Fund exceeded revenues in 2005. The Trust Fund’s uncommitted balance, which was about $1.9 billion at the end of fiscal year 2005, depends on the revenues flowing into the fund and the appropriations made available from the fund for various spending accounts. Policy choices, structural changes in the aviation industry, and external events have affected revenues flowing into and out of the fund. For example, the uncommitted balance has been declining in recent years because Trust Fund revenues for the last 5 years have been less than FAA’s forecasted levels. Figure 3 shows the fluctuations in the Trust Fund’s uncommitted balance since its inception. In addition to Trust Fund revenues, in most years General Fund revenues have been used to fund FAA. The General Fund contribution has varied greatly, ranging from 0 percent to 59 percent of FAA’s budget (see fig. 4). From fiscal year 1997, the year when existing Trust Fund excise taxes were authorized, through fiscal year 2006, the General Fund contribution has averaged 20 percent of FAA’s total budget. About $2.6 billion was appropriated for fiscal year 2006 from the General Fund for FAA’s operations. This amount represents about 18 percent of FAA’s total appropriation. The National Civil Aviation Review Commission (Commission) issued a Congressional report in 1997 analyzing several issues, including alternative funding means to meet the needs of the nation’s aviation system. The Commission’s report identified a number of concerns with FAA’s funding structure as it existed at the time the Commission began its work. To address these concerns, the Commission made several unanimous recommendations, including that FAA’s revenues be more closely linked to the costs of services provided to support ATC activities, including capital investments. The Commission also recommended that General Fund revenues be used to fund aviation security and safety activities and government use of the air traffic system, and that GA operators continue to pay a fuel tax, although perhaps at a higher rate. Some stakeholders support the current excise tax system, stating that it has been successful in funding FAA, has low administrative costs, and distributes the tax burden in a reasonable manner. Other stakeholders, including FAA, state that under the current system, the disconnect between the revenues contributed by users and the costs they impose on the NAS raises revenue adequacy, equity, and efficiency concerns. Trends in, and FAA’s projections of, both inflation-adjusted fares and average plane size suggest that the revenue collected under the current funding system has fallen and will continue to fall relative to FAA’s workload and costs, supporting revenue adequacy concerns. Comparisons of revenue contributed and costs imposed by different flights provide support for equity and efficiency concerns. However, the extent to which revenue and costs are linked depends critically on how the costs of FAA’s services are assigned to NAS users. Thus, assessing the extent to which the current approach or any other approach aligns costs with revenues would require completing an analysis of costs, using either a cost accounting system or cost finding techniques to distribute costs to the various NAS users. FAA stated that it has made substantial progress in designing a cost accounting system, implementing it throughout its lines of business, and modifying it to determine costs by user group. Some stakeholders believe that maintaining the current funding structure for FAA is appropriate because it has been successful in funding FAA for many years, suggesting that there is no urgent reason to change it. According to these stakeholders, the revenues collected from users under the current funding system, along with General Fund revenues provided by Congress, have been sufficient for the United States to develop a safe and efficient aviation system. As the number of air travelers grew, so did revenues going into the Trust Fund. Even though revenues fell during the early years of this decade as the demand for air travel fell, they began to rise again in fiscal year 2004 (see fig. 5); FAA estimates that revenues will continue to increase. In addition, these stakeholders state that administrative costs of the current system are relatively low. Another argument put forward by some industry stakeholders and analysts for maintaining the current funding structure is that this structure provides a reasonable allocation of the funding burden between commercial aviation and GA. With the current funding structure, system users who are subject to the commercial taxes—including commercial airlines, air taxis, and many fractional ownership operations—contribute about 97 percent of the tax revenue that accrues to the Trust Fund. The remaining GA operators, including those who operate purely private corporate and individual aircraft, contribute about 3 percent. Representatives of the GA segment of the industry contend that collecting the bulk of the user- contributed revenues from the commercial segment is appropriate because the ATC system exists at its current size to accommodate the demands of commercial aviation and GA users should not be asked to contribute more than the incremental costs that result from also providing services to GA aircraft. Although the incremental costs are not precisely known, GA representatives have told us that they believe that the revenues currently collected from fuel taxes are a rough approximation of the incremental costs that FAA incurs from providing services to GA aircraft. According to FAA, all cost studies to date concluded that GA users pay less than the costs they impose on the system, while commercial aviation users pay more than the costs they impose on the system. The disconnect between sources of Trust Fund revenues and FAA costs under the current funding system raises concerns that the current system will not produce adequate revenue in the future to keep pace with FAA’s workload increases and, consequently, FAA’s costs. The principle of revenue adequacy requires a funding system to produce revenues commensurate with workload changes over time. However, under FAA’s current funding system, increases in FAA’s workload will not necessarily be accompanied by revenue increases because users are not directly charged for the costs they impose on FAA from their use of the NAS. Rather, Trust Fund revenues are primarily dependent on the prices of tickets (the domestic ticket tax) and the number of passengers on a plane (the domestic ticket tax, the domestic passenger segment tax, and the international passenger tax); neither of these factors are directly related to workload, which is driven by flight control and safety activities. Long-term industry trends and FAA forecasts of declines in air fares and the growing use of smaller aircraft support revenue adequacy concerns. To illustrate the disconnect between revenues and costs, table 1 provides an example of revenues generated by different aircraft making similar flights. The use of multiple flights by smaller aircraft to carry the same number of travelers as one larger aircraft increases FAA’s workload, but will not necessarily be accompanied by increased revenues from system users to fund FAA’s additional costs associated with the workload increase. This example shows the taxes that would be generated from transporting 105 passengers from Los Angeles to San Francisco by (1) one flight using a common narrow-body jet (Boeing 737), and (2) three flights using a common regional jet (CRJ-200). In this case, the narrow-body jet has the capacity to carry 132 passengers, while each regional jet has the capacity to carry 48 passengers. As the table shows, differences in FAA’s workload are not reflected in revenues. FAA states, all other factors being equal (e.g., time of flight), that the total ATC costs of the three regional jet flights would be about three times the cost of one narrow-body flight. Revenues from the three regional jet flights, however, total only about $37, or 3 percent, more than the revenue generated by the one narrow-body jet flight. Revenue increases are not linked to cost increases because, under the current system, revenues are primarily influenced by the number of passengers, the average price of tickets, and the amount of fuel used—not the costs imposed on FAA through the use of its services. The disconnect between revenues and workload can work both ways; increases in the number of passengers on planes (e.g., larger planes or higher load factors) or increases in fares can result in higher revenues relative to workload. In fact, load factors have increased over the past several years, and fares have increased over the past year. However, long- term trends and FAA’s projections for both domestic fares and plane size suggest that Trust Fund revenues have declined relative to FAA’s workload and will likely continue to do so for the next several years. Trends in average fares suggest that the Trust Fund is collecting less revenue relative to workload than in the past, and FAA’s projections suggest that this decline will continue. Since the passenger ticket tax is a percentage of the ticket price, reductions in the average ticket price result in lower ticket tax revenues relative to FAA’s workload. Domestic airfares, adjusted for inflation, have steadily declined over the past 25 years, from an average of $233 in 1981 to $148 in 2005 (see fig. 6). This reduction represents an average decline of about 1.9 percent per year. Even though there have been increases in fares over the past year, FAA projects average fares will continue to decline over time. In FAA’s most recent forecast, inflation-adjusted domestic yields—a proxy measure for fares— are projected to decline approximately 8.5 percent over the next 10 years. Trends in the average size of airplanes also suggest that the Trust Fund is collecting less revenue relative to workload than in the past, and FAA’s projections suggest that this decline will continue (see fig. 7). Since smaller planes carry fewer passengers and burn less fuel, reductions in average plane size mean lower ticket tax, segment tax, and fuel tax revenue accrues to the Trust Fund relative to FAA’s workload. This decline in the average number of seats per aircraft is the result of airlines’ moving toward a substantially greater reliance on regional and narrow-body jets. Scheduled capacity (available seat miles) increased 29 percent from 1996 through 2005. During this time, wide-body jet capacity fell 42 percent, narrow-body jet capacity grew 35 percent, and regional jet capacity grew over 2900 percent. As a result, regional jets accounted for nearly 10 percent of scheduled capacity in 2005, up from less than 1 percent in 1995. In addition to projecting growth in commercial flights, FAA is projecting substantial growth in GA traffic, which will also add to FAA’s workload. Some aviation stakeholders have expressed concerns that the current approach to collecting funds from users through excise taxes creates inequities because the revenue contributions of different flights are not directly linked to the costs of the services that these flights receive from FAA. As noted, factors that influence the revenue contribution that a commercial flight makes to the Trust Fund are the number of passengers, the average price of tickets, and the amount of fuel used. None of these factors, however, are directly related to the cost of the ATC services that a flight receives from FAA. Table 2 shows FAA’s estimates of the revenue contributions made by various flights. Since FAA estimates that similar flights impose similar costs on the agency, the substantial differences in the revenue contributions of these flights raise issues of fairness. One equity issue is that similar commercial flights may contribute very different amounts of revenue. As shown in this example, a 767 flight contributes more than twice as much as two similar 737 flights. There is also a difference between the contributions for the two similar 737 flights; one flight contributes 14 percent more than the other flight. Concerns also exist about the fairness of the distribution of the funding burden between commercial airlines and GA operators. Domestic commercial passenger flights, and some flights typically considered GA flights that carry commercial passengers, are subject to, among other potential excise taxes, the passenger ticket tax, the passenger segment tax, the cargo/mail tax, and the fuel tax. GA flights (excluding those that carry commercial passengers) are subject only to a fuel tax. As a result, the revenue contributions of similar commercial and GA flights may be substantially different. For example, the taxes that the Trust Fund would receive from two different types of business jet flights would be substantially less than the taxes received from similar commercial flights (see table 2). Although the commercial and GA flights might receive the same services from FAA, raising equity concerns because of the large difference in revenue contribution, there is debate over whether GA and commercial flights should be assigned the same costs for similar flights because parties disagree on how to assign the fixed costs associated with the ATC system. Representatives of the commercial aviation industry favor assigning those costs to all system users in proportion to their use of the system. Representatives of GA, on the other hand, state that the system exists at its present size to serve the needs of the commercial aviation industry and that GA should be assigned only the incremental costs of serving GA (i.e., those costs that would not otherwise exist). Without a consensus on how to assign ATC costs to users, it is not possible to assess the extent to which the current approach or any other results in a distribution of the funding burden between commercial airlines and GA operators that approximates the distribution of costs attributable to those groups. Some stakeholders have also raised concerns that the current funding system does not provide aircraft operators with incentives to use FAA services in the most efficient manner. For users to make efficient decisions about their use of the NAS, their price for using the system (the taxes or charges they pay) should accurately reflect the costs their use imposes on the system. These prices, along with other factors influencing supply and demand, will influence users’ decisions about the type, size, and number of aircraft to operate, and when and where to operate them. Given the importance of some of these other factors to users’ decisions about using the NAS, the influence of prices charged for FAA’s services on these decisions may be comparatively small for some users. As discussed previously, FAA states that under the current funding system the taxes collected from users do not accurately reflect the costs those users impose on the system; some flights likely pay more than the costs they impose, while others likely pay less. These price differences suggest that the current funding structure creates incentives for inefficient use of the NAS. Users who pay more in taxes than the costs they impose may make less than optimal use of the system, while those who pay less than the costs they impose may make more than optimal use of the system. An airline’s decision about how many flights to offer in a given market illustrates how the current system does not provide incentives for efficient use of the system. In this example (the same one used for the revenue adequacy discussion), an airline is deciding how many daily flights it should provide for the Los Angeles to San Francisco market (see table 3). It estimates that the market demand at the fare it is charging totals 105 passengers per day, and it faces the choice of providing the market with one daily flight with a narrow-body jet (Boeing 737), or three daily flights with a regional jet (CRJ-200)—all flight choices are assumed to depart during peak periods. In this scenario, the revenue collected from the three regional jet flights—$1,215—is about 3 percent more than the revenue collected from the one narrow-body jet flight—$1,178. FAA states however, that each flight would impose similar costs on the agency, so FAA’s costs would be roughly 3 times more to handle the three regional jet flights than to handle the one medium jet flight. In this example, however, there is little financial incentive ($37) for the airline to limit its imposition of additional costs on FAA by using one flight instead of three flights. This situation is made worse during times when the NAS is congested. There are two issues associated with congestion. The first is plane size; if all other factors are equal, such as demand for air travel, it is more efficient to serve congested airspace with larger planes because they can move more passengers per flight. Second, when congestion is a factor, efficiency requires consideration of the delay costs imposed on other system users. Charging similar flights equally, regardless of plane size, and incorporating congestion costs, would create financial incentives to improve efficiency. Alternative funding options for collecting revenues from NAS users present both advantages and disadvantages. The degree to which alternative funding options could address concerns about the current excise system ultimately depends on the extent to which the contributions required from users actually reflect the costs they impose on the system. Given the diverse nature of FAA’s activities, a combination of alternative options may offer the most promise for linking revenues and costs. Switching to any alternative funding option would raise administrative and transition issues. For example, any cost-based funding system would require FAA to complete the appropriate cost analysis using either a cost accounting system or cost finding techniques. Some stakeholders who support the adoption of direct user charges also support a change in FAA’s governance structure—for example, commercializing air navigation services—but we found no evidence that the adoption of direct charges would require a governance change. The six funding options considered here include two that would modify the current excise tax structure and four that would adopt more direct charges to users. Without more detailed information and an understanding of the costs different flights impose on the NAS, any assessment of the current system or alternative funding options is only preliminary. The degree to which alternative funding options could address revenue adequacy, equity, and efficiency concerns, relative to the current system, ultimately depends on the extent to which the contributions required from users actually reflect the costs they impose on the system. More precise assessments of the current or alternative funding options are possible only if cost finding techniques are used throughout FAA. The two options we reviewed that would modify the current excise tax structure are relying solely on a fuel tax and increasing the passenger segment tax to replace the passenger ticket tax. One possible modification to the current system would be to increase the current aviation fuel taxes—which levy a specific amount per gallon of fuel—to replace the revenue lost by eliminating the remaining excise taxes and charges. Advocates of reliance on a fuel tax funding system state that it is appealing compared to the current system because there is a correlation between the time a plane spends in the system and the amount of fuel a plane uses. To the extent that time in the system is related to cost, this relationship creates at least a partial link between revenues and costs, which could partially address the revenue adequacy, equity, and efficiency concerns about the current system. In addition, advocates of the fuel tax state that a fuel tax is inexpensive and simple to administer. Under the current system the Internal Revenue Service (IRS) is responsible for collecting fuel taxes at the point of sale, and these funds are then deposited to the Treasury, which then credits the Trust Fund. FAA has no responsibility for collecting the revenue. Thus, transitioning to an all-fuel- tax funding system would be relatively easy, since the administrative system is already in place. Furthermore, the tax is easy for consumers to understand, and compliance is simple and inexpensive. From a revenue adequacy perspective, fuel taxes compare favorably with other existing excise taxes because they are more directly linked to workload. Thus, all things being equal, increases in workload over time would likely result in fuel tax revenue increases. Nonetheless, two factors that lead to lower fuel consumption will erode the ability of a fuel tax to generate revenue over time. First, while the incentive created through the tax to conserve fuel will promote more efficient use of the system, it will lead to lower fuel consumption, which will reduce revenues. Second, technological advances that increase the fuel efficiency of airplanes will reduce fuel consumption relative to FAA’s workload, leading to lower revenues relative to FAA’s workload; the new 787 aircraft and a recent effort to outfit planes with winglets are examples of these advances. Thus, it is likely that the fuel tax rate would have to be raised from time to time to be adequate in the long run. The extent to which a fuel tax would address equity issues appears to be limited. Although FAA states that there is a correlation between the time a plane spends in the NAS and fuel consumption, the extent to which fuel consumption correlates with costs imposed on FAA has not been established. First, there may be a relationship between time in the system and en-route control costs, but the relationship between time in the system and the costs of other FAA activities, such as terminal costs, is not obvious. Second, even if the fuel tax were limited to funding en-route costs, the connection between fuel consumption and those costs appears to be incomplete. For example, since heavier planes burn more fuel per mile than lighter planes, they would be required to contribute more for spending the same amount of time in the system. As with equity issues, the potential for a fuel tax to address efficiency issues appears limited because the connection between revenues and costs is incomplete. A fuel tax can create an incentive for operators to minimize their fuel consumption (e.g., by flying at off-peak times to avoid congestion delays) and, therefore, their time in the NAS. To the extent that time in the system correlates with costs imposed, this incentive can lead to improved efficiency. However, any relationship between time in the system and costs imposed on FAA appears to be limited to en-route control costs. A second option that represents a modification of the current system is to increase the current passenger segment tax to replace revenues lost by eliminating the current passenger ticket tax. Under this option, all other current excise taxes would remain unchanged, implying no change to revenues collected from cargo carriers and GA operators. This option would likely increase the tax differential between passengers traveling on one-stop (or more than one-stop) flights and those traveling on nonstop flights on the same route. As a result, there might be a shift in travelers’ demand toward more nonstop service, which might, in turn, lead airlines to operate more nonstop service. Because there is a partial link between the number of segments an airline operates and the cost of the services FAA provides to that carrier, this option might have some advantages over the present tax structure in terms of revenue adequacy, efficiency and equity. However, because there is no link to the cost of some of the other services that FAA provides, these advantages are limited. Compared to the present funding structure, this option might address concerns about revenue adequacy over time, but many of the concerns associated with the current system would likely remain. One way in which a passenger segment tax might better correlate to FAA’s workload is that commercial flights that include a stop require more terminal services from FAA than nonstop flights, and taxes based on the number of passenger segments traveled will increase as the number of stops increases. In addition, the current passenger segment tax is indexed to the Consumer Price Index so that it is adjusted each year to account for inflation, which preserves the purchasing power of the revenues collected. However, other services that FAA provides could increase without any increase in passenger segment tax revenues. For example, if the average distance of commercial flights increases, the cost of providing en-route services will rise, but the passenger segment taxes paid will not rise because they are not based on distance traveled or time in controlled airspace. Furthermore, passenger segment taxes apply only to commercial flights, so they have no advantage over ticket taxes in providing revenue adequate to fund cost increases associated with providing services to cargo and GA aircraft. In addition, there would be no improvement in providing adequate revenue for safety and security expenditures. Compared to ticket taxes, higher flight passenger segment taxes have the potential to increase equity by better aligning revenues with costs, and they create some additional incentives for efficient use of FAA services. However, these effects are likely to be limited because the tax revenues are aligned only to some cost elements and the tax applies only to commercial aircraft. With increased passenger segment taxes, the difference in the amount of taxes commercial airlines would have to pay for one-stop service compared with nonstop service would be greater. This greater difference in taxes might represent an improvement in equity compared to the present funding system because one-stop flights require more terminal and approach services from FAA than nonstop flights. This greater difference in taxes could also create an incentive to provide more nonstop service. Substituting nonstop for one-stop service could reduce the airlines’ need for FAA’s terminal and approach services. However, this incentive could be quite small relative to other factors that influence airlines’ service-offering decisions, so the effect on efficiency could also be quite small. In addition, airlines would have no additional incentive to be efficient in their use of en-route services because the passenger segment tax is not linked to time in controlled airspace, and there would be no change from the current structure in incentives for cargo and GA operators. Administrative and transition issues would be minimal, since this option would require only a change in the current tax per flight segment and the elimination of the passenger ticket tax. The four funding options we reviewed that would involve more direct charges to users include weight/distance charges, en-route charges, flight segment charges, and certification charges. Charges based on weight and distance traveled are used by a number of foreign air navigation service providers and are supported by the International Civil Aviation Organization. As suggested by the name, this option would base charges to users on the weight of the plane and the distance it travels within the NAS. According to their advocates, weight/distance charges are more appealing than the current system because they would establish a more direct relationship between revenues and costs by incorporating distance into the formula, thereby creating an incentive to limit excess use of FAA’s ATC en-route services. In addition, advocates say, weight/distance charges would strike a balance between basing charges on the ability-to-pay principle and more directly linking costs and revenues by incorporating both weight and distance in the distribution of costs among users. A weight/distance charge, relative to the current funding system, would be likely to improve the revenue adequacy of the system. Revenue adequacy is addressed by the incorporation of a cost component into the weight/distance formula. Generally, air navigation service providers that use a weight/distance formula regularly adjust the cost component to ensure that revenues match costs. For example, FAA’s counterpart in France—la Direction Générale de l’Aviation Civile—annually adjusts the cost component of its weight/distance formula on the basis of en-route charges. This adjustment ensures that revenues not only cover costs, but also do not exceed costs. As with the fuel tax, the extent to which a weight/distance charge would address equity issues appears to be limited. While there may be a relationship between the distance a plane travels in the NAS and the costs it imposes, the introduction of the weight component into the formula weakens any such connection. For example, since heavier planes would be charged more than lighter planes, they would be required to contribute more for traveling the same distance in the system, even though they may not impose greater costs on the ATC system. If a relationship between weight and distance in the system and costs imposed can be established, it is likely to be limited to en-route control costs. There is no obvious relationship between the weight/distance formula and other FAA activities—terminal control services and safety activities. Since the connection between revenues and costs is incomplete because of the weight component, the potential for a weight/distance charge to address efficiency issues also appears limited. The distance component of a weight/distance charge creates an incentive for operators to minimize their use of the NAS. To the extent that distance in the system correlates with costs imposed, this incentive could improve efficiency. However, the correlation between distance and costs imposed is limited by the introduction of the weight component. Furthermore, the relationship between distance in the system and the costs imposed on FAA is likely to be limited to en-route control costs, excluding consideration of the costs associated with terminal control and safety activities. Implementing a weight/distance charge would also involve significant administrative and transition issues. FAA would have to determine how to administer a weight/distance charging system for which it does not currently have the organizational capacity. FAA stated that one option would be to contract the billing out to a private party, much as European Union countries such as France contract out their billing to Eurocontrol. En-route charges would be based on the time users spend in the NAS or the distance they travel through the NAS. According to their advocates, en- route charges are more appealing than the current system because they would create a more direct relationship between revenues and costs. Therefore, compared to the current system, advocates say en-route charges would (1) better ensure that revenues are adequate to cover costs over time, (2) address equity issues, and (3) create incentives for efficient use of the current system. An en-route charge, relative to the current funding system, would be likely to improve the revenue adequacy of the system. As with weight/distance charges, en-route charges could address revenue adequacy concerns by incorporating a cost component into the charging formula that could be regularly adjusted to reflect any changes in costs. This approach could ensure, over time, that revenues match costs. As with other funding options discussed here, the ability of en-route charges to address equity and efficiency issues raised by the current system appears to be limited. According to FAA, there is a strong relationship between time and distance in the system and en-route costs imposed by users. Thus, if en-route charges were limited to funding en- route control costs, they might address equity issues raised by the current system by equating charges to costs imposed, depending on how costs are assigned. Furthermore, en-route charges for en-route control would create clear financial incentives to use the system more efficiently; less use of the system would lead to proportionately lower charges. However, there is no obvious relationship between time or distance in the system and other FAA activities—terminal control services and safety activities. As a result, if en-route charges were used to fund all FAA activities, their ability to address equity and efficiency issues is unclear. Implementing en-route charges would also involve significant administrative and transition issues. FAA would have to develop the organizational capacity to administer and collect en-route charges, which would include completing the appropriate cost analysis using either a cost accounting system or cost finding techniques. Flight segment charges to users would be based on the departures and landings that aircraft make at various airports throughout the NAS. According to their advocates, flight segment charges are more appealing than the current system because they would establish a more direct relationship between revenues and costs. Therefore, compared to the current system, advocates say that flight segment charges would (1) better ensure that revenues are adequate to cover costs over time, (2) address equity issues, and (3) create incentives for efficient use of the current system by directly connecting charges with costs imposed by users. A flight segment charge, relative to the current funding system, would be likely to improve the revenue adequacy of the system. As with weight/distance charges, flight segment charges could address revenue adequacy concerns by incorporating a cost component into the charging formula that could be adjusted regularly to reflect any changes in costs. This approach could ensure that, over time, revenues match costs. As with other funding options discussed here, the ability of flight segment charges to address equity and efficiency issues raised by the current system appears to be limited. FAA states that there is a strong relationship between departures and landings in the system and costs imposed by flights for terminal control handled by TRACONs. Thus, if flight segment charges were limited to funding terminal control costs, they might address equity issues raised by the current system by equating charges to costs imposed, depending on how costs were assigned. Furthermore, flight segment charges for terminal control would create clear financial incentives to use the system more efficiently: less use of the system would lead to proportionately lower charges. However, there is no obvious relationship between flight segments and other FAA activities—en-route control and safety activities. As a result, if flight segment charges were used to fund all FAA activities, their ability to address equity and efficiency issues would be limited. Implementing flight segment charges would involve administrative and transition issues similar to those associated with en-route charges. FAA would have to develop the organizational capacity to administer and collect flight segment charges and complete the appropriate cost analysis using either a cost accounting system or cost finding techniques. Certification charges to users would cover specific safety services provided by FAA, such as certificates for air worthiness, air operators, and air agencies; registration for air personnel, aircraft, and medical personnel; designees and delegations; and international training. According to their advocates, certification charges would be more appealing than the current system because they would establish a direct relationship between revenues and costs, which would address the revenue adequacy, equity, and efficiency concerns associated with the current system. Certification charges have the potential to fulfill revenue adequacy requirements for safety costs over time because they are directly linked to workload; charges would be assessed for each certificate issued. Thus, as workload changed over time (increasing or decreasing), so would the revenue from certification charges. In addition, any certification system would likely have the flexibility to adjust charges as costs changed. Certification charges, however, could not support all of FAA’s funding requirements, so this option would have to be used in combination with other revenue sources. According to FAA officials, there is a clear relationship between certification charges and the specific safety activities for which users would be charged. Thus, if certification charges were limited to funding the associated safety costs, they would address equity issues raised by the current system by equating charges to costs imposed; this equity improvement, however, would be limited to funding for safety activities. Furthermore, certification charges would likely create financial incentives to use the system efficiently, since charges would increase in proportion to use. FAA raises the concern that imposing certification charges for safety services would adversely affect safety because such charges would create incentives to avoid the use of safety services and, in some cases, ATC services. Our review of available data from five air navigation service providers in other countries found that since their air traffic control services were commercialized and charges were implemented, the safety of the services remained the same or improved. For example, data from New Zealand and Canada show fewer incidents of loss of separation (the distance required between planes) since commercialization. Implementing certification charges would involve administrative and transition issues similar to those associated with en-route and flight segment charges. FAA would have to develop the organizational capacity to administer and collect certification charges and complete the appropriate cost analysis using either a cost accounting system or cost finding techniques. Using a combination of workload-related taxes or charges to fund FAA might best address the revenue adequacy, equity, and efficiency concerns associated with the current funding structure, given that the costs of FAA’s ATC and safety activities are driven by different factors. No single option that we reviewed creates a direct link between revenues and all components of FAA’s activity costs. Fuel taxes, weight/distance charges, or en-route charges based on time or distance spent in the NAS could be used to create a more direct link with FAA’s costs of providing en-route ATC services. A segment tax for passengers or a flight segment charge could be used to create a more direct link with the costs of FAA’s terminal services. Certification charges could be used to create a more direct link with the costs of FAA’s various safety-related activities. Thus, some combination of options, such as en-route charges to fund en-route costs, flight segment charges to fund terminal control costs, and certification charges to fund some safety costs, might best address concerns with the current system by providing a better link between revenues and costs than any of these options used separately. According to one stakeholder, however, the administrative expense of using multiple funding options might outweigh the benefits of such an approach. According to FAA, other air navigation service providers, such as those in the European Union, have been able to administer direct charges without incurring excessive administrative costs. In discussing alternative funding options, some stakeholders have stated that if user charges are adopted, users should have more input into FAA’s operation, citing the “user pays, user says” principle. To many stakeholders, this principle implies that the adoption of direct user charges would require a change in FAA’s governance structure that could limit congressional influence on the agency while expanding the influence of airlines and other users. Many stakeholders support such a change, pointing out that many countries that rely on direct charges to fund aviation activities have commercialized their air navigation service providers. We did not find any evidence that a change in FAA’s governance structure would be required if direct charges were adopted. Federal law provides general authority for federal agencies to institute user charges except when otherwise prohibited. In FAA’s case, Congress has specifically prohibited the agency from instituting any new user charges under this general authority in every DOT appropriation act since 1998. Furthermore, under the current funding system, users already provide most of the revenue used to fund FAA programs through excise taxes. Adopting direct charges would change the manner in which revenues are collected from users, but would not necessarily change the aggregate contribution from users. Since users pay most of FAA’s program costs now, it is unclear what additional role users should play in FAA’s decision-making under an alternative system. Recent reforms in France’s Direction Générale de l’Aviation Civile illustrate how a government agency has moved toward a cost-based system of charges to fund the air navigation services it provides without changing the underlying governance structure. The French organization’s activities fall into two broad divisions —safety and regulation, and ATC. Safety and regulation are funded through a combination of general government support and specific user charges. For example, there are charges for pilots’ licenses, medical certificates, inspections, and aircraft registration. ATC activities are split into two categories—en-route control and terminal control. For en-route control, France must abide by the European Union’s regulations, which are based on principles established by the International Civil Aviation Organization. This approach incorporates a weight/distance formula that is used to determine charges for specific aircraft based on their activity. Although the formula distributes charges across aircraft differently by incorporating weight as a factor, the amount of the charges is based on cost data that are verified by the European Union. Eurocontrol actually bills users of the system; all European Union countries collect en-route charges through this organization. Terminal control charges are not directly based on cost factors, but are billed along with the en-route control charges through Eurocontrol. Allowing FAA to use debt financing for capital projects have advantages and disadvantages. Many stakeholders have identified the use of debt financing—such as bonds—as a means of funding FAA capital projects, such as components of NGATS or existing ATC facilities and equipment. Some stakeholders believe debt financing is attractive because it could provide FAA with a stable source of revenue to fund capital development and, at the same time, spread the costs out over the life of a capital project as its benefits are realized. If Congress approved the use of debt financing for FAA, the agency could borrow through the Treasury or directly from the private capital market, depending on what authority Congress provided. Debt-financing raises significant concerns, however, because it encumbers future resources and because expenditures from debt proceeds may not be subject to the congressional oversight that appropriations receive. In addition, debt financing is subject to federal budget scoring rules and raises issues associated with borrowing costs that are particularly important in light of the federal government’s long- term fiscal imbalance. According to its supporters, debt financing has a number of advantages, one of which is that it could provide FAA with a stable source of revenue to fund capital development. FAA officials state that the uncertainty associated with the appropriation process makes planning for large, complex, and expensive ATC systems difficult. Another advantage cited is that debt financing would allow the costs of capital projects to be repaid as the benefits are received, better aligning costs and benefits. Finally, supporters of debt financing, including an investment firm, state that the private capital market may offer disciplinary mechanisms that may encourage FAA to manage itself more efficiently. The discipline occurs because, to receive funding for projects, FAA would need to adhere to bond covenants, which are rules that govern how FAA will pay obligations. One investment firm noted, however, that projects could be overcapitalized, or “gold plated,” if FAA were given the authority to borrow without caps on the number and costs of projects it funds. For example, a significant amount of debt could be issued for projects with minimal marginal benefits to users. As a result, an investment firm noted, there may need to be a governing board with multiple aviation stakeholders, including airlines, airports, and air traffic controllers, to determine which capital projects are needed and how they will be funded. Treasury officials also question whether the private capital market will provide any market discipline to FAA debt obligations because investors may perceive that the obligations are backed by the federal government, and not just agency revenues. Treasury officials further noted that they could perform credit analyses similar to those done by private investment firms, which, when combined with statutory borrowing caps and other credit terms and conditions, would serve to protect the financial interests of the general taxpayer. To borrow from the Treasury, FAA would need borrowing authority from Congress. There are various ways Congress can provide borrowing authority, each with different legal, financial, and structural implications. For example, some government entities generate their own revenue to pay for borrowing costs, whereas others pay with appropriations. Some government entities with borrowing authority are federal agencies, such as the Bonneville Power Administration (BPA), while others are independent establishments, such as the U.S. Postal Service. Once borrowing authority is granted, the Treasury sets the terms and conditions for borrowing. FAA could borrow from the Treasury, using revenue options such as taxes, user fees, or appropriations to repay the debt, depending on the type of bond. Figure 8 describes the process for borrowing from the Treasury. In borrowing from the private capital market, FAA could issue general revenue (GR) or general obligation (GO) bonds. Both types of bonds would require FAA to pay interest and principal to bond holders, but the revenue sources used to make these payments would differ. A GR bond requires taxes or user fees to pay the interest and principal, while a GO bond uses expected appropriations. Several nonfederal government entities currently borrow from the private capital market using GR and GO bonds. In aviation, most commercial airports issue GR bonds for airport capital improvements that are backed by general revenues from the airport, including aircraft landing fees, concessions, and parking fees, for airport capital improvements. In surface transportation, some states issue grant anticipation revenue vehicle (GARVEE) bonds backed by anticipated federal apportionments to fund highways. However, the eligibility of a GARVEE bond for reimbursement with federal apportionments does not constitute a commitment by the federal government to provide for paying the principal or interest on the bond. The Department of Transportation, which oversees the GARVEE program, reimburses the state for debt service expenses as part of the annual federal-aid obligation authority. Figure 9 describes the process for borrowing from the private capital market. For FAA to borrow from the private capital market, Congress would need to give the agency statutory authority. Depending on how Congress writes the statute, FAA could use any revenue option—taxes, user fees, or appropriations—to secure the bond. According to some representatives of investment banks and Treasury officials, no organizational changes for FAA, such as a change to a government corporation or corporate entity, would be needed. Currently, some government corporations borrow from the private capital market, including the Tennessee Valley Authority (TVA). TVA is an independent, wholly owned federal corporation established by the Tennessee Valley Authority Act of 1933 that sells bonds in the private capital market to finance its capital improvements for power programs. TVA pays for its operations and debt service with revenues from its energy sales. Since TVA first issued bonds, Moody’s Investors Service and Standard & Poor’s have assigned TVA’s bonds their highest credit rating— Aaa/AAA. TVA does not receive a direct federal guarantee, although the interest rate charged by the private capital market suggests that there is an implied federal guarantee. Debt financing is subject to federal budget scoring rules and raises issues regarding borrowing costs that are particularly important in light of the federal government’s long term structural fiscal imbalance. How the borrowing authority is carried out will affect both budget scoring and costs. When an agency uses borrowing authority to finance a capital project, budget authority and obligations are recorded in the budget when the investments are made. Current budget scoring rules require that budget authority and obligations for the full cost of capital projects be scored upfront in the year that the obligations are made. Over time, the outlays will equal the budget authority and obligations that were scored upfront. As an example, if FAA borrowed $5 million with a 10 year bond to purchase air traffic control equipment, the $5 million would be scored as budget authority and obligations in the year or years in which FAA signed the contract or contracts to purchase the equipment, and not distributed annually over 10 years. Since this budget treatment is the same as if appropriations were obtained, there is little scoring incentive for an agency to borrow. Among the negative consequences of not scoring all government activities in the year in which obligations are made, according to CBO, is that the federal government’s obligations are understated. A Treasury official said the Treasury is supportive of budget scoring, noting that if the borrowing is for a purely governmental purpose, then that activity should be scored according to federal budget scoring rules. We have also reported that up- front budget scoring for capital projects should be maintained, since the budget should reflect the government’s commitments up front. If FAA was granted borrowing authority, the associated costs would likely be higher if the agency borrowed directly from the private capital market instead of through the Treasury. According to Treasury and representatives of investment firms, the federal government’s costs associated with debt financing for FAA’s capital projects would likely be lower if FAA borrowed through the Treasury than if FAA borrowed directly from the private capital market because the Treasury would likely be charged a lower interest rate to borrow money. Interest rates charged to FAA would likely be higher because bonds issued by FAA would likely be viewed as a greater credit risk compared to Treasury bonds because Treasury’s bonds are backed by the full faith and credit of the U.S. government, whereas FAA debt would not be. In addition, if FAA borrowed directly from the private capital market, the transaction costs of borrowing would likely be higher than if FAA borrowed through the Treasury; investment banks that serve as debt underwriters charge fees for these services, while the Treasury would charge a minimal administrative fee, if any. Treasury officials told us that it is the agency’s long-standing policy that all debt issued by federal entities, including FAA, should be issued solely to the Treasury because centralized financing of all such debt through the agency is the least expensive, most efficient means of financing this debt. The costs to the government associated with funding FAA’s capital spending through appropriations would be comparable to the costs of borrowing through the Treasury. The costs of borrowing from the private sector are based, in part, on how risky the revenue is that will be used for bond interest payments. Although all revenue options—taxes, user fees, and appropriations—can be used to repay borrowings, each option has a different risk profile. The Treasury noted that if FAA were to borrow from the private capital market against revenues that were subject to appropriations, there would most likely be a risk premium added to the credit rating to compensate for the risk that appropriations may not be provided. This risk premium would make borrowing more expensive. However, representatives from investment firms we interviewed noted that FAA may receive a high credit rating given that ATC services are essential and FAA has a monopoly in providing them. If a capital project has a high degree of “essentiality,” then it is assumed that the government will pay for the project through appropriations if that is the revenue source. Representatives of an investment firm we interviewed also noted that FAA may receive an implied federal guarantee because it is a federal agency. However, representatives of another investment firm we interviewed also said that many of FAA’s assets may have a low degree of marketability. That is, lenders may have difficulty selling an asset in the market in case of a bond default because there may be few willing buyers in the market for it. Borrowing costs are particularly important in light of the federal government’s long-term fiscal imbalance. As the baby boom generation ages, mandatory federal commitments to health and retirement programs will consume an ever-increasing share of the nation’s gross domestic product and federal budgetary resources, placing severe pressures on all discretionary programs, including those that fund defense, education, and transportation. Our simulations show that by 2040, revenues to the federal government might barely cover interest on the debt—leaving no money for either mandatory or discretionary programs—and that balancing the budget could require cutting federal spending by as much as 60 percent, raising taxes by up to 2½ times their current level, or some combination of the two. Accordingly, any program or policy change that may increase costs requires sound justification and careful consideration before adoption. We previously reported that agencies with authority to borrow were financing a large portion of their programs with debt and were repaying their debt with appropriations or new borrowing, rather than through revenue collections. As a result, we recommended that only those agencies that would, in all likelihood, be able to repay their borrowing through revenue collections be granted authority to borrow. We provided a draft of this report to DOT and Treasury for review and comment. We received comments from DOT through an e-mail from FAA’s Director of the Office of Aviation Policy and Plans on September 11, 2006, and from Treasury through an e-mail from the Deputy Assistant Secretary of Government Financial Policy on September 8, 2006. Neither DOT nor Treasury explicitly agreed or disagreed with our observations, and both raised a number of concerns. DOT stated that, in its opinion, although a change in FAA’s governance may not be statutorily required, it may be important as a matter of policy. DOT stated that because air navigation service providers are by nature monopoly providers, users need assurance that their concerns are taken into account in cost control and investment decisions, particularly under a system that more closely ties users’ contributions to the costs of the system. DOT stated that an alternative governance mechanism, along with user fees, could give system users a structured advisory role in how moneys are spent, costs are allocated, and charges are set to recover those costs, while still retaining the inherently governmental decision-making authority within FAA and DOT. In addition, DOT maintained that a governance mechanism specifically designed to give users input into investment decisions and cost recovery would add a valuable layer of discipline in optimizing the system to accommodate users’ needs most efficiently. In contrast, according to DOT, a system in which FAA/DOT could charge fees to cover costs with no meaningful stakeholder involvement would be much less attractive to the stakeholders. Finally, DOT stated, such an arrangement is fully consistent with the position of the International Civil Aviation Organization, which calls for user charges to be set in consultation between the service provider and the user community. DOT may want to encourage Congress to consider the issue of governance structure. However, we did not include an analysis of governance issues in the scope of our review; therefore, we did not provide a more detailed discussion of the issue in this report. DOT stated that our discussion of the need to analyze FAA’s costs implied FAA has not developed any cost accounting or cost allocation systems. Although we agree that FAA has made progress in implementing a cost accounting system, its current accounting system is not able to provide the information required for a cost allocation analysis. Therefore, in our view, our report does not mischaracterize the status of FAA’s cost accounting system by stating that an analysis of the extent to which the current funding approach, or alternative funding approaches, aligns costs with revenues would require the completion of a cost accounting system or the use of cost finding techniques. Our point is that this capability would be needed to operate under a cost-based user charge system. DOT stated that it believes user fees would provide greater revenue stability than taxes because user fees could be set up to be adjusted periodically without changes in the law, thus providing greater flexibility in aligning revenues to cover costs. Nonetheless, we continue to believe that revenue stability is not likely to vary much across the funding options. Significant decreases in the demand for air travel would decrease revenue regardless of whether the current funding structure is maintained or any of the options are adopted. Furthermore, increasing direct user charges while air travel demand was falling would increase costs for aircraft operators at the same time as their revenues were declining and might be no easier than increasing excise taxes. DOT also provided some clarifying and technical comments, which we incorporated where appropriate. According to Treasury, GAO raised several critical issues, but did not provide any analysis that would help policymakers judge reform options. Specifically, Treasury expressed concern that we did not (1) provide a more comprehensive discussion of FAA costs and cost shares, including any available cost information that provides insight into the issue, (2) evaluate FAA’s efforts to implement cost accounting, and (3) state whether FAA’s cost accounting program is likely, when completed, to generate cost information that is useful in determining a fair and efficient distribution of costs among users. We agree with Treasury that a more detailed analysis of FAA costs and cost shares should be conducted to inform the FAA reauthorization debate, and that this information would improve the analysis of specific alternative funding options. FAA’s current accounting system is not able to provide the information required for a cost allocation analysis. We believe that using partial cost information, as suggested by Treasury, would not be appropriate. Moreover, conducting a comprehensive cost analysis was beyond the scope of this report. Treasury also said that our report repeats claims made by interest groups without evaluating them, giving the sense that each argument is equally valid, even though policymakers need some way to evaluate them. This was not the objective of the report. We provided a basis for evaluating the current and alternative funding options by outlining criteria, including revenue adequacy, equity, and efficiency, and discussing the implications of these criteria with respect to specific funding options. Treasury raised concerns that a number of statements were attributed to “some stakeholders,” rather than the specific groups or individuals that made the statements, noting that attribution helps the reader evaluate the statements. In response, we added some attribution as appropriate. Treasury also noted its long-standing policy that all debt issued by federal entities, including FAA, should be issued solely to the Treasury, because centralized Treasury financing of all such debt is the least expensive, most efficient means of financing this debt. Treasury further maintained that market discipline would not be applied to FAA debt obligations issued directly to the private capital market because investors would perceive the obligations were backed by the federal government. We added language to the report to clarify Treasury’s position on these issues. Treasury also provided some clarifying and technical comments, which we incorporated where appropriate. As agreed with your offices, unless you announce the contents of this report earlier, we plan no further distribution until 30 days from the date of this letter. At that time, we will send copies of this report to interested congressional committees; the Secretary of Transportation; the Administrator, FAA; the Secretary of the Treasury; and the Director, OMB. Copies will also be available to others upon request and at no cost on GAO’s Web site at www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-3834 or dillinghamg@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. To accomplish all of our objectives, we reviewed relevant research, including GAO products, academic research, congressional testimony, industry group publications, and stakeholders’ responses to questions FAA asked them about its funding. We also interviewed officials from government agencies, including the Federal Aviation Administration (FAA), the Office of Management and Budget (OMB), the Congressional Budget Office (CBO), and the Department of the Treasury (Treasury); representatives of aviation industry groups, including the Air Transport Association, the Aircraft Owners and Pilots Association (AOPA), and the National Business Aviation Association; and academic and financial experts. In addition, as discussed in the following paragraphs, we performed further work to accomplish each objective. To assess the advantages and concerns that have been raised about the current approach to collecting revenues from national airspace system (NAS) users to fund FAA and the extent to which the available evidence supports the concerns, we examined FAA budget data, Airport and Airway Trust Fund (Trust Fund) revenue data, FAA forecasts, data reported to the Department of Transportation (DOT) on aircraft size and airfares (DOT Form 41 data), and FAA aviation activity data. We used data on tax revenues associated with different types of flights to assess the link between increases in FAA’s workload and increases in Trust Fund revenue. We obtained the FAA budget, Trust Fund, forecast, and aviation activity data from FAA. To assess the reliability of these data, we interviewed knowledgeable officials and reviewed the quality control procedures FAA applies to these data, and subsequently determined that the data were sufficiently reliable for our purposes. We obtained the DOT Form 41 data from BACK Aviation Solutions, a private contractor that provides these data to interested parties. We used these data to examine trends in aircraft size and airfares because of their impact on the relationship between Trust Fund revenues and FAA’s workload. To identify potential alternative funding options for FAA and criteria for comparing these options, we obtained information on the experience of foreign air navigation service providers by reviewing relevant GAO reports and other literature and interviewing officials at Eurocontrol and France’s FAA counterpart, la Direction Générale de l’Aviation Civile. We also interviewed representatives of Air France, AOPA-France, the International Air Transport Association, the Association of European Airlines, and Aéroports de Paris. Through our literature review and these interviews, we identified longer-run revenue adequacy, equity, efficiency, and administrative considerations as appropriate criteria for assessing the current and alternative funding options. We considered both modifications to the current excise tax structure and various forms of direct charges for FAA services as possible alternatives to the current tax structure. In selecting options for analysis, we considered whether there was a link between the option and some element of FAA’s workload. To identify the advantages and disadvantages of authorizing FAA to use debt financing for capital projects, we reviewed the borrowing authorities of other U.S. governmental entities, including the Tennessee Valley Authority and the Bonneville Power Administration. We conducted our work from May 2005 through August of 2006 in accordance with generally accepted government auditing standards. In addition to the contact named above, the following individuals made key contributions to this report: Ashley Alley, Christine Bonham, Jay Cherlow, Tammy Conquest, Colin Fallon, Carol Henn, David Hooper, Maureen Luna-Long, Maren McAvoy, Rich Swayze, and Matt Zisman.

The Federal Aviation Administration (FAA), the Airport and Airway Trust Fund (Trust Fund), and the excise taxes that support the Trust Fund are scheduled for reauthorization at the end of fiscal year 2007. FAA is primarily supported by the Trust Fund, which receives revenues from a series of excise taxes paid by users of the national airspace system (NAS). The Trust Fund's uncommitted balance decreased by more than 70 percent from the end of fiscal year 2001 through the end of fiscal year 2005. The remaining funding is derived from the General Fund. This report focuses on the portion of revenues generated from users of the NAS and addresses the following key questions: (1) What advantages and concerns have been raised about the current approach to collecting revenues from NAS users to fund FAA, and to what extent does available evidence support the concerns? (2) What are the implications of adopting alternative funding options to collect the revenues contributed by users that fund FAA's budget? (3) What are the advantages and disadvantages of authorizing FAA to use debt financing for capital projects? This report is based on interviews with relevant federal agencies, including FAA, the Office of Management and Budget, and the Congressional Budget Office. GAO also obtained relevant documents from these agencies, other key stakeholders, and academic and financial experts. Some stakeholders support the current excise tax system, stating that it has been successful in funding FAA, has low administrative costs, and distributes the tax burden in a reasonable manner. Other stakeholders, including FAA, state that under the current system there is a disconnect between revenues contributed by users and the costs they impose on the NAS that raises revenue adequacy, equity, and efficiency concerns. Trends and FAA projections in both inflation-adjusted fares and average plane size suggest that the revenue collected under the current funding system has fallen and will continue to fall relative to FAA's workload and costs, supporting revenue adequacy concerns. Comparisons of revenue contributed and costs imposed by different flights provide support for equity and efficiency concerns. The extent to which revenues and costs are linked, however, depends critically on how costs are allocated. Thus, to assess the extent to which the current approach or other approaches aligns costs with revenues would require completing an analysis of costs, using either a cost accounting system or cost finding techniques to assign costs to NAS users. The implications of adopting alternative funding options to collect revenue from NAS users and address concerns about the current excise tax system vary depending on the extent to which users' revenue contributions reflect the costs those users impose on FAA. This report considers six selected funding options, including two that modify the current excise tax structure and four that adopt more direct charges to users. Given the diverse nature of FAA's activities, a combination of alternative options may offer the most promise for linking revenues and costs. Switching to any alternative funding option would raise administrative and transition issues. Some stakeholders who support the adoption of direct user charges also support a change in FAA's governance structure, but GAO found no evidence adoption of direct charges requires this. Authorizing FAA to use debt financing for capital projects would have advantages and disadvantages. Some stakeholders identify debt financing as attractive because it could provide FAA with a stable source of revenue to fund capital developments, while at the same time spreading the costs out over the life of a capital project as its benefits are realized. Debt financing raises significant concerns, however, because it encumbers future resources, and expenditures from debt proceeds may not be subject to the congressional oversight that appropriations receive. Concerns regarding borrowing costs, oversight, and encumbering future resources are particularly important in light of the federal government's long-term structural fiscal imbalance. The Departments of Transportation and Treasury provided comments and technical clarifications on a draft of this report which we have incorporated or responded to as appropriate. DOT's comments focused on governance reforms required to adopt a user fee approach, and whether we accurately described the status of FAA's accounting system. Treasury's raised concerns about the level of analytical development for the options and associated issues. Data was not available to conduct the analysis Treasury suggested, and we agree necessary. However, we believe the report provides useful information to facilitate debate on the options.

On September 30, 1991, 8 months after his inauguration as Haiti’s first democratically elected president, Jean-Bertrand Aristide was overthrown by a military coup. On September 18, 1994, after 3 years of economic and diplomatic pressure, including the threat of direct U.S. military intervention, Haiti’s military regime relinquished power and allowed Aristide’s return to office in October 1994. Among other things, this arrangement allowed for the nonviolent entry of U.S. troops and for holding parliamentary elections in a free and democratic manner. In accordance with Haiti’s constitution precluding two consecutive presidential terms, President Aristide indicated that he would hand over power to an elected successor when his term expired in February 1996. From September 1994 through March 1995, the multinational force of about 20,000 U.S. troops and 4,100 military and support personnel from other countries was deployed to Haiti to establish a “safe and secure environment.” On March 31, 1995, responsibility for continuing the mission was transferred to the U.N. Mission in Haiti, which had about 6,900 troops at the time of transfer. U.S. troops comprised about half the U.N. force. International observers documented numerous irregularities in the first round of parliamentary and local elections, held in June 1995. Subsequent runoff and partial elections and the December presidential election proceeded more smoothly, and the electoral council showed more willingness to cooperate with its technical assistance advisors. Technical weaknesses persisted, however, and a shortened electoral period precluded some scheduled activities before the presidential election. The elections proceeded without violence and Haitians were free to exercise their voting rights; however, some instances of violence and intimidation were reported. Also, President Aristide’s ambiguity about his intentions to step aside to his successor may have created some confusion among voters and those who might have financially supported other candidates. In December 1994, the Aristide government appointed a nine-member provisional electoral council that carried out the 1995 parliamentary, municipal, and presidential elections. The electoral council administered elections through nine offices covering Haiti’s geographic departments. Administration was further delegated to the 133 communal offices as election material distribution and collection points for the more than 10,000 polling stations, each assigned 5 pollworkers. Haiti’s 1995 electoral law temporarily set the number of electoral districts at 83, corresponding to the number of seats in the Chamber of Deputies. Some districts were combined to share representation. The first round of parliamentary and local elections was a large and complicated undertaking. The elections were due to be held in December 1994, but they were difficult to organize within 2 months of President Aristide’s return. The elections were rescheduled twice and held on June 25, 1995. USAID estimates that about 97 percent of the potential voting population were registered; this is difficult to substantiate because Haiti lacks current and reliable census data. More than 11,000 candidates ran for about 2,200 seats. These seats included 18 of the 27-member Senate, all 83 members of the Chamber of Deputies, all municipal councils elected as cartels of 3 candidates (133 mayors and 266 deputy mayors), and all 1,695 seats on 565 local community councils. OAS estimated that about 40 to 45 percent of Haiti’s 3.5-million registered voters went to more than 10,000 polling stations; the International Republican Institute (IRI) estimated the turnout at 30 to 40 percent, and the electoral council announced that 51 percent had voted. The election required more than 17 million ballots be printed to accommodate the numerous contests and candidates. International observers of several organizations noted that, although the June 1995 elections were generally peaceful—deadly violence and intimidation that had historically marred Haitian elections were largely absent—the voting process was disorganized and had many technical difficulties and irregularities. For example, many polling stations opened late or did not open at all because they had not received registration lists, ballots, or other election materials due to confusion over logistics. Also, the electoral council added several hundred polling stations just days before the election, causing confusion in delivering election materials and for potential voters who did not know where to vote. Some candidates reported being on the wrong ballot and some ballots had missing photographs of candidates and party emblems. OAS received complaints on 38 ballots out of the 799 different ballots printed, and reported that 32 complaints had merit and 6 were groundless. According to USAID, most ballot problems originated with the electoral council and not the printer. One communal electoral council office was burned and several were damaged by fire after the elections when losing parties tried to burn the ballots and tally sheets; ballot security both before and after actual voting was a concern. Observers and U.S. officials noted that pollworkers were inadequately trained to carry out their responsibilities, and many were not paid on time. Observers found no evidence of organized fraud, but generally noted that the proliferation of ballots and the myriad of candidates were overwhelming in a largely illiterate country with weak infrastructure and little experience in election administration. Given concerns over serious irregularities and the validity of the results in certain problem areas, the electoral council decided to hold makeup elections in those areas. These partial elections took place on August 13, 1995, in 21 communal districts. On September 17, runoff elections were held for 8 Senate and 65 deputy seats; in October, additional partial elections were held in areas that had experienced irregularities. The Platform Politique Lavalas, the coalition supported by President Aristide, dominated the election results, winning 17 of the 18 contested Senate seats, 66 of 83 deputy seats, and a majority of local seats. Alleging that the June elections were marred by fraud, manipulation, and inefficiency at the hands of the Lavalas-dominated electoral council, about two-thirds of the almost 30 parties participating in the June elections boycotted the subsequent partial and runoff elections and called for an annulment of the June results and replacement of the electoral council. Nevertheless, some candidates of the boycotting parties remained on the ballots, and five candidates of the three major boycotting parties won deputy seats. All but one of the parties that boycotted the parliamentary and municipal elections extended the boycott through the presidential election. Our field observations and reports of U.S. officials and international observer groups indicated that Haiti’s electoral administration improved with each election after the June race. Much of the improvement was attributed to having fewer candidates and ballots; the lower voter turnout, while disappointing to observer groups, made voting procedures and vote counting easier to implement. Observers also noted that pollworkers were better trained and prepared and vote counting at the polling stations went more smoothly. Voting materials and registration lists were generally received on time at the polling stations, and the delivery of ballots to communal and department electoral offices for vote counting was more orderly than in June. In addition, political party pollwatchers were present at most polling stations and electoral offices. The electoral council cooperated more fully with representatives of the United States, international organizations, and other donors after the June elections. The turning point was President Aristide’s appointment of a new electoral council president on July 27. According to U.S. embassy and USAID officials, officials of U.S. nongovernmental organizations, and observers’ reports, the new council president made concerted efforts to strengthen the electoral process and improve relations with the donor community. The chief of the U.N. technical assistance team also noted an improved relationship with the electoral council and greater willingness to consider the team’s advice. Enhancing the electoral council’s transparency was the press center established by the National Democratic Institute for International Affairs (NDI) before the September 17 run-off election. The electoral council held regular press briefings at the center, particularly during election periods; presented its electoral budget during a press conference; and televised the lottery held for candidate name placement on the presidential ballot. The new council president also held meetings to coordinate the various civic education activities that the United States and other donors planned for the presidential election. Nevertheless, despite the overall technical improvements noted after the June elections, our observations and the reports of international observers and U.S. officials noted several persisting weaknesses in electoral administration. On December 17, many polling stations did not receive electoral registration lists or received inaccurate ones. OAS observed 49 polling stations where voters with a valid card were prevented from voting when their names did not appear on the list and reported that unlisted voters were observed at 599 polling stations, mostly in the department covering Port-au-Prince. IRI observers also reported irregularities in the electoral lists. As an emergency measure, the electoral council president declared at 11 a.m. that all persons possessing a valid voting card could vote at the polling station indicated on the card, even if their names were not listed at that location. While increasing the potential for voting more than once, this was likely mitigated by the pollworkers’ notations of those who voted, as well as the use of indelible ink on each voter’s thumb. In addition to the electoral list problems, OAS reported some additional irregularities that it characterized as serious. These included the premature signing of the vote count the morning of the election in one department, attributed by OAS to inadequate training, and suspicions of ballot stuffing. The latter was a concern at certain polling places with large numbers of recorded votes where no massive presence of voters had been observed during the day. OAS reported these localities to the electoral council. OAS and IRI reported numerous other irregularities, generally minor, such as failure to count unmarked ballots or to post results at the polling stations. Overall, OAS reported major irregularities in about 2 percent of the 3,134 polling stations visited and observed minor irregularities in 29 percent. Observers generally stated that the irregularities did not appear to be the result of organized fraud and did not have a significant impact on the election’s outcome. Rene Preval won with 87.9 percent of the vote and low voter turnout, reported by the electoral council at 27.8 percent and estimated somewhat lower by OAS. Haiti’s constitution calls for presidential elections to be held the last Sunday in the November preceding the scheduled February inauguration; in 1995, that date would have been November 26. Due to several delays in scheduling the parliamentary and local elections, it was difficult to organize and administer an election by that date following the last partial elections in October. December 17 was set as the latest possible date allowing for vote counting and contesting and a possible runoff before the February 7, 1996, inauguration. The U.N. technical assistance team produced a detailed electoral calendar that indicated the optimal time needed was 110 days. This calendar included the time required for cleaning up the registration lists, registering candidates, campaigning, printing and distributing ballots, counting votes, and announcing and contesting results, for both a first round and a runoff. However, the electoral council made its formal announcement of the December 17 presidential election and published the electoral calendar on November 6. A compressed schedule was therefore necessary. The U.N. team subsequently produced a 60-day calendar that saved time by shortening the period allowed for some activities and eliminating the computerization of the voter registration lists. (According to USAID, it never considered computerizing the voter registration lists as a viable option due to the lack of resources.) Due to the compressed schedule, less than 4 weeks were allowed for candidates to campaign. Some candidates told us that they were adversely affected by the compressed schedule because they had insufficient time to raise funds and organize their campaigns. The shortened electoral period also affected several assistance programs. For example, the electoral council’s party pollwatcher program did not take place as planned. In response to opposition parties’ concerns over technical problems and alleged fraud in the June elections, the electoral council established an Electoral Monitoring Unit as an adjunct to the pollwatchers attached to the individual parties. The unit was to consist of pollwatchers nominated by the candidates as a resource pool to monitor the presidential election and record complaints. Each candidate was permitted to nominate 750 names; this figure equated roughly to the number of polling stations divided by the number of candidates. However, the unit was not fully functional because the candidates did not provide all the planned pollwatchers in time for the election. NDI had originally planned to train 610 party pollwatchers at the departmental and communal levels; in turn, these pollwatchers were to train the remaining 10,250 needed at the polling station level. However, NDI was able to train only 338 participants at the departmental and communal levels due to the lack of time and the candidates’ inability to submit their full quota of names; several candidates did not submit any names. OAS observers reported the presence of Electoral Monitoring Unit pollwatchers in only 20 percent of the polling stations. The short electoral calendar also did not allow for certain scheduled civic and voter education activities and, according to some election observers, may have been one factor causing the low turnout. For example, NDI canceled its civic education program because it was unable to find a suitable Haitian nongovernmental counterpart within the time available. This program was aimed at the middle class which, according to NDI, traditionally has not voted in Haitian elections. IRI also canceled plans to train political party pollwatchers due to lack of time and interest on the part of the Haitian political parties. USAID indicated its belief that the compressed electoral calendar did not degrade the civic education activities or contribute to low voter turnout. USAID said that voter education programs carried out by the electoral council, the International Foundation for Electoral Systems (IFES), the European Union, and the U.S. Military Information Support Team provided election information to the voters. USAID also noted that Gallup polls taken in November and December 1995 indicated that 72 percent of the respondents knew where to register, 87 percent had seen voter education posters and messages, and 79 percent knew the date of the election. According to USAID, voter fatigue, a lack of candidates who captured the public’s imagination, and staunch support for President Aristide were probably more important factors contributing to low voter turnout than any limitations on voter education activities caused by the compressed electoral calendar. Nonetheless, OAS noted in its presidential election report that the low turnout can be attributed to a variety of factors, including the limited impact of the awareness campaign. The Haitian elections, for the most part, proceeded without serious incidents, and most observers agreed that the technical irregularities likely had little impact on the outcomes. However, some observers were concerned about less tangible problems within the electoral environment that, nonetheless, raise questions about the tenuous nature of democracy in Haiti. For example, violence broke out following the November 7 murder of a Lavalas deputy who was also a cousin of President Aristide. Aristide gave an emotional eulogy, denouncing the international community for not doing enough to disarm those associated with the coup regime and calling on the people to disarm their communities. In mid-November, protests began in Port-au-Prince and other parts of Haiti. Homes were burned, roadblocks erected, and individuals and media were threatened and assaulted, allegedly by Aristide supporters. On November 20, Aristide called for a national dialogue, and the violence abated. On December 12, the house of presidential candidate Leon Jeune was attacked by gunfire, but no one was injured. IRI investigated and reported on alleged acts of intimidation during the 1995 electoral periods. An additional factor in the uncertain electoral climate was Aristide’s perceived lack of commitment to the presidential election. While he repeatedly assured the international community that he intended to hold elections and hand power to his successor, his public statements on this subject were often vague. Many supporters called for him to extend his term for 3 more years to make up the time he lost in exile. At the national dialogue meetings, he indicated he would consider remaining in office if that was what the Haitian people wanted. He publicly endorsed Preval 2 days before the election; his earlier statements indicated that he did not want to influence the outcome of the election early in the campaign. However, some observers and opposition leaders were concerned that President Aristide’s ambiguity created confusion among the voters and those who might financially support Preval’s opponents. A third concern was the boycott by some opposition parties and their assertions that Lavalas, as the party in power, had unfair advantages over the opposition parties. IRI said that it had documented Lavalas’ use of state resources to finance its campaign, but this has not been documented by other observer groups. Under Haitian law, parties may receive some government assistance for campaigning, but funding for the presidential candidates was precluded by Haiti’s precarious economic situation, according to a memorandum from the Prime Minister to the electoral council. The Haitian government provided limited free television and radio air time. Some observers also asserted that the absence of several major opposition parties meant little competition for Lavalas and assured its victory. Other officials opined, however, that the opposition was weak and fragmented and was trying to gain legitimacy through a boycott after its loss at the June polls. The U.S. government spent about $18.8 million in financial support for Haiti’s parliamentary, local, and presidential elections. Other donors contributed about $9 million. U.S. diplomatic and aid officials also made diplomatic efforts and monitored the electoral process closely to resolve or minimize problems. As of April 15, 1996, USAID grantees spent about $15.1 million to support Haiti’s electoral process from the June 1995 parliamentary and local elections through the December 1995 presidential election. More than half of this assistance—about $9.1 million—was expended under a grant to the United Nations to finance technical assistance and budget support to Haiti’s electoral council. Technical assistance was provided by a team of U.N. election experts in Haiti. The remaining USAID funds were grants to four U.S. nongovernmental organizations for election observation, assistance, and support. IRI fielded pre-election and election observation missions. IFES trained pollworkers and procured a total of 31 million ballots, NDI conducted political party-strengthening activities, and the American Institute for Free Labor Development (AIFLD) participated in voter registration. IFES, NDI, and AIFLD also conducted various civic education activities. In addition, the State Department granted $29 million to OAS for nationwide human rights monitoring and reporting; about $3.7 million was spent for OAS observation of the 1995 Haitian elections. Table 1 summarizes expenditures by each grantee. In addition to financial support, the United States made diplomatic efforts to assure that the elections were held and a successor to Aristide inaugurated by February 1996. Our review of embassy cable traffic and discussions with embassy officials revealed extensive U.S. monitoring of the electoral process and U.S. efforts to ensure that problems were addressed and schedule delays minimized. For example, the electoral process was often a primary agenda item for the Ambassador’s weekly meetings with President Aristide and for other meetings between embassy officials and their Haitian counterparts. Several high-level U.S. delegations visited Haiti during the electoral periods; the Deputy Secretary of State mediated negotiations between Aristide’s Lavalas party and the boycotting opposition parties in August 1995. In addition, USAID held daily and weekly meetings with the U.N. technical assistance team and U.N. Mission in Haiti officials to keep the process on track. As of April 1996, the USAID Office of the Inspector General was conducting an audit of USAID’s internal controls for accounting for U.S. election support funds for Haiti and will report separately on its findings. The Inspector General staff generally found adequate controls over funds expended by AIFLD, IFES, IRI, and NDI. The grant agreement and funding arrangements for U.N. technical and budget support to the electoral council contained accountability weaknesses that impeded detailed oversight and limited USAID’s ability to influence how grant funds were spent. The United Nations required the electoral council to hire an accounting firm to maintain its records and to account for its expenditures through September 30, 1995. This contract was subsequently extended to account for all electoral council donor funds and expenditures. As of February 1996, the accounting firm had reported on expenditures only through August 31, 1995. Due to these delays, USAID provided $30,000 for the firm to hire additional staff for its review. The Inspector General was unable to audit the expenditure of funds provided to the OAS for elections assistance because OAS, as an international organization, denied U.S. auditors access to its accounting records. However, the OAS internal auditors conducted an audit of OAS funds expended in Haiti and plans to issue their report this summer. USAID’s ability to monitor and influence the use of funds provided to the United Nations for its support to the electoral council was impeded by the grant agreement’s weak accountability and reporting requirements and a multilateral trust fund arrangement that precluded detailed donor oversight. The initial grant agreement had been negotiated in Washington, D.C., under standard reporting requirements applying to all U.N. trust funds. When additional funds were needed for the December election, the USAID mission included a clause in the amendment requiring U.N. quarterly financial and progress reports and a trust fund audit. However, U.N. officials in New York would not sign the amendment with the extra conditions. They maintained that all trust funds had to comply only with the standard annual financial reporting requirement. Although the amendment was signed without the additional report and audit conditions, USAID mission officials noted that raising the oversight issue resulted in focusing more U.N. attention on the Haiti trust fund. For example, U.N. headquarters requested from the U.N. Mission in Haiti information on controls over trust fund advances to the electoral council. The Chief of the U.N. technical assistance team in Haiti said that the team’s role was to provide technical election advice to the electoral council, which the council was free to consider or reject. This assistance was accomplished primarily through the development of election schedules and budgets and daily contact with electoral council staff. USAID officials in Haiti told us that they would have preferred more proactive efforts by the U.N. technical assistance team, but their leverage was limited. USAID met regularly with the U.N. technical assistance team and other U.N. and OAS officials involved in the electoral process to attempt to mitigate the team’s passive assistance role, urging both stronger program discipline and greater financial accountability. Weak accountability requirements did not seriously damage the electoral process, but did weaken USAID’s ability to require actions beyond the limited scope of the grant agreement. For example, the U.N. team declined to implement the recommendations of a joint August 1995 U.N./USAID study on the election programs in Haiti. The study recommended, among other things, that the United Nations provide financial and management consultants to the electoral council and assist the council in developing a data base for its operations and an analysis of lessons learned from the June election. (Two weeks before the December presidential election, the U.N. team agreed to use a newly arrived French technical assistance contractor to help the electoral council improve its management. However, the contractor was also working on management problems at the council’s departmental office covering Port-au-Prince, and the recommendation went unimplemented.) USAID program and Inspector General officials told us that the U.N. and the Haitian electoral council cooperated with the Inspector General’s staff by providing (1) records to support some of the larger U.N. expenditures associated with its initial election efforts and (2) summary reports indicating how grant funds were being spent. The electoral council granted auditors access to its accounting records, bank account and disbursement records, and the reports of its independent accounting firm. As a country with a long history of repressive and brutal military dictatorships, Haiti has a human rights situation that continues to concern international human rights observers and the U.S. Congress and executive branch. Human rights experts estimate that at least 3,000 individuals were killed for political reasons by the coup regime after President Aristide’s ouster in 1991. The number of politically motivated killings and abuses has decreased dramatically since the intervention of international forces and the return of President Aristide, but allegations of political murder and abuse continue to plague Haiti. Since October 1994, human rights monitors have reported that about 20 murders may have been politically motivated. Other factors, such as robbery, were ruled out, and all victims were targeted and killed in execution style. About half of these victims had been former army members or otherwise were considered Aristide opponents. A prominent case was the March 1995 murder of Mireille Durocher Bertin, an attorney and outspoken critic of President Aristide, and her client. The Haitian government asked the Federal Bureau of Investigation (FBI) to assist in its investigation of the Bertin case. The FBI concluded its investigation in June 1995, but it was unable to interview Haitian government and interim police officials under impartial conditions. The FBI Deputy Assistant Director testified that investigators did not find sufficient evidence to attribute responsibility to specific individuals for the Bertin murder, but they developed definitive evidence linking the murder to other recent execution-style killings. The FBI briefed Haitian government investigators on the results of its investigation in December 1995. The Haitian government had made no progress in investigating alleged cases of political killings as of March 1996. While it established an investigative unit specifically to review cases of politically motivated murders and assigned government attorneys, it had not provided the support and direction needed, and the unit had not undertaken any investigative work. OAS has noted overall improvement in the human rights situation since the coup regime relinquished power, but has also pointed out persisting weaknesses in the criminal justice system and possible excessive use of force by the Haitian National Police. Responding to U.S. pressure to bring closure to alleged human rights violations and congressional restrictions on aid, the Haitian Ministry of Justice established a Special Investigative Unit in October 1995. The unit is charged with investigating cases of alleged political murders, mostly of prominent political and business leaders, that took place between 1988 and 1995. These include more than 20 cases that occurred following President Aristide’s return and about 30 committed during the coup period. The unit was staffed by 10 new Haitian National Police officers with little training or experience in investigative work. The State Department has proposed that full staffing capability would be 40 police investigators. Five civilian police monitors from the U.N. Mission in Haiti were assigned to the unit to provide technical assistance and on-the-job training. Additionally, the State Department contracted for two U.S. investigators to provide technical assistance and report to the State Department on the unit’s progress and good faith efforts. The Ministry of Justice assigned an investigative attorney and a prosecuting attorney to the unit’s cases. When we visited the Special Investigative Unit on October 25, 1995, about 2 weeks after it was established, it had not yet received dossiers or other case documentation from the Haitian authorities, the OAS/U.N. International Civilian Mission, or Haiti’s Truth Commission for human rights investigation. The OAS/U.N. International Civilian Mission subsequently provided the unit summary information, and the Haitian government provided some case files. By December 1995, the unit had 20 dossiers. At that time, the unit prioritized the first 18 cases to be investigated and later added the Bertin case to the list, bringing the priority cases to 19. Fourteen of these cases occurred before Aristide’s October 1994 return; some went back as far as 1988. As of March 1996, no investigative work had been accomplished. According to a State Department official, the unit had fallen into disuse, and the Haitian investigators had been reassigned to other cases. The unit lacked complete documentation on many cases, particularly forensic data, and Haitian witnesses are traditionally fearful of providing information to authorities. Progress on human rights investigations was slowed after the Prime Minister resigned in October and the newly elected National Assembly confirmed a new Prime Minister and cabinet. According to senior State Department officials, the primary obstacle to the unit’s progress was the lack of clear direction and support from the Haitian government. They expressed the expectation that President Preval would take human rights investigations seriously and have urged him to do so. As of April 22, 1996, according to one State Department official, the unit’s progress had improved significantly in a short period. This official stated that a chief for the unit had been assigned, the prosecuting attorney was working full time with the unit, and the investigators were conducting routine investigative work, such as interviewing witnesses and tracking down vehicle license plates. USAID has acknowledged that the Special Investigative Unit accomplished little before the inauguration of President Preval in February, but said that the Haitian government has demonstrated a greater commitment to the unit and some progress had been made in investigating the Bertin murder case under Preval’s leadership. The OAS/U.N. International Civilian Mission has investigated various human rights issues since its return to Haiti in October 1994, including about 20 execution-style killings since the return of President Aristide. We reviewed the mission’s reports and interviewed the Executive Director on each of our four field trips to Haiti. Reporting appeared adequate; however, we could not determine the quality and completeness of investigations because we were denied access to case files by the Executive Director due to concerns over witness confidentiality. In addition to the execution-style killings, the OAS/U.N. International Civilian Mission has monitored cases of “summary” justice, in which suspected criminals were caught and killed by local citizens, and abuses by state agents, such as the Haitian National Police and the interim police. The mission reported several cases in which excessive force may have been used, including nine people killed by Haitian National Police officers since its deployment in June 1995. (The Washington Office on Latin America reported similar findings.) The mission’s February 1996 report also concluded that key deficiencies remain in the criminal justice system. These include inadequate training, unethical behavior in certain instances by police and judicial officials, lack of material resources, and use of preventive detention. This detention, when combined with judicial delays, meant that only 12 percent of prison detainees had been convicted in a court of law. The report noted, however, that the overall human rights situation has improved dramatically since the period of the coup regime, when thousands of politically motivated murders and other abuses allegedly took place. The mission also observed gradual improvements in the administration of justice, including penal reform and the establishment of a magistrate academy for judicial training. In commenting on a draft of this report, USAID said that the report overall presents a fair and balanced assessment of U.S. assistance for the Haitian elections. USAID also offered several clarifications and technical corrections, as well as updated information, that we have incorporated throughout the report as appropriate. Appendix I provides more detailed information on the amounts and types of election and election-related assistance undertaken by the grantees. USAID’s comments are reprinted in appendix II. Although the State Department indicated that it had no specific comments, its letter is reprinted in appendix III. To obtain information for this report, we traveled to Haiti four times in 1995 and observed the September and December elections with teams from OAS, IRI, and the U.S. embassy. In Haiti we met with officials from the U.S. embassy, USAID, the U.N. technical assistance team, the Haitian electoral council, NDI, AIFLD, IRI, and the OAS Electoral Observation Mission. We interviewed the U.N. Secretary-General’s Special Envoy and Chief of Mission in Haiti, the Chief of the OAS/U.N. International Civilian Mission, and the Chief of the U.N. civilian police monitoring unit. We also interviewed leaders of three boycotting opposition parties and the Lavalas party, five presidential candidates, and an official from the presidential palace. We discussed U.S. election assistance programs in Haiti with officials of the State Department, USAID, the four nongovernmental organization grantees, and the Haitian government. We also reviewed documentation such as grant agreements and scopes of work, USAID reports and election updates, embassy cables, election observation team reports, and election calendars. We reviewed issues related to human rights investigations through interviews with State Department and embassy officials, the OAS/U.N. International Civilian Mission Chief, the Human Rights Watch/National Coalition for Haitian Refugees program officer in Haiti, and U.N. civilian police monitors assigned to the Special Investigative Unit. We also reviewed embassy cables and correspondence from the Haitian Ministry of Justice. We coordinated our work with the staff of the USAID Office of the Inspector General and shared preliminary findings throughout the review. We conducted our review between August 1995 and March 1996 in accordance with generally accepted government auditing standards. Unless you announce its contents earlier, we plan no further distribution of this report until 14 days after its issue date. At that time, we will send copies to other interested congressional committees, the Secretary of State, and the Administrator of USAID. Copies will be provided to others upon request. If you have any questions concerning this report, please contact me at (202) 512-4128. Major contributors to this report are listed in appendix IV. As of April 1996, the United States had spent about $18.8 million to support Haiti’s electoral process from the June 1995 parliamentary and local elections through the December 1995 presidential election. These funds were disbursed mostly through the U.S. Agency for International Development (USAID), which provided grants to the United Nations and various nongovernmental organizations for direct elections support and elections-related support activities. The Organization of American States (OAS) spent about $3.7 million from a State Department grant for election monitoring. More than half of USAID’s assistance was provided through a $9.4-million Elections Support Project grant to a U.N. trust fund, which was also financed by other donors, including Canada, the European Union, and France. About $292,000 obligated under this grant remained unspent as of April 1996. From the trust fund, more than half was provided to the electoral council for budget support to administer the elections. The remainder was spent to cover the costs of the U.N. technical assistance team in Haiti, the purchase of some high-costs items such as vehicles for the electoral council and its regional offices, and for related activities. USAID’s Elections Support Project also granted about $3.76 million to the International Foundation for Electoral Systems (IFES) for ballot procurement, pollworker training, civic education, and a candidate registration data base. IFES had expended all but $60,000 from that grant as of April 1996. IFES received an additional pollworker training grant for $231,926 for the parliamentary elections from USAID’s Bureau for Global Programs. In what was described as a “goodwill gesture,” IFES also computerized the list of polling stations for the electoral council for the December presidential election. USAID provided additional assistance for election-related activities to three U.S. nongovernmental organizations under its Democracy Enhancement Project. A total of about $3.6 million was granted to the International Republican Institute (IRI) for elections observation; the National Democratic Institute for International Affairs (NDI) for political party strengthening and civic education; and the American Institute for Free Labor Development (AIFLD) for labor union participation in voter registration and civic education. These grants date back to 1991 and include other democracy-related activities; we have focused on the 1995 election activities. IRI received a total of $931,132 to train political party pollwatchers, field election observation missions, and document these observations. For the parliamentary and local elections, IRI fielded five observation missions and wrote two reports, including a report documenting its assessment of the problematic June elections. For the December presidential election, IRI fielded four observer delegations and produced four election “alerts” and two reports, including a final report on the presidential election. According to USAID, IRI spent about $655,000 on 1995 election activities. NDI grant amounts since 1991 totaled about $1.25 million. NDI’s program for the parliamentary and local elections consisted of the creation of an Electoral Information Center in September to serve as a press center and information clearinghouse, a civic education campaign of radio and television debates (done in conjunction with a Haitian nongovernmental organization); political party and consensus-building seminars; and political party pollwatcher training. For the presidential election, NDI’s program primarily consisted of a civic education campaign of televised roundtables, training seminars for journalists, press conferences, and election-day radio broadcasts from around the country. Various organizations, including the electoral council, the U.N. Mission in Haiti, the U.S. Presidential Delegation, and IRI, used the Electoral Information Center’s facilities to disseminate information. In August 1995, NDI sent three political party leaders to an NDI-sponsored conference in Africa on managing election-related disputes. According to USAID, NDI spent about $865,000 on 1995 election activities, and about $230,000 remained in total unexpended grant obligations. AIFLD has received $1,485,786 in grant obligations since 1991. According to USAID, AIFLD spent about $600,000 on 1995 election activities that included fielding a monitoring delegation and supporting the activities of several trade union confederations. These funds were administered by AIFLD. For the parliamentary and local elections, AIFLD’s program consisted of helping to organize a nonpartisan trade union election commission to plan and execute election-related activities; a civic education campaign of seminars, radio advertisements, and candidate forums designed to register voters and encourage voting; and a union pollwatcher training program. AIFLD also fielded an election-monitoring group and reported on the parliamentary and local elections. For the presidential election, AIFLD carried out a civic education program of radio advertisements, banners, and forums, including holding a candidate forum designed to familiarize trade union leaders with the candidates and their views and wrote a report on the presidential elections. According to a USAID report and discussions with mission officials, USAID was disappointed with AIFLD’s election assistance program for the parliamentary and local elections, saying that it was unable to measure any output for AIFLD’s election work. Specifically, USAID’s complaints centered on (1) AIFLD’s lack of financial and program reporting; (2) USAID’s perception that AIFLD was not carrying out its program; (3) AIFLD’s overhead costs, which USAID viewed as excessive; and (4) AIFLD’s absence from several key donor meetings in May and August 1995. USAID also questioned the cost-effectiveness of AIFLD’s June 1995 union pollwatcher monitoring group, saying that AIFLD delegates received their observer credentials too late to be of any use. USAID did not believe AIFLD’s claim to have registered 800,000 voters for the June 1995 elections because it was not backed by any verifiable data. In August 1995, USAID commissioned an evaluation of AIFLD’s program in Haiti. This evaluation concluded that the management of AIFLD’s program was deficient; that it lacked adequate planning, monitoring, reporting, and accounting systems; and that USAID’s money could have been spent more effectively. The report also concluded that AIFLD’s program had helped preserve and develop the Haitian trade union movement, particularly during the years of Aristide’s exile and that AIFLD had played a significant role in registering voters for the June elections. The report recommended the establishment of a work plan, an improved flow of financial and program information, a short-term focus on civic education, and documentation of AIFLD-assisted trade union accomplishments. As a result, USAID reduced AIFLD’s budget for the presidential election and more narrowly focused the program on civic education. USAID believes the resulting program was more successful, particularly AIFLD’s “candidate forum,” which gave labor leaders a chance to meet and discuss substantive issues with the presidential candidates. AIFLD officials acknowledged that the parliamentary and local elections program could have been better managed, but said that USAID had understated the program’s accomplishments. Specifically, AIFLD pointed to the establishment of the Trade Union Election Commission, labor’s involvement in the Tripartite Commission discussing privatization and other issues in Haiti, and its claim to have helped register 800,000 people during the spring 1995 registration period as being important accomplishments. AIFLD officials also said that USAID’s decision to allocate funding for short periods reduced program effectiveness. AIFLD admitted that its June 1995 monitors did not receive observer credentials in time, but stated that they still served in an unidentified capacity. Between 1992 and 1994, the State Department granted $29 million to the OAS/U.N. International Civilian Mission in Haiti for human rights monitoring. Since October 1994, the mission has conducted related programs in the areas of civic education, administration of justice, and medical services. About $3.7 million was spent for election observation and reporting by the OAS Electoral Observation Mission. The OAS Electoral Observation Mission brought in outside observers for the elections, but most of its observers were OAS/U.N. International Civilian Mission human rights monitors who were seconded to election observation. OAS observers totaled 293, 174, and 320, respectively, for the June, September, and December elections. Both monitoring units issued regular reports on their findings. The OAS/U.N. International Civilian Mission issued monthly human rights reports and periodic press releases and progress reports. The OAS Electoral Observation Mission issued press releases and reports following each election and a final report on all elections. This unit also informed the electoral council of its findings, but maintained that its mandate did not include providing technical assistance or monitoring enforcement. The following is GAO’s comment on USAID’s letter dated May 16, 1996. 1. USAID attached to its letter several points of clarification, technical corrections, and updated information that have been incorporated throughout the report as appropriate. Oliver G. Harter The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed U.S. efforts to foster democratic elections and increased respect for human rights in Haiti, focusing on: (1) how the elections in Haiti were conducted; (2) the nature and extent of U.S. support for Haitian elections; (3) whether election assistance funds were properly controlled and spent; and (4) the progress made in investigating allegations of politically motivated killings. GAO found that: (1) during Haiti's parliamentary and local elections in June 1995, international observers noted various irregularities, but subsequent elections were less troubled; (2) most observers agreed that the presidential elections were generally peaceful, citizens were free to vote, organized fraud was not evident, and technical irregularities did not affect the election's outcome; (3) the U.S. government spent about $18.8 million in support of Haiti's parliamentary, local, and presidential elections, including $9.1 million through a United Nations trust fund, $6 million by U.S. nongovernmental organizations, and $3.7 million to support the efforts of the Organization of American States to observe the elections; (4) without U.S. financial and diplomatic support, it is unlikely that the elections would have been held in time to inaugurate the president's successor in February 1996; (5) the Agency for International Development (AID) Inspector General found that adequate controls existed over the use of election support funds granted to the four U.S. nongovernmental organizations; and (6) the human rights situation in Haiti remains fragile and continues to concern the United States and international organizations, despite dramatic improvements.

Since 2000, legacy airlines have faced unprecedented internal and external challenges. Internally, the impact of the Internet on how tickets are sold and consumers search for fares and the growth of low cost airlines as a market force accessible to almost every consumer has hurt legacy airline revenues by placing downward pressure on airfares. More recently, airlines’ costs have been hurt by rising fuel prices (see figure 1). This is especially true of airlines that did not have fuel hedging in place. Externally, a series of largely unforeseen events—among them the September 11th terrorist attacks in 2001 and associated security concerns; war in Iraq; the SARS crisis; economic recession beginning in 2001; and a steep decline in business travel—seriously disrupted the demand for air travel during 2001 and 2002. Low fares have constrained revenues for both legacy and low cost airlines. Yields, the amount of revenue airlines collect for every mile a passenger travels, fell for both low cost and legacy airlines from 2000 through 2004 (see figure 2). However, the decline has been greater for legacy airlines than for low cost airlines. Legacy airlines, as a group, have been unsuccessful in reducing their costs to become more competitive with low cost airlines. Unit cost competitiveness is key to profitability for airlines because of declining yields. While legacy airlines have been able to reduce their overall costs since 2001, these were largely achieved through capacity reductions and without an improvement in their unit costs. Meanwhile, low cost airlines have been able to maintain low unit costs, primarily by continuing to grow. As a result, low cost airlines have been able to sustain a unit cost advantage as compared to their legacy rivals (see figure 3). In 2004, low cost airlines maintained a 2.7 cent per available seat mile advantage over legacy airlines. This advantage is attributable to lower overall costs and greater labor and asset productivity. Weak revenues and the inability to realize greater unit cost-savings have combined to produce unprecedented losses for legacy airlines. At the same time, low cost airlines have been able to continue producing modest profits as a result of lower unit costs (see figure 4). Legacy airlines have lost a cumulative $28 billion since 2001 and are predicted to lose another $5 billion in 2005, according to industry analysts. Since 2000, as the financial condition of legacy airlines deteriorated, they built cash balances not through operations but by borrowing. Legacy airlines have lost cash from operations and compensated for operating losses by taking on additional debt, relying on creditors for more of their capital needs than in the past. In the process of doing so, several legacy airlines have used all, or nearly all, of their assets as collateral, potentially limiting their future access to capital markets. In sum, airlines are capital and labor intensive firms subject to highly cyclical demand and intense competition. Aircraft are very expensive and require large amounts of debt financing to acquire, resulting in high fixed costs for the industry. Labor is largely unionized and highly specialized, making it expensive and hard to reduce during downturns. Competition in the industry is frequently intense owing to periods of excess capacity, relatively open entry, and the willingness of lenders to provide financing. Finally, demand for air travel is highly cyclical, closely tied to the business cycle. Over the past decade, these structural problems have been exacerbated by the growth in low cost airlines and increasing consumer sensitivity to differences in airfares based on their use of the Internet to purchase tickets. More recently airlines have had to deal with persistently high fuel prices—operating profitability, excluding fuel costs, is as high as it has ever been for the industry. Airlines seek bankruptcy protection for such reasons as severe liquidity pressures, an inability to obtain relief from employees and creditors, and an inability to obtain new financing, according to airline officials and bankruptcy experts. As a result of the structural problems and external shocks previously discussed, there have been 160 total airline bankruptcy filings since deregulation in 1978, including 20 since 2000, according to the Air Transport Association. Some airlines have failed more than once but most filings were by smaller carriers. However, the size of airlines that have been declaring bankruptcy has been increasing. Of the 20 bankruptcy filings since 2000, half of these have been for airlines with more than $100 million in assets, about the same number of filings as in the previous 22 years. Compared to the average failure rate for all types of businesses, airlines have failed more often than other businesses. As figure 5 shows, in some years, airline failures were several times more common than for businesses overall. With very few exceptions, airlines that enter bankruptcy do not emerge from it. Of the 146 airline Chapter 11 reorganization filings since 1979, in only 16 cases are the airlines still in business. Many of the advantages of bankruptcy stem from legal protection afforded the debtor airline from its creditors, but this protection comes at a high cost in loss of control over airline operations and damaged relations with employees, investors, and suppliers, according to airline officials and bankruptcy experts. Contrary to some assertions that bankruptcy protection has led to overcapacity and under pricing that have harmed healthy airlines, we found no evidence that this has occurred either in individual markets or to the industry overall. Such claims have been made for more than a decade. In 1993, for example, a national commission to study airline industry problems cited bankruptcy protection as a cause for the industry’s overcapacity and weakened revenues. More recently, airline executives have cited bankruptcy protection as a reason for industry over capacity and low fares. However, we found no evidence that this had occurred and some evidence to the contrary. First, as illustrated by Figure 6, airline liquidations do not appear to affect the continued growth in total industry capacity. If bankruptcy protection leads to overcapacity as some contend, then liquidation should take capacity out of the market. However, the historical growth of airline industry capacity (as measured by available seat miles, or ASMs) has continued unaffected by major liquidations. Only recessions, which curtail demand for air travel, and the September 11th attack, appear to have caused the airline industry to trim capacity. This trend indicates that other airlines quickly replenish capacity to meet demand. In part, this can be attributed to the fungibility of aircraft and the availability of capital to finance airlines. Similarly, our research does not indicate that the departure or liquidation of a carrier from an individual market necessarily leads to a permanent decline in traffic for that market. We contracted with Intervistas/GA2, an aviation consultant, to examine the cases of six hub cities that experienced the departure or significant withdrawal of service of an airline over the last decade (see table 1). In four of the cases, both local origin-and-destination (i.e., passenger traffic to or from, but not connecting through, the local hub) and total passenger traffic (i.e., local and connecting) increased or changed little because the other airlines expanded their traffic in response. In all but one case, fares either decreased or rose less than 6 percent. We also reviewed numerous other bankruptcy and airline industry studies and spoke to industry analysts to determine what evidence existed with regard to the impact of bankruptcy on the industry. We found two major academic studies that provided empirical data on this issue. Both studies found that airlines under bankruptcy protection did not lower their fares or hurt competitor airlines, as some have contended. A 1995 study found that an airline typically reduced its fares somewhat before entering bankruptcy. However, the study found that other airlines did not lower their fares in response and, more importantly, did not lose passenger traffic to their bankrupt rival and therefore were not harmed by the bankrupt airline. Another study came to a similar conclusion in 2000, this time examining the operating performance of 51 bankrupt firms, including 5 airlines, and their competitors. Rather than examine fares as did the 1995 study, this study examined the operating performance of bankrupt firms and their rivals. This study found that bankrupt firms’ performance deteriorated prior to filing for bankruptcy and that their rivals’ profits also declined during this period. However, once a firm entered bankruptcy, its rivals’ profits recovered. Under current law, legacy airlines’ pension funding requirements are estimated to be a minimum of $10.4 billion from 2005 through 2008. These estimates assume the expiration of the Pension Funding Equity Act (PFEA) at the end of this year. The PFEA permitted airlines to delay the majority of their deficit reduction contributions in 2004 and 2005; if this legislation is allowed to expire it would mean that payments due from legacy airlines will significantly increase in 2006. According to PBGC data, legacy airlines are estimated to owe a minimum of $1.5 billion this year, rising to nearly $2.9 billion in 2006, $3.5 billion in 2007, and $2.6 billion in 2008. In contrast, low cost airlines have eschewed defined benefit pension plans and instead use defined contribution (401k-type) plans. However, pension funding obligations are only part of the sizeable amount of debt that carriers face over the near term. The size of legacy airlines’ future fixed obligations, including pensions, relative to their financial position suggests they will have trouble meeting their various financial obligations. Fixed airline obligations (including pensions, long term debt, and capital and operating leases) in each year from 2005 through 2008 exceed total cash balances of these same legacy airlines by a substantial amount. Legacy airlines carried cash balances of just under $10 billion going into 2005 (see figure 7). These airlines fixed obligations are estimated to be over $15 billion in both 2005 and 2006, over $17 billion in 2007, and about $13 billion in 2008. Fixed obligations in 2008 and beyond will likely increase as payments due in 2006 and 2007 may be pushed out and new obligations are assumed. If these airlines continue to lose money this year as analysts predict, this picture becomes even more tenuous. The enormity of legacy airlines’ future pension funding requirements is attributable to the size of the pension shortfall that has developed since 2000. As recently as 1999, airline pensions were overfunded by $700 million based on Security and Exchange Commission (SEC) filings; by the end of 2004 legacy airlines reported a deficit of $21 billion (see figure 8), despite the termination of the US Airways pilots plan in 2003. Since these filings, the total underfunding has declined to approximately $13.7 billion, due in part to the termination of the United Airline plans and the remaining US Airways plans. The extent of underfunding varies significantly by airline. At the end of 2004, prior to terminating its pension plans, United reported underfunding of $6.4 billion, which represented over 40 percent of United’s total operating revenues in 2004. In contrast, Alaska reported pension underfunding of $303 million at the end of 2004, or 13.5 percent of its operating revenues. Since United terminated its pensions, Delta and Northwest now appear to have the most significant pension funding deficits—over $5 billion and nearly $4 billion respectively—which represent about 35 percent of 2004 operating revenues at each airline. The growth of pension underfunding is attributable to 3 factors. Assets losses and low interest rates. Airline pension asset values dropped nearly 20 percent from 2001 through 2004 along with the decline in the stock market, while future obligations have steadily increased due to declines in the interest rates used to calculate the liabilities of plans. Management and labor union decisions. Airline management has funded their pension plans far less than they could have. For example, PBGC examined 101 cases of airline pension contributions from 1997 through 2002; these cases covered 18 pension plans sponsored by 5 airlines.During this time, $28.2 billion dollars could have been contributed to these pension plans on a tax-deductible basis; actual contributions amounted to $2.4 billion, or about 8.5 percent of what they could have contributed, despite earning profits in 1997-2000 (see figure 9) The maximum deductible contribution was made in only 1 of the 101 pension contribution cases examined by PBGC. In addition, management and labor have sometimes agreed to salary and benefit increases beyond what could reasonably be afforded. For example, in the spring of 2002, United’s management and mechanics reached a new labor agreement that increased the mechanics’ pension benefit by 45 percent, but the airline declared bankruptcy the following December. Pension funding rules are flawed. Existing laws and regulations governing pension funding and premiums have also contributed to the underfunding of defined benefit pension plans. As a result, financially weak plan sponsors, acting within the law, have not only been able to avoid contributions to their plans, but also increase plan liabilities that are at least partially insured by PBGC. Under current law, reported measures of plan funding have likely overstated the funding levels of pension plans, thereby reducing minimum contribution thresholds for plan sponsors. And when plan sponsors were required to make contributions, they often substituted “account credits” for cash contributions, even as the market value of plan assets may have been in decline. Furthermore, the funding rule mechanisms that were designed to improve the condition of poorly funded plans were ineffective. Other legal plan provisions and amendments, such as lump sum distributions and unfunded benefit increases may also have contributed to deterioration in the funding of certain plans. If large numbers of participants in an underfunded plan elect to receive their pension benefits in a lump sum, it can create the effect of a “run on the bank” and exacerbate the possibility of a plan’s insolvency as plan assets are liquidated more quickly than expected. Plan funding can also be worsened by unfunded benefit increases. When a pension plan is underfunded and the plan sponsor is also in poor financial condition, there is an incentive, known as moral hazard, for the plan sponsor and employees to agree to pension benefit increases because at least part of the benefit increases may be insured by PBGC. Finally, the premium structure in PBGC’s single-employer pension insurance program does not encourage better plan funding. While PBGC premiums may be partially based on plan funding levels, they do not consider other relevant risk factors, such as the economic strength of the sponsor, plan asset investment strategies, the plan’s benefit structure, or the plan’s demographic profile. In addition, current pension funding and pension accounting rules may also encourage plans to invest in riskier assets to benefit from higher expected long-term rates of return. The cost to PBGC and participants of defined benefit pension terminations has grown in recent years as the level of pension underfunding has deepened. When Eastern Airlines defaulted on its pension obligations of nearly $1.7 billion in 1991, for example, claims against the insurance program totaled $530 million in underfunded pensions and participants lost $112 million. By comparison, the US Airways and United pension terminations cost PBGC $9.6 billion in combined claims against the insurance program and reduced participants’ benefits by $5.2 billion (see table 2). In recent pension terminations, active and high salaried employees generally lost more of their promised benefits compared to retirees and low salaried employees because of statutory limits. For example, PBGC generally does not guarantee benefits above a certain amount, currently $45,614 annually per participant at age 65. For participants who retire before 65 the benefits are even less; participants that retire at age 60 are currently limited to $29,649. Commercial pilots often end up with substantial benefit cuts when their plans are terminated because they generally have high benefit plans and are also required by FAA to retire at age 60. Far fewer nonpilot retirees are affected by the maximum payout limits. For example, at US Airways fewer than 5 percent of retired mechanics and attendants faced benefit cuts as a result of the pension termination. Tables 3 and 4 summarize the expected cuts in benefits for different groups of United’s active and retired employees. It is important to emphasize that relieving legacy airlines of their defined benefit funding costs will help alleviate immediate liquidity pressures, but does not fix their underlying cost structure problems, which are much greater. Pension costs, while substantial, are only a small portion of legacy airlines’ overall costs. As noted previously in figure 3, the cost of legacy airlines’ defined benefit plans accounted for a 0.4 cent, or 15 percent difference between legacy and low cost airline unit costs. The remaining 85 percent of the unit cost differential between legacy and low cost carriers is attributable to factors other than defined benefits pension plans. Moreover, even if legacy airlines terminated their defined benefit plans it would not fully eliminate this portion of the unit cost differential because, according to labor officials we interviewed, other plans would replace them. Widely reported recent large plan terminations by bankrupt sponsors such as United Airlines and US Airways and the resulting adverse consequences for plan participants and the PBGC have pushed pension reform into the spotlight of national concern. The effect of various proposals to reform pension requirements on airlines, PBGC, and plan participants will vary. The funding relief afforded by PFEA will expire at the end of this year and many agree that the current rules are flawed and must be fixed. Various proposals have been made to correct these rules and shore up the PBGC guaranteed plans, and these proposals are still being debated. The administration has proposed tightening the funding rules among other changes. Some of the legacy airlines with large shortfalls have endorsed another bill in the Senate for a 25-year payback period if current plans are frozen. However, one legacy airline that has better funded its plan, while supporting a longer payback period, opposes freezing their plan. While the airline industry was deregulated 27 years ago, the full effect on the airline industry’s structure is only now becoming evident. Dramatic changes in the level and nature of demand for air travel combined with an equally dramatic evolution in how airlines meet that demand have forced a drastic restructuring in the competitive structure of the industry. Excess capacity in the airline industry since 2000 has greatly diminished airlines’ pricing power. Profitability, therefore, depends on which airlines can most effectively compete on cost. This development has allowed inroads for low cost airlines and forced wrenching change upon legacy airlines that had long competed based on a high-cost business model. The historically high number of airline bankruptcies and liquidations is a reflection of the industry’s inherent instability. However, this should not be confused with causing the industry’s instability. There is no clear evidence that bankruptcy has contributed to the industry’s economic ills, including overcapacity and underpricing, and there is some evidence to the contrary. Equally telling is how few airlines that have filed for bankruptcy protection are still doing business. Clearly, bankruptcy has not afforded these companies a special advantage. Bankruptcy has become a means by which some legacy airlines are seeking to shed their costs and become more competitive. However, the termination of pension obligations by United Airlines and US Airways has had substantial and wide-spread effects on the PBGC and thousands of airline employees, retirees, and other beneficiaries. Liquidity problems, including $10.4 billion in near term pension contributions, may force additional legacy airlines to follow suit. Some airlines are seeking legislation to allow more time to fund their pensions. If their plans are frozen so that future liabilities do not continue to grow, allowing an extended payback period may reduce the likelihood that these airlines will file for bankruptcy and terminate their pensions in the coming year. However, unless these airlines can reform their overall cost structures and become more competitive with low cost competition; this will be only a temporary reprieve. As we have previously reported, the Congress should consider broad pension reform that is comprehensive in scope and balanced in effect.Revising plan funding rules is an essential component of comprehensive pension reform. For example, we testified that Congress should consider the incentives that pension rules and reform may have on other financial decisions within affected industries. Under current conditions, the presence of PBGC insurance may create certain “moral hazard” incentives—struggling plan sponsors may place other financial priorities above “funding up” its pension plan because they know PBGC will pay guaranteed benefits. Further, because PBGC generally takes over underfunded plans of bankrupt companies, PBGC insurance may create an additional incentive for troubled firms to seek bankruptcy protection, which in turn may affect the competitive balance within the industry. In light of the intrinsic problems facing the defined benefit system, meaningful and comprehensive pension reform is required to ensure that workers and retirees receive the benefits promised to them. Ideally, effective reform would incorporate many elements, among them: improving the accuracy of plan funding measures while minimizing complexity and maintaining contribution flexibility; revising the current funding rules to create incentives for plan sponsors to adequately finance promised benefits; developing a more risk-based PBGC insurance premium structure and providing incentives for sponsors to fund plans adequately; addressing the issue of underfunded plans paying lump sums and granting modifying PBGC guarantees of certain plan benefits; resolving outstanding controversies concerning hybrid plans by safeguarding the benefits of workers regardless of age; and improving plan information transparency for pension plan stakeholders without overburdening plan sponsors. The various proposals for comprehensive reform advanced by the Administration and various members of Congress could be a critical first step in addressing part of the long-term stability of the private defined benefits system. While we understand the legacy airline’s liquidity pressures and their request for assistance, the uncertain efficacy of industry-specific relief needs to be weighed against the potential effects on both the industry and the government. At this point, because of a lack of a thorough understanding of those effects, particularly as they might change under various specific legislative proposals, we would suggest proceeding carefully, relying on sound fiduciary principles as a guide. This concludes my statement. I would be pleased to respond to any questions that you or other Members of the Subcommittee may have at this time. For further information on this testimony, please contact JayEtta Hecker at (202) 512-2834 or by e-mail at heckerj@gao.gov; or Barbara Bovbjerg at (202) 512-7215 or by e-mail at bovbjergb@gao.gov. Individuals making key contributions to this testimony include Joe Applebaum, Paul Aussendorf, Anne Dilger, David Eisenstadt, Charles Ford, Charles Jeszeck, Steve Martin, George Scott, Richard Swayze, and Pamela Vines. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Since 2001, the U.S. airline industry has confronted unprecedented financial losses. Two of the nation's largest airlines--United Airlines and US Airways--went into bankruptcy, terminating their pension plans and passing the unfunded liability to the Pension Benefit Guaranty Corporation (PBGC). PBGC's unfunded liability was $9.6 billion; plan participants lost $5.2 billion in benefits. Considerable debate has ensued over airlines' use of bankruptcy protection as a means to continue operations, often for years. Many in the industry and elsewhere have maintained that airlines' use of this approach is harmful to the industry, in that it allows inefficient carriers to reduce ticket prices below those of their competitors. This debate has received even sharper focus with pension defaults. Critics argue that by not having to meet their pension obligations, airlines in bankruptcy have an advantage that may encourage other companies to take the same approach. GAO's testimony presents preliminary observations in three areas: (1) the continued financial difficulties faced by legacy airlines, (2) the effect of bankruptcy on the industry and competitors, and (3) the effect of airline pension underfunding on employees, retirees, airlines, and the PBGC. U.S. legacy airlines have not been able to reduce their costs sufficiently to profitably compete with low cost airlines that continue to capture market share. Internal and external challenges to the industry have fundamentally changed the nature of the industry and forced legacy airlines to restructure themselves financially. The changing demand for air travel and the growth of low cost airlines has kept fares low, forcing these airlines to reduce their costs. They have struggled to do so, however, especially as the cost of jet fuel has jumped. So far, they have been unable to reduce costs to the level of their low-cost rivals. As a result, legacy airlines have continued to lose money--$28 billion since 2001. Although some industry observers have asserted that airlines undergoing bankruptcy reorganization contribute to the industry's financial problems, GAO found no clear evidence that historically airlines in bankruptcy have financially harmed competing airlines. Bankruptcy is endemic to the industry; 160 airlines filed for bankruptcy since deregulation in 1978, including 20 since 2000. Most airlines that entered bankruptcy have not survived. While bankruptcy may not be detrimental to the health of the airline industry, it is detrimental for pension plan participants and the PBGC. The remaining legacy airlines with defined benefit pension plans face over $60 billion in fixed obligations over the next 4 years, including $10.4 billion in pension contributions--more than some of these airlines may be able to afford given continued losses. Various pension reform proposals may provide some immediate liquidity relief to those airlines, but at the cost shifting additional risk to PBGC. Moreover, legacy airlines still face considerable restructuring before they become competitive with low cost airlines.

Depot maintenance is a key part of the total DOD logistics effort and is a vast undertaking, supporting millions of equipment items, 53,000 combat vehicles, 514,000 wheeled vehicles, 372 ships, and 17,300 aircraft of over 100 different models. Depot maintenance requires extensive shop facilities, specialized equipment, and highly skilled technical and engineering personnel to perform major overhaul of weapon systems and equipment, to completely rebuild parts and end items, to modify systems and equipment by applying new or improved components, or to manufacture parts unavailable from the private sector. DOD’s depot maintenance facilities and equipment are valued at over $50 billion. DOD annually spends about $15 billion—or about 6 percent of its $243 billion fiscal year 1996 budget—on depot maintenance activities. About $2 billion of this amount includes contractor logistics support, interim contractor support, and funds for labor associated with the installation of some major modifications and parts of software maintenance, which are contracted to the private sector using procurement, rather than operation and maintenance funds. The DOD depot system, which is actually comprised of four systems,employs about 89,000 DOD civilian personnel, ranging from laborers to highly trained technicians to engineers and top-level managers. Our recent report on closing maintenance depots provides a history of each of the services’ depot systems. While the number of depot personnel has been reduced by over 40 percent relative to when the DOD depot system was at its peak in 1987, depot facilities and equipment have not been similarly downsized. At the time of the 1995 BRAC process, the DOD depot system had 40 percent excess capacity, based on an analysis of maximum potential capacity for a 5-day week, one 8-hour-per-day shift operation. The Air Force, which had not closed a U.S. depot since the 1960s, had 45 percent excess capacity. Currently, there are 29 major DOD depot maintenance facilities—Army depots, Air Force logistics centers, naval aviation depots, naval shipyards, naval warfare centers, and Marine Corps logistics bases—that perform depot maintenance work—of which 10 are in the process of being closed as DOD maintenance depots as a result of BRAC decisions. Additionally, DOD uses over 1,300 U.S. and foreign commercial firms to support its depot maintenance requirements. Statutes and regulations influence the mix of maintenance work performed by the public and private sectors. For example, as early as 1974, legislation prescribed a specific dollar value mix for public and private sector performance of alteration, overhaul, and repair work for naval vessels. Since then, workload allocation decisions have been influenced by percentage goals found in DOD policy guidance and legislation. DOD Directive 4151.1, “Use of Contractor and DOD Resources for Maintenance of Materiel,” directed the services to plan for not more than 70 percent of their depot maintenance to be conducted in DOD depots to maintain a private sector industrial base. The most basic of the legislative mandates governing the performance of depot-level workloads is 10 U.S.C. 2464, which provides for a “core” logistics capability to be identified by the Secretary of Defense and maintained by DOD unless the Secretary waives DOD performance as not required for national defense. Traditionally, core was defined as the capability, including personnel, equipment, and facilities, to ensure timely response to a mobilization, national contingency, or other emergency requirement. The composition and size of this core capability are at the heart of the depot maintenance public-private mix debate. Other statutes affect the extent to which depot-level workloads can be converted to private sector performance. Two of the most significant are 10 U.S.C. 2466 and 10 U.S.C. 2469. The first prohibits the use of more than 40 percent of the funds made available in a fiscal year for depot-level maintenance or repair for private sector performance: the so-called “60/40” rule. The second provides that DOD-performed depot maintenance and repair workloads valued at not less than $3 million cannot be changed to performance by another DOD activity without the use of “merit-based selection procedures for competitions” among all DOD depots and that such workloads cannot be changed to contractor performance without the use of “competitive procedures for competitions among private and public sector entities.” In recent years DOD has sought relief from both these two statutes. DOD and the Congress are defining the role of DOD depots in the post cold war era, much in the same way the roles of U.S. war-fighting forces are being reshaped. The new model for managing depot maintenance has not yet emerged. However, given DOD’s depot maintenance policy report, the model apparently will be a mix between public and private sector capabilities, but with a clear shift toward greater reliance on the private sector. DOD’s March 1996 Depot-Level Maintenance and Repair Workload Report projected a significant increase in the depot work that will be privatized. Further, since the services periodically reevaluate their core workload requirements, it is unknown how much more of their current work will be determined to be non-core and privatized. Unless effectively managed, including downsizing of remaining depot infrastructure, a major shift in depot workloads to the private sector would exacerbate existing excess capacity in the DOD depot maintenance system. Historically, depot maintenance on wartime critical DOD systems has been largely performed in DOD depots. Based on both cost and risk factors, the general DOD policy was to rely on DOD depots to provide a cost-effective and reliable source of support for wartime readiness and sustainability. With some exceptions, peacetime maintenance of weapon systems with wartime taskings was performed in DOD depots. This peacetime workload constituted depot maintenance core. Core was determined by quantifying the depot work that would be generated under war scenarios and then computing the amount of peacetime work needed to employ the number of people necessary to support the anticipated wartime surge. Peacetime workload was composed of a mix of high and low-surge items allowing employees to transfer from low surge workload to high surge workload during war. While there were always a number of potential war scenarios, the depots were sized to support a sustained global war. During the cold war, there was not much pressure to move work from DOD depots to the private sector. Military leaders expressed a clear preference for retaining much of their work in DOD depots, which were highly flexible and responsive to changing military requirements and priorities. The quality of the DOD depots was high and users were generally well-satisfied with the depots’ work. Further, the threat of a global war and the resulting stress on the logistics system were constant reminders of the need to maintain the flexibility and responsiveness the depot system provided. Historically, DOD has reported that about 70 percent of its depot maintenance work was performed in DOD depots. In our 1994 testimony before the Readiness Subcommittee of the House Armed Services Committee, we stated that the private sector more likely receives about 50 percent of the DOD depot maintenance budget. We noted that a portion of the funds expended on the maintenance workload assigned to the public sector ultimately was used for private sector contracts for parts and materiel, maintenance and engineering services, and other goods and services. Additionally, some types of depot maintenance activities, such as interim contractor support and contractor logistics support, were not included in previously reported statistics. Our review of data in DOD’s March 1996 workload report indicates that by fiscal year 1997, the mix will be about 64 percent in the public sector and 36 percent in the private sector. Further analysis indicates that the data does not include funds reported by the services for interim contractor support, contractor logistics support, or goods and services that the DOD depots ultimately buy from the private sector. Including these funds would change the mix to about 53 percent in the public sector and 47 in the private sector. While the Department’s projection for the public-private mix in 2001 is 50 percent in each sector, our analysis indicates that it is actually about 37 percent in the public sector and 63 in the private sector. Further, since the services are conducting risk analyses to further define their minimum core capability, the DOD depots’ share of funding could be reduced even further. With the end of the cold war and the subsequent declines in defense spending, there are increased pressures to privatize more depot maintenance work. Those declines affected force structure and the public and private activities supporting force structure. As acquisition programs began to decline, a growing concern arose over the impact on the defense industrial base. Particular concern focused on how that industrial base could be maintained without the large development and production programs of the past, and attention began to shift to DOD depot workloads as a potential source of work to keep the industrial base viable. Advocates of more private sector involvement argue that a shift toward the private sector would not only help keep the private sector production base healthy during a period of reduced weapon procurement but also could result in lower costs, since the private sector could provide depot maintenance at lower cost than the public sector. Proponents of the DOD depot system believe the DOD depots have provided a quality, responsive, and economical source of repair. They note that DOD maintenance policy for many years has supported the outsourcing of depot maintenance work when it was determined to be cost-effective to do so. Further, they contend there are substantial differences between developing and producing new systems and maintaining fielded ones and that the dollars spent on maintenance, while not small, cannot fill the void created by declining production dollars. Section 311 of the National Defense Authorization Act for Fiscal Year 1996 is an indication of congressional intent regarding the continued need for DOD depots: It is the sense of Congress that there is a compelling need for the Department of Defense to articulate known and anticipated core maintenance and repair requirements, to organize the resources of the Department of Defense to meet those requirements economically and efficiently, and to determine what work should be performed by the private sector and how such work should be managed. Section 311 also directed the Secretary of Defense to develop a comprehensive policy on the performance of depot-level maintenance and repair for the Department of Defense that maintains the core capability described in 10 U.S.C. 2464 and report to the Senate Committee on Armed Services and House Committee on National Security. The section further directed that in developing the policy, the Secretary should include certain elements such as interservicing, environmental liability, and exchange of technical data. The Congress supports preserving a DOD depot maintenance system to support core requirements. With no additional BRACs scheduled, the Department was charged with developing a depot maintenance policy that provides adequate workloads to ensure cost efficiency and technical proficiency in time of peace. We are analyzing DOD’s depot maintenance policy and workload analysis reports, as required by Section 311, and will be reporting our findings by May 18, 1996. However, as requested, I am providing our observations to date on the policy report. First, it provides an overall framework for managing DOD depot maintenance activities. Second, it sets forth a clear preference for moving workload to the private sector, which will likely result in a much smaller core capability than exists today. Third, it is not consistent with congressional guidance in one key area—the use of public-private competitions. Fourth, the policy provides substantial latitude in implementation. As a result, the precise affect of this policy on such factors as public-private mix, cost, and excess capacity remain uncertain. In response to the congressional requirement for a comprehensive statement of depot maintenance policy, DOD provided an overall framework for managing DOD depot maintenance activities. The policy report reiterates some past policies and identifies some new initiatives for depot-level maintenance. It references other directives, publications, memorandums, and decisions and notes that DOD plans to develop an updated single publication with applicable maintenance policy guidance. Our assessment is based on observations to date about the policy report and other related documents. The policy report clearly states that the Department has a preference for privatizing maintenance support for new systems and for outsourcing non-core workload. It represents a fundamental shift in the historical policy of relying on DOD depots to provide for the readiness and sustainment of wartime tasked weapon systems. Section 311 of the authorization act states that the DOD policy should provide that core depot-level maintenance and repair capabilities are performed in facilities owned and operated by the United States. It also states that core capabilities include sufficient skilled personnel, equipment, and facilities that are of the proper size to ensure a ready and controlled source of technical competence, and repair and maintenance capability necessary to meet the requirements of the National Military Strategy and other requirements, and to provide for rapid augmentation in time of emergency. Core, as set forth in the policy and workload reports, no longer means that wartime work will be performed primarily by DOD depots. DOD’s core concept is for its depots to perform maintenance requirements that the service secretaries identify as too risky for the private sector to perform. In determining core workloads, the DOD policy calls for maintaining only “minimum capability”—which does not necessarily mean an actual workload for a depot. What once was calculated as core is now called pre-risk core. For those mission essential workloads that historically would dictate retention of a core capability, the services will conduct a risk assessment to determine if the work should be made available for competition within the private sector. The policy guidance provides some limited criteria for performing a risk assessment, but DOD has not yet developed guidelines for making those assessments in a consistent manner. It is unclear the extent measured criteria or subjective judgement will be used for such assessments. In a similar vein, DOD’s policy on depot maintenance seeks to severely limit the use of DOD depots for new weapon systems. Section 311 provides for the performance of maintenance and repair for any new weapon systems defined as core in facilities owned and operated by the United States. On the other hand, the Department reported to the Congress in August 1995 that it intended to privatize depot maintenance for new systems and reported in its January 1996 depot maintenance privatization initiative that it intended to freeze the transition of new workloads to DOD depots. The policy report and other recently issued DOD guidance, such as DOD Instruction 5000.2, also show that DOD’s maintenance concept for new and modified systems will minimize the use of DOD depots. This preference, in combination with DOD’s minimum core concept and limited public-private competitions, if not effectively managed—including reducing infrastructure and developing competitive markets—would likely result, over the long term, in DOD depots becoming an economic liability rather than a cost-effective partner in the total DOD industrial base. The DOD policy report states that the Department will provide for cost efficiency, sufficient workload, and technical proficiency in its depots. However, accomplishing this objective will be difficult given that the depots already are underutilized and the policy providing for additional outsourcing would exacerbate that situation, unless there are additional depot closures. Further, the report does not provide a clear indication, aside from recognizing ongoing BRAC actions, on how the Department intends to downsize to minimum core. While we are in the process of reviewing the policy report for consistency with congressional direction and guidance, our observation to date is that the report is inconsistent in one key area—the use of public-private competitions for allocating non-core depot maintenance workloads. Section 311(d)(5) of the act provides that in cases of workload in excess of the workload to be performed by DOD depots, DOD’s policy should provide for competition “between public and private entities when there is sufficient potential for realizing cost savings based upon adequate private-sector competition and technical capabilities.” DOD’s report provides a policy that is inconsistent with this instruction. According to DOD, it will engage in public-private competition for workloads in excess of core only when it determines “there is not adequate competition from private sector firms alone.” The report did not clarify what would constitute adequate competition. Under this policy, DOD depots would be used sparingly for public-private competitions and DOD depots cannot compete for all non-core workloads, where adequate private sector competition exists, even though the DOD depots could offer the most cost-effective source of repair. We have reported that public-private depot maintenance competitions can be a beneficial tool for determining the most optimum cost-effective source of repair for non-core workloads. As noted in our recent reports on the Navy’s depot maintenance public-private competition programs for ships and aviation, we found that these competitions generally resulted in savings and benefits and provided incentives for DOD depot officials to reengineer maintenance processes and procedures, to develop more cost-effective in-house capability and to ensure that potential outsourcing to the private sector is more cost effective than performing the work in DOD depots. We recognize that DOD’s public-private depot maintenance competition program raised concerns about the reliability of DOD’s depot maintenance data and the adequacy of its depot maintenance management information systems. These deficiencies are not insurmountable. As we noted in prior reports, many of the problems were internal control deficiencies that can be addressed with adequate top-level management attention. We also noted that some corrective actions have already been undertaken and additional improvements can be made. Further, we recommended that the Defense Contract Audit Agency be used to certify internal controls and accounting policies and procedures of DOD depots to assure they are adequate for identifying, allocating, and tracking costs of depot maintenance programs and to ensure proper costs are identified and considered as part of the bids by DOD depots. DOD has stated that it plans to use the Defense Finance and Accounting Service to review and certify the accounting systems of DOD depots. The policy report provides wide implementation latitude in a number of key areas. For example, it provides for a DOD depot capability, but the ultimate extent of such capability, and hence DOD depot requirements, could be substantially reduced depending on future core workload assessments of privatization, readiness, sustainability, and technology risks. Depending on implementation, the policy’s preference for privatization and the lack of a clear and consistent methodology for determining risks will likely lead to significant amounts of workload previously designated as core being reclassified as non-core and privatized. For example, with respect to the Aerospace Guidance and Metrology Center, the Air Force is privatizing depot maintenance operations involving 627,000 direct labor hours of work—100 percent of which had been previously defined as core—stating that because the workload is being privatized-in-place, the risk is manageable. It is unclear how risky that privatization may turn out to be, particularly in light of the contractor’s interest in divesting itself of its defense business. However, a similar rationale is being used to support other in-place privatizations. With this predilection, it is likely that future core will represent something far different than it did in the past. For example, DOD’s March 1996 workload report noted that core would ensure “that the Air Force establishes and retains the capabilities needed to assure competence in overseeing depot maintenance production that has both public and private sector elements”—a significantly different mission than that historically envisioned for DOD’s core capability. Further, DOD’s March 1996 report to Congress, Improving the Edge Through Outsourcing, included intermediate maintenance of DOD weapons and equipment—another function traditionally considered core—as one which the Department will now consider privatizing. The policy also provides wide latitude in several areas where the decision for determining public or private sector source of repair is based on an assessment of what is “economical” or “efficient.” For example, the policy states that non-core workloads be made available for only private sector competition when it is determined that the private sector can provide the required capability with acceptable risks, reliability, and efficiency. This efficiency requirement does not require the inclusion of the public sector to ensure that privatization is the most cost-effective option. The underlying assumption behind DOD’s depot maintenance privatization initiative is the expectation that savings of 20 percent will be achieved and these savings will be made available to support the services’ modernization programs. Our analysis indicates that this assumption is unsupported. The data cited by Department officials to support this savings assumption is the Report of the Commission on Roles and Missions of the Armed Forces. In May 1995 the CORM concluded that 20 percent savings could be achieved by the privatization of various commercial activities and recommended that DOD transfer essentially all depot maintenance to the private sector. The Commission rejected the notion of core and recommended that DOD (1) outsource all new support requirements, particularly the depot-level logistics support of new and future weapon systems and (2) establish a time-phased plan to privatize essentially all existing depot-level maintenance. In its August 1995 response to the Congress on the CORM report, DOD noted that the Department agreed with the Commission’s recommendation to outsource a significant portion of its depot maintenance work, including depot maintenance activities for new systems. However, the DOD response noted that DOD must retain a limited core depot maintenance capability to meet essential wartime surge demands, promote competition, and sustain institutional expertise. We found that the Commission’s assumptions on savings from privatization generally were based on reports of projected savings from public-private competitions for various commercial activities as part of the implementation of OMB Circular A-76. These commercial activities reviews included various base operating support functions, such as family housing, real property and vehicle maintenance, civilian personnel administration, food service, security and law enforcement, and other support services. While these activities were varied in nature, they had similarities in that they generally involved low-skilled labor; required little capital investment; generally involved routine, repetitious tasks that could readily be identified in a statement-of-work; and had many private sector offerors who were interested and had the capability to perform the work. Our review of A-76 competitions and public-private competitions for depot-level maintenance found that the conditions under which A-76 competitions resulted in lower private sector prices were often not present or applicable to depot maintenance. Specifically, we found that: Reengineered government activities won about half of the A-76 competitions because they could provide the work cheaper. Our work shows that for public-private competitions involving depot maintenance activities, a program authorized by the Congress and implemented independently from A-76, DOD depots won 67 percent of the non-ship competitions. Public-private competitions for ships provided a unique situation wherein private sector offerors could bid marginal or incremental costs while DOD depots were required to bid full costs—a condition which, in concert with the more competitive nature of the ship repair market, led to the public shipyards not being competitive. When the private sector won A-76 competitions, savings were significantly higher than when the government function was performed by military personnel. The additional costs of military pay and benefits when coupled with productivity losses incurred for additional duties resulted in decreased competitiveness of the military personnel assigned to these duties. Depot maintenance, on the other hand, is performed almost exclusively with civilian personnel. The A-76 competitions did not involve activities comparable to depot maintenance—which is far more complex, less repetitious, and involves many unique systems not found in the private sector. Problems associated with statements of work in A-76 competitions resulted in cost increases for privatized work because of contract modifications to more explicitly define required work—a condition we also identified in our review of DOD’s public-private program for depot maintenance. The impact of this cost growth for depot maintenance competitions can be illustrated by submarine repair competitions. While the average award amount for private shipyards was 16 percent less than that for competitions won by the public sector, greater cost growth in the private sector resulted in the average actual costs being about the same. While the A-76 commercial activity competitions resulted in savings, the savings were not readily quantifiable, did not consider the cost of the competition or the administration of the contracts, and for those competitions that were audited, savings were often less than projected. We had similar findings in our review of public-private competitions for depot maintenance. Additionally, we found that for the non-ship depot maintenance competitions won by a DOD depot, the DOD depots’ bids averaged 40 percent less than the lowest private sector offeror. Where we observed cost growth in the limited number of depot competitions we analyzed, the growth was not sufficient to result in the DOD depots’ costs exceeding the bid of the lowest private sector offeror. The A-76 competitions were conducted in a highly competitive private sector market—frequently involving 4 or more offerors, with 10 percent of the competitions involving 11 or more offerors. Savings were much higher for those A-76 competitions won by the private sector where there were 5 or more private sector offerors. Our review of DOD’s 95 non-ship depot maintenance public-private competitions showed the private sector market to be significantly less competitive. Twenty-two of the competitions had no private offerors and 33 had only one. Only 28 of these competitions had three or more offerors, while the number of offerors averaged less than two per competition. Recognizing the influence of competition on achieving savings from privatization, we analyzed the competitiveness of DOD’s non-ship depot maintenance repair contracts. We asked 12 DOD buying commands to identify depot maintenance contracts that were open during 1995. They identified 8,452 contracts valued at $7.3 billion and, based on high dollar value, we selected 240 contracts valued at $4.3 billion to analyze the commands’ use of competitive procedures for the contracted workloads. The following table shows the results of our analysis. As shown, the 12 buying commands awarded (1) 182, or 76 percent, of the contracts through sole-source negotiation; (2) 49, or 20 percent, through full and open competition, and (3) 9, or 4 percent, by limited competition. The 49 fully competitive awards accounted for about 51 percent of the total dollar value while the 182 sole-source contracts accounted for about 45 percent of the dollar value. In reviewing the number of offerors for the 49 contracts valued at $2.2 billion that were awarded through full and open competition, we found that the commands averaged 3.6 offers for the 49 contracts—ranging from a low of only 2 offers to a high of 10. For 30 of the 49 contracts—about 86 percent of the $2.2 billion—the number of offers was 4 or less. Five contracts valued at $525.8 million had only two offers, while only 19 contracts valued at $309.4 million had five or more offers. We also found that a large portion of the dollar value of the contracts went to a relatively small number of contractors. Although the total number of contractors involved in the 240 contracts was 71, 13 of these contractors had most of the workload, about 76 percent of the $4.3 billion. Three of these 13 contractors had workload valued at $1.3 billion, about 30 percent of the $4.3 billion. Our analysis of depot maintenance contracts showed that the private sector market was more competitive for certain types of systems and equipment than for others. For example, awards for repair of ground vehicles, trucks, airframes, engines, and other items were more often competitive while sole-source contracts were prevalent for fire control systems, communications and radar equipment, electronic components, and other components. We found that the buying commands sometimes used both DOD depots and private sector sources for repair of a limited number of items. To make price comparisons, we looked at 414 items that buying activities identified as being maintained in both sectors. For 62 percent of the items, the contract price was higher than the price for the same item repaired in a DOD depot. We also analyzed the impact of other conditions relevant to creating a competitive environment. Regarding the ability to clearly define the service to be provided, the buying commands reported that depot maintenance activities present a difficult challenge. For much of the depot maintenance work, specific tasks that must be done, spare and repair parts that will be required, and the type and skill-level of the labor required cannot be identified until the equipment or component is inducted into the repair facility for inspection and repair. Our review of depot maintenance contracts showed the difficulty in constraining cost growth in this environment—particularly when cost-type contracts are used. It also showed the large costs normally associated with drafting statements of work, conducting the competitions, and administering the contracts. At one buying activity which obligates about $180 million per year for depot maintenance contracts, we found sole-source contracts were used 100 percent of the time—many of which were also cost reimbursable. Officials said they did not have the manpower, technical data, technical manpower, or contracting skills to use competitive contracting. Additionally, officials noted that the process for qualifying repair sources is difficult and time-consuming. There have been a number of recent initiatives to privatize depots recommended for closure or realignment by BRAC. The most prominent among these so-called “in-place” privatization initiatives involve the Aerospace Guidance and Metrology Center, a depot recommended for closure by the 1993 BRAC Commission and located on Newark Air Force Base, Ohio, and the Sacramento and San Antonio Air Logistics Centers, which were recommended for closure by the 1995 BRAC Commission and are located on McClellan Air Force Base, California, and Kelly Air Force Base, Texas, respectively. We previously reported that, although it may be several years before the total cost of privatizing the Aerospace Guidance and Metrology Center’s depot maintenance workload can be identified, our preliminary analysis indicated that this privatization will likely increase, rather than decrease, depot maintenance costs. In addition, our recent analysis of 254 contract items disclosed that (1) unit costs were higher after privatization for 201, or about 79 percent, of the items and (2) overall, there was a net cost increase of $6.01 million for the 254 items. Further, although the Air Force is projecting annual savings of $5 million for the last 4 years of the 5-year contract, we found that the Air Force did not include all relevant costs in its analysis. For example, our analysis showed that the Air Force’s estimated prices for eight contract items did not include such items as material costs totalling $15 million. We also reported on the potential impact of privatizing the San Antonio Air Logistics Center’s engine workload in place rather than transferring the work to the Oklahoma City Air Logistics Center. Specifically, we reported that consolidating San Antonio’s engine workload with Oklahoma City’s engine workload would reduce Oklahoma City’s overhead rate for engine work by as much as $10 an hour and would result in an estimated annual savings of $76 million. As requested by Chairman Spence, we are conducting a more thorough review of the Department’s privatization-in-place initiatives, particularly those underway at San Antonio and Sacramento. Our preliminary observations on these initiatives follow. The BRAC Commission’s July 1995 report to the President noted that the decision to close the Sacramento and San Antonio Air Logistics Centers was a difficult one to make, but was necessary given the Air Force’s significant excess depot capacity and limited defense resources. The Commission report also concluded that these actions should save about $151.3 million over the 6-year implementation period and $3.5 billion over 20 years. Since this announcement, DOD has moved forward with its privatization efforts at these locations, including the announcement that contracts for five prototype workloads are to be awarded by the close of 1997. When the President forwarded the BRAC Commission recommendations to the Congress, he stated that his intent was to privatize the work in place or in the local communities in order to (1) avoid the immediate costs and disruption in readiness that would result from the relocation of the centers’ missions, (2) mitigate the impact on the local communities, and (3) preserve important defense work forces. The administration also decided to delay the centers’ closures until the year 2001 to further mitigate the adverse impact on the local communities. Our analysis indicates that delaying the centers’ closures until 2001 could increase net costs during the 6-year BRAC implementation period by hundreds of millions of dollars, primarily because it would limit the Air Force’s ability to achieve recurring savings to offset expected closure costs. Additionally, although the closures’ potential impact on local communities and readiness is a valid concern, actions can be taken to limit the impact. For example, the Sacramento community’s successful conversion of the Sacramento Army Depot to private use has demonstrated that this conversion, although difficult, can be accomplished. Further, according to Navy depot maintenance officials, on-going efforts to quickly close three aviation depots have had no significant impact on readiness. Our preliminary analysis also indicates that privatizing the two centers’ depot maintenance workloads in place is likely to be a more costly alternative than transferring the workloads to the three remaining centers. One reason for this is that there are substantial costs associated with privatization-in-place that do not apply to DOD maintenance depots. For example, our analysis indicates that unique requirements such as the cost of proprietary data rights, contractor profits, and contractor oversight could add 20 percent, or more, to the cost of performing the work. Further, the cost plus contract that will likely be used is not conducive to generating significant private sector economies, a situation already unfolding at the Aerospace Guidance and Metrology Center. More significantly, our analysis indicates that privatization-in-place eliminates the opportunity to consolidate workloads at the remaining centers and to, thereby, achieve substantial “economy of scale” savings and other efficiencies. The Air Force’s five air logistics centers currently have approximately 57.3 million direct labor hours of depot maintenance capacity to accomplish about 29.3 million hours of workload (projected fiscal year 1999)—leaving a projected excess capacity of 49 percent in 1999. The BRAC decision to close the San Antonio and Sacramento Air Logistics Centers provides the Air Force the opportunity to redistribute workload to the remaining three air logistics centers, thereby reducing excess capacity within its depot system to about 8 percent. Our analysis indicates that redistributing 8.2 million hours of work from Sacramento and San Antonio to the remaining centers would allow the Air Force to achieve annual savings of as much as $182 million. According to financial management officials at the receiving air logistics centers, one-time workload transition costs of about $475 million would be required to absorb the additional workloads, indicating that net savings would occur within 2-1/2 years of the transition completion. On the other hand, if the remaining centers do not receive additional workload, they will continue to operate with significant excess capacity, becoming more and more inefficient and more and more expensive as their workloads continue to dwindle due to downsizing and privatization initiatives. Finally, various statutory restrictions may affect the extent to which depot-level workloads can be converted to private-sector performance—through privatization-in-place or otherwise—including 10 U.S.C. 2464, 10 U.S.C. 2466, and 10 U.S.C. 2469. While each of these statutes has some impact on the allocation of DOD’s depot-level workload, 10 U.S.C. 2469 constitutes the primary impediment to privatization in the absence of a public-private competition. Competition requirements of 10 U.S.C. 2469 have broad application to all changes to the depot-level workload valued at $3 million or more currently performed at DOD installations, including Kelly and McClellan. The statute does not provide any exemptions from its competition requirements and, unlike most of the other laws governing depot maintenance, does not contain a waiver provision. Further, there is nothing in the Defense Base Closure and Realignment Act of 1990—the authority for the BRAC recommendations—that, in our view, would permit the implementation of a recommendation involving privatization outside of the competition requirements of 10 U.S.C. 2469. The determination of whether any single conversion to private-sector performance conforms to the requirements of 10 U.S.C. 2469 depends upon the facts applicable to the particular conversion. We do not have DOD’s position regarding how it plans to comply with the statutory restrictions. While DOD has stated that it will structure these conversions to comply with existing statutory restrictions, details of the Department’s privatization plans for Kelly and McClellan are still evolving. Further, “in-place” privatizations at Newark, Kelly, and McClellan are now the subject of litigation. In March 1996, the American Federation of Government Employees filed a lawsuit challenging these privatization initiatives, contending that they violate the public-private competition requirements of 10 U.S.C. 2469 and other depot maintenance statutes. While our analysis of DOD’s depot policy report continues, we believe there are several points the Congress needs to consider as it contemplates the repeal of 10 U.S.C. 2466 and 10 U.S.C. 2469—two statutes that influence the allocation of depot maintenance workload between the public and private sectors. First, the policy does not provide for participation of DOD depots in depot maintenance competitions for non-core workload as directed by the Congress. Second, since the policy provides wide latitude during implementation, likely outcomes of the policy change are difficult to predict. Third, cost savings are likely achievable from some depot privatization, but not in the percentages and scope predicted by the CORM. Fourth, privatization-in-place does not appear to be cost-effective given the excess capacity in DOD’s depot maintenance system. Given these considerations, the Congress needs to assure itself that any new policy has the intended required features and that a process is in place to monitor readiness, sustainability, and cost considerations. Further, the effective implementation of the new policy will require a further downsizing of the Department’s remaining depot maintenance infrastructure and the development of more competitive private sector markets. Thank you Mr. Chairman that completes my statement. I would be happy to answer questions at this time. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO discussed the privatization of defense depot maintenance activities. GAO noted that: (1) the Department of Defense's (DOD) evolving depot maintenance policy includes a public-private mix and shifts work to the private sector where feasible; (2) depot privatization could worsen excess maintenance capacity and inefficiencies if not carefully managed; (3) DOD's policy report provides an overall framework for managing depot maintenance activities and substantial implementation flexibility, but the policy is not consistent with congressional guidance providing for public-private competition for non-core workloads; (4) privatizing depot maintenance is not likely to achieve the 20-percent savings DOD projects, since the 20-percent savings were for commercial-type activities that more readily led to competition which produced the reported savings; (5) most non-ship depot maintenance private-public competitions have been won by the public sector and averaged fewer than two competitors; (6) DOD plans to privatize in place and delay downsizing and closure of two air logistics centers will probably cost more than closing them and relocating their workloads to underutilized defense or private facilities; and (7) statutes governing competition and base closures may have to be repealed or amended before DOD can proceed with its privatization efforts.

GPS is a global positioning, navigation, and timing system consisting of space, ground control, and user equipment segments that support the broadcasts of military and civil GPS signals. These signals each include positioning and timing information, which enables users with appropriately-equipped GPS receivers to determine their position, velocity, and time, 24 hours a day, in all weather, worldwide. GPS is used by all branches of the military to guide troop movements, assist with logistics support and situational awareness, and synchronize communications networks. In addition, weapon systems, including munitions, are guided to their targets by GPS signals, and GPS is used to locate military personnel in distress. Early in the development of GPS, its scope was expanded to include complementary civil capabilities such as civil, maritime, and land navigation. Space, ground control, and user equipment segments are needed to take full advantage of GPS capabilities. The GPS space segment consists of a constellation of satellites that broadcast encrypted military signals and civil signals. In recent years, because numerous satellites have lasted longer than anticipated, the constellation has grown well beyond the minimum requirement of 24 satellites to approximately 40 satellites of various generations, with 8 in residual status. The satellites are operated by a master control station that regularly updates navigation signals on the satellites. Using these navigation signals, GPS military and civilian user equipment receivers determine a user’s location. Figure 1 below illustrates how GPS satellites, ground control, and user equipment function together as an operational system. The GPS ground control segment primarily consists of software deployed at a master control station at Schriever Air Force Base, Colorado, and at an alternate master control station at Vandenberg Air Force Base, California. The ground control software is supported by 6 Air Force and 11 National Geospatial-Intelligence Agency monitoring stations located around the globe along with four ground antennas with uplink capabilities. Information from the monitoring stations is processed at the Master Control Station to determine satellite clock and orbit status. The GPS user equipment segment includes military and civilian GPS receivers. These receivers determine a user’s position and time by calculating the distance from four or more satellites using the navigation signals on the satellites to determine its location. Military GPS receivers are designed to utilize the encrypted military GPS signals that are only available to authorized users, including military and allied forces and some authorized civil agencies. Civilian—including commercial— receivers use the civil GPS signal, which is publicly available worldwide. In 2000, DOD began an effort to modernize the space, ground control, and user equipment segments of GPS to enhance the system’s performance, accuracy, and integrity. To that end, the Air Force is now in the process of developing a new generation of GPS III satellites, OCX, and M-code capable MGUE receivers. GPS III satellites are planned to supplement and eventually replace the constellation of GPS satellites now in orbit; these satellites consist of multiple versions or generations developed and launched over the years. The first GPS III satellite was originally expected to be available for launch in April 2014; however, due to development problems it is now expected to be ready for launch in May 2017. A complete GPS III satellite has not yet been tested, and the program is now rebaselining its cost estimates as a result of the schedule delay and associated increased costs. The GPS ground control segment is being modernized under the OCX program. OCX is required because the existing GPS control system, Operational Control System (OCS) cannot control—and therefore enable—the modernized features of the two latest generations of GPS satellites—IIR-M and IIF—currently in orbit. The Air Force plans to develop OCX in blocks, with each block delivering upgrades as they become available. Block 0 is intended to support the launch and initial testing of GPS III satellites; block 1 is designed to command and control the GPS II and III satellites and basic modernized signals; and block 2 is to enable the full modernized M-code signal and support, monitor, and control additional navigation signals. OCX block 1 is needed to incorporate GPS III satellites into the operational constellation to sustain required levels of GPS signal coverage, because the legacy OCS system cannot support the GPS III satellites. OCX is also required to enable military and civil use of modernized GPS signals. In particular, the military cannot use M-code signals for enhanced warfighting until OCX block 1 is delivered. The Air Force began the OCX program in 2007 with a technology development phase, referred to as phase A. The Air Force awarded phase A contracts, for approximately $160 million each, to Northrop Grumman in Redondo Beach, California, and Raytheon Intelligence and Information Systems in Aurora, Colorado to produce competitive prototypes. Following the competitive down-select between these firms’ prototypes, the Air Force awarded the development contract to Raytheon in February 2010. This $886 million contract covered the development of OCX blocks 1 and 2 (as discussed in this report, block 0 was added as a contract modification later), with an option to begin preliminary work on blocks 3 and 4 which are to provide additional capabilities to support follow-on, upgraded versions of GPS III satellites. At the formal start of development (milestone B), the Air Force estimated the total OCX acquisition costs—including other costs such as expenditures on technology development prior to 2010 and annual management support and enterprise integrator services—at $3.5 billion. According to Air Force documentation, enterprise integrator services are required to ensure GPS enterprise coordination among the ground, space, and user equipment segments. These costs do not include funding contributions from civilian agencies to support OCX; according to Air Force documentation, the Department of Transportation and other agencies are to provide resources to DOD to develop and operate GPS civil capabilities. For the purpose of this report, OCX development costs refer to the costs of the development contract with Raytheon. Military GPS receivers are also being modernized under the MGUE program. The Air Force was directed by the Assistant Secretary of Defense for Networks and Information Integration in August 2006 to develop M-code capable GPS receiver cards to meet military services’ needs. In January 2011, the Ike Skelton National Defense Authorization Act for Fiscal Year 2011 directed that DOD not obligate or expend funds to procure GPS user equipment after fiscal year 2017 unless that equipment is capable of receiving M-code. Figure 2 below shows an illustration of a MGUE receiver card. The modernized receiver cards are to provide U.S. forces with enhanced position, navigation, and time capabilities, while improving resistance to existing and emerging threats, such as jamming. The Air Force plans to develop MGUE in two increments. The first increment, now under way and expected to cost about $1.7 billion, utilizes three prime contractors. All three contractors—L-3 Interstate Electronics Corporation, Raytheon Space and Airborne Systems, and Rockwell Collins—are developing receiver cards for ground environments. Raytheon and Rockwell Collins are also developing combined aviation/maritime receiver cards for use in air and sea environments. The Air Force plans to build on the work conducted in Increment 1 to develop a more compact receiver card in Increment 2 that can be used when size, weight, and power need to be minimized and that can serve as an “engine” for future military GPS receivers. At this time, the Air Force is also exploring the possibility of using Increment 1 technologies to support munitions, handheld devices, and space applications that it previously anticipated supporting with Increment 2. The Air Force initially began development of M-code in fiscal year 2003. It then transitioned that work to the modernized user equipment (MUE) program in 2006. According to GPS program updates, MUE was troubled by issues such as underestimation of software complexity, longer than anticipated software development time, and more difficult than expected software and hardware integration. These issues, among others, resulted in significant cost and schedule growth. The MUE program yielded aviation and ground prototypes of M-code-capable components, cost about $498 million, and spanned about 7 years, ending in 2013. Concurrent with and based on MUE development work, the Air Force initiated the MGUE program in 2011. While DOD policy on GPS user equipment and procurement in 2006 indicated the Air Force was to develop MGUE to production-ready status, the Air Force currently plans to develop MGUE Increment 1 ground and aviation/maritime receiver cards to the point of production representative test articles. It then intends for the MGUE program office to provide funding to the military services so that they can acquire, integrate, and operationally test the receiver cards on service-specific “lead platforms.” These platforms are expected to serve as pathfinders for the military services’ ground, aviation, and maritime environments and are currently designated by the services as follows: Army–Defense Advanced GPS Receiver (DAGR) Distributed Device (D3) onboard a Stryker ground combat vehicle; Air Force–B-2 Spirit aircraft; Navy–DDG-51 Arleigh Burke Destroyer hosted by GPS-based Positioning, Navigation, and Timing Service (GPNTS); and Marine Corps–Joint Light Tactical Vehicle (JLTV). Once operational testing is complete, the military services will then be responsible for procuring MGUE Increment 1 for their weapons systems. DOD indicates that the military services anticipate procurement of about 207,000 MGUE Increment 1 receiver cards. Historically, full fielding of user equipment has lagged behind the anticipated availability of GPS satellite and ground systems. For example, in 2009 we reported that the Air Force expected GPS satellite and ground systems to be available in 2013 to transmit and process the M-code signal, but that fielding of M- code equipment to all designated military users would not be complete until 2025. OCX development contract costs have more than doubled since the contract was awarded in February 2010, increasing by approximately $1.1 billion to $1.98 billion, and the program’s schedule has roughly doubled over estimates at contract award. The Air Force awarded the contract to begin OCX development but did not follow key acquisition practices such as completing a preliminary design review before development start as called for by best practices. In addition, key requirements, particularly for cybersecurity, were not well understood by the Air Force and contractor at the time of contract award. The contractor, Raytheon, experienced significant software development challenges from the onset, but the Air Force consistently presented optimistic assessments of OCX progress to acquisition overseers. Figure 3 below shows select key events related to the OCX program since the development contract was awarded in 2010. Further, the Air Force complicated matters by accelerating OCX development to better synchronize it with the projected completion time lines of the GPS III satellite program, but this resulted in disruptions to the OCX development effort. As Raytheon continued to struggle developing OCX, the program office paused development in late 2013 to fix what it believed were the root causes of the development issues, and significantly increased the program’s cost and schedule estimates. However, progress reports to DOD acquisition leadership continued to be overly optimistic relative to the reality of OCX problems. OCX issues appear to be persistent and systemic, raising doubts whether all root causes have been adequately identified, let alone addressed, and whether realistic cost and schedule estimates have been developed. Figure 4 below shows how OCX costs have grown and the schedule delayed since contract award in February 2010. The Air Force awarded the OCX development contract to Raytheon in February 2010 for $886 million; blocks 1 and 2 were forecast for completion in August 2015 and March 2016, respectively. The development contract was awarded before completing a milestone B decision formally authorizing the start of development. In addition, the program did not complete a preliminary design review (PDR) as called for by best practices; the Air Force subsequently acknowledged in the OCX acquisition strategy approved in September 2012 that the contract was awarded earlier than normal. Figure 5 below depicts key events in DOD’s typical acquisition process and the corresponding knowledge called for in GAO’s prior work on best product development practices. As shown in the figure, our prior work has identified several proven management practices that, if fully implemented, can help DOD minimize cost overruns by ensuring programs are established after matching requirements and resources. The Air Force did not conduct a milestone B review prior to awarding the development contract, missing an opportunity to ensure the program began on a sound foundation. A milestone B review is an important point in a program where requirements and resources—that is, technology, design, time, and funding—should be properly matched to make sure the program can be executed as planned. The Under Secretary of Defense for Acquisition, Technology, and Logistics (USD AT&L) approved the Air Force’s request in January 2010 to award the OCX development contract prior to milestone B. The Air Force and USD AT&L decided to award the contract early for several reasons. First, the Air Force believed that the 3- year competitive technology development effort from 2007-2010, where two contractors (Northrop Grumman and Raytheon) worked on proof-of- concept prototypes for OCX, successfully demonstrated initial functionality and reduced risk. Second, the Air Force wanted to reduce the costs of carrying two contractors through PDR. Finally, the Air Force and USD AT&L believed that splitting efforts between two contractors was slowing progress on the program, and decided that down-selecting to one contractor prior to milestone B would help accelerate OCX completion. According to Raytheon officials, a contributing factor was the need to align OCX development with that of GPS III development, the development contract for which was awarded 21 months earlier, in May 2008. In addition, prior to development start, the program did not complete a PDR—which assesses the maturity of the preliminary design and confirms that the system is ready to proceed into detailed design with acceptable risk. As GAO’s best practices work has shown, PDR is a critical step at which customer needs are balanced with available resources. Programs that are launched prior to completing a preliminary design tend to experience more problems compared to programs where launch occurs after completing a preliminary design. Figure 6 below depicts how programs generally have better outcomes when they are launched after conducting relatively more detailed systems engineering: particularly, completing a preliminary design before formally launching. Programs launched after only completing a notional design—a general concept, unconstrained by resources, of what the product will look like and what it might be capable of—often experienced problems and needed more time or money than had been estimated at program launch. OCX program office and Raytheon officials stated that certain OCX requirements were not well understood at the time of contract award. In particular, Raytheon officials stated that the company did not understand the extent to which it would be required to implement Information Assurance (IA) requirements until as late as 2013, 3 years after beginning OCX development. The purpose of IA, also referred to as cybersecurity, is to ensure that DOD systems can resist and continue to operate during cyber-attacks by managing risks and implementing safeguards. OCX was required to be compliant with the DOD directive that prescribes IA policies, responsibilities, and procedures, among other things, for DOD information technology and defense components. In light of increasing cyber threats and given that satellite ground systems are the most vulnerable components to potential attacks, the Air Force plans for OCX to have an improved IA capability over the current GPS control system. According to program and Raytheon officials, OCX is going to be one of the first large-scale programs within DOD to fully implement IA requirements. In addition, Raytheon officials described cybersecurity threats as continuously evolving, and that both Raytheon and the Air Force have had to adapt their interpretation over time of how to meet IA requirements on OCX development to address changing threats. Program office and Raytheon officials noted that past DOD acquisition programs routinely waived some of the IA requirements in prior programs, and that Raytheon entered OCX development with the expectation that some IA requirements would be waived as in the past. According to Raytheon officials, at the time OCX development began, neither the Air Force nor Raytheon had the experience—and therefore, the knowledge— of developing such a complex IA-intensive program. Given the importance of GPS to the military and civil communities and with the increase in cybersecurity threats, the Air Force did not waive any IA requirements for OCX. Consequently, Raytheon found that it had greatly underestimated the cost and time to meet these requirements. According to program officials, most of the requirements issues were resolved in early 2015. The OCX program held PDR in August 2011—more than a year after contract award and after the Air Force had spent over $1 billion on the overall OCX program since 2007 including phase A technology development—with the Air Force affirming that the OCX architecture and design was solid and that the program was ready to begin formal system development. However, at PDR, Raytheon did not prepare—and the Air Force did not assess—a preliminary design for the entire OCX system because Raytheon followed an iterative software development process unlike the traditional, “waterfall” software development process. An iterative approach is a common industry practice that consists of developing software in a series of iterations and blocks, where developers go through multiple cycles of breaking down requirements, and designing, coding, and integrating software. By contrast, in the traditional waterfall approach the system is fully designed before coding and testing follow in a linear sequence. In theory, an iterative approach allows for a balanced and efficient use of resources when developing complex systems. Smaller iterative development cycles—as opposed to a lengthy, linear, waterfall development process—allow for capabilities to more easily be added incrementally, lessons learned to be incorporated, and early integration testing to be conducted to minimize cost and schedule risk. To achieve this, effective systems engineering—the process of deriving and allocating requirements for individual software iterations—is key. The Air Force, acknowledging Raytheon’s use of an iterative approach in late 2010, tailored its preliminary design review criteria—which originally were based on a traditional, waterfall approach—to only review the two iterations completed by Raytheon at that point. For the remaining six iterations, Raytheon had only completed the initial allocation of requirements at a high level. However, as Raytheon encountered problems in software development, it began deferring difficult work to later iterations. Raytheon was experiencing difficulties developing OCX, but the Air Force presented optimistic assessments of progress to USD AT&L. By October 2011, an independent OSD-chartered review team—one of a series of independent reviews—warned of severe software development problems, particularly with mounting deferred work, ineffective integration testing, and overestimated software productivity rates, and predicted a 15 to 19 month delay to block 1 delivery. According to an official with USD AT&L, these independent assessments are designed to identify program cost, schedule, and performance risks and provide feedback to the program manager and the milestone decision authority. The official further noted that the assessments were considered helpful, but essentially nonbinding advice from department experts to the program manager. The independent team followed up in February 2012 for a more detailed assessment of OCX and predicted a higher, 24-month slip to block 1. However, at the annual GPS enterprise review (AGER) with USD AT&L in April 2012, the Air Force acknowledged the software issues but projected a shorter, 11-month delay citing various corrective actions already taken, including greater oversight. This projection proved to be overly optimistic. In light of the challenges experienced to that point, at the April 2012 AGER, USD AT&L postponed the scheduled milestone B decision for OCX and directed the program to return for a formal Defense Acquisition Board (DAB) review within 120 days and to report on progress. In the midst of mounting OCX problems, the Air Force disrupted the ongoing software development effort. As it became apparent that OCX block 1 would not be ready in time to support the then-projected launch time frames of the first GPS III satellite (August 2015 and May 2014 respectively), USD AT&L directed the Air Force in January 2011 to separate out development of the satellite launch and initial testing portion of the block 1 software. Dubbed block 0, this subset of software is to contain the capabilities needed to launch and test the initial GPS III satellites. The Air Force modified the OCX contract in January 2012 to implement this change, which included bringing forward IA capabilities originally scheduled for later iterations. Block 0 is a temporary measure because the Air Force anticipated that some GPS III satellites would need to be launched before the entirety of block 1 was to be developed. Block 0 does not allow the GPS III satellites to be incorporated into the overall GPS constellation and used by the military or civil community—which requires block 1 implementation. Further, this decision to accelerate the GPS III launch and test capability was based on Air Force assessments of GPS III progress that appeared overly optimistic, as GAO found at the time. Nevertheless, according to Raytheon officials, the creation of block 0 and acceleration of some software capabilities caused it to have to revamp the OCX software development plan 2 years into development— not in accordance with best practices in software development, which call for stabilizing requirements and design prior to coding. Our prior work in this area shows that too many changes to requirements can result in additional, sometimes unmanageable risk. Meanwhile, Raytheon continued to struggle with developing OCX, specifically with implementing IA requirements. The contractor was overly optimistic in its initial estimates of the work associated with incorporating open source and reused software, and, according to the Air Force, did not appear to follow IA screening or software assurance processes as required, for example, incorporating open source software without ensuring that it was IA-compliant. The Air Force stated that it was not aware that Raytheon’s software assurance processes were not in compliance with the OCX software development plan until it performed an audit of Raytheon’s secure coding process in August 2012. According to Raytheon officials, however, the development contract did not specifically address the extent that open source software had to be scanned for IA compliance. Nevertheless, this led to significant rework and added cost to remediate the security vulnerabilities and meet IA standards. In addition, Raytheon’s systems engineering was incomplete, resulting in an inability to build code as planned and work being consistently deferred to later iterations. The Air Force noted that it became aware of these systems engineering issues early in development and took some corrective actions such as defining the completion of certain components; however, these actions were focused only on high-priority OCX components. Nevertheless, in July 2012, the Air Force presented a positive progress report at the mandated DAB review, stating essentially that all technical issues had been identified and mitigated, the design fully validated, and an executable and realistic schedule put in place after a thorough internal examination of the program. While the program stated that sufficient margin had been incorporated into the cost and schedule estimates, the program’s estimates for block 1 delivery—between February and October 2016—were still optimistic compared to the time frames (November 2016 to March 2017) forecast by the independent team in October 2011. The Air Force formally completed milestone B for OCX in November 2012— more than 2 years after contract award. At this point, the Air Force developed its first detailed cost estimate for the program as required by DOD policy, forecasting that OCX development would cost approximately $1.6 billion, which, according to the Air Force, was its first, formal realization of the magnitude to which the contract was initially underbid. According to the OCX program office and independent OSD-chartered reviews, Raytheon’s incomplete systems engineering led to continuous rework and deferred requirements to later iterations. In addition, the Air Force made significant changes to certain requirements, particularly with updating the specifications for OCX’s connections to other government systems, and M-code signal requirements. As a result, Raytheon officials estimated that nearly two-thirds of the requirements baseline as of PDR had changed by mid-2012. In March 2013, an independent OSD-led team praised the program for the corrective steps taken, but pointed to a rapidly deteriorating delivery schedule as a result of software development taking much longer than planned and a high defect rate, among other factors. The independent team projected block 1 completion between February and July 2018, a schedule slip of 16 to 21 months over the estimate at milestone B. DOD’s contract performance reviewer, Defense Contract Management Agency (DCMA), highlighted major technical difficulties in OCX development in its monthly analyses from the June to October 2013 time frame, including missed software iteration milestones, concurrent systems engineering and software development, and a high defect rate, among other issues. In addition, DCMA reported as early as August 2013 that the Air Force and Raytheon were going to soon begin an over target baseline (OTB) process where DOD determines that contract budgets are unrealistic and formally increases the program’s budget. An OTB is intended to allow for more realistic budget and work estimates and, therefore, more meaningful performance measurement against the updated budget. However, at the September 2013 AGER, the Air Force again presented an optimistic assessment of the program to USD AT&L, stating that OCX was on track to meet milestone B cost and schedule estimates, IA challenges had been identified, and key metrics had been established to ensure progress. The program projected just a 2-month delay to block 1 (to December 2016). Meanwhile, the Air Force noted that an excessive amount of rework was occurring at the time, and it directed Raytheon in November 2013 to pause development and complete a greater level of systems engineering for block 1. In December the Air Force and Raytheon began an OTB process to identify root causes and corrective actions and establish revised cost and schedule goals for OCX, and notified USD AT&L that block 1 delivery was likely to slip by approximately 9 months, to September 2017. The OTB concluded in June 2014, and identified root causes, many of which were similar to those identified at previous AGERs and independent assessments: incomplete systems engineering, inadequate process discipline, and IA implementation difficulties due to complexity. At this point, the Air Force’s estimate of the program’s development cost had grown to $1.7 billion—about 6 percent more than the milestone B estimate and nearly twice the initial contract estimate in 2010. The Air Force also added a total of 2 years to the program over its estimate at the September 2013 AGER with block 1 delivery moved to November 2018. Despite the ongoing problems with OCX development, the Air Force, faced with the statutory deadline to deliver M-code capable receivers— and consequently, OCX support for M-code—by fiscal year 2018, added some uncertainty to the OCX development plan. The Air Force agreed with Raytheon during the OTB to concurrently develop a portion of block 2 M-code capability during the block 1 effort; however, the Air Force noted in technical comments on a draft of this report that Raytheon is under no contractual obligation to do so. Complicating matters, the Air Force made significant updates to M-code signal requirements in 2014, contributing to additional OCX requirements changes. At the time of contract award in 2010, the Air Force had only developed a preliminary M-code specification, and it took time for the Air Force to fully mature M- code requirements and specify the functionality required of OCX. According to Raytheon officials, this had a significant effect on the IA requirements and design of certain components of OCX. The OCX program’s path forward following the conclusion of the OTB process in June 2014 depended on resolving certain problems in order to meet revised cost and schedule targets. In particular, the Air Force assumed that the contractor would (1) resolve software defects quickly, (2) bring greater discipline to software development processes, and (3) achieve higher software productivity rates than previously demonstrated. However, latest available data as of May 2015 shows that the contractor has not yet resolved these issues, increasing the risk that additional OCX cost and schedule growth is likely. A key expectation underpinning the post-OTB baseline was that Raytheon would be able to sharply reduce the number of new defects and resolve them within 6 months. However, the opposite took place. Raytheon uncovered defects at a faster pace than it could resolve them, causing further delays. By October 2014, just 4 months after the OTB, DCMA reported that the defect resolution rate was unsustainable, needing a continual increase in cost and schedule, with the trend likely to continue. In addition, DCMA found that the majority of defects were identified during the later phases of software development at which point it is harder and much more expensive to resolve than if found earlier. Latest available DCMA reports, as of May 2015, showed that Raytheon had yet to bring the defect rate within planned levels. Raytheon officials noted in early July 2015 that the defect discovery rate and backlog have been greatly reduced as OCX prepares to begin testing on iteration 1.5. However, data on the number of defects open and resolved show that the defect backlog as of July 2015 was still more than three times that predicted in February 2015. In addition, DCMA has pointed out that the test activities will likely identify new defects, potentially increasing the backlog further. DCMA further noted that the current defect backlog consists of more complex and difficult defects, which will require considerable effort to close compared to those resolved so far. The persistently high defect rate for OCX may be a result of as-yet unidentified systemic issues. An October 2014 independent review noted that, given the difficulty in resolving defects as planned, systemic issues may remain but that neither the OCX program office nor Raytheon had conducted detailed engineering assessments to determine if there were any systemic issues. As GAO’s prior work has shown, effective defect management requires a realistic schedule in that it takes time to be able to fully identify, analyze, prioritize, and track defects. Without investing the time and resources to conduct detailed engineering assessments, the program cannot know if any systemic issues are causing the persistently high software defect rate. The high defect rate is a symptom of continued struggles with the root cause of undisciplined processes at Raytheon, noted to-date by the OCX program office, independent OSD-chartered review teams, and DCMA. For example, Raytheon has had difficulty establishing consistent software development environments—the computer infrastructure including hardware, operating systems, and databases—across the OCX program. Developers built each of the environments with different hardware and operating system versions and settings. According to Raytheon officials, this was partly because the Air Force’s requirements called for multiple tailored environments to save on hardware costs, but that this drove complexity and posed technical difficulties to Raytheon. This meant that the contractor could not deploy a given software build onto all the environments without a lot of rework—and consequently, time and expense. Following the June 2014 OTB, the Air Force and Raytheon reported that they had taken steps to correct this issue, such as using automated tools to ensure consistency. However, latest available data, as of May 2015, showed that inconsistent environments were still contributing to a high defect rate and consequent cost growth and delays. Raytheon noted that it had greatly reduced the amount of time required to deploy software builds onto a given environment to 3 days as of July 2015; however, this was still well short of the 1-day limit called for by the October 2014 independent review. Another example of undisciplined processes at Raytheon is its peer review process. DCMA reported in a November 2014 audit that peer reviews—shown by our prior work to be a crucial quality assurance component of software development—were inconsistent and less effective than planned. The OCX software development plan identifies peer reviews as an integral part of the software development process, the main purpose of which is to identify and report defects in software artifacts—work products such as code, software class libraries, and design models—as early as possible in the development life cycle. Effective peer reviews depend, in part, on clear instructions, compliance, and standards, as well as consistent processes. DCMA reported Raytheon having problems with all of these items. In its audit, DCMA noted that the work instructions were often unclear, contradictory, or contained loopholes; many peer reviews were non-compliant with the work instructions; and a high percentage of reviews were held “virtually” where users review artifacts independently instead of during a formal, in- person meeting as presumed in the OCX software development plan. DCMA concluded that the quality of the peer review process was questionable because peer-reviewed software artifacts resulted in an abnormally high number of defects being discovered later. According to Raytheon officials, Raytheon has incorporated changes to its peer review instructions in its software development plan, which is awaiting Air Force approval as of July 2015 and will govern future software development. The Air Force and Raytheon also assumed overly optimistic software productivity rates—considering Raytheon’s track record on OCX—when developing the post-OTB schedule. For example, although the post-OTB baseline added nearly 2 years to the prior plan, both the Air Force and Raytheon assumed productivity rates for iterations 1.6 and 1.7 that were approximately two-thirds higher than the rate achieved for iteration 1.5, which was the most complex software effort undertaken to that point. As of the latest available data, from February 2015, Raytheon’s productivity rate had increased above the rate achieved for 1.5 but was roughly only half the forecast increase because of continued difficulties, increasing risk that OCX delivery will be further delayed. According to DCMA, the program has had a history of being overly optimistic with forecasting schedules, for example, using data from Raytheon’s prior experience on other programs that have consistently proven inaccurate when applied to OCX, and assuming efficiencies based on learning curves which have not materialized due to staff turnover, process changes, and poor retention of lessons learned from past development difficulties. This tendency of overestimating software productivity rates was noted as far back as October 2011 as part of an independent OSD-led assessment that highlighted several OCX development problems. Raytheon noted that corrective actions were taken as part of the OTB, such as revalidating and completing systems engineering, establishing a common environment infrastructure, and validating IA implementation during block 0 tests will result in software productivity improvement for iterations 1.6 and beyond beginning in 2016. USD AT&L may not have adequate insight into the full extent of OCX development problems given the Air Force’s consistently optimistic assessments of the program’s progress. The October 2014 advisory independent review estimated that block 1 was likely to take approximately 2 years longer than the June 2014 OTB estimate, putting probable block 1 delivery around November 2020. DOT&E also expressed concerns about OCX delays in a memorandum sent to USD AT&L in early November 2014, citing the negative effect of those delays on the Air Force’s ability to deliver overall GPS capability. In addition, DOT&E called the Air Force’s schedules for the overall GPS enterprise, including that of OCX, “inaccurate, implausible, and incoherent” given OCX development difficulties to date. Nevertheless, at the ensuing fiscal year 2015 AGER that was held in November 2014, the Air Force acknowledged additional development difficulties but noted that OCX systems engineering was improving and would lead to better software development in the future. At that time, the OCX program office forecast block 1 completion in August 2019—a delay of 9 months over the OTB estimate, but optimistic compared to the independent review’s assessment. One month later, in December 2014, the Air Force presented its official update for the program and forecast a block 1 completion date of July 2019. It also estimated that contract costs had grown to nearly $2 billion—representing cost growth of 16 percent in just 5 months over the OTB estimate, and more than 120 percent over initial contract estimates in 2010. Following the November 2014 AGER, in December 2014, USD AT&L expressed concern about the continued deterioration in the cost performance of OCX and stated that the program’s trajectory must be corrected. USD AT&L directed the Air Force to provide by January 2015 a “deep dive” program review of OCX focused on the status of and ways to improve program execution. At this review, which occurred in February 2015, the Air Force acknowledged the program’s volatile cost and schedule history and that the contractor’s schedules were aggressive relative to the risk and amount of work remaining. However, the Air Force also highlighted signs that program execution was stabilizing and noted that it had reduced program risk to medium-low by including an additional $331 million and 3 months above contractor estimates. To address USD AT&L’s directive to detail OCX’s path forward and key decision points should OCX continue to sustain cost and schedule growth, the Air Force established a process whereby it would closely monitor key software development events roughly every 6 months beginning in July 2015 and report any deviations from cost performance to the Air Force’s senior acquisition executive as well as to USD AT&L as needed. While the closer monitoring of key software development events provides USD AT&L with opportunities to spot cost growth on a more timely basis than at the annual AGERs, there is little reason to believe that OCX systemic problems have been adequately addressed. Notwithstanding the Air Force’s optimistic report to USD AT&L in February 2015, additional OCX cost growth is quite likely. First, our analysis of detailed earned value data from November 2013 through November 2014 showed that the program office significantly underestimated the anticipated cost of resolving the risks that could affect OCX development and may have also significantly underestimated the extent of risks. Second, DCMA’s reporting of earned value performance for the program shows a sustained deterioration in program cost and schedule performance immediately following the OTB—where cost and schedule variances were reset to zero in order to begin measuring performance against the new baseline. Cost and schedule variances measure the differences in expectations between the value of work accomplished in a given period with the value of the work expected in that period. Negative variances indicate that the program is either overrunning cost or performing less work than planned; conversely, positive variances indicate the program is either underrunning cost or performing more work than planned. Figure 7 below shows the cumulative cost and schedule variances from October 2013 through last available data as of May 2015, and how the OCX program has been both overrunning cost and performing less work than planned immediately after the OTB concluded. While the OTB process completed in June 2014, the program formally reset the cost and schedule baselines at the end of July. Further, DCMA’s June 2015 analysis forecast that OCX costs were likely to increase to $2.15 billion based upon, among other things, the higher than expected defect rate and poor comprehension of requirements by Raytheon. From November 2014 through last available data as of May 2015, Raytheon depended on shifting between approximately 100 and 180 additional staff than planned to resolve defects, slowing work on later iterations. According to DCMA, Raytheon will likely need even more staff in the future to maintain schedule, and, consequently, incur additional cost growth. The Air Force’s current schedule estimates, forecasting block 1 completion in July 2019, are still optimistic by at least a year compared to the October 2014 independent team’s assessment that block 1 will most likely be delivered in November 2020. Figure 8 below summarizes the differences in Air Force estimates for the total number of months to complete block 1 and those predicted by independent reviews. In addition, our prior work on space acquisitions has shown that unrealistic estimates of the achievability of planned schedules, among other things, directly contributed to unrealistic cost estimates and, consequently, distorted management decisions, increased risks to mission success, and virtually guaranteed program delays. By contrast, senior leaders in successful organizations in the commercial sector actively encouraged program managers to share bad news about their programs. These organizations took pains to ensure program estimates are complete and accurate. Based on the persistently high software defect rate, continued undisciplined processes, lower than expected software productivity rates, the downward trend in cost and schedule performance following the OTB, and lack of realism in Air Force depictions of OCX progress, root causes do not appear to have been adequately addressed or perhaps even fully identified. Until all root causes of OCX problems are fixed, the program is likely to continue to struggle to achieve desired outcomes. The Air Force has revised the MGUE acquisition strategy several times in attempts to develop M-code capability. Even so, the military services are unlikely to have sufficient knowledge about MGUE design and performance to make informed procurement decisions starting in fiscal year 2018 because it is uncertain whether an important design review will be conducted prior to that time and because operational testing will still be under way. The Air Force has revised the MGUE acquisition strategy several times as it pursued the program’s development. The latest strategy of record eliminated a key design assessment, the critical design review (CDR), and it is uncertain whether a recent revision to that strategy will include a CDR. As a result, the military services may face a knowledge gap about MGUE’s design stability and maturity. As we have noted in our prior work, positive acquisition outcomes typically require the use of a knowledge- based approach to product development that demonstrates high levels of knowledge before significant commitments are made. We determined that, in keeping with that approach, knowledge gained through a CDR helps supplant risk over time by ensuring that a product’s design will meet customer requirements as well as cost, schedule, and reliability targets. As part of that work, we also found that a knowledge deficit early in a program can cascade through design and production, leaving decision makers with less knowledge to support decisions about when and how best to move into subsequent acquisition phases that commit more budgetary resources. Additionally, we found that demonstrating a stable and mature design, typically via a CDR, is generally considered a prerequisite by leading commercial firms and successful DOD programs for moving forward with a program because it assesses final product design and provides assurance that product specifications have been captured in detailed design documentation. The 2011 preliminary strategy for MGUE called for development of separate ground, air, and sea receiver cards within a traditional DOD three-phase acquisition process that encompassed separate technology development, system development, and production phases. The program’s 2012 baseline strategy was similar and rolled development of air and sea cards together based on the Navy’s assessment that aviation card development could support the maritime environment. In 2014, the Air Force received approval from USD AT&L to revise the MGUE acquisition strategy again by bypassing a formal system development phase and combining the development and production decision points. USD AT&L approved the associated Air Force acquisition strategy document in April 2015, but just two months later an official from AT&L indicated that—based on a June 2015 memorandum from the Assistant Secretary of Defense for Acquisition—those milestones would be split apart again. According to the AT&L official, it was determined to be impractical to combine the milestones because there were many events that needed to be conducted between them. The official indicated that USD AT&L plans to oversee the program until the development decision point, at which time it would then delegate program oversight to the Air Force. The last MGUE acquisition strategy of record eliminated a key assessment of MGUE design maturity, the CDR, which we have found in our prior best practices work is typically held mid-way through a program’s development phase and is essentially an assessment of whether a product’s design meets the customer’s requirements. The Air Force stated that a CDR was unnecessary because, among other things, detailed design work normally approved at CDR was completed for MGUE’s PDR in September 2014 and hardware and software designs and cost, schedule, and performance risks typically reviewed at CDR were assessed as part of that PDR. It also anticipated other events would cover and go beyond the purpose of a CDR, including a compatibility certification process to ensure MGUE receiver cards were compatible with the GPS satellite signal; security validation to ensure contractors’ designs met security certification criteria; initial and final verification reviews to assess product maturity and deficiencies; and utilizing multiple contractors with multiple designs from which the military services can choose to meet their requirements. However, the MGUE program carries design-related risks that would typically be revisited within the CDR context. An April 2015 assessment of the MGUE PDR by the Office of the Undersecretary of Defense for System Engineering noted, for example, that some security and information assurance design details were not addressed as part of PDR and had been deferred to the security verification review scheduled to finish in late summer/early fall 2015. The Undersecretary’s office also pointed out that the refinement of security countermeasures may result in later design changes. In addition, the office stated that MGUE PDR interface designs with the military services’ lead platforms may not be rigorous enough to account for implementing those designs across various operating environments. Moreover, it is unclear whether MGUE designs meet the military services’ requirements. According to Army program officials, the Army has identified a set of performance gaps for ground and aviation receiver cards between the MGUE program’s technical requirements document and Army operational requirements, including power and thermal incompatibility issues between the ground receiver card and the platform the cards are to support. The MGUE program said that it is unaware of any data showing the ground cards do not meet power and thermal requirements; analysis shows the cards meet service requirements for power as captured in MGUE’s capability development document; the power and thermal issues may be caused by the host platforms rather than the ground cards; and as such, the lead platform program offices would need to make modification to their power supplies or platforms to resolve the issues. Army officials, in contrast, said that the MGUE capability development document does not contain specific power consumption limits and that the Air Force’s GPS Directorate had recently changed power limitations by unilaterally editing the MGUE technical requirements document, against significant objection by the Army. A DOT&E official explained to us that the Air Force believes it has addressed this issue by clarifying the specification to be an average maximum power requirement for MGUE ground cards but also emphasized that this clarification relieves the MGUE contractors of meeting more stringent instantaneous power limits. As a result, the official said, there are potential adverse effects to host platform interface designs which may then require additional development and integration by the military services in order to adapt their platforms to MGUE. The military services are unlikely to have the knowledge to begin informed MGUE procurement at the start of fiscal year 2018 because operational testing that will provide knowledge about MGUE’s operational effectiveness and suitability for ground, air, and sea environments will not yet have been completed. Prior to operational testing, the Air Force is conducting activities and demonstrations that it believes reduce program risk by providing knowledge about MGUE performance. Those efforts, however, have limitations which call into question their value in assessing MGUE’s ability to meet the military services’ needs. Based on the MGUE program’s April 2015 acquisition strategy document, as shown in Figure 9 below, formal integration and operational testing for the MGUE aviation receiver card on the Air Force’s B-2 aircraft abuts fiscal year 2018; that same effort for the Army’s DAGR D3/Stryker extends into fiscal year 2018; and it carries into fiscal year 2019 on the GPNTS/Navy’s DDG-51 Arleigh Burke-class destroyer and Marine Corps’ JLTV. In accordance with DOD guidance, these operational tests mark an important point in the assessment of MGUE design and performance— the first time production representative test articles are planned to be incorporated on the lead platforms and tested in a realistic environment. Among other things, per DOD guidance on test and evaluation, the operational tests are expected to serve as field tests that assess the receiver cards’ ability to satisfy the military services’ requirements; current capabilities, including operational benefits or burdens; and the need for further development of the cards to correct performance deficiencies. The Air Force believes that completion of operational testing is not required for the military services to begin informed procurement planning because data from prior test events can be leveraged. It stated that the MGUE program has participated in various testing and integration risk reduction activities prior to the time lines shown in figure 9, and that those efforts can count towards MGUE integration. Those activities include, according to the Air Force, successful demonstration of technologies on the Army’s Raven unmanned aerial vehicle; demonstration of MGUE receiver cards in operationally relevant environments on a surrogate C-12 aircraft flying similarly enough to the B-2 lead platform to serve as a surrogate; and lead platform-based risk reduction events via prototype GPS units that will be incorporated into the B-2. The Air Force said that testing on surrogate platforms is intended as early risk reduction activity rather than emulation of lead platform environments. Additionally, it stated that it plans to conduct subsystem-level integration on lead platforms prior to system-level integration, and that these efforts also would provide the military services with further information to asses MGUE performance. Some of the military services, however, believe that the integration activities to date have been more limited than indicated by the Air Force and that they are now responsible for more of the MGUE development effort than indicated earlier in DOD policy. Army officials stated, for example, that the MGUE program began initial contact with the DAGR D3 program office in fiscal year 2014, but that integration with the DAGR D3/Stryker per se has not yet started due to lack of MGUE devices for platform integration. Instead, they said, “fit check” tests were conducted in early 2015 and showed the DAGR D3 is unable to provide sufficient power to two of the three MGUE contractors’ ground receiver cards. The officials also said that the Army planned to perform thermal testing with the DAGR D3/Stryker, but such testing has not yet been executed due to lack of MGUE devices. They emphasized that they have begun planning for MGUE procurement but cannot initiate actual procurement at the start of fiscal year 2018 for several reasons, including the need to modify as- built MGUE devices to close performance gaps, resolve power and thermal issues by either modifying one MGUE card design or the integration on some 100 platforms; and developing and implementing a means for MGUE and the currently-fielded Selective Availability Anti- Spoofing Module GPS receiver to coexist and interact in the field. In addition, both Army and Navy officials stated that, based on their understanding of DOD’s 2006 policy on GPS user equipment and procurement, the Air Force was to develop MGUE to production-ready status. They noted that, in order to bring MGUE to true production-ready status, their respective military services will need to do further development after the Air Force completes its MGUE development effort. Moreover, the demonstrations and risk reduction activities that the Air Force points to may not establish MGUE performance to the extent that the Air Force believes. For example, DOT&E—the DOD independent test authority charged with ensuring a program’s operational testing confirms operational effectiveness and suitability in combat use—expressed serious concerns about the Air Force’s characterization of the results from one of the past demonstrations. In a November 2014 memorandum to USD AT&L, DOT&E emphasized that the Air Force had overstated MGUE development maturity and that the demonstrations to-date had achieved more mixed results than the Air Force indicated. DOT&E noted, for example, that the testing with the Army’s Raven did not use final development models; had employed a limited subset of realistic threats; successfully flew its intended flight pattern in just 7 out of 51 attempts; and that none of those flights was conducted in the presence of electronic jamming. It went on to say that developmental activity with the Raven should in no way be construed as having the level of rigor of operational testing or even formal developmental testing. Furthermore, Army program officials stated that the Army’s lead platform had been changed from the Raven to the DAGR D3/Stryker not only because the Army does not intend to procure additional Ravens but also because the Raven does not adequately stress the capabilities of the ground receiver card. In addition, DOT&E concluded in its 2014 memorandum that MGUE cannot be considered effective until it is successfully integrated on host platforms. According to a DOT&E official, subsystem integration does not guarantee proper integration into a larger system and may not necessarily demonstrate integration and performance in the intended operational environment—particularly in light of the fact that the MGUE test strategy is inherently risky in that only 4 lead platforms are designated to represent the operating conditions for over 100 platforms service-wide. He also emphasized—as did DOT&E’s 2014 memorandum —that DOD has found integration historically challenging. Our prior work on integration and testing has also shown that integration is a common risk in system development, and that commercial firms reduce risk by capturing design and manufacturing knowledge early. Our previous work pointed out, for example, that systems integration problems can occur even though the various components performed successfully on previous systems. We have found, in illustration, that the Air Force’s C-17 aircraft program intended to use current, available, and proven technology to minimize development costs and structure a low technical risk effort but that the integration of technologies was a major engineering and management task that eventually contributed to significant cost increases. In addition, we have also determined that it is during integration and testing that problems are likely to be found. Furthermore, as we pointed out in a 2000 study of best acquisition practices for testing among commercial firms, test weaknesses invariably cause negative program outcomes, such as cost increases, schedule delays, or performance shortfalls. Our work on best commercial acquisition practices also shows that such negative outcomes can be avoided by accumulating knowledge prior to beginning production. In a 2002 study of best commercial practices, for example, we found that leading commercial firms reduce program risk by demonstrating fully- integrated prototypes prior to making production decisions. Such demonstrations, we determined, help the firms decide when to make the transition from product development to production and to ensure that transition is smooth. The military services can postpone procuring MGUE if they lack sufficient knowledge about MGUE design and performance after fiscal year 2017. They can, as provided by the National Defense Authorization Act for Fiscal Year 2011, request a waiver of such procurement. Specifically, the Secretary of Defense may waive this limitation upon a determination that suitable M-code capable user equipment is either not available or DOD does not require that user equipment be capable of receiving M-code from GPS. An official from USD AT&L stated that the waiver process has not yet been established. In 2010, we examined GPS satellite reliability data and reported that a delay of 2 or more years to the launches of GPS III satellites—originally projected to begin launching in April 2014—would likely reduce the constellation size below the minimum requirement of 24 operational GPS satellites by fiscal year 2018. Since that report, while the GPS III and OCX programs have been delayed, the GPS constellation has proven to be more reliable than previously expected because the Air Force successfully extended the life of existing satellites, primarily by modifying the satellites’ battery charging procedures. However, the Air Force now needs OCX block 1 to be operational by late 2019 to launch and incorporate GPS IIIs into the constellation. Given the ongoing development problems with OCX, delivery by that date—and consequently, overall constellation reliability—is at some risk. The Air Force is exploring contingency plans, including modifying the current ground system to launch and operate GPS IIIs. However, the modified ground system would operate GPS III satellites as current legacy GPS satellites, thereby not utilizing all the capabilities on the new satellites. Moreover, this approach would increase the risk that several IIIs may be launched before they can be fully tested with OCX. Delays to OCX delivery also mean that M-code functionality—which requires OCX block 1 to be fully enabled—is not scheduled to be deployed until 2019 at the earliest, and likely for another decade or more until the military services can widely deploy modernized MGUE receivers. The performance standards for (1) the standard positioning service provided to civil and commercial GPS users, and (2) the precise positioning service provided to military GPS users, commit the U.S. government to at least a 95 percent probability of maintaining a constellation of 24 operational GPS satellites. In September 2010, we reported that delays to the launch of the GPS III satellites—originally projected to begin in April 2014—could affect the long-term probability of maintaining the minimum constellation size. Based on satellite reliability data provided by the Air Force, we had predicted that, for example, a 2- year delay in the production and launch of the first and all subsequent GPS III satellites would reduce the guaranteed size of the constellation (at the 95 percent confidence level) to about 18 satellites by around fiscal year 2018. Since our 2010 report, the launch availability date for the first GPS III satellite has, in fact, slipped by about 3 years, to May 2017, mainly due to development difficulties with its navigation payload. Moreover, as discussed earlier, the delivery of OCX block 1—required to operate the IIIs—has been delayed by about 4 years, to mid-2019. However, the 3- year delay in GPS III’s delivery has not resulted in a predicted significant reduction in the size of the GPS constellation largely because the expected life of the existing generations of GPS satellites has risen dramatically. Figure 10 below displays the probability of maintaining a 24- satellite GPS constellation as a function of time based on March 2014 satellite reliability data and launch schedules—the latest date for which a complete set of approved parameters is available. The launch schedule has been adjusted to reflect the successful launches of four GPS IIF satellites through June 2015. The expected lifetimes of the GPS IIR and IIR-M satellites—the types of satellites that currently comprise the majority of the current constellation—have risen dramatically since our 2010 report, from 14.8 and 10.5 years, respectively, to 20.4 and 17.5 years, respectively. This increase in expected lifetime is due mainly to the Air Force’s implementation of a modified charging procedure that will prolong the life of the satellites’ batteries. Spacecraft batteries—like all other rechargeable batteries—are subject to cycle life, that is, they can only be charged and discharged so many times before they will not charge efficiently anymore. Because satellite batteries were predicted to be one of the primary life-limiting components of these satellites, the Air Force developed and tested a procedure in 2012 to reduce the battery charge rates during certain times of the year. The Air Force estimated that this new procedure would increase the expected lifetimes of the IIR and IIR-M satellites by 1 to 2 years. Based on recent telemetry data, the new procedure has proved successful. In 2014, the Air Force announced that this modified battery charging procedure has added, in aggregate, 27 additional years of operational life to the GPS IIR and IIR-M satellites. Moreover, according to Air Force representatives, additional information suggests that this estimate is probably conservative, and that the next formal review of satellite reliability data will likely reveal an additional increase in the expected life of the IIR satellites. Because these satellites form the mainstay of the current GPS constellation, and because most of them are still expected to be operational in 2017, when the first GPS III satellite is planned to be available for launch, an increase in the expected life of these older satellites significantly improves the Air Force’s ability to meet the 24-satellite performance standard despite the 3-year slip in GPS III and 4-year slip in OCX. The expected lifetimes of the GPS IIA satellites—the oldest version of GPS satellites on orbit—have also risen since our 2010 report, from 16.5 years to 19.7 years, due to positive trends in on orbit performance. However, the effect of this increase on the Air Force’s ability to meet the 24-satellite performance standards is less significant than the effect of the IIR/IIR-M life expectancy increase. The current constellation availability analysis assumes each of these satellites is only available to replace other satellites in the unlikely event of a large number of satellite failures until OCX is operational. Because OCX is not expected to be capable of operating the IIA satellites, the analysis assumes that any remaining IIA satellites would be turned off by July 2019, the date that OCX is scheduled to be required to be ready to transition to operations. Despite the delays to GPS III and OCX, the GPS constellation is expected to meet the constellation size requirement if OCX block 1 is available by July 2019 as needed. However, further OCX delays could affect overall constellation availability, bringing it below the 95-percent performance standard for maintaining a 24-satellite constellation for a year or more. For example, if OCX block 1 is not ready to transition to operations until November 2020—as the most recent (October 2014) independent review of OCX estimated—the Air Force could experience a 15-month period during which it would not meet its commitment to maintaining a 24-satellite constellation with a 95 percent confidence level. While GPS III satellites could still be launched beginning in 2017, they would not be added to the constellation until OCX comes on line or until a contingency operational control capability able to command GPS III satellites is available. Figure 11 depicts the predicted constellation size based on OCX block 1 being potentially delayed until November 2020. GPS capabilities are unbalanced, with satellite capabilities outstripping those supported by the ground and user equipment segments. In particular, the Air Force has launched two generations of satellites—IIR-M and IIF—capabilities of which are still not fully utilized because they have been transmitting three signals, including M-code, that are not supported by the current GPS ground control system, OCS. GPS IIFs are nearing the end of their launch, with the last three scheduled to be launched between July 2015 and February 2016. GPS IIIs are needed to continue to sustain the constellation. Since OCX block 1 is needed to operate the GPS III satellites as part of the constellation, the Air Force is preparing contingency plans in the event that OCX block 1 is delayed beyond August 2019, the need date for the first GPS III satellite. The Air Force has pinpointed one option as the most viable— modifying the current OCS system to operate the GPS III satellites but at the older, IIF satellite level of functionality; this option would deliver M- code support without all modernized GPS III functionality, specifically support for the L1C signal. Under this plan, the Air Force would use OCX block 0 to launch and initially test the GPS III satellites, and use the modified OCS to control the satellites’ navigation signals as part of the overall constellation. As figure 12 shows, successfully implementing this contingency plan should enable the Air Force to meet its commitment to maintaining a 24-satellite constellation with a 95 percent confidence level, even if the OCX block 1 delivery date were to slip to November 2020 (or later), but the plan would require funding and other resources. The Air Force estimated the cost of this option at $105 million as of February 2015; however, pursuing this option means that the Air Force may launch up to 5 GPS III satellites without fully testing them with OCX block 1, increasing the risk that issues may be found during testing without the ability to fix the satellites already launched. M-code initial operational capability is defined as having 18 M-code capable satellites on orbit, the control segment able to command/upload M-code capabilities to the space segment, and MGUE receivers fielded across the military services to utilize M-code capabilities operationally. Full operational capability occurs when 24 M-code satellites are on orbit, and a larger portion of MGUE is fielded. The GPS constellation will likely include 18 M-code capable satellites by September 2015, but M-code capabilities will not be available to users— assuming MGUE receivers can be fielded—until OCX delivery in mid- 2019. The launching of M-code capable satellites has been stretched over a much longer period than originally envisioned because of delays to the IIF, III, OCX, and MGUE programs. However, because the oldest of the M-code capable satellites are lasting longer than originally predicted, the effect of this stretched deployment has been mitigated. In particular the oldest M-code capable satellite—the first IIR-M satellite, launched in September 2005—has a better than a 75 probability of still being operational in October 2020, and better than a 50 percent probability of still being operational in October 2023. Nevertheless, DOD cannot take full advantage of M-code capability until MGUE receivers are deployed in sufficient numbers across the services. As noted earlier, the earliest MGUE operational test will not be completed until the end of fiscal year 2017 and the latest at the end of fiscal year 2019. Accordingly, some of the military services will not be able to make informed fielding decisions until fiscal year 2020 at the earliest, assuming operational testing goes as planned. Furthermore, as discussed earlier, the Air Force plans to deliver production representative test articles, which the military services will then acquire, integrate, and operationally test only on selected military platforms. As GAO reported in 2009, each of the military services would still need to add the new user equipment to other platforms, which could take an additional 10 or more years based on the need to perform such activities as coordinating installation with the platforms’ maintenance and upgrade schedules. GPS is a global utility that is integral to U.S. national security and civilian use. As a result, any decision about GPS has far-reaching consequences. A combination of many factors, including technical challenges, poor contractor execution and program management, and ineffective acquisition oversight have all put GPS modernization at significant risk. OCX is the key to enabling the full GPS capability that both the military and civilian communities depend on, from command and control of the satellite constellation including the new GPS III satellites, to allowing military receivers to take advantage of M-code signals for more robust warfighting capability, to enabling advanced civil GPS signals. However, by any measure, OCX development has been mired in development difficulties resulting in steady cost growth and schedule delays. Moreover, despite a 7 month pause ending in mid-2014, OCX has yet to turn the corner on resolving the problems that have affected the program since development began in 2010. Independent observers such as OSD-led teams, DCMA, and DOT&E have raised red flags about the effectiveness of the Air Force’s oversight and Raytheon’s ability to deliver promised outcomes. The Air Force has compounded matters by consistently presenting overly optimistic assessments about OCX progress— demonstrating a pattern of marginalizing warnings about OCX delays presented by independent assessors, likely because their assessments are considered advisory in nature. Driving towards unrealistic timeframes is compounding the program’s inability to meet stated cost and schedule goals. Five years into what was originally estimated to be a 5-year effort, OCX is still roughly 5 years away from completion. Without comprehensively identifying systemic causes for OCX problems, DOD cannot have high-confidence cost and schedule estimates for OCX. The Air Force and Raytheon have noted that OTB corrective actions will begin paying off in early 2016 as block 1 development resumes in earnest. But given the lack of success in prior attempts to turn the program around, the Air Force could benefit from external expertise and guidance on what is necessary to address systemic issues. Since the enactment of the statute directing DOD to generally procure only M-code capable user equipment after fiscal year 2017, the Air Force has struggled to deliver a MGUE acquisition strategy that would allow the military services to comply with that direction. As it stands now, the Air Force essentially truncates its MGUE development work and hands the result of its efforts off to the military services to continue development on their respective platforms. At the point the Air Force provides funds to the military services for them to acquire and test MGUE, it also transfers the onus of the development work to the military services’ shoulders because MGUE development cannot be considered complete until the cards work in the lead platforms. Unfortunately, the military services will absorb an added development burden because the Air Force plans to transfer its work without conducting design and performance assessments that could help the services decide the extent to which the cards are ready to integrate and test and how much additional work they will need to do before beginning to procure the cards. Fortunately, many GPS satellites now on orbit have served the nation particularly well by working considerably longer than expected. Even so, it cannot be presumed that they will continue to do so. Therefore, it is critical that the modernized GPS III, OCX, and MGUE development efforts succeed sooner rather than later. Of those three programs, OCX is now the pacing item for modernization due to its many past delays and probable future delays. Until the OCX program trajectory is corrected, those delays are likely to pose significant risks to sustaining the GPS constellation, and consequently, delivering GPS capability to the military community. To better position DOD as it continues pursuing GPS modernization, to have the information necessary to make decisions on how best to improve that modernization, and to mitigate risks to sustaining the GPS constellation, we recommend that the Secretary of Defense take the following five actions: Convene an independent task force comprising experts from other military services and defense agencies with substantial knowledge and expertise to provide an assessment to USD AT&L of the OCX program and concrete guidance for addressing the OCX program’s underlying problems, particularly including: A detailed engineering assessment of OCX defects to determine the systemic root causes of the defects; Whether the contractor’s software development procedures and practices match the levels described in the OCX systems engineering and software development plans; and Whether the contractor is capable of executing the program as currently resourced and structured. Develop high confidence OCX cost and schedule estimates based on actual track record for productivity and learning curves. Direct the Air Force to retain experts from the independent task force as a management advisory team to assist the OCX program office in conducting regular systemic analysis of defects and to help ensure OCX corrective measures are implemented successfully and sustained. Put in place a mechanism for ensuring that the knowledge gained from the OCX assessment is used to determine whether further programmatic changes are needed to strengthen oversight. To allow the military services to fully assess the maturity of the MGUE design before committing test and procurement resources, incorporate a CDR in the Air Force’s MGUE development effort. We provided a draft of this report to the Department of Defense for review and comment. In its written comments, reproduced in appendix II, DOD concurred with the four OCX-related recommendations calling for a more robust independent review of the program, stating that the intent of those recommendations has been met by OSD-chartered independent reviews conducted to-date, which included both Air Force and OSD staff, and other DOD activities governing OCX oversight, such as the use of independent cost estimates. DOD partially concurred with the recommendation calling for incorporating a CDR in the MGUE program’s development effort, noting that adding a CDR would delay the program amidst its efforts to compress the acquisition process to deliver MGUE capability by fiscal year 2018. DOD also provided technical comments, which we have incorporated as appropriate. While DOD concurred with our recommendations to put OCX on a better path forward, it responded that it had essentially been following the intent of our recommendations all along, requiring no further action from the department. These comments provide little confidence that the department intends to fully implement our recommendations to fix the development problems that have beset the program since its inception. As we noted in our report, the independent reviews conducted to-date have been nonbinding and advisory in nature. Even as the OSD-led reviews have warned of significant delays and inadequate insight into systemic root causes of development problems, the Air Force has repeatedly developed unrealistic timeframes, ostensibly with OSD’s knowledge. If business continues as usual without swift and thoughtful action, OCX will likely continue on its path of demonstrating poor cost and schedule outcomes. We continue to stand by our recommendations calling for a fresh review—this time an in-depth and comprehensive critical review of the program—to identify the true root causes of OCX development difficulties and to ensure the Air Force implements the corrective actions. Regarding DOD’s response to our recommendation that DOD hold a CDR for MGUE, we note that DOD’s rationale for skipping this best practice step is based on a desire to accelerate fielding of the units. However, our past work has consistently shown that taking shortcuts and skipping important knowledge points in the acquisition process generally results in an inability to deliver promised cost and schedule outcomes. To minimize any potential disruption to development efforts, a CDR could be held after production-representative test articles are delivered. The results of a CDR will show whether the MGUE design is stable. Rushing into lead platform testing without an approved and stable design means that DOD will have to concurrently test and correct the receivers’ design at a point where resolving issues are typically more expensive than earlier in development—not in accordance with best practices that are important to achieving cost and schedule outcomes with a high degree of confidence. We are sending copies of this report to the appropriate congressional committees, the Secretary of the Defense, the Secretary of the Air Force, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-4841 or by email at chaplainc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix III. In May 2014 House Report No. 113-446, accompanying H.R. 4435, the Howard P. “Buck” McKeon National Defense Authorization Act for Fiscal Year 2015, noted the requirements stated in section 913 of the Ike Skelton National Defense Authorization Act for Fiscal Year 2011 (Public Law No. 111–383) requiring the Department of Defense (DOD) to purchase M-code capable user equipment during the fiscal years after fiscal year 2017 and included a provision for us to report on DOD’s progress in deploying M-code capability. The House report also provided that our report assess current and planned investments; whether key milestones are being met; the projected ability to the meet the requirements in section 913 of Public Law No. 111–383; and an identification of the challenges that Global Positioning System (GPS) faces and possible recommendations on how to make the program more successful in delivering M-code capabilities. Additionally, in June 2014, Senate Report No. 113-176, accompanying the Carl Levin National Defense Authorization Act for Fiscal Year 2015, included a provision for us to review the cost, scope, and schedule of the GPS III operational control system, including synchronization with the launch of the GPS III constellation with recommendations for improvement. In response, for this report our objectives were to assess (1) the extent to which DOD is meeting cost, schedule, and performance requirements for next generation operational control system (OCX), (2) the progress DOD has made in delivering M-code capable military GPS user equipment (MGUE) by the end of fiscal year 2017, and (3) the challenges DOD faces in synchronizing the development of GPS III satellites, OCX, and MGUE to deploy M-code. To assess the extent to which DOD is meeting cost, schedule, and performance requirements for OCX, we reviewed program and contractor cost and schedule documentation, including program acquisition baselines, earned value metrics, and test plans. Throughout this report, we focused on the costs of the OCX development contract instead of the full program acquisition costs because the latter includes prior, technology development expenditures as well as costs of management support and enterprise integration support services. We analyzed the progress made against planned program milestones and reviewed technical documentation such as software development plans to gain insights into OCX progress. In addition, we reviewed briefings and schedule documentation provided by program and contractor officials to determine changes in OCX cost, schedule, and performance over time. These documents included annual GPS enterprise reviews, OCX program assessments, and program status briefings. We also interviewed officials from the Air Force Space and Missile Systems Center GPS and OCX program offices, OCX prime contractor Raytheon, and Defense Contract Management Agency (DCMA) officials charged with oversight of the OCX contractor efforts to identify and assess cost and schedule issues facing the program’s development efforts, major program risks, and technical challenges. Finally, we interviewed officials from DOD’s Office of Cost Assessment and Program Evaluation and Director, Operational Test and Evaluation office to discuss cost, schedule, and performance challenges for OCX. When program documents identified program events by fiscal quarter rather than by month, we used the last month of the given quarter as the date of the event. To determine the progress DOD has made in delivering M-code capable MGUE by the end of fiscal year 2017, we reviewed and analyzed program plans and documentation related to cost, schedule, acquisition strategy, technology development, and major challenges to delivering MGUE Increment 1. We then compared the information we obtained to GAO’s criteria for best practices in system development. To assess the program’s progress and challenges, we held discussions with and received information from officials at the Air Force Space and Missile Systems Center GPS and MGUE program offices; Office of the Undersecretary of Defense for Acquisition, Technology, and Logistics; DOD’s Office of Cost Assessment and Program Evaluation; and DOD’s Director, Operational Test and Evaluation. Additionally, to understand their respective development efforts and challenges, we interviewed contractor officials from the three MGUE prime contractors, L-3 Interstate Electronics Corporation, Raytheon Space and Airborne Systems, and Rockwell Collins Inc. as well as Defense Contract Management Agency officials overseeing those contractors. To identify the military services’ concerns about MGUE development, schedule, and integration and testing with their respective platforms, we interviewed officials from the lead platform program offices for the Army’s Defense Advanced GPS Receiver Distributed Device/Stryker, Air Force’s B-2 aircraft, Navy’s Arleigh Burke Class DDG-51 ship, and Marine Corps Joint Light Tactical Vehicle. As with OCX, when program documents identified program events by fiscal quarter rather than by month, we used the last month of the given quarter as the date of the event. We did not assess MGUE Increment 2 because it has not yet reached its technology development milestone. To identify the challenges faced in synchronizing GPS III satellite, OCX, and MGUE to deploy M-code capability, we built on our work assessing the plans, schedules, and challenges of the OCX and MGUE programs by reviewing integrated master schedules and analyzing approved GPS constellation reliability parameters. We also interviewed cognizant officials within Air Force Space Command (AFSPC) and the Aerospace Corporation, which supports the Air Force’s Space and Missile Systems Center (SMC), to evaluate predicted reliability of the GPS satellite constellation. We did not assess cost, schedule, or performance for GPS III satellites. To assess the status of the GPS constellation, we interviewed officials from the Air Force SMC GPS program office and AFSPC. To assess the risks that a delay in the acquisition and fielding of GPS III satellites could result in the GPS constellation falling below the 24 satellites required by the standard positioning service and precise positioning service performance standards, we employed a methodology very similar to the one we had used to assess constellation performance in 2009 and 2010. We obtained information from the Air Force predicting the reliability for 57 GPS satellites—each of the 39 current (on-orbit as of March 2015) and 18 future GPS satellites—as a function of time. Each satellite’s total reliability curve defines the probability that the satellite will still be operational at a given time in the future. It is generated from the product of two reliability curves—a wear-out reliability curve defined by the cumulative normal distribution, and a random reliability curve defined by the cumulative Weibull distribution. For each of the 57 satellites, we obtained the two parameters defining the cumulative normal distribution, and the two parameters defining the cumulative Weibull distribution. For each of the 18 unlaunched satellites we included in our model, we also obtained a parameter defining its probability of successful launch, and its current scheduled launch date. The 18 unlaunched satellites include 3 IIF satellites and 15 III satellites; launch of the final III satellite we included in our model is scheduled for March 2025. Using this information, we generated overall reliability curves for each of the 57 GPS satellites. We discussed with Air Force and Aerospace Corporation representatives, in general terms, how each satellite’s normal and Weibull parameters were calculated. However, we did not analyze any of the data used to calculate these parameters. Using the reliability curves for each of the 57 GPS satellites, we developed a Monte Carlo simulation to predict the probability that at least a given number of satellites would be operational as a function of time, based on the GPS launch schedule as of December 2014. We conducted several runs of our simulation—each run consisting of 10,000 trials—and generated “sawtoothed” curves depicting the probability that at least 24 satellites would still be operational as a function of time. We then used our Monte Carlo simulation model to examine the effect of a delay to the delivery of OCX block 1 until November 2020, which would delay the introduction of GPS III satellites into the operational constellation. We then reran the model based on this assumption, and calculated new probabilities that at least 24 satellites would still be operational as a function of time. Finally, we simulated the effect of the Air Force’s proposed contingency plan, which would enable GPS III satellites to be added to the operational constellation, with limited functionality, prior to the delivery of OCX block 1. We conducted this performance audit from August 2014 to September 2015 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the contact named above, Dr. Nabajyoti Barkakati, Chief Technologist; Art Gallegos, Assistant Director; Jason Lee, Assistant Director; Karen Richey, Assistant Director; Jay Tallon, Assistant Director; Marie P. Ahearn; Pete Anderson; Brandon Booth; Brian Bothwell; Raj Chitikila; Tana Davis; Roxanna Sun; and Hai V. Tran made key contributions to this report.

The satellite-based GPS provides positioning, navigation, and timing data to users worldwide. The Air Force is modernizing the satellite, ground control, and user equipment segments to enhance GPS performance. The Senate and House Armed Services Committee reports accompanying bills for the National Defense Authorization Act for Fiscal Year 2015 included provisions for GAO to review the status of OCX development and DOD's efforts to field M-code signal capability. This report addresses (1) the extent to which DOD is meeting cost, schedule, and performance requirements for OCX; (2) the progress DOD is making in delivering M-code capable MGUE by the end of fiscal year 2017; and (3) the challenges DOD faces in synchronizing the development of GPS III, OCX, and MGUE to deploy M-code. To conduct this work, GAO analyzed program documents such as acquisition strategies; reviewed oversight reporting; assessed constellation reliability metrics; and interviewed officials from DOD programs and contractors. The Air Force has experienced significant difficulties developing the Global Positioning System (GPS) next generation operational control system (OCX) and consistently overstated progress to the Office of the Secretary of Defense (OSD) compared to advisory independent assessments it received. It needs $1.1 billion and 4 years more than planned to deliver OCX due to poor acquisition decisions and a slow recognition of development problems. The Air Force began OCX development in 2010 prior to completing preliminary development reviews in contrast with best acquisition practices. It accelerated OCX development in 2012 to meet optimistic GPS III satellite launch timeframes even as OCX development problems and costs grew, and then paused development in 2013 to address problems and resolve what it believed were root causes. However, as the figure below shows, OCX cost and schedule growth have persisted due in part to a high defect rate, which may result from systemic issues. Further, unrealistic cost and schedule estimates limit OSD visibility into and oversight over OCX progress. The Air Force is implementing the military GPS user equipment (MGUE) program to develop for the military services GPS receiver cards capable of receiving the military-code (M-code) signal—which can help users operate in jamming environments. The Air Force has revised MGUE's acquisition strategy several times in its quest to develop the cards. Even so, the military services are unlikely to have sufficient knowledge to make informed procurement decisions starting in fiscal year 2018 because operational testing that provides valuable information about MGUE performance will not be complete until fiscal year 2019. The current GPS constellation has proven to be much more reliable than the Air Force predicted when GAO last reported on it in 2010. Given delays to OCX, the Air Force has prepared contingency plans for sustaining the GPS constellation, but these plans may not deliver the full range of GPS capability. Initial M-code capability will not be available until OCX delivery in mid-2019 at the earliest and full M-code capability is likely at least a decade away—once the services are able to deploy MGUE receivers in sufficient numbers. Until the OCX program trajectory is corrected, additional delays to it may likely pose significant risks to sustaining the GPS constellation and delivering GPS capability. GAO recommends that DOD obtain a more robust independent assessment of OCX to identify and resolve root causes, and ensure MGUE design is stable to inform testing and procurement decisions. DOD concurred on OCX but stated that actions taken to date are sufficient. DOD partially concurred on MGUE. GAO believes all recommended actions are necessary to address systemic problems.

In 1986, the Congress replaced CSRS with FERS for federal employees hired beginning January 1, 1984, in part to (1) recognize the inclusion of federal employees under Social Security and (2) reduce federal pension costs. Among the concerns of congressional deliberators in crafting FERS were that its retirement benefits be comparable with those under CSRS and enable employees to maintain their standard of living in retirement. To accomplish these and other goals, FERS provides a retirement benefit that comprises three components: a basic FERS annuity, Social Security payments, and TSP payments. The total income from these sources is meant to help individuals to receive retirement benefits comparable with CSRS benefits and commensurate with their retirement income goals. The basic FERS annuity is similar to CSRS in that it guarantees a specific monthly retirement benefit based on age, length of creditable service, and the average of the highest 3 consecutive years’ salaries. However, the FERS annuity is lower because its benefit formula credits each year of service generally at 1 percent while CSRS service credits range from 1.5 to 2 percent per year of service. In addition, cost-of-living adjustments authorized by FERS are lower and generally are not provided before age 62. Unlike the FERS basic annuity, the benefit provided under Social Security’s benefit formula declines as a proportion of individuals’ preretirement earnings as their earnings increase. For example, a person aged 62, with a certain lifetime earnings pattern and earnings of $20,000 in his or her final year of employment, would receive Social Security benefits that represent about 35 percent of those earnings. In contrast, a person aged 62, with a certain lifetime earnings pattern and earnings of $75,000 in his or her final year of employment, would receive a benefit that represents just about 17 percent of those earnings. Pension professionals believe that to maintain roughly the same living standard in retirement, individuals’ income needs generally range from 60 to 80 percent of their preretirement annual pretax earnings. Among other things, retirees typically pay less taxes, do not have work-related expenses such as daily commuting costs and clothing needs, may no longer have dependent children, and may have their mortgages paid. TSP is administered by the Federal Retirement Thrift Investment Board, which is an independent agency. The Board consists of five part-time members who are appointed by the President. TSP’s daily activities are carried out by a staff headed by an executive director selected by the Board. Retirement benefits from TSP are the flexible component of FERS because they depend on the amount that is in each employee’s account at retirement. Thus, TSP can help FERS-covered employees to save toward a total retirement benefit that is commensurate with their retirement income goals. Employees under FERS are automatically enrolled in TSP because federal agencies are required to contribute an amount equal to 1 percent of their employees’ salaries to the plan. In addition, employees can make voluntary contributions up to 10 percent of their salaries: agencies match the first 3 percent on a dollar-for-dollar basis and the next 2 percent at 50 cents to a dollar, for a 5 percent total agency contribution; additional employee contributions are not matched, but all contributions and earnings thereon are tax deferred. CSRS employees may also participate in TSP by contributing up to 5 percent of their salaries; while there is no agency match, the contributions and earnings are tax deferred. However, all employee contributions are limited to a statutory inflation-adjusted cap, which was $8,994 in 1993. TSP contributions can be invested in a federal government securities fund (G fund), a commercial bond fund (F fund), and a commercial large capitalization stock fund (C fund). The C and F funds are passively managed index funds that track changes in a certain body of securities in the stock and bond markets. These investment options were specified in TSP’s statute, which also provided for adding investment options, via amendments, at the request of TSP’s Board. In addition, TSP’s law restricted the amounts that could be invested in the C and F funds through 1990.With the lifting of the restriction in 1991, employees have increased their contributions to the C and F funds. For example, in January 1991 about 5 percent and 2 percent of contributions were going into the C and F funds, respectively, while in August 1994 the comparable rates were 35 percent and 10 percent. In January 1995, TSP contributions and earnings were invested as shown in table 1. TSP’s three funds have had different average annual rates of return since 1987. The C fund has averaged 12.5 percent, a higher return than the F and G funds’ average earnings of about 8.0 percent each over the 7 years of plan experience. The C and F funds also have been more volatile than the G fund as shown in figure 1. Figure 1 shows that a $1,000 investment in the C fund on January 1, 1987, would grow to $2,452 over the following 7 years based on actual annual rates of return. Similarly, $1,000 investments in the F and G funds would grow to $1,836 and $1,868, respectively, over the same period. The higher returns available from the C fund also connote the somewhat higher risks inherent in a stock portfolio. Thus, the retirement income TSP ultimately provides a participant will depend on how much the individual has contributed and on the rates of return earned on those contributions. Since returns and risks are related, the ability to diversify investments among stocks and bonds is an important factor for participants in a program such as TSP because it allows them to tailor their investment portfolios to reflect the level of risk they are willing to assume. The proportion of FERS-covered employees contributing to TSP has steadily increased. For example, in September 1987 some 219,000 FERS-covered employees (about 38 percent) were making voluntary contributions to TSP; whereas, in September 1994 about 942,000 (76 percent) were doing so. However, the degree of voluntary participation in TSP has varied considerably among salary ranges as shown in table 2. Most of the 300,000 (24 percent) FERS-covered employees who did not make any voluntary contributions were lower-paid workers. Historically, such employees have been less likely to make voluntary contributions than have employees in the middle and higher salary ranges. However, as the table shows, the lower salary ranges have shown the greater increase over time in the percentage of individuals who make contributions. Overall, in 1993 FERS-covered employees making voluntary contributions were deferring an average of 5.7 percent of their salaries compared with 3.7 percent in 1987. The deferral rates varied from 4.4 percent of their salaries for low-wage employees to 7.2 percent for the highest-wage employees as table 3 shows. Also, as with the percentage of employees making contributions, deferral rates vary among salary groups. The deferral rate among employees in the lower salary range has also increased the least compared with the rates of the other employees since 1987—about 27 percent compared with over 50 percent for all but the highest salary range. The 41-percent increase in the highest salary range may be partly due to the statutory inflation-adjusted cap on annual contributions, which was $8,994 for 1993. Our analysis showed a disparity in the extent to which higher- and lower-paid employees under FERS may need to contribute to TSP to achieve total FERS retirement benefits that would be commensurate with their preretirement standard of living. In general, lower-paid workers may achieve retirement income goals, or total benefits that are in the range of 60 to 80 percent of final annual earnings, with minimal TSP deferral rates, while higher-paid workers need to defer at correspondingly higher rates. A July 1986 Congressional Research Service report included illustrative comparisons of the replacement rates under FERS and CSRS for various retirement assumptions and TSP benefits from (1) just the mandatory agency 1-percent contribution and (2) employee voluntary contributions of 5 percent. In our analysis, we updated the Congressional Research Service’s illustration for employees retiring after 30 years of service at age 62. Our analysis showed that such employees with earnings in the lower salary ranges might achieve a level of FERS benefits that would be within 60 to 80 percent of final annual earnings with just their agencies’ mandatory 1-percent contribution but that employees in the higher salary ranges would not. However, using conservative assumptions of TSP returns of 6.1 percent, contributions of 5 percent throughout their careers would also provide higher-paid employees with an overall FERS replacement rate within this range as shown in table 4. Again, the disparity in the total replacement rates largely results from the varying level of benefits that Social Security provides to individuals in different earnings brackets. As table 4 shows, the Social Security replacement rate is just 14 percent for an employee aged 62 with final pretax wages of $100,000 but over twice as high (35 percent) for someone with final wages of $20,000. Furthermore, table 4 shows that a 5-percent deferral provides a total FERS replacement rate for higher-paid workers that is in the lower end of the range that pension professionals believe is needed (that is, the 64 and 68 percent shown in table 4). These lower replacement rates may not reflect such individuals’ retirement income goals and, consequently, these employees would need to contribute more than 5 percent to TSP to achieve a higher level of total FERS benefits. In general, the lower TSP’s investment earnings are the more an individual would need to contribute in order to reach a certain total FERS replacement rate goal; conversely, higher TSP returns would provide individuals with a higher retirement income than they projected as their goal at a given deferral rate. For example, using TSP’s actual average rate of return of 8.95 percent for the period 1988 to 1994 produces TSP replacement rates that are about 50 percent higher than those shown in table 4. The TSP Board produces and provides to federal agencies a variety of educational materials for their employees. Among other things, these leaflets, pamphlets, and brochures emphasize the monetary benefits of TSP, such as the advantages of tax deferral, the effects of compounding, and the higher returns possible from beginning to make contributions early in one’s career. In addition, these materials inform employees about the pros and cons, including potential risks, of investing in each of TSP’s three funds and the earnings history of each fund. However, TSP’s educational materials are not explicit in discussing the importance of employee TSP contributions in achieving total FERS retirement benefits that would be commensurate with preretirement living standards, that is, benefits in the range of 60 to 80 percent of earnings. For example, the materials do not include illustrative examples of FERS replacement rates at varying TSP deferral rates and their effect on total FERS benefits. Private sector plans have such examples in their educational materials. Were TSP’s Board to revise its materials to include that type of example, it would need to do so in collaboration with the federal Office of Personnel Management (OPM), which has some responsibility for overall FERS education, including establishing training programs for agency retirement counselors. In May 1995, TSP’s Board decided to seek legislation that would add two investment options: an indexed domestic small capitalization equity fund and an indexed international equity fund. The Board selected these funds because they add diversity and provide the opportunity for greater returns than the current options though at somewhat increased risk. Adding the two funds would make TSP’s number of investment options and mix more like those provided under private sector section 401(k) plans. TSP’s Board began looking into the possibility of increasing the number of investment options in 1992 after the statutory restrictions on C and F fund investments expired. Among other things, the Board reviewed the investment options generally available under section 401(k) plans and the returns and risks associated with them. On average, most private sector section 401(k) plans offer four or more investment options that include a number of bond and stock funds of varying risk. The Board’s actions to broaden TSP’s investment options are consistent with pension professionals’ beliefs that employees should have a variety of investment options encompassing a range of risks and returns to provide the opportunity for higher earnings that would increase their retirement nest eggs. The new options would allow TSP participants to diversify their investments. The new funds would complement the C fund, which has historically outperformed the G and F funds by an average of about 4.5 percentage points since 1987. Proposed legislation to add the options was introduced in the Senate on July 27, 1995, and in the House of Representatives on September 12, 1995. TSP was designed to provide one source of retirement income for FERS-covered employees. However, unlike the two other FERS components whose benefits are determined by formula and are constant for individuals with the same work histories, TSP’s benefits will vary according to the amounts that employees have contributed and the investment returns on those contributions. Because of the effects of Social Security’s benefit formula, higher-paid workers will be more dependent on TSP income than lower-paid workers in maintaining their standard of living in retirement. TSP’s educational materials, however, are not explicit in making this distinction. These materials should explain and provide examples of contribution rates and their relationship to preretirement earnings and potential retirement income. TSP was also designed to be a retirement savings vehicle for federal employees that is similar to section 401(k) plans for workers in the private sector. The addition of the indexed domestic small capitalization equity and indexed international equity funds will provide federal employees the same opportunity that those in the private sector have for tailoring their investment portfolios to reflect the returns they seek and the risks they are willing to undertake. We recommend that to help ensure that TSP participants have investment opportunities similar to those available under comparable private sector plans, the Congress enact legislation adding the two investment options sought by TSP’s Board. We recommend that the Board, in collaboration with OPM, include in TSP’s educational materials (1) an explanation of TSP’s pivotal role in enabling employees under FERS to achieve their retirement income goals and (2) explicit illustrations of the effects of TSP deferral rates on total FERS benefits. The Federal Retirement Thrift Investment Board provided written comments on a draft of this report (see app. II). The Board disagreed with our recommendation that TSP’s educational materials include an explanation of TSP’s role in FERS and explicit examples of the effect of TSP deferral rates on total FERS benefits. The Board stated that such actions by TSP would constitute employee education about FERS, which is an OPM responsibility under the FERS statute. The Board noted that its educational materials are replete with illustrations that show the dramatic effect of contributions and investment earnings on the size of an employee’s TSP account. However, the Board added that the materials do not analyze or explain the impact that employee TSP accounts will have on total FERS retirement income because FERS legislation gave that responsibility to OPM. Also, the Board provided some technical comments that we incorporated in the report as appropriate. While OPM has some responsibility for FERS education, such as establishing training programs for agency retirement counselors, we do not agree that authority to educate employees on the effects of TSP investments on their total FERS benefits is vested exclusively in OPM. We continue to believe that the Board is in a better position to develop educational materials that include explicit examples of TSP’s potential effects on FERS retirement income. Such examples would demonstrate TSP’s pivotal role in the context of FERS, particularly given the effect of Social Security’s benefit formula. For example, an OPM booklet on FERS includes examples of replacement rates for four individuals retiring at various ages, with differing work histories of federal and nonfederal service, and with TSP deferral rates of 3 and 5 percent. However, while the examples are helpful in showing the increased benefits derived from contributions at 5 percent compared with 3 percent, they are not explicit in demonstrating TSP’s significance in overall FERS benefits at retirement. Without its FERS context, we believe the value of TSP’s educational materials to the individual employee is greatly diminished. Furthermore, TSP is the appropriate source for such information because it periodically contacts all employees who participate in the plan—including those not making any voluntary contributions. Accordingly, we believe that TSP should prepare such educational materials. OPM officials stated that the Board could do so in collaboration with OPM. As arranged with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 5 days after its issue date. At that time, we will send copies of this report to other congressional committees and members with an interest in this matter and to others upon request. Our review was performed under the direction of Donald C. Snyder, Assistant Director. Other contributors were Endel P. Kaseoru, Evaluator-in-Charge, and evaluators Carolina M. Morgan and Gregory Curtis. If you or your staff have any questions about this report, please call me on (202) 512-7215 or Mr. Snyder on (202) 512-7204. We calculated illustrative FERS replacement rates for each of the program’s three components—the basic FERS annuity, Social Security benefits, and TSP—for employees retiring with 30 years of service at age 62, the average federal retirement age in 1994 for regular retirements. To make our calculations, we simulated the salary histories of five hypothetical federal employees and estimated the annuities they would receive under certain assumptions. The time frame for our analysis was 1986 through 2015. To produce the salary histories for our model, we used wage growth rates that are consistent with federal General Schedule salaries. The workers in our model began their federal careers in 1986 at entry-level salaries for GS-2, –3, –5, –7, and –9 and retired in January 2016 at age 62 with final annual salaries, as measured in 1995 dollars, of $20,000, $30,000, $45,000, $75,000, and $100,000. We first created an inflation-adjusted earnings history for these workers and then converted it to current year earnings using the actual inflation rates from 1986 to 1995 and 3.4 percent thereafter. To determine employees’ FERS annuities, we used the basic FERS annuity formula in the law. However, while the formula computes the benefit at 33 percent of the average of the highest 3 consecutive years’ salaries, the replacement rate is less than 33 percent because the estimated wages grow in each of the 3 years prior to retirement; thus, the 3-year average used to calculate the annuity is lower than the final year’s wages. To calculate Social Security benefits, we used the “ANYPIA” software program provided by the Social Security Administration’s Office of the Actuary. In applying this program, we used the alternative I assumptions of future economic activity from the 1994 report of the Board of Trustees of the Federal Old Age and Survivors Insurance and Disability Insurance Trust Funds. The alternative I assumptions are conservative, and thus they produced replacement rates that were lower by 1 to 5 percentage points than the rates produced by alternatives II and III. To calculate the TSP replacement rates, we estimated the balance in the individuals’ accounts at retirement based on employee and agency contributions of 5 percent each and the agency-only 1-percent contribution. For our baseline analysis, we assumed that the accounts earned a conservative return of 6.1 percent, the same rate the Congressional Research Service used in its analysis. We also calculated replacement rates using a weighted average of actual TSP returns from 1988 to 1994 of 8.95 percent. This higher annual rate of return produced TSP replacement rates that were about 50 percent higher for each salary level. We then calculated an annuity for each account balance using a worksheet in TSP’s annuities booklet. We assumed an increasing single life annuity at 6-percent interest, the rate used in TSP’s worksheet. The replacement rates we computed, shown in table 4, vary by final year wage because each had a different growth rate over the 30 years we modeled. We also tested different rates of wage growth, returns on TSP, and the FERS annuity and found the results were consistent across the five final salaries we modeled. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (301) 258-4097 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO provided information on the Federal Employee Retirement System (FERS), focusing on: (1) the extent to which employees under FERS voluntarily contribute to the Thrift Savings Plan (TSP); (2) how well TSP educational materials address the importance of employee participation; and (3) whether there should be additional TSP investment options. GAO found that: (1) as of September 1994, about 76 percent of FERS employees voluntarily contributed an average of 5.7 percent of their salaries to TSP; (2) most of the non-contributing employees were in lower pay grades; (3) lower paid workers contribute less of their salaries to TSP because social security benefits are proportionally higher for lower income workers; (4) these workers may not need to contribute as much to TSP in order to maintain their preretirement standard of living; (5) mid- and high-pay level workers need to contribute at least 5 percent of their salaries over their careers to achieve 60 to 80 percent of preretirement income; (6) although TSP educational materials extensively discuss the plan's financial aspects, they do not explicitly discuss the importance of employee participation in TSP; (7) the TSP Board is seeking legislation that would add two stock fund options to the three investment options it already offers in order to make these options more similar to those available in private-sector plans; and (8) although they carry a higher investment risk, these new investment options could potentially produce higher earnings for plan participants.

Legislation related to telework began to emerge from Congress in the 1990s. For example, beginning in 1992, Congress provided funding to GSA to establish the first federal telework centers. Three years later, Congress permanently authorized federal agencies to spend money to install telephone lines and related equipment and pay monthly charges for federal workers authorized to work at home, in accordance with OPM guidelines. Within the legislation that evolved from 1992 through 2009, the most significant congressional action was the enactment of Section 359 of Pub. L. No. 106-346 in October 2000. This section required each executive-branch agency to establish a telework policy “under which eligible employees of the agency may participate in telecommuting to the maximum extent possible without diminished employee performance.” It also directed OPM to provide that the law’s requirements were applied to 25 percent of the federal workforce by April 2001 and to an additional 25 percent of the federal workforce in each subsequent year, until 2004 when the law was to be applied to 100 percent of the federal workforce. From 2005 through 2009, federal workforce participation in routine telework has remained low. In calendar year 2009, according to OPM’s latest survey of federal agency telework coordinators, less than 6 percent of federal employees employed by the 79 agencies that responded to the survey teleworked at least 1 day per month, while less than 4 percent of the federal workforce employed by these agencies teleworked at least 1 day per week. The estimated percentage of employees teleworking at least 1 day per month, relative to the number of the federal employees employed by the agencies that responded to the survey, has remained between 5 and 7 percent, since calendar year 2005. The legislation for telework provided both OPM and GSA with leadership roles in the implementation of telework in the federal government. However, in 2003, we reported that the lack of coordination between OPM and GSA had resulted in executive agencies receiving conflicting messages on several telework-related topics, including emergency closings of government offices. These conflicting messages had created confusion for federal agencies in implementing their individual telework programs. To provide federal agencies with consistent, inclusive, unambiguous support and guidance related to telework, we recommended that OPM and GSA better coordinate their efforts. In response, OPM and GSA agreed later that year on a Memorandum of Understanding (MOU). Under this MOU, OPM and GSA agreed to the respective responsibilities listed in table 1. The agencies also agreed in the MOU to continue to share draft telework documents to ensure mutual concurrence. Recently, Congress passed a framework for implementing a comprehensive federal telework program. Congress passed the Telework Enhancement Act of 2010 in November 2010, and the President signed it into law in December 2010. The law requires each executive agency to designate a telework managing officer, establish a telework policy, and submit an annual report to the Chair and Vice Chair of the Chief Human Capital Officers (CHCO) Council on the agency’s efforts to promote telework. Under the act, OPM is to play a leading role in helping agencies implement the new telework provisions. The law requires OPM to provide policy and policy guidance for telework in several areas, including pay and leave, agency closure, performance management, official worksite, recruitment and retention, and accommodations for employees with disabilities. In developing its telework policy and policy guidance, OPM is to consult with FEMA, GSA, and the National Archives and Records Administration (NARA) relative to their designated areas of policy responsibility, as listed in table 2. NARA provides guidance to agencies on ensuring the federal government’s essential records are, among other things, secure and accessible to key federal personnel during emergencies. The Telework Enhancement Act also includes several provisions related to the potential use of telework during emergencies. The law requires agencies to incorporate telework into their COOP plans. OPM is to report annually to Congress and provide its assessment of each agency’s progress toward its goals, such as the effect of telework on emergency readiness. Finally, the act requires the Director of OPM to research the utilization of telework that identifies best practices and recommendations for the federal government and review the outcomes associated with an increase in telework. Agencies with jurisdiction over such matters as energy consumption, urban transportation patterns, and planning the dispersal of work during periods of emergency, shall work cooperatively with the Director, as necessary, to carry out these research responsibilities. OPM, GSA, FEMA, and FPS offer a host of guidance on telework or telework-related emergency planning. Several of these guidance documents have expanded significantly in recent years, broadening the scope of the topics that they address and describing broader responsibilities for the lead agencies. After Congress required agencies to establish a telework policy in 2000, OPM provided several telework guidance documents to agencies. OPM’s major telework guidance is contained in the Guide to Telework in the Federal Government. According to Federal Continuity Directive 1 (FCD 1), OPM is also responsible for developing and promulgating personnel guidance to support the operation of federal executive-branch agencies during emergencies. OPM has also issued regulations providing uniform instructions to agencies on making payments to employees evacuated due to, among other reasons, natural disasters and pandemic health crises. OPM also has the lead role, by mutual agreement with other agencies in the region, in determining when to close federal offices and dismiss employees in the Washington, D.C., area. According to OPM, these procedures, referred to as the Dismissal Guide, ensure a coordinated response to areawide disruptions, which is important given the concentration of federal employees in this area. According to an OPM official, its dismissal and closure procedures are updated annually. The latest procedures were issued in December 2010. While these procedures directly apply only to executive agencies in the Washington, D.C. area, OPM officials explained that agency leaders in other regions of the United States use OPM’s procedures as a model. In addition, OPM’s guidance on managing human-capital resources during emergencies and continuity events was included in FEMA’s directive to executive agencies on developing and implementing continuity plans. Table 3 lists examples of OPM guidance relating to the use of telework during various types of emergencies, as provided in OPM and FEMA documents. Since 2000, OPM has expanded the scope of its guidance on the use of telework in the federal government. For example, the telework guidance OPM issued in 2001 listed several topics agencies should address in their telework policies. The current version of OPM’s guidance for routine telework, the 2011 Telework Guide, provides guidance for managers and employees on an array of topics, such as the telework agreement, purchasing equipment, communications, and performance management. This guide also provides advice on using telework during emergencies, continuity events, and pandemics. Compared to the 20 06 Telework Guide, the immediate predecessor to the 2011 Guide, the 2011 Guide added  a summary of the major sections of the Telework Enhancement Act;  the checklist items that OPM used to evaluate telework policies to help agency officials assess and revise their policies;  guidance on pay, leave, and work-schedule flexibilities, including the new option of unscheduled telework; and telework and reasonable accommodations for employees with disabilities. During the same period, OPM expanded some of its emergency-related pay and leave policies. For example, OPM broadened its regulations on pay during an evacuation to include, in the event of a pandemic health crisis, permission for agencies to order employees to work from home (or an alternative location), regardless of whether they have a telework agreement. As the lead agency in the management of federal workplaces, GSA provides governmentwide workplace guidance related to telework, emergency planning, and continuity planning and operations. Under its authority to provide guidance, assistance, and oversight on the establishment and operation of alternative workplace arrangements, such as telework, GSA has issued guidelines, through Federal Management Regulation (FMR) Bulletins, on the use of information and telecommunication technology to support telework and the implementation and operation of alternative workplace arrangements, as shown in table 4. GSA, under its responsibility to operate, maintain, and protect the buildings and grounds it controls, also leads the federal Occupant Emergency Program, which consists of short-term emergency- response programs for particular facilities. Under this program, each agency or facility develops and maintains an occupant emergency plan (OEP) which details the procedures for safeguarding lives and property, such as evacuation or shelter-in-place, in response to a wide range of emergencies. At the agency level, the OEP and the continuity plan are the principal emergency-response plans. According to GSA’s regulations, GSA approves OEPs for GSA-controlled facilities. Lastly, as described in FCD 1, GSA is also responsible for assisting FEMA and other federal agencies in continuity planning and operations. GSA’s ongoing responsibilities include coordinating the provision of facilities to support the continuity of the executive-branch agencies and providing data on alternate facilities. GSA issued regulations on telework in 2005. These regulations included references to existing statutory provisions on telecommuting and OPM’s 2001 guidance regarding telework and telework centers. In these regulations, GSA also described the statutory obligation of agencies to consider whether the need for space could be met by alternative workplace arrangements and offered to assist agencies with alternative workplace arrangements. In 2006, GSA issued an FMR Bulletin that provided significantly expanded guidance to agencies on using alternative workplace arrangements. In addition to what had been stated in the regulation, the bulletin provided detailed information on implementing alternative work arrangements, such as teleworking and telework centers. The bulletin addressed such topics as the agency’s provision of equipment, payment of telework-related expenses, and factors to consider in using alternative workplace arrangements, often with citations to related federal regulations or other guidance. GSA issued another FMR bulletin the following year that provided agencies with recommendations on the type of information technology (IT), telecommunications, and peripheral services and equipment, as well as security, needed to adequately support telework. Like its predecessor, the bulletin also provided citations to related federal regulations or guidance issued by other federal agencies. FEMA and FPS provide governmentwide guidance to federal agencies on responding to emergencies. FEMA is responsible for leading the nation in developing a national preparedness system, including serving as the executive branch’s lead agent on matters concerning the continuity of national operations. To help federal agencies develop continuity plans and programs, FEMA offers direction in the FCD 1. According to this directive, each federal agency has continuity responsibilities and certain agencies have leadership responsibility for providing governmentwide guidance or services. For instance, all federal agencies are responsible for, among other things, developing plans to ensure the continuation of their essential functions and incorporating continuity requirements into their daily operations. Among other continuity guidance FEMA has issued, FEMA has also provided agencies with a COOP plan template to help them develop their plans. Both FEMA documents, listed in table 5, include guidance relating to the potential use of telework. FPS is authorized to protect the buildings, grounds, and property that are under the control and custody of GSA, as well as the persons on the property. Towards that end, FPS provided agencies with a guide, template, and instructions on developing an OEP and guidelines on responding to various emergency situations. The OEP guide refers to the potential use of telework during emergencies. A GSA official indicated that those federal agencies that have delegated authority to own and operate their own, non-GSA facilities, are required to have their own version of the OEP. He said that most of these agencies follow FPS’s guidance in developing their OEPs. As with some of the OPM and GSA guidance documents, the current version of FEMA’s guidance on continuity planning, FCD 1, issued in 2008, has expanded since the original version was issued in 2004. For example, in the 2008 version of FCD 1, the description of GSA’s responsibilities included such additional duties as coordinating the provision of executive-branch facilities to support continuity operations. Similarly, the description of OPM’s additional responsibilities include providing guidance to agencies on developing personnel policies that address continuity plans and procedures, as well as alternate work options, and coordinating continuity efforts before, during, and after an emergency with the Federal Executive Boards (FEB). In addition, FEMA broadened the scope of the guidance to include such topics as using risk management to maximize an agency’s readiness, and aligning acquisition and budgeting to support the continuity program. Several recent reports from OPM, inspectors general, and GAO have identified several examples of potential problems with incorporating telework into emergency or routine operations at various agencies. These examples identified problems in such operational areas as planning, IT, personnel readiness, and program monitoring. For instance, in 2010 OPM evaluated the telework policies of 72 executive-branch departments and agencies to determine whether the policies provided a foundation for effective telework programs. Each policy was evaluated against two objectives: (1) whether the policy could be clearly understood and easily used and (2) whether the policy included elements essential to the development and support of an effective program, with respect to program implementation, participant responsibilities, and program operations. For each evaluation objective, OPM developed a checklist of items that OPM evaluators used to assess the agencies’ telework policies. One of the policy checklist questions asked the evaluators to assess whether the telework policy “references agency emergency policies (e.g., COOP and pandemic).” According to the report, OPM evaluators scored 25 of the 73 agency telework policies (35 percent) as not referencing the agencies’ emergency policies at all. Other recent reports indicated potential problems in such operational areas as IT, personnel, and program monitoring: In 2010, we reported that federal wireless networks were increasingly vulnerable to attack and that information on the networks was vulnerable to unauthorized use and disclosure. In 2009, on the basis of our review of a 2007 Department of Homeland Security (DHS) report, we concluded that, in the event of a protracted emergency where 40 percent or more of the population was absent from school or work, residential users in most locations in the United States, including federal teleworkers, would likely experience congestion when attempting to use the Internet.  While OPM’s Telework Guide indicates that the potential to use telework during an emergency depends on agencies implementing routine telework as broadly as possible, OPM’s latest telework survey, as noted earlier, reported that less than 6 percent of federal employees employed by the 79 agencies responding to the 2009 survey teleworked at least 1 day per month. In 2010, the Inspector General for the U.S. Nuclear Regulatory Commission (NRC) reported that, as of October 2009, most of NRC’s offices had not identified all of the individuals needed to perform essential functions and high-priority tasks while teleworking during a pandemic.  According to an OPM official, few agencies in the Washington, D.C., area during the winter of 2011 were able to provide data on employees’ use of telework during an emergency. In issuing its Dismissal Procedures in December 2010, OPM requested that agencies in this area provide data on the use of telework following any OPM-announced dismissal or closure. However, by March 2011, OPM withdrew its request because, according to an OPM official, few agencies in the area provided the requested data. Our review of the OPM, GSA, FEMA, and FPS governmentwide guidance on telework and telework-related emergency planning found that none of the documents provided a definition of what constitutes incorporating telework into continuity and emergency planning or operations, or a cohesive set of practices that agencies could use to achieve this type of incorporation. Such practices would address the wide range of factors that could affect the potential use of telework during emergency operations such as planning, training, IT infrastructure (equipment, software, and security), testing, facilities, data collection, and program monitoring. OPM, in partnership with GSA, has conducted an annual survey of the executive-branch agencies since 2001 to ascertain the status of telework and gauge agency progress in various aspects of their telework programs, such as participation, policy, eligibility, cost savings, and technology. Since 2002, OPM has used the survey results to prepare its annual report to Congress on the status of telework in the federal government. Since 2004, the survey has asked agency telework coordinators about whether telework had been integrated or incorporated into the agency’s emergency or continuity plans, or both. In the latest survey, the 2009 Telework Survey, the question was worded as follows: “Telework has been integrated into your agency emergency preparedness / COOP plans.” The survey asks the coordinators to respond with either Yes or No. A review of the responses to this question, for calendar years 2004 through 2009, shows that the number of agencies reporting they had integrated telework into their emergency preparedness or continuity plans increased from 27 of 78 agencies responding for calendar year 2005 (or 35 percent), to 57 of 79 agencies responding for calendar year 2009 (or 72 percent). OPM reported that the 2009 Telework survey results were an example of an encouraging use of telework implementation practices. However, OPM’s survey instrument does not describe what OPM means by “integrating” telework into emergency or continuity planning and operations. The survey does not provide a definition or citation to another document. This is also the case for the 2004 through 2008 surveys. One reference the telework coordinators might have turned to was OPM’s 2006 Telework Guide, the version in use at the time of the 2007 through 2009 surveys. However, the 2006 guide did not provide a definition or describe a set of practices required to integrate telework into emergency or continuity planning and operations, or refer to other federal guidance containing such a definition or set of practices. This lack of a definition or description calls into question the reliability of the survey results for assessing agencies’ progress. As GAO has previously indicated, survey questions should use unambiguous language and concrete terms, and specify the conditions that the respondents are to report on. Without a common understanding of what OPM means by “integration,” the agency telework coordinators who responded to the survey would have applied their own understanding of what integration means. As a result, OPM officials could not describe what agencies meant when they reported they had integrated telework into agency emergency preparedness / COOP plans. The question is also vague in that it does not define what it means by emergency preparedness plans and seems to treat emergency preparedness plans the same as continuity plans. As noted earlier, agencies may have several different types of emergency preparedness plans. The question does not explain whether the agency is to reply “Yes,” if telework has been “integrated” into just one emergency preparedness plan, or only if telework has been “integrated” into all of the emergency preparedness plans, as well as the COOP plan. Lastly, the response choice of Yes or No does not permit the coordinator to report interim progress, as would be possible if the response choices reflected various stages of integration. The Telework Enhancement Act of 2010 now requires that federal agencies incorporate telework into their COOP plans. However, the act itself does not define the standards for adequate incorporation. Nevertheless, we found that three lead telework agencies and FPS have provided some governmentwide guidance that could help agencies in their efforts to incorporate telework into their continuity and emergency plans and operations. In reviewing several current OPM, GSA, FEMA, and FPS guidance documents, we found a number of practices that could help agencies incorporate telework into aspects of their continuity or emergency planning. Appendix II provides a listing of practices that OPM, GSA, FEMA, and FPS have suggested in various telework or emergency- related guidance documents. These practices address a variety of operational areas, such as human capital, training, facilities, testing, and technology infrastructure. However, it would be difficult for an agency to use these practices to help achieve telework incorporation and assess their progress. First, the practices are included in guidance documents that are principally concerned with matters other than incorporating telework into emergency plans. For example, as noted earlier, OPM’s Dismissal Guide describes the announcements and procedures that federal executive agencies in the Washington, D.C., area are to follow when there are work disruptions in the region, but also reminds the agencies that they have a responsibility to ensure that their equipment and technical support have been tested, and their IT infrastructure can support a large number of teleworkers simultaneously. However, the guide does not provide a cross-reference to additional OPM or GSA guidance that would help agencies use IT equipment and services to support telework, or suggest consultation with the agencies’ chief information officers (CIO). Moreover, according to their 2003 MOU, OPM and GSA agreed to work together to help agencies use IT to support telework and facilitate the CIOs involvement in related planning. Second, these practices are scattered across a wide range of documents, so it is difficult to be certain that agencies would consider all operational areas required to fully incorporate telework into emergency or continuity planning. For example, as illustrated in appendix II, OPM guidance dealing with a range of human-capital topics, such as pay and leave and human-resource flexibilities, also includes practices that could help agencies incorporate telework into their continuity and emergency planning. But without lead agency collaboration and consensus on a comprehensive set of practices on incorporating telework into emergency and continuity planning, agencies and OPM cannot be sure that each agency has considered all of the relevant areas of its operations that may need to be adapted to support teleworkers during an emergency. It is also not unusual for a lead agency to draw practices, in whole or in part, from other lead agencies’ guidance. While this helps the reader become aware of the potential crosscutting implications of the guidance, the description of these practices often does not provide a reference to more-specific guidance available from the other lead agency, which would also be helpful. For example, FEMA’s FCD 1 suggests repeatedly that agencies consider using telework to support continuity operations, but does not reference OPM or GSA telework guidance. OPM and FEMA have both recently released updated guidance pertaining to telework and emergency planning, and both agencies reached out to some of their stakeholder communities. Following the extended closings of federal agencies in the Washington, D.C., area during the February 2010 snowstorms, OPM updated the Dismissal Guide to introduce “unscheduled telework,” a new option for federal employees to telework, to the extent possible, when severe weather conditions or other circumstances disrupt commuting. OPM officials believe this option will help maintain the productivity and resilience of the federal workforce during periods of heavy snow accumulation, national security events, and other regional emergencies and help ensure the safety of employees. This new option had effects in areas of policy and guidance, such as pay and leave, agency IT capacity, telework data collection, and emergency operations. Although only applicable to executive-branch agencies in the Washington, D.C., area, OPM officials indicated that the guide also serves as a reference and model for closures and dismissals due to snow and other emergencies in other regions. The guide notes that FEBs coordinate similar dismissal or closure procedures in other major metropolitan areas. Officials from OPM’s Pay and Leave Office led the development of the 2010 Dismissal Guide. The OPM officials’ goal was to coordinate among stakeholders and issue the guide before the arrival of winter storms. According to these officials, they developed potential revisions and began the process of coordination among many stakeholders on October 28, 2010, by briefing the deputy Chief Human Capital Officers (CHCO) on the new concepts and asking for their reactions. During November 2010, OPM officials continued to share concepts and drafts, receive comments, and incorporate revisions to the draft guide. During this period, they sought additional reviews of the draft guide from agency human-capital officials. The Executive Director of the CHCO Council distributed the draft for agency comment to the deputy CHCOs, requesting that they share the draft with the points of contacts within their agencies. In addition, OPM officials said they consulted with unions, the Metropolitan Washington Council of Governments, and the FEMA Office of National Capital Region Coordination, among others. In December 2010, the OPM officials completed their stakeholder reviews and OPM completed its internal clearance process to issue the guide. They presented the new Guide to the CHCO Council members and the White House staff on December 14 and held a media briefing on the new guide when it was issued on December 15. On the same day, OPM officials reported they also briefed the FEBs on the new guide. FEMA also provided updated guidance in 2010 to agencies on developing their continuity plans. In 2004 FEMA issued a template to help agencies develop continuity plans. The template provided general guidance, sample text, and a template containing all the elements of a viable continuity plan. This enabled agencies to insert information from the template into their continuity plans as they deemed appropriate. FEMA began developing its latest version of the template in 2009. By May 2009, FEMA National Continuity Programs (NCP) officials had developed the updated template sufficiently to share an initial version of it with departments and agencies to prepare for the 2009 continuity exercise called Eagle Horizon. FEMA NCP officials continued to update the template during 2010. According to NCP officials, the template was discussed during Interagency Continuity Working Group meetings and Continuity Advisory Group meetings. In September and October 2010, NCP officials e- mailed more than 100 members of the interagency community, including both continuity coordinators and continuity planners, requesting comments on the updated template draft. In addition, NCP officials provided the GSA Office of Emergency Response and Recovery, which has responsibility for policy guidance on alternate continuity facilities, the opportunity to review and comment on the template. FEMA finalized the continuity template in February 2011 and posted it on its website. The template provided agencies with more-detailed guidance for each section of the plan, including proposed language, citations for supporting authorities and a glossary of terms. This template included guidance on the use of alternative facilities, including telework. Although OPM officials shared drafts of the Dismissal Guide with other federal agencies, including GSA, the coordination between OPM and GSA did not involve working together to involve the agencies’ CIO community. As noted earlier, to address coordination issues between OPM and GSA, in 2003, the two agencies committed to an MOU. Under the MOU, GSA agreed to develop innovative workplace policy and guidance and to seek resolution to issues regarding IT support for telework where concerns about home IT equipment continue to be a barrier to the successful implementation of telework programs. OPM is to work with GSA to reach appropriate groups of CIOs to facilitate their greater involvement in agency long-term planning. The Dismissal Guide noted that the new option of “unscheduled telework” could occasion large numbers of employees teleworking simultaneously. Consequently, the Dismissal Guide advised agencies that they should ensure that their IT infrastructure was in place to support this. IT infrastructure includes equipment, software, and security. However, in 2009, we reported that 12 agencies, including GSA, as indicated in table 6, had done some, little, or no testing of the ability of their IT infrastructure to handle telework or similar arrangements during a protracted emergency, such as a pandemic influenza. Without adequate testing of IT infrastructure, agencies would not know whether they could rely on unscheduled telework to sustain operations during emergencies. Some agencies may have limited IT capacity to support large numbers of employees attempting to telework simultaneously, and might have to take steps to distribute system load until the agency capacity is increased. For example, the Department of Energy’s (DOE) Inspector General reported in 2010 that DOE’s COOP plan estimates peak employee absenteeism during a pandemic event could be as high as 40 percent, or approximately 3,000 employees at DOE headquarters; however, DOE’s remote access server can only accommodate 800 concurrent users. GSA’s last assessment of the telework technology environment and the technologies required to support 25 to 50 percent of the federal workforce teleworking was conducted in 2006. At that time, GSA concluded, “In general most organizations do not provide their teleworkers the same level of access to agency applications, data, and technical support as their office workers, which can hinder a teleworker’s ability to perform all job duties from their telework site.” Despite the importance of agency IT capacity to support telework during emergencies, OPM and GSA did not work together, as outlined in their MOU, to reach out to the CIO community regarding potential agency capacity limitations. At the time of development of the new policy, each of the OPM and GSA officials responsible for coordinating on the MOU did not recognize the opportunity to involve the CIO community. The OPM official thought the agency telework coordinators and CHCOs were aware of the IT capacity issue and would consider this in implementing their telework programs. The GSA official thought the policy was targeted to the human-resources community and the technology issue would be more effectively dealt with in a document directed to the IT community. Consequently, GSA officials did not offer any governmentwide guidance on ways to address infrastructure limitations or provide direct assistance to agencies regarding the adequacy of their IT infrastructure. The senior OPM official responsible for telework programs acknowledged that consulting with the CIO Council prior to issuing the Dismissal Guide would have been worthwhile. Going forward, coordination will become even more important because the Telework Enhancement Act of 2010 requires OPM to consult with GSA on policy and policy guidance for telework in the areas of telework technology and equipment. The Telework Enhancement Act of 2010 requires agencies to incorporate telework into their COOP plans. However, there is no governmentwide definition or cohesive set of practices for incorporating telework into COOP plans. Recent reports offered several examples of challenges agencies may face in supporting the use of telework during emergencies in such key areas as planning, IT, and personnel readiness. The absence of a definition of what constitutes incorporation of telework into emergency and continuity plans did not create these potential problems in readiness. However, a definition accompanied by a set of practices that acknowledged the broad range of operational areas affected by incorporating telework into emergency preparedness might help agencies identify, avoid, or address any operational problems. Such practices would help agencies take the appropriate steps, measure their progress, and know when they have achieved success. Some practices already in OPM, GSA, FEMA, and FPS guidance documents could provide a starting point. The lack of a definition for incorporating telework into emergency and continuity planning has also affected the reliability of recent OPM telework survey results and assessment of agency progress. Since 2004, OPM has used its annual survey of agency telework coordinators to report on the status of agencies’ telework program and policies, including agencies’ incorporation of telework into emergency plans. However, problems with OPM’s telework survey methodology, such as a lack of a definition for incorporation or integration into continuity and emergency planning, and vagueness in question construction, call into question the reliability of reported results with respect to this issue. Without reliable data, OPM is unable to assess the extent to which the federal government as a whole, or an individual agency, is making progress on incorporating teleworking into emergency planning and operations. Over the years, governmentwide telework and emergency-related guidance from OPM, GSA, FEMA, and FPS has broadened in scope to include references to the potential use of telework during emergencies. In addition, these guidance documents suggest practices drawn from other lead agencies’ guidance. While these guidance features can heighten agency officials’ awareness of the crosscutting considerations bearing on the potential use of telework during emergencies, without lead-agency collaboration and consensus on the crosscutting passages in the guidance, coordination with a key policy area may be missed. In a recent instance, while OPM coordinated with other agencies on its new Dismissal Guide, the guide did not address important areas of agency IT operations––despite the existence of an MOU between OPM and GSA intended to improve interagency coordination specifically on this issue. In developing future telework policy and policy guidance relating to emergency and continuity planning and operations, OPM will need to consult with agencies responsible for key policy areas, now including NARA, to ensure all key policy areas are fully considered. To enhance the potential use of telework during emergencies, we recommend that the Director of OPM, in consultation with agencies responsible for key policy areas specified under the Telework Act and with other agencies providing governmentwide guidance on emergency preparedness, such as FPS, take the following three actions:  Develop (1) a definition of what constitutes incorporating telework in emergency and continuity plans and (2) a cohesive set of practices that agencies should implement to achieve successful incorporation.  Revise OPM’s data-collection methodology to ensure agencies and OPM report reliable results on the extent to which agencies have incorporated telework into their emergency and continuity planning and operations.  Establish an interagency coordination process among OPM, FEMA, FPS, GSA, and NARA to ensure all major areas of agency operations are considered when OPM issues new or updated guidance related to using telework during emergencies. We provided a draft of this report to the Secretary of Homeland Security, the Administrator of GSA, and the Director of OPM for review and comment. The Director of OPM and the Director of the Departmental GAO/OIG Liaison Office for DHS provided written comments, which we have reprinted in appendixes III and IV. In summary, OPM concurred with our recommendations and highlighted a number of actions the agency has under way or plans to undertake in response. OPM and DHS also provided technical comments, which we incorporated as appropriate. GSA had no comments on the draft. We are sending copies of this report to the congressional committees with jurisdiction over DHS and GSA, and their activities; the Secretary of Homeland Security; the Administrator of GSA; and the Director of Office of Management and Budget. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you have any questions about this report, please contact me at (202) 512-6543 or steinhardtb@gao.gov. Key contributors to this report are listed in appendix V. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. This report (1) describes the guidance that the Office of Personnel Management (OPM), General Services Administration (GSA), Federal Emergency Management Agency (FEMA), and Federal Protective Service (FPS) have issued pertaining to the use of telework during emergencies; (2) assesses the extent to which OPM’s recent reviews of the agencies’ telework policies and programs address the incorporation of telework into continuity plans, and the extent to which OPM, FEMA, GSA, and FPS offer guidance on incorporating telework into emergency and continuity planning; and (3) assesses the extent to which OPM and FEMA coordinated with other agencies on the development of their recently released guidance documents pertaining to the use of telework during emergencies. To address these three objectives, we reviewed governmentwide telework and emergency-related statutes. We also reviewed regulations issued by OPM and GSA, and related governmentwide guidance that OPM, GSA, FEMA, and FPS had issued over the past 10 years. Lastly, we conducted interviews with key officials from each of these agencies regarding each objective. We conducted additional data collection and analyses to answer selected objectives, as described below. To assess the extent to which OPM’s recent reviews of the agencies’ telework policies and programs addressed the incorporation of telework into continuity plans, we reviewed OPM’s annual telework survey and its 2010 evaluation of agency telework policies. We compared the survey question relating to whether telework had been integrated into the agency’s emergency preparedness / continuity of operations (COOP) plans to generally accepted survey methodology. We also reviewed the survey results for this question from the first year it was included in the survey, in 2004, through the most recent survey, conducted in 2009. We also compared OPM’s description of the methodology it used to evaluate agency telework policies to GAO guidelines for developing and using checklists. We also reviewed the cumulative scores that OPM evaluators assigned to the checklist item—whether the telework policy “references agency emergency policies (e.g., COOP and pandemic).” In addition to reviewing OPM’s recent telework assessments, we reviewed recent GAO and inspector general reports to identify examples of problems agencies might be having with potentially using telework during emergencies. To identify practices suggested by OPM, GSA, FEMA, and FPS for incorporating telework into continuity or emergency planning, we reviewed these agencies’ related governmentwide regulations and guidance, and identified suggested practices for incorporating telework into continuity or emergency planning. We compared these practices to practices that GAO previously suggested for the same purpose and identified examples of OPM, GSA, FEMA, and FPS suggested practices that were similar to GAO-suggested practices. We conducted this performance audit from February 2010 through July 2011 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In reviewing several current guidance documents, we found that the Office of Personnel Management (OPM), General Services Administration (GSA), Federal Emergency Management Agency (FEMA), and Federal Protective Service (FPS) have suggested to federal agencies, in various telework or emergency-related guidance documents, several practices for incorporating telework into various aspects of continuity or emergency planning. In addition to the contact named above, William Doherty, Assistant Director, and Patricia Farrell Donahue, analyst-in-charge, led the development of this report. Sharon Hogan and Robert Gebhart made significant contributions to this report. Gregory Wilmoth and Tom Beall assisted with the design and methodology. Karin Fangman provided legal counsel. Robert Love, Len Benning, and James Sweetman provided technical assistance. William Trancucci verified the information in the report. Federal Work/Life Programs: Agencies Generally Satisfied with OPM Assistance, but More Tracking and Information Sharing Needed. GAO-11-137. Washington, D.C.: December 16, 2010. Human Capital: Telework Programs Need Clear Goals and Reliable Data. GAO-08-261T. Washington, D.C.: November 6, 2007. Human Capital: Greater Focus on Results in Telework Programs Needed. GAO-07-1002T. Washington, D.C.: June 12, 2007. Continuity of Operations: Agencies Could Improve Planning for Telework during Disruptions. GAO-06-740T. Washington, D.C.: May 11, 2006. Continuity of Operations: Selected Agencies Could Improve Planning for Use of Alternate Facilities and Telework during Disruptions. GAO-06-713. Washington, D.C.: May 11, 2006. Agency Telework Methodologies: Departments of Commerce, Justice, State, the Small Business Administration, and the Securities and Exchange Commission. GAO-05-1055R. Washington, D.C.: September 27, 2005. Human Capital: Key Practices to Increasing Federal Telework. GAO-04-950T. Washington, D.C.: July 8, 2004. Facilities Location: Progress and Barriers in Selecting Rural Areas and Using Telework. GAO-03-1110T. Washington, D.C.: September 4, 2003. Human Capital: Further Guidance, Assistance, and Coordination Can Improve Federal Telework Efforts. GAO-03-679. Washington, D.C.: July 18, 2003.

When historic snowstorms forced lengthy closings of federal offices in the National Capital Region in 2010, thousands of employees continued to work from their homes, making clear the potential of telework in mitigating the effects of emergencies. GAO was asked to (1) describe the guidance lead agencies have issued pertaining to the use of telework during emergencies; (2) describe Office of Personnel Management (OPM) and other assessments related to agencies' incorporation of telework into emergency or continuity planning, and the extent to which the lead agencies have provided definitions and practices to support agency planning; and (3) assess the extent to which OPM and the Federal Emergency Management Agency (FEMA) coordinated with other agencies on recent guidance documents. To address these objectives, GAO reviewed relevant statutes, regulations, guidance documents, and OPM's telework survey methodology, and interviewed key officials of agencies providing telework and telework-related emergency guidance. OPM, the General Services Administration (GSA), FEMA, and the Federal Protective Service (FPS) offer a host of telework and telework-related emergency guidance. These lead agencies provide advice to other federal agencies through regulations, directives, guides, bulletins, and other documents. Several of these guidance documents have expanded significantly in recent years, broadening the scope of the topics that they address and describing broader responsibilities for the lead agencies. The Telework Enhancement Act of 2010 requires agencies to incorporate telework policies into their continuity of operations plans, but recent OPM reviews and other agency reports identify potential problems agencies may face in achieving this incorporation in various operational areas. GAO's review of the OPM, GSA, FEMA, and FPS governmentwide guidance on telework or telework-related emergency planning found that none of the documents provide a definition of what constitutes incorporating telework into continuity and emergency planning or a cohesive set of practices that agencies could use to achieve this type of incorporation. Additionally, this lack of a definition or description calls into question the reliability of the results of a survey OPM annually conducts to assess agencies' progress. In reviewing several lead-agency guidance documents, GAO found a number of practices, in areas such as information technology (IT) infrastructure and testing, that could help agencies incorporate telework in aspects of their continuity or emergency planning. However, because the practices are scattered among various documents principally concerned with other matters, it would be difficult for an agency to use these practices to help achieve telework incorporation and assess its progress. Both OPM and FEMA coordinated the development of their recent guidance. OPM updated its Washington, D.C., area dismissal and closure procedures to introduce "unscheduled telework," a new option for federal employees to telework when emergencies disrupt commuting. While developing these procedures, OPM officials reported coordinating with GSA, agency human-capital officials, FEMA, unions, and the Metropolitan Washington Council of Governments, among others. However, OPM and GSA did not work together to reach out to agency chief information officers regarding potential agency capacity limitations. Consequently, officials did not offer any governmentwide guidance on ways to address IT infrastructure limitations or provide direct assistance to agencies regarding the adequacy of their IT infrastructure. In February 2011, FEMA provided agencies with more-detailed guidance for developing continuity plans. According to FEMA officials, in 2010 they shared a draft of the guidance with the interagency community, including both continuity coordinators and continuity planners, and GSA. GAO recommends that OPM consult with other lead agencies to develop a definition and cohesive set of practices for incorporating telework into emergency and continuity planning; improve its related data collection; and establish an interagency coordination process for guidance. OPM concurred with GAO's recommendations.

Information security is a critical consideration for any organization that depends on information systems and computer networks to carry out its mission and is especially important for a government corporation such as FDIC, which has responsibilities to oversee the financial institutions that are entrusted with safeguarding the public’s money. While the use of interconnected electronic information systems allows the corporation to accomplish its mission more quickly and effectively, their use also exposes FDIC’s information to various internal and external threats. Cyber-based threats to information systems and cyber-related critical infrastructure can come from sources internal and external to the organization. Internal threats include errors as well as fraudulent or malevolent acts by employees or contractors working within an organization. External threats include the ever-growing number of cyber- based attacks that can come from a variety of sources such as hackers, criminals, and foreign nations. Potential attackers have a variety of techniques at their disposal, which can vastly enhance the reach and impact of their actions. For example, cyber attackers do not need to be physically close to their targets, their attacks can easily cross state and national borders, and cyber attackers can preserve their anonymity. Further, the interconnectivity among information systems presents increasing opportunities for such attacks. Indeed, reports of security incidents from federal agencies are on the rise, increasing by more than 650 percent from fiscal year 2006 to fiscal year 2010. Specifically, the number of incidents reported by federal agencies to the United States Computer Emergency Readiness Team (US-CERT) has increased dramatically over the past 4 years: from 5,503 incidents reported in fiscal year 2006 to about 41,776 incidents in fiscal year 2010. Compounding the growing number and kinds of threats are the deficiencies in security controls on the information systems at federal agencies, which have resulted in vulnerabilities in both financial and nonfinancial systems and information. These deficiencies continue to place assets at risk of inadvertent or deliberate misuse, financial information at risk of unauthorized modification or destruction, and critical operations at risk of disruption. Accordingly, we have designated information security as a governmentwide high risk area since 1997, a designation that remains in force today. The Federal Information Security Management Act (FISMA) requires each agency to develop, document, and implement an agencywide information security program to provide information security for the information and systems that support the operations and assets of the entities, using a risk-based approach to information security management. FDIC was created by Congress to maintain the stability of and public confidence in the nation’s financial system by insuring deposits, examining and supervising financial institutions, and resolving troubled institutions. Congress created FDIC in 1933 in response to the thousands of bank failures that had occurred throughout the late 1920s and early 1930s. FDIC identifies, monitors, and addresses risks to the Deposit Insurance Fund when a bank or thrift institution fails. The Bank Insurance Fund and the Savings Association Insurance Fund were established as FDIC responsibilities under the Financial Institutions Reform, Recovery, and Enforcement Act of 1989, which sought to reform, recapitalize, and consolidate the federal deposit insurance system. The act also designated FDIC as the administrator of the Federal Savings & Loan Insurance Corporation Resolution Fund, which was created to complete the affairs of the former Federal Savings & Loan Insurance Corporation and liquidate the assets and liabilities transferred from the former Resolution Trust Corporation. The Bank Insurance Fund and the Savings Association Insurance Fund merged into the Deposit Insurance Fund on February 8, 2006, as a result of the passage of the Federal Deposit Insurance Reform Act of 2005. FDIC relies extensively on computerized systems to support its mission, including financial operations, and to store the sensitive information that it collects. The corporation uses local and wide area networks to interconnect its systems and a layered approach to security defense. To support its financial management functions, FDIC relies on many systems, including a corporatewide system that functions as a unified set of financial and payroll systems that are managed and operated in an integrated fashion, a system to calculate and collect FDIC deposit insurance premiums and Financing Corporation bond principal and interest amounts from insured financial institutions; a Web-based application that provides full functionality to support franchise marketing, asset marketing, and asset management; a system to request access to and receive permission for the computer applications and resources available to its employees, contractors, and other authorized personnel; and a primary receivership and subsidiary financial processing and reporting system. FDIC also relies on other computerized systems in deriving its estimates of losses from loss-sharing agreements. This complex estimation process was developed and implemented in order to manage the significant number of loss-sharing agreements that have been created as a result of the current financial crisis. The process uses databases containing information on loss-sharing agreements and asset valuations, software programs that use information from the databases and other sources to calculate the estimated losses, data and programs stored in FDIC’s document sharing system, a Web service used to exchange valuation information with outside contractors, and several manual processing steps. In addition, in order to reduce the risk that a material misstatement will not be detected, FDIC relies heavily on supervisory review and oversight controls in the process. We have previously reported that this process is complex, is not fully documented, and involves multiple manual data entries. In a separate report, we have made an additional recommendation to FDIC to improve the documentation around this process. Under FISMA, the Chairman of FDIC is responsible for, among other things, (1) providing information security protections commensurate with the risk and magnitude of the harm resulting from unauthorized access, use, disclosure, disruption, modification, or destruction of the entity’s information systems and information; (2) ensuring that senior agency officials provide information security for the information and information systems that support the operations and assets under their control; and (3) delegating to the corporation’s Chief Information Officer the authority to ensure compliance with the requirements imposed on the agency under FISMA. The Chief Information Officer is responsible for developing and maintaining a corporatewide information security program and for developing and maintaining information security policies, procedures, and control techniques that address all applicable requirements. The Chief Information Officer also serves as the authorizing official with the authority to approve the operation of the information systems at an acceptable level of risk to the corporation. The Chief Information Security Officer reports to the Chief Information Officer and serves as the Chief Information Officer’s designated representative. The Chief Information Security Officer is responsible for the overall support of assessment and authorization activities; for the development, coordination, and implementation of FDIC’s security policy; and for the coordination of information security and privacy efforts across the corporation. Although FDIC had implemented numerous controls over its systems, it had not always implemented access and other controls to protect the confidentiality, integrity, and availability of its financial systems and information. A key reason for these weaknesses is that the corporation did not always fully implement key information security program activities, such as effectively developing and implementing security policies. Although these weaknesses did not individually or collectively constitute a material weakness or significant deficiency in 2010, they still increase the risk that financial and other sensitive information could be disclosed or modified without authorization. A basic management objective for any organization is to protect the resources that support its critical operations and assets from unauthorized access. Organizations accomplish this by designing and implementing controls that are intended to prevent, limit, and detect unauthorized access to computer resources (e.g., data, programs, equipment, and facilities), thereby protecting them from unauthorized disclosure, modification, and loss. Specific access controls include system boundary protections, identification and authentication of users, authorization restrictions, cryptography, protection of sensitive system resources, and audit and monitoring procedures. Without adequate access controls, unauthorized individuals, including intruders and former employees, can surreptitiously read and copy sensitive data and make undetected changes or deletions for malicious purposes or for personal gain. In addition, authorized users could intentionally or unintentionally modify or delete data or execute changes that are outside of their authority. Boundary protection controls logical connectivity into and out of networks and controls connectivity to and from network-connected devices. Any connections to the Internet or to other external and internal networks or information systems should occur through controlled interfaces (for example, proxies, gateways, routers and switches, firewalls, and concentrators). Many networked systems allow remote access to the information systems from virtually any remote location; thus, it is imperative that remote access paths be appropriately controlled and protected using a method such as a virtual private network (VPN). In addition, networks should also be appropriately configured to adequately protect access paths between systems; this can be accomplished through the use of access control lists and firewalls. National Institute of Standards and Technology (NIST) guidance states that agencies should establish trusted communication paths between users and the agency’s information systems, that firewalls should be configured to provide adequate protection for the organization’s networks, and that the information transmitted between interconnected systems should be controlled and regulated. FDIC had not always controlled the logical and physical boundaries protecting its information and systems. Examples are as follows:  Certain network devices, servers, and workstations on FDIC’s internal network were not always configured to sufficiently restrict access or to fully secure connections.  Firewalls controlling traffic between segments of FDIC’s internal network did not sufficiently control certain types of network traffic.  Boundary protection controls were configured in a manner that limited the effectiveness of monitoring controls. As a result of these deficiencies, FDIC faces an increased risk that individuals could gain unauthorized access to its financial systems and information. A computer system must be able to identify and authenticate the identity of a user so that activities on the system can be linked to that specific individual and to protect the system from inadvertent or malicious access. When an organization assigns a unique user account to a specific user, the system is able to distinguish that user from others—a process called identification. The system must also establish the validity of the user’s claimed identity by requesting some kind of information, such as a password, which is known only by the user—a process called authentication. NIST guidance states that an organization should manage information system authenticators by changing the default content of authenticators (e.g., passwords) when installing an information system. Also, FDIC policy states that passwords should be changed periodically. FDIC had effectively implemented controls for identifying and authenticating users on certain systems. For example, it had implemented controls to effectively detect and change default vendor-supplied user accounts and passwords in installed software and had ensured that passwords for privileged accounts on certain servers were changed in accordance with its policy. However, FDIC had not consistently enforced other identification and authentication user controls. Examples are as follows:  Passwords for certain privileged accounts on a system supporting financial processing were not configured in accordance with FDIC policy. Additionally, two of the accounts were using the same password.  Password settings for certain accounts on a system supporting the loss-share loss estimation process were not configured in accordance with FDIC policy.  Systems supporting financial processing were not always configured with sufficiently strong identification and authentication controls. As a result of these deficiencies, FDIC is at an increased risk that an individual with malicious intentions could gain inappropriate access to its financial systems and information. Authorization is the process of granting or denying access rights and privileges to a protected resource, such as a network, system, application, function, or file. A key component of granting or denying access rights is the concept of “least privilege,” which refers to granting a user only the access rights and permissions needed to perform official duties. To restrict a legitimate user’s access to only those programs and files needed, organizations establish user access rights: allowable actions that can be assigned to a user or to groups of users. File and directory permissions are rules that are associated with a particular file or directory, regulating which users can access it—and the extent of their access rights. To avoid unintentionally giving a user unnecessary access to sensitive files and directories, an organization should give careful consideration to its assignment of rights and permissions. NIST guidance states that access to information systems should be allowed only for authorized users and only for the tasks necessary to accomplish the work, in accordance with the organization’s missions and business functions. In addition, NIST guidance states that agency information systems should separate user functionality from functions necessary to administer databases, network components, workstations, or servers. FDIC policy requires that the access to information technology (IT) resources be periodically reviewed to ensure that access controls remain consistent with existing authorizations and current business needs. Also, the Division of Resolutions and Receiverships requires user access to the document sharing system supporting the loss-share estimation process to be reviewed every 3 months. FDIC had implemented controls to restrict user access to certain resources. For example, it had configured access control lists on servers dedicated to network management to restrict access to only those users who required it, controlled access to sensitive files of critical network devices, and limited user access rights to a business application supporting resolution and receivership activities to only those roles necessary for personnel to perform their duties. However, other deficiencies in authorization controls placed FDIC’s financial information and systems at risk. Examples are as follows:  The Division of Resolutions and Receiverships had not documented a procedure describing how access to the Web service used in the loss- share loss estimation process was to be reviewed, including requirements for conducting reviews at regular intervals or retaining documentation of reviews.  The Division of Resolutions and Receiverships had not reviewed access to the document sharing system every 3 months in accordance with its policy; instead, it had conducted a review only once during 2010.  FDIC had given users access to sensitive resources on certain systems supporting financial processing that they did not need to accomplish their work. As a result, FDIC faces an increased risk that a user could gain inappropriate access to computer resources, circumvent security controls, and deliberately or inadvertently read, modify, or delete financial information and other sensitive information. Cryptography underlies many of the mechanisms used to enforce the confidentiality and integrity of sensitive information. A basic element of cryptography is encryption. Encryption can be used to provide basic data confidentiality and integrity by transforming plain text into cipher text using a special value known as a key and a mathematical process known as an algorithm. If encryption is not used, user identification (ID) and password combinations will be susceptible to electronic eavesdropping by devices on the network when they are transmitted. The National Security Agency and NIST recommend encrypting network services, and NIST guidance states that passwords should be encrypted while being stored and transmitted. NIST guidance also states that the use of encryption by organizations can reduce the probability of unauthorized disclosure of information and that government systems should use sufficiently strong encryption in order to establish and maintain secure communication links between information systems and applications. FDIC had implemented controls to encrypt certain sensitive information on its systems. For example, it had restricted the use of unencrypted protocols on the mainframe and had required that sensitive information stored on user workstations or mobile devices be encrypted. However, FDIC had not always ensured that sensitive financial information transmitted over and stored on its network was adequately encrypted. Specifically, FDIC had not always used sufficiently strong encryption on two systems supporting the loss-share loss estimation process and had not always strongly encrypted stored passwords on certain financial systems. As a result of these deficiencies, FDIC is at an increased risk that an individual could capture information such as user IDs and passwords and use them to gain unauthorized access to data and system resources. To establish individual accountability, monitor compliance with security policies, and investigate security violations, the capability to determine what, when, and by whom specific actions have been taken on a system is needed. Organizations accomplish this by implementing system or security software that provides an audit trail for determining the source of a transaction or attempted transaction and by monitoring user activity. To be effective, organizations should (1) configure the software to collect and maintain a sufficient audit trail for security-relevant events; (2) generate reports that selectively identify unauthorized, unusual, and sensitive access activity; and (3) regularly monitor and take action on these reports. NIST guidance states that organizations should track and monitor access by individuals who use elevated access privileges, review and analyze information system audit records for indications of inappropriate or unusual activity, and report the findings to designated organization officials. FDIC had ensured that default installation user accounts were no longer used on certain servers and had configured its mainframe logging controls efficiently. However, FDIC’s audit and monitoring of security- relevant events on key financial systems was not always sufficient. For example, FDIC had not always sufficiently configured logging controls on a system that supported the loss-share loss estimation process or on several network devices. As a result of these deficiencies, FDIC faces an increased risk that unauthorized activity or a policy violation on its systems and networks would not be detected. In addition to access controls, organizations should use policies, procedures, and techniques for securely segregating incompatible duties, configuring information systems, and ensuring continuity of computer processing operations in the event of a disaster or unexpected interruption to ensure the confidentiality, integrity, and availability of its information. However, FDIC’s systems were not always in full compliance with these policies, procedures, and techniques, leaving them vulnerable to intrusions. Segregation of duties refers to the policies, procedures, and organizational structure that help ensure that one individual cannot independently control all key aspects of a process or computer-related operation and thereby gain unauthorized access to assets or records. Often, segregation of incompatible duties is achieved by dividing responsibilities among two or more organizational groups, which diminishes the likelihood that errors and wrongful acts will go undetected because the activities of one individual or group will serve as a check on the activities of the other. Inadequate segregation of duties increases the risk that erroneous or fraudulent transactions could be processed, improper program changes implemented, and computer resources damaged or destroyed. According to NIST, in order to maintain separation of duties, personnel who administer access control functions should not also be responsible for administering audit functions. FDIC’s Division of Resolutions and Receiverships had not always separated audit responsibilities from administration of access to loss- share and asset valuation data and programs. Specifically, the FDIC access administrators for both the external Web service and the document sharing system used in the loss-share loss estimation process were also responsible for approving and reviewing user access to the systems. As a result, the access administrators had the ability to grant inappropriate levels of access to loss-share and asset valuation data and programs without being detected, placing the data and programs at risk of unauthorized access, misuse, modification, or destruction. Configuration management is another important control that involves the identification and management of security features for all hardware and software components of an information system at a given point and systematically controls changes to that configuration during the system’s life cycle. An effective configuration management process includes procedures for (1) identifying, documenting, and assigning unique identifiers (for example, serial number and name) to a system’s hardware and software parts and subparts, generally referred to as configuration items; (2) evaluating and deciding whether to approve changes to a system’s baseline configuration; (3) documenting and reporting on the status of configuration items as a system evolves; (4) determining alignment between the actual system and the documentation describing it; and (5) developing and implementing a configuration management plan for each system. In addition, establishing controls over the modification of information system components and related documentation helps to prevent unauthorized changes and ensure that only authorized systems and related program modifications are implemented. This is accomplished by instituting policies, procedures, and techniques that help make sure all hardware, software, and firmware programs and program modifications are properly authorized, tested, and approved. According to NIST, organizations should document approved configuration-controlled changes to information systems, retain and review records of the changes, audit activities associated with the changes, and coordinate and provide oversight for configuration change control activities through a mechanism such as a change control board. NIST also recommends that agencies configure their systems to reflect the most restrictive mode possible consistent with operational requirements and employ malicious code protection mechanisms to detect and eradicate malicious code transported by electronic mail, electronic mail attachments, or other common means. FDIC had not applied appropriate configuration management controls to many of the special purpose programs and data in the loss-share estimating process. Although FDIC had documented activities for development, testing, and production for three of the programs used to calculate the estimates of losses due to loss-sharing agreements and had assigned responsibility for the different activities, it had neither documented approved changes to the programs prior to implementation nor retained records of the changes made. While the corporation had documented plans for tracking changes to these three programs, the plans had not been implemented. Additionally, the corporation had not documented plans for controlling changes to a program that generated a key dataset or to two other programs used to validate the data contained in a key database used in the loss-share loss estimation process. Furthermore, FDIC had not applied version control or change control to the database for the loss-share cost estimates. Moreover, a workstation used to execute one of the key calculation programs had configuration weaknesses that could allow it to be compromised. Until FDIC fully implements configuration management and configuration change controls to these data and programs, increased risk exists that changes to the programs could be unnecessary, may not work as intended, or may result in the unintentional loss of data or program integrity, or that individuals, both internal and external to the corporation, could exploit configuration weaknesses and gain unauthorized access to financial or other sensitive data and systems. Patch management is a critical process that can help alleviate many of the challenges in securing computing systems. Malicious acts can range from defacing a Web site to taking control of an entire system, thereby being able to read, modify, or delete sensitive information; disrupt operations; or launch attacks against other organizations’ systems. After a vulnerability has been validated, the software vendor may develop and test a patch or workaround to mitigate the vulnerability. Incident response groups and software vendors issue regular information updates on the vulnerability and the availability of patches. NIST guidance states that a comprehensive patch management process should include prioritization of the order in which vulnerabilities are addressed, with a focus on high- priority systems such as those essential for mission-critical operations. FDIC had patched many of its systems and had ensured that much of its software was up-to-date. For example, it had retired critical network devices that were not supported by their manufacturers, updated patch levels for third-party software running on two UNIX servers, and removed an obsolete version of third-party software running on a Windows server. However, FDIC had not consistently updated its financial systems and servers with critical patches or kept its software up-to-date, including systems supporting the loss-share loss estimation process. For example, certain servers supporting financial processing were running a version of software that was unsupported for patch updates, and several workstations used in the loss-share loss estimation process were missing patches and were running software that was no longer supported by the manufacturer. Additionally, certain workstations were missing operating system patches. As a result of these deficiencies, FDIC is at an increased risk that unpatched vulnerabilities could allow its information and information systems to be compromised. Contingency planning, which includes developing contingency, business continuity, and disaster recovery plans, should be performed to ensure that when unexpected events occur, essential operations can continue without interruption or can be promptly resumed, and that sensitive data are protected. NIST guidance states that organizations should develop and implement contingency plans that describe activities associated with backing up and restoring the system after a disruption or failure. The plans should be updated and include information such as contact, resources, and description of files in order to restore the application in the event of a disaster. In addition, the plans should be tested to determine their effectiveness and the organization’s readiness to execute the plans. Officials should review the test results and initiate corrective actions. FDIC’s Information Technology Security Risk Management Program requires contingency plans and disaster recovery plans to be developed and tested for all sensitive applications (both major and nonmajor) and general support systems; the plans should address measures to be taken in response to a disruption in availability due to an unplanned outage. Although FDIC had developed contingency plans for its major systems and had also conducted testing on these plans, it had not documented plans for recovering the automated and semiautomated processes supporting the loss-share loss estimation process. Although the security plan for one of FDIC’s general support systems included the document sharing system and one of the key databases supporting the process, the corporation had not documented or tested contingency plans that addressed restoring the computer programs, workstations, and datasets supporting the preparations of the estimates of losses and costs due to loss-sharing agreements or of the workspaces within the document sharing system where loss-share and asset valuation information and programs are stored. As a result, FDIC may not be able to effectively recover the data and programs in the loss-share loss estimation process and resume normal operations after a disruption. An underlying reason for the information security weaknesses noted in the previous section is that, while FDIC has developed and documented a comprehensive corporate information security program, including documenting an information security risk management policy, developing security policies and procedures, documenting system security plans, and periodically testing information security controls, the corporation had not fully implemented its information security program. Specifically, it had not fully implemented its security policies and had not completed actions to remediate certain control weaknesses. In addition, FDIC had not applied security management controls to the programs and data in the loss-share loss estimation process. An entitywide information security management program is the foundation of a security control structure and a reflection of senior management’s commitment to addressing security risks. The security management program should establish a framework and continuous cycle of activity for assessing risk, developing and implementing effective security procedures, and monitoring the effectiveness of these procedures. Without a well-designed program, security controls may be inadequate; responsibilities may be unclear, misunderstood, or improperly implemented; and controls may be inconsistently applied. FISMA requires each agency to develop, document, and implement an information security program that, among other things, includes  periodic assessments of the risk and magnitude of harm that could result from the unauthorized access, use, disclosure, disruption, modification, or destruction of information and information systems;  policies and procedures that (1) are based on risk assessments, (2) cost effectively reduce information security risks to an acceptable level, (3) ensure that information security is addressed throughout the life cycle of each system, and (4) ensure compliance with applicable requirements;  plans for providing adequate information security for networks, facilities, and systems;  periodic testing and evaluation of the effectiveness of information security policies, procedures, and practices, to be performed with a frequency depending on risk, but no less than annually, and that includes testing of management, operational, and technical controls for every system identified in the agency’s required inventory of major information systems; and  a process for planning, implementing, evaluating, and documenting remedial actions to address any deficiencies in its information security policies, procedures, or practices. FDIC had developed and documented a comprehensive corporate information security program that was consistent with FISMA requirements and had implemented some elements of its program, but had not fully implemented other elements. Specifically:  FDIC had developed and documented an IT security risk management policy that required all sensitive applications to periodically be assessed for the risk and magnitude of harm that could result from vulnerabilities and potential threats.  FDIC had not fully implemented its policies requiring that users be provided with only the minimum level of access required to allow them to perform their duties and that its computer security information response team monitor the progress of security patching activities by reviewing reports on the status of implementation. In addition, it had not fully implemented its policies for frequency of password changes and for storage of passwords.  FDIC had developed and documented security plans for all of the major systems we reviewed that addressed policies and procedures for providing management, operational, and technical controls, and had documented requirements for physically securing FDIC facilities.  FDIC had conducted annual periodic testing and evaluation of the effectiveness of the management, operational, and technical controls for the major systems we reviewed.  Although FDIC had established a process for planning, implementing, evaluating, and documenting remedial actions to address information security weaknesses, and had completed actions to remediate 26 of the 33 control weaknesses we identified in our calendar year 2009 audit, the corporation had not yet completed actions to correct or mitigate 7 of the previously reported weaknesses. For example, FDIC had not separated or partitioned the data network from the voice network, developed and documented policies and procedures for assigning access to systems and databases where application controls could be compromised, or fully implemented its monitoring program. In addition, FDIC had not received an independent audit report from the provider of its Web service in a timely manner. FISMA information security requirements apply not only to an agency’s own systems but also to information systems used or operated on its behalf by a contractor or other agency, such as an external service provider. According to OMB, service providers are required to provide client organizations with an audit report that describes whether internal controls were designed to achieve specified objectives, have been placed into operation, and are operating effectively. Previously known as Statement on Auditing Standards (SAS) 70 reports, since June 15, 2011, they have been known as Statement on Standards for Attestation Engagements (SSAE) 16 reports. OMB also states that such reports should be provided within a reasonable time frame so that auditors of client organizations may use them during their financial statement audits. However, the provider of the Web service used to exchange information with valuation contractors did not provide FDIC with a SAS 70 report until March 2011, more than 8 weeks after the end of the financial reporting period and more than 5 months after the end of the period that the SAS 70 audit covered. Until all key elements of its information security program are fully implemented, FDIC may not have assurance that controls over its financial systems and information are appropriately designed and operating effectively. FDIC had not applied key controls in its information security program to the loss-share loss estimation process. OMB Circular A-130, Appendix III, requires federal agencies to implement and maintain an automated information security program, including planning for adequate security of each system, assessing risks, and reviewing security controls. OMB Circular A-127 requires that federal financial management systems, which include core financial systems as well as any automated and manual processes, procedures, data, hardware, and software that support financial management, be subject to the requirements of Circular A-130. However, FDIC had not applied key controls in its information security program to the automated and semiautomated processes used to support the preparation of the estimates of losses and costs due to loss- sharing agreements. Specifically, FDIC had not  assessed the risks associated with the information and programs involved to identify potential threats and vulnerabilities as well as possible countermeasures and mitigating controls, and had not included the programs in the risk assessment of any of its general support systems;  documented the management, technical, or operational security controls intended to protect the programs in system security plans, and had not included the programs in the system security plans of any general support system; or tested any security controls for the programs, and had not included the programs when testing the security controls of other general support systems. FDIC had not applied these controls because the Division of Resolutions and Receiverships developed the process independently, in order to be able to manage the large increase in bank failures and the extensive use of loss-sharing agreements resulting from the current financial crisis. In doing so, the Division of Resolutions and Receiverships had not used FDIC’s existing IT management framework—which requires these controls to be put into place—to develop and manage the process. During 2010, FDIC had mitigated the effect of these weaknesses on financial reporting by implementing compensating management and reconciliation controls in this process. However, because of ongoing financial institution failures and the lack of information security management controls around the process, the financial information processed by the programs involved—representing a nearly $39 billion impact on the corporation’s financial statements—continues to be at risk of unauthorized disclosure, modification, or destruction. FDIC has made significant progress in correcting or mitigating previously reported information security weaknesses, but other control weaknesses continue to unnecessarily put FDIC’s systems at an increased risk from internal and external threats. A key reason for these weaknesses is that the corporation had not fully implemented key elements of its information security program, such as effectively implementing security policies, conducting risk assessments, documenting security management plans, documenting contingency plans, testing security controls, or implementing an effective continuous monitoring program. FDIC had made improvements in its information security controls and had mitigated the potential effect of its remaining weaknesses on financial reporting by implementing compensating management and reconciliation controls during 2010, enabling us to conclude that FDIC had resolved the significant deficiency over information systems that we had reported in our 2009 audit. However, the weaknesses—both old and new—continue to challenge the corporation in its efforts to ensure the confidentiality, integrity, and availability of financial and sensitive information. Until FDIC further mitigates known information security weaknesses in access controls and other information system controls and fully implements its information security program, the corporation will continue to face an increased risk that sensitive financial information and resources will not be sufficiently protected from inadvertent or deliberate misuse, improper disclosure, or destruction. We recommend that the Acting Chairman take the following two actions to enhance FDIC’s information security program:  Direct the Director of the Division of Resolutions and Receiverships and the Chief Information Officer to develop, document, and implement appropriate information security activities in the loss-share loss estimation process, such as assessing and mitigating risks, managing and controlling the configurations of programs and databases, evaluating the effectiveness of security controls, and ensuring that data and programs can be recovered after a disruption.  Direct the Chief Information Officer to work with the external Web service provider to obtain a more timely delivery of the provider’s SSAE 16 report (previously known as a SAS 70 report), or to obtain other means of assurance of internal controls. We are also making 38 new recommendations to address 37 new findings in a separate report with limited distribution. These recommendations consist of actions to implement and correct specific information security weaknesses related to access controls, segregation of duties, configuration management, and contingency planning identified during this audit. In providing written comments (reprinted in app. II) on a draft of this report, the Deputy to the Chairman and Chief Financial Officer of FDIC stated that FDIC was pleased to accept our acknowledgment of the significant progress made toward correcting and mitigating our previously reported weaknesses. In addition, he indicated that the corporation plans to implement improvements to address our recommendations, and discussed the actions that FDIC has taken or plans to take to review and improve controls over the loss-share loss estimation process, to obtain timely delivery of appropriate audit reports from current and future service providers, and to conduct additional due diligence activities to obtain assurance of the service provider’s internal controls. In responding to our draft recommendation that FDIC develop, document, and implement appropriate information security controls over the automated and semiautomated processes within the loss-share loss estimation process, the Deputy to the Chairman stated that although FDIC agrees that the loss-share business processes and the data associated with these processes deserve proper controls assessment and protection, the corporation will not necessarily treat the processes and data as a separate FDIC system. The Deputy to the Chairman further stated that FDIC is currently taking steps to improve the information security controls around the process. The intent of our draft recommendation was not to suggest that FDIC treat the data and programs supporting the loss-share loss estimation process as a separate information system. We agree that it may not be appropriate for FDIC to treat these data and programs as a separate information system, as they are stored, processed, and executed across multiple systems. Rather, our intent was to recommend that appropriate information security control activities be incorporated into the process. Accordingly, we have clarified our recommendation to state that the Acting Chairman direct the Director of the Division of Resolutions and Receiverships and the Chief Information Officer to develop, document, and implement appropriate information security activities in the loss-share loss estimation process, such as assessing and mitigating risks, managing and controlling the configurations of programs and databases, evaluating the effectiveness of security controls, and ensuring that data and programs can be recovered after a disruption. We are sending copies of this report to the Chairman and Ranking Member of the Senate Committee on Banking, Housing, and Urban Affairs; Chairman and Ranking Member of the House Financial Services Committee; and other interested parties. In addition, this report will be available at no charge on the GAO Web site at http://www.gao.gov. If you have any questions regarding this report, please contact Gregory C. Wilshusen at (202) 512-6244 or Dr. Nabajyoti Barkakati at (202) 512- 4499. We can also be reached by e-mail at wilshuseng@gao.gov and barkakatin@gao.gov. Key contributors to this report are listed in appendix III. The objective of our audit was to determine the effectiveness of the Federal Deposit Insurance Corporation’s (FDIC) controls protecting the confidentiality, integrity, and availability of its financial systems and information. To do this, we examined FDIC information security policies, plans, and procedures; tested controls over key financial applications; and interviewed key agency officials in order to (1) assess the effectiveness of corrective actions taken by FDIC to address weaknesses we previously reported and (2) determine whether any additional weaknesses existed. This work was performed in support of our opinion on internal control over the preparation of the calendar year 2010 and 2009 financial statements of two funds administered by FDIC. To determine whether controls over key financial systems were effective, we considered the results of our evaluation of FDIC’s actions to mitigate previously reported weaknesses and performed new audit work at FDIC facilities in Arlington, Virginia, and Washington, D.C. We concentrated our evaluation primarily on the controls for financial applications and enterprise database applications associated with the New Financial Environment; the Assessment Information Management System; the Communication, Capability, Challenge, and Control System (4C) application; the programs, data, and systems supporting the preparation of the estimates of losses and costs due to loss-sharing agreements, and the general support systems. Our selection of the systems to evaluate was based on consideration of systems that directly or indirectly support the processing of material transactions that are reflected in the funds’ financial statements. Our evaluation was based on GAO’s Federal Information System Controls Audit Manual, which contains guidance for reviewing information system controls that affect the confidentiality, integrity, and availability of computerized information. Using National Institute of Standards and Technology (NIST) standards and guidance and FDIC’s policies, procedures, practices, and standards, we evaluated controls by  observing methods for providing secure data transmissions across the network to determine whether sensitive data were being encrypted; testing and observing physical access controls to determine if computer facilities and resources were being protected from espionage, sabotage, damage, and theft;  evaluating the control configurations of selected servers and database inspecting key servers and workstations to determine whether critical patches had been installed or were up-to-date; and  examining access responsibilities to determine whether incompatible functions were segregated among different individuals. Using the requirements of the Federal Information Security Management Act (FISMA), which establishes key elements for an effective agencywide information security program, we evaluated FDIC’s implementation of its security program by reviewing FDIC’s risk assessment process and risk assessments for key FDIC systems that support the preparation of financial statements to determine whether risks and threats were documented consistent with federal guidance;  analyzing FDIC’s policies, procedures, practices, and standards to determine their effectiveness in providing guidance to personnel responsible for securing information and information systems;  analyzing security plans to determine if management, operational, and technical controls were in place or planned and that security plans were updated;  analyzing security testing and evaluation results for six key FDIC systems to determine whether management, operational, and technical controls were tested at least annually and based on risk; and  examining remedial action plans to determine whether they addressed vulnerabilities identified in FDIC’s security testing and evaluations. We also discussed with key security representatives and management officials whether information security controls were in place, adequately designed, and operating effectively. To determine the status of FDIC’s actions to correct or mitigate previously reported information security weaknesses, we identified and reviewed its information security policies, procedures, and guidance. We reviewed prior GAO reports to identify previously reported weaknesses and examined FDIC’s corrective action plans to determine which weaknesses FDIC reported as being corrected. For those instances where FDIC reported it had completed corrective actions, we assessed the effectiveness of those actions. We conducted this audit from November 2010 to August 2011, in accordance with generally accepted government auditing standards. We conducted our data collection, analysis, and assessment procedures in support of the financial audit between November 2010 and March 2011. We conducted supplemental audit procedures to prepare this report from March 2011 to August 2011. The generally accepted government auditing standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objective. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objective. In addition to the individuals named above, Lon Chin, David Hayes, Charles Vrabel, and Christopher Warweg, Assistant Directors; Gary Austin; Angela Bell; William Cook; Saar Dagani; Nancy Glover; Rosanna Guerrero; Jason Porter; Michael Stevens; and Shaunyce Wallace made key contributions to this report.

The Federal Deposit Insurance Corporation (FDIC) has a demanding responsibility enforcing banking laws, regulating financial institutions, and protecting depositors. Because of the importance of FDIC's work, effective information security controls are essential to ensure that the corporation's systems and information are adequately protected from inadvertent misuse, fraudulent use, or improper disclosure. As part of its audits of the 2010 financial statements of the Deposit Insurance Fund and the Federal Savings & Loan Insurance Corporation Resolution Fund administrated by FDIC, GAO assessed the effectiveness of the corporation's controls in protecting the confidentiality, integrity, and availability of its financial systems and information. To perform the audit, GAO examined security policies, procedures, reports, and other documents; tested controls over key financial applications; and interviewed key FDIC personnel. Although FDIC had implemented numerous controls in its systems, it had not always implemented access and other controls to protect the confidentiality, integrity, and availability of its financial systems and information. FDIC has implemented controls to detect and change default user accounts and passwords in vendor-supplied software, restricted access to network management servers, developed and tested contingency plans for major systems, and improved mainframe logging controls. However, the corporation had not always (1) required strong passwords on financial systems and databases; (2) reviewed user access to financial information in its document sharing system in accordance with policy; (3) encrypted financial information transmitted over and stored on its network; and (4) protected powerful database accounts and privileges from unauthorized use. In addition, other weaknesses existed in FDIC's controls that were intended to appropriately segregate incompatible duties, manage system configurations, and implement patches. An underlying reason for the information security weaknesses is that FDIC had not always implemented key information security program activities. To its credit, FDIC had developed and documented a security program and had completed actions to correct or mitigate 26 of the 33 information security weaknesses that were previously identified by GAO. However, the corporation had not assessed risks, documented security controls, or performed periodic testing on the programs and data used to support the estimates of losses and costs associated with the servicing and disposal of the assets of failed institutions. Additionally, FDIC had not always implemented its policies for restricting user access or for monitoring the progress of security patch installation. Because FDIC had made progress in correcting or mitigating previously reported weaknesses and had implemented compensating management and reconciliation controls during 2010, GAO concluded that FDIC had resolved the significant deficiency in internal control over financial reporting related to information security reported in GAO's 2009 audit, and that the remaining unresolved issues and the new issues identified did not individually or collectively constitute a material weakness or significant deficiency in 2010. However, if left unaddressed, these issues will continue to increase FDIC's risk that its sensitive and financial information will be subject to unauthorized disclosure, modification, or destruction. GAO recommends that FDIC take two actions to enhance its comprehensive information security program. In commenting on a draft of this report, FDIC discussed actions that it has taken or plans to take to address these recommendations.

Mass transit and passenger rail systems provided 10.7 billion passenger trips in the United States in fiscal year 2008. The nation’s mass transit and passenger rail systems include all multiple-occupancy vehicle services designed to transport customers on local and regional routes, such as transit buses, heavy rail, commuter rail, and light rail services, and the interconnected facilities and vehicles feeding into the transit systems. Buses are the most widely used form of transit, providing almost two- thirds of all passenger trips. Heavy rail systems––subway systems like New York City’s transit system and Washington, D.C.’s Metro—typically operate on fixed rail lines within a metropolitan area and have the capacity for a heavy volume of traffic. Commuter rail systems typically operate on railroad tracks and provide regional service (e.g., between a central city and adjacent suburbs). Light rail systems are typically characterized by lightweight passenger rail cars that operate on track that is not separated from vehicular traffic for much of the way. Mass transit and passenger rail systems in the United States are typically owned and operated by public sector entities, such as state and regional transportation authorities. Amtrak, which reported that it provided 25.8 million passenger trips in fiscal year 2007, operates the nation’s primary intercity passenger rail and serves more than 500 stations in 46 states and the District of Columbia. Amtrak operates over a 22,000 mile network, primarily over leased freight railroad tracks. In addition to leased tracks, Amtrak owns about 650 miles of track, primarily on the “Northeast Corridor” between Boston and Washington, D.C., which carries about two-thirds of Amtrak’s total ridership. Stations are owned by Amtrak, freight carriers, municipalities, and some private entities. Amtrak also operates commuter rail services in certain jurisdictions on behalf of state and regional transportation authorities. Figure 1 identifies the geographic distribution of rail transit systems and Amtrak within the United States. Though not indicated on the map, all of these cities also have bus transit systems. To date, U.S. mass transit and passenger rail systems have not been attacked by terrorists. However, these systems have received heightened attention as several alleged terrorists’ plots have been uncovered, including multiple plots involving systems in the New York City area. Worldwide, mass transit and passenger rail systems have been the frequent target of terrorist attacks. According to the Worldwide Incidents Tracking System maintained by the National Counter-Terrorism Center, from January 2004 to July 2008, there were 530 terrorist attacks worldwide against mass transit and passenger rail targets, resulting in over 2,000 deaths and over 9,000 injuries. Terrorist attacks include a 2007 attack on a passenger train in India (68 fatalities and over 13 injuries); the 2005 attack on London’s underground rail and bus systems (52 fatalities and over 700 injuries); and the 2004 attack on commuter rail trains in Madrid (191 fatalities and over 1,800 injuries). In January 2008, Spanish authorities arrested 14 suspected terrorists who were allegedly connected to a plot to conduct terrorist attacks in Spain, Portugal, Germany, and the United Kingdom, including an attack on the Barcelona metro subway system. The most common means of attack against mass transit and passenger rail systems has been improvised explosive devices (IED), with many of these attacks delivered by suicide bombers. According to transit agency officials, certain characteristics of mass transit and passenger rail systems make them inherently vulnerable to terrorist attacks and therefore difficult to secure. By design, mass transit and passenger rail systems are open (i.e., have multiple access points, hubs serving multiple carriers, and, in some cases, no barriers to access) so that they can move large numbers of people quickly. The openness of these systems can leave them vulnerable because operator personnel cannot completely monitor or control who enters or leaves the systems. In addition, other characteristics of mass transit and passenger rail systems—high ridership, expensive infrastructure (more so for passenger rail than bus), economic importance, and location in large metropolitan areas or tourist destinations––also make them attractive targets for terrorists because of the potential for mass casualties, economic damage and disruption. Moreover, some of these same characteristics make them difficult to secure. For example, the number of riders passing through a subway system––especially during peak hours—may make the sustained use of some security measures, such as airport style passenger screening checkpoints, difficult because the measures could disrupt scheduled service. In addition, multiple access points along extended routes may make securing each location difficult because of the costs associated with such actions. Balancing the potential economic impacts of security enhancements with the benefits of such measures is a difficult challenge. Securing the nation’s mass transit and passenger rail systems is a shared responsibility requiring coordinated action on the part of federal, state, and local governments; the private sector; and passengers who ride these systems. Since the September 11, 2001 attacks, the role of federal agencies in securing the nation’s transportation systems has continued to evolve. In response to the September 11th terrorist attacks, Congress passed the Aviation and Transportation Security Act of 2001, which created TSA within DOT and conferred to the agency broad responsibility for overseeing the security of all modes of transportation, including mass transit and passenger rail. In 2002, Congress passed the Homeland Security Act, which established DHS, transferred TSA from DOT to DHS, and assigned DHS responsibility for protecting the nation from terrorism, including securing the nation’s transportation systems. Within TSA, the office of Transportation Sector Network Management (TSNM) leads the unified effort to protect and secure the nation’s intermodal transportation systems, with divisions dedicated to each transportation mode, including mass transit and passenger rail. Within TSA’s Office of Security Operations, the Office of Multi-modal Oversight manages the Surface Transportation Security Inspection Program which coordinates with TSNM to develop and implement security programs, including strategies for conducting and implementing assessments and other actions in mass transit and passenger rail. In addition, TSA’s Office of Intelligence (TSA- OI) is responsible for collecting and analyzing threat information related to the transportation network, which includes all modes of transportation. TSA is supported in these efforts by other DHS entities such as the NPPD and the Federal Emergency Management Agency’s (FEMA) Grant Programs Directorate and Planning and Assistance Branch. The NPPD is responsible for coordinating efforts to protect the nation’s most critical assets across all 18 industry sectors, including surface transportation. FEMA’s Grant Programs Directorate is responsible for managing DHS grants for mass transit. FEMA’s Planning and Assistance Branch is responsible for assisting transit agencies with how to conduct risk assessments. TSA has issued requirements related to the security of mass transit and passenger rail systems. Specifically, in May 2004, TSA issued security directives that mandated passenger rail agencies and Amtrak to implement certain security measures, such as periodically inspecting passenger rail cars for suspicious or unattended items and reporting potential threats or significant security concerns to appropriate law enforcement authorities and TSA. In addition to these requirements, in August 2007, the 9/11 Commission Act was signed into law, which included provisions that task TSA with security actions related to mass transit and passenger rail security. Among other things, these provisions include mandates for developing and issuing reports on TSA’s strategy for securing public transportation, conducting and updating security assessments of mass transit systems, and establishing a program for conducting security exercises for transit and rail agencies. While TSA is the lead federal agency for overseeing the security of all transportation modes, DOT continues to play a key supporting role in securing mass transit and passenger rail systems. In a 2004 memorandum of understanding (MOU) and a 2005 annex to the MOU, TSA and FTA agreed that the two agencies would coordinate their programs and services, with FTA playing a supporting role by providing technical assistance and assisting DHS with implementation of its security policies, including collaborating in developing regulations affecting transportation security. In particular, FTA has played a role in coordinating and funding security training programs for mass transit and passenger rail employees, and provided dedicated funding to three federal training providers to implement mass transit and passenger rail employee training programs. Additionally, FTA administers the State Safety Oversight program and may withhold federal funding for states’ noncompliance with regulations governing state safety oversight agencies. As part of this program, state safety oversight agencies are responsible for reviewing and approving rail transit agencies’ safety and security plans, among other activities. FTA also promotes mass transit and passenger rail safety and security by providing funding for research, technical assistance, and technology demonstration projects. In addition to FTA, DOT’s FRA also has regulatory authority over commuter rail operators and Amtrak and employs over 400 inspectors who periodically monitor the implementation of safety and security plans at these systems. FRA regulations require railroads that operate intercity or commuter passenger train service or that host the operation of that service to adopt and comply with a written emergency preparedness plan approved by FRA. State and local governments, mass transit and passenger rail operators, and private industry are also important stakeholders in the nation’s mass transit and passenger rail security efforts. State and local governments, in some cases, own or operate a significant portion of mass transit and passenger rail systems. Consequently, the responsibility for responding to emergencies involving systems that run through their jurisdictions often falls to state and local governments. Although all levels of government are involved in mass transit and passenger rail security, the primary responsibility for securing the systems rests with the mass transit and passenger rail operators. These operators, which can be public or private entities, are responsible for administering and managing transit activities and services, including security. They can also directly operate the security service provided or contract for all or part of the total service. We discuss security actions taken by federal agencies and mass transit and passenger rail system operators later in this report. In recent years, we, along with the Congress, the executive branch, and the 9/11 Commission have recommended that federal agencies with homeland security responsibilities utilize a risk management approach to help ensure that finite national resources are dedicated to assets or activities considered to have the highest security priority. We have concluded that without a risk management approach, there is limited assurance that programs designed to combat terrorism would be properly prioritized and focused. Thus, risk management, as applied in the homeland security context, can help to more effectively and efficiently prepare defenses against acts of terrorism and other threats. Homeland Security Presidential Directive 7 (HSPD-7) directed the Secretary of Homeland Security to establish uniform policies, approaches, guidelines, and methodologies for integrating federal infrastructure protection and risk management activities. Recognizing that each sector possesses its own unique characteristics and risk landscape, HSPD-7 designates federal government sector specific agencies (SSA) for each of the critical infrastructure sectors that are to work with DHS to improve critical infrastructure security. On June 30, 2006, DHS released the National Infrastructure Protection Plan (NIPP), which created—in accordance with HSPD-7—a risk-based framework for the development of SSA strategic plans. As the SSA for transportation, TSA developed the Transportation Systems—Sector Specific Plan (TS-SSP) in 2007 to document the process to be used in carrying out the national strategic priorities outlined in the NIPP and the National Strategy for Transportation Security (NSTS). The TS-SSP contains supporting modal implementation plans for each transportation mode, including mass transit and passenger rail, which provides information on current efforts to secure mass transit and passenger rail, as well as TSA’s overall goals and objectives related to mass transit and passenger rail security. The NIPP defines roles and responsibilities for security partners in carrying out critical infrastructure and key resource (CI/KR) protection activities through the application of risk management principles. Figure 2 illustrates the several interrelated activities of the risk management framework as defined by the NIPP. The NIPP requires that federal agencies use this information to inform the selection of risk-based priorities and the continuous improvement of security strategies and programs to protect people and critical infrastructure by reducing the risk of acts of terrorism. Within the risk management framework, the NIPP also establishes baseline criteria for conducting risk assessments. According to the NIPP, risk assessments are a qualitative and/or quantitative determination of the likelihood of an adverse event occurring and are a critical element of the NIPP risk management framework. Risk assessments can also help decision makers identify and evaluate potential risks so that countermeasures can be designed and implemented to prevent or mitigate the potential effects of the risks. The NIPP characterizes risk assessment as a function of three elements: Threat: The likelihood that a particular asset, system, or network will suffer an attack or an incident. In the context of risk associated with a terrorist attack, the estimate of threat is based on the analysis of the intent and the capability of an adversary; in the context of a natural disaster or accident, the likelihood is based on the probability of occurrence. Vulnerability: The likelihood that a characteristic of, or flaw in, an asset, system, or network’s design, location, security posture, process, or operation renders it susceptible to destruction, incapacitation, or exploitation by terrorist or other intentional acts, mechanical failures, and natural hazards. Consequence: The negative effects on public health and safety, the economy, public confidence in institutions, and the functioning of government, both direct and indirect, that can be expected if an asset, system, or network is damaged, destroyed, or disrupted by a terrorist attack, natural disaster, or other incident. Information from the three elements that assess risk—threat, vulnerability and consequence—can lead to a risk characterization and provide input for prioritizing security goals. Since 2004, federal and industry stakeholders have conducted assessments of individual elements of risk—threat, vulnerability and consequence— and this information has informed TSA’s mass transit and passenger rail security strategy. However, TSA could strengthen its approach by using and combining this information to conduct a risk assessment of the mass transit and passenger rail system and by updating its strategy to include characteristics that we identified as desirable practices for successful national strategies and to more fully address elements that are outlined in Executive Order 13416: Strengthening Surface Transportation Security. While federal and industry stakeholders have conducted assessments of individual elements of risk—threat, vulnerability, and consequence—TSA could strengthen its security approach by using and combining this information to conduct a risk assessment of the mass transit and passenger rail system. A risk assessment, as required by the NIPP, involves assessing each of the three elements of risk and then combining them together into a single analysis. Since 2004, federal agencies have conducted a range of assessment activities related to the individual elements of risk to help determine their strategy for securing mass transit and passenger rail systems, and provided guidance to mass transit and passenger rail agencies on how to conduct assessments of individual elements of risk. For example, DHS’s threat assessments considered potential threats to the mass transit and passenger rail system, while vulnerability assessments focused on mass transit and passenger rail systems’ security conditions or specific infrastructure such as tunnels. In addition to DHS assessments, DOT provided assistance to mass transit and passenger rail agencies on how to conduct threat and vulnerability assessments, and transit agencies have reported conducting risk assessments for their own systems or assets. See table 1 for a summary of federal and industry stakeholders’ assessment activities related to individual elements of risk. As table 1 shows, DHS developed the Strategic Homeland Infrastructure Risk Assessment (SHIRA) that assessed risk across 18 CI/KR sectors. To develop SHIRA, DHS collaborated with members of the intelligence community to determine threats against various systems and assets in the 18 CI/KR sectors. TSA then assessed the vulnerabilities and consequences that resulted from these threat scenarios and provided this information to HITRAC. Although TSA contributed to DHS’s risk assessment effort, it has not conducted its own risk assessment of mass transit and passenger rail systems. TSA officials explained that the threat scenarios that SHIRA provided were general and not specific to mass transit and passenger rail. TSA could, however, use the information it provided for SHIRA to support a risk assessment. Table 1 also shows that mass transit and passenger rail agencies, including Amtrak, have reported conducting risk assessments of their own systems. Officials from 26 of 30 of the transit systems we visited stated that they had conducted their own assessments of their systems, including risk assessments. For example, one transit agency official stated that the agency had conducted risk assessments of its stations since 2003 and had updated them every 2 years. The official explained that his agency uses the risk assessment results to conduct cost benefit analyses that the agency uses before instituting new programs or purchasing equipment for its system. He also said that the assessments help the agency track risk reduction as a result of its security investments. Additionally, Amtrak officials stated that they conducted a risk assessment of all of their systems. As part of the assessment, Amtrak contracted with a private consulting firm to provide a scientific basis for identifying critical points at stations that might be vulnerable to IED attacks or that are structurally weak. Amtrak officials also stated that they gather and analyze threat information obtained from various classified and unclassified sources such as DHS, TSA, and the Federal Bureau of Investigation’s Joint Terrorism Task Force. Transit agencies have also received assistance in the form of either guidance or actual risk assessments from several federal and industry stakeholders. Table 2 identifies the various assistance programs available to transit agencies for risk assessment efforts. As table 2 shows, federal and industry stakeholders also provided assistance to transit agencies on how to assess risk. For example, FTA provided on-site technical assistance to the nation’s 50 largest transit agencies (i.e., those transit agencies with the highest ridership) on how to conduct threat and vulnerability assessments, among other technical assistance needs, through its Security and Emergency Management Technical Assistance Program (SEMTAP). According to FTA officials, although FTA continues providing technical assistance to transit agencies, the on-site SEMTAP program concluded in July 2006. Furthermore, FTA officials stated that on-site technical assistance was transferred to TSA when TSA became the lead agency on security matters for mass transit and passenger rail. Also, from 2004 though 2007, the former DHS Office of Domestic Preparedness (ODP), through a private consulting firm, provided assistance to transit agencies on how to conduct risk assessments through the Mass Transit Technical Assistance Program. Within this program, ODP developed a Transit Risk Assessment Methodology (TRAM) tool kit that provided transit agencies with an instrument to compare relative risks of terrorism against critical assets to better identify and prioritize security enhancements to reduce those risks. We reported in 2005 that officials from transit agencies participating in the Mass Transit Technical Assistance Program valued it and stated that the program was successful in helping them to devise risk-reduction strategies to guide security-related investments. Subsequently, since the restructuring of ODP in 2007, this program has been transferred to FEMA’s Planning and Assistance Branch where it has continued assisting transit agencies with risk assessments. However, according to FEMA’s Chief of the Planning and Assistance Branch, because of the high cost of the program—$300,000 to $600,000 per transit agency—the rate of assistance to transit agencies has decreased annually. Also, FEMA is trying to convert the focus of the program from technical assistance to training. As such, FEMA plans to educate transit agencies on how to conduct risk assessments. Additionally, the same FEMA official reported that FEMA is also in the process of conducting a pilot project with one transit agency to evolve the program and the tool kit to an all hazards focus. Furthermore, recognizing the value of this program, officials from four of the 30 transit agencies we interviewed have since contracted with the same private consulting firm that ODP and FEMA used to update security plans or conduct a cost-benefit analysis of new programs or equipment. TSA has reported conducting annual threat assessments of the mass transit and passenger rail systems, and these assessments have provided TSA with an array of information about potential threats to the systems. TSA is responsible for conducting and issuing an annual threat assessment report for the mass transit and passenger rail systems. While it has been widely reported that no specific threats to the mass transit and passenger rail systems currently exist, it has been noted that terrorists tend to target these systems, as overseas attacks on mass transit and passenger rail systems have demonstrated. TSA’s Mass Transit Modal Annex identified numerous potential threats to mass transit, including placing a vehicle bomb near a station or track or introducing an IED or lower-yield explosive in a station, train, or bus, or laying explosives on a track. Deploying conventional or improvised explosives would likely result in scores of casualties. Since IEDs were used in the majority of the recent overseas attacks against mass transit and passenger rail systems, TSA and other experts are concerned that extremists may be motivated to employ similar tactics to target mass transit and passenger rail systems. In its Modal Annex, TSA also noted that the threat to heavy and commuter rail (i.e., underground, subway, elevated, rapid rail, or metro) is higher than the threat to buses and light rail (i.e., street cars, trolley) because of the accessibility of the large numbers of people typically found in the confined spaces of a rail system. Several DHS components, including TSA, conducted assessments related to vulnerability and consequence since 2004, which have highlighted areas for security improvement. For example, DHS S&T conducted vulnerability assessments of transit tunnels as well as assessments of the potential consequences that various types of explosives threats would have on tunnel structures (which showed that improving evacuation plans and emergency response efforts, among other things, would improve public safety). Additionally, TSA has gathered vulnerability data through such programs as the Baseline Assessment for Security Enhancement (BASE). TSA officials reported that the BASE assesses the security posture of a mass transit or passenger rail system against the Security and Emergency Management Action Items and is TSA’s primary source of vulnerability information. For example, through initial assessments of the BASE program, TSA officials identified the need for increased security training at mass transit and passenger rail systems. Furthermore, FEMA has calculated consequence information for mass transit and passenger rail by using proxy data, such as population and national infrastructure indices. This information has been incorporated into the Transit Security Grant Program (TSGP). TSA officials also reported using population density and ridership data as information for consequence assessments for mass transit and passenger rail systems and stated that they consider the number of potential casualties when determining consequence, and as a result, have chosen to focus their security efforts on the mass transit and passenger rail systems carrying the most passengers. Officials also mentioned that other factors such as the nature of the infrastructure (underground, underwater tunnels), time of day, and number of mass transit and passenger rail lines are also considered when assessing consequence. TSA has used these various threat, vulnerability, and consequence assessments to inform its security strategy for mass transit and passenger rail—the Mass Transit Modal Annex. TSA reported that its efforts to inform its strategy included using information from TSA-OI’s annual mass transit threat assessment report to, for example, highlight the greater threats to underground and underwater passenger rail segments within a transit system. TSA also reported incorporating into its strategy information identified through its BASE reviews, such as the need for increased security training at mass transit and passenger rail systems. While TSA reported using these various assessments to inform its mass transit and passenger rail security strategy, it could further strengthen its approach for securing these systems by combining the results from these assessments to conduct a risk assessment of the mass transit and passenger rail systems. Both the NIPP and TS-SSP establish a risk management framework that includes a process for considering threat, vulnerability, and consequence assessments together to determine the likelihood of a terrorist attack and the severity of its impact. The NIPP states that after the three elements of risk have been assessed, they are factored numerically and combined mathematically to provide an estimate of the expected loss considering the likelihood of an attack or other incident. It also states that when numerical values are not practical, scales could be used to estimate threat, vulnerability, and consequence. Thus, risk can be measured either quantitatively (i.e., numerically) or qualitatively (i.e., descriptively). However, rather than using the methodology established in the NIPP for assessing risk, TSA officials stated that the agency uses an intelligence-driven approach to make strategic investment decisions across the transportation system. Within this intelligence-driven approach for the sector, TSA also developed a tactical, threat-based process known as Objectively Measured Risk Reduction (OMRR) at the program level to help each of its individual divisions manage their day-to-day security operations. These approaches differ from the NIPP in part because they rely primarily on intelligence information to identify threats, prioritize tactics, and guide long-term investments, rather than systematically assessing the vulnerabilities and consequences of a range of threat scenarios. In March 2009, we recommended that TSA work with DHS to validate its risk management approach by establishing a plan and time frame for assessing the appropriateness of TSA’s intelligence-driven risk management approach for managing risk and document the results of this review once completed. TSA concurred with this recommendation. TSA officials stated that they plan to revise and reissue the TS-SSP, as required by DHS, to reflect the adoption of their intelligence-driven methodology. As on June 2009, TSA reported that the update of the TS-SSP is ongoing, with the goal of completing the effort in 2009. Until TSA works with DHS to validate its risk management approach, TSA lacks assurance that its approach provides the agency and DHS with the information needed to guide investment decisions to ensure resources are allocated to the highest risks. Moreover, as we reported in March 2009, although intelligence is necessary to inform threat assessments, it does not provide all of the information needed to assess risk, in particular information related to vulnerability and consequence assessments. In addition, the intelligence- driven approach that TSA uses may be limited because, in contrast with practices adopted by the intelligence community, TSA officials do not plan to assign uncertainty or confidence levels to the intelligence information it uses to identify threats and guide long-range planning and strategic investment. Both Congress and the administration have recognized the uncertainty inherent in intelligence analysis and have required analytic products within the intelligence community to properly caveat and express uncertainties or confidence in analytic judgments. Furthermore, while intelligence can and does help the U.S. security community on an operational or tactical level, uncertainty in intelligence analysis limits its utility for long-range planning and strategic investment. Without expressing confidence levels in its analytic judgments, it will be difficult for TSA to correctly prioritize its tactics and long-term investments based on uncertain intelligence. In March 2009, we recommended that the Assistant Secretary of TSA work with the Director of National Intelligence to determine the best approach for assigning uncertainty or confidence levels to analytic intelligence products and apply this approach to intelligence products. TSA officials agreed that they do not have a risk assessment and expressed the desire to conduct one; however, they reported that a lack of resources and other factors made completing a risk assessment challenging. For example, TSA officials stated that comprehensive vulnerability and consequence assessments are cost-prohibitive and time-intensive to conduct. Specifically, according to TSA officials, the Security Analysis and Action Program conducted by surface inspectors, a program that identifies, among other things, transit agencies’ vulnerabilities can take days to complete resulting in a large resource investment. However, the 9/11 Commission Act requires TSA to use existing relevant assessments developed by federal and industry stakeholders, as appropriate, to develop a risk assessment for rail, including passenger rail. Furthermore, as suggested by the NIPP, agencies should consider existing risk measures when assessing risk. In addition to using the information for SHIRA, TSA could use other risk assessments, such as industry stakeholders’ risk assessments and federal and industry stakeholders’ guidance on how to conduct risk assessments, to potentially support a risk assessment of the mass transit and passenger rail systems. Despite the challenges that TSA officials reported, it is important to note that risk assessment is an accepted and required practice with a long history of use in a wide variety of public and private sector organizations. Completing a risk assessment would provide TSA greater assurance that it is directing its resources toward mitigating the highest priority risks. Moreover, other agencies conduct risk assessments based on threat, vulnerability, and consequences and have overcome the challenges TSA cited. For instance, within DHS, the U.S. Coast Guard, and FEMA use risk assessment methodologies to inform resource allocation. The U.S. Coast Guard, which is responsible for securing the maritime transportation mode, conducts risk assessments using its Maritime Security Risk Analysis Model (MSRAM). Coast Guard units use the Maritime Security Risk Analysis Model to assess the risk of terrorist attack based on scenarios—a combination of target and attack mode— in terms of threats, vulnerabilities, and consequences to more than 18,000 targets. The model combines these assessments and provides analysis to identify security priorities and support risk management decisions at the strategic, operational, and tactical levels. The tool’s underlying methodology is designed to capture the security risks facing different types of targets spanning every DHS CI/KR industry sector, allowing comparison between different targets and geographic areas at the local, regional, and national levels. In conducting assessments, the Coast Guard Intelligence Coordination Center quantifies threat as a function of intent (the likelihood of terrorists seeking to attack), capability (the likelihood of terrorists having the resources to attack), and presence (the likelihood of terrorists having the personnel to attack). Intelligence Coordination Center officials stated that the Coast Guard uses MSRAM to inform allocation decisions, such as the local deployment of resources and grants. In June 2008, we reported that FEMA used a reasonable risk assessment methodology—based on a definition of risk as a function of threat, vulnerability, and consequence—to determine grant funding allocations under the Homeland Security Grant Program. We found that this program utilized a reasonable methodology to assess risk and allocate grants to states and urban areas even though its assessment of vulnerability was limited. The risk assessment methodology used by FEMA is based on assessments of the threat, vulnerability, and consequence of a terrorist attack to each state and the largest urban areas. FEMA’s methodology estimates the threat to geographic areas based on terrorists’ capabilities and intentions, as determined by intelligence community judgment and data on credible plots, and planning and threats from international terrorist networks. Because this threat information is recognized as uncertain, threat accounts for 20 percent of the total risk to a geographic area, while vulnerability and consequence account for 80 percent. Moreover, the NIPP states that implementing protective programs based on risk assessment and prioritization enables DHS, sector-specific agencies, and other security partners to enhance current CI/KR protection programs and develop new programs where they will offer the greatest benefit. By conducting a risk assessment, TSA would be able to better prioritize risks as well as more confidently assure that its programs are directed toward the highest priority risks. TSA’s Mass Transit Modal Annex contains some information that is consistent with our prior work on characteristics of a successful national strategy and that is called for by Executive Order 13416: Strengthening Surface Transportation Security. However, the Modal Annex could be strengthened by including additional information that could help TSA and other implementing parties better leverage their resources to achieve the strategy’s vision of protecting mass transit and passenger rail systems from terrorist attacks. In February 2004, we identified six characteristics of successful national strategies. Additionally, the Executive Order calls for the Secretary of Homeland Security to develop modal annexes for each transportation sector that includes certain elements, many of which are similar to the national strategy characteristics. Table 3 provides a brief description of five of the national strategy characteristics and relevant Executive Order elements that are discussed further below. The Modal Annex contains information related to three of the characteristics we identified as desirable characteristics for a successful national strategy: (1) purpose, scope and methodology; (2) organizational roles, responsibilities, and coordination; and (3) integration and implementation. For example, the organizational roles, responsibilities, and coordination characteristic, which is also an element in Executive Order 13416, calls for agencies to identify which organizations are to implement the strategy, their roles and responsibilities, and the mechanisms for collaborating. The Modal Annex generally addresses this characteristic as it identifies relevant stakeholder roles and responsibilities. Specifically, the Modal Annex states that TSA has primary responsibility for ensuring security for mass transit and passenger rail while other federal and industry stakeholders, such as the FTA, FRA, FBI, private sector, and transit labor representatives have partnership roles. The Modal Annex also describes stakeholders’ collaboration efforts. For example, it describes FTA, FRA, APTA, and transit operators’ involvement in the development and implementation of security standards and directives. See appendix III for more information on the characteristics the Modal Annex includes. The Modal Annex, however, could be strengthened by addressing the other two desirable characteristics of an effective national strategy: (1) goals, subordinate objectives, activities, and performance measures and (2) resources and investments. Both of these could be useful in achieving the vision articulated in the Modal Annex of securing the mass transit and passenger rail systems. In conformance with this characteristic, the Modal Annex identifies sector- wide goals that apply to all modes of transportation as well as subordinate objectives specific to mass transit and passenger rail systems. For instance, one of TSA’s transportation sector goals is to enhance resiliency of the U.S. transportation system and presents three subordinate objectives to demonstrate how the agency intends to meet this goal. Further, for each subordinate objective, TSA presents information to explain what TSA, other federal components, or industry stakeholders are doing to meet the subordinate objective. For example, the agency identifies its Explosives Detection Canine Teams as an activity to accomplish assessing, managing, and reducing risk associated with key modes, links, and flows within critical transportation systems. Table 4 provides a complete list of the TSA’s goals and their subordinate objectives for the mass transit and passenger rail systems. While the Modal Annex identifies goals, objectives, and activities, it does not contain measures or targets on the effectiveness of the operations of the security programs identified in the Modal Annex. For example, one of TSA’s security programs listed in the Modal Annex—Security Technology Deployment––aligns under one of the sector goals: prevent and deter acts of terrorism using or against the transportation system. However, the Modal Annex contains no measures or targets to assess the effectiveness of this program in achieving this goal. In August 2006, we reported that performance measures are an important tool to communicate what a program has accomplished and provide information for budget decisions. Further, we noted that it is desirable for these measures to be as effective as possible in helping to explain the relationship between resources expended and results achieved because agencies that understand this linkage are better positioned to allocate and manage their resources effectively. Although the Modal Annex does not contain specific measures or targets, it does call for developing measures of effectiveness to evaluate mass transit and passenger rail efforts to mitigate risk and increase the resilience of systems and assets. TSA has developed performance measures to track the progress that the surface transportation security program has made in conducting activities to enhance the security of the mass transit and passenger rail systems. Specifically, TSA’s Surface Transportation Security Inspection Program fiscal year 2009 Annual Inspection Plan identifies annual and quarterly performance metrics for conducting mass transit and passenger rail-related assessments that TSA plans nationwide: number of inspections conducted per 1,000 inspector work hours on mass transit, passenger rail, and freight rail systems and number of BASE reviews conducted at the top 100 largest transit agencies. While these measures are useful in tracking activities or actions taken, they are output measures that do not fully inform TSA about how various actions have impacted the security of mass transit and passenger rail systems’ goals and objectives. For example, TSA has so far reported the progress in its Visible Intermodal Prevention and Response (VIPR) program in terms of the number of VIPR operations TSA conducted, but has not yet developed measures or targets to report on the effectiveness of the operations themselves. However, in June 2009, TSA program officials reported that they are planning the introduction of additional performance measures for no later than the first quarter of fiscal year 2010. These measures would gather information on (1) interagency collaboration by collecting performance feedback from federal, state and local security, law enforcement, and transportation officials prior to and during VIPR deployments; and (2) stakeholder views on the effectiveness and value of the VIPR deployment. In February 2009, TSA reported plans to introduce its first outcome measures for its mass transit and passenger rail security programs. For example, TSA plans to introduce a performance measure for its BASE review program. TSA officials reported that they plan to calculate this measure by comparing the results from the first and second round BASE reviews for the nation’s top 100 largest transit mass transit and passenger rail systems. TSA also reported plans to introduce additional outcome performance measures in the future, including an overall risk reduction measure tied to the BASE program. Implementing these new performance measures and including them in future updates of the Mass Transit Modal Annex should better inform decision makers at TSA on the effect of its programs in securing mass transit and passenger rail. While the Modal Annex identifies how TSA has allocated funds available to different transit agencies, the Modal Annex provides relatively few details on how grant resources should be targeted. Also, the Modal Annex contains little information on resources and costs associated with mass transit and passenger rail security programs. For example, the Modal Annex identifies as its third sector-goal, as shown on table 4, improve the cost-effective use of resources for transportation security; however, it provides few details on the costs, types, or levels of resources associated with implementation of the security programs that are aligned with this goal. Furthermore, the Modal Annex describes risks to the mass transit and passenger rail systems by discussing overseas attacks and the potential consequences of such attacks in the United States. However, the Modal Annex does not provide information on the cost of the consequences of such attacks and is silent on risk assessment efforts. TSA officials acknowledged the lack of this information and the need to include it in future updates of the Modal Annex. While providing cost estimates may be difficult to do, including resources and costs, to the extent possible, would help implementing parties allocate budgets according to priorities and constraints, and would help stakeholders shift such investments and resources as appropriate. Since 2004, federal and industry stakeholders have implemented several key actions to strengthen the security of the nation’s mass transit and passenger rail systems and federal actions have generally been consistent with TSA’s security strategy. However, federal efforts are largely in the early stages and opportunities exist for TSA to strengthen some programs. Since 2004, federal stakeholders have taken a number of key actions to secure mass transit and passenger rail systems and TSA has been the primary federal agency involved in implementing these actions. In general, these actions can be categorized into three areas: (1) deploying surface inspectors and other personnel to conduct voluntary security assessments and security operations at the nation’s largest mass transit and passenger rail systems; (2) establishing and implementing coordination mechanisms between federal entities and mass transit and passenger rail industry stakeholders; and (3) coordinating with the DHS Science and Technology Directorate (DHS S&T) to develop and test new security technology appropriate for deployment in mass transit and passenger rail systems. Since 2004, TSA’s primary security activity for mass transit and passenger rail has been conducting voluntary security assessments of the nation’s top 100 largest mass transit and passenger rail systems through its BASE program. TSA has used the BASE results to inform the development of security enhancement programs and to determine priorities for allocating mass transit and passenger rail security grants. In addition, through its VIPR program and its National Explosive Detection Canine Team Program (NEDCTP), TSA has deployed personnel and explosive detection canine teams to augment mass transit and passenger rail systems’ security forces to conduct hundreds of random and event-based security operations as a show of force to deter potential terrorist attacks at key mass transit and passenger rail stations. Federal agencies have taken other actions as well to strengthen security by enhancing coordination with transit industry stakeholders. For example, TSA established the monthly Transit Policing and Security Peer Advisory Group (PAG) and FTA initiated the semi-annual Transit Safety and Security Roundtables, both of which provide forums for TSA and mass transit and passenger rail systems, including Amtrak, to share security information and ideas. Additionally, FTA has enhanced mass transit and passenger rail security by funding the development and delivery of security training curriculum and programs for mass transit and passenger rail system employees, and by developing a list of recommended security and emergency action items for mass transit security programs, which it later updated in collaboration with TSA. TSA also collaborates with DHS S&T to pursue research, development, and testing of new security technology appropriate for deployment in mass transit and passenger rail systems. In 2006, DHS reorganized its security technology research and development structure, and under the new structure, TSA is to identify technology priorities to address security gaps and communicate these priorities to DHS S&T, which in turn is to conduct technology research, development, and testing. Table 5 provides descriptions of key federal programs and activities, initiated since 2004, mostly by TSA and FTA, to enhance mass transit and passenger rail system security. For a more extensive list of federal programs and activities, see appendix IV. Federal actions to secure mass transit and passenger rail systems generally have been consistent with those that TSA outlined in its security strategy for mass transit, the Mass Transit Modal Annex. The Modal Annex describes TSA’s strategic objectives and associated federal programs and activities to meet these objectives. For example, one objective calls for conducting security readiness assessments, which TSA has been doing since August 2006 through its BASE review program. Another objective calls for a public awareness program, which TSA reported implementing through its Employee Awareness Poster Program. See appendix V for a list of all of the Modal Annex mass transit objectives and TSA’s reported actions to achieve these objectives. Mass transit and passenger rail systems, including Amtrak, reported taking key actions since 2004 to improve their security. Most systems reported making operational enhancements to their security programs, such as adding security personnel or transit police. Moreover, some of the largest systems have implemented varying types of random passenger or baggage inspection screening programs. These programs include deploying security personnel at checkpoints to conduct visual observation of passengers for suspicious behaviors as well as non-invasive baggage checks. Since 2004, Amtrak reported taking additional actions to secure its system, focusing particularly on securing stations on its Northeast Corridor. Among other things, Amtrak introduced new passenger and baggage screening operations, increased its own explosive detection canine capacity, and deployed an armed mobile tactical team to respond to threats and conduct deterrent operations. Further, Amtrak provided security training to all of its frontline employees and conducted additional security risk assessments on its system as the baseline for developing its corporate security strategy. Officials from 24 of 25 passenger rail systems we interviewed and Amtrak also reported taking actions to strengthen the security of their systems in response to TSA’s 2004 passenger rail security directives. These actions included removing trash receptacles from high-risk platform areas and deploying explosive detection canine units to patrol their systems. Amtrak also initiated identification checks for adult passengers. However, TSA’s security directives contained limited requirements for passenger rail, and TSA has not enforced their implementation. Additionally, TSA released a report summarizing results of the BASE reviews it had conducted of mass transit and passenger rail systems during fiscal year 2007. This report showed that almost all transit agencies reported providing some type of security training to their frontline employees; however, the extent of the training provided varied greatly—with a majority providing an introductory level of safety and security training for new hires, but not refresher training. Many mass transit and passenger rail agencies also reported making capital improvements to secure their systems. For example, since 2004, 19 of the 30 transit agencies we interviewed had embarked on programs to upgrade their existing security technology, including upgrading closed- circuit television at key station locations with video surveillance systems that alert personnel to suspicious activities and abandoned packages and installing chemical, biological, radiological, nuclear, and explosives detection equipment and laser intrusion detection systems in critical areas. For bus transit agencies, capital improvements have included installing automatic vehicle location tracking, silent alarms, and engine disabling systems to counter potential hijacking threats. While mass transit and passenger rail systems as a whole have taken actions to enhance their security, TSA’s BASE reviews indicated that rail transit agencies were implementing a wider range of security programs than bus only transit agencies. For example, according to TSA’s initial findings from its BASE reviews of the 50 largest transit agencies, conducted during fiscal year 2007, rail transit agencies implemented more of the TSA/FTA security and emergency management action items than bus-only systems. TSA officials attributed the differences to three factors. First, passenger rail agencies have been required to comply with FTA’s triennial State Safety Oversight audits that require passenger rail agencies to have both a safety and security plan in place and TSA’s 2004 security directives. In contrast, bus-only transit agencies have not been required to implement such FTA security requirements, and no federal agency has issued bus-specific security requirements or directives. Second, bus-only transit agencies tend to be smaller than rail only or rail and bus transit agencies and have fewer financial resources available to invest in security activities. Finally, because passenger rail has been the target of recent high profile terrorist attacks overseas and rail is considered a higher security risk to terrorist attack than bus-only systems, passenger rail transit security has received greater focus—both by the transit industry and the federal government. As part of its research and development (R&D) strategy, DHS has been exploring new explosive detection technologies, particularly those that deter, detect, defeat, and protect against the use of IEDs in or around transit infrastructure. Accordingly, DHS technology pilot projects for mass transit and passenger rail have sought to identify and develop technologies that can effectively detect explosive weapons or compounds while causing minimal delays to passengers, such as fare card vending machines capable of detecting explosive residue on passengers’ bodies or bags (see figure 3). Although DHS has worked to develop some security technologies specific to mass transit and passenger rail systems, most technologies that it has pursued could work across different transportation modes, including aviation, maritime, mass transit, and passenger rail. DHS has also pursued several infrastructure protection projects that address the threat of IEDs, with a particular focus on addressing the vulnerabilities of underground and underwater transit tunnels. Unlike its role in commercial aviation, TSA does not procure or deploy security technologies for mass transit and passenger rail systems. Instead, TSA partners with mass transit and passenger rail systems to conduct pilot projects and demonstrations of commercially available technologies and technologies from DHS laboratories. The mass transit and passenger rail systems themselves determine which security technologies to procure and deploy. See appendix VI for a list of ongoing and completed TSA and DHS mass transit and passenger rail security-related technology pilot programs. Since 2007, TSA, like other DHS components, has been responsible for articulating the technology needs of all transportation sector end-users— including mass transit and passenger rail agency operators—to DHS S&T for development. TSA has taken some initial actions to reach out to mass transit and passenger rail systems regarding their security R&D needs; however, these efforts could be expanded and improved by more fully leveraging existing forums to solicit a wider range of input. This effort is important because, as we reported in September 2004, stakeholders are more likely to use research results if they are involved in the R&D process from the beginning. The Mass Transit Modal Annex states that DHS S&T and TSA will identify security technology needs in full partnership with the mass transit community. To achieve this, TSA officials told us that TSA leverages existing forums for communication, such as the semi-annual Transit Security Roundtables, to identify technology capability gaps and to solicit input and feedback on its technology priorities. Additionally, in 2008, TSA headquarters officials reported that they sought input from transit industry representatives through the Transit Policing and Security Peer Advisory Group and the Mass Transit Sector Coordinating Council Security Technology Working Group. Nonetheless, in a September 2008 draft report, the Mass Transit Sector Coordinating Council Security Technology Working Group reported that other than occasional telephone discussions, there was no ongoing structure that brought the federal government and transit industry together to discuss transit security technology priorities, needs, and areas of potential interest for technology advancement and research. In September 2004, we recommended that DHS and TSA improve their outreach to the transportation industry (including mass transit and passenger rail systems) to ensure that the industry’s R&D security needs have been identified and considered. DHS agreed that this recommendation was key to a successful R&D program and since that time, DHS and TSA have made some preliminary efforts to outreach on R&D security issues. However, by continuing to expand these efforts and getting input early on in the project selection process, TSA should be able to ensure that DHS has adequately considered and addressed the full scope of the industry’s R&D needs. TSA has taken initial actions to share information on available security technologies, but could strengthen its approach by providing more information to support transit agencies that are considering deploying new security technologies. Consistent with a recommendation we made in September 2005, TSA established the Public Transit Portal of DHS’s Homeland Security Information Network (HSIN), a secure Web site that serves as a clearinghouse of information on available security technologies that have been tested and evaluated by DHS, in addition to providing security alerts, advisories, and information bulletins. In February 2009, TSA reported that it had established HSIN accounts for 75 of the 100 largest mass transit and passenger rail systems. However, officials from 11 of 17 mass transit and passenger rail systems who discussed HSIN told us that they did not use it for guidance on available security technologies when considering security technology investments. These officials said that they did not use HSIN when considering such investments because HSIN did not contain product details that would support these decisions, including details on product capabilities, maintenance, ease of use, and the suitability of the products in a bus or rail venue. We reviewed HSIN and found that for a given security product, TSA’s listing provides a categorical definition (such as video motion analysis), a sub-category (such as day/night camera), and the names of products within those categories. However, HSIN neither provides nor indicates how transit agencies can obtain information beyond the product’s name and function. A senior program official with TSA’s TSNM mass transit division told us that mass transit and passenger rail system officials would already know whom to contact at TSA for more information on a product. However, the official acknowledged that the product listing could be enhanced by including the contact information of the TSA officials capable of providing that information. In the absence of more detailed information on security- related technologies, officials from 19 of 30 mass transit and passenger rail systems we interviewed told us that they either (1) asked other operators about their experiences with a particular technology; (2) performed their own research via the Internet or trade publications; or (3) performed their own testing. Making the results of research testing available to industry stakeholders could be a valuable use of federal resources by reducing the need for multiple industry stakeholders to perform the same research and testing. The senior TSA program official with the TSNM mass transit division also acknowledged that HSIN contained limited technology information but noted that the site’s content was largely in the early stages of development. The official attributed some of the limitations to TSA’s reluctance to provide substantive details regarding any particular product, since TSA officials did not want to be perceived as endorsing any particular vendor. Nonetheless, TSA stated that its goal for HSIN was to provide a way for transit agencies to share, receive, and find information on security technology as well as to provide a technology database with performance standards and product capabilities so that mass transit and passenger rail agencies would be well prepared to interact with vendors. Although TSA has set this goal for HSIN, there was no set deadline for the content-related improvements. By taking action to address mass transit and passenger rail agencies’ need for more information, TSA could help provide transit agencies with a consolidated source of information on security technologies and help ensure that limited resources are not used to duplicate research and testing efforts. In response to mass transit and passenger rail industry concerns about its VIPR program, TSA reported taking steps to work with the industry to improve the effectiveness of the program. TSA conducts VIPR operations as a way to introduce security measures (such as random bag searches) at mass transit and passenger rail systems to deter potential terrorist threats, augment local security forces, and promote the visibility of TSA resources. TSA, to date, has conducted over 800 VIPR operations at mass transit and passenger rail systems. TSA also reported that almost all operations were deployed on a random basis or to enhance security at special events or on holidays, rather than in response to specific threat information. Mass transit and passenger rail system officials we interviewed had varying opinions on the effectiveness of the VIPR operations that TSA had conducted on their systems. For example, security and management officials from 5 of the 30 mass transit and passenger rail systems we visited told us that they generally welcomed the additional security resources that the VIPRs provided. In contrast, officials from four other mass transit and passenger rail systems reported that because they were already deploying their own transit police and security personnel on their systems on a daily basis, the addition of a largely unarmed VIPR team on a single day did not add significant security value especially with the additional planning and costs incurred by these operations. In response to VIPR planning and implementation concerns raised by large mass transit and passenger rail systems, in October 2007 TSA issued a Concept of Operations (CONOPS) for its VIPR program that established general guidelines for the planning and execution of a VIPR deployment. TSA developed the guidance in coordination with members of the Transit Policing and Security Peer Advisory Group and issued the guidance to both its field personnel and the mass transit industry. The CONOPS includes general guidelines for 10 core components of collaboration, such as coordination, planning, and communications. In June 2008, the DHS Inspector General (DHS-IG) reported on VIPR planning and implementation concerns and noted that transit system officials reported that TSA’s issuance of the VIPR guidance had led to improvements that addressed many of the VIPR implementation concerns. Nevertheless, our review of TSA after-action reports for 104 VIPR operations TSA conducted from November 2007 through July 2008 on mass transit and passenger rail systems—a nine month period after TSA issued the CONOPS guidance— identified insufficient interoperable radio communications as a key challenge faced during many VIPR operations. According to the after-action reports, TSA’s key challenge has been ensuring that its VIPR teams have reliable interoperable radio communications—both among TSA personnel and with local law enforcement. According to the CONOPS, ensuring interoperable radio communications between VIPR team members and local law enforcement is essential to the safe and effective execution of VIPR programs, including ensuring their ability to communicate information on potential threats encountered during operations. However, in almost half of the after-action reports we reviewed (49 of 104), VIPR participants reported that a lack of reliable communications equipment had hindered their ability to conduct real-time communications with local law enforcement. This challenge has existed since TSA expanded the VIPR program into mass transit and passenger rail systems, where cell phone or other communications systems that previously worked in airports did not effectively operate in a transit environment. In many cases, TSA field personnel reported requests for new interoperable radio systems, but had not had those requests fulfilled by TSA headquarters. These reports indicated the need for a more comprehensive solution in which TSA procures communications systems capable of real time interoperability with security partners in mass transit and passenger rail systems. TSA managers of the VIPR program acknowledged the challenges that the VIPR program had experienced since it expanded into mass transit and passenger rail systems and stated that the agency was taking actions to address them. Examples include: Communications Improvements: TSA reported deploying additional communications equipment to field locations and working with DHS S&T to test new technologies for enhancing communications capability and interoperability in a mass transit or passenger rail environment. Coordination and Awareness: TSA reported that it developed and made available to mass transit and passenger rail systems a brochure with information on scheduling and deploying VIPR operations, including a description of the different options available for systems in utilizing VIPR teams and the planning and operational roles and responsibilities of participating TSA personnel. Further, to improve nationwide coordination of VIPR operations, TSA established a coordination center dedicated solely to VIPR operations and has established dedicated mobile VIPR teams in 10 cities. TSA has reported that it plans to expand the number of these teams nationwide by 2010. Training TSA Personnel: TSA reported in February 2009 that the agency had begun requiring VIPR team personnel to participate in system orientation and safety training from mass transit and passenger rail systems where they deploy in order to familiarize VIPR team members with both the transit agency’s physical structure and operating procedures. TSA also reported offering additional training on surface-based law enforcement tactics and legal authorities. Because TSA plans to further expand the VIPR program in 2009, effectively implementing these actions should better ensure that TSA uses its limited security resources to maximize the security benefit of VIPR operations in mass transit and passenger rail. In February 2007, TSA established a training program to assist mass transit and passenger rail agencies in expanding security training for their frontline transit employees. However, opportunities exist for TSA to strengthen its process for ensuring consistency in the performance of non- federal training vendors that mass transit and passenger rail agencies use to obtain training through the program. After TSA’s initial BASE reviews revealed wide variations in the extent of training that transit agencies were providing to their employees, including limited recurrent training, TSA established a Mass Transit Security Training program to provide curriculum guidelines for basic and follow-on security training areas— training programs and courses largely developed and funded by FTA. It also specified areas in which particular categories of employees should receive recurrent training as well as a matrix tool to enable transit agencies to determine the costs and timelines for implementing the training. To support delivery of the training courses, TSA aligned the program with the DHS Transit Security Grant Program. The Transit Security Grant Program has made transit agency grant funding for security training a top priority and offers mass transit and passenger rail agencies the option of using grant funding to cover costs for training to employees that is supplied by either (1) training providers that are federally funded or sponsored or (2) other training providers. While TSA has reported that the Mass Transit Security Training Program is providing opportunities for mass transit and passenger rail systems to expand security training to their employees, senior officials from FTA’s Safety and Security Office expressed concern that TSA had not established the necessary criteria to effectively manage the program. According to TSA’s Mass Transit Security Training Program guidance, TSA allows transit systems to obtain DHS grant funding to contract with private security training vendors if TSA has determined that the performance of the vendors’ training curriculum and delivery services is equal to those of the federally sponsored providers. As a result, TSA assumed new responsibility for evaluating whether these security training vendors met the performance standards of federally sponsored training providers and whether they could be used by transit agencies for training under the Transit Security Grant Program. However, opportunities exist for TSA to strengthen its process for making this evaluation. According to TSA, transit agency requests to use non-federally funded or sponsored training vendors under the Transit Security Grant Program are reviewed by TSA’s mass transit training specialist and by FEMA’s Grants Program Directorate for approval. This review includes an analysis of course documentation, such as a description of the course syllabus, cost estimate, and justification for why the course was the preferred solution. However, both FTA and TSA officials acknowledged that additional criteria are needed for TSA to properly evaluate the selection of the training vendors. As the lead federal agency for developing and implementing mass transit employee safety training programs since 1971, FTA is in the process of issuing guidance that could be relevant to TSA’s evaluation of training vendors. According to FTA’s 2009 Training Curriculum Development Guidelines, scheduled for release in 2009, criteria for evaluating the quality of training services should include, among other things, a review of the credentials of the instructors who would deliver the training course, the training vendor’s experience in providing the security course, and any performance evaluations or feedback obtained from organizations and students who previously received training from the vendor. Additionally, as we reported in March 2004, agencies should try to develop clear criteria when determining whether to contract with vendors for training. We identified factors that agencies should consider include the prior experience, capability, and stability of the vendors offering the training. Since implementing the Mass Transit Security Training Program in 2007, TSA reported that about 50 mass transit and passenger rail systems had applied for Transit Security Grant Program funding for employee security training, including one agency that applied to use training vendors that are not federally funded or sponsored. However, more applications for this option are expected as additional grant funding for training becomes available. TSA and FTA officials both noted their preference for transit agencies to use federally-sponsored training providers and expressed concerns that increased demands on the providers may make scheduling training with federally funded or sponsored providers more difficult. Enhancing criteria for evaluating the quality of training services could strengthen DHS’s ability to ensure that the grant money DHS is awarding to mass transit and passenger rail agencies is consistently funding sound and valid security training programs for these employees. In October 2005, we reported that collaborating agencies can identify opportunities to leverage each other’s resources, thus obtaining additional benefits that would not be available by working separately. By coordinating the enhancement of these criteria with other agencies conducting similar efforts, such as FTA, TSA could also leverage the expertise of other agencies to better ensure its efforts result in sound criteria. In March 2009, TSA reported that it had implemented some of the 9/11 Commission Act’s provisions related to mass transit and passenger rail security. While most mass transit and passenger rail industry security actions have been voluntary to date, the 9/11 Commission Act sets forth mandatory requirements for federal and industry stakeholders, and implementing those requirements may pose challenges for TSA and industry stakeholders, particularly for TSA’s Surface Transportation Security Inspection Program. TSA has more than doubled the size of its Surface Transportation Security Inspection Program over the past year, but has not completed a workforce plan to address current and future program needs, and surface inspectors have reported concerns with organizational changes that TSA has made to the program that may affect implementation of new responsibilities. Additionally, officials from the mass transit and passenger rail industry have reported concerns with the cost and feasibility of implementing pending 9/11 Commission Act regulations. The 9/11 Commission Act, enacted in August 2007, contains many provisions that task TSA with implementing various actions related to surface transportation, including mass transit and passenger rail security. Among other things, these provisions identify mandates for developing and issuing reports on TSA’s strategy for securing public transportation, conducting and updating security assessments of mass transit systems, and establishing a program for conducting security exercises for transit and rail agencies. In March 2009, TSA reported that it had satisfied some provisions of the 9/11 Commission Act pertaining to mass transit and passenger rail, including some through actions that had been taken prior to the enactment of the 9/11 Commission Act. For example, TSA reported that it had issued a report on the transportation security enforcement process and that its Mass Transit Modal Annex satisfied the requirement to develop a strategy for securing public transportation. However, TSA also reported that it had not yet implemented a number of other 9/11 Commission Act provisions, including several requiring TSA to issue regulations that would place new requirements on the mass transit and passenger rail industry. The 9/11 Commission Act requires TSA to develop and issue several different regulations for mass transit and passenger rail, including regulations for employee security training programs and requiring high- risk rail carriers to develop and implement security plans. TSA reported that it was in the process of developing these regulations and that for some required regulations it had sought feedback from the transit industry as it developed new regulations. However, as of March 2009, TSA had missed several legislative deadlines for issuing the required mass transit and passenger rail regulations, and in some cases had not established time frames for when it would ultimately do so. For example, TSA was required to issue interim regulations outlining requirements for a mass transit employee security training program by November 2007, with final regulations due by August 2008. TSA was also required to issue regulations by August 2008 requiring high-risk rail carriers to develop and implement security plans. However, TSA did not meet these deadlines. TSA reported that deadlines in the act for developing and issuing new regulations have been difficult to meet because of different factors, including the comprehensive scope of the requirements, the need to coordinate them with various entities, and a lack of resources for completing certain tasks. See table 6 below for a list of key selected 9/11 Commission Act mass transit and passenger rail provisions mandating actions by TSA, along with TSA’s reported status in doing so, as of March 2009. Develop and implement the National Strategy for Public Transportation Security. TSA reported that the Mass Transit Modal Annex to the Transportation System -Sector Specific Plan meets this requirement. Establish a task force to complete (by Feb. 2008) a risk assessment of a terrorist attack on railroad carriers, and based on the assessment, develop and implement the National Strategy for Railroad Transportation Security. TSA reported that the task force has been established and that the National Strategy for Railroad Transportation Security is under development. For passenger rail, TSA reported that the Mass Transit Modal Annex to the Transportation System -Sector Specific Plan meets. the requirement to develop and implement a security strategy. Review and update FTA security assessments of high-risk public transportation agencies, require high-risk public transportation agencies to develop security plans and review, amend as necessary, and approve the security plans. TSA reported that FTA security assessments were provided to TSA and that TSA used the BASE program to update the assessments. TSA stated that it will use the BASE results in developing regulations implementing this requirement. TSA reported that a regulatory project has been initiated. Conduct security assessments to determine the specific needs of local bus- only transportation systems. TSA reported that the assessments have been completed and that information is being prepared for use by the transportation system operators. Issue regulations (by Aug. 2008) that require each railroad carrier, including passenger rail carriers, determined to be high-risk to conduct a vulnerability assessment and to prepare, submit for approval, and implement a security plan. TSA reported that a regulatory project has been initiated. Establish a program for conducting security exercises for public transportation agencies and for railroad carriers. Establish a program for conducting security exercises for railroad carriers, including passenger rail carriers. TSA reported that its Intermodal Security Training and Exercise Program meets this requirement. The agency further reported a multi-phased, multi-jurisdictional pilot of this exercise program was held in the National Capitol Region from January through June 2008, northern New Jersey in September 2008, and Los Angeles from February to June 2009. Issue interim (by Nov. 2007) and final regulations (by Aug. 2008) for a public transportation security training program and issue regulations (by Feb. 2008) for a security training program for frontline railroad, including passenger railroad, employees. TSA reported that a consolidated regulatory project including public transportation, railroad, and over-the-road bus has been initiated. TSA reported that it anticipates issuing a Notice of Proposed Rulemaking in late calendar year 2009 or early calendar year 2010. Complete a name-based security background check for all public transportation frontline employees and frontline railroad employees. TSA reported that it has begun to develop a project plan for a rulemaking needed to satisfy this requirement, but that significant funding and time will be required to meet this requirement. TSA reported that it tracks the implementation status of mass transit and passenger rail security provisions of the 9/11 Commission Act on a monthly basis as part of a DHS-managed working group and was identifying processes needed to implement the provisions. TSA provided us with progress reports for completing these provisions which, in certain cases, identified challenges it faced in doing so, including a lack of resources. But the reports did not include a plan for addressing these challenges or milestones for implementing several 9/11 Act Commission provisions, as called for by project management best practices. TSA officials reported that before they could move forward on the 9/11 Commission Act requirements, they needed to allow the new administration time to review TSA’s efforts to date. However, until TSA develops a plan with milestones, it will be difficult to provide reasonable assurance that the provisions of the act are being developed and that a strategy is in place for overcoming identified challenges. While the majority of industry actions to secure mass transit and passenger rail have been taken on a voluntary basis, the pending 9/11 Commission Act regulations outline a new approach that sets forth mandatory requirements, the implementation of which may create challenges for TSA and industry stakeholders. With the exception of the 2004 passenger rail security directives, TSA had not, until recently, imposed security requirements on the mass transit and passenger rail industry. Instead TSA took a collaborative approach in encouraging passenger rail systems to voluntarily participate and address security gaps through its BASE review program. With TSA’s pending issuance of regulations required by the 9/11 Commission Act, TSA will fundamentally shift this approach, and establish a new regulatory regime for mass transit and passenger rail security. Once TSA issues the pending regulations for mass transit and passenger rail security, TSA’s Surface Transportation Security Inspection Program would have responsibility for enforcing industry compliance—further expanding and evolving the roles and responsibilities these inspectors have for mass transit and passenger rail, in addition to their responsibilities for other surface modes, such as freight rail, highway, and motor carrier security. TSA officials have raised concerns about their ability to meet the growing inspection requirements for mass transit and passenger rail and other surface modes that will be incurred by the new regulations required by the 9/11 Commission Act, particularly because TSA’s Surface Transportation Security Inspection Program is already challenged to meet its existing workload. For example, 10 of 11 Surface Transportation Security Inspection Program field office supervisors— Assistant Federal Security Directors for Surface Transportation (AFSD- S)—whom we interviewed reported that while they were meeting their primary inspection responsibilities for mass transit and other surface modes, resource constraints were routinely leading them to delay secondary activities, such as conducting stakeholder outreach with mass transit and passenger rail agencies. These field office supervisors attributed their resource constraints to a significantly expanded Surface Transportation Security Inspection Program workload from fiscal year 2006 through 2008 without a corresponding increase in its workforce. During this time, TSA expanded the responsibilities of the surface transportation security inspectors to include additional surface transportation modes, including conducting various voluntary security inspections for mass transit bus and freight rail, and participating in VIPR operations. TSA’s Surface Transportation Security Inspection Program is at risk of being unable to meet its expanding responsibilities if it does not plan for how to meet them. TSA reported that the agency had been appropriated funding to hire an additional 125 surface inspectors that would more than double its surface inspector workforce—including 75 in fiscal year 2008 and 50 more in fiscal year 2009—and planned to complete their hiring, training, and deployment by the end of fiscal year 2009. TSA reported plans to largely dedicate its newly hired surface inspectors to conducting VIPR activities, assessing security activities on surface modes, and monitoring newly issued freight rail security rules, such as ensuring a secure chain of custody for certain hazardous materials. However, as reported by the DHS Inspector General, beyond supporting current activities, the additional manpower TSA plans to put into its Surface Transportation Security Inspection Program may provide only limited relief. As a result, even with these additional resources, TSA’s surface inspectors may face challenges in fulfilling their responsibilities. GAO guidance on strategic human capital management reinforces that high performing organizations conduct workforce planning and analysis to identify and prepare for current and future workforce needs. Accordingly, a workforce plan that includes an analysis of a program’s workforce needs can help to ensure that the program has the right amount of resources to achieve program goals, allowing program managers to spotlight areas for attention before problems develop. In February 2009, we reported that TSA did not have a human capital or other workforce plan for its Transportation Security Inspection Program, but the agency had plans to conduct a staffing study to identify the optimal workforce size to address its current and future program needs. TSA reported that it had hired a contractor to conduct a full workforce analysis of its security inspectors, including both its aviation and surface inspectors, to determine the number needed to fulfill expanded roles and responsibilities and ensure effective deployment. TSA reported that it had initiated the study in January 2009 to be completed in late fiscal year 2009. This study, if completed, should provide TSA with a more reasonable basis for determining the surface inspector workforce needed to achieve its current and future workload needs in light of the new requirements of the 9/11 Commission Act. Surface inspectors have raised concerns about recent organizational changes that TSA has made to the Surface Transportation Security Inspection Program that may affect the implementation of its expanded roles and responsibilities. These concerns were reported by Surface Transportation Security Inspection Program field officials we interviewed, two recent DHS-IG reports, and an internal TSA report prepared by several Surface Transportation Security Inspection Program field officials. Specifically, in April 2008, TSA announced plans to expand the number of Surface Transportation Security Inspection field offices nationwide, from 22 to 54. Under a re-organized reporting structure, TSA placed 31 of the 32 new field offices under the command of Federal Security Directors and Assistant Federal Security Directors for Inspections—aviation-focused positions that historically have not had an active role in conducting mass transit, passenger rail, or other surface transportation inspection duties. TSA’s Surface Transportation Security Inspection Program headquarters officials continue to set strategy and annual goals, while in most field offices the surface inspectors report to the Federal Security Directors and Assistant Federal Security Directors for Inspections, who have day-to-day management lead and hiring responsibilities for surface inspectors. Reported field official concerns include: Balancing aviation and surface transportation priorities: A January 2008 report that 6 of 12 of TSA’s Assistant Federal Security Directors for Surface submitted to TSA headquarters cited concerns that placing the Surface Transportation Security Inspection program under the Federal Security Directors had resulted in the surface transportation mission being diluted by TSA’s aviation mission. The report also stated that the current reporting line of surface inspectors is less efficient and may create confusion among surface inspectors, because Federal Security Directors’ priorities and needs differ from those of the surface program. Establishing and maintaining credibility with industry stakeholders: Eight of the 11 Assistant Federal Security Directors for Surface we interviewed reported concerns that Federal Security Directors were not sufficiently focused on mass transit and passenger rail and the different challenges that surface inspectors face in overseeing the industry’s voluntary participation in non-regulatory security assessment activities. For example, one Assistant Federal Security Director for Surface commented that Federal Security Directors had tasked aviation security officers to participate in surface assessments and that doing so had caused some frustration among transit agency officials because of their lack of knowledge about the transit environment. A June 2008 DHS-IG report also noted surface inspectors’ concerns that Federal Security Directors were hiring surface inspectors who had no prior surface transportation experience, and that in some cases, Assistant Federal Security Directors reported that they were not included in hiring decisions. TSA disagreed with the DHS-IG and Assistant Federal Security Directors reports’ findings that the present Surface Transportation Security Inspection Program field office command structure had inhibited the program’s effectiveness. For example, TSA did not concur with the DHS- IG’s recommendation that TSA place the Surface Transportation Security Inspection Program under the direct authority of a TSA headquarters official responsible for surface transportation, rather than under the Federal Security Directors. TSA reported that they had selected their current command structure because Federal Security Directors were best equipped to make full use of the security network in their geographical location because they frequently interacted with state and local law enforcement and mass transit operators, and were aware of vulnerabilities in these systems. While TSA has not yet issued the new 9/11 Commission Act regulations for mass transit and passenger rail, 12 of 30 mass transit and passenger rail agencies we interviewed raised potential implementation concerns associated with one expected regulatory requirement regarding training for mass transit and passenger rail employees. Among other comments, mass transit and passenger rail agency officials reported that unless these new regulations were accompanied by funding to address implementation costs, they would be challenged to comply since mass transit agencies face tight budgetary constraints. For example, one transit agency official reported in feedback to TSA that an agency with 5,000 employees would incur labor costs of $1.5 million to have its employees participate in an 8- hour training program. Another transit agency official reported that it would be an achievement to get 30 to 40 percent of frontline employees through training in a year due, in part, to the costly overtime for backfilling those employees’ positions while they are in training. Additionally, these 12 agencies also reported concerns about the logistical feasibility of implementing the training requirement. For instance, under the act, mass transit and passenger rail agencies would be required to complete security training for all of their frontline employees within one year of DHS’s approving the transit agency’s training program. However, several mass transit and passenger rail agencies reported having thousands of employees and said it would be difficult to schedule training for all employees within one year without disrupting operations because they did not have the staff needed to backfill the positions of the employees undergoing training. As terrorist attacks on mass transit and passenger rail systems overseas have made clear, even with a variety of security precautions in place, mass transit and passenger rail systems that move high volumes of passengers on a daily basis remain vulnerable to attack. Since 2004, TSA has introduced a variety of initiatives aimed at enhancing the security of the nation’s mass transit and passenger rail systems, including conducting security assessments, implementing new security programs, and implementing some provisions of the 9/11 Commission Act. However, given the importance of the mass transit and passenger rail systems, the inherent vulnerabilities that could be exploited by terrorist threats, and the broadening requirements that will result in a shift to a regulatory approach, addressing management and coordination challenges should help ensure that current and future actions effectively improve the security of these systems. TSA has taken key steps to help secure the nation’s mass transit and passenger rail systems; however, additional actions to more effectively target resources would strengthen TSA’s security approach. To ensure that TSA’s efforts best prioritize and address risks, TSA should conduct a risk assessment for the mass transit and passenger rail systems that combines the results of threat, vulnerability, and consequence assessments. Until the overall risk to the entire system is identified through such an assessment, TSA cannot best determine how and where to target its limited resources to achieve the greatest security gains. TSA’s 2007 Mass Transit Modal Annex represents a positive step toward documenting TSA’s strategy for securing the mass transit and passenger rail systems, but further refinements to the strategy documented in future updates to the Modal Annex would help ensure that it provides all stakeholders with a clear and measurable path forward. For example, including relevant performance metrics will allow stakeholders to better evaluate their progress in achieving the strategy’s vision. In addition, incorporating information on what the strategy will cost, to the extent possible, would help implementing parties allocate budgets according to priorities and constraints, and would help stakeholders shift such investments and resources as appropriate. Federal and industry efforts to work together in securing mass transit and passenger rail in the absence of any significant federal security regulations have been commendable. In particular, TSA’s BASE reviews have been a positive step as they enhanced the awareness of security vulnerabilities at mass transit and passenger rail agencies throughout the country, while strengthening relationships among transit stakeholders. Notwithstanding TSA’s progress, TSA’s efforts remain largely in the early stages and opportunities exist for TSA to strengthen the implementation of some of its security programs. Expanding its outreach with mass transit and passenger rail officials will be particularly important for TSA in gathering security technology information and disseminating it to the systems and enabling officials to identify and deploy new security technologies to better secure their systems. In addition, with HSIN, TSA already has a venue in place for expanding the dissemination process and should explore the feasibility of populating this site with better and more relevant technology information to help meet the needs of mass transit and passenger rail agencies regarding information on available security technology. Such action should help ensure that limited resources are not used to duplicate research and testing efforts. Finally, by providing guidance and funding to cover mass transit and passenger rail agency costs for providing employee security training, TSA has taken steps to reduce a key vulnerability it identified during its BASE reviews. However, with the anticipated increase in demand for employee security training, it is important that TSA have an effective evaluation process in place to ensure it is consistently funding sound and valid training programs for mass transit and passenger rail agencies seeking funding to pay for non- federal training providers. Finally, a significant transition lies ahead. While TSA reports making progress in implementing the provisions of the 9/11 Commission Act, the agency has fallen behind in issuing required mass-transit and passenger- rail security regulations. The implementation of these regulations will be a fundamental shift in approach for TSA as it assumes more of a regulatory role in securing mass transit and passenger rail. This shift—combined with an expanding Surface Transportation Security Inspector workforce that has more than doubled in size in the past year and shifting deployment and field reporting structures—will challenge TSA to manage its new responsibilities. However, this transition will be important for both TSA and industry stakeholders to manage successfully to ensure that new requirements are met and that TSA and stakeholders continue to work together to secure mass transit and passenger rail. One approach that could help DHS manage these many changes is to develop a schedule with milestones for implementing the remaining 9/11 Commission Act requirements pertaining to mass transit and passenger rail. Without such a plan, it will be difficult for TSA to provide reasonable assurance that the provisions of the act are being implemented and that a strategy is in place for overcoming identified challenges. We recognize the inherent challenges to securing these systems given the continuing terrorist threat, openness of the system, and difficulties posed by attempting to secure and patrol numerous points of entry. However, given the criticality of mass transit and passenger rail systems to our way of life and the economy, and the inherent risks to them, TSA should continue to strive to strengthen its security efforts for the systems. To help ensure that the Transportation Security Administration is successfully prioritizing resources and collaborating with federal and industry stakeholders in implementing actions to secure the mass transit and passenger rail systems from acts of terrorism, and that its strategy is consistent with the characteristics of a successful national strategy, we are making six recommendations to the Assistant Secretary for the Transportation Security Administration: To help ensure that the federal strategy to secure the mass transit and passenger rail systems considers assessment information within the context of risk, TSA, as the sector-specific agency for mass transit and passenger rail, should conduct a risk assessment that integrates all three elements of risk—threat, vulnerability, and consequence. As part of this assessment, TSA should, to the extent feasible, fully leverage existing assessment information from its own sources as well as those provided by other federal and industry stakeholders, as appropriate, and use this information to inform its security strategy. To better achieve the security strategy laid out in its Mass Transit Modal Annex—TSA’s security strategy for the mass transit and passenger rail systems—TSA should, to the extent feasible, incorporate into future updates of the Modal Annex the characteristics of a successful national strategy and the elements outlined in Executive Order 13416, including: measuring the agency’s and industry’s performance in achieving the goals of preventing and deterring acts of terrorism and enhancing the resiliency of mass transit and passenger rail systems and incorporating information on what the strategy will cost along with the specifying the sources and types of resources and investments needed, and identifying where those resources and investments should be targeted. To help ensure that DHS security technology research and development efforts reflect the security technology needs of the nation’s mass transit and passenger rail systems, TSA should expand its outreach to the mass transit and passenger rail industry in the planning and selection of related security technology research and development projects. To meet the needs of mass transit and passenger rail agencies regarding information on available security technologies, TSA should explore the feasibility of expanding the security technology product information on the Public Transit Portal of the Homeland Security Information Network, and consider including information such as product performance in a rail or bus venue, cost, maintenance needs, and other information to support mass transit and passenger rail agencies purchasing and deploying new security technologies. To better ensure that DHS consistently funds sound and valid security training delivery programs for mass transit and passenger rail employees, TSA should consider enhancing its criteria for evaluating whether security training vendors meet the performance standards of federally sponsored training providers and whether the criteria could be used by transit agencies for training under the transit security grant program. As part of this effort, TSA should consider coordinating with other federal agencies that have developed criteria for similar programs, such as the Federal Transit Administration. To better ensure DHS’s ability to satisfy the provisions of the 9/11 Commission Act related to mass transit and passenger rail, DHS should develop a plan with milestones for implementing provisions of the 9/11 Commission Act related to mass transit and passenger rail security. We provided a draft of this report to DOT, Amtrak, and DHS for review and comment. DOT did not provide comments. Amtrak provided written comments on June 16, 2009. In its letter, Amtrak provided additional information on security actions they were taking, noted collaboration with federal agencies, and expressed some concern about the cumbersome nature and cost share requirements of the Transit Security Grant Program. Amtrak’s comments are presented in appendix VII. DHS provided written comments on June 17, 2009, which are presented in appendix VIII. In commenting on the report, DHS stated that it concurred with all six recommendations and identified actions planned or under way to implement them. In comments related to our first recommendation, that DHS conduct a risk assessment that integrates all three elements of risk, DHS stated that it recognized the importance of conducting risk assessments to inform agency priorities, security enhancement programs, and resource allocations. It also reported that in addition to the various assessments already completed and the BASE reviews conducted on a continuous cycle, an assessment pilot program is planned for later in 2009. Under this pilot, TSA will evaluate the effectiveness of and provide lessons learned from its new risk assessment tool for mass transit and passenger rail to enhance the tool’s capability prior to its implementation. In comments related to our second recommendation that DHS incorporate in future updates of the Modal Annex the characteristics of a successful national strategy and the elements outlined in Executive Order 13416, DHS reported that it planned to revise its Mass Transit Modal Annex and incorporate these characteristics and elements to improve its ability to measure agency and industry progress toward achieving mass transit and passenger rail security performance goals. In response to our third recommendation that TSA expand its outreach to the mass transit and passenger rail industry in the planning and selection of related security technology research and development projects, DHS reported on several planned coordination efforts including its intent to coordinate the Modal Annex update with mass transit and passenger rail stakeholders, and work to ensure that stakeholders have ample opportunities to provide input on security technology development and testing priorities. With regard to our fourth recommendation that TSA explore the feasibility of expanding the security technology product information on the Public Transit Portal of the Homeland Security Information Network, DHS reported that it expects to expand both the scope and quality of security technology information provided to stakeholders through the Public Transportation Information Sharing and Analysis Center—which has a principal objective of aligning and integrating analytical and information sharing activities with relevant federal processes to enhance the information-sharing environment for the mass transit and passenger rail community. In comments related to our fifth recommendation that TSA consider enhancing its criteria for evaluating security training vendors under the Transit Security Grant Program and consider coordinating with other federal agencies that have developed such criteria, DHS stated that TSA will work with FTA through an existing joint working group to develop criteria for reviewing new vendor-provided training courses. Lastly, with regard to our sixth recommendation that DHS develop a plan with milestones for implementing provisions of the 9/11 Commission Act related to mass transit and passenger rail security, DHS reported that TSA will produce a plan that identifies necessary actions and sets milestones to evaluate its effectiveness in meeting statutory requirements associated with the 9/11 Commission Act. DHS and Amtrak also provided us with technical comments, which we considered and incorporated into the report where appropriate. As agreed with your office, unless you publicly announce the contents of this report, we plan no further distribution for 30 days from the report date. At that time, we will send copies of this report to the Secretary of Homeland Security, the Secretary of Transportation, Amtrak, interested congressional committees, and other interested parties. The report will also be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions concerning this report, or wish to discuss these matters further, please contact me at (202) 512-3404 or berrickc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix IX. The objectives of this report were to: (1) determine the extent that federal and industry stakeholders assessed or supported assessments of the security risks to mass transit and passenger rail since 2004, and how, if at all, TSA used risk assessment information to inform and update its security strategy; (2) describe key actions, if any, that federal and industry stakeholders implemented or initiated, since 2004, to strengthen the security of mass transit and passenger rail systems, the extent to which federal actions were consistent with TSA’s security strategy, and what challenges, if any, TSA faces in implementing these actions; and (3) describe TSA’s reported status in implementing provisions of the Implementing Recommendations of the 9/11 Commission Act of 2007 related to mass transit and passenger rail security, and discuss challenges, if any, TSA and the mass transit and passenger rail industry face in implementing the actions required by the act. To determine the extent that federal and industry stakeholders have assessed or supported assessments of the security risks to mass transit and passenger rail since 2004, and how, if at all, TSA has used risk assessment information to inform and update its security strategy, we obtained and analyzed various reports that address some or all elements of security risk (threat, vulnerability, and consequence) from DHS component agencies, including TSA, the DHS Office of Infrastructure Protection within the National Protection and Programs Directorate (NPPD), and the Homeland Infrastructure Threat Reporting and Analysis Center (HITRAC). We also reviewed information on risk-related assessments conducted by federal agencies, including mass transit and passenger rail security vulnerability assessments conducted by DOT’s Federal Transit Administration (FTA), and a variety of federally developed security risk assessment tools for the mass transit and passenger rail industry. Additionally, we reviewed TSA’s Baseline Assessment for Security Enhancement (BASE) review checklist and fiscal year 2007 BASE report of the results of BASE reviews that TSA conducted at 44 of the top 50 largest mass transit and passenger rail agencies—by ridership—as a measure of TSA’s efforts to gather vulnerability information. We gathered information on TSA’s consequence assessments through interviews with Transportation Sector Network Management (TSNM) officials. We also interviewed TSNM officials in order to assess how risk assessments were informing TSA’s security strategy for mass transit and passenger rail and then compared TSA’s actions to GAO and DHS reports on risk assessment. Because of the scope of our work, we relied on TSA to identify its assessment activities but did not assess the extent to which its assessment activities meet the National Infrastructure Protection Plan criteria for threat, vulnerability, and consequence assessments. To further assess risk assessment efforts, we interviewed federal officials from DHS’s HITRAC, TSA’s Office of Intelligence, TSA’s Surface Transportation Security Inspection Program, and, at DOT, FTA’s Office of Safety and Security to understand what additional assessment information or assistance on risk assessments was available to either TSA or the transit agencies. Further, we interviewed security officials from Amtrak and 30 mass transit and passenger rail agencies across the nation. This sample allowed us to meet with agencies of varying sizes and types to determine their perspectives on federal and mass transit and passenger rail industry risk assessment efforts to date (see objective 2 for how these agencies and cities were selected). Additionally, we reviewed TSA’s strategic planning document—the Mass Transit Modal Annex to the Transportation System - Sector Specific Plan (TS-SSP) issued in May 2007—and identified federal guidelines for developing a risk-based security strategy. Specifically, to determine the extent to which TSA’s strategy conformed to requirements and best practices, we reviewed relevant statutory requirements of the Government Performance and Results Act of 1993 (GPRA) that included general requirements that are applicable in the establishment of government strategies and programs, and the Implementing Recommendations of the 9/11 Commission Act of 2007, which included requirements for establishing a security strategy. For example, we reviewed existing Executive Directives, including Homeland Security Presidential Directives 1, 7, and 8, and Executive Order 13416: Strengthening Surface Transportation Security to determine the extent to which TSA’s Mass Transit Modal Annex conformed to these requirements. We also analyzed executive guidance documents outlining best practices for effectively implementing a risk management framework, and in particular, risk assessment best practices, including both the DHS National Infrastructure Protection Plan (NIPP) and the TS-SSP. We also reviewed GAO best practice criteria for developing a successful national strategy and compared the Mass Transit Modal Annex against it. Finally, to identify the extent to which TSA is measuring its performance in implementing its mass transit and passenger rail security programs, we reviewed DHS and TSA documents, including the Modal Annex, DHS Critical Infrastructure and Key Resources Annual Reports, and Surface Transportation Regulatory Activities Plan, as well as the Office of Management and Budget’s Program Assessment Rating Tool (PART), which assessed the adequacy and effectiveness of these program measures. To identify and describe the key actions federal and industry stakeholders have implemented or initiated since 2004 to strengthen the security of mass transit and passenger rail systems, the extent to which federal actions are consistent with TSA’s security strategy, and the challenges, if any, that TSA has faced in making these actions effective, we interviewed officials from DHS and DOT. From DHS, we interviewed officials from TSA’s Surface Transportation Security Inspection Program within the Office of Security Operations, TSA’s Office of Security Technology, and, within TSA’s Office of Law Enforcement, the Federal Air Marshal Service (which plays a lead role in implementing VIPR Operations). We also interviewed officials from DHS’s Science and Technology Directorate and Federal Emergency Management Agency’s Grants Programs Directorate. Within DOT, we interviewed officials from FTA’s Office of Safety and Security and also the Federal Railroad Administration. We also interviewed officials from the three federally sponsored mass transit employee training providers—the National Transit Institute, Transportation Safety Institute, and Johns Hopkins University—to obtain information on training they offered to mass transit and passenger rail employees and their perspectives on TSA’s Mass Transit Security Training Program. To obtain information on industry security actions and perspectives on federal mass transit and passenger rail security actions, we conducted site visits at, or held teleconferences with, officials representing 30 mass transit and passenger rail systems across the nation—representing 75 percent of the nation’s total mass transit and passenger rail ridership— based on information we obtained from the Federal Transit Administration’s National Transit Database and the American Public Transportation Association (APTA). We selected this non-probability sample of mass transit and passenger rail systems and cities because of their high levels of ridership, geographic dispersion, experience with TSA security assessments, eligibility for grant funding, and expert recommendation. Because we selected a non-probability sample of mass transit and passenger rail agencies, the information obtained from these site visits cannot be generalized to all transit agencies nationwide. However, we determined that the selection of these sites was appropriate for our design and objectives and that the selection would provide valid and reliable evidence. The information we obtained provided us with a broad overview of the types of actions taken to strengthen security. Table 1 lists the mass transit and passenger rail systems we interviewed. We also interviewed Amtrak headquarters officials and visited three Amtrak station locations in the Northeast Corridor. During site visits to mass transit and passenger rail agencies, we interviewed security officials, toured stations and other facilities such as control centers, and observed security practices. Further, we interviewed TSA surface transportation security inspectors from the 13 field offices responsible for overseeing the passenger rail and mass transit systems we visited, and in one case, observed the inspectors conduct a BASE review of a mass transit system. We also interviewed state officials with homeland security responsibilities, representatives of the American Public Transportation Association, and where applicable, regional transportation authority officials. To determine the extent to which federal and industry actions were consistent with TSA’s security strategy, we reviewed TSA and FTA documentation describing ongoing programs and compared them with the strategic objectives, programs, and actions TSA described in its Mass Transit Modal Annex. To further assess federal and industry actions, and identify potential challenges, we reviewed DHS and DOT documents relevant to federal and industry stakeholder actions to secure passenger rail and mass transit systems. For example, we reviewed program documentation for TSA’s Mass Transit Security Training Program, as well as FTA’s 2009 Final Draft of Training Curriculum Development Guidelines and federal transit employee training curricula. We also reviewed TSA’s Surface Transportation Security Inspection Program Standard Operating Procedures, strategic and annual plans, and documentation of completed mass transit and passenger rail security assessments. We also reviewed TSA’s Concept of Operations (CONOPS) for its Visible Intermodal Prevention and Response (VIPR) program—TSA’s program for deploying security personnel to augment security on mass transit and passenger rail systems—to identify guidelines TSA has established for implementing the program. We then obtained a list of VIPR operations, by location and date, that TSA reported conducting in mass transit and passenger rail systems immediately following its issuance of the CONOPS in October 2007. We obtained and matched this list with information found in electronic copies of all TSA VIPR operation plan after-action reports (AAR)—describing the results and challenges encountered during VIPR operations that TSA conducted at mass transit and passenger rail systems from November 2007 through July 2008. We chose to review after-action reports for this period to determine the impact of guidance TSA issued in October 2007, to improve its implementation of the VIPR program. Both the initial list and after-action reports identified 108 VIPR operations; however, we reviewed 104 of these. Two analysts independently coded the challenges noted on each of the reports. They discussed differences until agreement could be reached on the most appropriate challenge category. We also conducted a site visit to the Transportation Security Operations Center to interview officials with TSA’s Federal Air Marshal Service within the Office of Law Enforcement—which manages the VIPR program—to discuss the challenges identified in the after-action reports. We also obtained access to, and reviewed, the DHS Homeland Security Information Network - Public Transit Portal secure Web communication system to identify the type and extent of security technology information that TSA had made available to industry users of the system, and identified and reviewed best practices applicable to R&D programs identified by leading research organizations, such as the National Research Council of the National Academy of Sciences, in order to establish criteria for evaluating federal and industry coordination in research and development efforts. We also reviewed two DHS-IG reports and found the quality of the methods used to develop these reports sufficient for use as a source in this report. For the final objective, to determine the status of TSA’s implementation of 9/11 Commission Act requirements for mass transit and passenger rail, and challenges, if any, that TSA and the mass transit and passenger rail industry face in meeting these requirements, we reviewed the 9/11 Commission Act to identify DHS and industry requirements related to mass transit and passenger rail security. We also reviewed TSA status reports outlining the agency’s reported status in satisfying various 9/11 Commission Act provisions related to mass transit and passenger rail security. However, we did not verify the accuracy of TSA’s reported status in implementing these 9/11 Commission Act requirements. To identify potential challenges TSA and the mass transit and passenger rail industry may face in implementing various 9/11 Commission Act requirements, we interviewed TSA headquarters officials from the Transportation Security Network Management—Mass Transit division, including the Deputy General Manager, and officials from TSA’s Surface Transportation Security Inspection Program, including the headquarters based Program Chief, and Surface Transportation Security Inspectors from 13 of 54 field offices, including 11 of 12 Assistant Federal Security Directors for Surface. We also reviewed TSA program documents relating to its inspection program including strategic and annual inspection plans, standard operating procedures, and memorandums and directives documenting organizational and staffing plans. Moreover, to obtain information on industry perspectives of potential challenges, we interviewed officials from 30 mass transit and passenger rail systems and Amtrak as well as APTA. In addition, we obtained and reviewed various reports which discuss the federal or industry role in implementing the 9/11 Commission Act, including recent reports issued by the DHS-IG, Congressional Research Service, and a January 2008 report prepared by six Assistant Federal Security Directors for Surface. We conducted this performance audit from September 2007 through June 2009 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. The following list of voluntary Security and Emergency Management Action Items is an update to the Federal Transit Administration's Top 20 Security Program Action Items originally released in January 2003. The update has been developed by FTA and the Department of Homeland Security's Transportation Security Administration and Office of Grants & Training in consultation with the public transportation industry through the Mass Transit Sector Coordinating Council, for which the American Public Transportation Association serves as Executive Secretary. The updated action items address current security threats and risks that confront transit agencies, with particular emphasis on priority areas where gaps need to be closed in security and emergency preparedness programs. Though this update consolidates the previous 20 items into 17, the purpose, scope, and objectives remain consistent. 1. Establish written security plans and emergency management plans. 2. Define roles and responsibilities for security and emergency management. 3. Ensure that operations and maintenance supervisors, forepersons, and managers are held responsible for security issues under their control. 4. Coordinate security and emergency management plan(s) with local and regional agencies. 5. Establish and maintain a security and emergency training program. 6. Establish plans and protocols to respond to the DHS Homeland Security Advisory System threat levels. 7. Implement and reinforce a Public Security and Emergency Awareness Program. 8. Conduct tabletop and functional drills. 9. Establish and use a risk management process to assess and manage threats, vulnerabilities and consequences. 10. Participate in an information sharing process for threat and intelligence information. 11. Establish and use a reporting process for suspicious activity (internal and external). 12. Control access to security critical facilities with ID badges for all visitors, employees, and contractors. 13. Conduct physical security inspections. 14. Conduct background investigations of employees and contractors. 15. Control access to documents of security critical systems and facilities. 16. Ensure existence of a process for handling and access to Sensitive Security Information (SSI). 17. Conduct security program audits. To help the federal government develop sound national strategies, we have previously identified six desirable characteristics of successful national strategies, including (1) purpose, scope, and methodology of the strategy; (2) risk assessment; (3) goals, subordinate objectives, activities, and performance measures; (4) resources and investments; (5) organizational roles, responsibilities, and coordination; (6) integration and implementation. We discussed four of these characteristics in the body of the report, and below we discuss the other two characteristics. Where applicable, we link relevant sections of Executive Order 13416 to highlight the importance of these measures to strengthen the passenger rail and mass transit security national strategy. This characteristic addresses why the strategy was produced, the scope of its coverage, and the process by which it was developed. For example, a strategy might discuss the specific impetus that led to its creation, such as statutory requirements, executive mandates, or other events—such as terrorist attacks. In addition to describing what the strategy is meant to do and the major functions, mission areas, or activities it covers, a national strategy would ideally also outline its methodology. For example, a strategy might discuss the principles or theories that guided its development, what organizations or offices drafted the document, whether it was the result of a working group, or which parties were consulted in its development. TSA’s Mass Transit Modal Annex identifies the purpose and scope of the Modal Annex and references several principle documents used to develop the Modal Annex—including the Presidential Executive Order 13416: Strengthening Surface Transportation Security, the Transportation System- Sector Specific Plan (TS-SSP), and the National Infrastructure Protection Plan. It also describes the process or methodology that was used and who developed the Annex. For example, the Modal Annex states that TSA’s vision is to provide a secure, resilient transit system that leverages public awareness, technology, and layered security programs while maintaining the efficient flow of passengers and encouraging the expanded use of the nation’s transit services. The Modal Annex also discusses the scope and type of various federal and industry mass transit and passenger rail security efforts and aligns them with three broad DHS security goals for the transportation sector, as outlined in the TS-SSP. In addition, the Modal Annex references the National Infrastructure Protection Plan as a source for developing security programs for mass transit and passenger rail systems, and it also discusses several domestic and international terrorist attacks that have occurred as evidence of the various security risks to the mass transit and passenger rail systems. Furthermore, the Modal Annex explains the methodology used in its development, as called for in our prior work on characteristics of a national strategy. In addition to referencing the National Infrastructure Protection Plan and the TS-SSP as literatures providing the principles or theories that guided the development of the Modal Annex, the Modal Annex also describes the process and information that were used to develop the strategy and identified entities that contributed to its development. For example, the strategy describes how mass transit and passenger rail security programs and initiatives are developed and implemented and how they are aligned with the overall transportation sector goals and objectives and mass transit and passenger rail modal strategies and objectives. Also, the Modal Annex identifies the Transit, Commuter and Long Distance Rail Government Coordinating Council (TCLDR-GCC), the Mass Transit Sector Coordinating Council (SCC), the Critical Infrastructure Partnership Advisory Council, and TSA’s Mass Transit Division as entities involved in developing the transportation security strategic policy. This characteristic addresses both how a national strategy relates to other strategies’ goals, objectives, and activities and to subordinate levels of government and their plans to implement the strategy. For example, a national strategy could discuss how its scope complements, expands upon, or overlaps with other national strategies, such as DHS efforts to mitigate transportation risks. Also, related strategies could highlight their common or shared goals, subordinate objectives, and activities. Similarly, the Executive Order requires that the Modal Annex identify existing security guidelines and requirements. To meet these requirements and because protecting the mass transit and passenger rail systems is a shared responsibility among many stakeholders, the Modal Annex could identify regulations and programs that affect the security of the mass transit and passenger rail systems. In addition, a strategy could address its relationship to other agency strategies using relevant documents from implementing organizations, such as strategic plans, annual performance plans, or annual performance reports that GPRA requires of federal agencies. A strategy might also discuss, as appropriate, various strategies and plans produced by the state, local, or private sectors and could provide guidance such as the development of national standards to more effectively link together the roles, responsibilities, and capabilities of the implementing parties. TSA’s Modal Annex delineates mechanisms to facilitate stakeholders coordination, specifically the TCLDR-GCC and the Mass Transit SCC, discusses other relevant industry security plans, and identifies regulations and programs such as the regulation on designating a rail security coordinator and security programs related to public awareness and training that affect the security of the mass transit and passenger rail systems. The Modal Annex also addresses its relationship with strategic documents or activities of other federal agencies that have a role in mass transit and passenger rail security, such as those that guide FTA, which has a supporting role along with TSA for protecting mass transit and passenger rail systems. For example, the Modal Annex mentions how FTA’s activities, such as the State Safety Oversight Agencies audit program and FTA’s Section 5307 grant program fit into TSA’s overall strategy for securing mass transit and passenger rail systems. The Modal Annex also mentions DHS-DOT collaborative efforts through their memorandum of understanding such as the development of public transportation annex delineating areas of coordination to assist transit agencies in prioritizing and addressing security related needs. In addition, the Modal Annex points out how it relates to the National Strategy for Transportation Security required by the Intelligence Reform and Terrorism Prevention Act of 2004. For example, it explains how TSA’s effort in building security force multipliers through security training for front-line employees of mass transit and passenger rail systems directly supports the National Priorities, the National Preparedness Goal, and the National Strategy for Transportation Security. By providing such information, the agency is identifying linkages with other developed strategies and other organizational roles and responsibilities. This appendix expands on the list of key actions identified in the body of the report in table 5. This table presents a more comprehensive list of federal actions taken to enhance mass transit and passenger rail security since 2004. In addition to the contact named above, Dawn Hoff, Assistant Director, and Daniel Klabunde, Analyst-in-Charge, managed this assignment. Jay Berman, Martene Bryan, Charlotte Gamble, and Su Jin Yon made significant contributions to the work. Chuck Bausell and Rudy Chatlos, assisted with design, methodology, and data analysis. Lara Kaskie and Linda Miller provided assistance in report preparation; and Tracey King provided legal support.

Terrorist incidents worldwide have highlighted the need for securing mass transit and passenger rail systems. The Department of Homeland Security's (DHS) Transportation Security Administration (TSA) is the primary federal entity responsible for securing these systems. GAO was asked to assess (1) the extent to which federal and industry stakeholders have assessed risks to these systems since 2004, and how TSA has used this information to inform its security strategy; (2) key actions federal and industry stakeholders have taken since 2004 and the extent to which federal actions are consistent with TSA's security strategy, and the challenges TSA faces in implementing them; and (3) TSA's reported status in implementing 9/11 Commission Act provisions for mass transit and passenger rail security. GAO reviewed documents including TSA's mass transit and passenger rail strategic plan, and interviewed federal officials and industry stakeholders from 30 systems and Amtrak--representing 75 percent of U.S. mass transit and passenger rail ridership. Since 2004, federal and industry stakeholders have conducted assessments of individual elements of risk--threat, vulnerability and consequence--for mass transit and passenger rail systems and this information has informed TSA's security strategy; however, TSA has not combined information from these three elements to conduct a risk assessment of these transportation systems. By completing a risk assessment, TSA would have reasonable assurance that it is directing its resources toward the highest priority needs. Further, while TSA's mass transit and passenger rail security strategy contains some information, such as goals and objectives, that is consistent with GAO's prior work on characteristics of a successful national strategy, it could be strengthened by including performance measures to help TSA track progress in securing these systems, among other things. Federal and industry stakeholders have taken several key actions to strengthen the security of mass transit and passenger rail systems since 2004, and while federal actions have been generally consistent with TSA's security strategy, TSA faces coordination challenges, and opportunities exist to strengthen some programs. TSA has deployed surface inspectors to assess industry security programs and worked with DHS to develop security technologies, among other actions. Mass transit and passenger rail systems, including Amtrak, also reported taking actions to increase security, such as implementing passenger and baggage screening programs. Although TSA has taken steps to enhance its efforts, it can further strengthen security programs by, for example, expanding its efforts to obtain and share security technology information with industry. By improving information sharing with industry, TSA can help to ensure that its and industry's limited resources are used more productively to secure mass transit and passenger rail systems. As of March 2009, TSA reported implementing some of the 9/11 Commission Act provisions related to securing mass transit and passenger rail such as developing a strategy for securing transportation, but had missed deadlines, for example, for issuing new regulatory requirements for mass-transit and passenger-rail employee security training. In addition, TSA's progress reports that track its implementation of 9/11 Act provisions lack milestones to guide this effort as called for by project management best practices. Additionally, in some cases, TSA progress reports identify challenges to meeting 9/11 Act provisions, but these reports do not include a plan for addressing these challenges. Until TSA develops a plan with milestones, it will be difficult for TSA to provide reasonable assurance that the act's provisions are being implemented and that a plan is in place for overcoming challenges that arise. Additionally, officials from almost half of the mass transit and passenger rail systems GAO visited reported concerns with the potential costs and the feasibility of implementing pending employee security training requirements.

OMB’s Uniform Guidance was issued on December 26, 2013, and became effective on December 26, 2014. The Uniform Guidance establishes the principles for defining, calculating, and negotiating ICRs for federally funded research. The guidance describes the classification and types of allowable indirect costs; methods of allocating such costs; reasonableness of claimed costs; and exclusions and descriptions of unallowable cost elements, such as alcohol and bad debt. The Uniform Guidance also provides the basis for OMB to systematically collect information from federal agencies on all their federal financial assistance programs and establishes federal policies for providing this information to the public. NSF implements the Uniform Guidance using its Proposal and Award Policies and Procedures Guide and Grant General Conditions. The guide consists of NSF’s proposal preparation and submission guidelines as well as NSF’s policy and procedures used to award, administer, and monitor grants and cooperative agreements. NSF awards are subject to the Grant General Conditions, which list awardee responsibilities, obligations, and rights, and other conditions for awards. In addition, NSF’s CAAR staff use the Indirect Cost Rate Proposal Review Standing Operating Guidance as a guide for reviewing and negotiating indirect cost proposals from the awardees for which NSF has cognizance. The internal guidance includes descriptions of the three stages of ICR proposal review. Briefly these stages are: 1. Intake and Adequacy Review: This stage consists of steps for processing an incoming proposal including verifying that NSF has rate cognizance for the awardee and entering the proposal into the MTD, conducting initial reviews, and assigning the proposal to staff for review and rate negotiation. 2. Proposal Analysis and Rate Negotiation: This stage consists of steps for reconciling total expenditures in the financial statements with the total expenditures in the indirect cost proposal, reviewing indirect costs, verifying unallowable costs, preparing a trend analysis (including the organization’s indirect costs and rates for the previous 2 years, if available), and negotiating the indirect rate with the awardee. 3. Rate Approval and Issuance: This stage consists of steps for approving the indirect rate, preparing and transmitting the rate agreement to the awardee for signature, and closing out the proposal review by updating the MTD and filing working papers in appropriate files. Additionally, NSF’s internal guidance includes tools to assist awardees in preparing ICR proposals and to expedite NSF’s review process. These tools, as explained in further detail later in the report, are to be used in coordination with federal regulation and NSF guidance. For fiscal year 2000 through fiscal year 2016, the percentage of total annual award funding that NSF budgeted for indirect costs varied from year to year. This variation was based on various factors, such as by the types of activities supported by the awards and the types of organizations receiving the awards. Budgeted indirect costs on NSF awards ranged from 16 to 24 percent of the total annual amounts the agency awarded for fiscal years 2000 through 2016. The percentage fluctuated during this period, though it has generally increased since reaching a low point in 2010. In fiscal year 2016, NSF awards included approximately $1.3 billion budgeted for indirect costs, or about 22 percent of the total $5.8 billion that NSF awarded. Budgeted indirect costs were 16 percent of total annual amounts twice during the 17-year period—in fiscal years 2002 and 2010—and reached their highest point of 24 percent of total annual amounts in fiscal year 2015. Figure 1 illustrates annual funding for direct and indirect costs over the 17-year period. Indirect costs on individual awards within a given year varied more widely than the year-to-year variation for all NSF awards. Most NSF awards included indirect costs in their budgets—for example, the budgets for about 90 percent of the 12,013 awards that NSF made in fiscal year 2016 included indirect costs. Our analysis indicated that the funding for indirect costs ranged from less than 1 percent to 59 percent of the total award. Among the organizations whose ICR files we examined at NSF, examples of awards with indirect costs of less than 1 percent included awards to organizations that provided funding for (1) student travel to a networking workshop for women on computer and network systems and (2) a university workshop on modeling magnetic interactions between stars and planets. In contrast, examples of awards with indirect costs that ranged from 50 to 59 percent included awards to (1) a small business to study the atmosphere and improve current models of air quality and climate change and (2) a policy research organization to study how to broaden participation of underrepresented groups in the academic disciplines of science, technology, engineering, and mathematics. NSF officials told us that the overall composition of NSF’s awards portfolio varies from year to year and that indirect costs as a percentage of budgeted total award costs for any given year will reflect that variation. In particular, officials said that variation in indirect costs among individual awards—and thus variation in total award costs from year to year—can be due to several factors: Type of award supported activity: NSF’s awards support various types of activities, and NSF allows awardees to budget for indirect costs on these activities to varying degrees. For example, it does not allow indirect costs on its awards for stipends and travel of participants in NSF-sponsored conferences or training projects. In contrast, NSF allows administrative and clerical salaries to be allowed as indirect costs. However, these salaries may be considered direct costs if, for example, administrative or clerical services are integral to an activity or the costs are not recovered as an indirect costs. Type of research: NSF supports a range of research activities, some of which require investment in expensive infrastructure such as a telescope to study the universe. Which research activities are funded each year depends on a variety of considerations including the types of proposals submitted, the objectives of scientific research, the outcome of the merit review process, and available funding. Type of disciplinary field: NSF supports research in biological sciences, engineering, and social sciences, among others, and the level of indirect costs associated with awards in these fields can vary, according to NSF officials. In our analysis of awards made in fiscal year 2016, we found that the indirect costs varied among the NSF directorates that focus on different disciplinary fields. For example, budgeted indirect costs as a percentage of total annual amounts of awards were 22.5 percent in the directorate for engineering and 26.3 percent in the directorate for geosciences. Type of organization: NSF’s data categorized awardees as federal, industry, small business, university, or other—a category that includes nonprofits and individual researchers. Figure 2 illustrates our analysis on the average percentage of total awards budgeted for indirect costs in fiscal year 2016, by type of awardee. As shown in the figure, university awardees had the highest average indirect costs— about 27 percent of the total amount of awards—and federal awardees had the lowest average indirect costs—about 8 percent of the total amount of awards. According to NSF officials, certain types of projects, such as those carried out at universities, typically involve more indirect costs than others. They said that this outcome is because, for example, of the universities’ expense for maintaining scientific research facilities, which may be included as an indirect cost in awards. Universities accounted for about 91 percent of the approximately $1.3 billion NSF budgeted for indirect costs in fiscal year 2016. Awards to organizations for which NSF had cognizance (e.g., non-profits, professional societies, museums, and operators of large shared-use facilities) averaged lower budgeted indirect costs than awards to organizations for which other federal agencies had cognizance (e.g., universities for which HHS or DOD have cognizance). According to NSF officials, this variance resulted from differences in the organizations and the types of awards they receive and does not reflect differences in how agencies negotiate ICRs. As shown in figure 3, our analysis of NSF data indicates that on average, NSF budgeted about 23 percent of award amounts for indirect costs on awards to organizations for which NSF did not have indirect cost cognizance and about 11 percent for indirect costs on awards to organizations for which NSF had cognizance. In fiscal year 2016, NSF made over 90 percent ($5.4 billion of $5.8 billion) of its awards to organizations for which it did not have cognizance. NSF has developed internal guidance for setting ICRs, but NSF staff have not consistently followed it. In addition, the internal guidance on supervisory review does not include certain details and procedures. Specifically, the guidance does not include the criteria to be used by the supervisor for risk assessment and mitigation and the steps for reviewing and documenting the work performed by NSF staff when setting ICRs. It also does not include procedures for implementing new provisions issued under the Uniform Guidance and for performing oversight of the ICRs set by Interior on NSF’s behalf. Standards for Internal Control in the Federal Government states that management should design control activities, such as controls to ensure the accuracy and completeness of an entity’s data, and implement the controls through guidance, including guidance on each office’s responsibility for an agency’s operational processes. NSF has designed control activities for setting ICRs and has implemented them through internal guidance, such as Indirect Cost Rate Proposal Review Standing Operating Guidance. Our review of a nongeneralizable sample of seven NSF ICR agreement files showed that NSF staff followed many parts of the agency’s internal guidance. For example, as required by the internal guidance, NSF staff reviewed awardees’ cost policy statements to obtain an understanding of awardees’ accounting systems, cost allocation methods, and types of costs charged as either direct or indirect prior to setting ICRs. However, NSF staff did not follow two aspects of NFS’s internal guidance for setting ICRs: Use of tools and templates for setting ICRs: In accordance with NSF’s internal guidance, NSF has developed tools and templates along with procedures for using these tools to help staff conduct consistent reviews of the ICR proposals. For example, NSF’s internal guidance includes procedures for NSF staff to use a standard document checklist to verify that the awardee’s ICR proposal package is complete and that all required documents have been submitted. However, in our review of seven NSF ICR agreement files, we found that in all cases, staff did not use the standard document checklist to verify that the awardee had submitted all required documents. Instead, NSF staff used five different versions of the document checklist to track the receipt of the documents. Although these five versions of the document checklist contained similar awardee documentation, we also identified differences. For example, some checklists required certifications of lobbying costs and indirect costs, and others did not. According to the Uniform Guidance, awardees are required to submit these certifications with their proposals for ICRs. The staff were not consistently using the tools and templates NSF had developed because, according to NSF officials, NSF had not yet required them to do so. Without fully implementing this aspect of their guidance by requiring staff to use the standard tools and templates, the agency does not have assurance that NSF staff are consistently collecting required documentation from awardees and that ICRs are being set in accordance with federal requirements. Updating the Monitoring Tracking Database with current information: We identified instances where NSF staff did not consistently follow internal guidance for updating the MTD with current data about (1) the awardees for which NSF has cognizance and (2) the status of ICR proposals. NSF’s internal guidance requires staff to verify in the MTD that NSF has cognizance over the awardee prior to negotiating an ICR and to update information such as the date of receipt of the ICR proposal. In reviewing MTD reports, we identified 6 of 102 awardees for which NSF was the cognizant agency that were not included on a list of awardees for which NSF had cognizance. NSF officials confirmed that NSF was the cognizant agency for the six awardees and that the MTD had not been updated, which resulted in the awardees being incorrectly omitted from the database report. Additionally, we identified instances where NSF staff had not followed guidance to update the current status of awardees’ proposals, including instances where the MTD was missing either the received date or both the received and closed dates. NSF officials said that such errors resulted from either the agency’s incomplete reconciliation of the database, which NSF normally conducts on an annual basis after the end of the fiscal year, or from data entry errors. In order to achieve the agency’s objectives and adhere to requirements in federal internal control standards, it is essential for management to have accurate and complete ICR operational data. With accurate operational data for management to use in their decision making process, NSF could better ensure that it is managing the process for setting ICRs efficiently and in accordance with its internal guidance. NSF’s internal guidance includes various details on its control activities for setting ICRs, such as details on confirming the mathematical accuracy of the rates proposed and reconciling the total costs in the proposal, both allowable and unallowable, to the total costs shown on audited financial statements. As described in Standards for Internal Control in the Federal Government, including an appropriate level of such details allows for effective monitoring of an organization’s control activities. Specifically, through monitoring, management can assess the quality of work performed and promptly resolve any issues identified. However, we identified two areas of supervisory activities in which the procedures established by NSF did not include this level of detail. In particular, NSF’s existing internal guidance on supervisory activities, which are a key part of the agency’s control activities for setting ICRs, did not include details on (1) the criteria to be used by the supervisor to assess the risk level of a proposal and determine the types of review steps to be performed by staff for each risk level and (2) the steps that the supervisor needs to take when reviewing and documenting the work performed by NSF staff to set ICRs. According to NSF officials, risk assessments are typically performed by the supervisor for each proposal submitted and are intended to provide the basis for determining the scope of the ICR review, including the steps to be performed to mitigate the identified risks. NSF’s internal guidance includes procedures for performing risk assessment, such as reviewing past issues in the ICR negotiation for an awardee and changes in the awardee’s accounting system. According to NSF officials, part of conducting the risk assessment includes the supervisor’s categorizing the identified risk in levels of high, medium, or low and determining steps for staff to perform to mitigate risks at each level. In our sample of seven NSF ICR agreement files we reviewed, the supervisor used various criteria for assessing risks, such as NSF’s funding levels and the awardee’s rate history. Depending on the risk level, the supervisor directed staff to perform additional review steps, such as transaction testing. However, NSF’s internal guidance does not include these details on the criteria for the supervisor to use when categorizing the level of risk and the steps for mitigating the risks at each level. Supervisory review is a type of control activity that aids in providing reasonable assurance that staff follow agency procedures for setting ICRs. NSF’s internal guidance includes broad procedures for supervisory review, such as requiring the supervisor to review the staff’s completed proposal package and the applicable rates and document any concerns identified during the review; however, the guidance does not include details on what steps the supervisor needs to perform when (1) reviewing the completed proposal package and applicable rates, and (2) annotating the results of the review in the working papers. For example, the internal guidance does not include procedures requiring the supervisor to ensure that staff have adequately performed and documented key controls identified in the internal guidance, such as analyzing trends in awardees indirect costs. In our review of seven NSF ICR agreement files, we found that the files included documents supporting key controls such as trend analysis; however, we did not find any documentation that the supervisor had reviewed the work performed by staff and annotated the results of the reviews. In a September 2016 report, we identified similar issues in which agencies lacked detailed supervisory procedures, resulting in supervisors approving rates that were set by staff who did not perform control activities required by agencies’ internal guidance. NSF officials explained that the office that sets ICRs is relatively small— consisting of a single supervisor and several staff who review ICR proposals—and that as a result, supervisory activities are not as fully documented in guidance as they would be in a larger office. Additionally, the officials stated that because of the complexity of the ICR setting process, the supervisor directly discusses any concerns about a completed proposal package with the staff instead of documenting such concerns. However, more detailed internal guidance on supervisory activities could help NSF management ensure that ICRs are set consistently and in accordance with federal guidance and decrease the risk that the supervisor could approve rates that were not properly executed by staff—for example, when a new person assumes the supervisory position, as occurred in 2016. We found that NSF internal guidance does not include (1) procedures for implementing new provisions issued under the Uniform Guidance and (2) procedures in its internal guidance for performing oversight of the ICRs set by Interior on NSF’s behalf. The Uniform Guidance became effective for grants awarded on or after December 26, 2014, and NSF implemented the Uniform Guidance through its policies and procedures. The Uniform Guidance included several new provisions for research organizations, such as an option to apply for a onetime extension of an ICR for up to 4 years and an option to use a de minimis ICR of 10 percent if the organization has not previously had a negotiated ICR with the cognizant agency. However, NSF’s internal guidance for setting ICRs does not include specific procedures that NSF staff should perform to implement certain aspects of the Uniform Guidance’s new provisions. For example, the internal guidance does not specify criteria for determining whether an awardee is eligible for an extension. NSF officials stated that they updated the internal guidance to include Internet links to the Uniform Guidance and that they expect NSF staff to speak to a supervisor for clarification on questions about applying the new provisions. However, the links in the internal guidance directed staff to outdated OMB guidance rather than to the Uniform Guidance. NSF officials stated that they had incorrectly added links in the guidance. By adding procedures in its internal guidance for implementing the new provisions, NSF could better ensure that NSF staff will apply the provisions correctly in accordance with the Uniform Guidance. Standards for Internal Control in the Federal Government states that management should establish and operate monitoring activities for its internal control system and evaluate the results. In addition, when an entity engages an external party to perform certain operational processes, management retains responsibility for monitoring the effectiveness of internal control activities performed by the external party. NSF has engaged an external party to provide assistance in negotiating ICRs. Specifically, in 2009, NSF entered into an interagency agreement with Interior to negotiate ICRs for a portion of the awardees for which NSF has cognizance. For the first 10 months of fiscal year 2017, Interior had negotiated ICRs for approximately 33 percent of the awardees for which NSF had cognizance. According to NSF officials, Interior, acting as an agent for NSF, has the responsibility for reviewing ICR proposals and setting ICRs directly with awardees, following OMB requirements. In addition, NSF’s internal guidance states that Interior will set ICRs in accordance with NSF protocols. However, according to NSF officials and internal guidance, Interior’s approval and signing of ICRs are largely independent of NSF, and in addition, NSF conducts limited monitoring of the ICRs that Interior negotiates. For example, NSF officials said that they review monthly summaries of rates negotiated by Interior and completed rate agreements and that on request, they meet with Interior’s supervisors and participate in problem resolution. However, they do not review Interior’s ICR files, such as checking the adequacy of documentation submitted by awardees or the accuracy and reasonableness of the calculations and resulting rates proposed by the awardee. NSF officials stated that it was their understanding that by entering into the interagency agreement, they delegated the authority for performing ICR proposal reviews and setting ICRs to Interior. However, under Standards for Internal Control in the Federal Government, NSF still retains cognizance and oversight responsibilities for the ICRs set by Interior. By including procedures in its internal guidance for overseeing work performed by Interior, NSF could ensure that the ICRs set by Interior comply with federal guidance and NSF protocols. Setting ICRs in accordance with federal guidance is important to ensuring that federal agencies do not pay more than their share of awardees’ indirect costs. NSF has developed internal guidance to help ensure that ICRs are set appropriately, and NSF staff follow many parts of the guidance. However, staff have not consistently followed the guidance for using certain tools and templates for setting ICRs or guidance for updating the agency’s database to reflect the status of awardees and their ICR proposals—for example, because NSF has not yet required them to do so. Details on supervisory activities are also not included in the guidance, including the criteria used by the supervisor for assessing the risk level of a proposal and determining specific steps for mitigating risks, and steps supervisors take for their reviews of work performed to set ICRs. Additionally, the guidance does not include procedures necessary to carry out new provisions of the Uniform Guidance and to monitor the ICRs set by Interior on NSF’s behalf to ensure that they comply with federal guidance and NSF protocols. NSF officials described ways that staff implement these procedures even though the procedures are not fully detailed or included in guidance. However, including the missing details and procedures in NSF’s internal guidance, and requiring staff to follow the guidance, could help NSF ensure that staff properly and consistently negotiate ICRs and that the rates negotiated comply with applicable federal guidance, which in turn would help ensure that the funding provided for indirect costs does not unnecessarily limit the amount available for research. We are making the following three recommendations to NSF: The Director of NSF should require staff to follow written internal guidance for (1) using tools and templates NSF has developed for the process for setting indirect cost rates and (2) updating the agency’s database to reflect the status of awardees for which NSF has cognizance and of indirect cost rate proposals. (Recommendation 1) The Director of NSF should add details to NSF’s internal guidance for setting indirect cost rates specifying (1) the criteria to be used by the supervisor for assessing the level of risk and steps for mitigating the risks at each level and (2) the steps for supervisory review of the process for setting indirect cost rates and documentation of the results of the review. (Recommendation 2) The Director of NSF should add procedures to NSF’s internal guidance for (1) implementing the applicable new provisions of the Uniform Guidance, including updating links to OMB guidance, and (2) monitoring the indirect cost rates that the Department of Interior sets on NSF’s behalf. (Recommendation 3) We provided a draft of this report to NSF and Interior for review and comment. In its comments, reproduced in appendix I, NSF concurred with our recommendations and described actions it would take to address them. These actions include updating and adding details and procedures to its internal guidance for setting ICRs, reviewing and updating the tracking database on a quarterly basis, and working with Interior to establish procedures for monitoring ICRs set by Interior on NSF’s behalf. NSF stated that these actions will improve NSF’s protocols for negotiating ICRs. Interior stated that it did not have comments on our draft report. We are sending copies of this report to the appropriate congressional committees; the Director of the National Science Foundation; Secretary of the Department of the Interior; and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact John Neumann at (202) 512-3841 or neumannj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. In addition to the contacts named above, key contributors to this report were Joseph Cook, Assistant Director; Kim McGatlin, Assistant Director; Rathi Bose; Ellen Fried; Cindy Gilbert; Ruben Gzirian; Terrance Horner, Jr.; Diana Lee; David Messman; Kathryn Smith; and Sara Sullivan. NIH Biomedical Research: Agencies Involved in the Indirect Cost Rate- Setting Process Need to Improve Controls. GAO-16-616. Washington, D.C.: September 2016. Federal Research Grants: Opportunities Remain for Agencies to Streamline Administrative Requirements. GAO-16-573. Washington, D.C.: June 2016. Grants Management: Programs at HHS and HUD Collect Administrative Cost Information but Differences in Cost Caps and Definitions Create Challenges. GAO-15-118. Washington, D.C.: December 2014. Biomedical Research: NIH Should Assess the Impact of Growth in Indirect Costs on Its Mission. GAO-13-760. Washington, D.C.: September 2013. University Research: Policies for the Reimbursement of Indirect Costs Need to be Updated. GAO-10-937. Washington, D.C.: September 2010. National Institutes of Health Extramural Research Grants: Oversight of Cost Reimbursements to Universities. GAO-07-294R. Washington, D.C.: January 2007.

NSF awards billions of dollars to institutions of higher education (universities), K-12 school systems, industry, science associations, and other organizations to promote scientific progress by supporting research and education. NSF reimburses awardees for direct and indirect costs incurred for most awards. Direct costs, such as salaries and equipment, can be attributed to a specific project that receives an NSF award. Indirect costs, such as the costs of operating and maintaining facilities, are not directly attributable to a specific project but are necessary for the general operation of an awardee's organization. For certain organizations, NSF also negotiates ICR agreements, which are then used for calculating reimbursements for indirect costs. ICR negotiations and reimbursements are to be done in accordance with federal guidance and regulation and NSF policy. GAO was asked to review the amount of NSF funding for indirect costs and NSF's negotiation of ICRs. This report examines (1) what is known about indirect costs on NSF awards over time, and (2) the extent to which NSF has implemented guidance for setting ICRs for organizations over which it has cognizance. GAO reviewed relevant regulations, guidance, and agency documents; analyzed budget data and a nongeneralizable sample of nine ICR files from fiscal year 2016 selected based on award funding; and interviewed NSF officials. For National Science Foundation (NSF) awards during fiscal years 2000 through 2016, budgeted indirect costs varied from 16 to 24 percent of the total annual amounts the agency awarded. The percentage fluctuated during this period, though this percentage generally has increased since reaching a low point in 2010. The variation from year-to-year was based on various factors such as by the types of activities supported by the awards and the types of awardee organizations receiving the awards. Note: Award funding has not been adjusted for inflation. NSF has developed internal guidance for setting indirect cost rates (ICR) but has not consistently implemented this guidance and has not included certain details and procedures, in particular: NSF has not consistently implemented its guidance because it has not yet required NSF staff to follow aspects of its guidance, such as using a documentation checklist that NSF developed to verify that an awardee's ICR proposal package is complete. NSF did not include details on supervisory activities, such as the criteria to be used by the supervisor of the ICR process for assessing an ICR proposal's risk level and mitigating risks at each level. NSF did not include certain procedures, such as for implementing new provisions of federal guidance on setting ICRs. NSF officials described ways that staff implement procedures even though the procedures are not fully detailed or included in guidance. Nevertheless, with complete guidance that includes the missing details and procedures and that is consistently followed, NSF could better ensure that ICRs are set consistently and in accordance with federal guidance on indirect costs and with federal internal control standards. GAO recommends that NSF take three actions to improve its guidance for setting ICRs, including adding certain details and procedures. NSF concurred with GAO's recommendations and described plans to address them.

The federal government has helped provide affordable housing to low- income households since the 1930s. Since then, a number of federal housing programs have subsidized the construction of housing for the poor, provided rental assistance to tenants in existing privately owned housing, and insured mortgages for both single- and multifamily properties. Today, HUD administers the majority of federal housing assistance programs in urban areas, and USDA’s Rural Housing Service (RHS) implements housing programs in rural areas. In general, both HUD and USDA programs target families at lower income levels. HUD programs target families with incomes that are: extremely low (no more than 30 percent of an area’s median), very low (no more than 50 percent of an area’s median), and low (no more than 80 percent of an area’s median). USDA programs also target families with incomes that are very low and low. In addition, some USDA programs target families with moderate incomes (no more than 115 percent of an area’s median). HUD and USDA provide many types of housing assistance programs, including single-family programs, multifamily programs, rental assistance programs, and public housing. Housing developments can be assisted by multiple programs. For example, a loan or mortgage on a multifamily property may be insured through a HUD or USDA program, and the property may have tenants that receive rental assistance from these agencies. Federal housing assistance can generally be categorized as follows: Single-family programs that provide mortgage insurance, loan guarantees, or direct loans for homeowners and grants or loans for home repairs or modifications; Multifamily programs that provide loans, subsidies, mortgage insurance or loan guarantees, or a combination of these to support the development and rehabilitation of rental properties, including: Production programs that provide federal funds to construct or substantially rehabilitate units for households with extremely low to moderate incomes; and Mortgage insurance/loan guarantee programs that provide incentives for lenders to finance rental housing by reducing risk; Rental assistance programs, which can be used for multifamily and some single-family housing and generally pay property owners the difference between 30 percent of a household’s adjusted income and its rent, including: Tenant-based rental assistance that provides vouchers for eligible tenants to rent privately owned apartments or single-family homes and that tenants can use in new residences if they move; and Project-based rental assistance that is attached to specific properties and is available to tenants only when they are living in units at these properties; and Public housing, which is provided through HUD and offers units for eligible tenants in properties administered by public housing authorities. The RHS implements all of the USDA programs we reviewed. Two offices administer the HUD programs covered in this report: The Office of Housing, which administers multifamily subsidy and mortgage insurance programs, multifamily production programs, and a variety of single-family programs; and The Office of Public and Indian Housing, which administers the Public Housing, Housing Choice Voucher, and Section 8 Moderate Rehabilitation Programs. HUD has specific goals for increasing housing opportunities for the elderly. As outlined in its fiscal year 2004 Annual Performance Plan, these goals include (1) increasing the availability of affordable housing for the elderly, (2) increasing the number of assisted-living units, (3) increasing the number of elderly households living in privately owned, federally assisted multifamily housing served by a service coordinator, and (4) increasing elderly families’ satisfaction with their Section 202 units. USDA does not have specific goals related to the elderly in its fiscal year 2004 Annual Performance Plan. However, USDA does have the broad objective of improving the quality of life of rural families by financing, among other things, quality housing. To reach this objective, USDA has established two housing-related goals: (1) to increase financial assistance to rural households to buy a home, and (2) to increase the number of minority homeowners. We identified a total of 23 federal housing programs that are targeted at or have special features for the elderly: 2 that are intended for the elderly only, 3 that are targeted to the elderly and disabled, and another 18 that have special features, such as properties designated for elderly occupants and income adjustments that lower elderly households’ rental payments. Some or all units in many multifamily and public housing properties are designated for the elderly. Data on the number of elderly served are not available for each program; however, analysis of available data shows that the elderly occupied at least 1.3 million units provided through several of these programs. Limited information is available about the effectiveness of most housing programs in assisting elderly households. HUD and USDA offer 23 housing programs that either target or have special features for the elderly. Two programs—HUD’s multifamily Section 202 program and USDA’s single-family Section 504 Rural Housing Repair and Rehabilitation Grants program—are currently intended for elderly beneficiaries only. Three multifamily programs target the elderly and the disabled: HUD’s ALCP and two mortgage insurance programs (Section 231 and Section 232). Although not targeted to the elderly, another 18 programs have special features for them. For example, public housing and some multifamily programs allow properties to be wholly or partially designated for occupancy by the elderly. Also, rental assistance programs and USDA single-family programs make adjustments to elderly households’ incomes when determining program eligibility or calculating benefits. These programs are briefly described in Figure 1, and in more detail in Appendix III. Although most of the 23 programs are intended for households with low, very low, or extremely low incomes, 7 of the multifamily insurance programs may provide benefits for households of all income levels (that is, there are no income restrictions or limits). Among these 23 programs, 13 have approximately 943,000 units designated for the elderly, based on available data (fig. 2). Some public and multifamily housing programs may designate properties wholly or partially for the elderly, but the number of units set aside varies across programs. As figure 2 shows, the Section 202 and Section 231 programs have designated most of their units for the elderly (84 percent and 90 percent, respectively), consistent with these programs’ targeting of the elderly or the elderly and disabled. The Section 202 program designates the most units for the elderly—about 268,000 units. HUD, USDA, and other agencies also administer other housing programs for which the elderly are eligible, but which do not have special features for the elderly. For example, under the Low-Income Housing Tax Credit program, the largest active multifamily housing production program, states receive annual allocations of tax credits and distribute them at their discretion. In their guidelines for the distribution of tax credits, some states have established preferences for properties intended for the elderly, but the federal government does not require such preferences. Similarly, HUD’s single-family mortgage insurance program does not have features that apply only to elderly borrowers. HUD's Home Equity Conversion Mortgage program, which targets only elderly homeowners, did not meet our definition of housing assistance for this review. This program allows elderly homeowners to borrow against the equity in their homes and defer repayment for as long as they occupy their homes, but the money can be used for any purpose, not just housing. Appendix I explains our methodology for determining which programs to include, and Appendix IV provides brief descriptions of some programs that were excluded from this review. Elderly households occupied at least 1.3 million units provided through HUD and USDA rental assistance, public, and multifamily housing programs for which data on occupancy were available between April and June 2004 (fig. 3). The household counts by program do not match the unit counts presented in figure 2 and Appendix III, because some units may have been vacant or current data on tenants may have been incomplete. The data include households in units designated for the elderly as well as other elderly households receiving assistance through the programs. Overall, elderly households comprised approximately 30 percent of the households participating in programs for which data were available. In addition to the elderly households in public and multifamily housing programs, at least 69,650 elderly households received loans or grants through 4 USDA single-family programs from October 1995 through April 2004, the time period for which we were able to obtain data (fig. 4). Specifically, USDA’s Section 504 grants program, which provides home repair grants to elderly homeowners, made at least 40,697 grants to elderly recipients during that period. Also, the Section 502 Rural Housing Loans (Direct), Section 502 Direct Housing Natural Disaster Loans, and Section 504 Rural Housing Repair and Rehabilitation Loans programs, which offer loans for home purchase or repair, had at least 28,953 elderly borrowers. USDA generally did not have data on the age of borrowers in its Section 502 guaranteed loan program, which had more than 167,500 active loans as of April 2004. Program effectiveness can be assessed in a number of ways—for example, by evaluating the extent to which programs reach or serve intended beneficiaries or examining the efficiency with which they serve those that they do reach. In evaluation research, “effectiveness” is often defined in terms of the achievement of program goals or outcomes. While HUD and USDA have not established specific goals for each housing program that offers assistance for the elderly in their annual Performance and Accountability Reports, these reports contain limited information on some programs’ performance—that is, the extent to which a program met specific goals and objectives or contributed to the attainment of larger overarching goals. The Office of Management and Budget (OMB), has used its standardized Program Assessment Rating Tool (PART) to assess five of the programs, finding that three had not demonstrated results, one was ineffective, and one was moderately effective. While we also identified academic studies of some programs we reviewed, they generally did not evaluate the effectiveness of a program in terms of the extent to which it reached its goals. For example, some academic studies analyzed the impact of the public housing and housing choice voucher programs upon property values. However, affecting property values is not a goal of these programs. Finally, we also identified information on how effectively some of the housing programs are implemented. Of the two programs that are targeted to the elderly, only HUD’s Section 202 program is included in the agencies’ fiscal year 2003 Performance and Accountability Reports. HUD addresses this program under its overall goal of improving housing options for the elderly. According to HUD’s Performance and Accountability Report, HUD exceeded its goal of approving 250 Section 202 and Section 811 projects to start construction during fiscal year 2003, approving 334 projects. While HUD exceeded this goal, the number of projects approved for construction is a limited measure of the effectiveness of the Section 202 program. USDA’s fiscal year 2003 Performance and Accountability Report did not address the effectiveness of the Section 504 program. OMB rated the Section 202 program in its fiscal year 2004 PART assessment. PART is designed to assess the effectiveness of federal programs through a series of diagnostic questions intended to provide a consistent approach to rating federal programs. Drawing on available performance and evaluation information, the questionnaire attempts to determine the strengths and weaknesses of federal programs with a particular focus on individual program results. According to this assessment, the results of the Section 202 program had not been demonstrated, and: The program lacked evidence showing the overall level of impact that the program had on poor elderly individuals; HUD had not established quantifiable long-term performance goals with outcomes for this program, so progress could not be measured; The program had produced about 6,000 units per year, yet there is a need for over a million units for the elderly. We also have reported specifically on how effectively the Section 202 program has been managed. Our 2003 report on the Section 202 program outlined several factors that prevented efficient and effective implementation. Appendix II contains an update on HUD’s efforts to improve these deficiencies. HUD’s fiscal year 2003 Performance and Accountability Report also provides information on two of the three HUD programs targeted to the elderly and disabled—the Assisted Living Conversion Program and the Section 232 mortgage insurance program. According to this report, HUD: Met its goal for increasing the number of assisted living units for the elderly by adding 2,618 units or beds to the estimated 18,000 already in place in 325 properties insured by the Section 232 mortgage insurance program; Exceeded its goal of converting 10 properties through the ALCP by converting 13 properties, adding an additional 407 assisted living units. While performance and accountability information does not directly address the extent to which programs that have special features for the elderly—such as HUD’s public housing, project-based rental assistance, and housing choice voucher programs—are effective in assisting the elderly, it does provide additional context. For example, HUD’s fiscal year 2003 Performance and Accountability Report has an indicator to track the share of units receiving assistance through these programs that are occupied by the elderly, but HUD has not established goals for this indicator because housing providers have discretion regarding admissions policies. USDA’s Office of Inspector General (OIG) has reported on how effectively USDA’s multifamily housing programs are implemented. For example: In 2004, the OIG reported that USDA’s Rural Housing Service missed by three projects its target of having no more than 140 multifamily housing projects with accounts more than 180 days past due. According to the report, this indicator is a measure of how effectively and efficiently the multifamily housing loan portfolio is being managed. Also in 2004, the OIG reported that RHS had not implemented all of the policy changes that it had agreed to implement to better monitor the owners of Section 515 developments. As a result, according to the OIG, rural rental housing funds remained vulnerable to theft and abuse. In addition, we identified two studies that provided information related to the effectiveness of housing choice vouchers in assisting the elderly and the effectiveness of the Section 502 Rural Housing Loans (Direct) Program in helping the elderly become homeowners and obtain better quality housing: A 2001 study on the Housing Choice Voucher Program showed that the elderly had less success finding and leasing a unit than other household types, possibly because elderly renters may have difficulty looking at multiple units. A 1999 study of USDA’s Section 502 Rural Housing Loans (Direct) Program, which provides loans to very-low and low-income rural residents for the purchase or repair of single-family homes, reported the results of a 1998 survey and found that of respondents with at least one elderly person on the mortgage, 30 percent were first-time homebuyers, and almost 90 percent reported that their current home was of better quality than their previous home. Finally, OMB’s fiscal year 2004 PART assessments included information about the effectiveness of 4 of the 18 programs that provide special features for the elderly—the Housing Choice Voucher Program, project- based rental assistance, and the Sections 515 and 521 programs (table 1). While Performance and Accountability Reports can be helpful in assessing program effectiveness, their usefulness is limited. For example, USDA’s fiscal year 2003 report provides little useful information on the effectiveness of USDA’s housing programs in assisting the elderly because of the lack of specific goals and objectives related to improving housing options for the elderly. Similarly, HUD’s fiscal year 2003 report identifies such goals for only 3 of the 23 programs we reviewed. Further, HUD’s goals are not necessarily specific; for example, although information was available on HUD’s Section 232 program, targeted goals were not established for this program, such as increasing the number of assisted- living units through Section 232 insurance by a specific percentage. Without such specific criteria, HUD management and outside evaluators lack the essential information needed to assess this aspect of the program’s effectiveness. In addition, GAO has previously reported that the usefulness of PART assessments is limited, for the following reasons: Many PART questions contained subjective terms that were open to interpretation. We noted that such subjective terminology could influence program ratings by permitting OMB staff’s views about a program’s purpose to affect assessments. OMB assigned overall program ratings and individual section scores. Overall ratings encourage a determination of the effectiveness of a program even when performance data are unavailable, the quality of those data is uneven, or they convey a mixed message on performance. OMB inconsistently defined appropriate measures—outcomes versus outputs—for programs. Most other studies we identified on various aspects of the programs we reviewed did not evaluate either the programs’ overall effectiveness, or their specific effect on the elderly. We also found some academic studies that provided information related to effectiveness but that were based on a prohibitively small sample of elderly program participants or did not use reasonably current data. The overall lack of information on the effectiveness of the programs we reviewed in assisting the elderly may be due to the fact that, as GAO has previously reported, many agencies lack the capacity to undertake program evaluations to assess a federal program’s contributions to results. Agencies’ capability to gather and use performance information has posed a persistent challenge. Finally, we did not evaluate the extent to which the programs we reviewed addressed the needs of eligible elderly households because (1) complete data on the number of elderly households occupying units provided by federal housing programs were not available, and (2) eligibility criteria for each program varied, making it difficult to establish the number of eligible elderly households. Generally, HUD and USDA’s housing assistance programs are not required to provide supportive services to the elderly. Of the 23 housing assistance programs that we reviewed, 4 required the owners of properties developed under the programs to ensure that supportive services were available. HUD has programs that link the elderly to or provide them with supportive services. Two of these programs, the Service Coordinator and the Resident Opportunities and Self Sufficiency (ROSS) programs, link residents with appropriate supportive services that are available in the community. In addition, the Congregate Housing Services Program funds meals and other needed services in public and multifamily housing properties, and the Neighborhood Networks program provides resources for establishing computer centers at such sites. Owners of public and multifamily housing may also provide supportive services by establishing partnerships with public and private organizations in the community. HUD’s Section 202 program, the ALCP, and Section 232 Mortgage Insurance, and USDA’s Section 515 Congregate Housing Program, which is a sub-program of the Section 515 program (see Appendix III) require property owners to make supportive services available to their residents. Generally, HUD and USDA do not provide funding for these services. The property owners typically obtain other funds to provide supportive services or must ensure that appropriate services are available in the community (see table 2). While USDA does not generally provide funding for supportive services for residents of federally assisted housing, HUD has two programs that link residents of public and multifamily properties developed under HUD programs to supportive services, and two that provide supportive services. None of these four programs is just for the elderly, but they either can be used in properties designated for the elderly or have funding specifically for the elderly. The Service Coordinator Program, for example, provides funding for managers of multifamily properties designated for the elderly and disabled to hire coordinators to assist residents in obtaining supportive services from community agencies. These services, which may include personal assistance, transportation, counseling, meal delivery, and healthcare, are intended to help the elderly live independently and to prevent premature and inappropriate institutionalization. Service coordinators can be funded through competitive grant funds, residual receipts (excess income from a property), or rent increases. According to HUD’s fiscal year 2003 Performance and Accountability Report, service coordinators were serving more than 111,000 units in elderly properties. Similarly, HUD’s ROSS grant program links residents with appropriate services. This program differs from the Service Coordinator Program in that it is designed specifically for public housing residents. The ROSS program has five funding categories, including the Resident Service Delivery Models for the Elderly and Persons with Disabilities (Resident Services) and the Elderly/Disabled Service Coordinator Program. Resident Services funds can be used to hire a project coordinator; assess residents’ needs for supportive services and link residents to federal, state, and local assistance programs; provide wellness programs; and coordinate and set up meal and transportation services. The Elderly/Disabled Service Coordinator Program has not provided new grants since 1995 but still services existing grants. The Congregate Housing Services Program provides grants for the delivery of meals and nonmedical supportive services to elderly and disabled residents of public and multifamily housing, including USDA’s Section 515 housing. While HUD provides up to 40 percent of the cost of supportive services, grantees must pay at least 50 percent of the costs, and program participants pay fees to cover at least 10 percent. Like the Elderly/Disabled Services Coordinator Program under ROSS, the Congregate Housing Services Program has provided no new grants since 1995, but Congress has provided funds to extend expiring grants on an annual basis. In addition, the Neighborhood Networks program encourages property owners, managers, and residents of HUD-insured and -assisted housing to develop computer centers. Although computer accessibility is not a traditional supportive service for the elderly, a senior HUD official told us that having computers available enhances elderly residents’ quality of life. HUD does not fund each center's planned costs, but encourages property owners to seek cash grants, in-kind support, and donations from sources such as state and local governments, educational institutions, private foundations, and corporations. Elderly residents of public and federally subsidized multifamily housing can also receive supportive services through partnerships between property owners and local organizations and through programs provided by the Department of Health and Human Services (HHS). For example, property owners can establish relationships with local nonprofit organizations, including churches, to ensure that residents have access to the services that they need. At their discretion, property owners may establish relationships that give the elderly access to meals, transportation, and housekeeping and personal care services. In site visits to HUD and USDA multifamily properties, we found several examples of such partnerships: In Greensboro, North Carolina, Dolan Manor, a Section 202 housing development, has established a relationship with a volunteer group from a local church. The volunteer group provides a variety of services for the residents, such as transportation. In Plain City, Ohio, residents of Pleasant Valley Garden, a Section 515 property, receive meals five times a week in the community’s senior center (a $2 donation is suggested). A local hospital donates the food and a nursing home facility prepares it. Volunteers, including residents, serve the meals. The senior center uses the funds collected from the lunch for its activities. In addition, local grocery stores donate bread products to the senior center daily. The United Way provides most of the funding for the senior center. In Guthrie, Oklahoma, Guthrie Properties, a Section 515 property, has established a relationship with the local Area Agency on Aging. The agency assists residents of Guthrie Properties in obtaining a variety of services, including meals and transportation to a senior center. Some elderly residents of public and federally subsidized housing may also obtain health-related services through programs run by HHS. For example, HHS’s Public Housing Primary Care Program provides public housing residents with access to affordable comprehensive primary and preventive health care through clinics that are located either within public housing properties or in immediately accessible locations. The program awards grants to public and nonprofit private entities to establish the clinics. The organizations must work with public housing authorities to obtain the physical space for the clinics and to establish relationships with residents. Currently, there are 35 grantees, 3 of which are in rural areas. According to a program administrator, although clinics are not specifically geared toward elderly-designated public housing, they can be established at such properties. Elderly residents of federally subsidized housing may also be eligible for the Medicaid Home and Community-Based Services (HCBS) Waiver Program, which is administered by HHS’s Centers for Medicare and Medicaid Services. Through this waiver program, individuals eligible for Medicaid can receive needed health care without having to live in an institutional setting. HUD has identified the use of these waivers as an innovative model for assisting the frail elderly in public housing. In addition, eligible elderly residents of federally subsidized housing may also receive health care through the Program of All-Inclusive Care for the Elderly (PACE), which is also administered by the Centers for Medicare and Medicaid Services. Like the HCBS waiver program, this program enables eligible elderly individuals to obtain needed services without having to live in an institutional setting. The program integrates Medicare and Medicaid financing to provide comprehensive, coordinated care to older adults eligible for nursing homes. Figure 5 provides information on the housing assistance programs that can use federally funded supportive services programs that assist the elderly. Although the potential for duplication exists, HUD and USDA have established policies and procedures to guide the development of multifamily housing for the elderly in rural areas. A 1991 agreement between the agencies and subsequent guidance to HUD and USDA field offices established a framework for coordinating efforts to provide housing assistance to low-income rural households. As noted, HUD develops rental housing for the elderly in rural areas through its Section 202 program, and USDA can develop such housing through its Section 515 program. In addition to obtaining one another’s input on proposed developments, HUD field office and state USDA office staffs assess the markets in areas where a new development is proposed. Site visits to HUD and USDA field offices in three states revealed that while staff did not consistently follow coordination procedures, each office did analyze market conditions in the proposed locations. In addition, the potential for unnecessary overlap and duplication between these programs has been limited by funding levels and the geographic areas in which HUD and USDA develop new housing. Policies and procedures designed to coordinate HUD and USDA efforts to develop rental housing in rural areas have been in place since the early 1990s. In 1991 HUD and USDA signed a Memorandum of Understanding and agreed to maintain an on-going working relationship to address issues related to providing housing assistance to rural areas in a cooperative, cost effective, and nonduplicative manner. As a result of this agreement, HUD and USDA issued guidance, in 1991 and 1992 respectively, specifying how the agencies should coordinate. Specifically, the guidance outlined coordination procedures that each agency should follow when reviewing applications for funds to develop rental units. For HUD, this policy applied to several programs, including the Section 202 program. For USDA, this policy applied specifically to the Section 515 program, which can be used to develop properties for families or the elderly. Among the programs we included, HUD’s Section 202 program and USDA’s Section 515 program are the only two that actively produce rental units for the elderly in rural areas. While neither the original agreement between HUD and USDA nor the resulting guidance has been updated since the early 1990s, the Section 515 program instructions and the Section 202 program’s annual Notice of Funding Availability, which announces the availability of funding as well as program requirements, describe current procedures. According to the Section 515 program instructions, the purpose of the coordination effort is to (1) foster better communication, (2) obtain additional documentation to determine market feasibility, (3) prevent overdevelopment of subsidized housing, and (4) prevent adverse effects on proposed or existing units that provide similar types of rental housing. The program instructions also state that state USDA offices will forward basic loan information on Section 515 loan applications that are selected for further processing to the applicable HUD field office. HUD field office staff will identify any pending, authorized, or existing units in the market area and provide comments, positive or negative, on the proposed market area to USDA within 2 weeks. When HUD staff have concerns about market feasibility or the impact of a proposed project on existing or authorized HUD units, they also provide documentation to support their concerns. HUD has established a similar process for notifying USDA of proposed developments. HUD’s fiscal year 2004 Notice of Funding Availability for the Section 202 program requires HUD to seek USDA’s input on Section 202 applications, giving USDA the opportunity to respond if it has concerns about the demand for additional assisted housing or possible harm to existing housing in the same market area. In addition to seeking input from one another on the markets in which new rural rental housing is proposed, both HUD and USDA assess market conditions when they evaluate applications for Section 202 and Section 515 funds. For example, HUD’s guidance instructs HUD’s economists to evaluate the markets in which all Section 202 applications propose development to determine if sufficient demand for the units exists and to assure that any new units will not have a long-term adverse impact on existing assisted housing for the elderly. If this analysis shows that sufficient demand for the units proposed in an application does not exist, then the application cannot be funded. Similarly, according to USDA officials, USDA state office staffs analyze market data to determine need and demand for the units proposed in Section 515 applications. Applications are not approved if: another rural rental housing loan has already been selected for further processing in the same market; a previously authorized USDA, HUD, Low-Income Housing Tax Credit, or similar type of project in the same market area has not been completed, has not reached its projected occupancy level, or is experiencing high vacancies; or the need in the market area is for additional rental assistance and not for additional housing units. We visited HUD field offices and USDA state offices in North Carolina, Ohio, and Oklahoma to determine whether and how these offices were following these policies and procedures. We chose these states because they had the largest numbers of approved Section 202 grants in rural areas and Section 515 loan awards in fiscal years 2002 and 2003 (see Appendix I). Our observations from the site visits are not necessarily representative of all field offices. Overall, the HUD and USDA field offices in the three states we visited did not consistently follow the policies and procedures designed to facilitate coordination in fiscal years 2002 or 2003. For example, while local USDA officials in North Carolina and Oklahoma obtained input from HUD on Section 515 loan applications in fiscal years 2002 and 2003, USDA officials in Ohio did not. According to a senior official from USDA’s Ohio state office, the agency did not seek HUD’s input on the sites funded in fiscal years 2002 or 2003 because USDA determined that market demand existed for the units. Moreover, this official stated that they had sought HUD’s input on Ohio’s list of “designated places”—cities, towns, and communities for which USDA could approve new Section 515 development. Based on input from HUD and USDA field offices, USDA state offices can remove places from the list if a market for additional rental housing does not exist. Since HUD officials had not raised concerns about the two places on Ohio’s list for which Section 515 housing was proposed in fiscal years 2002 and 2003 and funds were ultimately allocated, USDA officials did not think that it was necessary to request their input again. Only one of the HUD offices visited sought USDA’s input on Section 202 grant applications in fiscal year 2002, and none sought USDA’s input in fiscal year 2003. According to HUD officials in the field offices we visited, HUD offices did not seek input for various reasons. For example, in North Carolina and Oklahoma there were staffing changes. In Ohio, funds were awarded to a site that had been funded in fiscal year 2002. Since contact was made with USDA officials in fiscal year 2002 regarding this site, HUD officials did not see a need to contact them again in fiscal year 2003. Although coordination between the HUD and USDA offices we visited was inconsistent in fiscal years 2002 and 2003, we found that these offices based their funding decisions on market analyses. For example, HUD economists evaluated the markets in which all Section 202 applications proposed development. If the economists determined that a sufficient market for development did not exist, the application was not funded. Similarly, in each state USDA office we visited, officials explained that if they determined that the market for Section 515 development was insufficient in a place where development was proposed, they would not fund the application. Possibly as a result of these market analyses, we did not identify any examples of HUD and USDA providing unnecessarily duplicative housing assistance for the elderly. Several factors limit the potential for unnecessarily duplicative Section 202 and Section 515 housing for the elderly in rural areas: funding constraints, geographic restrictions, and, in some areas, demand for additional rental units. First, the way that Section 202 funding is allocated and the amount of Section 515 funding limit the number of units these programs produce in rural areas. HUD generally allocates only a portion of Section 202 funds to nonmetropolitan areas, which are more likely than metropolitan areas to be considered rural and thus eligible for USDA funds. In fiscal years 2002 through 2004, for example, HUD set aside enough funds for each of its local offices to fund a minimum of five units in nonmetropolitan areas and allocated 15 percent of all funds appropriated for the Section 202 program to these areas. And although the Section 515 program provides funding exclusively in rural areas, funding for this program has fallen sharply since its peak in 1979. During the peak funding years, the program produced more than 20,000 new units annually. Since 2000, fewer than 2,000 new units have been produced annually. Also, not all new Section 515 units are for the elderly, further reducing the potential for overdevelopment of elderly housing in rural areas. Second, geographic restrictions on areas where Section 515 funds can be used limit the extent to which the Section 515 and Section 202 programs can overlap. Section 202 properties can be developed anywhere in the United States. In contrast, not only must Section 515 properties be in rural areas, but also new development can occur only in designated places. State USDA offices develop lists of places with a need for multifamily rental housing, and invite Section 515 applications for these places. According to the USDA officials we interviewed, this list can be refined through input from HUD and USDA field offices. As a result, new Section 515 properties could potentially be developed and overlap with new Section 202 development only in a small number of areas. Finally, officials at the HUD field offices and state USDA offices we visited told us that oversaturating a market with both HUD and USDA units for the elderly was not a concern for several reasons, including a high demand for such units. In some cases, they said, existing elderly properties had waiting lists, and new properties generally rented quickly. According to other officials, given the limited number of elderly units that could be constructed with available funding, overlap and duplication between HUD and USDA housing for the elderly was not a concern. We provided a draft of this report to USDA and HUD for their review and comment. USDA had no comments on the report. HUD provided comments in a letter from the Assistant Secretary for Housing – Federal Housing Commissioner (see Appendix V). The letter stated that the report does not give HUD’s programs full credit for their contributions in assisting the elderly. HUD’s specific comments in this regard, and our responses, are as follows: HUD commented that the report should include more detail on partnership arrangements and overall supportive services provided to the elderly. We believe that the report covers these issues at a level of detail appropriate to our objective. The report includes information on the four federal housing assistance programs that require that supportive services be made available to elderly residents, and four programs that can be used to either link residents to supportive services or provide services directly. In addition, it includes examples of private partnerships and health care programs to convey a broader sense of the supportive services that can be available to the elderly. HUD noted that the draft did not include data on units designated for a mixed population of the elderly and disabled. We initially determined that the 6,004 units that fall into this category were not significant enough to merit inclusion. However, in response to HUD’s comment, we have included this data as a note to Figure 2. HUD stated that the report fails to give Home Equity Conversion Mortgages (HECM) sufficient credit. As detailed in the Objectives, Scope, and Methodology (see Appendix I), this report focuses upon programs that met our definition of housing assistance. While the HECM program did not meet this definition, both the body of the report and Appendix IV acknowledge HECM as a program that assists elderly families. HUD agreed that GAO used the appropriate definition of elderly households, but stated that the report should acknowledge that other households may have members aged 62 or older (that are not the head, co-head, or spouse). While we did not initially provide data on such households because they are not, by definition, elderly households, we did add a note to Figure 3 to acknowledge that housing assistance programs can also benefit this group. As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days from the report date. At that time, we will send copies to the Secretaries of Agriculture and Housing and Urban Development. We will make copies available to others upon request. This report will also be available at no charge on GAO’s Web site at http://www.gao.gov. Please call me at (202) 512-8678 if you or your staff have any questions about this report. Key contributors to this report are listed in Appendix VI. To determine the extent to which federal housing assistance programs provide benefits to elderly households, we first identified the relevant programs through a literature search, review of the Catalog of Federal Domestic Assistance, and consultation with Department of Housing and Urban Development (HUD) and Department of Agriculture (USDA) officials. As used in this report, housing assistance programs are those that: subsidize mortgage interest rates, rent, or housing repair or provide mortgage insurance, loan guarantees, or direct loans for single- family or multifamily housing; and support the construction, rehabilitation, or purchase of multifamily housing or assisted living facilities. We excluded programs that are administered by government-sponsored enterprises such as Fannie Mae, Freddie Mac, and Federal Home Loan Banks; are used exclusively to fund nursing homes or supportive housing for the homeless; are not used exclusively to provide housing assistance; or lack special features that apply to the elderly. We then reviewed laws and regulations to determine which housing assistance programs were (1) targeted to the elderly as the only beneficiaries, (2) targeted to the elderly and disabled only, and (3) were not targeted to the elderly but had special features for them. Appendix III describes the programs we included, and Appendix IV describes some programs we excluded, even though they can benefit elderly households. For each of the programs we included, we interviewed agency officials and reviewed laws, regulations, handbooks, and other documentation to describe the programs and how they benefit the elderly. We used various HUD and USDA databases to analyze the extent to which the programs within our scope either designated units for the elderly or were occupied by elderly households. Specifically: To determine the number of units under HUD’s multifamily programs, we used data from HUD’s Multifamily Portfolio Reporting Database as of April 1, 2004. HUD provided corrected unit counts for 32 properties in August 2004. Unit counts were not available for 42 percent of the Section 207 Manufactured Home Parks properties, 57 percent of Section 232 properties, and 23 percent of Section 542(b) and (c) properties, so we could not produce a reliable count of the total number of units for these programs. For the other HUD multifamily programs, fewer than 2 percent of the properties were missing unit counts (except that 4.5 percent of Section 207/223(f) properties were missing unit counts), so we considered the data sufficient for our purposes. We combined these data with the results of HUD’s Multifamily Housing Inventory Survey, conducted between June 2002 and January 2003, to determine the number of multifamily units designated for the elderly under each program. To determine the number of elderly households living in properties under HUD’s multifamily programs, we used data from HUD’s Tenant Rental Assistance Certification System as of May 25, 2004. However, we were only able to determine the number of households receiving project-based rental assistance. Not all units under HUD’s multifamily program receive project-based rental assistance, and occupancy data on these unassisted units were not available. As a result, the numbers that we report for these programs do not reflect the total number of elderly households in these programs but only the number of elderly households receiving project-based rental assistance. To determine the number of elderly households living in public housing or receiving rental assistance through the Housing Choice Voucher or Section 8 Moderate Rehabilitation programs, we used data from HUD’s Public and Indian Housing Information Center as of June 3, 2004. We excluded outdated data (29 percent of the records) by including in our analysis only households whose records had been updated within the last 15 months. To determine the number of public housing units designated for the elderly, we used a July 14, 2004, Designated Housing Plan Status Report—a spreadsheet that HUD uses to track public housing authorities that have requested or been approved to formally designate units for the elderly. For USDA’s Section 515 program, we used data from USDA’s Multi- Family Integrated System as of April 30, 2004, and July 13, 2004, to determine the number of units, units designated for the elderly, and units occupied by elderly households (including households receiving Section 521 rental assistance). To determine the number of Section 538 units, we used (1) a report from USDA’s Guaranteed Loan System as of June 3, 2004, covering Section 538 properties guaranteed beginning in fiscal year 2000 and (2) a June 16, 2004, report listing data on Section 538 properties guaranteed through fiscal year 1999 but not maintained in USDA’s central data systems. Data on the occupants of Section 538 properties were not available. For USDA’s Section 502 and 504 single-family programs, we used data from USDA’s Dedicated Loan Origination and Servicing System and Guaranteed Loan System as of April 30, 2004, to determine the total number of loans and grants and the number of borrowers or grantees who were elderly at the time they applied for a loan or grant. However, the Guaranteed Loan System only had the primary or secondary borrower age for about 2 percent of the Section 502 Guaranteed Loans, a percentage that was insufficient for our purposes. In order to assess the reliability of the program data described above, we reviewed related documentation and interviewed agency officials and contractors who worked with these databases. In addition, we performed internal checks to determine the extent to which the data fields were populated and the reasonableness of the values contained in the data fields. During our internal checks, for household counts based on the age of household members, we identified 0.01 percent of cases where the age appeared to be erroneous due to unreasonably high values. We concluded that the data we used were reliable for purposes of this report. To provide information on the effectiveness of the programs within our scope in assisting the elderly, we reviewed studies and reports by federal agencies, research institutions, and the HUD and USDA Offices of Inspector General. We reviewed the methodologies used in relevant studies to ensure that the results reported were reasonable. We excluded studies that did not focus on the effectiveness of a program in assisting the elderly, were more than 15 years old, or were not focused upon the goals of the program. We also reviewed the Office of Management and Budget’s Program Assessment Rating Tool assessments for programs within our scope. In addition, we used HUD and USDA fiscal year 2003 Performance and Accountability Reports to determine if program goals or performance measures specific to the elderly had been established for these programs. For those programs with such goals, we provided information on whether the goal was reached. For those programs that did not have these goals, we summarized available information related to the effectiveness of the programs in assisting the elderly. To describe the types of supportive services that federal housing assistance programs provide for the elderly, we first reviewed laws and regulations to determine which of the programs within our scope were required to ensure that supportive services were available. Next, we identified supportive services programs that could be used with various housing assistance programs, whether or not the housing assistance program required such services. For example, while public housing authorities that manage public housing are not required to provide supportive services, housing authorities may implement such services voluntarily. We identified supportive services programs by reviewing literature and descriptions of housing assistance programs and interviewing administrators of the housing assistance programs within our scope, as well as representatives of advocacy organizations and professional associations interested in elderly housing issues and supportive services. We obtained descriptions of these supportive services programs by interviewing officials from HUD, USDA, and the Department of Health and Human Services. We also reviewed program descriptions, notices of funding availability, and other documentation to develop descriptions of these supportive service programs. To determine how HUD and USDA avoid overlap and duplication in programs that offer similar types of housing assistance to the elderly, we first determined which programs were actively producing new units in the same areas. We found that both HUD and USDA were actively producing new multifamily rental units in rural areas through the Section 202 and Section 515 programs, respectively. While Section 202 units can be constructed in metropolitan areas, Section 515 units cannot. As a result, we focused our analysis upon rural areas. We reviewed agreements, policies, and procedures established by HUD and USDA to coordinate the development of subsidized housing in rural areas and interviewed HUD and USDA officials responsible for administering these programs. We also visited HUD and USDA field offices in Greensboro and Raleigh, North Carolina; Columbus, Ohio; and Oklahoma City and Stillwater, Oklahoma to determine whether and how the policies and procedures for coordinating were being followed. We identified these states as having received, when both programs are considered together, the most Section 202 grants in rural areas and the most Section 515 new construction loans made in fiscal years 2002 and 2003. To determine the status of HUD’s efforts to improve administration of the Section 202 program (Appendix II), we interviewed HUD officials and reviewed related documentation to identify steps HUD had taken to implement the recommendations we made in our 2003 report on the Section 202 program. We also obtained a HUD report on the number of delayed Section 202 properties as of November 10, 2004, and compared this report with data we presented in our 2003 report. We conducted our work primarily in Baltimore, Maryland; Greensboro and Raleigh, North Carolina; Columbus, Ohio; Oklahoma City and Stillwater, Oklahoma, and Washington, D.C., between December 2003 and December 2004, in accordance with generally accepted government auditing standards. Our May 2003 report to the Senate Special Committee on Aging noted the significance of the Department of Housing and Urban Development’s (HUD) Section 202 Supportive Housing for the Elderly (Section 202) Program, which subsidizes the development of rental housing and provides rental assistance for elderly households with very low incomes. Among other things, the report found that many Section 202 properties encountered delays before beginning construction, and we made several recommendations to improve the program’s timeliness and oversight. Specifically, we recommended that HUD: Evaluate the effectiveness of the current methods for calculating the capital advances that project sponsors receive and make any necessary changes to these methods, based on this evaluation, so that capital advances adequately cover the development costs of Section 202 projects consistent with HUD’s project design and cost standards; Provide regular training to ensure that all field office staff are knowledgeable of and held accountable for following current processing procedures required to approve projects to begin construction; Update the Section 202 program handbook to reflect current processing Improve the accuracy and completeness of information that field staff enter in the program’s database system and expand the system’s capabilities to track key project processing stages. At the time we completed our work, HUD had made some progress but had not fully implemented these four recommendations. First, in our 2003 report, we found that construction of Section 202 properties was sometimes delayed, in part because HUD awarded inadequate capital advance amounts. As a result, sponsors had to put off construction while they sought additional funding. We recommended that HUD evaluate the effectiveness of its methods for calculating capital advances and make any changes necessary to cover the development costs of Section 202 projects. HUD commissioned a study to examine how the development cost limits used to calculate capital advance amounts compared with indicators of local construction costs and to recommend any needed changes in limits for high-cost areas. A HUD official said that the agency received the results of the study in the fall of 2004, but had not determined whether to make any changes in its methods for calculating capital advances. In addition, HUD had not implemented the two recommendations related to training field staff and updating the Section 202 program handbook. We concluded in our 2003 report that providing adequate formal training for field office staff responsible for reviewing Section 202 properties before approving them to begin construction and issuing an updated program handbook could reduce delays in approving projects for construction by ensuring that staff were accountable for applying and interpreting HUD policies and procedures consistently. HUD agreed with these recommendations. According to senior agency officials, HUD’s goal at the time we completed our work was to provide formal training to field staff in fiscal year 2005 in the technical implementation of the agency’s new rules for using Low-Income Housing Tax Credits or other mixed financing to help fund Section 202 properties. These officials also said that HUD’s goal was to update its program handbooks in fiscal year 2005 to reflect these new rules and other updates to policies and procedures. Although HUD had not fully implemented our recommendations for improving the timeliness of the Section 202 program, the number of delayed Section 202 properties had declined. In our 2003 report, we found that construction on 169 properties that had received Section 202 capital advance awards had been delayed as of the end of fiscal year 2002. By November 2004, all but 18 of these properties had been approved for construction, according to our analysis of a report prepared by HUD. These 18 properties had been awarded capital advances in fiscal years 1998 through 2000. An additional 108 projects funded in fiscal years 2001 and 2002 had been delayed, for a total of 126 delayed projects as of November 10, 2004. Senior HUD officials said that management staff in headquarters were monitoring the progress of these delayed projects by meeting quarterly with field office managers to discuss steps that could be taken to help the projects proceed to construction. Finally, HUD had identified needed enhancements to its program database but had not implemented the improvements, as we suggested. To improve HUD’s oversight of the Section 202 program, we recommended that HUD improve the accuracy and completeness of information entered into its program database and expand the system’s capabilities to track key stages of the development process. Senior HUD officials said that the agency had hired a new contractor in the summer of 2004 to work on the program database and had developed a list of needed enhancements that would address our recommendation. The list included improving the system’s ability to track properties’ progress, correcting data errors, and automating reports. However, the officials did not have a timeline for when they expected the enhancements to be complete. This appendix presents information on 23 federal programs we identified that provide housing assistance to the elderly. The programs, which are administered by the departments of Agriculture (USDA) and Housing and Urban Development (HUD), are organized alphabetically by agency into three categories. The first category includes programs that are targeted to the elderly, the second programs that are targeted to the elderly and disabled, and the third programs that are not targeted to the elderly or disabled but have special features for the elderly. This appendix includes active programs as well as programs that no longer actively produce or subsidize new units yet still fund existing units. In general, both HUD and USDA programs target families at lower income levels. HUD programs target families with incomes that are: extremely low (no more than 30 percent of an area’s median), very low (no more than 50 percent of an area’s median), and low (no more than 80 percent of an area’s median). USDA programs also target families with incomes that are very low and low. In addition, some USDA programs target families with moderate incomes (no more than 115 percent of an area’s median). According to HUD and USDA officials, the terms “family” and “household” are generally used interchangeably. HUD’s definition of family specifically includes elderly families, which are families whose head, spouse, or sole member is a person who is at least 62 years of age. It may include two or more persons who are at least 62 years of age living together or one or more persons who are at least 62 years of age living with one or more live- in aides. In general, USDA’s definition of an elderly household also includes the disabled—that is, the head, spouse, or sole member is at least 62 years old or is a disabled person of any age. However, for the Section 504 grants and Section 538 programs, nonelderly disabled households are not included in the definition of elderly. Project-based rental assistance provides subsidies for tenants in specific properties so that the subsidy is not portable if a tenant moves. Tenant- based rental assistance provides vouchers for eligible tenants to rent single or multifamily units. Through both project-based and tenant-based forms of rental assistance, the renter generally pays 30 percent of adjusted income towards rent. Service coordinators are individuals that can generally be hired to work in public or other federally subsidized multifamily housing to assist residents in obtaining supportive services. For each program, we identify the federal agency responsible for administering the program, the type of assistance provided, and the type of housing. We also provide brief descriptions of (1) the program’s purpose and objectives; (2) how the program is administered; (3) eligibility requirements; (4) special features for the elderly; (5) supportive services provided; and (6) available data on the extent to which the program targets or serves the elderly. We obtained the information for the summaries from the Catalog of Federal Domestic Assistance, program fact sheets, program handbooks, various HUD and USDA databases, and agency officials. This appendix describes some additional housing programs that can benefit the elderly but that either did not meet our definition of housing assistance or were not considered by agency officials to be “key” in assisting the elderly. The appendix does not include all of the housing assistance programs that serve the elderly. A variety of federal agencies are responsible for these programs, including the departments of Agriculture (USDA), Health and Human Services (HHS), Housing and Urban Development (HUD), and Veterans Affairs (VA). The programs are organized in alphabetical order according to their administering agency. These programs provide a direct loan, guaranteed loan, or project grant assistance to construct, enlarge, extend, or otherwise improve community facilities, such as medical clinics, schools, fire and rescue stations, and child care centers for public use in rural areas. These programs finance a range of service centers for the elderly, including nursing homes, boarding care, assisted care facilities, adult day care, and intergenerational care facilities. To be eligible, applicants must be entities such as city, county, and state agencies; private nonprofit corporations; or federally recognized tribal governments. Priority is given to projects that will enhance public safety or provide health care facilities. This program provides loans or project grants to provide decent, safe, and sanitary low-rent housing for domestic farm laborers. Loans are available to farmers, family farm partnerships, family farm corporations, or associations of farmers. Loans and grants are available to states, public or private nonprofit organizations, federally recognized Indian Tribes, and nonprofit corporations of farm workers. Grants are available to eligible applicants only when there is a pressing need and when such facilities cannot be obtained without grant assistance. Loans are usually for 33 years at 1 percent interest, and grants may cover up to 90 percent of development costs. This program provides grants to sponsoring organizations for the repair or rehabilitation of housing for very low- and low-income rural residents. To be eligible, an applicant must be a state or political subdivision, public nonprofit corporation, Indian tribal corporation authorized to receive and administer housing preservation grants, private nonprofit corporation, or a consortium of such entities. Organizations may use less than 20 percent of the Housing Preservation Grant funds for program administration purposes, such as to hire personnel and pay necessary and reasonable administrative expenses. Eighty percent or more of the funds must be used for loans, grants, or other assistance on individual homes, homeowners, rental properties, or cooperatives to pay any part of the cost for repair or rehabilitation of structures. This federally funded program helps low-income households meet their home heating and cooling needs. The program provides funding to states, federally- or state- recognized Indian tribes and tribal organizations, and insular areas, such as Puerto Rico and Guam, in the form of block grants for home energy assistance, energy crisis intervention or assistance, and low-cost residential weatherization and other energy-related home repair. The program targets (1) households with a high energy burden (those households with the lowest incomes and highest home energy costs), and (2) vulnerable households, including those with frail older individuals, individuals with disabilities, and very young children. According to the Congressional Research Service, this program is the largest source of federal financial assistance for state and local governments’ community development and neighborhood revitalization activities. The program’s objective is to develop viable urban communities by providing housing and expanding economic opportunities, principally for individuals with low to moderate incomes. For example, Community Development Block Grant funds have been used to rehabilitate affordable senior housing, construct senior centers, and provide services such as congregate meals and transportation. This program enables elderly homeowners to withdraw some of the equity in their home in the form of monthly payments for life or a fixed term, or in a lump sum, or through a line of credit. This reverse mortgage program allows families to stay in their home while using some of its equity. The total income that an owner can receive through the program is the maximum claim amount, which is calculated with a formula including the age of the owner, the interest rate, and the value of the home. The borrower remains the owner of the home and may sell it and move at any time, keeping the sales proceeds that exceed the mortgage balance. No repayment is required until the borrower moves, sells, or dies. This program is a federal formula block grant to state and local governments designed to create affordable housing for low-income households. The program provides funds to states and localities to build, buy, and rehabilitate affordable housing for rent or homeownership. Also, funds can be used to provide direct rental assistance to low-income people. For rental housing, at least 90 percent of HOME funds must benefit low- and very low-income families at 60 percent of the area median income; the remaining ten percent must benefit families below 80 percent of the area median income. Assistance to homeowners and homebuyers must go to families below 80 percent of the area median. HOME gives grantees the flexibility to use a variety of mechanisms to fund housing projects that meet local priorities, and is routinely combined with other public and private financing for affordable housing such as Housing Choice Vouchers and the Low-Income Housing Tax Credit. This program provides counseling to consumers on seeking, financing, maintaining, renting, or owning a home. The Housing Counseling Assistance Program enables individuals wanting to rent or own housing— whether through a HUD program, a Veterans Affairs program, other Federal programs, a State or local program, or the regular private market— to get the counseling needed to make their rent or mortgage payments and to be a responsible tenant or owner in other ways. The counseling is provided by HUD-approved housing counseling agencies. There are three strategic goals for the program: (1) to improve the quality of renter and homeowner education, (2) to develop a reliable stream of funding and resources for counseling agencies, and (3) to enhance coordination among local housing providers. This program promotes homeownership among families with low to moderate incomes by providing mortgage insurance for the purchase or refinancing of a principal residence. This program provides mortgage insurance to protect lenders, such as mortgage companies, banks and savings and loan associations, against the risk of default on loans to qualified buyers. Insured loans may be used to purchase new or existing one- to four- family homes, as well as to refinance debt. The insurance allows homebuyers to finance up to 97 percent of the home’s cost through their mortgage. This program enables homebuyers and homeowners to finance the purchase (or refinancing) of a house that is at least a year old and the cost of its rehabilitation through a single mortgage. The cost of rehabilitation must be at least $5000, but the total value of the property must still fall within the mortgage limit for the area. This program provides federal insurance for mortgage loans on multifamily rental projects located in urban renewal areas and areas where local governments have undertaken designated revitalization activities. The purpose of Section 220 is to encourage the development of quality rental housing in urban areas targeted for overall revitalization, and to insure lenders against loss on mortgage defaults. The maximum amount of the mortgage loan may not exceed 90 percent of the estimated replacement cost for new construction or 90 percent of the estimated cost of the repair and the estimated value of the property before the repair for substantial rehabilitation. This program insures loans for terms of up to 30 years for the purchase of a unit in a condominium building. The building must contain at least four dwelling units and can be detached or semidetached, a rowhouse or walk-up, or an elevator structure. The insurance covers loans made by lending institutions such as mortgage companies, banks, and savings and loan associations. This program is designed to allow very low-income adults with disabilities to live independently in the community by funding the development of rental housing with appropriate supportive services. HUD provides interest-free capital advances to nonprofit sponsors to help finance the development of supportive housing. The advance does not have to be repaid as long as the housing remains available for very low-income persons with disabilities for at least 40 years. The program also provides project rental assistance, which covers the difference between the operating costs of the development as approved by HUD and the tenants’ contribution toward rent (usually 30 percent of adjusted income). Each project must have a supportive services plan. The services offered may vary with the target population but could include case management, training in independent living skills, and assistance in obtaining employment. This program facilitates the financing of improvements to homes and other existing structures and the building of new nonresidential structures. The maximum loan amount is $25,000 for improving a single family home or for improving or building a nonresidential structure. For improving a multifamily structure, the maximum loan amount is $12,000 per family unit, not to exceed a total of $60,000 for the structure. HUD insures private lenders against losses of up to 90 percent of any single loan. This program provides funding to community agencies providing services to homeless veterans. The program’s purpose is to help homeless veterans achieve residential stability, increase their skill levels and income, and obtain greater self-determination by promoting the development of housing with supportive services. Only programs with supportive housing or service centers are eligible for the two levels of funding: grants and per diem. This program helps veterans, certain service personnel, and certain unmarried surviving spouses of veterans obtain credit for the purchase, construction, or improvement of homes on more liberal terms than are generally available to nonveterans. Lenders, such as mortgage companies, savings and loan associations, or banks, make the loans and VA provides the guarantee. The amount guaranteed varies with the amount of the loan and previous use of the program. With the current maximum guarantee, a veteran who has not previously used the benefit may be able to obtain a loan up to $240,000, depending on the borrower’s income level and the appraised value of the property. This program subsidizes the purchase, construction, rehabilitation, or refinancing of (1) owner-occupied housing for very low- to moderate- income households, and (2) rental housing in which very-low-income households can afford and will occupy at least 20 percent of the units. The Federal Home Loan Banks offer both grants and loans with below-cost interest rates. This program provides a favorably priced source of wholesale funds for any member involved in lending for community and economic development. The funds can be used for development of commercial projects, infrastructure improvements, or business that creates jobs. The funds are available to finance home purchases by families whose income does not exceed moderate-income levels, for purchase or rehabilitation of housing for occupancy by families whose income does not exceed moderate- income levels, or for commercial and economic development activities that benefit low- and moderate-income families or neighborhoods. This tax credit program, which IRS and the states administer jointly, is the principal federal program designed to support the development and rehabilitation of housing for low-income households. Under this program, states are authorized to allocate federal tax credits as an incentive to the private sector to develop low-income rental housing. In their guidelines for the distribution of tax credits, some states have established preferences for properties intended for the elderly, but such a preference is not a federal requirement. Annually, IRS allocates tax credits to each state. For 2005, the credit is equal to $1.85 per state resident. Investors that provided financing may take the tax credits annually for 10 years to offset federal taxes. At a minimum, the owner must agree to make (1) 20 percent of the property units affordable to households with incomes at or below 50 percent of the area median income or (2) 40 percent of the units affordable to households with incomes at or below 60 percent of the area median. In addition to those named above, Emily Chalmers, Natasha Ewing, Alison Martin, John McGrail, Marc Molino, Lisa Moore, John Mingus, and Julianne Stephens made key contributions to this report.

According to the 2003 American Housing Survey sponsored by the U.S. Department of Housing and Urban Development (HUD), nearly one-third of elderly households--those whose head was age 62 or older--were experiencing housing affordability problems. Further, a congressional commission reported in 2002 that investment in affordable housing is decreasing, although the elderly population is expected to increase. A number of federal housing programs provide assistance, including rent subsidies, mortgage insurance, and loans and grants for the purchase or repair of homes, to low-income renters and homeowners. These programs are administered primarily by HUD or the U.S. Department of Agriculture (USDA). GAO was asked to determine the extent to which federal housing programs provide benefits to elderly households, summarize information on the programs' effectiveness in assisting the elderly and supportive services, and determine how HUD and USDA avoid overlap and duplication in their programs. A total of 23 federal housing programs target or have special features for the elderly. Specifically, one HUD and one USDA program target the elderly exclusively, while three HUD programs target the elderly and disabled. The remaining 18 programs serve a variety of household types but have special features for elderly households, such as income adjustments that reduce their rents. The 13 programs for which data were available provide about 943,000 housing units designated for occupancy by the elderly. However, many programs also serve the elderly in undesignated units. Available occupancy data show that the elderly occupied at least 1.3 million units under rental assistance, public, and multifamily housing programs as of spring 2004. Information on the effectiveness of housing programs that assist the elderly is limited. HUD has an overall goal related to elderly housing, but not all individual programs that assist the elderly are explicitly linked to this goal. USDA does not have specific goals related to elderly housing. Most of the 23 housing assistance programs we reviewed are not designed to provide supportive services for the elderly. Four programs require the owners of program properties to ensure that services such as meals or transportation are available to their residents. In addition, HUD administers four programs--for example, the Service Coordinator Program--that can be used in conjunction with various housing programs to help the elderly obtain supportive services. Supportive services are also available to elderly residents of subsidized housing through partnerships between individual properties and local organizations. To avoid overlap and duplication in the development of rural housing for the elderly, HUD and USDA have established policies and procedures that require field offices from both agencies to notify their counterparts of applications to build new housing and consider each other's input on local market conditions. GAO visits to selected HUD field offices and state USDA offices revealed that staff were not consistently following these policies and procedures but were analyzing markets to ensure the need for proposed housing. Overall, however, funding and geographic constraints limit the potential for overlap and duplication in the construction of rural housing for the elderly.

The DOD purchase card program is part of the Governmentwide Commercial Purchase Card Program, which was established to streamline federal agency acquisition processes by providing a low-cost, efficient vehicle for obtaining goods and services directly from vendors. The purchase card can be used for both micropurchases and payment of other purchases. Although most cardholders have single purchase transaction limits of $2,500, some have limits of $25,000 or higher. The Federal Acquisition Regulation, Part 13, “Simplified Acquisition Procedures,” establishes criteria for using purchase cards to place orders and make payments. DOD has issued supplemental guidance to the Federal Acquisition Regulation that contain sections on simplified acquisition procedures. General Services Administration (GSA) reports show that DOD used purchase cards for nearly 11 million transactions, valued at almost $6.8 billion and representing nearly 45 percent of the federal government’s fiscal year 2002 purchase card activity. According to unaudited GSA data, the Army, Navy, and Air Force made about $2.7 billion, $1.9 billion, and $1.6 billion, respectively, in purchase card acquisitions during fiscal year 2002. Other DOD agencies, such as the Defense Logistics Agency and the Defense Finance and Account Service, made the remaining $564 million in purchase card acquisitions. The overall management of DOD’s purchase card program has been delegated to the DOD Purchase Card Joint Program Management Office, which is in the Office of the Assistant Secretary of the Army for Acquisition Logistics and Technology. At each service installation, personnel in three positions—program coordinator, cardholder, and approving official—are collectively responsible for providing reasonable assurance that purchase card transactions are appropriate and meet a valid government need. The installation program coordinator is responsible for the day-to-day management, administration, and oversight of the program, including developing local operating procedures, issuing and canceling cards, and providing training to cardholders and approving officials. Cardholders— members and civilian personnel—use purchase cards to order goods and services for their units and their customers, to be picked up or delivered to themselves or to an end user. The cardholders are responsible for recording the transactions in their purchase log, obtaining documented independent confirmation that the items have been received and accepted by the government, and notifying the property book-officer of accountable items received so that these items can be recorded in the accountable property records. Approving officials, who typically are responsible for more than one cardholder, are to review cardholders’ transactions and the cardholders’ reconciled statements and certify the official consolidated bill for payment. Approving officials are to ensure that (1) all purchases made by the cardholders within his or her cognizance are appropriate and that the charges are accurate and (2) the monthly summary statement is certified for payment on time by the Defense Finance and Accounting Service (DFAS). DFAS relies on the approving official’s certification of the monthly bill as support to make the payment. Between July 2001 and December 2002, we testified four times and issued four reports highlighting a weak control environment and breakdowns in specific internal controls over the purchase card program at the Army, Navy, and Air Force. Based on statistical sampling and selected reviews of at- risk transactions we identified through data mining, we reported that these weaknesses left the purchase card program at the three services vulnerable to fraudulent, improper, and abusive purchases. The testimonies and reports we issued pointed to common weaknesses. We identified (1) a proliferation of cardholders, (2) lack of documented evidence of training of cardholders and approving officials, (3) inadequate program monitoring, and (4) lack of disciplinary actions against cardholders who abused the purchase cards. We made recommendations to each of the services for improving the purchase card program. We reported that the proliferation of cardholders resulted in an unmanageable approving official span of control and excessive credit limits compared to historical spending. This problem originated from the fact that the services did not have specific policies governing the number of cards to be issued or criteria for identifying employees eligible for the privilege of cardholder status. Consequently, as of September 2002, the Air Force reported that it had about 77,000 purchase card accounts— translating to about 1 purchase card for every 7 employees. By contrast, the Navy, which in 2000 had 1 cardholder for every 3 employees in some of its units, had taken positive steps to reduce the number of its purchase cardholders to only about 1 cardholder for every 31 employees by September 2002. The proliferation of cardholders also resulted in a span of control problem for some approving officials. For example, at the end of fiscal year 2002, some officials at two Air Force installations had multiple job responsibilities in addition to being approving officials for more than 20 cardholders, making it difficult for them to systemically scrutinize each purchase card statement they had to certify for payment. We also found that the credit limits on the purchase cards exceeded procurement needs. We saw little evidence that limits were set based on an analysis of individual cardholders’ needs or past spending patterns. For example, at the Marine Corps, the credit limit as of March 2002 exceeded average fiscal year 2001 monthly expenditures by a ratio of 34 to 1, while at an Air Force location, the credit limit exceeded fiscal year 2001 monthly purchases by a ratio of 20 to 1. At the Army, we saw infrequently used cards that, nevertheless, had spending limits set at the maximum. In some cases, we were told that the monthly limits were based on anticipated peak spending to avoid possible limit changes. Limits that are higher than justified by the cardholder’s authorized and expected use unnecessarily increase the government’s exposure to fraudulent, improper, and abusive purchases. Limiting credit available to cardholders is a key factor in managing the purchase card program and in minimizing the government’s financial exposure. We reported that cardholders, approving officials, and/or agency program coordinators did not receive adequate training necessary to carry out their responsibilities. Specifically, we found that 51 percent of the fiscal year 2001 transactions at one Air Force location, 56 percent of the transactions at the Marine Corps, and as high as 87 percent of the transactions at one Navy command, were made by cardholders or approved for payment by approving officials for whom there was no documented evidence of either initial training or refresher training at the time the transactions were made. At the Army, cardholders received initial training, but were seldom provided refresher training as required by DOD guidance. Further, we noted that, even though the functions performed by the agency program coordinators, approving officials, and cardholders were substantially different, the training curriculum for the three positions was identical. The services did not have specific guidance or training concerning the role and responsibilities of agency program coordinators or approving officials. We reported that all of the military services needed to improve the quality of their monitoring and oversight of the purchase card program. At the time of our audits, the purchase card program offices of the military services did not systematically monitor the purchase card program. We also reported that when a military services’ purchase card program office or audit agency did uncover control weaknesses or improper and abusive or questionable activity, the results of those efforts were not always used to improve program management. We also noted in our reports and testimonies that individuals who misused the purchase card were not always subject to strong disciplinary action or consequences. For example, we found that cardholders who purchased and officials who authorized items with excessive cost or without documented government need, including designer brief cases, folios, and palm pilot carrying cases from Coach, Dooney and Bourke, and Louis Vuitton; personal clothing including golf shirts and ski clothing; food including beer, wine, and cigars; and Bose stereo headset and clock radios, were not disciplined for their actions. We reported that without disciplinary actions, improper, abusive, and questionable purchases like those mentioned above will likely continue. In response to the concerns we expressed about DOD’s management of the purchase card program, the Congress included Section 1007 in the National Defense Authorization Act for fiscal year 2003 (Public Law 107-314) and Section 8149 in the fiscal year 2003 DOD Appropriations Act (Public Law 107-248) to require DOD to take specific actions to improve the management of the purchase card program, and in particular the weaknesses we identified. As shown in table 2, these laws limit the number of purchase cards and require DOD to train purchase card officials, monitor purchase card activity, discipline cardholders who misuse the purchase card, and assess the credit worthiness of cardholders. During fiscal year 2003, DOD and the military services took actions to implement all of the requirements mandated by the fiscal year 2003 National Defense Authorization and DOD Appropriations acts. In several cases, although DOD and the services have issued policies and guidelines that implement the legislative mandates, sufficient time has not passed for the objective of the legislative mandate to be achieved. DOD has substantially reduced the number of purchase cards issued. According to GSA records, DOD had reduced the total number of purchase cards from about 239,000 in March 2001 to about 145,000 in March 2003. DOD also informed us that it manages the gross number of purchase and travel cards in accordance with the DOD Appropriations Act, 2003. To that end, DOD had reduced the total number of purchase and travel cards to about 1.23 million, about .27 million less than the 1.5 million statutory limit. DOD also issued policy guidance on April 25, 2002, to field activities to (1) perform periodic reviews of all purchase card accounts to reestablish a continuing bona fide need for each card account, (2) cancel accounts that were no longer needed, and (3) devise additional controls over infrequently used accounts to protect the government from potential cardholder or outside fraudulent use. The policy cited as an acceptable control for infrequently used cards the reduction of the spending limit to $1 until such time as the card is needed. To implement the requirement to train each purchase cardholder and each official with responsibility for overseeing the use of purchase cards, DOD’s Defense Acquisition University has made available several on-line, self- paced purchase card courses on its Web site. The on-line curriculum included a GSA module targeted to cardholders on how to use the card responsibly, a DOD course for cardholders and billing officials on the mandatory requirements and other guidelines of the purchase card program, and a GSA module aimed at providing advanced training to agency program coordinators who have completed basic training on the purchase card program. Further, on September 27, 2002, DOD issued a memorandum requiring all cardholders, approving officials, and certifying officials to complete the training module. To address the requirement that the Inspectors General of DOD and the military services periodically audit the program to identify potentially fraudulent, improper, and abusive uses of the purchase cards, as well as any patterns of improper cardholder transactions, DOD indicated that its Office of Inspector General and the Navy have prototyped and are now expanding a data-mining capability to screen for and identify high-risk card transactions (such as potentially fraudulent, improper, and abusive use of purchase cards including prohibited purchases) for subsequent investigation. According to DOD, this capability will eventually be implemented across the department. In addition, on June 27, 2003, the DOD Inspector General issued a report summarizing the results of in- depth review of purchase card transactions made by 1,357 purchase cardholders. The report identified 182 cardholders who potentially used their purchase cards inappropriately or fraudulently. With respect to the National Defense Authorization Act’s requirement to use strategic sourcing (i.e., that the Inspectors General identify categories of purchases that should be made by means other than purchase cards in order to better aggregate purchases and obtain lower prices), DOD issued a memorandum on June 5, 2003, reiterating a prior decision requiring all DOD components to review fiscal year 2002 purchase card transaction files and stratify the volume of purchases by vendors. According to the memorandum, these data will be used to determine if any componentwide contracts should be established to optimize purchasing power. DOD also indicated that each of the military departments have initiated a strategic sourcing plan, contract, or Blanket Purchase Agreement (BPA) to take advantage of purchase card demand (sales volume) data. As an example, DOD said that the Army had awarded a BPA for office supplies in 2002 to address long-standing concerns over cardholder compliance with mandatory sourcing requirements. Likewise, the Air Force entered into a BPA with a large provider of office supplies and anticipates others. The Navy is expected to make similar BPA arrangements when its sales volume analysis is completed. According to DOD, the strategic sourcing initiative is still in the infancy stage, but the department is committed to expanding opportunities to leverage its purchase card purchasing power. The issue of strategic sourcing of purchase card transactions is also the subject of an audit that we initiated at the request of the Chairman and Ranking Minority Member of the House Committee on Government Reform, Subcommittee on Government Efficiency, Financial Management and Intergovernmental Relations. With respect to establishing regulations that provide for appropriate adverse personnel actions or other punishment for misuse, abuse, or fraud with respect to purchase cards, DOD has issued disciplinary guidelines, separately, for civilian and military employees. In both updated guidelines, DOD continues to emphasize its policy that improper, fraudulent, abusive, or negligent use of a government charge card is prohibited. This includes any use of government charge cards at establishments or for purposes that are inconsistent with the official DOD business or with applicable regulations. The intent of the guide is to ensure that management emphasis is given to the important issue of personal accountability. The civilian guide has a sample range of potential charge card offenses and remedies or penalties for such offenses as shown in table 3. According to the disciplinary guidelines, there is no single response for all cases. Instead, a progression of increasingly severe disciplinary measures is often appropriate in the case of minor instances of misuse, but more serious cases may warrant the most severe sanctions in the first instance. The disciplinary guide for military employees indicates that actions available when military personnel misuse a purchase or travel card include counseling, admonishment, reprimand, nonjudicial punishment (Article 15, Uniform Code of Military Justice – UCMJ), court-martial, and administrative separation. In addition to corrective disciplinary actions, military personnel who misuse their government charge cards may have their access to classified information modified or revoked if warranted in the interests of national security. These guidelines emphasized that while the merits of each case may be different, timeliness, proportionality, and the exercise of good judgment and common sense are always important. Finally, with regard to the requirement that DOD evaluate the credit worthiness of cardholders, DOD told us that a senior focus group consisting of acquisition, financial management, and general counsel executives had concluded that there are conflicts between this legislation and the Fair Credit Reporting Act. The department is pursuing an alternative solution that would rely on a self-certification process by prospective cardholders. The legality and practicality of this alternative are being staffed and coordinated. This process, however, is in stark contrast to the standard industry practice of conducting credit checks on credit card applicants. According to information provided by representatives of the Army, Navy, and Air Force, the three services have either completed or initiated actions to implement nearly all of the 109 recommendations we made to improve the management of the purchase card program. As shown in table 4, we made 22 recommendations to the Army to improve its purchase card program and the Army provided us with information that it had implemented 18 of those recommendations and initiated actions to implement the remaining 4 recommendations. In addition, the Navy told us that it had implemented 38 of our 48 recommendations, and initiated actions to implement the 10 other recommendations. Similarly, the Air Force reported that it had implemented 24 of our 39 recommendations and initiated actions to implement the 15 other recommendations. The recommendations that the Army, Navy, and Air Force told us they have implemented related to issuing new purchase card policies and procedures, retraining cardholders and approving officials, and reducing the number of purchase card accounts to improve management of the purchase card program. The recommendations they have not fully implemented generally were those dealing with leveraging purchase card buying power, establishing servicewide databases for data mining, investigating suspected and known fraud cases, and linking the cardholders’, approving officials, and agency program coordinators’ performance appraisals to performance standards. The Air Force and Navy reported to us that they planned to complete implementation of most of the remaining GAO recommendations by June 2004. The Air Force planned to complete implementation of all of the partially completed recommendations by January 4, 2004. The Navy indicated that some of the recommendations would be implemented by June 2004. The Army and the Navy did not provide a date for when some of the partially completed recommendations would be implemented, but indicated that there was an ongoing effort to identify opportunities to leverage purchasing power, develop data mining, analysis, and investigation functions, and develop databases of known fraud cases to improve internal controls. Appendixes II, III, and IV summarize GAO recommendations and the military services’ representations of actions taken. We have not verified whether the military services are effectively implementing the policies and procedures that we recommended they establish and/or modify. In our purchase card reports and testimonies, we identified 51 cases where cardholders had used the government purchase card to make fraudulent or potentially fraudulent purchases and 120 cases where cardholders had made improper and abusive or questionable purchases. In general, when a court of law determined that a cardholder fraudulently used the purchase card, all the military services took strong disciplinary actions (i.e., assessed fines, and in the case of uniformed personnel, sentenced the individual to jail/confinement). We also found that the military services either took strong disciplinary actions or were actively investigating the cases we reported as potentially fraudulent. For example, our two Navy reports identified 26 fraudulent and potentially fraudulent transactions totaling more than $1,342,000. The Navy reported that in response, it fired six cardholders, reduced the grade of others, confined several uniformed serviceman to from 14 months to 60 months, and required repayment to the government of over $460,000. Other actions taken on fraudulent or potentially fraudulent transactions included suspending or revoking purchase card privileges, requiring repayment to the government for the cost of the items obtained, giving the items obtained back to the government, and written reprimands. In eight instances where no action was taken against cardholders we categorized as having used the purchase card in a fraudulent or potentially fraudulent manner, the military services and the credit card company determined that the fraud was committed by a third party, and the government had no responsibility for the charge. The military services were still investigating 15 cases for fraud. However, as shown in table 5, the military services often did not discipline the 120 individuals that we identified as having made improper, abusive, or questionable transactions. Further, the discipline, if it was imposed at all, was usually retraining. The responses the military services provided to our inquiries concerning disciplinary actions indicated that in three instances the cardholder had to repay the government for the cost of the improper, abusive, or questionable item(s) we identified. Of the remaining cardholders, 7 had their purchase card privileges revoked, 5 received verbal or written reprimands, and 6 had to return items that we deemed improper, abusive, or questionable to the government. Further, in their response to our inquiries concerning the disciplinary actions taken against cardholders who we identified as making improper, abusive, or questionable purchases, the military services stated that they did not take any action in over half of the transactions we identified. We believe that these items were imprudent use of tax dollars, but the military services claimed that policies existing at the time the purchases were made permitted the acquisitions. Therefore, the military services did not think that they had the authority to discipline the cardholders or approving officials. Rather, the military services told us they modified their policies and procedures to prohibit similar acquisitions in the future. The Navy, for example, told us that it had issued numerous e-mails and updated its policies to indicate that some products purchased in the past were now prohibited, and that it planned to better monitor purchases so that none of these purchases would occur in the future. While clarifying purchase card policies and procedures is appropriate, failure to take any disciplinary actions against individuals who purchased or authorized the purchase of items that clearly exceed the needs of the government (designer briefcases) or were excessive in cost ($350 clock radios) does not serve as a deterrent to future abuse or the waste of tax dollars. DOD and the military services have taken strong actions to improve the controls over the purchase card program. DOD has initiated actions to implement all of the requirements that were mandated in the fiscal year 2003 National Defense Authorization and DOD Appropriations acts. In addition, DOD and the military services have taken actions on nearly all of 109 recommendations that GAO made in its four reports on the purchase card program, and DOD has plans to have most of the recommendations implemented by June 2004. While the military services have generally taken strong disciplinary actions against cardholders who we identified as having made fraudulent or potentially fraudulent purchases, the military services generally have done little or nothing to discipline cardholders who have made improper, abusive, or questionable purchases. To help provide reasonable assurance that DOD holds cardholders and approving officials accountable for improper and abusive purchase card acquisitions, we recommend that the Secretary of Defense direct the service secretaries and the heads of DOD agencies to establish procedures to: monitor the results of purchase card reviews conducted by the military services and the DOD agencies, track whether the major commands and units are consistently applying DOD’s disciplinary guidelines to those who made and/or authorized improper or abusive acquisitions, and notify the appropriate officials at the major commands or units if DOD’s disciplinary guidelines are not being consistently applied. In comments on a draft of this report, reprinted in appendix V, DOD stated that while more needs to be done, it appreciated our recognition of the department’s efforts to address previously cited managerial and internal control deficiencies. In its response, DOD requested that we add some perspective to table 5 that shows the extent to which DOD had not taken disciplinary actions on purchases that we had characterized as improper, abusive, or questionable because the military services belief that they had documented policies that specifically authorized the purchases we questioned. To provide this additional perspective, we modified table 5 to separately identify the three transactions that we considered abusive or questionable that the military services believe were specifically authorized by existing Air Force regulations. While we believe that this differentiation is useful, we continue to question whether the purchase card was the appropriate vehicle to make the purchases we identified as abusive or questionable in our prior report. We also modified the report’s title to be focused on future program improvements. DOD did not comment on our recommendations to monitor implementation of the disciplinary guidance. We will send copies to interested congressional committees; the Secretary of Defense; the Under Secretary of Defense, Comptroller; the Under Secretary of Defense for Acquisition Technology and Logistics; the Secretary of the Army; the Secretary of the Navy; the Secretary of the Air Force; the Director of the Defense Finance and Accounting Service; and the Director of Management and Budget. We will make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. Please contact Gregory D. Kutz at (202) 512-9505 or kutzg@gao.gov, or John V. Kelly at (202) 512-6926 or kellyj@gao.gov if you or your staffs have any questions concerning this report. Major contributors to this report are acknowledged in appendix VI. This study responded to the legislative mandate in the conference report to the Bob Stump National Defense Authorization Act of 2003, that directs the Comptroller General to review the actions taken by the Department of Defense (DOD) to comply with the requirements of Section 1007 of the act and submit a report on those actions to the congressional defense committees no later than December 2, 2003. At the request of the committee, this report also summarizes the actions taken by the Army, Navy, and Air Force to respond to the legislative mandates in Section 8149 of fiscal year 2003 DOD Appropriations Act, and actions taken by the military services to implement the recommendations we made in four reports issued during fiscal years 2002 and 2003 aimed at improving the military services’ management of the purchase card program. Finally, the report also discusses the actions taken by the military services against individuals we identified in our testimonies and reports as having made potentially fraudulent, improper, abusive, or questionable purchase card transactions. To meet the objectives of this assignment, we requested that DOD and the military services provide us the (1) status of DOD and the military services’ efforts in implementing certain provisions of the National Defense Authorization Act for fiscal year 2003 and the fiscal year 2003 DOD Appropriations Act, (2) status of actions taken to implement the recommendations included in our four reports, and (3) administrative or disciplinary actions taken against individuals we identified as having made potentially fraudulent, improper, and abusive or questionable transactions. While we asked DOD and the military services to provide evidence documenting actions taken to improve the purchase card program and prevent individuals and companies from further obtaining fraudulent, improper, and abusive or questionable items with a DOD purchase card, we did not make any field visits to independently validate whether DOD had effectively implemented the reported changes. We briefed DOD managers, including DOD officials in the Office of the Undersecretary of Defense (Comptroller), and the Inspector General; Army officials in the Office of Deputy Chief of Staff for Logistics; Navy officials in the Office of the Assistant Secretary of the Navy for Research Development and Acquisition; and Air Force officials in the Office of the Deputy Chief of Staff for Installation and Logistics. We conducted our review from June through September 2003 in accordance with U.S. generally accepted government auditing standards. We requested comments on a draft of this report from the Secretary of Defense or his designee. We received written comments from the Director of DOD’s Purchase Card Joint Program Management Office, which are reprinted in appendix V. We have incorporated suggested changes as appropriate. Purchase Cards: Control Weaknesses Leave Army Vulnerable to Fraud, Waste, and Abuse (GAO-02-732, June 27, 2002) Overall program management and environment Address key control environment issues in Army-wide standard operating procedures. At a minimum, the following key issues should be included in the procedure: assessment of ongoing need for cards. Identified in Army Standard Operating Procedure (SOP) Section 15. Also reinforced by Army in memorandum issued May 22, 2002, requesting that heads of contracting activities ensure cards are issued only to individuals with bonafide needs and that the limits reflect actual needs and available funding. Identified in Army SOP Section 15. leaves the Army, is reassigned, or no longer has a valid need for the card. 3. Span of control of the approving officials. Identified in Army SOP Section 5. 4. Appropriate cardholder spending limits. Identified in Army SOP Sections 15 and 16. Also reinforced by Army in memorandum issued May 22, 2002, requesting heads of contracting activities to ensure cards are issued only to individuals with bonafide needs and that the limits reflect actual needs and available funding. Issued Memorandum endorsed by General John Keane, Vice Chief of Staff articulating the focus on the number of purchase card organizations for each card account (300), and the skill sets typically require a GS-11 and also required in-depth skills in financial and contracting policy and procedures with strong verbal communications skills. The DOD Concept of Operations (CONOPS) report has been updated to further identify skill sets for the billing official and cardholders. 6. Assess the adequacy of human capital resources devoted to the purchase card program, especially for oversight activities, at each management level, and provide needed resources. Memorandum signed by Vice Chief of Staff, July 8, 2002, directed Army commanders to provide adequate resources for purchase card program coordinators to ensure a system of strong internal controls. This was also reemphasized in the Army SOP. Identified in Army SOP Section 9. oversight system for program coordinators that includes standard activities and analytical tools to be used in evaluating program results. GAO observation on the status of recommendation 8. Develop performance measures and goals to assess the adequacy of internal control activities and the oversight program. Identified in Army SOP Section 9 and appendixes J and I. Also reinforced in Secretary of the Army memorandum dated January 28, 2003. 9. Require reviews of existing cardholders and their monthly spending limits to help ensure that only those individuals with valid continuing purchasing requirements possess cards and that the monthly spending limits are appropriate for the expected purchasing activity. These reviews should result in canceling unneeded cards Army-wide and especially at Fort Hood where we found a significant problem. September 2002 the Army had 101,398 cardholders. Army canceled 35,778 since September. Additionally, agency program coordinators are required to review this as part of their surveillance reviews as identified in the Army SOP. Identified in SOP Sections 12 and 18. Identified in SOP 13. 12. Independent review by an approving official of the cardholder’s monthly statements and supporting documentation. Identified in SOP Section 11. Identified in SOP Section 11. on the monthly statement with invoices and other supporting documentation and forwarding the reconciled statement to the designated disbursing office for payment as required by governmentwide and DOD regulations. Identified in SOP Section 12. invoices that support their purchases and provide the basis for reconciling cardholder statements. Identified in SOP Appendix E. checklists for approving officials to use in the monthly review of cardholders’ transactions. These procedures and checklists should specify the type and extent of review that is expected and the required review documentation. Identified in SOP Section 19. purchase card transaction files and require that compliance with record retention policy be assessed during the program coordinator’s annual review of each approving official. Identified SOP Section 8. implementation of coordination and reporting procedures to help ensure that accountable property bought with the purchase card is brought under appropriate control. Identified in SOP Section 18 and Appendix D. justification and approval of those planned purchases that are “questionable” that fall outside the normal procurements of the cardholder in terms of either dollar amount or type of purchase. 19. Analyze the procurements of continuing requirements through micropurchases and require the use of appropriate contracting processes to help ensure that such purchases are acquired at the best prices. Ongoing effort. If the Army identifies leveraging opportunities, they will be implemented through some form of contracting process. The Army issued 12 mandatory Blanket Purchase Agreements for office products and supplies in September 2002. The army has also teamed with the Army Comptroller’s office in awarding a support contract to assess the Army’s purchasing data to determine if leveraging opportunities exist. 20. Develop an Army-wide database on known fraud cases that can be used to identify potential deficiencies in existing internal control and to develop and implement additional control activities, if warranted or justified. The Army participates in the DOD charge card special focus group to look at this issue DOD-wide. However, the Army has teamed with the Army Criminal Investigative Command and the Public Affairs office to identify, report, and publish newsworthy fraud cases and to inform Army soldiers and Department of the Army civilian personnel, their supervisors, and the public of corrective actions taken to resolve misuse of the Army Purchase Card. GAO observation on the status of recommendation 21. Develop and implement an Army-wide data mining, analysis, and investigation function to supplement other oversight activities. This function should include providing oversight results and alerts to major commands and installations when warranted. Ongoing. Army will continue to work with the DOD Charge Card Focus Group. The Army participates in the DOD Charge Card Focus Group to look at this issue DOD-wide. DOD is currently working with the DOD IG to test a Navy prototype data- mining system. A July 2003 Draft Army Audit Report Audit of Army Government Purchase Card (using DOD IG data-mining techniques) stated that about 6 percent (281) of the 4,537 reviewed Army purchase card transactions were improper. Over half of those instances were instances of compromised purchase cards used by third parties for charges and in billing adjustments that returned about 98 percent of the improperly charged Army funds. This left about 3 percent of purchases that were improper, which is less than commercial industry standards of 4.2 percent identified in the 2003 Purchase Card Benchmark survey results, a VISA survey conducted by Palmer and Mahendra Gupta dated July 21, 2002. 22. Incorporate GAO recommendations, to the extent applicable, into the Charge Card Task Force’s future recommendations to improve purchase card policies and procedures throughout DOD. The Army participates in the DOD Charge Card Focus Group to look at these issues DOD-wide. Purchase Cards: Control Weaknesses Leave Two Navy Units Vulnerable to Fraud and Abuse (GAO-02-32, Nov. 30, 2001) 1. Establish specific policies and strategies governing the number of purchase cards to be issued with a focus on minimizing the number of cardholders. The revised eBusiness Operations Office Instruction (EBUSOPSOFFINST) 4200.1 incorporates the Department of Defense “Span of Control Goals” which resulted in approving officials having a reasonable number of cardholders. The Navy executes the DOD Purchase Card Program in a decentralized manner consistent with DOD policy. This allows individual commands to issue purchase cards to employees as mission requirements warrant. No less than semiannually, Agency Program Coordinators (APC) review the continuing need for each account under their purview. 2. Develop criteria for identifying employees eligible for the privilege of cardholder status. As part of the effort to develop these criteria, assess the feasibility and cost-benefit of performing credit checks on employees prior to assigning them cardholder responsibilities to ensure that employees authorized to use government purchase cards have demonstrated credit worthiness and financial integrity. The criterion on eligibility for cardholder’s duties has been developed and is incorporated in Department of Navy (DON) PC desk guides. The issue of credit checks was deferred to DOD. DOD is seeking additional legislative action required to implement credit checks. 3. Develop policies and strategies on credit limits provided to cardholders with a focus on minimizing specific cardholder spending authority and minimizing the federal government’s financial exposure. EBUSOPSOFFINST 4200.1, chapter 2, paragraph 3, defines DON policy. In addition, the EBUSOPSOFF monitors credit limits quarterly and takes action when it appears that existing credit limits exceed mission requirements. Credit limits are now a critical element in the revised semiannual review procedures. EBUSOPSOFFINST 4200.1, chapter 3, paragraph 9, addresses mandatory requirements for training. In addition, chapter 4, paragraph 1b.2, mandates that program compliance with applicable training be reported as part of the semiannual APC review. Major claimants have been reporting status of training completion via the semiannual review report and have reported corrective actions are necessary. Incorporate into purchase card training programs any relevant changes in policies and procedures made as a result of the recommendations in this report. Policy changes resulting from previous GAO audit recommendations were incorporated into revised desk guides and training modules as well as the September 2002 revision of the DON EBUSOPSOFFINST 4200.1. Furthermore, a second combined Purchase Card/Travel Card APC conference was held in San Diego Nov. 5-8, 2002. A third combined conference was held in Philadelphia, Mar 17-20, 2003. On Sept 27, 2002, the Navy commenced distribution of training CDs that contained four training modules for purchase cardholders and approving officials (AO). A total of 30,000 CDs were distributed at that time. In December. 2002, the remaining three training modules were completed and posted to the DON eBusiness Web site for downloading of complete training modules. A second version of the Navy training CD was released and distributed at the March 2003 APC conference. This version contains seven training modules and a desk guide for each module for use by APCs. Each module contains both a Citidirect (shore) and WINSAALTS (afloat) version. Investigate ways to maximize potential rebates, such as (1) working with Citibank to facilitate timely receipt of monthly purchase card statements and (2) reducing the time associated with mailing and receipt of hard copy billing statements. A plan for the “on-line statement process” (electronic certification) was presented to APCs at the March 2003 APC conference in Philadelphia. All general fund activities are expected to be performing electronic certification by Sept. 30, 2003. All others, including outside the continental United States (OCONUS), nonappropriated fund (NAF), and Navy working capital fund (NWCF) activities, are expected to be performing electronic certification by June 30, 2004. In April 2003, a Navy Working Capital Fund users conference convened to discuss issues, explore problem areas, and develop an implementation plan. The electronic certification tool provides Navy purchase card customers with the ability to significantly decrease payment timelines, thereby optimizing rebate amounts. 7. Establish effective policies and procedures for routinely calculating and verifying Citibank rebates. The Defense Contract Audit Agency (DCAA) was assigned the task of auditing the integrity of the rebate computation process. A number of systems issues have been discussed with the banks and DCAA has finalized its audit recommendations. The Navy and the PC Program Management Office are assessing the results. GAO observation on the status of recommendation 8. Develop guidance for routine distribution of rebate earnings to Navy units and activities. The Navy made a determination to retain the rebates at the department level in lieu of disbursing them to lower echelons. 9. Establish in Navy Supply Systems Command (NAVASUP) Instruction 4200.94 further guidelines for an effective internal review program, such as having reviewers analyze monthly summary statements to identify (1) potentially fraudulent, improper, and abusive purchases and (2) any patterns of improper cardholder transactions, such as purchases of food or other prohibited items. EBUSOPSOFFINST 4200.1, chapter 4, addresses this issue. 10. Revise NAVSUP Instruction 4200.94 to require that (1) written reports on the results of internal reviews along with any recommendations for corrective actions be prepared and submitted to local management and cognizant commands and (2) commands identify and report systemic weaknesses and corrective action plans to the Naval Supply Systems Command for monitoring and oversight. EBUSOPSOFFINST 4200.1, chapter 4, addresses this issue. Semiannual program reviews have been established and reports are being submitted to the DON EBUSOPSOFF. 11. Require purchase card agency program coordinators to report in writing to the unit commander and the Commander of Naval Supply Systems Command any internal control weakness identified during the semiannual program reviews. EBUSOPSOFFINST 4200.1, chapter 4, addresses this issue. Semiannual program reviews have been established and reports are being submitted to the DON EBUSOPSOFF. 12. Disclose systemic purchase card control weaknesses along with corrective action plans in the Secretary of the Navy’s Annual Statement of Assurance, prepared under 31 U.S.C. 3512 (d). The Navy included systemic purchase card weaknesses identified in the semiannual report in the Secretary of the Navy’s Annual Statement of Assurance. 13. Revise NAVSUP Instruction 4200.94 to eliminate ambiguous language suggesting that advance independent authorization of a purchase can be substituted for independent confirmation that goods and services ordered and paid for with a purchase card have been received and accepted by the government. EBUSOPSOFFINST 4200.1 is a comprehensive instruction that addresses the roles of each participant in the purchase card process, with specific guidance addressing the responsibilities of each program participant. EBUSOPSOFFINST 4200.1, chapter 2, section 4d—Approving Official Duties, states that the AO will “ensure proper receipt, acceptance, and inspection is accomplished on all items being certified for payment.” Additionally, EBUSOPSOFF 4200.1, chapter 3, section 7– Establishing Internal Management Controls, discusses the separation of functions between receipt and acceptance of goods and services. GAO observation on the status of recommendation 14. Implement procedures to require and document independent confirmation of receipt of goods and services acquired with a purchase card. EBUSOPSOFFINST 4200.1, chapter 3, paragraph 7c, under Establishing Internal Management Controls separation of function, addresses this issue. 15. Revise NAVSUP Instruction 4200.94 to require that (1) cardholders notify approving officials prior to payment that purchase card statements have been reconciled to supporting documentation, (2) approving officials certify monthly statements only after reviewing them for potentially fraudulent, improper, and abusive transactions, and (3) approving officials verify, on a sample basis, supporting documentation for various cardholders’ transactions prior to certifying monthly statements for payment. EBUSOPSOFFINST 4200.1 has been revised to reflect the issues noted. The specific provisions are cited below: (1) Chapter 2, Section 6e, Cardholders Duties— Review the monthly purchase card statement to ensure that all charges are proper and accurate; (2) Chapter 2, Section 6f, Cardholders Duties— Forward the monthly purchase card statement to the AO with the appropriate supporting documentation, (i.e., sales slips, documentation of receipt and acceptance, purchase log) promptly to maximize rebates and minimize prompt payment penalties; (3) Chapter 2, Section 4.c, Approving Official Duties—Notify the Commanding Officer and APC in the event of any suspected unauthorized purchase (purchases that would indicate noncompliance, fraud, misuse, and/or abuse); (4) Chapter 2, Section 4.b, Authorizing Official Duties—Verify supporting transaction documentation on all card accounts prior to certifying the monthly invoice. 16. The Navy Comptroller withdrew the June 3, 1999, policy memorandum or revised the policy guidance to be consistent with the preceding recommendation for revising payment certification guidance in NAVSUP Instruction 4200.94. The Navy Comptroller policy letter dated June 3, 1999, was rescinded effective March 12, 2002. 17. Monitor and confirm that purchase card transactions are recorded to projects that benefited from the goods and services or to relevant overhead accounts promptly, in accordance with internal control standards and federal accounting standards. Both Public Works Center, San Diego, and Naval Space and Warfare Systems Command (SPAWAR) Systems Center, San Diego, concurred and are complying. Internal operating procedures at both sites include guidance on the issue. 18. Revise NAVSUP Instruction 4200.94 to require that purchase card expenses be properly classified in the Navy’s detailed accounting records. EBUSOPSOFFINST 4200.1, chapter 2, paragraph 4, addresses this issue. GAO observation on the status of recommendation 19. Verify that the detailed purchase card transaction records reflect the proper object classification of expense. Both Public Works Center, San Diego, and SPAWAR Systems Center, San Diego, concurred and are complying. The Navy issued policy as an interim change to NAVSUP Instruction 4200.94—Standards of Compliance for Timely Recording and Classifying of Navy Purchase Card Commitments and Obligations, which reiterates existing Navy and DOD Financial Management Regulation policy on the issue. 20. Require and verify that accountable property obtained using a purchase card is promptly recorded in property records as it is acquired, in accordance with DOD and Navy policies and procedures. The EBUSOPSOFFINST 4200.1 definition of accountable property reads as follows: Accountable Property: A term used to identify property recorded in a formal property management or accounting system. Accountable Property includes all property purchased, leased (capital leases), or otherwise obtained, having a unit acquisition cost of $5,000 or more (land, regardless of cost), and items that are sensitive, or classified. Additional and/or separate records or other recordkeeping instruments shall be established for management purposes, or when otherwise required by law, policy, regulation, or Agency direction, including, but not limited to pilferable items (items that have a ready resale value or application to personal possession and which are, therefore, especially subject to theft). Additionally, EBUSOPSOFFINST 4200.1, chapter 2, sections 6d and f, require a detailed purchase log to identify all purchase card transactions, including defined “Pilferable Personal Property.” Purchase log data are also forwarded to the AO as part of its purchase review process. A similar change is included in the draft revision to Secretary of Navy (SECNAV) Instruction 7320.10, which is currently in the coordination phase. All compromised accounts are closed. compromised purchase card accounts. 22. Determine whether purchases of excessive cost, questionable government need, or both, such as items for personal use, including personal digital assistants (such as Palm Pilots) and flat screen computer monitors, that were identified by GAO, are proper government purchases. If not, the Commander should prohibit their purchase. EBUSOPSOFFINST 4200.1, chapter 4, paragraphs 1 and 2 require a monthly 100 percent APC review and a semiannual APC review that addresses this issue. EBUSOPSOFFINST 4200.1, Enclosure 2, contains a list of generally prohibited items. Due to differing mission requirements and unique requirements throughout the Department of the Navy and DOD, it is difficult to develop a general list of what items can be purchased with or without special justification. Ticket purchases to Disneyland may be an appropriate purchase not requiring special justifications within a Non-Appropriated Funded activity, but may require such documentation at an Appropriated Funded activity. These decisions are best left to the local command. documented justifications and procurement management approval for types of items that can be acquired with a government purchase card. GAO observation on the status of recommendation 24. Examine purchase card acquisition guidance to determine whether the purchase card is the right vehicle for acquiring certain goods and services, such as vehicle and equipment maintenance, installation of upgraded computer software, and other recurring or installationwide services, or whether these items should be subject to negotiated contracts. The cardholder training CD in use Navy-wide contains specific information on the requirement to verify other contracting sources prior to making all purchases. EBUSOPSOFFINST 4200.1 chapter 2, section 6.b, Purchase Cardholder Duties, includes a requirement to “screen all requirements for their availability from mandatory Government sources of supply.” Additionally, the Naval Facilities Engineering Command (NAVFAC) has undertaken an initiative to facilitate client ordering of Indefinite Delivery Indefinite Quantity (IDIQ) services from Base Operations Support (BOS) contracts using the Department of Defense Electronic Mall (DOD EMALL). The NAVFAC Electronic Facilities support Contracts (e-FSC) initiative was created to facilitate direct client ordering by governmentwide commercial purchase cards (GCPC), thereby streamlining the BOS IDIQ ordering process and providing better compliance with DFARS 213.270 (Use of the Governmentwide Commercial Purchase Card). Each NAVFAC contract’s IDIQ schedule that is posted to the DOD EMALL is from a competed contract that has satisfied Competition In Contracting Act (CICA) requirements. This distinguishes NAVFAC contracts on the DOD EMALL from blanket purchase agreements (BPA) and other contracting instruments since orders off of the IDIQ catalogs are not considered stand-alone (open-market) purchases. The e-FSC initiative is currently in its early stages. NAVFAC is in the process of adding the Payment by Third Party clause (48 C.F.R. 52.232-36) and an e-FSC requirement to all new BOS solicitations and selected existing BOS contracts from installations and regions across the DON. As new BOS IDIQ schedules continue to be uploaded to the DOD EMALL, NAVFAC anticipates that within the next year many BOS contracts at most major installations will be available for electronic ordering by GCPC. This strategic initiative is expected to result in significant labor-hour savings and expedite the order and delivery process. GAO observation on the status of recommendation 25. Work with the Under Secretary for Acquisition, Technology, and Logistics and DOD’s Purchase Card Joint Program Office to determine whether the purchase card should be used to acquire computers and other equipment or property items individually that could be more economically and efficiently procured through bulk purchases. The DOD EMALL is now available to the Navy and its use and availability are being articulated to Navy purchasers as a single point for commercial purchases, including computers, using the government purchase card. Most recently, a DOD EMALL representative spoke at the March 2003 APC conference. The DON EBUSOPSOFF is in the process of querying all Level III APCs to identify all strategic sourcing agreements in their respective claimancies. This information will then be shared Navywide. 26. Revise NAVSUP Instruction 4200.94 to make it consistent with the Federal Acquisition Regulation, 48 C.F.R. 13.301(a), which states that the “card may be used only for purchases that are otherwise authorized by law or regulation.” The clarifying guidance should specifically state that in the absence of specific statutory authority, purchases of items for the personal benefit of government employees, such as flowers or food, are not permitted and are therefore improper transactions. EBUSOPSOFFINST 4200.1, chapter 1, paragraph 4, contains clarifying guidance. 27. Prohibit splitting purchases into multiple transactions as required by the Federal Acquisition Regulation and emphasize this prohibition in purchase card training provided to cardholders and approving officials. EBUSOPSOFFINST 4200.1, chapter 1, paragraph 5a, addresses this specific issue. In addition, training modules emphasize the prohibition on split purchases, as do all monthly and semiannual program reviews. Also, APCs have an on-line tool to monitor split purchases. 28. Require approving officials to monitor monthly purchase card statements and identify and report to them regarding any split purchases and the names of cardholders who made the transactions. EBUSOPSOFFINST 4200.1, chapter 2, paragraph 4, requires the approving official to verify supporting documentation on all card accounts prior to certifying the monthly accounts. Detecting potential split purchases and notifying AOs to review these transactions will be a capability of the data-mining tool. The tool will push the suspected split purchase down to the AO for review. The data-mining tool will identify the cardholder(s) who are splitting purchase requirements, along with the disciplinary actions associated with the transaction. 29. Incorporate GAO recommendations, to the extent applicable, into the Commander of the Naval Supply Systems Command’s future revisions to NAVSUP Instruction 4200.94, to include specific consequences for noncompliance with these guidelines and not enforcing the guidelines. Guidance for actions that may be taken for noncompliance with the regulations have been incorporated as disciplinary guidelines in the draft revision of EBUSOPSOFFINST 4200.1A. GAO observation on the status of recommendation Purchase Cards: Navy Is Vulnerable to Fraud and Abuse but Is Taking Action to Resolve Control Weaknesses (GAO-02-1041, Sept. 27, 2002) On September 19, 2002, DON issued EBUSOPSOFFINST 4200.1 that mandates a maximum span of control of card accounts to approving officials (AO) of 7:1: this metric (span of control ratio of 7:1) is monitored by the Navy on a monthly basis and corrective action is taken as required. the number of cardholders who report to an approving official and make the changes necessary to prevent approving officials from having the responsibility of reviewing more cardholders than allowed by Navy and DOD policies. 31. Establish a database that maintains information on all purchase card training taken by cardholders, approving officials, and agency program coordinators. Require that agency program coordinators update that database whenever these purchase card program officials take training. DON EBUSOPSOFF is building and will maintain an automated centralized training database using an e- mail response mechanism. Training completion responses are currently being accumulated and held off-line until the centralized training database is completed. After completion, e-mail responses will be electronically processed and student records will be recorded in the database. cardholders, approving officials, and agency program coordinators tailored to the specific responsibilities associated with each of these roles. Role-based training for APCs, Aos, and cardholders have been developed and distributed to all participants on CD ROM and are also posted to the DON EBUSOPSOFF Web site for downloading. This version contains seven training modules and a desk guide for each module for use by APCs. Each module contains both a Citidirect (shore) and WINSAALTS (afloat) version. EBUSOPSOFFINST 4200.1 mandates a maximum 7:1 ratio (seven accounts for each AO). Additionally, approximately 1 year ago, a one-time purge was done to realign the hierarchies in accordance with this policy. Compliance is monitored by EBUSOPSOFF twice a month with data from Citidirect. When an AO is found to be operating outside the ratio, they are notified and are required to take corrective action. Also on October 29, 2001, DON issued a policy letter PC02-05 and PCPN #69 requiring all agency program coordinators to review an approving official’s overall workload and determine whether the approving official has the time necessary to perform the required review functions. If the determination is that an approving official does not have the necessary time, the APC will address this situation with the approving official’s commander or supervisor. approving official’s overall workload and determine whether the approving official has the time necessary to perform the required review functions. GAO observation on the status of recommendation 34. Establish job descriptions that identify responsibility and performance standards for cardholders, approving officials, and agency program coordinators. Established recommended guidelines in the DON PC desk guides. 35. Link the cardholders’, approving officials, and agency program coordinators’ performance appraisals to achieving their performance standards. The Office of the Secretary of Defense, Personnel and Readiness, has advised that inclusion of purchase card duties in the performance goals is solely a supervisory responsibility, just as the inclusion of other performance outcomes, and should not be separately mandated. The DON EBUSOPSOFF cannot mandate this requirement. Performance goals are established by supervisors and employees and are a reflection of the employee’s major duties/responsibilities and the desired performance outcomes based on those duties. The goals established and the performance appraisals given are unique to the individual. However, roles and responsibilities have been outlined in DON PC desk guides. Internal management controls have been identified (e.g., span of control ratios, credit limit determination, delinquency management, separation of functions). 36. Work with the Naval Audit Service and Command Evaluation staff to begin periodic audits of the purchase card program to provide Navy management at the command and unit levels an independent assessment of the control environment and whether the agency program coordinators, approving officials, and cardholders are adhering to control procedures. The DON EBUSOPSOFF and the Assistant Secretary of the Navy (Research Development & Acquisition) Acquisition Business Management (ABM) offices are engaged with the Naval Audit Service (NAVAUDSVC) to finalize a schedule of purchase card command assessments. Ongoing audits: NAVAUDIT Activity reviews Validate filters Rebates GAO Leveraging buying power DODIG Convenience checks OCONUS transactions 37. Identify vendors with which the Navy or Marine Corps uses purchase cards to make frequent purchases, evaluate Navy purchasing practices with those vendors, and forward the results of that evaluation to the Assistant Secretary of the Navy for Research, Development, and Acquisition to contract with them, when applicable, to optimize Navy purchasing power. The data mining of purchase card transactions was completed in September 2002 and May 2003. Total transactions by vendor were extracted from the bank database and forwarded to the Office of the ASN (ACQ) for further review to determine whether Navy- wide contracts should be established. The EBUSOPSOFFINST 4200.1 definition of accountable property reads as follows: Accountable Property: A term used to identify property recorded in a formal property management or accounting system. Accountable Property includes all property purchased, leased (capital leases), or otherwise obtained, having a unit acquisition cost of $5,000 or more (land, regardless of cost), and items that are sensitive, or classified. Additional and/or separate records or other recordkeeping instruments shall be established for management purposes, or when otherwise required by law, policy, regulation, or Agency direction, including, but not limited to pilferable items (items that have a ready resale value or application to personal possession and which are, therefore, especially subject to theft). Additionally, EBUSOPSOFFINST 4200.1, chapter 2, sections 6d and f, requires a detailed purchase log to identify all purchase card transactions, including defined “Pilferable Personal Property.” Purchase log data are also forwarded to the AO as part of the AO purchase card review process. A similar change is included in the draft revision of SECNAV Instruction 7320.10, which is currently in the coordination phase. Property” in SECNAV Instruction 7320.10 dated August 1, 2001, by eliminating the requirement that a portable item easily converted to personal use also be difficult to repair or replace, and specifically identify items such as computers, cameras, personal digital assistants, and audiovisual equipment as meeting the definition of being pilferable and thus accountable. Modify NAVSUP Instruction 4200.94 to provide cardholders, approving officials, and agency program coordinators detailed instructions on the following: 39. Timely and independent receiving and acceptance of items obtained with a purchase card and documenting the results of that process. Complete – This subject was addressed in the DONEBUSOPSOFF Instruction 4200.1 dated September 19, 2002. 40. Screening purchases for the availability from required vendors and documenting the results of the screening. Complete – This subject was addressed in the DONEBUSOPSOFF Instruction 4200.1 dated September 19, 2002. 41. Promptly reconciling the monthly purchase card statements to supporting documentation and documenting the results of that reconciliation. Complete – This subject was addressed in the DONEBUSOPSOFF Instruction 4200.1 dated September 19, 2002. 42. Promptly reviewing a cardholder purchase card statement by the approving official prior to certifying the statement for payment and documenting the results of that review. Complete – This subject was addressed in the DONEBUSOPSOFF Instruction 4200.1 dated September 19, 2002. GAO observation on the status of recommendation 43. Prompt cardholder notification to property accountability officer of the pilferable property obtained with the purchase card, and approving official responsibility for monitoring that the pilferable property has been recorded in the accountability records. Complete – This subject was addressed in the DONEBUSOPSOFF Instruction 4200.1 dated September 19, 2002. Potentially fraudulent, improper, and abusive or questionable purchases 44. Modify NAVSUP Instruction 4200.94 to require cardholders to maintain documented justification and advanced approval of purchases that fall outside the normal procurements of the cardholder in terms of either dollar amount or type of purchase. Complete – This subject was addressed in the DONEBUSOPSOFF Instruction 4200.1 dated September 19, 2002. 45. Establish a Navy-wide database of known purchase card fraud cases by type of fraud that can be used to identify deficiencies in existing internal control and to develop and implement additional control activities, if warranted or justified. This is a parallel effort to the Automated Review and Response Oversight Wizard (ARROW) data-mining tool discussed below. ARROW is in the early stages of development. Additionally, the Office of the DOD IG, Investigative Policy and Oversight, has established a Government Purchase Card Fraud Investigations database that is already operational. 46. Establish a Navy-wide data-mining, analysis, and investigation function to supplement other oversight activities. This function should include providing oversight results and alerts to major commands and installations when warranted. The EBUSOPSOFF, in partnership with DOD IG, is conducting the ARROW data-mining project. Phase I of the data-mining pilot was completed in June 2003 at the Marine Corps site at Camp LeJeune, Fayetteville, N.C. The automated process is functioning as planned and initial reaction to the process from the participants has been positive; however, the fraud indicators were not adequately validated in Phase I. Phase 2 development will focus on validating the proposed fraud indicators. The Phase 2 pilot will begin in January 2004. 47. Modify NAVSUP Instruction 4200.94 to include a schedule of disciplinary actions as a guide for taking action against cardholders who make improper or abusive acquisitions with the purchase card. A schedule of disciplinary actions has been incorporated in the revised EBUSOPSOFF instruction 4200.1A. 48. Incorporate GAO recommendations, to the extent applicable, into the Charge Card Task Force’s future recommendations to improve purchase card policies and procedures throughout DOD. The Navy sent this recommendation to OUSD for action. GAO observation on the status of recommendation Purchase Cards: Control Weaknesses Leave the Air Force Vulnerable to Fraud, Waste, and Abuse (GAO-03-292, Dec. 20, 2002) Overall program management and environment Direct the Assistant Secretary of the Air Force for Acquisition and the Deputy Assistant Secretary for Contracting to take the following actions: 1. Establish specific policies and strategies governing the number of purchase cards to be issued with a focus on minimizing the number of cardholders. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that the number of cards issued should be minimized. Item closed. 2. Direct all command and installation-level agency program coordinators to review purchase card use with a view towards eliminating unneeded purchase card accounts. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that Installation Purchase Card Managers eliminate unneeded purchase card accounts. Item closed Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that Installation Purchase Card managers, in conjunction with Financial Services officers, review all purchase cardholders with multiple accounts and eliminate those accounts existing to facilitate line-item accounting. Item closed. 4. Direct all agency program coordinators to review the number of cardholders who report to an approving official and make the changes necessary so that approving officials do not have responsibility for reviewing more cardholder accounts than allowed by Air Force and DOD policies. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that Installation Purchase Card managers and coordinators review the number of cardholders who report to an approving official and make the changes necessary so that approving officials do not have responsibility for reviewing more cardholder accounts than allowed by Air Force and DOD policies. Item closed. 5. Review existing credit limits and monthly spending and develop policies and strategies on credit limits provided to cardholders with a focus on minimizing specific cardholder spending authority and minimizing the federal government’s financial exposure. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that installation purchase card program managers shall to review existing credit and monthly spending limits against current spending patterns and determine if cardholder spending authority can be reduced in the interest of minimizing the federal government's financial exposure. 6. Deactivate purchase card accounts of alternate cardholders and approving officials when primary cardholders and approving officials are available. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that alternate cardholders and billing official accounts be suspended when primary cardholders and billing officials are available. Item closed. GAO observation on the status of recommendation 7. Establish specific training courses for cardholders, approving officials, and agency program coordinators tailored to the specific responsibilities associated with each of those roles. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that, in addition to already instituted mandatory training through the Defense Acquisition University for cardholders, billing officials, and financial services officers, all A/OPCs are required to take the A/OPC training developed by GSA. Item closed. 8. Require installation program coordinators to track and monitor corrective actions on purchase card audit and annual surveillance findings and provide periodic status reports to their installation contracting directors. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that Installation Purchase Card managers track and monitor corrective actions on purchase cards and annual surveillance findings and provide quarterly status reports to their installation Contracting Director. Item closed. oversight system for program coordinators that includes standard activities and analytical tools to be used in evaluating program results. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed mandatory use of the review checklist in the GPC Surveillance Guide. Item closed. 10. Require reports on annual surveillance results to include an assessment of control environment issues, including the ratio of cardholders to employees, ratio of approving officials to cardholder accounts, ratio of monthly credit limits to actual spending, and number of cardholders and approving officials requiring training. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that reports on annual surveillance results include an assessment of control environment issues, including the ratio of cardholders to employees, ratio of approving officials to cardholder accounts, ratio of monthly credit limits to actual spending, and number of cardholders and approving officials requiring training. Item closed. 11. Assess the adequacy of human capital resources devoted to the purchase card program, especially for oversight activities at each management level, and provide needed resources where appropriate. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that the Directors of Contracting address the adequacy of personnel devoted to the purchase card program, especially for oversight activities, at each management level, and work to increase manpower authorizations where appropriate. Direct the Assistant Secretary of the Air Force for Acquisition and the Deputy Assistant Secretary for Contracting to make the following revisions to Air Force Instruction 64- 117, Air Force Government-wide Purchase Card Program: 12. Correct faulty records retention guidance by referring to specific guidelines in the Federal Acquisition Regulation, National Archives and Records Administration federal records retention guidelines, DOD’s Financial Management Regulation, and other federal guidelines as appropriate. Per the Air Force response to the GAO final report, correction was incorporated into the December 6, 2002, revision to AFI 64-117. Item closed. GAO observation on the status of recommendation management and administrative records generated by installation program coordinators and approving officials, such as records of cardholder and approving official appointments and training, cardholder delegations of authority, and purchase card surveillances, to be retained for 3 years. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117, Air Force Government-wide Purchase Card Program, will be revised to require purchase card program management and administrative records generated by installation program coordinators and approving officials, such as records of cardholder and approving official appointments and training, cardholder delegations of authority, and purchase card surveillances, to be retained for 3 years. 14. Stipulate, in the body of the Instruction, that approving officials are required to have annual purchase card refresher training. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117, Air Force Government-wide Purchase Card Program, will be revised to specify that approving officials are required to have annual purchase card refresher training. 15. Require that the surveillance checklist, which is included in an appendix to the Air Force Instruction, be used to guide and document surveillance results. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117, Air Force Government-wide Purchase Card Program, will be revised to require that the surveillance checklist be used to guide and document surveillance results. 16. Require reports on the results of annual surveillances to be signed by installation contracting directors to demonstrate management oversight and “tone at the top.” Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117, Air Force Government-wide Purchase Card Program, will be revised to require reports on the results of annual surveillances to be signed by the contracting squadron commander/chief of the contracting office. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117, Air Force Government-wide Purchase Card Program, will be revised to require reports on surveillance results to be addressed to unit commanders. GAO observation on the status of recommendation 18. Require reports on surveillance results to include recommendations for unit commander action, where approving officials and cardholders have failed to follow Air Force policy—particularly policy related to federal regulations, such as micropurchase requirements and mandated sources of supply. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117, Air Force Government-wide Purchase Card Program, will be revised to require reports on surveillance results to include recommendations for unit commander action, where approving officials and cardholders have failed to follow Air Force policy related to federal regulations. To resolve noncompliance with requirements in law for proper certification of purchase card payments, we recommend that the Secretary of the Air Force take the following actions: 19. Direct the Assistant Secretary of the Air Force for Acquisition and the Deputy Assistant Secretary for Contracting to work with the Under Secretary of Defense (Comptroller) to resolve inconsistencies between DOD and Air Force policies and procedures for reconciling purchase card statements prior to payment. The Under Secretary of Defense (Comptroller) requested an opinion from the Deputy General Counsel (Fiscal) (DGC(F)) to determine whether “pay and confirm” is in compliance with Title 10, United States Code (U.S.C.), section 2784. In its response, the DGC(F) stated that the business practice of paying a purchase card statement of account before receipt of a reconciled statement and detailed supporting documentation is supported by governmentwide policy, and not otherwise prohibited by statute. Counsel did caution that the practice is contingent upon maintaining appropriate internal controls sufficient to ensure that the benefits associated with this practice outweigh the risk of loss. Item closed. The Under Secretary of Defense (Comptroller) requested an opinion from the Deputy General Counsel (Fiscal) (DGC(F)) to determine whether “pay and confirm” is in compliance with Title 10, United States Code (U.S.C.), section 2784. In its response, the DGC(F) stated that the business practice of paying a purchase card statement of account before receipt of a reconciled statement and detailed supporting documentation is supported by governmentwide policy, and not otherwise prohibited by statute. Counsel did caution that the practice is contingent upon maintaining appropriate internal controls sufficient to ensure that the benefits associated with this practice outweigh the risk of loss. Item closed. compliance with requirements in the law that DOD purchase card policies and procedures require reconciliation of purchase card statements prior to payment. GAO observation on the status of recommendation 21. Establish appropriate criteria, including types of items and dollar thresholds for documenting independent receipt and acceptance of items obtained with a purchase card. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. The Deputy Assistant Secretary of Contracting will revise Air Force Instruction 64-117 to provide cardholders, approving officials, and installation program coordinators appropriate criteria, including types of items and dollar thresholds for documenting independent receipt and acceptance of items obtained with a purchase card. documenting independent receiving, such as requiring the approving official or supervisor to sign and date the vendor invoice, sales receipt, or credit card receipt, or requiring the approving official to sign the cardholder’s monthly purchase log to verify that items noted as having been received were actually received. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. The Deputy Assistant Secretary of Contracting will revise Air Force Instruction 64-117 to provide cardholders, approving officials, and installation program coordinators with detailed instructions on procedures for documenting independent receiving, such as requiring the approving official or supervisor to sign and date the vendor invoice, sales receipt, or credit card receipt, or requiring the approving official to sign the cardholder’s monthly purchase log to verify that items noted as having been received were actually received. documentation of timely and independent receiving and acceptance of items obtained with a purchase card. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Deputy Assistant Secretary of the Air Force (Contracting) will revise Air Force Instruction 64- 117 to require cardholders to maintain documentation of independent receiving and acceptance of items obtained with a purchase card. A SAF/AQC letter, dated March 27, 2003, was sent to the purchase card points of contact at the Air Force major commands requesting that they direct their A/OPCs to review all accounts subject to automatic suspension in July 2002 due to lack of cardholder reconciliation and approving official review to ensure that they have been manually reconciled. Item closed. purchase card statements associated with accounts that were “shut down” (suspended) in July 2002 due to lack of cardholder reconciliation and approving official review. GAO observation on the status of recommendation 25. Verify that all potentially fraudulent and erroneous transactions that have been detected are disputed and properly resolved. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. The Deputy Assistant Secretary of Contracting will revise Air Force Instruction 64-117 to instruct cardholders, approving officials, and installation program coordinators to verify that all potentially fraudulent and erroneous transactions that have been detected are disputed and properly resolved. 26. Require timely cardholder notification to the property accountability officer of pilferable property, such as fax machines, digital cameras, and palm pilots obtained with the purchase card. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. The Deputy Assistant Secretary of Contracting will revise Air Force Instruction 64-117 to provide cardholders, approving officials, and installation program coordinators with detailed instructions to require timely cardholder notification to the property accountability officer of accountable pilferable property obtained with the purchase card. Air Force Policy Memo #03-C-05, issued March 2003, encouraged installation Contracting Officers to consider the benefits of central purchasing and receiving and acceptance of computer equipment by installation information technology units to facilitate recording computer equipment in accountable property records at the time it is received. Item closed. to consider the benefits of central purchasing and receiving and acceptance of computer equipment by installation information technology units to facilitate recording computer equipment in accountable property records at the time it is received. GAO observation on the status of recommendation 28. Define and list examples of sensitive and pilferable property purchased with a government purchase card, including cell phones, digital cameras, fax machines, palm pilots, and copiers and printers, and require prompt recording of these items in installation property systems. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Air Force Instruction 64-117 will be revised to define and list examples of sensitive and pilferable property purchased with a government purchase card, including cell phones, digital cameras, fax machines, palm pilots, and copiers and printers. Sub-paragraph 5.3.1 of DODI 5000.64 allows additional and/or separate records or other recordkeeping instruments when required by law, policy, regulation, Agency direction, or for management purposes (e.g., pilferable item, property hazardous to health and human safely). Property not meeting the minimum accountability threshold is still subject to appropriate internal controls which, depending on the property, can include an accountable property record. SAF/AQCP is working with USAF/ILGP, Materiel Management Policy Division, to establish clear accountability and/or visibility criteria that will meet the intent of GAO’s accountability concerns. recording all pilferable and sensitive property, including digital cameras, palm pilots, and cell phones, in installation- accountable property records. At a minimum, require installations to follow DOD policies and procedures on accountable property. Anticipate publication of revised AFI in March 2004. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. Current guidance (DODI 5000.64, AFI 33-112, AFI 23-111, and AFI 23-110) all indicate that organizational commanders must account for property issued to them or procured by them. These guidelines do not mandate a mechanism to ensure accountability is established for items procured from outside of the standard base supply system. SAF/AQCP is working with USAF/ILGP, Materiel Management Policy Division, to establish clear accountability and/or visibility criteria that will meet the intent of GAO’s accountability concerns. These changes will be incorporated into AFI 64-117. 30. Direct the Air Force Audit Agency and Air Force Office of Special Investigations to establish an Air Forcewide database of known fraud cases by type of fraud, including purchase card fraud, that can be used to identify systemic weaknesses and deficiencies in existing internal control and to develop and implement additional control activities, if warranted or justified. The Air Force Office of Special Investigations (AFOSI), in conjunction with the other Defense Criminal Investigative Organizations (DCIO), now reports information on initiated and ongoing Government Purchase Card (GPC) investigations quarterly to the Department of Defense Inspector General for macro-level analysis of systemic weaknesses in the GPC program DOD-wide. The DOD IG has been directed to develop a centralized purchase card database on known fraud cases and audit results that can be used to identify potential deficiencies in existing internal controls. The Air Force will evaluate the Air Force cases and audits to determine the effectiveness of existing internal controls and implement additional control activities, if warranted. known purchase card fraud cases by type of fraud, including vendor fraud and compromised accounts, that can be used to identify deficiencies in existing internal control and implement additional control activities, if warranted. 32. Identify vendors with which the Air Force used purchase cards to make frequent, recurring purchases, evaluate Air Force purchasing practices with those vendors, and where appropriate, develop contracts with those vendors to optimize Air Force purchasing power. Air Force Policy Memo #03-C-11, issued May 22, 2003, directed that A/OPCs identify vendors with which they used purchase cards to make frequent, recurring purchases, evaluate purchasing practices with those vendors, and where appropriate, develop contracts with those vendors to optimize Air Force purchasing power. Item closed. Anticipate publication of revised AFI in Mar 04. purchase card and revoke purchase cards issued to organizations that do not have authority to participate in the governmentwide purchase card program. Reported partially implemented pending March 2004 issuance of revised Air Force Instruction 64-117. The Deputy Assistant Secretary for Contracting will review organizational use of the purchase card and revoke purchase cards issued to organizations that do not have authority to participate in the governmentwide purchase card program. However, AF/HC does not agree that the Chaplain Service had no authority to use GPCs. DODD 1015.1. recognizes Chaplain Religious Funds and states that “funds are administered and managed in accordance with separate DOD Component regulations” (Par. 2.2. and 2.2.11.). Based on DODD 1015.1, AFI 52- 101 (May 19, 1997) was issued that stated “The International Merchant Purchase Authorization Card (IMPAC) is the official Chaplain Service funds credit card” (Para. 4.3.). AF/HC will recommend reinstatement of the Chaplain Funds into the revised publication of DODD 1015.1, Establishment, Management, and Control of Nonappropriated Fund Instrumentalities. AFI 52- 101 is in the process of being updated to reflect the current DOD and AF policies regarding the GPC. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that installation purchase card managers cancel convenience check privileges of cardholders who have misused convenience checks more than once. Item closed. 35. Require accounting adjustments to be made to correct transactions that were charged to the wrong appropriation account with respect to fiscal year and purpose of the expenditures. A SAF/AQC letter, dated March 27, 2003, was sent to SAF/FMP requesting that an accounting adjustment be made to correct any GPC transactions that were charged to the wrong appropriations account with respect to fiscal year and items purchased. Item closed. Force-wide policy as a guide for taking disciplinary actions with respect to cardholders and approving officials who make or approve fraudulent, improper, or abusive purchase card transactions. The Deputy Assistant Secretary of the Air Force (Contracting) does not make Air Force-wide policy as a guide for taking disciplinary actions with respect to cardholders and approving officials who make or approve fraudulent, improper, or abusive purchase card transactions. Guidelines for procedures regarding the violation of Air Force GPC procedures are already contained in AFI 64- 117. In addition, the Deputy Assistant Secretary of the Air Force (Contracting) has issued a memorandum requiring a summary of each case of purchase card fraud and each instance of repeated misuse of the purchase card and a quarterly briefing by the contracting squadron commander to the installation commander including the disciplinary action taken. Item closed. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that installation purchase card managers remind all cardholders and billing officials that they are “accountable officials” in accordance with Attachment 2, paragraph 1.b. of AFI 64-117, and as such, may be pecuniarily liable for erroneous payments (see DOD Financial Management Regulation, Volume 5, Chapter 33, August 1998, page 33-1) and may be required to reimburse the government for any unauthorized or erroneous purchase card transaction that was not disputed within the 60-day grace period. In addition, all “benefiting individuals” who have requested personal items to be purchased for their use may also be required to reimburse the government for such purchases. Item closed. officials to reimburse the government for any unauthorized or erroneous purchase card transactions that were not disputed. GAO observation on the status of recommendation reimburse the government for the cost of any personal items that they requested or directed a cardholder to purchase for them. Air Force Policy Memo #03-C-05, issued March 18, 2003, directed that installation purchase card managers remind all cardholders and billing officials that they are “accountable officials” in accordance with Attachment 2, paragraph 1.b. of AFI 64-117, and as such, may be pecuniary liable for erroneous payments (see DOD Financial Management Regulation, Volume 5, Chapter 33, August 1998, page 33-1) and may be required to reimburse the government for any unauthorized or erroneous purchase card transaction that was not disputed within the 60-day grace period. In addition, all “benefiting individuals” who have requested personal items to be purchased for their use may also be required to reimburse the government for such purchases. Item closed. 39. Incorporate GAO recommendations, to the extent applicable, into the Charge Card Task Force’s future recommendations to improve purchase card policies and procedures throughout DOD. This recommendation was directed to the Under Secretary of Defense (Comptroller), not to the Air Force. Staff making key contributions to this report were Francine DelVecchio, Gail Luna, Jerrod O’Nelio, Harold Reich, John Ryan, Quan Thai, and Gary Wiggins. The General Accounting Office, the audit, evaluation and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to e-mail alerts” under the “Order GAO Products” heading.

This study responds to a legislative mandate, which directs the Comptroller General to review the actions taken by the Department of Defense (DOD) to implement provisions included in the Bob Stump National Defense Authorization Act for fiscal year 2003 (Public Law 107-314) concerning management of the purchase card program. This study also discusses DOD efforts to implement provisions in the DOD Appropriations Act for fiscal year 2003 (Public Law 107-248) as well as recommendations and the status of disciplinary actions taken against individuals identified in prior GAO reports as having used the government purchase card for potentially fraudulent, improper, and abusive or questionable purposes. DOD has initiated actions to implement all of the requirements in the National Defense Authorization Act for fiscal year 2003 and the DOD Appropriations Act for fiscal year 2003. While it has largely completed revamping its policies and other requirements, it still had considerable work to complete in order to implement managerial and oversight mechanisms, such as strategic sourcing, monitoring, and auditing. However, to implement the legislative requirement that DOD evaluate credit worthiness prior to issuing a purchase card, DOD is allowing cardholders to self-certify their credit worthiness rather than conducting credit checks on cardholders, as is typically done in the private sector. DOD started actions to implement nearly all of the 109 GAO recommendations, some of which may closely relate to the legislative provisions. DOD and the military services have taken disciplinary actions against cardholders whom a court of law determined had fraudulently used their purchase cards. They have also started to educate cardholders and approving officials on the proper use of the purchase card. The military services have not taken strong disciplinary actions against cardholders GAO identified as making improper and abusive or questionable purchase card acquisitions. The military services determined that many of these purchases did not directly violate existing policies. Consequently, the services modified these policies to provide a basis for disciplinary actions for similar purchases in the future.

Unmanned aircraft systems generally consist of (1) multiple aircraft, which can be expendable or recoverable and can carry lethal or non-lethal payloads; (2) a flight control station; (3) information and retrieval or processing stations; and (4) in some cases, wheeled land vehicles that carry launch and recovery platforms. DOD categorizes these systems based on key characteristics including weight and operating altitude. While there were many small, less expensive unmanned aircraft in DOD’s portfolio, our review focused on the larger, more costly programs. At that time, these programs accounted for more than 80 percent of DOD’s total investment in unmanned aircraft from fiscal year 2008 through fiscal year 2013. DOD’s 2011 budget request indicates that the department plans to invest nearly $25 billion from 2010 through 2015 in development and procurement of the unmanned aircraft systems we reviewed. Table 1 details many of the key characteristics and funding requirements of those systems. See appendix I for additional program data. Despite the proven success of unmanned aircraft on the battlefield and the growing demand for the aircraft, these acquisitions continued to incur cost and schedule growth. The cumulative development cost for the 10 programs we reviewed increased by over $3 billion, or 37 percent, from initial estimates. While 3 of the 10 programs had little or no development cost growth and one had a cost reduction, six experienced substantial growth ranging from 60 to 264 percent. This cost growth was in large part the result of changes in program requirements and system designs after initiating development. Many of the programs began system development with unclear or poorly defined requirements, immature technologies, and unstable designs—problems we have frequently found in other major acquisition programs. For example, in 2001, the Air Force began the Global Hawk program based on knowledge gained from a demonstration program, and planned to incrementally integrate more advanced technologies over time. Within a year, however, the Air Force fundamentally restructured and accelerated the program to pursue a larger, unproven airframe with a multimission capability that relied on immature technologies. The final design of the new airframe required more substantial changes than expected. These changes ultimately drove development costs up nearly threefold. Procurement costs also increased for 6 of the 7 systems that reported procurement cost data. Although in large part the cost increases were due to the planned procurement of additional aircraft, many programs had also experienced unit cost increases independent of quantity. As detailed in table 2, overall procurement unit costs increased by 12 percent on average, with three programs experiencing unit cost growth of 25 percent of more. The Reaper and Shadow had unit cost growth despite increased quantities. Reaper’s unit costs increased in part because requirements for missiles and a digital electronic engine control were added—resulting in design changes and increased production costs. Unit cost increases in the Shadow program were largely the result of upgrades to the airframe that were needed to accommodate the size, weight, and power requirements for integrating a congressionally mandated data link onto the aircraft. Furthermore, the Army is retrofitting fielded systems with capabilities that it had initially deferred, such as a heavy fuel engine. A number of programs had experienced problems in both testing and performance, requiring additional development that contributed to the cost growth noted above. Four programs had experienced delays of 1 to nearly 4 years in achieving initial operational capability. Some of these delays resulted from expediting limited capability to the warfighter, while others were the result of system development and testing problems. For example, early demonstration and production Global Hawks were rushed into operational service. Program officials noted that as a result, the availability of test resources and time for testing were limited, which delayed the operational assessment of the original aircraft model by 3 years. Similarly, in February 2009, the Air Force reported that initial operational testing for the larger, more capable Global Hawk aircraft and the program’s production readiness review had schedule breaches. Air Force officials cite the high level of concurrency between development, production, and testing; poor contractor performance; developmental and technical problems; system failures; and bad weather as key reasons for the most recent schedule breach. Consistent with DOD’s framework for acquiring unmanned systems, some of the tactical and theater-level unmanned aircraft acquisition programs we reviewed had identified areas of commonality to leverage resources and gain efficiencies. For example, the Army and Marine Corps achieved full commonality in the Shadow program. In assessing options for replacing an aging tactical unmanned aircraft system, the Marine Corps determined that the Army’s Shadow system could meet its requirements for reconnaissance, surveillance, and target acquisition capabilities without any service-unique modifications. An official from DOD’s Office of Unmanned Warfare emphasized that the Marine Corps believed that Shadow represented a “100 percent” solution. The Marine Corps also found that it could use the Army’s ground control station to pilot the Shadow aircraft as well as other Marine Corps unmanned aircraft. A memorandum of agreement was established in July 2007 to articulate how the Marine Corps and the Army would coordinate to acquire Shadow systems. By forgoing any service-unique modifications in order to achieve a high level of commonality, the Marine Corps avoided the costs of developing the Shadow. Additionally, the Marine Corps and Army are likely to realize some benefits in supporting and maintaining the systems because the components are interchangeable. The Army’s Shadow program office agreed that commonality has allowed the two services to realize economies of scale while meeting each service’s needs. According to an official at the Navy, the Marine Corps has been able to realize savings or cost avoidance in other areas such as administration, contracting, and testing, although quantitative data on these savings were not available. In some cases, the services had collaborated to identify common configuration, performance, and support requirements, but ultimately were not maximizing efficiencies. For example, the Army and Navy had different data link requirements for their respective variants of Fire Scout, primarily because of the Army’s requirement for its variant to operate within the Future Combat Systems network. According to the Fire Scout contractor, the Army’s system could have been equipped with the same data link as the Navy Fire Scout, as well as the Army’s Shadow and Sky Warrior systems, and placed into service sooner. Though the services had not agreed on a common data link, the Army and Navy had settled on common Fire Scout requirements for the air vehicle, engine, radar, navigation, and some core avionics subsystems requirements. The services had also agreed to use one contract to procure the airframe. However, in an information letter sent to members of Congress on January 11, 2010, the Army noted that it had terminated the Fire Scout portion of its FCS contract—following a decision by the Office of the Secretary of Defense (OSD) to cancel the FCS program—because analysis indicated that an improved Shadow system could meet future Army requirements, and the Fire Scout was no longer needed. Cancellation of the Army Fire Scout could lead to increased unit cost for the Navy variant. Although the Navy BAMS and Air Force Global Hawk programs had identified commonalities between their airframes, the two programs had established different payload, subsystem, and ground station requirements. The Navy anticipated spending more than $3 billion to modify the Global Hawk airframe and ground stations, and to integrate Navy-specific payloads, including the radar. In addition, we found that the Navy had an opportunity to achieve greater efficiency in BAMS production. While production of the first two BAMS aircraft was planned to occur at the same California facility that produces Global Hawk, the remaining aircraft were expected to be produced at a facility in Florida. We pointed out that this approach might create duplication in production by staffing and equipping two facilities to conduct essentially the same work. At the time of our review the Navy had not assessed the costs or benefits of establishing a second production facility, and according to contractor officials, the official business case analysis would not be conducted for several years. Therefore, it was unclear whether any benefits of a second production facility would outweigh costs, such as additional tooling and personnel. In contrast to the examples of the Shadow, Fire Scout, and BAMS / Global Hawk programs above, the Army and Air Force missed opportunities to achieve commonality and efficiencies between their Sky Warrior and Predator programs. In 2001, the Army began defining requirements for a replacement to the aging Hunter unmanned aircraft system, and decided to pursue the development of Sky Warrior. Both the Air Force and the Joint Staff responsible for reviewing Sky Warrior’s requirements and acquisition documentation raised concerns about duplicating existing capability— specifically, capability provided by Predator. Nevertheless, the Army program received approval to forgo an analysis of alternatives that could have determined whether or not existing capabilities met its requirements. The Army noted that such an analysis was not needed and not worth the cost and effort. Instead, it conducted a source selection competition and began the Sky Warrior development program in 2005, citing battlefield commanders’ urgent need for the capability. The development contract was awarded to the same contractor working with the Air Force to develop and produce Predators and Reapers. Since the Sky Warrior is a variant of the Predator, the two aircraft are assembled in the same production facility. Despite the establishment of a memorandum of understanding in 2006, direction from the Deputy Secretary of Defense in 2007 to combine their programs, and a subsequent memorandum of agreement, the Army and Air Force maintained separate programs and at the time of our review, had achieved little commonality. While several of the unmanned aircraft programs we examined had achieved commonality at the airframe level, service-centric acquisition processes and ineffective collaboration resulted in service-unique subsystems, payloads, and ground control stations. Despite DOD’s efforts to encourage a joint approach to identifying and prioritizing warfighting needs and to emphasize the need for commonality among the programs, we noted that the individual services continued to drive requirements and make independent resource allocation decisions. In many cases, the services had established requirements so specific that they demanded service-unique solutions, thereby precluding opportunities for commonality. Within DOD’s funding system, each service has the responsibility and authority to prioritize its own budget, allowing it to make independent funding decisions to support unique requirements. Therefore, once a service concludes that a unique solution is warranted, the service has the authority to budget for that unique solution, to the exclusion of other solutions that might achieve greater commonality and efficiencies. While we recognized that service-unique requirements appeared to be necessary in some cases, one OSD official we spoke with emphasized concerns that some of the services’ distinctions in requirements could lead to duplication and inefficiencies. However, OSD had not quantified the potential costs or benefits of pursuing various alternatives, including commonality. In 2007, OSD established the Unmanned Aircraft Systems Task Force and the Office of Unmanned Warfare primarily to facilitate collaboration and encourage greater commonality among unmanned aircraft programs. While the two groups act as advisors and have implemented OSD’s recommendations regarding areas where further commonality might be achieved key officials from these groups emphasized to us that they do not have direct decision-making or resource allocation authority. OSD repeatedly directed the Army and Air Force to collaborate on their Sky Warrior and Predator programs, but the services continued to pursue unique systems. In response to OSD direction to merge their unique signals intelligence payload efforts into a single acquisition program, the Army and Air Force concluded that continuing their separate programs was warranted, and recommended that OSD direct an objective, independent organization—such as a federally funded research and development center—to conduct a business case analysis to assess the impact of merging the two programs. Table 3 summarizes OSD’s directions and the services’ responses over the past few years. Congress and OSD took additional action in 2009 aimed at increasing collaboration and commonality among unmanned aircraft programs. In section 144 of the Duncan Hunter National Defense Authorization Act for Fiscal Year 2009, Congress directed “he Secretary of Defense, in consultation with the Chairman of the Joint Chiefs of Staff, establish a policy and an acquisition strategy for intelligence, surveillance, and reconnaissance payloads and ground stations for manned and unmanned aerial vehicle systems. The policy and acquisition strategy shall be applicable throughout the Department of Defense and shall achieve integrated research, development, test, and evaluation, and procurement commonality.” In an acquisition decision memorandum issued on February 11, 2009, the Under Secretary of Defense for Acquisition, Technology and Logistics identified the opportunity to adopt a common unmanned aircraft ground control station architecture that supports future capability upgrades through an open system and modular design. Similar to OSD’s approach to ground control stations, the Air Force Unmanned Aircraft Systems Task Force expected future unmanned aircraft to be developed as open, modular systems to which new capabilities could be added instead of developing entirely new systems each time a new capability is needed. Since July 2009 when our report was issued, DOD has made several key investment decisions regarding unmanned aircraft systems that will likely impact those estimates. In general, these decisions reflect increased emphasis on developing more advanced unmanned aircraft capabilities and acquiring larger numbers of specific systems, but they do not appear to focus on increasing collaboration or commonality among systems. The 2010 Quadrennial Defense Review (QDR) reported that “U.S. forces would be able to perform their missions more effectively—both in the near-term and against future adversaries—if they had more and better key enabling capabilities at their disposal.” The QDR report included unmanned aircraft systems among these key enablers, and emphasized the importance of rapidly increasing the number and quality of unmanned aircraft systems—among other enablers—to prevail in today’s wars, and to deter and defeat aggression in anti-access environments. The report also noted that: the Air Force is going to increase the total number of Predator/Reaper aircraft it plans to buy; the Army will accelerate the production of its Predator-class Sky Warrior system; and the Navy will conduct field experiments with prototype versions of its Unmanned Combat Aircraft System, which, the QDR points out, offers the potential to greatly increase the range of strike, and intelligence, surveillance, and reconnaissance (ISR) operations from the Navy’s carrier fleet. As part of DOD’s fiscal year 2011 budget development process, OSD made several unmanned aircraft-related adjustments to the services’ budget submissions. As part of those adjustments, OSD: Directed the Army to stop development and initial fielding of its Fire Provided the Air Force an additional $344 million from FY2011 to FY2015 to develop, procure, and integrate counter-communication and counter-improvised explosive device jamming pods onto 33 MQ-9 Reaper aircraft, and directed the Air Force to present its assessment of platforms for this capability by June 1, 2010; Provided an additional $1.8 billion from FY2011 through FY2015 to purchase an additional 74 MQ-9 Reaper aircraft; Added $2 billion to the Navy budget from FY2013 to FY2015 to define requirements and develop unmanned carrier based capability, and directed the Navy to develop an execution plan by March 30, 2010; Added $201.6 million to the Global Hawk procurement budget to procure 19 Block 40 aircraft by 2015, and 22 total; Added $270.5 million for development and procurement of Global Hawk satellite communication terminals; Added $2.4 billion over the Future Years Defense Program to the Army’s Extended Range Multi-Purpose (Sky Warrior) Aircraft budget to procure an additional 12 aircraft and 5 ground stations (one company) per year from 2011 through 2015. In concert with the QDR and the fiscal year 2011 budget, DOD also published its first submission of a long-range, fixed-wing aviation procurement plan. Among other things, the plan addresses DOD’s strategy for meeting the demand for persistent, unmanned, multi-role ISR capabilities by: Emphasizing “long-endurance, unmanned ISR assets—many with strike capabilities—to meet warfighter demands; Projecting an increase in the number of platforms in this category from approximately 300 in 2011 to more than 800 in 2020, nearly 200 percent increase; Noting the “replacement of Air Force Predators with more capable Establishing a specific category for Unmanned Multi-role Surveillance and Strike systems, that distinguishes those systems from other types of aircraft, such as fighters and bombers; Noting that the department will continue to adapt the mix of unmanned and manned systems as security needs evolve; and Noting that unmanned systems are being considered as future long- range strike platforms and future fighter / attack aircraft. In closing, recent experience in Iraq and Afghanistan has proven that unmanned aircraft are extremely valuable to the warfighter, and it is clear that more are needed. However, DOD will continue to be challenged to meet this increasing demand within available resources. Many of DOD’s larger unmanned aircraft acquisition programs have experienced cost growth, schedule delays, and performance shortfalls, while not enough have achieved the efficiencies one might expect from commonality. DOD recognizes that to more effectively leverage its acquisition resources, it must achieve greater commonality among the military services’ various unmanned system programs. However, in many cases the services have preferred to pursue unique solutions. In general, the military services continue to establish unique requirements and prioritize resources while foregoing opportunities to achieve greater efficiencies. As a result, commonality has largely been limited to system airframes, and in most cases, has not been achieved among payloads, subsystems, or ground control stations. Opportunities for identifying commonality are greatest when requirements are being established. Therefore, as the department continues to develop and procure unmanned aircraft systems, it must take more care in setting requirements for those systems. Rather than looking for unique solutions to common problems, DOD must increasingly find common solutions to those problems. However, we recognize that commonality is not a panacea, and in some cases, given legitimate differences in operating environments or mission needs, may not make sense. We also recognize that achieving commonality is not always easy, especially given the strong service-driven acquisition processes and culture within the department. Therefore, in our July 2009 report we recommended that DOD (1) direct an objective, independent examination of unmanned aircraft requirements and report a strategy to Congress for achieving greater commonality among systems and subsystems, and (2) require future unmanned aircraft programs to take an open systems approach to product development and to clearly demonstrate that potential areas of commonality have been analyzed and identified. We believe that these steps could help overcome these barriers and could go a long way to ensuring that DOD maximizes efficiency as it continues to greatly increase emphasis on developing and acquiring more capable and larger quantities of unmanned aircraft. For further questions about this statement please contact Michael J. Sullivan at (202) 512-4841. Individuals making key contributions to this statement include Bruce Fairbairn, Assistant Director; Travis Masters; Rae Ann Sapp; Leigh Ann Nally; Laura Jezewski; and Susan Neill. This appendix contains 3 tables that provide additional information about the 8 unmanned aircraft systems assessed in our July 2009 report. Table 4 contains the combined total development and procurement funding DOD has requested in its fiscal year 2011 budget submission for each of the programs. The budget data is presented in then year dollars and may not add precisely due to rounding. Tables 5 and 6 detail many of the key characteristics and compare the capabilities of the systems discussed in this statement.

For the last several years, the Department of Defense (DOD) has planned to invest billions of dollars in development and procurement of unmanned aircraft systems. In its fiscal year 2011 budget request the department indicated a significant increase in these investments, expecting to need more than $24 billion from 2010 through 2015. DOD recognizes that to leverage its resources more effectively, it must achieve greater commonality among the military services' unmanned aircraft system acquisition programs. This testimony is based primarily on GAO's July 2009 report (GAO-09-520) which examined 10 unmanned aircraft acquisition programs: eight unmanned aircraft systems--Global Hawk, Reaper, Shadow, Predator, Sky Warrior, Fire Scout, Broad Area Maritime Surveillance, and Unmanned Combat Aircraft System-Demonstration; and two payload development programs--Multi-Platform Radar Technology Insertion Program, and Airborne Signals Intelligence Payload. The testimony focuses on: 1) the cost, schedule, and performance progress of the 10 programs as of July 2009; 2) the extent to which the military services collaborated and identified commonality among the programs; 3) factors influencing the effectiveness of the collaboration; and, 4) recent DOD investment decisions related to these acquisitions. Most of the 10 programs reviewed had experienced cost increases, schedule delays, performance shortfalls, or some combination of these problems. The programs' development cost estimates increased by more than $3 billion collectively, or 37 percent, from initial estimates. Procurement funding requirements for most programs also increased, primarily because of increases in numbers of aircraft being procured, changes in system requirements, and upgrades and retrofits to fielded systems. Procurement unit costs increased by an average of 12 percent, with three aircraft programs experiencing unit cost increases of 25 percent or more. Four programs reported delays of 1 year or more in delivering capability to the warfighter. Global Hawk, Predator, Reaper, and Shadow had been used in combat operations with success and lessons learned, but had been rushed into service in some cases, leading to performance issues and delays in development and operational testing and verification. Programs collaborated and identified areas of commonality to varying degrees. The Marine Corps was able to avoid the cost of initial system development and quickly deliver useful capability to the warfighter by choosing to procure existing Army Shadow systems. The Navy expected to save time and money on Broad Area Maritime Surveillance (BAMS) by using Air Force's Global Hawk airframe, and payloads and subsystems from other programs. However, Army and Air Force had not collaborated on their Sky Warrior and Predator programs, and might have achieved greater savings if they had, given that Sky Warrior is a variant of Predator and being developed by the same contractor. DOD encouraged more commonality between these programs. Although several programs achieved airframe commonality, service-driven acquisition processes and ineffective collaboration were key factors that inhibited commonality among subsystems, payloads, and ground control stations, raising concerns about potential inefficiencies and duplication. Despite DOD's efforts to emphasize a joint approach to identifying needs and commonality among systems, most of the programs assessed continued to pursue service-unique requirements. The services also made independent resource allocation decisions to support their unique requirements. DOD had not quantified the costs and benefits associated with pursuing commonality among these programs, and efforts to collaborate had produced mixed results. However, in order to maximize acquisition resources and meet increased demand, Congress and DOD have continued to push for more commonality. Since July 2009, DOD has made several investment decisions regarding unmanned aircraft systems, which in general, reflect increased emphasis on developing advanced capabilities and acquiring larger numbers of specific systems. However, the decisions do not appear to focus on increasing collaboration or commonality among the programs.

The SSO program covers all states with fixed guideway systems operating in their jurisdictions. FTA defines a rail fixed guideway system as any light, heavy, or rapid rail system, monorail, inclined plane, funicular, trolley, or automated guideway that is not regulated by the Federal Railroad Administration (FRA) and is included in FTA’s calculation of fixed guideway route miles, or receives funding under FTA’s formula program for urbanized areas, or has submitted documentation to FTA indicating its intent to be included in FTA’s calculation of fixed guideway route miles to receive funding under FTA’s formula program for urbanized areas. Figure 1 shows the types of systems that are included in the SSO program. Figure 1 shows the types of systems that are included in the SSO program. In the SSO program, state oversight agencies are responsible for directly overseeing rail transit agencies. As of December 2009, 27 state oversight agencies exist to oversee rail transit in 26 states. According to FTA, states must designate an agency to perform this oversight function at the time FTA enters into a grant agreement for any “New Starts” project involving a new rail transit system, or before a transit agency applies for FTA formula funding. States have designated several different types of agencies to serve as oversight agencies, including state departments of transportation, public utilities commissions, or regional transportation funding authorities. FTA has a set of rules that an oversight agency must follow, such as developing a program standard that transit agencies must meet, reviewing transit agencies’ safety and security plans, conducting safety audits, and investigating accidents. In the program, rail transit agencies are mainly responsible for meeting the program standards that oversight agencies set out for them, which generally include developing a separate safety and security plan, developing a hazard management process, reporting accidents to oversight agencies within 2 hours, and other similar tasks. Under the program, FTA provides limited funding to oversight agencies in only limited instances, generally for travel or training. While oversight agencies are to include security reviews as part of their responsibilities, TSA also has security oversight authority over transit agencies. (See fig. 2 showing roles and responsibilities of participants in the program.) FTA’s role in overseeing safety and security of rail transit is relatively limited. FTA relies on a staff member in its Office of Safety and Security to lead the SSO program. A program manager is responsible for the SSO program along with other duties. Additional FTA staff within the Office of Safety and Security assist with outreach to transit and oversight agencies and additional tasks. FTA regional personnel are not formally involved with the program’s day-to-day activities, but officials from FTA regional offices help address specific compliance issues that occasionally arise and help states with new transit agencies establish new oversight agencies. FTA also relies on contractors to do many of the day-to-day activities, ranging from developing and implementing FTA’s audit program of state oversight agencies to developing and providing training classes on system safety. Rail transit has been one of the safest modes of transportation in the United States. For example, according to DOT, in 2008, 57.7 people were injured traveling in motor vehicle accidents per 100 million miles traveled and 5.5 people were injured in commuter rail accidents per 100 million miles traveled. For rail transit, the rate was 0.5 people injured per 100 million miles traveled. The injury rate on rail transit has varied from 0.2 to 0.9 injuries per 100 million miles traveled since 2002. Also, the Washington Metro Red Line accident this summer marked the first fatalities involving a collision between two rail cars on a U.S. rail transit system in 8 years. However, according to FTA officials, the recent major incidents in Boston, San Francisco, and Washington have increased their concern about rail transit safety. In addition, FTA states that the number of derailments, worker injuries, and collisions has increased on rail transit systems as a whole in the last several years. Our 2006 report found that officials from the majority of oversight and transit agencies with whom we spoke stated that the SSO program enhances rail transit safety. Officials at several transit agencies cited improvements in reducing the number of derailments, fires, and collisions through actions undertaken as a result of their work with state oversight agencies. However, despite this anecdotal evidence, FTA had not definitively shown that the program had enhanced safety because it had neither established performance goals nor tracked performance. Also, FTA had not audited each state oversight agency in the previous 3 years, as the agency had stated it would. Therefore, FTA had little information with which to track oversight agencies’ performance over time. We recommended that FTA set and monitor performance goals for the SSO program and keep to its stated schedule of auditing state oversight agencies at least once every 3 years. Although FTA officials pointed out that tracking safety performance would be challenging in an environment where fatalities and incidents were low, they agreed to implement our recommendation. FTA assigned the task to a contractor and said that it would make auditing oversight agencies a priority in the future. We also found that FTA faced several challenges in assuring the effectiveness of the program and recommending improvements to transit agency safety practices. Funding challenges limited staffing levels and effectiveness. Officials at several state oversight agencies we spoke with stated that since FTA provided little to no funding for rail transit safety oversight functions, and because of competing priorities for limited state funds, they were limited in the number of staff they could hire and the amount of training they could provide. While FTA requires that states operate safety oversight programs, capital and operating grants are not available to support existing state oversight agencies once passenger service commences. FTA, however, has begun to provide training for state oversight agency staff. With the current financial crises most states are experiencing, states face increasing challenges in providing adequate funding for state oversight agencies. Also, in our 2006 report, we found that 10 state oversight agencies relied on the transit agencies they oversaw for a portion of their budgets. In those cases, the oversight agencies required that the transit agency reimburse the oversight agency for its oversight expenses. Expertise varied across oversight agencies. The level of expertise amongst oversight staff varied widely. For example, we found that 11 oversight agencies had staff with no previous career or educational background in transit safety or security. Conversely, another 11 oversight agencies required their staff to have certain minimum levels of transportation education or experience, such as having 5 years of experience in the safety field or an engineering degree. In the agencies in which oversight officials had little or no experience in the field, officials reported that it took several years before they became confident that they knew enough about rail transit operations to provide effective oversight— a process that new staff would likely have to repeat when the current staff leave their positions. Officials from 18 of the 24 oversight agencies with whom we spoke stated that additional training could be useful in providing more effective safety oversight. FTA, under the current system, does not have the authority to mandate a certain level of training for oversight agency staff. In response to our prior recommendation, FTA has created a recommended training curriculum and is encouraging oversight agency staff to successfully complete the curriculum and receive certification for having done so. Staffing levels varied across oversight agencies. The number of staff that oversight agencies devoted to safety oversight also varied. For example, we found that 13 oversight agencies dedicated less than one full- time equivalent (FTE) staff member to oversight. While in some cases the transit agencies overseen were small, such as a single streetcar line, we found one state that estimated it devoted 0.1 FTE to oversight of a transit agency that averaged 200,000 daily trips. Another state devoted 0.5 FTE to overseeing five different transit systems in two different cities. To help ensure that oversight agency staff were adequately trained for their duties, we recommended that FTA develop a suggested training curriculum for oversight agency staff and encourage those staff to complete it. FTA implemented our recommendation and over 50 percent of state oversight agencies have staff who have completed at least the first tier of this training. Still, the number of staff devoted to safety oversight remains potentially problematic. FTA currently does not require that states devote a certain level of staffing or financial resources to oversight; without additional funding from the federal government or another source, and due to the fiscal difficulties most states are now experiencing, it is unlikely states will independently increase staffing for safety oversight. FTA, however, has asked many SSO agencies to perform formal manpower assessments to ensure they have adequate resources devoted to oversight functions. Enforcement powers of oversight agencies varied. The individual authority each state oversight agency has over transit agencies varies widely. While the SSO program gives state oversight agencies authority to mandate certain rail safety practices, it does not give them authority to take enforcement actions, such as fining an agency or shutting down operations. Some states have given their oversight agencies such authority, however. In our 2006 report, we stated that 19 of 27 oversight agencies had no punitive authority, such as authority to issue fines, and those that did have such authority stated that they rarely, if ever, used it. While taking punitive action against a rail transit agency could be counterproductive (by, for instance, withholding already limited funding), several oversight agency officials told us the threat of such action could potentially make their agencies more effective and other DOT modal administrations with safety oversight authority can level fines or take other punitive action against the entities they oversee. Confusion existed about agency responsibilities for security oversight. Our 2006 report also found that the transit and oversight agencies were confused about the role TSA would take in overseeing security and what role would be left to the state oversight agencies, if any. We made recommendations to TSA and FTA to coordinate their security oversight activities. The agencies agreed and FTA officials reported they are now coordinating their audits with TSA. DOT is planning to propose major changes in FTA’s role that would shift the balance of federal and state responsibilities for setting safety standards for rail transit agencies and overseeing their compliance with those standards. Based on information provided to us by DOT, the department plans to propose a new federal safety program for rail transit, at an unspecified future date, with the following key elements: FTA, through legislation, would receive authority to establish and enforce minimum safety standards for rail transit systems not already regulated by FRA. States could become authorized to enforce the federal minimum safety standards by submitting a program proposal to FTA and receiving approval of their program. In determining whether to approve state safety programs, FTA would consider a state’s capability to undertake rail transit oversight, including staff capacity, and its financial independence from the transit systems it oversees. DOT would provide federal assistance to approved state safety programs. Participating states could set more stringent safety standards if they choose to do so. In states that decide to “opt out” of participation or where DOT has found the program proposals inadequate, FTA would oversee compliance with and enforce federal safety regulations. These changes would give FTA the authority to directly regulate rail transit safety and, in cooperation with the states, to oversee and enforce compliance by rail transit systems with these regulations. These changes would bring its authority more in line with that of other modal administrations within DOT. For example, FRA, Federal Motor Carrier Safety Administration, Federal Aviation Administration, and Pipeline and Hazardous Materials Safety Administration promulgate regulations and technical standards that govern how vehicles or facilities in their respective modes must be operated or constructed. In addition, each of these agencies use federal or state inspectors, or a combination of both, to determine compliance with the safety regulations and guidance they issue. Finally, these agencies can mandate corrective actions and levy fines to transportation operators, among other actions, for noncompliance with regulations. The new program DOT is planning to propose has the potential to address some challenges and issues we cited in our 2006 report. The consideration of staffing levels in deciding whether to approve states’ proposed programs and the provision of funds to approved programs could increase levels of staffing. Requiring that participating states not receive funds from transit agencies would make the state agencies more independent of the transit agencies they oversee. Providing FTA and participating states with the authority to enforce minimum federal safety standards across the nation’s transit systems could help ensure compliance with the standards and improved safety practices, and might prevent some accidents as a result. While the new program, as envisioned by DOT, may have some potential benefits, our work on the SSO program, other transit programs, and regulatory programs suggests there are a number of issues Congress may need to consider in deciding whether or how to act on DOT’s proposal. Roles of the states versus FTA. The following questions would need to be considered when determining whether changes are needed in the balance of federal versus state responsibility for establishing rail transit safety: Are uniform federal standards and nationwide coverage essential to achieving rail transit safety? Which level of government, state or federal, has the capacity to do the job at hand, taking into account such factors as resources and enforcement powers? In addition, shifting federal-state responsibilities for oversight of rail transit safety would bring a number of operational challenges. These include finding the appropriate level of FTA oversight of state programs and allocating costs between the federal government and the states. The new oversight system to be proposed would potentially involve major changes in the way states interact with FTA in overseeing transit safety. The new balance of state and federal responsibilities could take some time for transit agencies to adjust to, especially those that would now be reporting directly to federal officials. Adequate staff with needed skills. FTA would need to ensure it has adequate qualified staff to oversee safety under the new program, especially in states that opt out of participating in the new program. FTA’s current safety staff is very small as is the staff devoted to rail transit safety oversight in most state agencies. Building the capability within FTA, its contractors, and these state agencies to develop and carry out the envisioned program would pose a number of challenges. However, the actions FTA has taken in response to our 2006 recommendation to institute a training curriculum for oversight agency staff, would give it a head start on this process. Enforcement. Congress would need to determine which enforcement mechanisms to authorize FTA to use and FTA would need to develop an enforcement approach that makes the best use of these enforcement mechanisms. Other DOT modal administrations with safety oversight responsibilities, such as the Federal Aviation Administration and FRA, are authorized to issue fines or civil penalties to operators that violate regulations. However, transit agencies are usually publicly owned and face many financial challenges. As a result, fines and penalties could be counterproductive to enhancing safety when funding is at a premium and local riders or taxpayers ultimately could bear the cost of fines. Other enforcement tools are options. For example, FRA may order a locomotive, freight car, or passenger car out of service or may send warning letters to individuals if a safety violation is found, among other enforcement actions. Cost. According to FTA officials, their estimates of the total cost of the new program the department plans to propose are very preliminary. Better estimates of what, if any, costs that states would bear under the new system will also be important before moving forward with this proposal. This could include considering any estimated costs the federal government would incur under various scenarios based on how many states opt out and how many new federal employees or contractors would be required under each scenario to act as trainers, inspectors, and administrative staff. Currently, states bear most of the costs for transit safety oversight. Determining these additional costs would be added as the federal and state governments face significant increasing fiscal pressures. Further, it is uncertain how the program will be paid for. Congress will need to determine if riders, states, those who pay taxes to the Highway Trust Fund, or the Department of the Treasury, or a combination of sources, would bear the cost of this program. In addition to the issues that Congress may need to address, FTA would face some challenges in implementing a new system of transit safety oversight. These include: Variations in the different types of transit. The U.S. rail transit system consists of several different types of vehicles, from heavy and light rail to monorails and funiculars or inclined planes. These vehicles operate on different kinds of track with different power sources and can vary from new modern vehicles to vehicles that are 30 or more years old. Setting federal safety regulations for these varying systems could be a lengthy process and could require multiple parallel rulemakings. Transition to the new system. If the new safety oversight system is approved, it will take some time to transition to the new system. States currently performing safety oversight that opt out in favor of federal oversight will likely need to continue to perform their oversight functions until FTA has additional staff and an enforcement mechanism in place. However, a state may be less likely to replace staff who leave or ensure staff in place stay adequately trained if the state is in the process of giving over its oversight responsibilities to FTA. While the likely effect of this may be minimal, this situation could create the possibility of relaxed oversight during the transition period. As part of our ongoing review of challenges to improving rail transit safety, we will review states’ and FTA’s current efforts to oversee and enhance rail transit safety as well as DOT’s efforts to strengthen the federal role in overseeing rail transit safety. Mr. Chairman, this concludes my prepared statement. I would be pleased to respond to any questions that you or other Members of the Subcommittee might have. For further information on this statement, please contact David J. Wise at (202) 512-2834 or wised@gao.gov. Contact points for our Congressional Relations and Public Affairs offices may be found on the last page of this statement. Individuals making key contributions to this testimony were Catherine Colwell, Judy Guilliams-Tapia, and Raymond Sendejas, Assistant Directors; Timothy Bober; Martha Chow; Antoine Clark; Colin Fallon; Kathleen Gilhooly; David Goldstein; Joah Iannotta; Hannah Laufe; Sara Ann Moessbauer; and Stephanie Purcell. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Rail transit generally has been one of the safest forms of public transportation. However, several recent notable accidents are cause for concern. For example, a July 2009 crash on the Washington Metro Red Line resulted in nine deaths. The federal government does not directly regulate the safety of rail transit. Through its State Safety Oversight program, the Federal Transit Administration (FTA) requires states to designate an oversight agency to directly oversee the safety of rail transit systems. In 2006, GAO issued a report that made recommendations to improve the program. The Department of Transportation (DOT) is planning to propose legislation that, if passed, would result in a greater role for FTA in regulating and overseeing the safety of these systems. This statement (1) summarizes the findings of GAO's 2006 report and (2) provides GAO's preliminary observations on key elements DOT has told us it will include in its legislative proposal for revamping rail transit safety oversight. It is based primarily on GAO's 2006 report, an analysis of key elements of DOT's planned proposal through review of documents and interviews with DOT officials, and GAO's previous work on regulatory programs that oversee safety within other modes of transportation. GAO's 2006 report was based on a survey of the 27 state oversight agencies and transit agencies covered by FTA's program. GAO provided a draft of this testimony to DOT officials and incorporated their comments as appropriate. GAO's 2006 report found that officials from the majority of the state oversight and transit agencies stated that the State Safety Oversight program enhances rail transit safety but that FTA faced several challenges in administering the program. For example, state oversight agencies received little or no funding from FTA and had limited funding for staff. In fact, some required that the transit agencies they oversaw reimburse them for services. Also, expertise, staffing levels, and enforcement powers varied widely from agency to agency. This resulted in a lack of uniformity in how oversight agencies carried out their duties. As of 2006, 13 oversight agencies were devoting the equivalent of less than one full-time employee to oversight functions. Also, 19 oversight agencies GAO contacted lacked certain enforcement authority, such as authority to issue fines, and those that did have such authority stated that they rarely, if ever, used it. DOT is planning to propose major changes in FTA's role that would shift the balance of federal and state responsibilities for oversight of rail transit safety. According to DOT officials, under this proposal, the agency would receive authority to establish and enforce minimum standards although states still could maintain an oversight program. States could become authorized to enforce these standards if FTA determines their program capable and financially independent of the transit system they oversee. FTA would provide financial assistance to approved programs. Such changes would have the potential to address challenges GAO cited in its 2006 report. For example, providing funding to participating state agencies could help them maintain an adequate number of trained staff, and providing FTA and participating states with enforcement authority could help better ensure that transit systems take corrective actions when problems are found. Congress may need to consider several issues in deciding whether or how to act on DOT's proposal. These include determining whatlevel of government has the best capacity to oversee transit safety, ensuring that FTA and state oversight agencies would have adequate and qualified staff to carry out the envisioned program, and understanding the potential budgetary implications of the program.

The diversion and abuse of prescription drugs are associated with incalculable costs to society in terms of addiction, overdose, death, and related criminal activities. DEA has stated that the diversion and abuse of legitimately produced controlled pharmaceuticals constitute a multibillion-dollar illicit market nationwide. One recent example of this growing diversion problem concerns the controlled substance oxycodone, the active ingredient in over 20 prescription drugs, including OxyContin, Percocet, and Percodan. OxyContin is the number one prescribed narcotic medication for treating moderate-to-severe pain in the United States. Currently, a single 20-milligram OxyContin tablet legally selling for about $2 can be sold for as much as $25 on the illicit market in some parts of Kentucky. Combating the illegal diversion of prescription drugs while ensuring that the pharmaceuticals remain available for those with legitimate medical need involves the efforts of both federal and state government agencies. The Controlled Substances Act of 1970 provides the legal framework for the federal government’s oversight of transactions involving the sale and distribution of controlled substances at the manufacturer and wholesale distributor levels. The states address these issues through their regulation of the practice of medicine and pharmacy. The Controlled Substances Act established a classification structure for drugs and chemicals used in the manufacture of drugs that are designated as controlled substances. Controlled substances are classified by DEA into five schedules on the basis of their medicinal value, potential for abuse, and safety or dependence liability. Schedule I drugs—including heroin, marijuana, and hallucinogens such as LSD and PCP—have a high potential for abuse and no currently accepted medical use. Schedule II drugs—including methylphenidate (Ritalin) and opiates such as hydrocodone, morphine, and oxycodone—have a high potential for abuse among drugs with an accepted medical use and may lead to severe psychological and physical dependence. Drugs on schedules III through V have accepted medical uses and successively lower potentials for abuse and dependence. Schedule III drugs include anabolic steroids, codeine, hydrocodone in combination with aspirin or acetaminophen, and some barbiturates. Schedule IV contains such drugs as the antianxiety medications diazepam (Valium) and alprazolam (Xanax). Schedule V includes preparations such as cough syrups with codeine. All scheduled drugs except those in schedule I are legally available to the public with a prescription. Under the act, DEA provides legitimate handlers of controlled substances—including manufacturers, distributors, hospitals, pharmacies, practitioners, and researchers—with registration numbers, which are used in all transactions involving controlled substances. Registrants must comply with a series of regulatory requirements relating to drug security and accountability through the maintenance of inventories and records. Although all registrants, including pharmacies, are required to maintain records of controlled substance transactions, only manufacturers and distributors are required to report their transactions involving schedule II drugs and schedule III narcotics, including sales to the retail level, to DEA. The data provided to DEA are available for use in monitoring the distribution of controlled substances throughout the United States, in identifying retail-level registrants that received unusual quantities of controlled substances, and in investigations of illegal diversions at the manufacturer and wholesaler levels. Although data are reported to DEA regarding purchases by pharmacies, the act does not require the reporting of dispensing information by pharmacies at the patient level to DEA. State laws govern the prescribing and dispensing of prescription drugs by licensed health care professionals. State medical practice laws generally delegate the responsibility of regulating physicians to state medical boards, which license physicians and grant them prescribing privileges. In addition, state medical boards investigate complaints and impose sanctions for violations of the state medical practice laws. States regulate the practice of pharmacy based on state pharmacy practice acts and regulations enforced by the state boards of pharmacy. The state boards of pharmacy are also responsible for ensuring that pharmacists and pharmacies comply with applicable state and federal laws and for investigating and disciplining those that fail to comply. According to the National Association of Boards of Pharmacy, all state pharmacy laws require that records of prescription drugs dispensed to patients be maintained and that state pharmacy boards have access to the prescription records. State prescription drug monitoring programs varied in their objectives and operation. While all programs were intended to help law enforcement identify and prevent prescription drug diversion, some programs also included education objectives to provide information to physicians, pharmacies, and the public. Program operation also varied across states, in terms of which drugs were covered and how prescription information was collected. Which agency, such as a pharmacy board or public health department, was given responsibility for the program also varied across states. Additionally, methods for analyzing the data to detect potential diversion activity differed among state programs. State monitoring programs are intended to facilitate the collection, analysis, and reporting of information on the prescribing, dispensing, and use of prescription drugs within a state. The first state monitoring program was established in California in 1940, and the number of programs has grown slowly. We reported that the number of states with programs has grown from 10 in 1992 to 15 in 2002; the number of programs stands at 16 in 2004. We found that state programs varied in their objectives. All states used monitoring programs primarily to assist law enforcement in detecting and preventing drug diversion, and but some also used the programs for educational purposes. Programs assisted law enforcement authorities both by providing information in response to requests for assistance on specific investigations and by referring matters to law enforcement officials when evaluations of program data revealed atypical prescribing or dispensing patterns that suggested possible illegal diversion. The programs evaluated prescribing patterns to identify medical providers who may have been overprescribing and inform them that their patterns were unusual. They also identified patients who may have been abusing or diverting prescription drugs and provided this information to practitioners. For example, the programs in Nevada and Utah sent letters to physicians containing patient information that could signal potential diversion activity, including the number and types of drugs prescribed to the patient during a given time period and the pharmacies that dispensed the drugs. Monitoring programs have also been used to educate physicians, pharmacies, and the public about the existence and extent of diversion, diversion scams, the drugs most likely to be diverted by individuals, and ways to prevent drug diversion. Monitoring programs also differed in operational factors, some of which have cost implications. These factors included the choice of controlled substance schedules monitored, approaches to analyzing and using data, computer programming choices, number and type of staff and contractors, turnaround times and report transmittal methods, and number and type of requests for information. State programs varied in the controlled substances they covered, in part because of differences in available resources and other state-specific factors such as level of drug abuse. Two of the states we studied— Kentucky and Utah—covered schedules II through V. These states’ program officials told us that covering those schedules allowed them flexibility to respond if drugs on other schedules became targets for diversion. Most experts agree that covering all controlled substance schedules prevents drug diverters from avoiding detection by bypassing schedule II drugs and switching to drugs in other schedules. States used different approaches to analyze the prescription information they received. A few states used a proactive approach, routinely analyzing prescription data collected by the programs to identify individuals, physicians, or pharmacies that had unusual use, prescribing, or dispensing patterns that could suggest potential drug diversion, abuse, or doctor shopping. Trend analyses were shared with appropriate entities, such as law enforcement, practitioners, and regulatory and licensing boards. In contrast, most state programs generally used the prescription data in a reactive manner to respond to requests for information. These requests may have come from physicians or from law enforcement or state officials based on leads about potential instances of diversion. According to state program officials, most programs operated in a reactive fashion because of the increased amount of resources required to operate a proactive system. Some state programs had electronic reporting systems, while others were paper-based. If data are reported electronically, there are ongoing computer maintenance and programming choices and their attendant costs. Similarly, some state programs engaged private contractors to collect and maintain the data, while others did so in-house. If a private contractor collects the raw data from dispensers and converts them to a standardized format, the program pays annual contracting costs for database maintenance. Kentucky and Nevada privately contracted with the same company to collect data for their program databases. Utah, in contrast, collected and maintained drug dispensing data in-house, using its own software and hardware. The number and type of staff a state chose to operate its monitoring program also varied. In 2002, Kentucky’s program employed four full-time and four part-time staff to help ensure the accuracy of its reports, including a pharmacist-investigator who reviewed each report before it was sent. Nevada’s program operated with one employee because a private contractor collected the data. In contrast, in 2002 Utah’s program, with three full-time employees and no private contractor, had one program administrator who collected all dispensing data, converted them to a standardized format for monitoring, and maintained the database. The two other staff answered requests. If the program seeks to provide more timely responses to report requests, such as same-day responses, the costs involved in returning the response to the requester may increase. For example, in 2001 Kentucky spent up to $12,000 in 1 month for faxing reports. Monitoring program officials from Kentucky, Nevada, and Utah told us in 2002 that they estimated 3- to 4- hour turnaround times for program data requests, and all mainly used faxing, rather than more costly mailing, to send reports to requesters. Same-day responses may be preferable for physicians who want the prescription drug history for a patient being seen that day and for law enforcement users who need immediate data for investigations of suspected illegal activity. As users become more familiar with the benefits of monitoring program report data, requests for information and other demands on the programs may increase. In Kentucky, Nevada, and Utah, use had increased substantially, mostly because of an increase in the number of requests by physicians to check patients’ prescription drug histories. In Kentucky, these physician requests increased from 28,307 in 2000, the first full year of operation, to 56,367 in 2001, an increase of nearly 100 percent. Law enforcement requests increased from 4,567 in 2000 to 5,797 in 2001, an increase of 27 percent. Similarly, Nevada’s requests from all authorized users also increased—from 480 in 1997, its first full year, to 6,896 in 2001, an increase of about 1,300 percent. Additionally, as drug marketing practices change and monitoring programs mature, the operational needs may shift as well. For example, states face new challenges with the advent of Internet pharmacies, because they enable pharmacies and physicians to anonymously reach across state borders to prescribe, sell, and dispense prescription drugs without complying with state requirements. In addition, if users want program reports to reflect more timely information, dispensing entities would have to report their data at the time of sale, rather than submitting data biweekly or monthly, to capture the most recent prescription dispensing. If users want to be alerted if a certain drug, practitioner, or pharmacy may be involved in a developing diversion problem, programs would have to initiate periodic data analysis to determine trends or patterns. Such program enhancements would entail additional costs, however, including costs for computer programming, and data analysis. States that are considering establishing or expanding a monitoring program face a variety of other challenges. One challenge is the lack of awareness of the extent to which prescription drug abuse and diversion is a significant public health and law enforcement problem. States also face concerns about the confidentiality of the information gathered by the program, voiced by patients who are legitimately using prescription drugs and by physicians and pharmacists who are legitimately prescribing and dispensing them. Another challenge states face is securing adequate funding to initiate and develop the program and to maintain and modify it over time. We found that states with monitoring programs have experienced considerable reductions in the time and effort required by law enforcement and regulatory investigators to explore leads and the merits of possible drug diversion cases. We also found that the presence of a monitoring program in a state may help reduce illegal drug diversion there, but that diversion activities may increase in contiguous states without programs. The ability of the programs to focus law enforcement and regulatory investigators who are working on suspected drug diversion cases on specific physicians, pharmacies, and patients who may be involved in the alleged activities is crucial to shortened investigation time and improvements in productivity. States that do not have programs must rely on tips from patients, practitioners, or law enforcement authorities to identify possible prescription drug abuse and diversion. Following up on these leads requires a lengthy, labor-intensive investigation. In contrast, the programs can provide information that allows investigators to pinpoint the physicians’ offices and pharmacies where drug records must be reviewed to verify suspected diversion and thus can eliminate the need to search records at physicians’ offices and pharmacies that have no connection to a case. In each of the three states we studied, state monitoring programs led to reductions in investigation times. For example, prior to implementation of Kentucky’s monitoring program, its state drug control investigators took an average of 156 days to complete the investigation of alleged doctor shoppers. Following the implementation, the average investigation time dropped to 16 days, or a 90 percent reduction in investigation time. Similarly, Nevada reduced its investigation time from about 120 days to about 20 days, a reduction of 83 percent, and a Utah official told us that it experienced an 80 percent reduction in investigation time. Officials from Kentucky, Nevada, and Utah told us in 2002 that their programs may have helped reduce the unwarranted prescribing and subsequent diversion of abused drugs in their states. In both Kentucky and Nevada, an increased number of program reports were being used by physicians to check the prescription drug use histories of current and prospective patients when deciding whether to prescribe certain drugs that are subject to abuse. Law enforcement officials told us that they view these drug history checks as initial deterrents— a front-line defense—to prevent individuals from visiting multiple physicians to obtain prescriptions, because patients are aware that physicians can review their prescription drug history. For an individual who may be seeking multiple controlled substance prescriptions, the check allows a physician to analyze the prescription drug history to determine whether drug treatment appears questionable, and if so, to verify it with the listed physicians. In Kentucky, a physician could request a drug history report on the same day as the patient’s appointment, and usually received the report within 4 hours of the request. In 2002, Kentucky’s program typically received about 400 physician requests daily, and provided data current to the most recent 2 to 4 weeks. The presence of a monitoring program may also have an impact on the prescribing of drugs more likely to be diverted. For example, DEA ranked all states for 2000 by the number of OxyContin prescriptions per 100,000 people. Eight of the 10 states with the highest numbers of prescriptions— West Virginia, Alaska, Delaware, New Hampshire, Florida, Pennsylvania, Maine, and Connecticut—had no monitoring programs, and only 2 did— Kentucky and Rhode Island. Six of the 10 states with the lowest numbers of prescriptions—Michigan, New Mexico, Texas, New York, Illinois, and California—had programs, and 4—Kansas, Minnesota, Iowa, and South Dakota—did not. Another indication of the effectiveness of a monitoring program is that its existence in one state appears to increase drug diversion activities in contiguous states without programs. When states begin to monitor drugs, drug diversion activities tend to spill across boundaries to states without programs. One example is provided by Kentucky, which shares a boundary with seven states, only two of which had programs in 2002—Indiana and Illinois. As drug diverters became aware of the Kentucky program’s ability to trace their drug histories, they tended to move their diversion activities to nearby nonmonitored states. OxyContin diversion problems worsened in Tennessee, West Virginia, and Virginia—all contiguous states without programs—because of the presence of Kentucky’s program, according to a 2001 joint federal, state, and local drug diversion report. Although monitoring programs can enhance the ability of states to detect and deter illegal diversion of prescription drugs, the number of states with such programs has grown only slightly over the past 12 years from 10 in 1992 to 16 in 2004. A lack of awareness of the magnitude of the problem; concerns about confidentiality on the part of patients, physicians, pharmacists, and legislators; and difficulty in accessing funding have kept the numbers of monitoring programs low. Cooperative efforts at the state and national levels are seeking to overcome these challenges and increase the number of states with programs. Mr. Chairman, this concludes my prepared statement. I would be pleased to respond to any questions you or other Members of the Subcommittee may have. For more information regarding this testimony, please contact Marcia Crosse at (202) 512-7119. Individuals making key contributions to this testimony include Martin T. Gahart, Roseanne Price, and Opal Winebrenner. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

The increasing diversion of prescription drugs for illegal purposes or abuse is a disturbing trend in the nation's battle against drug abuse. Diversion can include such activities as prescription forgery and "doctor shopping" by individuals who visit numerous physicians to obtain multiple prescriptions. The most frequently diverted prescription drugs are controlled substances that are prone to abuse, addiction, and dependence, such as hydrocodone (the active ingredient in Lortab and many other drugs) and oxycodone (the active ingredient in OxyContin and many other drugs). Some states use prescription drug monitoring programs to control illegal diversion of prescription drugs that are controlled substances. GAO was asked to examine (1) how state monitoring programs compare in terms of their objectives and operation and (2) the impact of state monitoring programs on illegal diversion of prescription drugs. This testimony is based on GAO's report, Prescription Drugs: State Monitoring Programs Provide Useful Tool to Reduce Diversion, GAO-02-634 (May 17, 2002). In that report, the programs in Kentucky, Utah, and Nevada were selected for more in-depth study because they were the most recently established programs at the time. GAO found that the 15 state monitoring programs in place in 2002 differed in their objectives and operation. The programs were intended to facilitate the collection, analysis, and reporting of information about the prescribing, dispensing, and use of controlled substances. They provided data and analysis to state law enforcement and regulatory agencies to assist in identifying and investigating activities potentially related to illegal drug diversion. The programs could be used by physicians to check a patient's prescription drug history to determine if the individual was doctor shopping to seek multiple controlled substances. Some programs also offered educational programs for the public, physicians, and pharmacists regarding the nature and extent of the problem and medical treatment options for abusers of diverted drugs. The programs varied primarily in terms of the specific drugs they covered and the type of state agency in which they were housed. Some programs covered only those prescription drugs that are most prone to abuse and addiction, whereas others provided more extensive coverage. In addition, most programs were administered by a state law enforcement agency, a state department of health, or a state board of pharmacy. GAO also found that state monitoring programs may have realized benefits in their efforts to reduce drug diversion. These included improving the timeliness of law enforcement and regulatory investigations. Each of the three states studied reduced its investigation time by at least 80 percent. In addition, law enforcement officials told GAO that they view the programs as a deterrent to doctor shopping, because potential diverters are aware that any physician from whom they seek a prescription may first examine their prescription drug utilization histories based on monitoring program data. For example, as drug diverters became aware of Kentucky's ability to trace their drug histories, they tended to move their diversion activities to nearby nonmonitored states.

FEMA’s workforce consists of employees hired under both Title 5 and the Stafford Act, as well as individuals from two other workforce components who are not FEMA employees, but who can be deployed for disaster response: the DHS Surge Capacity Force and FEMA Corps. Title 5 employees are permanent and temporary employees who form FEMA’s day-to-day workforce, and can be deployed as needed during a disaster. Stafford Act employees include CORE and Reservists hired specifically to support disaster-related activities on a temporary or intermittent basis. COREs are temporary employees with 2- to 4-year appointments, while Reservists work on an intermittent basis and are deployed as needed. Surge Capacity Force volunteers include employees of other DHS components who augment FEMA’s workforce in the event of a catastrophic disaster. Their first and only deployment as of May 2017 was to assist in response and recovery efforts after Hurricane Sandy in 2012. FEMA manages and coordinates the Surge Capacity Force program. FEMA Corps is a national service program managed by AmeriCorps National Civilian Community Corps (NCCC). FEMA Corps members support disaster response and recovery efforts and work under the direction of FEMA staff. As shown in table 1 below, FEMA’s available disaster response workforce, including Surge Capacity Force and FEMA Corps totaled over 22,600 in March 2017. FEMA’s workforce categories are subject to different employee rights and legal protections. For example, Title 5 employees are generally afforded notice and appeal rights, while Reservists serve in temporary intermittent positions and can be terminated at any time with or without cause. In 2012, FEMA issued a policy directive outlining procedures for reporting misconduct within the agency, conducting administrative investigations, and reporting actions taken in cases where the misconduct allegation was sustained. The FEMA Administrative Investigations Policy directive (FD 123-19) lays out misconduct responsibilities for employees, managers, and supervisors, and for the personnel management offices involved in managing misconduct. The directive also describes three types of administrative investigations: Managerial Inquiry: managers or supervisors may be assigned to conduct an inquiry when the allegation is not complex (e.g., tardiness, absence without leave) and involves a minimal number of witnesses. Office of the Chief Security Officer (OCSO) Investigation: OCSO personnel are assigned to conduct investigations of allegations that are criminal in nature, or have the potential to be criminal in nature, for example travel card fraud or a physical altercation. Independent Investigation: senior FEMA or other government officials may be assigned to conduct investigations into complex allegations that involve multiple offices and witnesses, senior employees, or prohibited personnel practices such as political coercion. The process for conducting investigations is further discussed in FEMA’s Administrative Investigations Policy manual (FEMA Manual 123-19-1). FEMA’s Employee Discipline Manual (FEMA Manual 255-3-1) describes the policies, procedures, and responsibilities for taking conduct-related discipline against permanent Title 5 employees. The manual discusses options for disciplinary actions, which include reprimands and suspensions of 14 days or less, and adverse actions, which include suspensions for 15 days or more, demotions, and removals. The manual also lays out employee grievance and appeal rights. The policies and procedures described in the manual apply solely to permanent Title 5 employees. As discussed later in this report, the discipline and appeals process varies for other workforce categories. Within FEMA, three primary offices are involved in reviewing, investigating, and adjudicating employee misconduct allegations. OCSO Internal Investigations Branch (IIB): conducts investigations related to more serious allegations, such as those that may involve potential criminal misconduct. The Office of the Chief Component Human Capital Officer (OCCHCO) Labor and Employee Relations Branch (LER): advises supervisors who conduct lower level investigations and inquiries, such as time and attendance violations, and provides recommendations on any counseling or any disciplinary or adverse action for all cases. The Office of the Chief Counsel (OCC) Personnel Law Branch (PLB): provides legal advice during investigations and conducts legal reviews of certain reports of investigation and all disciplinary and adverse actions. Representatives from these three offices form FEMA’s Administrative Investigations Directive (AID) Committee, which reviews misconduct allegations, assigns investigators, and tracks the status of open cases. DHS OIG also plays a role in reviewing and investigating certain misconduct allegations. DHS Management Directive 0801.1 requires officials from all DHS components, including FEMA, to refer certain categories of misconduct to DHS OIG for review, such as allegations of criminal misconduct against a DHS employee and any allegations of misconduct against senior employees. After reviewing the allegation, DHS OIG may elect to initiate an investigation or refer the case back to the component, such as FEMA, for review. DHS OIG receives complaints (from employees, supervisors, the public, and agency referrals) against employees in all DHS components through the DHS OIG Hotline, which is a resource for reporting corruption, fraud, waste, abuse, mismanagement, or misconduct. Table 2 below shows the average number of misconduct complaints received through the DHS OIG Hotline from fiscal year 2014 through 2016 by selected DHS component. These complaints represent alleged offenses only, and not final actions or adjudication results. FEMA receives allegations of employee misconduct from individuals both within and outside the agency. Individuals (including members of the public) can report these allegations through a number of different mechanisms, including, but not limited to, FEMA’s OCSO Tipline, LER specialists, and the DHS OIG Hotline. Employees can also report misconduct to their supervisors and union representatives. The AID Committee reviews most allegations, including those declined by DHS OIG and returned to FEMA. LER reviews complaints received directly from managers or employees to determine if they need to be discussed at the AID Committee meetings. Allegations may then be assigned to one of the three types of administrative investigations discussed earlier. Once completed, reports of investigation are forwarded to LER for review. LER recommends appropriate disciplinary or adverse action, if warranted. Figure 1 below outlines the general steps in FEMA’s misconduct process, including DHS OIG’s role in reviewing and investigating allegations. Variations in potential case outcomes and the appeals process by workforce category are discussed later in this report. FEMA has developed a policy outlining procedures regarding investigations of misconduct as documented in FEMA’s Administrative Investigations Policy directive. The directive applies to all FEMA personnel. While FEMA Corps and Surge Capacity Force members are not FEMA employees, OCSO officials stated that the investigations process is the same regardless of the workforce category. FEMA has also documented misconduct policies and procedures regarding options to address misconduct and appeals for Title 5 and CORE employees. For Title 5 employees the options to address misconduct include both disciplinary and adverse actions. FEMA Corps disciplinary policies and procedures are the responsibility of AmeriCorps NCCC and are documented in their member handbook. Figure 2 outlines the options to address misconduct and appeal rights for Title 5, CORE, and FEMA Corps as documented in their respective employee discipline and program manuals. According to LER officials, as of February 2017, most supervisors had been offered training on FEMA’s 2015 Employee Discipline Manual, which applies to Title 5 employees. Three more training sessions are planned for 2017. Additionally, PLB partnered with LER in 2016 to offer misconduct and documentation training for all supervisors. In 2016, PLB and LER completed 21 joint trainings. They held two more joint trainings in March 2017. FEMA has not documented misconduct policies and procedures for Surge Capacity Force members. DHS issued the Surge Capacity Force Concept of Operations in 2010, which outlines FEMA’s base implementation plan for the Surge Capacity Force. However, the document does not address any elements pertaining to Surge Capacity Force human capital management, specifically misconduct and disciplinary policies and procedures. According to the FEMA Surge Capacity Force Coordinator, despite the lack of documentation, any incidents of misconduct would likely be investigated by FEMA’s OCSO, which would then refer the completed report of investigation to the employee’s home component for adjudication and potential disciplinary action. Additionally, OCSO officials said that while they were unaware of any misconduct investigations involving Surge Capacity Force members to date, if an incident were to occur they would follow the procedures in the Administrative Investigations Policy directive and also notify the member’s home component of the incident. However, although no allegations of misconduct were made at the time, the FCO in charge of one of the Hurricane Sandy Joint Field Offices said he had not seen anything in writing or any formal guidance that documents or explains how the process would work and stated that he would have had to contact FEMA headquarters for assistance in determining how to address any misconduct. Furthermore, he noted that taking the time to figure out the Surge Capacity Force misconduct process would have detracted from time spent on the Joint Field Office’s mission. LER officials stated that they would be able to walk through the procedures with supervisors and managers if asked. However, because these procedures are not documented, the process is not transparent and information is not readily available. LER officials noted that while LER specialists can quickly answer questions, it would be beneficial for the procedures to be documented. A PLB official also told us that FEMA should have a written policy for potential Surge Capacity Force misconduct. Standards for Internal Control in the Federal Government advises management to develop and maintain documentation of its internal control system. This documentation can assist in management’s ability to communicate controls to personnel, and is a means to retain organizational knowledge as well as communicating that knowledge to external parties. Without documented guidance, FEMA cannot ensure that Surge Capacity Force misconduct is addressed adequately in a timely and comprehensive manner, which could negatively affect the extent to which Joint Field Offices can accomplish their mission after a disaster. Furthermore, FEMA’s 2015 update to the Surge Capacity Force Concept of Operations calls for increasing the size of the Surge Capacity Force workforce from its current size of approximately 6,000 volunteers to 36,000 volunteers and also does not address any human capital management issues, including misconduct. This planned expansion further underscores the importance of documenting policies and procedures in order to address potential misconduct. FEMA’s Reservist Program Manual lacks documented policies and procedures on disciplinary options to address misconduct and appeal rights for Reservists. Both LER and PLB officials told us that, in practice, disciplinary actions for Reservists are limited to reprimands and termination. According to these officials, FEMA does not suspend Reservists because they are an intermittent, at-will workforce deployed as needed to respond to disasters. FCOs and cadre managers have the authority to demobilize Reservists and remove them from a Joint Field Office if misconduct occurs, which may be done in lieu of suspension. Furthermore, LER and PLB officials also told us that, in practice, FEMA grants Reservists the right to appeal a reprimand or termination to their second-level supervisor. Although officials stated that FEMA is carrying out these policies and procedures in practice, the actions are not documented in the Reservist Program Manual. Standards for Internal Control in the Federal Government advises management to document policies for each unit in its area of responsibility. Each unit, with guidance from management, determines the policies necessary to operate, as well as documents policies in the appropriate level of detail to allow management to effectively monitor activity. According to FEMA officials, because of the at-will nature of their employment, Reservists are not subject to the same policies, procedures, and appeal rights as Title 5 employees, such as a suspension as an option to address misconduct. However, the procedures that are executed in practice are not reflected in the Reservist Program Manual. Without documented Reservist disciplinary options and appeals policies, supervisors and Reservist employees may not be aware of all aspects of the disciplinary and appeals process. Additionally, seven supervisors and managers in two of the three regions and one of the three cadres we spoke with noted that there is a perception of inconsistency and unfairness in the discipline process. For example, one manager told us that if cadre management provided more information on outcomes in addressing misconduct, it would help improve the perception that misconduct is being seriously addressed at FEMA. Clear documentation of the Reservist disciplinary options and appeals policies and procedures currently in practice would help to address the concerns of inconsistency. FEMA revised its employee disciplinary manual for Title 5 employees in 2015, and in doing so, eliminated the agency’s table of offenses and penalties. Tables of offenses and penalties are used by agencies to provide guidance on the range of penalties available when formal discipline is taken. They also provide awareness and inform employees of the penalties which may be imposed for misconduct. Since revising the manual and removing the table, FEMA no longer communicates possible punishable offenses to its entire workforce. Instead, information is now communicated to supervisors and employees on an individual basis. Specifically, LER specialists currently use a “comparators” spreadsheet with historical data on previous misconduct cases to determine a range of disciplinary or adverse actions for each specific misconduct case. The information used to determine the range of penalties is shared with the supervisor on a case-by-case basis; however LER specialists noted that due to privacy protections they are the only FEMA officials who have access to the comparators spreadsheet. PLB and LER officials stated that the new comparators spreadsheet is an improvement over the old table, which contained overly broad categories and had not been updated since 1981. According to officials, the comparators spreadsheet is easier to use and thus it is easier to ensure cases are consistent across the agency. Supervisors and managers we spoke with shared their perspectives on how offenses and penalties are communicated both through the prior table and the new comparators spreadsheet. Specifically, 11 supervisors and managers in all three regions and cadres we spoke with, as well as both union representatives we interviewed, cited the benefits of a table in communicating punishable offenses and the range of penalties. These benefits included transparency, consistency, and possible deterrence against engaging in misconduct. For example, one supervisor noted that a table of offenses and penalties is beneficial for both employees and supervisors because it removes ambiguity and makes the disciplinary process more transparent. Additionally, as discussed earlier, other supervisors and managers told us that there is a perception that misconduct cases are handled inconsistently. However, management officials in one region, and one supervisor in another region, noted they preferred the new case-by-case comparators system and did not see benefits to having a table of offenses and penalties. Another manager noted that a table was useful in certain cases with specific rules, such as travel card misuse; however, many cases are more complicated and in those situations a strict table is less effective. The remaining supervisors and managers did not offer their perspective on this topic. Standards for Internal Control in the Federal Government advises management to consider standards of conduct, assigned responsibility, and delegated authority when establishing expectations. Management establishes expectations of competence for key roles as well as for all personnel through policies within internal control systems. Because information about offenses and penalties is not universally shared with supervisors and employees, FEMA management are limited in their ability to set expectations about appropriate conduct in the workplace and to communicate consequences of inappropriate conduct. Additionally, a FEMA OCCHCO official noted that FEMA is considering re-introducing a table which would inform employees and supervisors of misconduct offenses and penalties. Communicating information about offenses and common ranges of penalties, such as in a summary table that does not include individual case information, could help to provide transparency for employees and their supervisors on the range of penalties to expect for different types of misconduct and mitigate the perception that misconduct is handled inconsistently across FEMA’s workforce. The three offices on the AID Committee involved in investigating and adjudicating employee misconduct complaints each maintain separate case tracking spreadsheets with data on employee misconduct to facilitate their respective roles in the process. OCSO collects data in a case tracking spreadsheet about employee misconduct complaints and investigations. For example, the spreadsheet contains fields with narrative descriptions of alleged offenses and investigation updates, the locations and FEMA regions where alleged offenses occurred, and the number of days an investigation was open. According to agency officials, all of the cases discussed during weekly AID Committee meetings are in the OCSO case tracking spreadsheet. LER also collects data on employee misconduct complaints in a case tracking spreadsheet. Specifically, LER records information on lower- level allegations that do not rise to the level of AID Committee review as well as adjudication information for allegations which were investigated by OCSO. In addition to employee misconduct complaints, LER specialists record all inquiries from supervisors in the LER case tracking spreadsheet, including questions about performance, grievances, and employee counseling. Fields in the LER case tracking spreadsheet include, for example, narrative descriptions of alleged offenses and case summaries, comments with case status updates, and the disciplinary or adverse actions taken, if any. PLB collects data on misconduct-related disciplinary or adverse actions it reviews in a case tracking spreadsheet. Fields in the PLB spreadsheet include narrative descriptions of the charge or issue, the employee position and type, and the disciplinary or adverse action taken. According to PLB officials, the data are based on information provided by LER. We analyzed data provided by OCSO in its case tracking spreadsheet and found that there were 595 complaints from January 2014 through September 30, 2016. The complaints involved alleged offenses of employee misconduct which may or may not have been substantiated over the course of an investigation. Some complaints involved multiple allegations of various offenses against multiple subjects. In order to better summarize the number and type of alleged offenses, we developed eight general categories (see table 3 below). Based on our analysis, the 595 complaints contained approximately 799 alleged offenses from January 2014 through September 30, 2016. As shown in figure 3 below, the most common type of alleged offenses were integrity and ethics violations (278), inappropriate comments and conduct (140), and misuse of government property or funds (119). For example, one complaint categorized as integrity and ethics involved allegations that a FEMA employee at a Joint Field Office was accepting illegal gifts from a FEMA contractor and a state contractor. Another complaint categorized as inappropriate comments and conduct involved allegations that a FEMA employee’s supervisor and other employees had bullied and cursed at them, creating an unhealthy work environment. Finally, a complaint categorized as misuse of government property or funds involved allegations that a former FEMA employee was terminated but did not return a FEMA-owned laptop. LER officials provided summary data on employee misconduct outcomes separately from the LER case tracking spreadsheet. According to the summary data, there were 546 disciplinary or adverse actions related to employee misconduct taken during the last three calendar years — from January 2014 through December 2016. Of those actions, the most common were removals or terminations (354), reprimands (115), suspensions of 14 days or less (64), and suspensions of 15 days or more (6). The most common employee types affected by actions were Reservist (235) and Title 5 (110). We also analyzed data provided by PLB in its case tracking spreadsheet and found that PLB reviewed 454 final disciplinary or adverse actions related to employee misconduct from January 2014 through September 30, 2016. The most common types of action were reprimands (144), suspensions ranging from 1 to 45 days (121), and terminations (118). The most common employee types affected by actions were Title 5 (265), CORE (180), and Reservist (6). There are several potential explanations for the differences between the LER and PLB data on employee misconduct outcomes. A PLB official stated that PLB does not review all disciplinary actions against Reservists; therefore, LER Reservist actions may not appear in the PLB spreadsheet. Further, LER officials stated that they recently started adjusting their recordkeeping and it was possible that some actions were not input by LER specialists into the case tracking spreadsheet. Differences in the data may also be related to limitations we identified in the spreadsheets, as discussed below. OCSO, LER, and PLB collect data on employee misconduct and outcomes, but limited standardization of fields and entries within fields, limited use of unique case identifiers, and a lack of documented guidance on data entry restricts their usefulness for identifying and addressing trends in employee misconduct. We found that there was limited standardization of fields based on our review of OCSO, LER, and PLB case tracking spreadsheets. For example, we attempted to summarize misconduct allegations by employee type using the OCSO spreadsheet. Out of approximately 704 subjects named in complaints from January 2014 through September 30, 2016, the most common employee type was unknown because either the information was not consistently available, or not enough information was available in the case tracking spreadsheet. In 2014 and 2016, the OCSO spreadsheet sometimes included information on the subject’s employee type embedded within narrative summary of allegation fields. In 2015, that information was sometimes included in the narrative fields and sometimes listed in a subject type field unique to that year. There was also limited standardization of entries within fields. For example, in all years the offense field in the OCSO spreadsheet consisted of text entries that described similar offenses in different ways, such as “Travel Card Violation” and “Travel Policy Violation”, or multiple offenses related to a specific complaint, such as “Fraud / Travel Card Violation”. The LER spreadsheet also contained examples of fields with limited standardization of entries. For example, the fields which described a subject’s pay grade consisted of varied text entries, such as “9”, “GS-09”, and “GS-9”. The PLB spreadsheet had similar issues with other fields. Limited standardization of fields and entries within fields restricts the usefulness of the data for identifying and addressing trends in employee misconduct because it makes timely evaluation, summarization, and verification of the data more difficult. We found that there was limited use of unique case identifiers in the OCSO, PLB, and LER case tracking spreadsheets. The OCSO spreadsheet contained several unique case identifier fields, including an IIB case number, OIG case number, and incident number. However, the LER and PLB spreadsheets provided to GAO did not contain these unique case identifier fields. A PLB official confirmed that the LER and PLB case tracking spreadsheets do not include case numbers as identifiers, only subject names. OCSO officials also stated that LER does not use the OCSO IIB case number or the OIG case number, although LER does have access to the OCSO IIB case number from AID Committee meetings. As a result, the spreadsheets do not share the same unique identifier, which makes it more difficult to track the status and outcome of cases across the three case tracking spreadsheets. OCSO, LER, and PLB officials stated that they do not have data dictionaries or documented guidance on data entry for their respective case tracking spreadsheets. A PLB official reported that data are entered into the PLB spreadsheet by a paralegal and spot-checked by a senior PLB official. LER officials noted that they did not provide data dictionaries or guidance because the original case tracker involved a simple Excel spreadsheet rather than a more complicated database system. However, a lack of documented guidance on data entry may make it more difficult to maintain and verify the completeness and accuracy of the data. For example, we identified variations in data entry into the LER case tracking spreadsheet across LER specialists. In 2015, of the 9 LER specialists who entered information into the case tracking spreadsheet, 6 specialists did not enter information into the employee pay grade field, 2 specialists sometimes entered information, and 1 specialist always entered information. Standards for Internal Control in the Federal Government advises management to process data into quality information which is appropriate, current, complete, accurate, accessible, and provided on a timely basis. Additionally, management should evaluate processed information, make revisions when necessary so that the information is quality information, and use the information to make informed decisions. As described above, FEMA employee misconduct data are not readily accessible and cannot be verified as accurate and complete on a timely basis. These limitations restrict management’s ability to process the data into quality information which can be used to identify and address trends in employee misconduct. For example, an OCSO official stated that senior OCSO officials recently requested employee misconduct information based on employee type, such as the number of Reservists. However, the data are largely captured in narrative fields, making it difficult to extract without manual review. LER officials stated that LER specialists do not regularly conduct trend analysis on employee misconduct cases, but if they notice an apparent trend such as a number of cases involving misuse and non- payment of travel credit cards they will discuss it during quarterly meetings with component management. Five supervisors and management officials from each of the three regions we spoke with said that information on trends and patterns in employee misconduct would be useful. For example, supervisors could use trend analysis to identify specific types of employee misconduct which have become more common, allowing them to send out policy guidance or schedule targeted trainings to help address the issue. LER and OCSO are taking steps to improve their case tracking spreadsheets. Specifically, LER officials reported that they began using Microsoft SharePoint software in January 2017 which includes drop-down selections for all fields except the subject name and comments. The officials stated that they switched to SharePoint in response to DHS requests for specific misconduct information as well as our requests, which raised their awareness of the LER spreadsheet’s limitations. Similarly, an OCSO official stated that OCSO planned to add new fields to their spreadsheet which will make it easier to generate reports and show trends. These are positive steps towards addressing the issues with employee misconduct data quality identified above. However, it is not clear that they will be sufficient to address each of the data limitations we identified and improve the ability to conduct trend analysis. For example, although the LER SharePoint spreadsheet does include an entry number field, it was unclear whether that would allow officials to track the status and outcomes of specific cases across all three case tracking spreadsheets. New fields and increased use of drop-down selections in the LER SharePoint spreadsheet may help improve standardization of fields and entries within fields, but LER specialists may differ in how they interpret fields without documented guidance on data entry. Additionally, it is unclear whether OCSO’s modifications will include increased use of drop-down selections or other means to standardize entries. An OCSO official stated that they are also exploring DHS OIG’s database software, and noted that, if it could be adapted for FEMA’s purposes, it would be a substantial improvement. There are a number of possible quality control measures that could be implemented to help improve the usefulness of FEMA’s data for use in identifying and addressing trends in employee misconduct. At a minimum, based on our analysis of FEMA’s misconduct spreadsheets, the offices could: add additional drop-down fields with standardized entries to make fields easier to summarize; add unique case identifier fields to improve the ability to track cases across the three program offices’ case tracking spreadsheets; develop guidance documents to ensure standardized data entry within each office, including a procedure for quality control checks; or consider adopting the use of database software, which could improve standardization and case tracking across offices. These actions, combined with routine reporting on misconduct trends, could improve upon the initial steps already underway at LER and OCSO and better enable FEMA to manage and report on misconduct information. In accordance with FEMA’s Administrative Investigations Policy directive, officials from OCSO, LER, and PLB conduct weekly AID Committee meetings to coordinate information on misconduct allegations and investigations. The committee reviews allegations, refers cases for investigation or inquiry, and discusses the status of investigations. According to PLB officials, FEMA’s process for addressing employee misconduct was ad hoc and informal prior to the release of the directive in 2012. For example, officials from key offices did not always meet and share information on a regular basis, as they do now. PLB officials noted that a new FEMA Chief Counsel in 2010 drove changes calling for a formalized directive and clear instructions for managing misconduct. In addition to the weekly AID Committee meetings, LER and PLB officials stated that they meet on a regular basis to discuss disciplinary and adverse actions and ensure that any penalties are consistent and defensible in court. Employee misconduct information is also shared directly with FEMA’s Chief Security Officer and Chief Counsel. For example, OCSO provides the Chief Security Officer with a monthly report of all open investigations. Additionally, PLB provides the Chief Counsel with a monthly report of significant employee litigation, and includes in that report a list of all significant investigations, such as those involving high-level employees as the subject or witness in a case. Within FEMA, these regular meetings and status reports provide officials from key personnel management offices opportunities to communicate and share information about employee misconduct. FEMA also provides DHS OIG with information on employee misconduct cases on a regular basis. DHS Management Directive 0810.1 requires that all DHS organizational elements, including FEMA, provide monthly reports to DHS OIG on all open investigations. OCSO complies with this requirement and sends a monthly report that includes updates on investigative activity to DHS OIG. OCSO officials stated that they also provide limited outcome information on a case-by-case basis if DHS OIG specifically requests this information. According to OCSO and DHS OIG officials, DHS OIG does not regularly provide FEMA with updates on ongoing investigations it conducts. However, OCSO officials stated they ask DHS OIG for status updates on these cases frequently, especially if the employee was placed on administrative leave during the investigation. Our review indicates that OCSO has not established effective procedures to ensure that all cases referred to FEMA by DHS OIG are accounted for and subsequently reviewed and addressed. As discussed earlier, FEMA is required to refer certain misconduct allegations to DHS OIG for review before taking any action. OCSO officials told us they follow the guidelines closely and refer all serious misconduct allegations to DHS OIG for initial review. They also noted that DHS OIG declines to investigate and refers most cases back to FEMA for action. As also discussed above, OCSO sends a monthly report of open investigations to DHS OIG. However, while these reports provide awareness of specific investigations, according to OCSO officials, neither office reconciles the reports to a list of referred cases to ensure that all cases are accounted for. We reviewed a non-generalizable random sample of employee misconduct complaints DHS OIG referred to FEMA for review and found that FEMA did not adequately track all referred complaints and therefore could not ensure that all complaints in the sample we selected were reviewed and addressed at the time of our inquiry. Specifically, we tracked a random sample of 20 fiscal year 2016 employee misconduct complaints DHS OIG declined to investigate and referred to FEMA for action. We found that FEMA missed 6 of the 20 complaints during the referral process and had not reviewed them at the time of our inquiry. DHS OIG referred 3 of the 6 complaints to FEMA a year or more prior to our review. As a result of our review, FEMA subsequently took action to review the complaints. The AID Committee recommended that OCSO open inquiries in 3 of the 6 cases to determine whether the allegations were against FEMA employees, assigned 2 cases to LER for further review, and closed 1 case for lack of information. According to an OCSO official, OCSO subsequently determined that none of the allegations in the 3 cases they opened involved FEMA employees and the cases were closed. The remaining 2 cases were open as of April 2017. The results from our sample cannot be generalized to the entire population of referrals from DHS OIG to FEMA; however, they raise questions as to whether there could be additional instances of misconduct complaints that FEMA has not reviewed or addressed. FEMA OCSO officials offered several possible explanations for why the complaints we identified were missed. According to these officials, they sometimes receive large batches of complaints through the FEMA Tipline – some of which may be duplicative. DHS OIG referrals also come in through the Tipline and officials said that it is possible that they may not always accurately identify and record all of the referrals for the AID Committee since some of them are misclassified and some do not involve FEMA employees. Standards for Internal Control in the Federal Government advises management to perform ongoing monitoring – including comparisons and reconciliations or other routine actions. Such activities, either undertaken internally within OCSO or in coordination with DHS OIG, could improve the process. A senior OCSO official agreed that reconciliation procedures would help ensure that all complaints referred by DHS OIG to FEMA are accounted for and noted that FEMA is working with DHS OIG to improve reporting processes and case reconciliation. Employee misconduct can detract from FEMA’s mission and negatively impact public perceptions of the agency, particularly when associated with disaster response efforts. Given the broad scope of FEMA’s mission and the growth and different categories that make up its workforce, an effective process is key to mitigating any negative employee misconduct effects. FEMA has taken actions to manage the employee misconduct investigation and adjudication process. Specifically, FEMA has developed and documented misconduct policies and procedures for most of its employees, and has established procedures for regular internal communication and coordination, as well as information sharing with DHS OIG. However, misconduct policies and procedures for Surge Capacity Force members and outcome options and appeals policies and procedures for Reservists are not documented, and FEMA does not communicate the range of penalties for offenses to all employees. Clear documentation establishing who is responsible for investigating and adjudicating misconduct is especially important given FEMA’s goal of significantly expanding the Surge Capacity Force. Because the Reservist disciplinary action options and appeals currently in practice are not documented, FEMA supervisors and Reservist employees may not be aware of all aspects of the process. Similarly, the lack of communication on the range of penalties for specific offenses to FEMA’s workforce limits management’s ability to set expectations about appropriate conduct in the workplace and to communicate consequences of inappropriate conduct. Clearly documented policies and procedures for all workforce categories and communication about offenses and penalties could help to better prepare management to address misconduct and to mitigate any perceptions that misconduct is handled inconsistently across FEMA’s workforce. In addition, while several FEMA offices collect data on employee misconduct allegations, investigations, and outcomes, limitations related to how the data are collected and managed restrict their usefulness for identifying and addressing trends in employee misconduct. Addressing these limitations by implementing quality control measures could improve FEMA’s ability to track misconduct cases and to identify potential problem areas and opportunities for targeted training. Moreover, developing reconciliation procedures to track cases referred from DHS OIG to FEMA could help reduce the risk that FEMA does not address all misconduct complaints. In order to improve employee misconduct policies and procedures, the Secretary of Homeland Security should direct the FEMA Administrator to take the following three actions: document policies and procedures to address potential Surge document Reservist disciplinary options and appeals policies and procedures that are currently in practice at the agency; and communicate the range of penalties for specific misconduct offenses to all employees and supervisors. In order to better identify and address trends in employee misconduct, the Secretary of Homeland Security should direct the FEMA Administrator to take the following actions: improve the quality and usefulness of the misconduct data it collects by implementing quality control measures, such as adding additional drop-down fields with standardized entries, adding unique case identifier fields, developing documented guidance for data entry, or considering the adoption of database software; and once the quality of the data is improved, conduct routine reporting on employee misconduct trends. In order to ensure that all allegations of employee misconduct referred by DHS OIG are reviewed and addressed, the Secretary of Homeland Security should direct the FEMA Administrator to develop reconciliation procedures to consistently track referred cases. We provided a draft of this report to DHS and FEMA for review and comment. DHS provided written comments which are reproduced in appendix II. In its comments, DHS concurred with our recommendations and described actions planned to address them. FEMA and DHS OIG also provided technical comments, which we incorporated as appropriate. With regard to our first recommendation, that FEMA document policies and procedures to address potential Surge Capacity Force misconduct, DHS stated that FEMA is developing a Human Capital plan for the Surge Capacity Force and will include policies and procedures relating to potential misconduct. DHS noted that these policies and procedures will take into account FEMA’s limited authorities over Surge Capacity personnel who are not FEMA employees. DHS estimated that this effort would be completed by June 30, 2018. This action, if fully implemented, should address the intent of the recommendation. With regard to our second recommendation, that FEMA document Reservist policies and procedures related to disciplinary options and appeals currently in practice at the agency, DHS stated that FEMA will update FEMA Directive 010-06, FEMA Reservist Program, to include procedures for disciplinary actions and appeals current in practice at the agency. DHS estimated that this effort would be completed by December 31, 2017. This action, if fully implemented, should address the intent of the recommendation. With regard to our third recommendation, that FEMA communicate the range of penalties for specific misconduct offenses to all employees and supervisors, DHS stated that FEMA’s OCCHCO is currently drafting a table of offenses and penalties and will take steps to communicate those penalties to employees throughout the agency once the table is finalized. DHS estimated that this effort would be completed by December 31, 2017. This action, if fully implemented, should address the intent of the recommendation. With regard to our fourth and fifth recommendations, that FEMA improve the quality and usefulness of its misconduct data by implementing quality control measures, and, once the quality of the data is improved, conduct routine reporting on employee misconduct trends, DHS stated that FEMA’s OCCHCO is working with the DHS OIG to develop a new case management system. The system will use drop-down fields with standardized entries and provide tools for trend analysis. Once the new system is implemented, DHS stated that FEMA will be able to routinely identify and address emerging trends of misconduct. DHS estimated that these efforts would be completed by March 31, 2018. These actions, if fully implemented, should address the intent of the recommendations. With regard to our sixth recommendation, for FEMA to develop reconciliation procedures to consistently track referred cases, DHS stated that once the new case management system described above is established and fully operational, FEMA will be able to upload all DHS OIG referrals into a single, agency-wide database. Additionally, FEMA will work with DHS OIG to establish processes and procedures that will improve reconciliation of case data. DHS estimated that these efforts would be completed by March 31, 2018. These actions, if fully implemented, should address the intent of the recommendation. We are sending copies of this report to the Secretary of Homeland Security and interested congressional committees. If you or your staff have any questions about this report, please contact me at (404) 679-1875 or CurrieC@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. The objectives of this report were to determine (1) the extent to which the Federal Emergency Management Agency (FEMA) has developed policies and procedures for addressing employee misconduct; (2) what data are available on FEMA employee misconduct cases and their outcomes, and the extent to which FEMA uses these data to identify and address trends in employee misconduct; and (3) the extent to which information regarding misconduct cases is shared within FEMA’s personnel management offices and with the Department of Homeland Security Office of the Inspector General (DHS OIG). To address objective one, we reviewed, where available, FEMA’s policies and procedures for reporting, investigating, and adjudicating allegations of misconduct across all of the agency’s workforce categories, including: Title 5 employees, Cadre of On-Call Response/Recovery Employees (CORE), Reservists, Surge Capacity Force members, and FEMA Corps members. Specifically, we reviewed investigation and discipline directives and manuals, as well as program directives and manuals, and interagency agreements related to FEMA Corps, to determine the extent to which FEMA has developed misconduct policies and procedures. At FEMA headquarters we interviewed senior officials from the Office of the Chief Security Officer (OCSO), Office of the Chief Component Human Capital Officer, Labor and Employee Relations Branch (LER), and Office of the Chief Counsel, Personnel Law Branch (PLB). Additionally, we interviewed the Surge Capacity Force and FEMA Corps Coordinators, as well as AmeriCorps officials, to discuss Surge Capacity Force and FEMA Corps misconduct policies and procedures. We interviewed cadre management officials from FEMA’s three largest cadres - Public Assistance, Individual Assistance, and Logistics - to discuss Reservist misconduct policies and procedures. We also spoke with union representatives from two FEMA bargaining units to gain their perspective on how misconduct policies and procedures are implemented for Title 5 and CORE employees. Additionally, we interviewed a non-probability sample of supervisors and managers in three FEMA regions. Specifically, we interviewed three FEMA Regional Administrators or their designated representatives; five Federal Coordinating Officers (FCO), and 11 Public Assistance, Individual Assistance, and Logistics Branch Chiefs from FEMA Regions 2,4, and 6. We selected these three regions based on factors such as geographic dispersion and regions that typically respond to different types of disasters, as well as those with the highest number of misconduct allegations reported to OCSO from January 2014 through September 30, 2016. The results of these interviews are not generalizable to all 10 of FEMA’s regions or all cadres; however, they provided us with both regional and Joint Field Office perspectives on policies and procedures for addressing employee misconduct and supervisory misconduct training. To address objective two, we reviewed and analyzed misconduct case tracking spreadsheets maintained by OCSO, LER, and PLB. We were unable to identify and account for possible duplication of the same complaints in the different spreadsheets. As such, we focused our analysis on individual spreadsheets rather than on aggregating information across all spreadsheets. Specifically, in order to summarize alleged misconduct offenses recorded in the OCSO case tracking spreadsheet, we first reviewed all of the data available on complaints which were received from January 1, 2014 through September 30, 2016. Next, we created eight offense category definitions based on prior GAO reports, discussions with stakeholders, and our review of the summary of allegations and offense fields. Finally, we assigned alleged offenses from the OCSO spreadsheet to the categories. We also attempted to review and summarize the subjects included in each complaint by employee type. However, our ability to do so was limited because that information was not consistently included in the OCSO spreadsheet. In order to summarize final disciplinary and adverse actions reviewed by PLB and recorded in the PLB case tracking spreadsheet, we reviewed all of the data available from January 2, 2014 through September 28, 2016. Next, we selected actions which were categorized as related to employee misconduct and excluded proposed actions. Finally, we summarized final actions by the action type and employee type using existing spreadsheet fields. We also reviewed information on past employee misconduct outcomes and disciplinary actions provided by LER. To assess the reliability of these data, we reviewed the three spreadsheets for any gaps and inconsistencies. We also interviewed agency officials from each office about how data are entered into the spreadsheets, who enters the data, whether they have guidance documents for data entry, and the process of assigning complaints to the offices. We identified limitations which we discuss in the report, but overall found the data in the spreadsheets sufficiently reliable to provide some general information on the nature and characteristics of employee misconduct complaints. In order to determine the extent to which FEMA currently uses these data to identify and address trends in employee misconduct, we interviewed OCSO, LER, and PLB officials. Additionally, we included related questions during the interviews with supervisors and managers in FEMA Regions 2, 4, and 6. Although information obtained from these interviews, as mentioned above, is not generalizable to all 10 regions or all FEMA cadres, it provided insights into how employee misconduct information is shared with FEMA field supervisors. To address objective three, we first reviewed FEMA and DHS OIG employee misconduct directives and manuals to identify any requirements for coordination among internal FEMA offices and with DHS OIG. Next, we interviewed senior OCSO, LER, PLB, and DHS OIG officials to determine the extent to which they coordinate and communicate misconduct information and to obtain their perspectives on information sharing related to misconduct investigations and outcomes. We analyzed documents, including case tracking spreadsheets and reports, to determine what and how frequently employee misconduct information is shared within FEMA and with DHS OIG. Finally, we used fiscal year 2016 data from DHS OIG’s Enforcement Data System to randomly select a non-generalizable sample of 20 FEMA employee misconduct complaints which were referred from DHS OIG to FEMA. We compared these complaints to the OSCO case tracking spreadsheet provided by FEMA. For complaints we were unable to locate, we requested that OCSO provide us with their statuses, including whether or not OCSO had taken action to review or investigate the complaints. Additionally, we asked OCSO to coordinate with LER and PLB and obtain information on the outcomes of all 20 complaints. While not generalizable to all complaints referred from DHS OIG to FEMA, the results of our review provided insight about FEMA’s procedures for tracking referred complaints and ensuring that all allegations of misconduct are addressed. For all three objectives, we reviewed the documents and information we gathered and evaluated them against Standards for Internal Control in the Federal Government. We conducted this performance audit from June 2016 to July 2017 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the contact named above, Ben Atwater (Assistant Director), Sarah Turpin (Analyst-in-Charge), David Alexander, Dominick Dale, Eric Hauswirth, Rianna Jansen, Stephen Komadina, Kristiana D. Moore, and Heidi Nielson made key contributions to this report.

FEMA is responsible for coordinating government-wide efforts in preparing for, responding to, and recovering from natural or man-made disasters, including acts of terror. The agency relies on permanent and disaster-related temporary employees and has a total workforce of over 22,000. Employee misconduct incidents can detract from FEMA's mission, damage the agency's reputation, and hamper its ability to respond to disasters and maintain public trust. GAO was asked to review employee misconduct at FEMA. This report examines: (1) the extent to which FEMA developed policies and procedures for addressing misconduct; (2) available data on FEMA misconduct cases and the extent to which FEMA uses the data to identify and address trends; and (3) the extent that misconduct cases are shared within FEMA and with DHS OIG. GAO reviewed FEMA procedures, analyzed misconduct data, and interviewed officials from FEMA HQ and three regions (selected based on geographic dispersion and number of misconduct allegations). GAO also analyzed a random, non-generalizable sample of 20 complaints referred from DHS OIG to FEMA to determine whether they were addressed. The Federal Emergency Management Agency (FEMA) has developed and documented misconduct policies and procedures for most employees, but not its entire workforce. Specifically, FEMA has developed policies and procedures regarding misconduct investigations that apply to all FEMA personnel and has also documented options to address misconduct and appeal rights for Title 5 (generally permanent employees) and Cadre of On-Call Response/Recovery Employees (temporary employees who support disaster related activities). However, FEMA has not documented misconduct policies and procedures for Surge Capacity Force members, who may augment FEMA's workforce in the event of a catastrophic disaster. Additionally, FEMA's Reservist (intermittent disaster employees) policies and procedures do not outline disciplinary actions or the appeals process currently in practice at the agency. As a result, supervisors and Reservist employees may not be aware of all aspects of the process. Clearly documented policies and procedures for all workforce categories could help to better prepare management to address misconduct and mitigate perceptions that misconduct is handled inconsistently. FEMA records data on misconduct cases and their outcomes; however, aspects of this data limit their usefulness for identifying and addressing trends. GAO reviewed misconduct complaints recorded by FEMA's Office of the Chief Security Officer (OCSO) from January 2014 through September 30, 2016, and identified 595 complaints involving 799 alleged offenses, the most common of which were integrity and ethics violations. FEMA reported 546 disciplinary actions related to misconduct from calendar year 2014 through 2016. In addition to OCSO, two other FEMA offices involved in investigating and adjudicating misconduct also record data. However, limited standardization of data fields and entries within fields, limited use of unique case identifiers, and a lack of documented guidance on data entry across all three offices restricts the data's usefulness for identifying and addressing trends in employee misconduct. Improved quality control measures could help the agency use the data to better identify potential problem areas and opportunities for training. FEMA shares misconduct case information internally and with the Department of Homeland Security Office of Inspector General (DHS OIG) on a regular basis; however, FEMA does not have reconciliation procedures in place to track DHS OIG referred cases to ensure that they are reviewed and addressed. GAO reviewed a random sample of 20 cases DHS OIG referred to FEMA in fiscal year 2016 and found that FEMA missed 6 of the 20 complaints during the referral process and had not reviewed them at the time of GAO's inquiry. As a result of GAO's review, FEMA took action to review the complaints and opened inquiries in 5 of the 6 cases (1 case was closed for lack of information). In 3 of these cases, officials determined that the complaints did not involve FEMA employees. The 2 remaining cases were open as of April 2017. While the results from this review are not generalizable to the entire population of referrals from DHS OIG to FEMA, they raise questions as to whether there could be additional instances of misconduct complaints that FEMA has not reviewed or addressed. Procedures to ensure reconciliation of referred cases across FEMA and DHS OIG records could help ensure that FEMA accounts for all complaints. GAO is making six recommendations with which DHS concurred, including that FEMA document policies and procedures for addressing Surge Capacity Force and Reservist misconduct, improve the quality and usefulness of its misconduct data, and develop reconciliation procedures to consistently track referred cases.

CMS and states share responsibilities for overseeing QHPs offered through the exchanges. Specifically, CMS is responsible for establishing minimum QHP certification standards that all QHPs must meet in order to participate in any exchange. Federal regulations require that all exchanges have procedures to certify QHPs annually to ensure compliance with federal requirements. To be certified as a QHP, a plan must meet certain minimum federal requirements, including those related to, for example, the coverage of certain benefits and limits on cost- sharing. In FFE states, CMS is responsible for overseeing compliance with these requirements; in states operating SBEs, the states are responsible for ensuring that plans comply. CMS is responsible for conducting oversight and monitoring of QHPs offered on the FFE, and also requires all SBEs to develop an oversight and monitoring program. In addition to meeting federal exchange-specific requirements, QHP issuers must also abide by state-specific insurance regulations that apply to all issuers offering health insurance products, as states are the primary regulators of health insurance. Specifically, all QHPs, whether offered on the FFE or SBE, must be offered by a health insurance issuer that is licensed and in good standing to offer insurance coverage in each state in which it offers QHPs. As a result, QHP issuers are subject to oversight by the states in which they offer QHPs. As part of this oversight, state departments of insurance manage complaint hotlines where enrollees can notify agencies of concerns related to any health insurance plan sold in that state, including QHPs offered on the FFE or an SBE. All exchanges are required to carry out certain consumer assistance functions. Specifically, CMS requires exchanges to operate a toll-free call center and website to address the needs of consumers and enrollees requesting assistance and to conduct outreach and educational activities. For example, CMS operates a Marketplace Call Center to assist the needs of consumers in states that utilize the FFE. In addition, all exchanges are required to have a “navigator” program to carry out public education activities, help consumers select a QHP, and offer QHP enrollees with assistance after their enrollment, among other things. CMS awards grants to organizations to serve as navigators for the FFE. All exchanges may also implement other “assister” programs that perform many of the same or similar functions as navigators. Navigators and other assisters are collectively referred to as “assisters.” Individuals purchasing coverage through the exchanges may be eligible to receive financial assistance to offset the cost of such coverage, and, according to CMS, over three-fourths of QHP enrollees obtain at least one form of such assistance. Eligibility for financial assistance is based on income and provided in the form of premium tax credits and cost-sharing subsidies. One form of assistance is the premium tax credit, which is generally available to income-eligible individuals who do not have access to health insurance that meets certain standards. The credit is designed to reduce an eligible individual’s cost of purchasing health insurance through the exchange and can be paid to an enrollee’s issuer in advance to reduce the enrollee’s monthly premium costs. The amount of the premium tax credit varies and is designed to provide larger credit amounts to those with lower incomes. QHP enrollees who qualify for and opt to receive advance payments of the premium tax credit based on their income and family size at the time of application must attest that they will file a federal tax return for the applicable plan year. Such enrollees must reconcile on their federal tax return the amount of advance payments received based on their actual reported income and family size for the year. Enrollees who qualify for premium tax credits may also be eligible to receive cost- sharing reduction assistance to help offset QHP enrollees’ out-of-pocket expenses, including by lowering their deductibles, coinsurance and co- payments. Available data from the five national surveys we identified through our literature review show that most QHP enrollees were satisfied overall with the plans they obtained through the exchanges. QHP enrollees have also expressed satisfaction, to varying degrees, with specific aspects of their plans, including their coverage, their choice of providers, and plan affordability, according to five national surveys we reviewed. QHP enrollees who obtained their coverage through the exchanges have reported overall satisfaction with their plans from 2014 through 2016, according to national surveys that we reviewed. Specifically, 65 percent or more of QHP enrollees surveyed expressed overall satisfaction with their plans in 2014 through 2016, according to three national surveys that asked this question of enrollees. (See table 1.) The overall satisfaction level of QHP enrollees was somewhat lower than or similar to those who were enrolled in employer-sponsored health insurance in 2015 and 2016. Specifically, one national survey reported that 86 percent of QHP enrollees were satisfied or somewhat satisfied with their current health plan in 2015, compared to 93 percent of employer-sponsored health insurance enrollees. The survey did not specifically report the reason for the difference in satisfaction levels. The same survey in 2016 reported that overall plan satisfaction among QHP enrollees was equivalent to those with employer-sponsored health insurance, although other national surveys that we reviewed reported that QHP enrollees were, for example, less satisfied with their choice of providers or less likely to report ease in affording their premiums, compared to those with employer-sponsored insurance. Two of the national surveys we reviewed examined QHP re-enrollment, which can provide additional context for QHP enrollees’ overall satisfaction with their plans. Specifically, one national survey found that 77 percent of adult QHP enrollees that re-enrolled in the same plan for 2016 reported satisfaction with their QHP. Another national survey of QHP enrollees reported that, in 2015, most—82 percent—of those re- enrolling selected a plan with the same insurance company and about half stayed with the same plan. Stakeholders we interviewed and other literature we reviewed also provided additional context for QHP enrollees’ overall satisfaction with their plans based on QHP re-enrollment and a stable volume of consumer complaints regarding health insurance. For example, while many factors, including financial incentives, may affect an enrollee’s decision to re-enroll in a plan, officials from one exchange office we interviewed told us that they consider plan re-enrollment as one important measure of QHP enrollee satisfaction because enrollees have the capacity to change QHPs annually. In addition, a statewide survey of Vermont QHP enrollees found that 9 percent of QHP enrollees renewing their coverage in 2015 switched plans, with 80 percent of renewing QHP enrollees reporting that their plan fit their needs very or somewhat well. Although the remaining four selected states in our review had not directly measured QHP enrollee satisfaction, officials from all of these states’ departments of insurance told us that QHP enrollees in their states have not reported significant problems that are unique to QHPs. In addition, department of insurance officials from two of these selected states told us that the volume of complaints they received for all health plans had not increased since health insurance became available through the exchanges in 2014. QHP enrollees who obtained their coverage through the exchanges have rated their health insurance coverage positively and generally expressed satisfaction with their choice of providers, according to national surveys we reviewed. Specifically, two national surveys reported that approximately 70 percent of QHP enrollees rated their health insurance coverage as good, very good, or excellent in 2016. QHP enrollees have also generally reported satisfaction with their choice of providers, according to four national surveys we reviewed. For example, one national survey reported that 74 percent of QHP enrollees noted satisfaction with their choice of primary care doctor in 2016, and a smaller portion of enrollees—59 percent—noted satisfaction with their choice of specialists. Two other national surveys reported that more than 75 percent of QHP enrollees surveyed were satisfied with the doctors included under their plan. The fourth national survey reported levels of dissatisfaction, stating that 14 percent of QHP enrollees reported dissatisfaction with their choice of doctors and other providers. (See table 2.) In addition to reporting satisfaction with their choice of providers, most QHP enrollees surveyed had used their health insurance coverage in 2015 or 2016, according to four national surveys that reported this information. For example, one national survey reported that about two- thirds of QHP enrollees reported using their plans to access care or purchase medication. Three national surveys found that over half of QHP enrollees reported having a regular or routine check-up. Another national survey reported that 62 percent of QHP enrollees who needed to see a specialist could do so within 2 weeks or less. Satisfaction with plan affordability among QHP enrollees who obtained their coverage through the exchanges was lower than for satisfaction with plans overall and for coverage and access. Nevertheless, about half or more of QHP enrollees surveyed reported satisfaction with their plan’s affordability, according to the five national surveys that we reviewed. For example, two national surveys reported that about half or more of QHP enrollees found it easy to afford their plan’s premium costs. One national survey found that 45 percent of QHP enrollees reported high levels of confidence in their ability to obtain affordable care. Another national survey reported rates of dissatisfaction, with 25 percent of QHP enrollees reporting being very or somewhat dissatisfied with the premiums they paid for their plans. Finally, one national survey reported that approximately 60 percent of QHP enrollees were satisfied with various plan costs in 2016, such as annual deductibles and copayment amounts. (See table 3). In addition to these five national surveys, other studies from our literature review reported similar data regarding QHP enrollees’ satisfaction with plan affordability. Specifically, one narrowly focused study reported that 87 percent of QHP enrollees surveyed found their coverage to be affordable on the basis of their monthly budget. Another study reported that many community stakeholders interviewed—including assisters, provider representatives, and department of insurance officials—stated that QHP enrollees could obtain care more easily and affordably than they could prior to the advent of the exchanges. Despite general satisfaction with plan affordability, one national survey of enrollees reported in 2016 that their satisfaction with certain plan costs had declined since 2014. Another national survey reported that price was a common reason why enrollees were dissatisfied with their QHP in 2015. Enrollee dissatisfaction with premium amounts prompts some to drop their coverage, according to experts and assisters we interviewed. Three of the national surveys we reviewed also reported that dissatisfaction with plan costs is a primary reason why QHP enrollees switch plans. HHS recently reported that those who switched plans for 2016 generally moved to lower-cost plans. Although available data show most QHP enrollees were satisfied overall with their plans, our interviews with stakeholders—including experts, assisters, state department of insurance and exchange officials—and our review of literature, also revealed concerns about some QHP enrollees’ ability to afford and access their care, and understand their QHP, among other things. Some enrollees have concerns about affording care before reaching their deductible, according to experts we interviewed and our review of literature. Specifically, some individuals have reported concerns affording care, or have been deterred from seeking care, because they found it too expensive to pay for their out-of-pocket expenses before reaching their deductibles, according to experts we interviewed. Two national surveys of QHP enrollees found that over a quarter of them had experienced financial difficulties paying for their out-of-pocket health care expenses in the prior year, with some enrollees reporting unmet health care needs due to cost. One national survey reported that in 2016, 25 percent of QHP enrollees reported higher-than-expected out-of-pocket costs after using their coverage. Cost is a driving factor in QHP enrollees’ selection of a plan. According to three national surveys of QHP enrollees, premiums, deductibles, and copayments were the top factors that consumers used when selecting a QHP in 2015. In addition, an HHS analysis of QHP selection in the FFE reported that enrollees tended to select QHPs with the lowest premiums among those offering similar levels of coverage. Indeed, high-deductible health plans remain popular options among QHP enrollees, potentially because these plans tend to have lower premiums. One national survey reported that in 2016, 46 percent of QHP enrollees chose a plan with a high deductible. While many consumers believe at the time of their enrollment that their QHP will be affordable, some enrollees become overwhelmed after seeking care when trying to balance the need to pay out-of-pocket costs in addition to monthly premiums and other life expenses, according to experts we interviewed. Two factors that may contribute to QHP enrollees’ dissatisfaction with plan affordability is that many QHP enrollees have lower incomes and have been previously uninsured; as such, these individuals may not have previously had to pay for their health care expenses or balance the need to pay for them along with other life expenses. While some QHP enrollees perceive their premium and cost-sharing amounts to be unaffordable, most have received federal subsidies that were designed to help make their coverage more affordable. Specifically, CMS reported that 84 percent of QHP enrollees were receiving advance payments of the premium tax credit, and 56 percent of QHP enrollees were receiving cost-sharing reduction assistance to help offset their out- of-pocket expenses, as of December 2015. According to one national survey, about 60 percent of QHP enrollees paid either nothing or less than $125 per month in premiums in 2015 and 2016—amounts reported as comparable to those for employer-sponsored coverage. Some QHP enrollees who obtained their coverage through the exchanges have faced problems accessing care after enrollment due to both midyear changes in QHP provider networks and the unavailability of accurate information about provider networks and formularies at the time of enrollment, according to experts we interviewed and our review of literature. For example, one report noted widespread confusion among consumers and providers about which providers were included in a plan’s network. In addition, a 2015 survey of assisters found that half of the assisters had encountered enrollees who sought help because their provider was not in-network. A report examining state regulation of QHPs found that in 2014, only a minority of states enforced rules about frequency in updating provider directories. Two recent studies of the accuracy of provider directories for QHPs offered in Maryland and Washington, D.C. found that about half of the psychiatrists listed in the provider directories could no longer be reached at the phone numbers directories listed. Officials from CMS and state departments of insurance, as well as other stakeholders, also told us that enrollees have faced challenges verifying their coverage or otherwise communicating with the issuer before receiving their insurance cards, which can result in treatment delays. Concerns have been expressed both by some experts and in literature we reviewed about QHP enrollees’ ability to obtain or continue care given the increased prevalence of QHPs with narrow networks. Issuers have increasingly begun to offer narrow network plans as a mechanism to lower premiums; these plans offer coverage for services through a smaller group of physicians or hospitals than the plan has covered in the past. For example, a narrow network plan may only offer in-network coverage through one local hospital. One analysis reported that QHPs offered on the exchanges included 34 percent fewer providers, on average, than plans offered outside the exchanges. Another report identified 16 states where at least half of all QHPs offered had narrow networks. While stakeholders have expressed concerns with these plans, consumers continue to enroll in them and indicate they are willing to choose a plan with a narrow network to reduce their premiums. For example, one national survey of QHP enrollees reported that over forty percent of those with the option for a narrow network plan in 2016 enrolled in such a plan. Another national survey found that nearly 60 percent of QHP enrollees said in 2015 that they would be willing or somewhat willing to accept a smaller network of hospitals or doctors in exchange for lower overall health care payments. Some QHP enrollees who obtained their coverage through the exchanges have faced difficulties understanding how to use their plans, according to our interviews with stakeholders and our review of literature. Specifically, about half of nationwide QHP enrollees surveyed in 2015 had a good understanding of their plan benefits and total health coverage costs at the time of enrollment, according to one national survey. One factor that may contribute to enrollees’ difficulty in understanding their plans is that because many of those who have obtained coverage through the exchanges were previously uninsured, they may be unaccustomed to health insurance terminology—words such as premiums, coinsurance, deductibles, and out-of-pocket maximums—as well as health insurance practices such as navigating plan networks and formularies. For example, a 2015 survey of assisters reported that about three-quarters of assisters noted that most or nearly all consumers who shopped for or enrolled in a QHP needed help understanding basic health insurance concepts such as deductibles and in-network services. Another study reported that assisters spent considerable time helping QHP enrollees understand how to use their plan, including by explaining key insurance terms, provider networks, the financial risks of using out-of-network care, and the use of appropriate care settings. For example, some QHP enrollees who were previously uninsured did not realize that they should no longer use a hospital emergency room as their primary care location. In addition to facing difficulties understanding general health insurance concepts, some QHP enrollees have found it challenging to understand exchange-specific terminology, according to our review of the literature. For example, a 2015 survey of Vermont QHP enrollees found that less than one-third of enrollees fully understood exchange-specific terms, such as advanced premium tax credit and cost-sharing reduction. Furthermore, according to experts and assisters we interviewed, some enrollees also face language barriers, which can compound their difficulty in understanding how to use their QHP. Some assisters we interviewed told us that some enrollees take time off from work in order to travel to their offices for help translating and understanding notices they receive from CMS and issuers. To varying degrees, QHP enrollees who obtained their coverage through the exchanges have also faced a range of other challenges related to their health insurance plans, according to assisters and state department of insurance and exchange officials we interviewed and literature we reviewed. For example, some assisters told us about difficulties that enrollees have faced in updating information with CMS, including modifying income information and adding family members to plans. An assister also told us that enrollees have faced difficulties obtaining information from CMS during the appeals process—for example, information about the status of appeals in progress, or the rationale for appeal decisions. In addition, officials from state departments of insurance, exchange offices, and assisters told us about other state- specific challenges that enrollees have faced. According to Vermont exchange officials and assisters we interviewed, QHP enrollees in that state had faced some challenges related to billing. For example, enrollees had received incorrect premium statements, bills for premiums that were already paid but not recognized by the system, or incorrect medical bills for services received, according to assisters in that state. In another example, state department of insurance officials and assisters in Montana told us that some individuals in that state had become dually enrolled in Medicaid and a QHP after the state expanded eligibility for its Medicaid program in 2016. This dual enrollment is problematic as individuals may be held liable for repaying certain exchange subsidies received during the period of duplicate coverage. Further, the federal government could be paying twice, subsidizing exchange coverage and reimbursing states for Medicaid spending for those enrolled in both. CMS and the five selected states in our review have monitored QHP enrollees’ post-enrollment experiences by reviewing information reported by consumers, through call centers and enrollee surveys, as well as by assisters. CMS and selected states use this information to ensure that enrollee issues are resolved and to improve educational resources and post-enrollment assistance for enrollees, among other purposes. In addition to monitoring QHP enrollee experiences through these methods, CMS and the selected states conduct activities to monitor QHPs. (See App. II). CMS uses information collected from enrollees through its call center to monitor QHP post-enrollment experiences. QHP enrollees and their representatives, such as assisters, may call the CMS exchange call center to request agency assistance in resolving concerns. Using its casework system, CMS tracks individual issues—referred to as cases— that require action on the part of an issuer, state, or CMS to resolve. In 2014 and 2015, agency officials assigned all cases to one of four broad categories of concerns—plan and issuer, tax filing, eligibility, or legal and administrative—as well as to subcategories within each category that describe the general nature of the issue. Cases related to post- enrollment issues may be included in any of these categories. To identify trends in cases and to ensure their timely resolution, CMS officials prepare and examine weekly and monthly reports that include information on the type and volume of cases received and resolved by category, among other information. According to our analysis of CMS exchange casework data, three-quarters of CMS’s casework in 2014 and 2015 was in the plan and issuer category, which includes post-enrollment concerns such as enrollee access to services or benefits, among other issues. (See table 4.) Appendix III includes more detailed information about CMS QHP casework in 2015. CMS shares relevant casework information with the appropriate agency, issuer, or state officials for research and timely resolution, depending on the type of action required, according to CMS officials we interviewed. For instance, CMS officials told us that cases noting concerns about issuer compliance are forwarded to CMS’s compliance team for further investigation. CMS officials told us that they work with individual issuers to ensure that cases are resolved in a timely manner and the causes of any casework trends are addressed. For example, CMS has a monthly call with issuers to discuss casework trends and strategies to improve consumer experiences, in addition to providing ongoing technical assistance with specific casework issues, according to CMS officials. In addition to reviewing casework to resolve individual enrollee concerns, CMS officials we interviewed reported reviewing casework data relevant to their oversight responsibilities of issuers. For example, prior to conducting any issuer compliance reviews, CMS officials told us that they review relevant casework data, such as complaints, for the issuer. In addition, CMS officials told us that, as of 2016, they have begun using casework information to identify issuers for compliance review, including by reviewing any outliers in volume or timely resolution of cases. The five selected states included in our review have also used information submitted directly by enrollees through state call centers or online complaint systems to monitor enrollee experiences in both FFE and SBE states. Officials from all five of the selected states’ departments of insurance we interviewed reported tracking consumer complaints at the issuer level and working to resolve reported issues. Four of the selected states’ departments of insurance did not have any mechanism to track QHP-related complaints separate from those of other plans, according to state department of insurance officials we interviewed. One state— Indiana—began tracking QHP-related complaints in 2016 in categories such as billing, claim delay, and pharmacy benefits. In addition to monitoring complaints reported directly to them, state department of insurance officials located in FFE states have access to CMS’s casework system for all issuers operating in their state. Officials from departments of insurance in two of the three FFE states included in our review told us that they routinely monitored casework data in CMS’s system. CMS and the two states operating SBEs included in our review have surveyed or plan to survey QHP enrollees to monitor their experiences. Specifically, CMS developed a survey, which was administered to a sample of QHP enrollees nationwide, including those in FFE and SBE states, about their experiences with their plans in 2015 and 2016. CMS designed the survey to capture accurate and reliable information from consumers about their experiences with the health care and services they had received through their QHP and to allow for effective oversight, among other purposes. The survey, which was beta-tested in 2015, was administered to enrollees of QHPs with more than 500 enrollees. It included a core set of questions for enrollees on key areas of care and service, including overall rating of their QHP, the availability of information about their health plan and costs of care, how well they were able to get needed care, and the accessibility of information in a needed language or format. CMS officials told us that they ultimately expect the results of their 2017 and future surveys to, among other things, inform the agency’s monitoring of enrollee post-enrollment experiences, as well as their monitoring of issuers beginning in 2017. For example, CMS officials told us that they expect to use survey results to identify issues in enrollee overall satisfaction and access to care. CMS officials told us that they had also shared relevant 2015 survey results, and plan to share 2016 results, with issuers and SBEs to help inform their understanding of enrollee experiences. SBEs in two of the five selected states in our review had either already surveyed statewide QHP enrollees about their post-enrollment experiences or had plans to do so. Specifically, as mentioned earlier in this report, Vermont QHP enrollees were surveyed in 2015 to assess their satisfaction with their QHP. Officials from the state’s exchange office told us that they used the survey results to inform their prioritization of work related to improving enrollee experiences, such as developing better methods to educate enrollees on financial literacy and health insurance information, and to work with issuers to ensure that consumers with complaints are using the appropriate channels for filing them. While they have no plans to repeat the survey, officials from the state’s exchange office told us that they plan to add questions to another statewide survey that is conducted every two years or develop a shorter survey as an attempt to monitor QHP enrollee experiences over time. Officials from Colorado’s exchange office told us in July 2016 that they were in the process of surveying statewide QHP enrollees in 2016 about their experiences with their plans, including those related to post-enrollment. CMS and states have also monitored enrollee experiences with information received from assisters. CMS receives some post-enrollment information from navigators and other assisters on an ongoing basis as it relates to enrollees in FFE states. For example, federally funded navigators are required to report the number of post-enrollment meetings they have held with QHP enrollees on a weekly basis, and, according to CMS, the agency plans to require such navigators to report more detailed information related to their post-enrollment work with enrollees. In addition, post-enrollment issues are occasionally discussed during weekly meetings that CMS officials hold with navigators to discuss their ongoing work, according to CMS officials and navigators we interviewed. While the requirement to report post-enrollment information to CMS is only applicable to federally funded navigators, agency officials told us that they occasionally receive some post-enrollment information from other assisters, or from consumer advocacy groups who work with them, on an informal basis. For example, CMS officials reported receiving some information from assisters about low levels of health literacy among QHP enrollees. CMS officials told us that they use information they receive from navigators and other assisters to help them troubleshoot FFE enrollee problems, clarify policy, and develop additional training or materials for dissemination. Specifically, CMS circulates weekly newsletters to federally funded navigator grantees that address current areas of interest among navigators. Recent topics in these newsletters have included conducting culturally competent outreach and the appeals process. One navigator we interviewed told us they found these newsletters helpful because they explained relevant issues and presented solutions. CMS officials told us that they have also developed webinars to address post- enrollment issues identified by assisters; recent webinars addressed topics such as helping consumers after the open enrollment period, transitioning from a QHP to other coverage, and assisting consumers during the tax-filing seasons. The five selected states in our review, including those using the FFE and operating an SBE, have also gathered some information about enrollees’ post-enrollment experience from assisters, including navigators, operating in their state, according to state department of insurance officials and assisters we interviewed. The amount of information that assisters shared with these selected state officials varied and tended to be informal, as the selected states’ departments of insurance do not require navigators and assisters operating in their state to report any information about consumers’ post-enrollment experiences, according to officials and assisters we interviewed. Exchange offices in the two selected states in our study that operated an SBE, Colorado and Vermont, required their state-funded assisters to routinely report information about the post-enrollment assistance they provided, according to officials and assisters, and, officials from the state exchange offices told us that they use this information to, among other things, identify and address any problems related to enrollees’ experiences with their QHPs, or identify training needs. To the extent that assisters report information about consumers’ post- enrollment experiences to state officials in either FFE or SBE states, the information they provide tends to be about individual issues as they work with consumers to address them, according to the assisters and officials from state departments of insurance and exchange offices we interviewed. However, we found that assisters operating in four of the five selected states included in our review have also shared information about trends in QHP enrollee post-enrollment experiences to state department of insurance and exchange officials. For example, according to an official at the Montana department of insurance, assisters informed state officials about QHP enrollees who were found to be dually enrolled in Medicaid and have worked with state department of insurance and CMS officials to address the issue. Similarly, assisters operating in North Carolina and Colorado also told us that they have shared information with their department of insurance and exchange office, respectively, about trends in consumers’ experiences, including those related to post-enrollment that the issuer has the responsibility to resolve. We provided a draft of this report to HHS for comment; HHS provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of Health and Human Services, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-7114 or dickenj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in Appendix IV. To examine what is known about the early experiences of enrollees in qualified health plans (QHP) obtained through the exchanges, we conducted a structured search of research databases using various combinations of relevant search terms including, “Affordable Care Act,” “qualified health plan,” “marketplace,” and “exchange,” to identify any literature published from January 1, 2014, through April 30, 2016, that reported on QHPs obtained through the exchanges. We then reviewed the abstracts for 643 articles and the full text of 275 of those articles to determine whether they included information about QHP enrollees’ post- enrollment experiences and otherwise met our inclusion criteria. Our inclusion criteria included journal articles and government publications, as well as policy briefs or papers. Based on these steps, we identified 5 nationally representative surveys whose results were published in 14 articles between June 19, 2014, and July 7, 2016, and then summarized the QHP enrollee experiences on which these articles reported. To assess the reliability of the data presented in these surveys, we interviewed or corresponded with the authors of all of the survey reports, reviewed supporting documentation to understand what the surveys measured, and we examined the data for apparent errors. Although the surveys had relatively low response rates, they each reported that their results are nationally generalizable within certain margins of sampling error. In addition, the surveys reported similar results with respect to enrollee experiences. Based on these steps, we found the data to be sufficiently reliable for our purposes. The key methodological attributes of the five surveys are presented in table 5. The Commonwealth Fund, Americans’ Experiences with ACA Marketplace Coverage: Affordability and Provider Network Satisfaction, Findings from the Commonwealth Fund Affordable Care Act Tracking Survey, February – April 2016 (New York: July 7, 2016). The Commonwealth Fund, Americans’ Experiences with ACA Marketplace and Medicaid Coverage: Access to Care and Satisfaction, Findings from the Commonwealth Fund Affordable Care Act Tracking Survey, February – April 2016 (New York: May 25, 2016). The Commonwealth Fund, Are Marketplace Plans Affordable? Consumer Perspectives from the Commonwealth Fund Affordable Care Act Tracking Survey, March-May 2015 (New York: Sept. 25, 2015). The Commonwealth Fund, To Enroll or Not To Enroll? Why Many Americans Have Gained Insurance Under the Affordable Care Act While Others Have Not (New York: Sept. 25, 2015). The Commonwealth Fund, Americans’ Experiences with Marketplace and Medicaid Coverage, Findings from the Commonwealth Fund Affordable Care Act Tracking Survey, March-May 2015 (New York: June 12, 2015). The Commonwealth Fund, Are Americans Finding Affordable Coverage in the Health Insurance Marketplaces? (New York: Sept. 18, 2014). The Commonwealth Fund, Gaining Ground: Americans’ Health Insurance Coverage and Access to Care after the Affordable Care Act’s First Open Enrollment Period, (New York: July 10, 2014). Deloitte Center for Health Solutions, 2016 Survey of US Health Care Consumers: A Look at Exchange Consumers. (Washington, D.C.: May 11, 2016). Deloitte Center for Health Solutions, Public Health Insurance Exchanges: Opening the Door for a New Generation of Engaged Health Care Consumers, 2015 Survey of US Health Care Consumers (Washington, D.C.: Aug. 3, 2015). Kaiser Family Foundation, Survey of Non-Group Health Insurance Enrollees, Wave 3 (Menlo Park, CA: May 20, 2016). Kaiser Family Foundation, Survey of Non-Group Health Insurance Enrollees, Wave 2 (Menlo Park, CA: May 21, 2015) Kaiser Family Foundation, Survey of Non-Group Health Insurance Enrollees (Menlo Park, CA: June 19, 2014). PerryUndem Research/Communication, GMMB, and the Robert Wood Johnson Foundation, Results From a Survey of Individuals Who Purchased Health Plans Through the Health Insurance Marketplace (October 2015). Urban Institute, Health Reform Monitoring Survey, Health Care Access and Affordability among Low- and Moderate-Income Insured and Uninsured Adults under the Affordable Care Act, (Washington, D.C.: April 21, 2016). CMS and the selected states we reviewed conduct oversight of qualified health plans (QHP) offered on the exchanges to ensure that they comply with federal standards. This oversight generally includes certifying that QHPs have met these federal standards before consumers enroll in the QHP, although CMS also conducts a post-certification review as part of its oversight to ensure that certified QHPs are ready for enrollees to use in the plan year. To ensure that issuers are continuing to meet standards during the plan year, CMS and states also conduct compliance reviews and other ongoing monitoring activities. CMS and state oversight activities vary, depending on whether states utilized the federally facilitated exchange (FFE) or a state-based exchange (SBE). In 2016, 34 states utilized the FFE and 17 states operated an SBE. In order for a QHP to be offered on the exchanges, CMS, the SBE, or state department of insurance officials must first certify that the QHP meets all relevant federal standards. Specifically, QHP issuers must be state licensed and meet a range of other standards in order for the plan to be offered on either the FFE or a SBE. For example, these other standards include serving a geographic area that is established without regard to racial, ethnic, language or health status factors and providing enrollees with access to a sufficient number and type of covered providers to assure all services will be accessible without unreasonable delay. For all QHPs offered on the FFE, CMS reviews plan information and is responsible for ensuring that the plan meets federal standards prior to the annual open enrollment period. Issuers submit an application with plan data to CMS for review, and CMS officials told us they review all applications for current and new issuers and send information to issuers with corrections prior to certification. Officials from two selected FFE states said that they conducted reviews for QHP certifications in parallel with CMS using the same federal exchange standards and submitted recommendations for QHP certification to CMS. CMS officials told us that they examine the information and recommendations submitted by states and may conduct additional reviews to investigate any concerns that state officials may have had during their certification review of plans. CMS also conducts post-certification reviews of QHPs offered on the FFE. Agency officials told us that these reviews are focused on high priority and consumer-focused areas to ensure that issuers continue to meet certification standards and that certified QHPs are ready for enrollees to use in the plan year. From 2014 to 2016, CMS officials said they conducted at least one post-certification review for all QHP issuers in FFE states. For example, to ensure that consumers have up to date and accurate formulary information specific to their QHP, since 2014 CMS officials reviewed formulary information and coverage displayed on selected issuer’s website. (See table 6 for the number of CMS post- certification reviews by focus area from 2014 to 2016.) SBEs are responsible for developing a process to certify the QHPs in their state to ensure compliance with federal standards and are responsible for certifying the plans prior to the annual open enrollment period. Exchange officials from the two selected SBE states confirmed that they have a process in place to review and certify QHPs using the federal standards. To ensure that QHPs offered on the FFE and SBEs are continuing to meet standards throughout the plan year, CMS and selected states also conduct compliance reviews to varying degrees to ensure compliance with federal and state requirements, among other ongoing monitoring activities. Specifically, CMS conducts compliance reviews of QHPs offered by issuers in the FFE to ensure compliance with exchange-related standards, and SBE states are required to have oversight processes in place to ensure compliance with the same standards. Officials in the selected state departments of insurance we reviewed, both in the FFE and SBE states, state that they oversee issuers selling health insurance in their state to ensure compliance with state requirements and certain other standards. To identify QHP issuers in the FFE for compliance reviews, CMS uses a risk-based process that leverages information gathered from CMS account managers who work directly with QHP issuers, the certification review process, and the issuers’ compliance histories, including their performance in addressing identified issues. These compliance reviews assess QHPs’ compliance with a range of federal exchange standards, such as the requirement for issuers to maintain state licensure. In 2014 and 2015, key priority areas for CMS reviews included whether QHP issuers were covering prescription drugs in accordance with federal regulations and the readability of health plan notices for enrollees. In 2015, CMS conducted compliance reviews of QHPs offered by 32 issuers located in 15 states, representing 14 percent of all issuers offering QHPs on the FFE that year, and, in 2014, CMS conducted compliance reviews of QHPs offered by 23 issuers located in 14 states, representing 13 percent of all issuers offering QHPs on the FFE that year. As a result of its 2014 compliance reviews, CMS identified a range of issues, including the following examples: Some issuers had been excluding information from their QHP provider directories about whether providers were accepting new patients. Some issuers had not developed a procedure for resolving certain types of QHP consumer concerns. Some issuers sent notices to QHP enrollees that omitted required information explaining how those with limited English proficiency can access language services to understand their health plan notice. Once compliance reviews are completed, CMS officials said account managers follow up with issuers during the benefit year to monitor and ensure the resolution of identified issues. CMS requires all states operating SBEs to implement oversight and monitoring policies and procedures for their exchanges as a way to help ensure compliance with federal standards. As part of both the application to implement an SBE and required annual reporting, exchange officials must demonstrate their readiness to conduct plan management and oversight under the same federal standards as required for the FFE, including QHP certification and ensuring ongoing QHP compliance. Exchange officials from our selected states told us they have processes in place to report annually to CMS, and they conduct oversight activities in varied ways. Colorado exchange officials told us they rely on issuers complying with their contracts with the exchange office, which are required for an issuer to offer QHPs on the exchange and include an agreement on standards such as QHP certification, market conduct, and resolving enrollee concerns. Vermont exchange officials said they have enhanced their oversight and monitoring program. Additionally, officials we interviewed from our selected states’ departments of insurance in the both FFE and SBE states told us that they conduct compliance reviews of QHP issuers in the same way for all issuers offering health plans issued in their state, both on and off the exchange, to ensure compliance with state insurance rules and federal health insurance market standards, which are generally applicable to all plans, whether offered on or off an exchange. The selected states’ compliance reviews vary in scope and frequency—for instance, officials from Colorado told us that compliance reviews are conducted on an ad- hoc basis if there is a complaint of potential non-compliance, and officials from Montana told us that they conduct compliance reviews using retrospective data from the previous four to five years, and also investigate federal and state standard violation allegations. Officials from the selected state departments of insurance also told us that they generally conduct their oversight and monitoring activities at the issuer level and therefore were unable to readily separate out data on QHPs or QHP enrollee experiences. An official from one state department of insurance told us that it was important to maintain a level playing field and keep monitoring standards and policies the same for both QHPs and non- QHPs. CMS operates a Marketplace Call Center to assist the needs of consumers in states that utilize the federally facilitated exchange (FFE). CMS records and tracks issues—referred to as cases—that require action on the part of an issuer, state, or CMS to resolve. According to CMS, cases may include requests, such as those related to an address change, complex questions—for example, relating to tax filings, as well as individual complaints. In 2015, all cases were assigned to one of four categories—plan and issuer concerns, tax filing issues, eligibility, and legal and administrative— and to subcategories within each category that describe the general nature of the case. CMS officials described the main case categories as follows: Plan and Issuer Concerns: This category includes cases where issuers have the capacity or responsibility to resolve cases, such as disenrollment or premium payment. Tax Filing Issues: Cases in this category involve enrollee issues related to their tax form. Exchange enrollees are required annually to reconcile the amount of premium tax credit (a federal subsidy that is applied towards qualified health plan premiums) allowed based on reported income with the amount of premium tax credit received in advance. Eligibility: Cases in this category primarily consist of issues that consumers experienced prior to enrolling in a qualified health plan, such as technical errors on the exchange Web site, or questions regarding eligibility for the advanced premium tax credit. Legal and Administrative: This category includes consumer allegations of fraud or inappropriate release of enrollee information. In 2015, most cases were assigned to the plan and issuer category and, within that category, issuer enrollment/disenrollment was the most frequently assigned subcategory. This subcategory included cases of consumers having concerns with being properly enrolled or dis-enrolled by an issuer, such as when an issuer has not processed enrollment information sent from the exchange in a timely manner. The tax filing issues category became the second most frequently assigned category in 2015 when enrollees were required to submit tax information related to their qualified health plan. CMS officials told us that a significant portion of the cases in this category dealt with enrollees requesting an extra copy of their tax form, disagreeing with the information on their tax form, or requesting to update their mailing address. The eligibility category primarily included issues related to consumer requests for special enrollment periods and questions relating to the advanced premium tax credit, according to CMS. Lastly, CMS officials reported that the legal and administrative category includes cases such as a consumer alleging fraud committed against them that is then used in their program integrity review process. Table 7 shows the number and percentage total of cases by subcategory within the four main categories in the casework system in 2015. In addition to the contact named above, Kristi Peterson, Assistant Director; Patricia Roy, Analyst-in-Charge; Laura Sutton Elsberg; Kate Nast Jones; and Joanna Wu made key contributions to this report. Also contributing were Leia Dickerson; Sandra George; and Laurie Pachter.

The Patient Protection and Affordable Care Act (PPACA), enacted in 2010, included provisions that were intended to make health insurance more available and affordable for individuals seeking coverage, including the establishment of health insurance exchanges. Health insurance was made available to individuals through the exchanges beginning in 2014. While PPACA contributed to an overall expansion in health insurance coverage, experts and consumer advocates have raised concerns about enrollees' experiences with QHPs, including access to providers and affordability of care. PPACA includes a provision for GAO to conduct an examination of exchange activities and QHP enrollees. This report describes (1) what is known about enrollee experiences with QHPs obtained through the exchanges during the first years of exchange operation, and (2) how CMS and selected states have monitored the post-enrollment experiences of those who obtained their QHPs through the exchanges. GAO examined federal and state laws, regulations, and reports, and conducted a literature review to identify original research on enrollees' experiences with QHPs obtained through the exchanges. GAO interviewed officials from CMS and five selected states—Colorado, Indiana, Montana, North Carolina, and Vermont—that varied in geography and whether the state or CMS operated the exchange on which QHPs were offered, as well as officials from stakeholder groups and consumer assisters. Available survey data show that most enrollees who obtained their coverage through the health insurance exchanges were satisfied overall with their qualified health plans (QHP) during the first few years that exchanges operated, according to five national surveys of QHP enrollees that GAO identified through its literature review. Specifically, most QHP enrollees who obtained their coverage through the exchanges reported overall satisfaction with their plans in 2014 through 2016, according to three national surveys. The surveys reported that QHP enrollees' satisfaction with their plans was either somewhat lower than or was similar to that of those enrolled in employer-sponsored health insurance in 2015 and 2016. To varying degrees, QHP enrollees expressed satisfaction with specific aspects of their plan, including their coverage and choice of providers, and plan affordability. Stakeholders—including experts, state departments of insurance, and others GAO interviewed—and literature GAO reviewed also revealed some concerns about QHP enrollee experiences. Some enrollees found it too expensive to pay for their out-of-pocket expenses before reaching their deductibles and have reported concerns about affording care or have been deterred from seeking care, according to experts. Some enrollees have faced difficulties understanding their QHP's coverage terminology and others have faced problems accessing care after enrollment, according to stakeholders and literature reviewed. These issues have also been identified in literature as longstanding concerns of the private health insurance market. The Centers for Medicare & Medicaid Services (CMS), an agency within the Department of Health and Human Services (HHS), and selected states GAO reviewed have monitored enrollees' post-enrollment experiences by reviewing information reported by consumers and consumer assisters. For example, CMS uses information collected from enrollees through its Marketplace Call Center—where exchange enrollees may call to request agency assistance in resolving concerns. CMS officials said that they use this information to identify trends in enrollees' post-enrollment experiences and ensure that enrollee concerns are resolved in a timely manner. They began using it in 2016 to identify issuers for compliance reviews. Similarly, officials from the five selected states' departments of insurance reported tracking consumer complaints by issuer and working to resolve all reported issues. CMS developed a survey that was administered to a sample of QHP enrollees nationwide in 2015 and 2016, to gather information about their experiences with their plans. According to CMS officials, the agency expects to use results of its 2017 and future surveys to inform its monitoring of issuers. In addition, QHP enrollees in Vermont were surveyed with respect to their satisfaction in 2015; state officials reported using the results to inform their prioritization of work. CMS and selected states also reported monitoring enrollee experiences with information received from consumer assisters—including navigators—who interact directly with QHP enrollees. CMS officials told GAO that they have used information received from federally funded navigators to troubleshoot enrollee problems, clarify policies, or develop additional training or materials for dissemination. HHS provided technical comments on a draft of this report, which were incorporated as appropriate.

The safe, efficient, and convenient movement of people and goods depends on a vibrant transportation system that meets those needs. Our nation has built a vast transportation system of roads, airways, railways, pipelines, transit, and waterways that facilitate commerce and improve our quality of life. The flow of people and goods is enormous: the nation moved about 5 trillion ton miles of freight and 5 trillion passenger miles of people in 2004. Spending for commercial, personal, and government transportation represents about 11 percent of the gross domestic product. Yet there is a price for this system: Increasing congestion on the ground and in the air delays the arrival of people and freight at their destinations and imposes economic losses. According to Department of Transportation estimates, congestion costs Americans roughly $200 billion each year. The system is expensive to maintain and improve. Total federal, state and local transportation expenditures are close to $200 billion annually. There is a human cost: over 44,000 people are killed in transportation- related accidents and over 2.5 million are injured each year. The transportation system is under considerable strain from these factors, and this strain is expected to increase as the demand to move people and goods grows resulting from population growth, technological change, and the increased globalization of the economy. For example, according to the Transportation Research Board, an expected population growth of 100 million people could double the demand for passenger travel by 2040. Moreover, this population growth will be concentrated in certain regions and states, further intensifying the demand for transportation in these areas. The Department of Transportation implements national transportation policy and administers most federal transportation programs. Its responsibilities are considerable and reflect the extraordinary scale, use, and impact of the nation’s transportation system. The department has multiple missions—primarily focusing on mobility and safety—that are carried out by several operating administrations. (See table 1.) For fiscal year 2008, the President’s budget requested $67 billion to carry out these and other activities. This budget request would support about 55,000 full-time-equivalent employees. The department carries out some activities directly, such as employing more than 15,000 air traffic controllers to coordinate air traffic to make certain that planes stay a safe distance apart. However, the vast majority of its activities are not under its direct control. For example, in recent years the Federal Highway Administration (FHWA) has provided state governments nearly $34 billion each year to build and improve roads and bridges and meet other transportation needs. However, for the most part state and local governments decide which transportation projects have high priority within their political jurisdictions. Similarly, while the National Highway Traffic Safety Administration (NHTSA) encourages the use of safety belts by the motoring public as a means of saving lives and reducing injuries, states determine whether and how to punish noncompliance. In other cases—notably most freight railways and pipelines—the infrastructure is owned and operated by private companies and the Department of Transportation regulates the safety of their transportation operations. In our view, Congress and the Department of Transportation face four major transportation challenges—financing the nation’s transportation system, improving mobility, improving safety, and managing the transformation of the air traffic control system. Another three issues are of continuing concern: building human capital strategies, fostering improved departmental financial management, and improving transportation security and emergency preparedness and response. The efficiency of the nation’s transportation infrastructure is threatened by increasing demand for transportation services, and revenue from traditional funding mechanisms may be unable to keep pace at current tax rates. In addition, the nation’s long-term fiscal challenges will constrain decision makers’ ability to use other revenue sources for transportation needs. As a result of these concerns, we designated financing the nation’s transportation infrastructure as a high-risk issue this year. Revenues to support the Highway Trust Fund—the major source of federal highway and transit funding—at the current fuel tax rate are eroding. While receipts for the Highway Trust Fund, which are derived from motor fuel and truck-related taxes, are growing, the federal motor fuel tax rate of 18.3 cents per gallon has not been increased since 1993 and inflation has eroded purchasing power. In addition, increased fuel efficiency and the advent of alternative-fuel vehicles will further erode trust fund receipts. While increases in vehicle travel will increase fuel tax revenues, funding already authorized in recently enacted highway and transit program legislation is expected to outstrip the growth in trust fund receipts. According to recent estimates from the Congressional Budget Office and the President’s budget, the trust fund balance will steadily decline and reach a negative balance of more than $14 billion by the end of fiscal year 2012. (See fig. 2.) To help remedy this situation, a commission—chaired by the Secretary of Transportation—will report later this year on recommendations to place the trust fund on a sustainable path. In addition, the Department of Transportation’s strategic plan suggests exploring tolling projects and private sector involvement to address funding constraints—ideas that some state and local governments are currently exploring. Federal aviation programs are also facing growing infrastructure demands with constrained resources, and a disruption in the flow of funds may jeopardize FAA’s ability to carry out its improvement programs. Demand for air travel has increased in recent years, with over 740 million passengers flying in fiscal year 2006. Failing to meet infrastructure challenges in aviation may have significant consequences, since aviation is an integral part of the economy. To meet anticipated increases in commercial aviation travel, FAA and aviation stakeholders are developing new systems to modernize and increase capacity, but it is uncertain whether the current funding system can generate sufficient revenues to meet these budgetary needs. FAA and some stakeholders have concerns that the costs of providing and modernizing air traffic control services might increase without a corresponding increase in revenues collected from users. Under one preliminary estimate of modernization costs, FAA’s budget requirements would, on average, exceed fiscal year 2006 appropriation levels by approximately $1 billion a year (in today’s dollars) through 2025. To better connect FAA’s revenues with the cost of air traffic control services that it provides, the President’s budget for fiscal year 2008 has proposed replacing, in fiscal year 2009, FAA’s current excise tax financing system, built largely around purchases of tickets and aviation fuel, with a cost-based user fee system. This new system would aim to recover the costs of providing air traffic control services through user fees for commercial operators and aviation fuel taxes for general aviation. However, some stakeholders believe that the current structure has been effective in funding FAA and can be successful in the future, although some modifications may be necessary. In addition, the President’s budget has proposed cutting and reallocating federal funds for developing projects at the nation’s 3,400 airports. FAA estimates the total cost for planned airport projects eligible for funding at approximately $42 billion (in nominal dollars) for fiscal years 2007 through 2011. FAA is also proposing that Congress allow airports to collect more revenue from other sources to help offset any reductions. Adding to uncertainty, the current excise taxes that largely fund FAA revenue are scheduled to expire at the end of September 2007, unless there is congressional action to renew them or provide an alternative source of funding to avoid a lapse of revenue in fiscal year 2008. Freight traffic is projected to grow substantially, putting strain on ports, highways, railroads, and airports, but current public planning and financing impede strategies to address capacity investment, and industry’s ability to fund its capacity increases to meet growth is largely uncertain. Freight mobility—the ability to move goods—is a driver of economic growth, and increasing congestion and unreliability of transportation systems can have severe economic consequences. In the future, Congress is likely to receive funding requests for additional freight projects and face decisions about the federal role in the nation’s freight infrastructure. While the federal government has made huge investments in our nation’s transportation infrastructure in the last 50 years, the expansion of this infrastructure has not kept pace with needs and the system is currently under great strain. Congestion across modes—estimated to cost $200 billion per year—is significant and is projected to worsen. For example, travel on roads is expected to increase by about 25 percent from 2000 to 2010, freight traffic is expected to increase by 92 percent from 2002 to 2035, and demand for air travel is expected to climb by about 35 percent from 2006 to 2015. To help address congestion concerns, the federal government spends billions of dollars each year to build, maintain, operate, and improve the nation’s aging transportation system. As congestion increases, federal policymakers face the challenge of ensuring that funds are used efficiently in order to prevent congestion from overwhelming the system. However, currently there is little assurance that the projects selected and funded best meet national goals for meeting the nation’s mobility needs. The department and Congress have recently taken a number of new actions to address this major threat to our nation’s economic growth and quality of life. In May 2006, the department announced a national strategy that will provide $175 million to local governments to demonstrate and test innovative ideas for curbing congestion. Certain large-scale pilot projects would be chosen based on their sponsors’ willingness to implement a comprehensive congestion reduction strategy, including congestion pricing, commuter transit services, and commitments from businesses to expand flexible work schedules. The strategy also includes initiatives to encourage private sector investment in transportation infrastructure, promote the use of operational and technological improvements, address major freight bottlenecks, and accelerate major aviation capacity projects, among other things. The department is also implementing a number of new initiatives to mitigate congestion that were called for in the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU), including programs to allow states to monitor, in real-time, traffic conditions on major highways and new funding for projects that have national or regional benefits. In addition, SAFETEA-LU established a commission that will report on ways to raise revenue for highway and transit projects and also reduce the costs of congestion. Finally, in the aviation arena, FAA is the primary implementer of a multiagency effort to transform the air traffic control system in order to safely handle projected growth in the demand for air travel. I will further discuss this effort later in my statement. While these steps are encouraging, successfully addressing the nation’s mobility challenges requires strategic and intermodal approaches and solutions. The nation faces a growing fiscal crisis that challenges it to fundamentally reexamine existing government programs and commitments and to make tough choices in setting priorities and linking resources to results. In particular, the Highway Trust Fund—the largest source of federal funding for transportation—was created in 1956 for the purpose of constructing the interstate highway system and, although that system is now complete, the basic construct of the program, in terms of financing and delivery mechanisms, has not changed. In addition, this and other federal transportation programs do not have mechanisms to link funding levels with the accomplishment of specific performance-related goals and outcomes related to mobility. Most highway grant programs are apportioned by formula, without regard to the needs or capacity of recipients. In addition, the preponderance of evidence suggests that federal-aid highway grants have influenced state and local governments to substitute federal funds for state and local funds that otherwise would have been spent on highways. State and local governments have broad flexibility to select most projects that receive federal funding. As such, there is little assurance that the projects selected and funded best meet national goals for addressing the nation’s mobility needs. Intercity passenger rail service is also at a critical juncture, with the existing system in poor condition and federal subsidies—over $1 billion annually in recent years—not targeted to the greatest public benefits, including congestion relief. Furthermore, transportation programs and funding mechanisms are largely stovepiped by modes of transportation. For example, while passenger and freight travel occurs on all modes, federal funding and planning requirements focus largely on highway, transit, and aviation passenger travel. This framework makes it difficult for intermodal projects and other modal projects (e.g., freight or passenger rail) to be integrated into the transportation system. We have found, for example, that the limited visibility that freight projects receive in the process for planning and prioritizing transportation projects as well as the lack of a comprehensive evaluation approach, such as a cost-benefit framework, impedes the implementation of improvements to better ensure that systemwide, multimodal solutions are considered and adopted where appropriate. It is unlikely that mobility can be enhanced unless major modes—air, highway, rail, transit, and water—are well connected. However, intermodal connections, such as multimodal passenger terminals and roads that link freight terminals and major highways, are among the transportation system’s weakest links. The critical issues facing Congress and the department to effectively address congestion problems and enhance the nation’s mobility include: How narrowly or broadly should the federal role be defined? Should federal programs be more closely aligned with specific national interests and purposes, such as interstate freight mobility? Should formulas be revised to better consider need, performance, capacity, and effort by states and localities? Can intermodal solutions be effectively carried out within the existing federal modal program framework, or is another model needed? Finally, a high priority should be maximizing the benefits of federal investments in transportation infrastructure and ensuring accountability for results. Each year, FHWA distributes billions of dollars—$34.2 billion in fiscal year 2006—to state governments for projects aimed at improving the nation’s highway systems. However, we have found that often formal analyses are not used in deciding among alternative projects, projects often do not meet anticipated outcomes, and evaluations of outcomes are not typically conducted. Furthermore, we have reported on the need for improving accountability for results in FHWA’s oversight of projects, through goals and performance measures, for example. The agency has made progress in this area, partly in response to mandated improvements in SAFETEA-LU, but a continued focus on efforts to improve accountability will be important. Finally, FAA provides funds to airport operators to provide increased capacity at the nation’s airports and has estimated that the total cost for planned projects eligible for federal grants for fiscal years 2007 through 2011 will be $42 billion. While changes the Administration is proposing may reduce the amount FAA provides, it will be essential to ensure that public benefits from these investments are maximized. Each year, tens of thousands of people are killed and millions are injured in transportation accidents in the United States. In 2005 alone, over 44,000 people were killed and over 2.5 million were injured in highway, aviation, railroad, transit, and pipeline accidents. (See fig. 3.) Motor vehicle crashes, in particular, exact an enormous personal and economic toll on this country and are the leading cause of death for people aged 3 through 33. While transportation safety has improved considerably over the past 4 decades, in recent years, fatalities have plateaued. Since the highest pay- off actions—such as improvements in vehicle crashworthiness and increases in seat belt use—have occurred, future progress will be more difficult. Of particular concern is the limited progress in improving safety on our nation’s roads, where about 95 percent of all transportation fatalities occur. Furthermore, motorcycle fatalities have steadily increased over the past decade (to over 4,500 in 2005). While nonhighway modes of travel are much safer, safety in these modes—such as aviation and rail—is also a major concern because when accidents occur, they can have catastrophic consequences. Projected increases in congestion across modes, as a result of population and economic growth, could cause a deterioration in transportation safety in the future despite vigorous efforts to reduce accidents. To address these problems, the department has designated improving safety as its highest priority. Its efforts to improve surface transportation safety are wide-ranging and include programs to change driver behaviors—such as alcohol use and speeding—that cause accidents; enhance motor vehicle safety; improve the safety of highway and intersection infrastructure; and improve safety performance in the motor carrier, railroad, transit, and pipeline industries. SAFETEA-LU established an incentive grant program—which has been quite successful—to encourage states to pass primary seat belt laws. SAFETEA-LU also mandated a number of other promising new initiatives, including a grant program for highway safety that provides states with flexibility to target funds to their most critical safety needs. Under this program, states are required to prepare strategic highway safety plans, based on an analysis of safety data, and to assess results. FAA focuses on improving safety in commercial aviation, in which accidents are rare but have the potential for a large loss of life, as well as in general aviation. The agency’s safety activities include air traffic control as well as certification and inspection of various participants in the aviation industry, such as commercial airlines, flight schools, and aircraft manufacturers. While the department’s many efforts to improve transportation safety are to be commended, certain areas require increased attention. In particular, improvements in data, performance measures, and evaluations are needed to determine whether programs are achieving intended results. For example, in reviewing certain programs of the Federal Motor Carrier Safety Administration aimed at improving driver behavior, we found that, in some cases, funds were being directed to initiatives that lacked information on whether they worked and that evaluations of program impacts were not planned for a number of years. In reviewing a NHTSA grant program to help states improve the quality of their traffic safety data, we found that the agency did not have an effective process in place for monitoring progress. We have also found that the effectiveness of the department’s efforts to oversee and improve the safety performance of airlines, truck companies, pipeline companies, and railroads is unclear because of limitations in data, performance measures, and evaluation. For example, agencies need to develop better measures of the direct results of their efforts—such as safety improvements made as a result of enforcement of safety standards—that contribute toward reductions in accidents. Performance measures and evaluations, supported by appropriate data, provide managers with information on program results that helps them make decisions that can improve performance, including decisions to refine programs and adjust policies and priorities. This information can also hold agencies accountable for the performance of their programs and support congressional oversight. While agencies have been making progress in this area in response to our recommendations as well as some mandated improvements in SAFETEA-LU, it is important that the department continue to improve information on the performance of its safety programs to have greater assurance that they are producing desired effects. Furthermore, the department’s ability to maintain the high level of safety in the aviation industry will depend to a large extent on FAA’s ability to hire, train, and deploy its primary workforce, including safety inspectors and air traffic controllers. FAA must overcome several key challenges in this area. Planned changes in the agency’s oversight approach for air carriers will result in workload shifts for its inspectors that will make it important for FAA to improve its staffing process. In addition, the agency plans to hire almost 12,000 new air traffic controllers by 2015 to replace retiring controllers and accommodate increases in air traffic and will need to train these new controllers and incorporate them into its workforce. The current approach to managing air transportation is becoming increasingly inefficient and operationally obsolete. In 2003, Congress authorized the creation of the JPDO, housed within FAA, to plan for and coordinate the transition to NextGen, a complex and ambitious multiagency undertaking that is intended to upgrade the system by 2025 to safely accommodate increased air traffic. As the primary implementer of the transition to NextGen, FAA faces challenges in moving from planning to implementation, including institutionalizing management reforms it has made in recent years, obtaining financial and technical resources and expertise, and collaborating with JPDO on planning efforts. If FAA does not meet these challenges, the realization of NextGen goals could be severely compromised. Without a timely transition to NextGen capabilities, JPDO officials estimate a future gap between the demand for air transportation and available capacity that could cost the U.S. economy billions of dollars annually. FAA has had systemic management and acquisition problems that have led us to designate its air traffic control modernization program as high-risk since 1995. However, FAA has made significant progress in recent years. For example, FAA established the Air Traffic Organization to operate and modernize the air traffic control system. This organization is headed by a Chief Operating Officer who has focused on implementing more businesslike management and acquisition processes to address cost, schedule, and performance shortfalls that plagued air traffic control modernization in the past. FAA has reduced organizational stovepipes, increased accountability for costs, and begun investment reviews of major acquisitions. FAA has reported meeting its acquisition cost and schedule goals for the last 3 years. JPDO has completed some initial planning necessary for implementing NextGen. For example, JPDO has been developing an enterprise architecture, or technical blueprint, that it expects will provide more clarity regarding its expectations for NextGen, thereby facilitating (1) coordination among JPDO’s partner agencies and private sector manufacturers, (2) alignment across agencies of research and development activities with the blueprint, and (3) integration of modernized systems in a way that minimizes overlap and duplication and maximizes integration. As we reported in November 2006, a limited, preliminary cost estimate concluded that FAA’s budget under a NextGen scenario would average about $15 billion per year through 2025, or about $1 billion more annually (in today’s dollars), on average, than FAA’s fiscal year 2006 appropriation. Despite its progress, as the key implementer of NextGen, FAA needs to institutionalize improvements made and continuously improve. For example, we recommended that, before making decisions to fund systems already in service, FAA re-evaluate projects’ alignment with strategic goals and objectives, but FAA’s acquisition management guidance does not clearly indicate if this is yet the case. The agency developed a cost estimating methodology, but has yet to implement it, as well as a framework for improving system management capabilities, but has yet to institutionalize it. Additionally, we recently recommended that FAA examine its strengths and weaknesses with regard to the technical expertise and contract management expertise necessary to transition to NextGen. In response, FAA is considering convening a blue ribbon panel to make recommendations, which we believe could help the agency begin to address this concern. JPDO faces challenges in coordinating agencies and continuing planning necessary for implementation of NextGen. For example, work remains to synchronize NextGen’s enterprise architecture with the partner agencies’ planning documents and to keep the necessary research and development on track. In addition, JPDO has yet to provide Congress with a valid, comprehensive estimate of the costs to JPDO partner agencies for the required research, development, systems acquisitions, and systems integration. Finally, continuing collaboration between JPDO and the Office of Management and Budget is needed to allow the budget agency to make funding decisions based on a unified NextGen program. The Congressional Research Service has pointed out that Congress may examine options to align the budgets of the agencies involved, given that JPDO does not have authority over funding, personnel, and resources. The department and the transportation sector as a whole face persistent human capital challenges that put their mission performance at risk. Building human capital strategies that will allow the department and the transportation sector to attract, hire, and retain an effective workforce is an overarching issue that directly affects their ability to respond to the challenges I have outlined today. In particular, both are confronted with an impending shortage of skilled people that threatens to have serious short- and long-term consequences. For example, FAA alone expects to lose about 10,000, or 70 percent, of its air traffic controllers over the next 10 years, mostly due to retirement. For the department and the transportation sector as a whole, the growing demand for transportation services will collide with the reality of fewer people entering transportation-related fields. Further complicating this shortage, changes in intergovernmental responsibilities for delivering transportation services, new travel patterns, advances in technology, and changed public expectations are redefining the competencies and skills that are needed. Increasingly, transportation will require more diverse, sophisticated management and technical competencies than ever before. The department has acknowledged that accomplishing its mission depends on a strategic approach to human capital, and it is taking steps to adopt such an approach. For example, in 2005, the agency piloted a program to expand entry-level hiring in mission-critical occupations. Also, in 2006, the agency increased its investments in human capital by 48 percent. Furthermore, the agency is working to align its human capital initiatives to meet the President’s Management Agenda. In the department’s current performance and accountability report, the Office of Management and Budget awarded the department top marks for current and prospective progress on its human capital initiative. However, the department has not convinced its workforce about these results. In the results of the 2006 federal employee human capital survey, the employees scored the department lower in each of the four broad areas than they did in 2004, when the survey was last conducted. Among the 36 federal agencies surveyed, the department finished in the bottom 10 for talent management and job satisfaction and in the bottom 3 for fostering a results-oriented performance culture and for leadership and knowledge management. The department will need to take further actions to address these issues, to improve its ability to respond to the challenges it faces. Across the transportation sector, transportation agencies are also taking steps to improve human capital practices, by identifying organizational and staff competency needs, as well as other gaps. They are also beginning to investigate nontraditional sources for qualified employees, such as highly qualified retirees from other organizations, as well as ways to develop individual competencies by training the existing workforce. While these efforts are promising, these agencies vary widely, and although each has its own unique capabilities and resources to address workforce needs, all have limited resources. Furthermore, few have addressed their future workforce needs comprehensively, which further complicates efforts to predict how many people in specific job categories for each type of agency will be needed in 5 or 10 years. In 2003, we cited financial management as a major challenge facing the Department of Transportation, specifically, identifying weaknesses in the accuracy and reliability of FAA’s financial information. In recent years, the department has made significant progress in managing its finances, including substantial improvements in FAA’s financial management systems and practices. Improvements have included installing a departmentwide financial system, including a new general ledger system and integrated property systems at FAA, as well as receiving unqualified opinions on its financial statements from auditors for several fiscal years in a row. As a result of this progress, in 2005, we removed FAA’s financial management from our high-risk list. While progress has been made, work remains to ensure that the Department of Transportation soundly manages its finances and accounts for its use of federal and other funds. For fiscal year 2006, the department received a qualified opinion on its financial statements and the auditors cited two material internal control weaknesses. This qualified opinion resulted from a material weakness at FAA relating to management’s inability to support the accuracy and completeness of a $4.7 billion account used for equipment and facility projects. The department’s Office of the Inspector General has reported that correcting this deficiency will be critical for FAA to meet its stated goal of sound financial management. The other material weakness involves the financial management, reporting, and oversight of the Highway Trust Fund agencies. During fiscal year 2006, trust fund agencies implemented significant improvements over several previously reported deficiencies. However, weaknesses remained in several areas, including a lack of policies and procedures to ensure more timely correction of any abnormal account balances and concerns with the preparation and analysis of financial statements. The Inspector General has listed several additional steps to further improve oversight of the trust fund, including better detection of improper payments and development of realistic project cost estimates. The size and interconnectedness of the nation’s transportation systems make it highly difficult to secure against attack. In 2003, we cited transforming transportation security as a major challenge facing the Department of Transportation. In recent years, Congress has shifted many of the department’s security responsibilities to the Department of Homeland Security, which now has primary responsibility for securing the nation’s transportation infrastructure, including aviation, railroad, pipeline, and other systems. The Department of Transportation has retained some involvement in securing transportation infrastructure, in part, due to overlap with its safety efforts involving freight, including the transportation of hazardous materials, and passenger rail. In light of these changes, the department faces the challenge of working with the Department of Homeland Security to clarify its remaining role in securing the nation’s transportation infrastructure. The sheer number of stakeholders involved in securing transportation modes can sometimes lead to communication challenges, duplication of effort, and confusion about roles and responsibilities. For example, the department’s safety standards have at times conflicted with the Department of Homeland Security’s security standards. Both departments have begun efforts to strengthen coordination and cooperation to promote the security of the transportation system. The departments have signed a memorandum of understanding to define broad areas of responsibility for each department and to delineate specific security related roles, responsibilities, resources, and commitments for mass transit, rail, and other matters. However, the departments’ coordination efforts in this area are ongoing. The department also coordinates with the Department of Homeland Security in developing protective measures affecting transportation and has statutory roles related to emergency preparedness, response, and recovery. This encompasses programs like FHWA’s Emergency Relief program, which provides funding to states to repair or reconstruct highways and roads damaged or destroyed in disasters. During times of disaster, the department plays a significant role as the lead and supporting agency for coordinating transportation support. In this role, it is primarily responsible for coordinating the provision of federal and civil transportation services, as well as the recovery and restoration of transportation infrastructure, among other things. In the future, the department will be tasked to further clarify its roles and responsibilities with the Department of Homeland Security in planning for and providing evacuation assistance. Catastrophic disasters like Hurricane Katrina demonstrate the importance of transportation preparedness and response to ensure the safe evacuation of citizens in emergencies when state and local governments are overwhelmed. Yet the department’s responsibilities in providing evacuation assistance have not been entirely clear. In addition, despite recent progress by the federal government in providing evacuation assistance, gaps remain. For example, the Department of Homeland Security has not yet clarified, in the federal government’s plan for disaster response, the leading, coordinating, and supporting federal agencies to provide evacuation assistance when state and local governments are overwhelmed, and what their responsibilities are. One White House report recommended that the Department of Transportation be designated as the agency responsible for developing the federal government’s capability to carry out mass evacuations when state and local governments are overwhelmed. Mr. Chairman, this concludes my prepared statement. I would be pleased to respond to any questions that you or other Members of the Subcommittee might have. For further information on this statement, please contact Patricia Dalton at (202) 512-2834 or Daltonp@gao.gov. Individuals making key contributions to this testimony were Matthew Cail, Judy Guilliams-Tapia, Marietta Mayfield, Margaret Vo, and James Ratzenberger. Description (expected completion) Airport capital development funding (early 2007) Federal role in overseeing and funding railroad bridge and tunnel projects (mid 2007) JayEtta Hecker (202) 512-2834 HeckerJ@gao.gov (mid 2007) The Federal Transit Administration’s New Starts program (mid 2007) The Federal Transit Administration’s implementation and oversight of the New Freedom program (TBD) Operational, capacity, and safety issues associated with the Airbus A380 (early 2007) Introduction of very light jets into the national airspace system (mid 2007) Approaches to the efficient use of existing transportation infrastructure (mid 2007) Freight bottlenecks (late 2007) Public-private partnerships in transportation (late 2007) Restructuring the federal-aid highway program (TBD) Trends and performance of state contracting with the private sector (TBD) Port preparedness and mobility of goods during natural disasters (early 2007) Katherine Siggerud (202) 512-2834 SiggerudK@gao.gov mobility challenges (mid 2007) Corporate Average Fuel Economy program policy options (mid 2007) Surface transportation compliance with the Americans with Disabilities Act of 1990 (late 2007) The public safety impact of the Transportation Security Administration’s modifications to the prohibited items list (early 2007) Implementation progress of the Uniform Carrier Registration program (TBD) Description (expected completion) Operational, capacity, and safety issues associated with the Airbus A380 (early 2007) Administration (late 2007) Identification of motor carriers that pose a high risk for crashes (mid 2007) Susan Fleming (202) 512-2834 FlemingS@gao.gov monitoring of unsafe motor carriers (mid 2007) Safety standards for older drivers (early 2007) Emerging trends in and challenges to preventing highway fatalities (late 2007) Next generation air transportation system Survey of Joint Planning and Development Office stakeholders (mid 2007) Gerald Dillingham (202) 512-2834 DillinghamG@gao.gov modernization program (mid 2007) The Federal Aviation Administration’s financial management efforts (mid 2007) Highway transit funding authority (mid 2007) Security and emergency preparedness and response Research and development of aviation passenger checkpoint screening technologies (mid 2007) Cathleen Berrick (202) 512-3404 BerrickC@gao.gov commercial vehicles (TBD,b) Port preparedness and mobility of goods during natural disasters (early 2007) Katherine Siggerud (202) 512-2834 SiggerudK@gao.gov transit agencies (TBD) report my not be pblicly ilble t thi time because it my contin ecrity enitive informtion. The following are the most pertinent GAO products to the topics discussed in this hearing statement since our 2003 report on management challenges facing the Department of Transportation. Other products can be found at GAO’s Website at www.gao.gov. Intercity Passenger Rail: National Policy and Strategies Needed to Maximize Public Benefits from Federal Expenditures. GAO-07-15. Washington, D.C.: November 13, 2006. Freight Railroads: Industry Health Has Improved, but Concerns about Competition and Capacity Should Be Addressed. GAO-07-94. Washington, D.C.: October 6, 2006. Aviation Finance: Observations on Potential FAA Funding Options. GAO-06-973. Washington, D.C.: September 29, 2006. National Airspace System Modernization: Observations on Potential Funding Options for FAA and the Next Generation Airspace System. GAO-06-1114T. Washington, D.C.: September 27, 2006. Highway Finance: States’ Expanding Use of Tolling Illustrates Diverse Challenges and Strategies. GAO-06-554. Washington, D.C.: June 28, 2006. Highway Trust Fund: Overview of Highway Trust Fund Estimates. GAO-06-572T. Washington, D.C.: April 4, 2006. Federal Aviation Administration: An Analysis of the Financial Viability of the Airport and Airway Trust Fund. GAO-06-562T. Washington, D.C.: March 28, 2006. Freight Transportation: Short Sea Shipping Option Shows Importance of Systematic Approach to Public Investment Decisions. GAO-05-768. Washington, D.C.: July 29, 2005. Highlights of an Expert Panel: The Benefits and Costs of Highway and Transit Investments. GAO-05-423SP. Washington, D.C.: May 6, 2005. Airport and Airway Trust Fund: Preliminary Observations on Past, Present, and Future. GAO-05-657T. Washington, D.C.: May 4, 2005. Highway and Transit Investments: Options for Improving Information on Projects’ Benefits and Costs and Increasing Accountability for Results. GAO-05-172. Washington, D.C.: January 24, 2005. Federal-Aid Highways: Trends, Effect on State Spending, and Options for Future Program Design. GAO-04-802. Washington, D.C.: August 31, 2004. Surface Transportation: Many Factors Affect Investment Decisions. GAO-04-744. Washington, D.C.: June 30, 2004. Transportation-Disadvantaged Populations: Actions Needed to Clarify Responsibilities and Increase Preparedness for Evacuations. GAO-07-44. Washington, D.C.: December 22, 2006. Federal Transit Administration: Progress Made in Implementing Changes to the Job Access Program, but Evaluation and Oversight Processes Need Improvement. GAO-07-43. Washington, D.C.: November 17, 2006. Intercity Passenger Rail: National Policy and Strategies Needed to Maximize Public Benefits from Federal Expenditures. GAO-07-15. November 13, 2006. Freight Railroads: Industry Health has Improved, but Concerns about Competition and Capacity Should Be Addressed. GAO-07-94. Washington, D.C.: October 6, 2006. Commercial Aviation: Programs and Options for the Federal Approach to Providing and Improving Air Service to Small Communities. GAO-06- 398T. Washington, D.C.: September 14, 2006. Public Transportation: New Starts Program Is in a Period of Transition. GAO-06-819. Washington, D.C.: August 30, 2006. Public Transportation: Preliminary Information on FTA’s Implementation of SAFETEA-LU Changes. GAO-06-910T. Washington, D.C.: June 27, 2006. Intermodal Transportation: Challenges to and Potential Strategies for Developing Improved Intermodal Capabilities. GAO-06-855T. Washington, D.C.: June 15, 2006. Commuter Rail: Commuter Rail Issues Should Be Considered in Debate over Amtrak. GAO-06-470. Washington, D.C.: April 21, 2006. Transportation Services: Better Dissemination and Oversight of DOT’s Guidance Could Lead to Improved Access for Limited English-Proficient Populations. GAO-06-52. Washington, D.C.: November 2, 2005. Intermodal Transportation: Potential Strategies Would Redefine Federal Role in Developing Airport Intermodal Capabilities. GAO-05-727. Washington, D.C.: July 26, 2005. Federal-Aid Highways: FHWA Needs a Comprehensive Approach to Improving Project Oversight. GAO-05-173. Washington, D.C.: January 31, 2005. Highway and Transit Investments: Options for Improving Information on Projects’ Benefits and Costs and Increasing Accountability for Results. GAO-05-172. Washington, D.C.: January 24, 2005. Federal-Aid Highways: Trends, Effect on State Spending, and Options for Future Program Design. GAO-04-802. Washington, D.C.: August 31, 2004. Underinflated Tires in the United States. GAO-07-246R. Washington, D.C.: February 9, 2007. Rail Safety: The Federal Railroad Administration Is Taking Steps to Better Target Its Oversight, but Assessment of Results is Needed to Determine Impact. GAO-07-149. Washington, D.C.: January 26, 2007. Aviation Safety: FAA’s Safety Efforts Generally Strong but Face Challenges. GAO-06-1091T. Washington, D.C.: September 20, 2006. Natural Gas Pipeline Safety: Integrity Management Benefits Public Safety, but Consistency of Performance Measures Should Be Improved. GAO-06-946. Washington, D.C.: September 8, 2006. Natural Gas Pipeline Safety: Risk-Based Standards Should Allow Operators to Better Tailor Reassessments to Pipeline Threats. GAO-06- 945. Washington, D.C.: September 8, 2006. Truck Safety: Share the Road Safely Pilot Initiative Showed Promise, but the Program’s Future Success Is Uncertain. GAO-06-916. Washington, D.C.: September 8, 2006. Rail Transit: Additional Federal Leadership Would Enhance FTA’s State Safety Oversight Program. GAO-06-821. Washington, D.C.: July 26, 2006. Federal Motor Carrier Safety Administration: Education and Outreach Programs Target Safety and Consumer Issues, but Gaps in Planning and Evaluation Remain. GAO-06-103. Washington, D.C.: December 19, 2005. Large Truck Safety: Federal Enforcement Efforts Have Been Stronger Since 2000, but Oversight of State Grants Needs Improvement. GAO-06- 156. Washington, D.C.: December 15, 2005. Highway Safety: Further Opportunities Exist to Improve Data on Crashes Involving Commercial Motor Vehicles. GAO-06-102. Washington, D.C.: November 18, 2005. Aviation Safety: FAA’s Safety Oversight System Is Effective but Could Benefit from Better Evaluation of Its Programs’ Performance. GAO-06- 266T. Washington, D.C.: November 17, 2005. Aviation Safety: System Safety Approach Needs Further Integration into FAA’s Oversight of Airlines. GAO-05-726. Washington, D.C.: September 28, 2005. Vehicle Safety: Opportunities Exist to Enhance NHTSA’s New Car Assessment Program. GAO-05-370. Washington, D.C.: April 29, 2005. Highway Safety: Improved Monitoring and Oversight of Traffic Safety Data Program are Needed. GAO-05-24. Washington, D.C.: November 4, 2004. Federal Aviation Administration: Challenges Facing the Agency in Fiscal Year 2008 and Beyond. GAO-07-490T. Washington, D.C.: February 14, 2007. Next Generation Air Transportation System: Progress and Challenges Associated with the Transformation of the National Airspace System. GAO-07-25. Washington, D.C.: November 13, 2006. Next Generation Air Transportation System: Preliminary Analysis of Progress and Challenges Associated with the Transformation of the National Airspace System. GAO-06-915T. Washington, D.C.: July 25, 2006. Air Traffic Control Modernization: Status of the Current Program and Planning for the Next Generation Air Transportation System. GAO-06- 653T. Washington, D.C.: June 21, 2006. Next Generation Air Transportation System: Preliminary Analysis of the Joint Planning and Development Office’s Planning, Progress, and Challenges. GAO-06-574T. Washington, D.C.: March 29, 2006. National Airspace System: Transformation will Require Cultural Change, Balanced Funding Priorities, and Use of All Available Management Tools. GAO-06-154. Washington, D.C.: October 14, 2005. National Airspace System: FAA Has Made Progress but Continues to Face Challenges in Acquiring Major Air Traffic Control Systems. GAO- 05-331. Washington, D.C.: June 10, 2005. Federal Aviation Administration: Stronger Architecture Program Needed to Guide Systems Modernization Efforts. GAO-05-266. Washington, D.C.: April 29, 2005. Aviation Security: TSA’s Staffing Allocation Model Is Useful for Allocating Staff among Airports, but Its Assumptions Should Be Systematically Reassessed. GAO-07-299. Washington, D.C.: February 28, 2007. Aviation Safety: FAA Management Practices for Technical Training Mostly Effective; Further Actions Could Enhance Results. GAO-05-728. Washington, D.C.: September 7, 2005. Human Capital: Agencies Need Leadership and the Supporting Infrastructure to Take Advantage of New Flexibilities. GAO-05-616T. Washington, D.C.: April 21, 2005. Federal-Aid Highways: FHWA Needs a Comprehensive Approach to Improving Project Oversight. GAO-05-173. Washington, D.C.: January 31, 2005. FAA Budget Policies and Practices. GAO-04-841R. Washington, D.C.: July 2, 2004. Federal Aircraft: Inaccurate Cost Data and Weaknesses in Fleet Management Planning Hamper Cost Effective Operations. GAO-04-645. Washington, D.C.: June 18, 2004. Highway Emergency Relief: Reexamination Needed to Address Fiscal Imbalance and Long-Term Sustainability. GAO-07-245. Washington, D.C.: February 23, 2007. Passenger Rail Security: Federal Strategy and Enhanced Coordination Needed to Prioritize and Guide Security Efforts. GAO-07-459T. Washington, D.C.: February 13, 2007. Aviation Security: Progress Made in Systematic Planning to Guide Key Investment Decisions, but More Work Remains. GAO-07-448T. Washington, D.C.: February 13, 2007. Transportation-Disadvantaged Populations: Actions Needed to Clarify Responsibilities and Increase Preparedness for Evacuations. GAO-07-44. Washington, D.C.: December 22, 2006. Passenger Rail Security: Evaluating Foreign Security Practices and Risk Can Help Guide Security Efforts. GAO-06-557T. Washington, D.C.: March 29, 2006. Undeclared Hazardous Materials: New DOT Efforts May Provide Additional Information on Undeclared Shipments. GAO-06-471. Washington, D.C.: March 29, 2006. Passenger Rail Security: Enhanced Federal Leadership Needed to Prioritize and Guide Security Efforts. GAO-05-851. Washington, D.C.: September 9, 2005. General Aviation Security: Increased Federal Oversight is Needed, but Continued Partnership with the Private Sector Is Critical to Long-Term Success. GAO-05-144. Washington, D.C.: November 10, 2004. Transportation Security: Federal Action Needed to Help Address Security Challenges. GAO-03-843. Washington, D.C.: June 30, 2003. Transportation Security Research: Coordination Needed in Selecting and Implementing Infrastructure Vulnerability Assessments. GAO-03- 502. Washington, D.C.: May 1, 2003. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

A safe, efficient, and convenient transportation system is integral to the health of our economy and quality of life. Our nation's vast transportation system of airways, railways, roads, pipelines, transit, and waterways has served this need, yet it is under considerable strain from (1) increasing congestion, (2) the large costs to maintain and improve it, and (3) the human cost of over 44,000 people killed and over 2.5 million injured each year in transportation-related accidents. The Department of Transportation implements national transportation policy and administers most federal transportation programs. For fiscal year 2008, the department has requested $67 billion to carry out these and other activities. While the department carries out some activities directly, such as employing about 15,000 air traffic controllers to make certain that planes stay a safe distance apart, it does not have direct control over the vast majority of activities that it funds, such as local decisions on the priority and placement of airports, public transit, and roads. In other cases, such as railways and pipelines, the infrastructure is owned and operated by industry. This statement presents GAO's views on major transportation challenges facing Congress and the department. It is based on GAO products, including recommendations made, and the products of others. Financing mechanisms for the nation's transportation system are under stress. Our nation's transportation infrastructure is threatened by increasing demand for transportation services, and revenue from traditional funding mechanisms for the nation's highway and aviation systems may be unable to keep pace at current tax rates. In addition, freight traffic is projected to grow substantially, but current planning and financing mechanisms impede public strategies to address needs. Our nation's mobility is threatened because the nation's infrastructure is under great strain. Congestion across modes (e.g., aviation, highways, and rail) is expected to worsen. However, funding by mode and the lack of performance-related goals result in little assurance that funds are being channeled to the most critical mobility concerns and that intermodal approaches can be integrated into the transportation system. Improvements in transportation safety are needed to reduce the number of deaths and injuries from transportation accidents--about 95 percent of which occur on our nation's roads. Increases in congestion across modes as a result of population and economic growth could cause deterioration in transportation safety despite departmental and state efforts to reduce accidents. The transition from the current air traffic control system to a broader and modernized system will be one of the department's most complex undertakings. In previous years, FAA has faced systemic management and acquisition problems that led us to designate its air traffic control modernization program as high risk. While the agency has made significant progress in recent years, a key challenge going forward will be to institutionalize these improvements and to continually improve. In addition, the department and the transportation sector face persistent human capital challenges due to an impending shortage of skilled people to meet changing transportation needs. Furthermore, despite recent improvements in financial management, the department received a qualified opinion on its 2006 financial statements. Finally, the department is working to clarify its role in transportation security and emergency preparedness and response.

The National Cemeteries Act of 1973 (P.L. 93-43) authorized NCS to bury eligible veterans and their family members in national cemeteries. NCS operates and maintains 114 national cemeteries located in 38 states and Puerto Rico. In fiscal year 1996, NCS performed about 72,000 interments and maintained more than two million burial sites and over 5,600 acres of land developed for interment purposes. NCS offers veterans and their eligible family members the options of casket interment and interment of cremated remains in the ground (at most cemeteries) or in columbarium niches (at nine cemeteries). NCS determines the number and type of burial options available at each of its national cemeteries. The standard size of casket grave sites, the most common burial choice, is 5 feet by 10 feet, and the grave sites are prepared to accommodate two caskets stacked one on top of the other. A standard in-ground cremains site is 3 feet by 3 feet and can generally accommodate one or two urns. The standard columbarium niche used in national cemeteries is 10 inches wide, 15 inches high, and 20 inches deep. Niches are generally arrayed side by side, four units high, and can hold two or three urns, depending on urn size. Figure 1 shows a columbarium and in-ground cremains sites at national cemeteries. Armed forces members who die while on active duty and certain veterans are eligible for burial in a national cemetery. Eligible veterans must have been discharged or separated from active duty under other than dishonorable conditions and have completed the required period of service. People entitled to retired pay as a result of 20 years’ creditable service with a reserve component of the armed services are also eligible. U.S. citizens who have served in the armed forces of a government allied with the United States in a war may also be eligible. The benefit of burial in a national cemetery is further extended to spouses and minor children of eligible veterans and of active duty members of the armed forces. A surviving spouse of an eligible veteran who later marries a nonveteran, and whose remarriage is terminated by death or divorce, is also eligible for burial in a national cemetery. Burial in a VA cemetery includes, at no cost to the veteran, one grave site for the burial of all eligible family members. Also included are the opening and closing of the grave, perpetual care of the site, and a government headstone or marker and grave liner. Veterans’ families are required to pay for services provided by funeral directors and additional inscriptions on the headstone or marker. Generally grave sites may not be reserved; space is assigned at the time of need on the basis of availability. In addition to burying eligible veterans and their families, NCS manages three related programs: (1) the Headstones and Markers Program, which provides headstones and markers for the graves of eligible people in national, state, and private cemeteries; (2) the Presidential Memorial Certificates Program, which provides certificates to the families of deceased veterans recognizing their contributions and service to the nation; and (3) the State Cemetery Grants Program, which provides aid to states in establishing, expanding, or improving state veterans’ cemeteries. In 1978, Public Law 95-476 authorized NCS to administer the State Cemetery Grants Program, under which states receive financial assistance to provide burial space for veterans and eligible dependents. State veterans’ cemeteries supplement the burial service provided by NCS. The cemeteries are operated and permanently maintained by the states. A grant may not exceed 50 percent of the total value of the land and the cost of improvements. The remaining amount must be contributed by the state. The State Cemetery Grants Program has funded the establishment of 28 veterans’ cemeteries, including three cemeteries currently under development, located in 21 states, Saipan, and Guam. The program has also provided grants to state veterans’ cemeteries for expansion and improvement efforts. While VA strongly encourages states to adopt the eligibility criteria applied to national cemeteries, states have been allowed to establish eligibility criteria for interments that differ from VA-established criteria, but only if their criteria are more restrictive than those established for national cemeteries. In other words, state veterans’ cemeteries cannot be used for the interment of people who are not eligible for burial in a national cemetery. Most states have a residency requirement, and some states restrict eligibility to veterans who were honorably discharged, had wartime service, or both. As the veteran population ages, NCS projects the demand for burial benefits to increase. NCS has a strategic plan for addressing the demand for veterans’ burials up to fiscal year 2000, but the plan does not tie its strategic and performance goals to external factors such as veterans’ mortality rates and preferences for burial options—that is, caskets, in-ground cremains, or columbaria niches. In addition, NCS’ strategic plan does not address long-term burial needs—that is, the demand for benefits during the expected peak years of veteran deaths, when pressure on the system will be greatest. Beyond the year 2000, NCS officials said they will continue using the basic strategies contained in the current 5-year plan. With the aging of the veteran population, veteran deaths continue to increase each year. For example, NCS projects annual veteran deaths will increase about 20 percent between 1995 and 2010, from 513,000 to 615,000, as shown in figure 2. Moreover, NCS projects that veteran deaths will peak at about 620,000 in 2008. The demand for veterans’ burial benefits is also expected to increase. For example, NCS projects annual interments will increase about 42 percent between 1995 and 2010, from 73,000 to 104,000. NCS projects that annual interments will peak at about 107,000 in 2008. According to its 5-year strategic plan (1996-2000), one of NCS’ primary goals is to ensure that burial in a national or state veterans’ cemetery is an option for all eligible veterans and their family members. The plan sets forth four specific strategies for achieving this goal. First, NCS plans to establish, when feasible, new national cemeteries. NCS is currently establishing five new national cemeteries, which are in various stages of development, and projects that all will be operational by 2000. A second strategy for addressing veterans’ burial demand is to develop available space for cremated remains. NCS plans to survey national cemeteries to determine what space is available for use as in-ground cremains sites, construct additional columbaria at eight existing cemeteries, and include columbaria at the five new cemeteries. Third, NCS plans to acquire land through purchase or donation. NCS plans to use this land to extend the burial capacity and service period of national cemeteries currently projected to run out of available grave sites. Fourth, NCS plans to encourage states to provide additional burial sites for veterans through participation in the State Cemetery Grants Program. According to the plan, NCS plans to identify and prioritize those states most in need of a veterans’ cemetery; design a marketing strategy for those states; visit a minimum of four of those states annually until all prioritized states have been visited; and participate in the state conferences of at least three veterans’ service organizations (for example, the American Legion and the Veterans of Foreign Wars) each year. In addition to the strategic and performance goals, the plan also discusses assumptions, such as veterans’ demographics (the projected increases in veteran deaths and interments), and external factors, such as resource constraints, that could delay achievement of the plan’s performance goals. However, the plan does not tie the strategic and performance goals to its assumptions. For example, while the plan includes some data on demographic trends in the veteran population, it does not explain how these data were used in setting strategic goals, or how they will be used to measure progress in achieving these goals. Neither does the plan tie its strategic and performance goals to external factors—such as preferences for VA, state, or private cemeteries and preferences for casket, in-ground cremains, or columbaria niche burial—that will affect the need for additional VA and state cemetery capacity. NCS tracks actual burial practices in national cemeteries, monitors trends in the private cemetery sector, and in 1992 surveyed veterans to determine their preferences for type of cemetery (national, state, or private) and burial option (casket or cremation burial). Despite NCS plans to ensure that burial in a national or state veterans’ cemetery is an available option, officials acknowledge that large numbers of veterans currently do not have access to a veterans’ cemetery within a reasonable distance of their place of residence. For example, NCS estimates that of the approximately 26 million veterans in 1996, about 9 million (35 percent) did not have reasonable access to a national or state veterans’ cemetery. According to NCS officials, most underserved areas are major metropolitan regions with a high concentration of veterans. With the completion of the five new cemeteries, NCS officials estimate that the percentage of veterans who will have reasonable access to a veterans’ cemetery will increase from about 65 percent in fiscal year 1996 to about 77 percent in fiscal year 2000. Although NCS has a 5-year strategic plan for addressing veterans’ burial demand during fiscal years 1996 through 2000, it is unclear how NCS plans to address the demand beyond 2000. For example, NCS has not developed a strategic plan to address veterans’ burial demand during the peak years of veteran deaths, when pressure on the system will be greatest. According to NCS’ Chief of Planning, although its strategic plan does not address long-term burial needs, NCS is always looking for opportunities to acquire land to extend the service period of national cemeteries. For example, NCS is working to acquire land for one of its west coast cemeteries that is not scheduled to run out of casket sites until the year 2011. Also, to help address long-range issues, NCS compiles key information, such as mortality rates, number of projected interments and cemetery closures, locations most in need of veterans’ cemeteries, and cemetery-specific burial layout plans. In addition, the planning chief pointed out that the Government Performance and Results Act requires a strategic plan to cover only a 5-year period. However, the Results Act allows an agency to extend its strategic plan beyond a 5-year period to address future goals. Although NCS’ strategic plan notes that annual veteran deaths are expected to increase about 20 percent between 1995 and 2010, the plan does not indicate how the agency will begin to position itself to handle this increase in demand for burial benefits. A longer planning period would provide the opportunity to develop strategies for obtaining funds, acquiring land, assessing veterans’ preferences, or all three. While NCS does not have a formal strategic plan to address veterans’ burial demand beyond the year 2000, NCS officials said they will continue using the basic strategies contained in the current 5-year plan. For example, NCS plans to enhance its relationship with states to establish state veterans’ cemeteries through the State Cemetery Grants Program. According to NCS’ Chief of Planning, NCS will encourage states to locate cemeteries in areas where it does not plan to operate and maintain national cemeteries. Since the State Cemetery Grants Program’s inception in 1978, fewer than half of the states have established veterans’ cemeteries primarily because, according to NCS officials, states must provide up to half of the funds needed to establish, expand, or improve a cemetery, as well as pay for all equipment and annual operating costs. Furthermore, the Director of the State Cemetery Grants Program told us that few states, especially those with large veteran populations, have shown interest in legislation that VA proposed in its 1998 budget submission in order to increase state participation. This legislation would increase the federal share of construction costs from 50 to 100 percent and permit federal funding for up to 100 percent of initial equipment costs. In fact, according to the Director, state veterans’ affairs officials said that they would rather have funding for operating costs than for construction. In addition, VA does not plan to request construction funds for more than the five new cemeteries, which will be completed by the year 2000, because of its commitment to deficit reduction. Officials said that even with the new cemeteries, interment in a national or state veterans’ cemetery will not be “readily accessible” to all eligible veterans and their family members. According to NCS officials, most underserved areas will be major metropolitan areas with high concentrations of veterans, such as Atlanta, Georgia; Detroit, Michigan; and Miami, Florida. As demand for burial benefits increases, cemeteries become filled, thus reducing the burial options available to veterans and their families. We developed a model to analyze the relative costs of three types of cemeteries. The analysis showed that over 30 years, the traditional casket cemetery would be the most expensive interment option. Our analysis also showed that there would be no significant difference in the costs of columbarium and in-ground cremains cemeteries. Although the development and construction costs are higher for a columbarium cemetery, operating costs are higher for an in-ground cremains cemetery. Table 1 compares the 30-year costs of these three types of cemeteries. (See app. II for a detailed cost comparison of the three types of cemeteries.) A cemetery providing only casket burials would be the most expensive interment option, costing, on average, over twice as much as columbarium or in-ground cremains cemeteries. We estimated that over a 30-year period, the casket cemetery would cost over $50 million, compared with about $21 to $23 million for either of the two cremation cemeteries. The difference in costs is due primarily to the higher land development and operations/maintenance costs of a casket cemetery. Specifically, providing 50,000 grave sites for 30 years would require developing about 115 acres at a cost of $8.4 million, compared with 34 acres for an in-ground cremains cemetery and 14 acres for a columbarium cemetery, costing about $2.5 million and $1 million, respectively. Over 30 years, the total operations and maintenance cost for a casket cemetery is three times as much as that for a columbarium cemetery and over twice as much as that for an in-ground cremains cemetery. As table 1 shows, providing burial services and maintenance activities for a 115-acre casket cemetery would result in higher nonlabor and labor costs. For example, it requires about 39 full-time staff to operate and maintain a casket cemetery, compared with about 21 full-time staff for an in-ground cremains cemetery and 14 full-time staff for a columbarium cemetery. Over 30 years, it would cost about the same to plan, design, construct, operate, and maintain a columbarium and an in-ground cremains cemetery with 50,000 burial spaces: $23 and $21 million, respectively. The development and construction cost is higher for a columbarium cemetery, but its operations and maintenance cost is lower than that of an in-ground cremains cemetery. As table 1 shows, over 30 years the development and construction cost for a columbarium cemetery would be, on average, about three times as much as that for an in-ground cremains cemetery. This difference in costs is primarily due to the cost of building the columbarium structure. The operations and maintenance cost of an in-ground cremains cemetery is almost twice as much as that of a columbarium cemetery. This cost difference can be attributed to the fact that columbarium cemeteries have fewer acres to maintain, resulting in lower nonlabor and labor costs. As existing national cemeteries reach their capacity, columbarium burial offers the most efficient option for extending cemetery service periods. We developed a model to analyze the cost of three interment options on the basis of the cost of developing a total of 1 acre of land, composed of parcels of land not contiguous to each other, in a cemetery nearing exhaustion of available casket grave sites. The analysis showed that the average burial cost would be lowest and the service delivery period the longest using columbarium interment. The analysis also showed that the average cost per burial would be about the same for columbarium niches as for in-ground cremains sites. However, columbarium interment would extend the service period by about 50 years, while in-ground cremains interment would extend the service period about 3 years and casket burials, about half a year. Casket burials would be the most expensive per burial and would have the shortest service period. At the end of fiscal year 1996, 57 of VA’s 114 national cemeteries had exhausted their supply of casket grave sites available to first family members, as shown in figure 3. Of these 57 cemeteries, 38 could accommodate casket burial of subsequent family members and interment of cremated remains of both first and subsequent family members. Nineteen could accommodate only subsequent family members—for either casket or cremated remains interment. According to NCS’ Chief of Planning, unless NCS acquires additional land, it projects that 15 cemeteries will totally deplete their inventory of casket grave sites for first family members by 2010, and another 16 cemeteries will do so by 2020. In total, by 2020, NCS projects that 88 of the 119 national cemeteries (74 percent) will no longer be able to accommodate casket burials of first family members. As less burial space is available, columbarium burial offers the most efficient interment option for extending the service period of existing cemeteries. Our analysis of the costs of three interment options, based on the development of 1 remaining acre of land, pieces of which were not contiguous to each other, showed that the average burial cost would be lowest using columbarium interment. For example, the average columbarium interment cost would be about $280, compared with about $345 for in-ground cremains burial and about $655 for casket burial, as shown in figure 4. Our analysis also showed that the service delivery period would be extended the most using the columbarium. For example, a total of 1 acre of land could accommodate about 87,000 columbarium niches and could extend the service delivery period for over 52 years, compared with about 3 years for about 4,800 in-ground cremains sites and about 1/2 year for about 870 casket sites, as shown in figure 5. Although NCS officials acknowledge that the columbarium option could extend the service delivery period of existing cemeteries, they said that it has been used to do so at only one national cemetery, which is located on the west coast. Furthermore, at the end of fiscal year 1996, only 9 of the 114 national cemeteries offered interment in a columbarium, while the majority of cemeteries provided casket and in-ground cremains sites. According to NCS officials, NCS has not made greater use of columbaria primarily because of their substantial up-front construction costs. Officials said they generally develop casket and in-ground cremains sites first because they believe the initial costs are less. However, our analysis showed that the total cost per burial would be lower for a columbarium because of its low operations and maintenance costs. Columbaria would be particularly useful in metropolitan areas where interment rates are high; past or projected cremation demand is significant; land is scarce, expensive, or both; and no state veterans’ cemetery exists to compensate for the lack of available national cemetery grave sites. For example, at one midwestern cemetery, NCS plans to add about 8,000 casket sites, but no cremation sites, to its last acres. With the additional casket sites, the cemetery is projected to deplete all burial spaces about the time veteran deaths peak, and no state veterans’ cemetery exists to compensate for the lack of burial spaces. However, by incorporating columbaria into 1/2 acre of land, this cemetery could continue to provide a burial option to thousands of additional veterans, who otherwise would have no burial option available to them within a reasonable distance of their homes, and keep the cemetery open well beyond the peak years. While historical data imply that the majority of veterans and eligible dependents prefer a casket burial, NCS national data show that the demand for cremation at national cemeteries is increasing. For example, while about 70 percent of veterans prefer a casket burial, veterans choosing cremation increased from about 20 percent of the veteran population in 1990 to nearly 30 percent in 1996, and NCS officials expect demand for cremation to continue to increase in the future. At cemeteries offering both types of interments, the ratio of casket to cremation interments varies significantly. For example, cremation accounts for over 40 percent of interments at some cemeteries and less than 5 percent at others. In addition, according to cemetery directors, veterans choosing cremation do not strongly prefer either in-ground burial or interment in a columbarium niche. The incidence of cremation also continues to increase in the general population. For example, cremation was chosen for about 14 percent of nationwide burials in 1985 and about 21 percent in 1995. The Cremation Association of North America (CANA) projects that cremations will account for about 40 percent of all burials by 2010. Like other interment options, cremation is an individual’s decision and is subject to influences such as culture, religion, geographic area of the country, and age and generational preferences. According to CANA, people choose cremation primarily because it is perceived as less expensive and simpler than traditional casket burial, it uses less land, and it offers more options for memorialization. Long-range planning is crucial to addressing veterans’ burial needs during the peak years and beyond. Although NCS has a 5-year strategic plan, it does not address veterans’ burial needs beyond the year 2000, when the demand for burial benefits will be greatest. Specifically, while the World War II veteran population is entering its peak years of need, many national cemeteries are depleting their inventory of available casket grave sites. As a result, additional burial sites are needed to help meet future burial demand. In some cases, state veterans’ cemeteries could reduce the negative impact of the loss of available casket spaces from a national cemetery. However, it does not appear that state veterans’ cemeteries will be able to accommodate all veterans seeking interment. Therefore, NCS needs to rely more on extending the service periods of its existing national cemeteries. Columbaria can more efficiently utilize available cemetery land at a lower average burial cost than the other interment options and can also extend the service period of existing national cemeteries. Using columbaria also adds to veterans’ choice of services and recognizes current burial trends. Although cremation will not be the preferred burial option for all veterans, identifying veterans’ burial preferences would enable NCS to better manage limited cemetery resources and more efficiently meet veterans’ burial needs. To better serve the American veteran, we recommend that the Secretary of Veterans’ Affairs instruct the director of the National Cemetery System to extend its strategic plan to address veterans’ long-term burial demand during the peak years of 2005 to 2010; collect and use information on veterans’ burial preferences to better plan for future burial needs; and identify opportunities to construct columbaria in existing cemeteries, for the purpose of increasing burial capacity and extending the cemeteries’ service periods. In commenting on a draft of this report, the Director of NCS stated that our recommendations appeared valid and represented the vision and performance of NCS in meeting the burial needs of veterans. He also said that NCS is currently executing many of the practices recommended by our report. For example, the NCS Director concurred with our recommendation that NCS develop plans to address veterans’ long-term burial demand during the peak years and stated that NCS is already performing long-term planning, as evidenced by numerous strategies and activities. We recognize that NCS has developed valuable information from such sources as the Management and Decision Support System and cemetery master plans to help it address long-range issues, but even with this information, NCS is unable to specify the extent to which veterans will have access to a national or state veterans’ cemetery during the peak years. NCS’ estimates of the percentage of veterans who will have access to a veterans’ cemetery stop at the year 2000. NCS needs to develop a strategic plan that links information such as mortality rates and the number of projected interments and cemetery closures, obtained from various sources, to its strategic goals, performance measures, and mitigation plans over the next 15 years. For example, one of NCS’ goals is to ensure that a burial option is available to all eligible veterans. Although NCS’ current strategic plan estimates a 20-percent increase in annual veteran deaths between 1995 and 2010, it does not indicate how NCS will begin to position itself to handle this increase in demand for burial benefits. Because of the lead time required to acquire land and develop some types of interment spaces, NCS needs to develop strategies that address such issues as (1) how many burial spaces will be needed at each cemetery to accommodate the projected demand for burial benefits during the peak years; (2) how NCS will acquire the additional burial spaces—for example, by purchasing adjacent land or maximizing existing land by using columbaria; and (3) when and how NCS will obtain funds, acquire land, and assess veteran preferences. In addition, while one of NCS’ strategies for meeting the projected burial demand includes encouraging states to build cemeteries, the Director of the State Cemetery Grants Program told us that few states, especially those with large veteran populations—such as New York, Florida, Texas, Ohio, and Michigan—would be swayed by proposed legislation that would increase the federal share of construction and equipment costs. NCS officials also acknowledged that their ability to persuade states to participate in the program is limited, because the states must take the initiative to request grant funds. We revised our previous recommendation to encourage NCS to extend its strategic plan to address veterans’ long-term burial demand during the peak years of 2005 to 2010. The NCS Director also concurred with our recommendation to collect and use information on veterans’ burial preferences to better plan for future burial needs. While the Director stated that NCS carefully tracks actual burial practices in national cemeteries and monitors trends in the private cemetery sector, and that these indexes offer a reliable method of planning for the future, he said that additional data on veterans’ preferences would assist NCS in its planning efforts. Therefore, he stated that NCS will include questions pertaining to personal burial preferences in the next VA National Survey of Veterans. Finally, the Director of NCS concurred with our recommendation to identify opportunities to construct columbaria in existing cemeteries for the purpose of increasing burial capacity and extending the service delivery period of these cemeteries. He asserted that NCS is already accomplishing what our recommendation was intended to achieve in that it (1) plans to add columbaria at eight existing cemeteries and five new cemeteries and (2) annually considers all sites that may warrant the establishment of columbarium units. We acknowledge, as stated in our report, that NCS plans to add columbaria at 8 of the 114 existing national cemeteries and include columbaria in its 5 new cemeteries. However, the intent of our recommendation was to encourage VA to identify opportunities to construct columbaria in cemeteries that are nearing depletion of casket grave sites for first family members or have already run out. This will involve at least 72 cemeteries by 2010. Although NCS acknowledges that columbaria could extend service at a cemetery that would otherwise be closed to veteran use, they have only been used for this purpose at one national cemetery. While the NCS Director stated in his comments that NCS considers the anticipated ratio of casket burial to cremains burial when planning for the future, during our review, NCS officials stated that they primarily use historical usage data. For example, at one cemetery, NCS planned to allocate more than 30 percent of the burial spaces for cremation sites, although the cremation rate for the state in which the cemetery was located was more than 50 percent in 1995, and projected to increase to more than 60 percent in 2000 and to about 80 percent in 2010. As our report states, by including other factors in the decision process, such as projected cremation demand, availability and cost of land, and availability of grave sites at state veterans’ cemeteries, officials may identify additional national cemeteries that warrant the establishment of columbaria. NCS also provided technical comments in an attached white paper. Comments 1 through 3 repeat points made in the letter. Comments 4 and 5 question the results of our analysis of the cost of extending the service period of existing cemeteries, since it was based on the maximum number of burial sites available in an acre of land. Specifically, NCS commented that it may not be feasible to devote a single 1-acre plot entirely to columbarium niches because using the “absolute maximum” would not allow space between structures. However, in our analysis we did not envision a single 1-acre plot. Rather, we assumed several parcels of land dispersed around the cemetery that totaled 1 acre of available burial space. Accordingly, we have revised our discussion to clarify this issue. Comment 6 questions our assumption that first family member interments would be evenly spaced over 30 years for all three modes of burial. Specifically, NCS suggests an analysis in which the annual interment rates are assumed to differ for the three alternatives (casket, in-ground cremains, and columbarium burials), reflecting current use patterns. However, our objective was to perform a cost comparison. For a valid cost comparison, the alternatives being compared must be evaluated in terms of the same outcome—in this case, to inter a given number of eligible veterans and their dependents according to a given schedule. The specific assumption we adopted—evenly spaced first family member interments for all alternatives—was previously suggested to us by NCS, and our analysis is similar to the one NCS used in its 1996 study. The type of analysis that NCS is now suggesting is outside the scope of our work. NCS offered other technical comments, which we incorporated where appropriate. NCS’ comments are included in their entirety in appendix III. We are sending copies of this report to the Secretary of Veterans Affairs and other interested parties. This work was performed under the direction of Irene Chu, Assistant Director. If you or your staff have questions about this report, please contact Ms. Chu or me on (202) 512-7101. Other major contributors to this report are listed in appendix IV. In this appendix we discuss the methodology, data sources, and principal assumptions that we used to characterize the relative long-term cost of each of three modes of interment: casket, in-ground cremains, and columbarium; project the outlays that would be required to construct and operate a cemetery that offers each of these modes of interment over a period of 30 years or more; and estimate the cost of these three types of interment on the basis of the development of a total of 1 acre of land composed of parcels of land not contiguous to each other in a cemetery nearing depletion of available burial sites. Our analysis builds on a study that the National Cemetery System (NCS) performed at the request of the Chairman, Subcommittee on Compensation, Pension, Insurance and Memorial Affairs, in February 1996. In that study, the Department of Veterans Affairs (VA) presented an analysis of the relative costs of casket and columbarium burial over a 20-year period. For the purpose of this report, we have updated and extended the NCS analysis, most notably by adding in-ground cremains burial as a third alternative, as requested by the analyzing costs over 30 years or more, thus recognizing that cost differences among the modes of interment will persist far into the future; analyzing the relative long-term costs of the three alternatives in the context of using available space in existing cemeteries, as well as in the context of developing new cemeteries; and using the present value method to evaluate the relative long-term costs of the three alternatives. Simple comparisons of cumulative outlays for the several modes of interment (casket, in-ground cremains, and columbarium) would provide a misleading picture of the relative costs of the respective options because the modes differ in the relative share of total cost that is incurred in the first years. Moreover, a dollar paid by the government today is more costly than a dollar paid at some future date, because it increases the burden of making interest payments on the national debt. It is standard practice among policy analysts to compare different payment streams by calculating the present value (also known as the lump-sum equivalent) of each stream. We developed two models. The first model was used to estimate the long-term cost of alternative burial modes in a new cemetery. The second model was used to estimate the long-term cost of alternative uses of available space in an existing cemetery. Each model consisted of three basic components: simulating the sequence of events whereby a cemetery is opened and burial sites are developed, placed into service, and maintained; attaching estimated costs to each of these events, so as to create a trajectory of costs over the whole time period; and calculating the present values of cost streams associated with each of the options being evaluated. Assumptions and Data We developed the assumptions and specified the data to be collected in consultation with NCS experts. Except as noted below, NCS officials supplied the data. We did not verify all of the data. What follows is, first, a description of the elements of the model for the analysis of the costs of a new cemetery designed for 50,000 burial sites, with burials to take place over a 30-year period. Second, we describe how we modified the data and assumptions for the second model, which analyzes the cost of adding to an existing cemetery. Land acquisition. We assumed that all land acquisition and development of architectural master plans and environmental impact statements would occur in the first year. Development of burial sites. NCS officials told us that burial sites would be developed in three phases, each of which would result in one-third (about 16,700) of the total number of burial sites. The first phase would occur in the second and third years. The second phase would occur in the eleventh through thirteenth years. The third phase would take place in the twenty-first through twenty-third years. Each of the three phases would involve outlays for design, land development, and equipment acquisition (see below). The construction of buildings would occur during the first two phases. First family member interments. Per NCS guidance, we assumed that first family member interments would commence in the fourth year and that they would be evenly spaced over the next 30 years (that is, there would be 1,667 first family member interments per year). Subsequent interments. We used the assumption, supplied by NCS officials, that subsequent interments would initially make up 2 percent of first family member interments and would increase linearly over time, so that in the thirtieth year (that is, the thirty-third year of the period of analysis), subsequent interments would make up 60 percent of first interments. These costs include the cost of site acquisition, site development (conducting environmental impact assessments, obtaining architect/engineer design services, and developing land), and construction of buildings (administration and maintenance facilities). Site acquisition. According to NCS officials, land in the vicinity of the Tahoma National Cemetery costs $10,000 per acre. They told us that a cemetery exclusively devoted to casket burial would require 114.8 acres, of which 57.4 acres would be used for grave sites and 57.4 acres for infrastructure (parking lots, driveways, buildings, landscaping, and so on). A cemetery devoted exclusively to in-ground cremains burial would require 34.3 acres (10.3 acres for burial sites and 24.0 acres for infrastructure). An all-columbarium cemetery would require 14.25 acres (0.57 acre for columbaria and 13.68 acres for infrastructure). Site development. The estimated cost for the environmental assessment aspect of site development is $100,000 for a casket cemetery, $17,150 for an in-ground cremains cemetery, and $7,250 for a columbarium cemetery. These estimates reflect NCS’ experiences with similar projects in the past. The architect/engineer design cost category covers such services as carrying out a topographic survey, an archeological exploration, and traffic impact studies. The cost of architect/engineer design services is assumed to be proportional to construction costs (land development plus buildings). The estimated cost of these services for phase 1 is $545,414 for the casket alternative, $246,249 for in-ground cremains sites, and $862,233 for columbaria. For phases 2 and 3, costs would be lower. Land development costs include site preparation (for example, grading; landscaping; and providing irrigation, roads, storm drainage, and utilities) and purchasing site furnishings (for example, benches and flagpoles). The estimated cost of land development is $102,298 per acre for all modes of interment. Thus, land development costs for the three alternatives are proportional to their respective acreage requirements, discussed above. Under each alternative, one-third of the total acreage would be developed in each of the three phases (years 2 through 3, 12 through 13, and 22 through 23). For a casket cemetery, outlays would amount to $3.91 million in each phase. For an in-ground cremains cemetery, the estimated cost is $1.17 million per phase. For a columbarium cemetery, the estimated cost is $0.49 million per phase. Construction of buildings. Buildings that would be constructed in phase 1 include a public information building, an administration building, a maintenance building, a vehicle storage building, and two committal service shelters. An additional committal service shelter would be constructed in phase 2. The three alternatives have different requirements for the size of the maintenance and vehicle storage buildings. Columbaria niches would be constructed in each phase, giving this mode the highest total construction cost. These costs include (1) the cost of purchasing initial and subsequent equipment; (2) salary and benefits for personnel to handle administration and interment issues (drafting contracts and correspondence; handling public inquiries, ceremonies, and outreach; scheduling burial services; opening/closing grave sites or niches; interring casket or cremated remains; setting headstones or placing markers; and restoring burial sections); (3) the cost of purchasing nonlabor items (fertilizer, seeds, headstones, markers, and grave liners); and (4) the cost of maintenance activities (keeping the grounds and facilities). Equipment. VA provided estimates of the equipment costs for the three modes. The initial costs were $736,674 for caskets, $443,003 for in-ground cremains sites, and $91,664 for columbaria—all purchased in year 3 of the first phase. Subsequent equipment purchases were assumed to be equal and to occur in year 3 of phases 2 and 3. We estimated their cost at $150,000 for caskets, $90,000 for in-ground cremains sites, and $18,000 for columbaria. Labor associated with administration and interments. We assumed that it would require 7.3 full-time-equivalent (FTE) general schedule (GS) employees, at an annual rate (pay and benefits) of $45,216 each, plus 6.7 FTE wage grade (WG) employees at a rate of $35,085 each, to conduct the 1,667 interments that are projected for each year under all three burial modes. VA said that the GS administrative and interment requirements would be the same for all three modes but that the WG labor associated with each mode would vary. According to NCS assumptions, the WG labor required for casket burials was 6.7 FTEs. We had to develop our own estimate—3 FTEs for in-ground cremains sites and .56 FTE for columbarium niches—because VA had no specified ratio for WG labor for the noncasket modes. We assumed subsequent interments would require a prorated amount of labor. That is, if subsequent interments in a given year are estimated to be 20 percent of first interments, we assumed that labor costs associated with subsequent interments would be equal to 20 percent of the labor costs associated with first interments. Put differently, we assumed that each subsequent interment would require as much labor as each first interment. Nonlabor costs. These costs include the costs of irrigating and purchasing fertilizer, seed, and other supplies. We used VA estimates to derive amounts for this category of costs. The amounts are small and proportional to the acreage developed. For the casket model, the nonlabor costs would be $389,000 in phase 1, increasing by $95,500 in phases 2 and 3 to a total of $580,000 by the 24th year. For in-ground cremains sites, we adjusted the cost in phase 1 by the ratio of acreage to arrive at a cost of $117,000, rising by $28,500 in phases 2 and 3 to a total of $174,000 in the 24th year (with rounding). For columbaria, the initial nonlabor cost was $57,000, rising $14,000 in phases 1 and 2 to a total of $85,000 in years 24 through 33. Outlays for headstones and markers are proportional to the number of first interments in a given year. These costs vary depending on the area of the country in which the headstones and markers are purchased. For this analysis, we used the middle price in the range of prices VA said they pay. For a casket burial, we assumed a headstone cost of $120; for an in-ground cremains burial, we assumed a grave marker cost of $70; and for a columbarium burial, we assumed a niche cover cost of $15. Casket burials require grave liners, at an estimated cost of $240 apiece. Labor associated with maintenance. VA uses the standard of 1 FTE per 10.7 developed acres for casket cemeteries. Using this ratio, under the casket scenario, we estimated that maintenance of developed acreage would require 3.5 WG FTEs during phase 1 (years 4 through 13), 7 FTEs during phase 2 (years 14 through 23), and 10.5 FTEs during phase 3 (years 24 through 33), at the annual pay rates stated above. We adjusted these WG labor requirements for the fewer acres in the other modes. For in-ground cremains burials, we estimated that maintenance of developed grave sites would require 1.1 FTEs during phase 1 and an additional 1.1 FTEs during phases 2 and 3. For columbaria, we estimated that maintenance of developed grave sites would require .4 FTE during phase 1, .9 FTE during phase 2, and 1.3 FTEs during phase 3. Further, there would also be labor costs associated with the maintenance of burial sites that have already been placed in service (that is, in which there has been a first family member interment). VA uses an estimate of 1 FTE per 7,844 developed grave sites in its planning for new cemeteries. Using this ratio, it would require about .2 FTE a year for the 30-year burial period in a casket cemetery. We adjusted this amount to reflect the lesser acreage of the other modes. For in-ground cremains sites, .04 FTE per year would be required; for columbaria, .002 FTE would be required. The cost differences among the three alternatives are proportional to the differences in the number of burial acres (as opposed to infrastructure acres) that each alternative requires. For each alternative, grave site maintenance costs would increase linearly for each succeeding year, because we assumed that the same number of first family member interments (1,667) would take place each year. We also analyzed the relative long-term cost of each of the three alternatives as it applied to extending the service period of an existing cemetery. For this model, we adopted the same assumptions, and used the same data, as for the model we used to analyze the long-term cost of a new cemetery, with the following modifications: We assumed the existence of an acre of land that had already been acquired—an acre composed of parcels of land that were not contiguous to each other—so that the cost of land acquisition was zero for all three alternatives. Similarly, we assumed that such costs as environmental assessment, architect/engineer design, land development, and construction of administration and maintenance buildings had already been incurred for the casket and in-ground cremains site estimates. We assumed that for columbaria, it would be necessary to incur the cost of constructing a set of niches, including architect/engineer design costs. For each of the three alternatives, we assumed that a total of 1 acre of land, pieces of which were not contiguous to each other, could be devoted to burial sites. That is, we assumed that the cemetery’s infrastructure (for example, roads) was complete and that there were no other obstacles (such as irregular topography) to the full use of the acre for burial sites. Thus, we assumed the theoretical maximum number of interment sites: 871 for caskets; 4,840 for in-ground cremains sites; and 87,000 for columbaria. Only costs that are incurred up to the time that the acre is closed to further first family member interments are accounted for. Because, as noted above, each of the three alternatives permits a different number of interment sites per acre, and because we are assuming that first family member interments will take place at a rate of 1,667 per year, the time at which the acre’s first family member interment sites are full will be different under the three alternatives (0.52 years for caskets; 2.9 years for in-ground cremains sites; and 52.2 years for columbaria). This simplifying assumption leads to an understatement of the cost of casket burial relative to that of the other alternatives, all other things equal. Future changes in cost factors. All costs are expressed in 1997 dollars. We assumed that although the costs of labor and materials could rise in the future, the relative prices would remain unchanged. Discount rate. We used a (real) discount rate of 3.21 percent. This rate is based on (1) a (nominal) long-term cost to the government of borrowing 6.71 percent, as represented by the interest rate on 30-year Treasury securities as of June 1997, and (2) a long-term inflation rate projection of 3.5 percent that was prepared by the Social Security Administration (SSA). Period of analysis. As agreed with your office, we analyzed cost data over a period that ends 30 years after the first interments (that is, 33 years), at which time the cemeteries are assumed to be full. Ideally, a cost analysis would consider the entire useful life of the project, given that differences in operating costs among the three modes of interment would persist even if there was no new development of burial sites or new first family member interments. For a cemetery, this time period is indefinite. Accordingly, we performed a sensitivity analysis in which the present value of costs for the three modes of interments was evaluated over a period of 53 years (that is, until 20 years had elapsed since the last first family member interments). We found that when costs were evaluated over the longer period, the cost would be $58.4 million for casket burial, $24.1 million for in-ground cremains burial, and $24.8 million for columbarium burial. The differences between costs for the 33-year and 53-year periods reflect differences in operating costs across the three modes of interment, especially the fact that columbaria would require far less costly maintenance than the other two types of interment. We provided information on a cemetery providing only casket interment, another providing only interment of cremated remains in columbarium niches, and a third providing interment of in-ground cremated remains. For each type of cemetery, this appendix provides 30-year undiscounted and present value cost estimates in 1997 dollars for development and construction and operations and maintenance. We also projected the cash outlays that would be required to construct and operate a cemetery that offered each of these modes of interment over a 30-year period (see fig. II.1). Costs were based on actual figures obtained from the most recent NCS construction project—Tahoma National Cemetery. The following tables present detailed data for each type of cemetery we analyzed. Table II.1: Cost Summary for a Cemetery Offering Only Casket Burial Not applicable. Nonlabor costs include the cost of purchasing such items as grass seed, pest control, grave liners, and headstones or markers. Not applicable. Nonlabor costs include the cost of purchasing such items as grass seed, pest control, and niche covers. Not applicable. Nonlabor costs include the cost of purchasing such products as grass seed, pest control, and markers. Donald C. Snyder, Assistant Director (Economist), (202) 512-7204 Jaqueline Hill Arroyo, Evaluator-in-Charge, (202) 512-6753 Jeffrey Pounds, Evaluator Timothy J. Carr, Senior Economist The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed the Department of Veterans Affairs' (VA) National Cemetery System (NCS), focusing on: (1) NCS' plans for addressing veterans' future burial demands; (2) the relative 30-year costs of three types of cemeteries: casket-only internment, cremated internment in columbarium niches, and in-ground internment of cremated remains; and (3) what NCS can do to extend the service period of existing national cemeteries. GAO noted that: (1) NCS projects that demand for veterans' burial benefits will increase; (2) NCS has adopted a 5-year strategic plan with the goal of ensuring that burial in a national or state veterans' cemetery is an available option for all veterans and their eligible family members; (3) strategies outlined in NCS' plan include: (a) establishing five new national cemeteries; (b) developing available space for cremated remains; (c) acquiring contiguous land at existing cemeteries; and (d) encouraging states to provide additional burial sites through participation in the State Cemetery Grants Program; (4) the strategic plan does not tie its goals to external factors, such as the mortality rate for veterans and veterans' relative preferences for burial options, that will affect the need for additional cemetery capacity; (5) it is unclear how NCS will address burial demand during the peak years when pressure on it will be greatest, since NCS has not developed a strategic plan for beyond 2000; (6) according to NCS' Chief of Planning, beyond 2000, NCS will continue using the basic strategies outlined in its current 5-year plan; (7) NCS plans to encourage states to establish veterans' cemeteries in areas where it does not plan to operate national cemeteries; (8) fewer than half of the states have established veterans' cemeteries; (9) states also have shown limited interest in a legislative proposal to increase state participation by increasing the share of federal funding; (10) GAO estimated the present value of the costs of three types of cemeteries, each with 50,000 burial sites, over a 30-year period; (11) planning, designing, constructing, and operating a cemetery of casket grave sites and no other burial options would be the most expensive interment option available; (12) the costs for a cemetery that offered only a columbarium and one that offered only in-ground cremains sites would be about the same; (13) while the cost of a casket-only cemetery would be over $50 million, the cost of a cremains-only cemetery would be about $21 million; (14) while the majority of veterans and eligible family members prefer a casket burial, cremation is an acceptable interment option for many, and the demand for cremation continues to increase; (15) as annual internments increase, cemeteries will reach their burial capacity, increasing the importance of making the most efficient use of available cemetery space; and (16) GAO's analysis of three interment options showed that columbaria offer the most efficient interment option because they would involve the lowest average burial cost and would significantly extend a cemetery's service period.

The F/A-22 is an air superiority aircraft with advanced features to make it less detectable to adversaries (stealth characteristics) and capable of high speeds for long ranges. It is being developed under contracts with Lockheed Martin Corporation for the aircraft and Pratt & Whitney Corporation for the engine. Because of potential cost increases, the Air Force established a team—the Joint Estimating Team—to review the total estimated cost of the F/A-22 program in 1996. This team reported that the cost of the F/A-22 production program could grow by $13.1 billion from the amount planned. In response to identified cost growth, Congress, in the National Defense Authorization Act for Fiscal Year 1998 established cost limits for the development and production phases of the F/A-22 program. The current production cost limit is $37.5 billion. “Air superiority” is the degree of air dominance that allows the conduct of operations by land, sea, and air forces without prohibitive interference by enemy aircraft. Defense (OSD) and the Air Force, disagreed over how many aircraft could be purchased for $43 billion. OSD believed that only 297 aircraft could be purchased for $43 billion while the Air Force believed 333 aircraft could be purchased for the same amount. DOD informed Congress of these divergent viewpoints in September 2001. The F/A-22 President’s Budget for fiscal year 2004 would transfer $876 million in production funding and reduce the number of aircraft to 276 to help fund estimated cost increases in development. As a result, the current production cost estimate is $42.2 billion, an amount that still exceeds the cost limit of $37.5 billion. To fully offset the $13.1 billion in estimated cost growth, the Air Force and contractors designed cost reduction plans. Since 1997, the Air Force has been identifying and implementing these plans. (See appendix IV for a list of the major categories of cost reduction plans designed to offset the cost growth estimated in 1997.) A direct relationship cannot be established between the cost reduction plans and specific areas of cost growth. The reason is that the plans generally offset cost growth in broad areas by enhancing production technology, improving manufacturing techniques, and improving acquisition practices. F/A-22 cost reduction plans are categorized as either “implemented” or “not yet implemented.” The Air Force’s and contractors’ criteria for determining if a cost reduction plan is implemented include whether the contractor has submitted a firm, fixed price proposal that recognizes the impact of the cost reduction; the impact of the reduction has been reflected in a current contract price or negotiated in an agreement; or the contractor has reduced the number of hours allocated to a task. Currently, $14 billion in cost reduction plans is considered “implemented.” Cost reduction plans are categorized as “not yet implemented” if the plans are well defined but none of the criteria listed above are met. Table 3 in appendix II shows the amounts the Air Force currently considers “implemented” and “not yet implemented.” Over the last 6 years, $17.7 billion in estimated production cost growth has been identified during the course of two program reviews. As a result, the estimated cost of the production program currently exceeds the congressional cost limit despite the establishment of cost reduction plans designed to offset a significant amount of this estimated cost growth. The effectiveness of these cost reduction plans has varied. During a review in 1997, the Air Force estimated cost growth of $13.1 billion. The major contributing factors to this cost growth were inflation, increased estimates of labor costs and materials associated with the airframe and engine, and engineering changes to the airframe and engine. These factors made up about 75 percent of the cost growth identified in 1997. (See appendix III for a complete list of cost growth categories identified in 1997.) In August 2001, DOD estimated an additional $5.4 billion in cost growth for the production of the F/A-22, bringing total estimated production costs to $43 billion. The major contributing factors to this cost growth were again due to increased labor costs and airframe and engine costs. These factors totaled almost 70 percent of the cost growth. According to program officials, major contractors’ and suppliers’ inability to achieve the expected reductions in labor costs throughout the building of the development and early production aircraft has been the primary reason for estimating this additional cost growth. (See appendix VI for a complete list of the categories and sources of cost growth identified in 2001.) The effectiveness of cost reduction plans has varied. The Air Force was able to implement cost reduction plans and offset cost growth in the first four production lot contracts awarded. Air Force projections for cost reduction plans show that expected offsets are also planned for the future production lot contracts to enable the production program to be completed within the current production cost estimate. However, the Air Force has not fully funded production improvement programs (PIPs), which are designed to offset cost growth by improving production processes. Consequently, planned offsets may not be achieved in the amount expected. The Air Force was able to implement cost reduction plans and offset cost growth in the first four production contracts awarded. The total offsets for these contracts slightly exceeded earlier projections by about $0.5 million. Table 1 compares previous planned offsets with implemented cost reduction plan offsets in the first four production contracts. Cost reduction plans exist but have not yet been implemented for subsequent production lots planned for fiscal years 2003 through 2010 because contracts for these production lots have not yet been awarded. If implemented successfully, the Air Force expects these cost reduction plans to achieve billions of dollars in offsets to estimated cost growth and allow the production program to be completed within the current production cost estimate of $43 billion. However, as we noted earlier in this report, this amount exceeds the congressionally imposed production cost limit of $37.5 billion. A production improvement program is a type of cost reduction plan whereby the government must make an initial investment to realize savings. The earlier the Air Force implements PIPs, the greater the impact on the cost of production. Examples of PIPs previously implemented by the Air Force include manufacturing process improvements for avionics, improvements in the fabrication and assembly processes for the airframe, and the redesign of several components to enable lower production costs. The Air Force reduced the funding available for investment in PIPs because of cost growth in production lots 1 and 2. The Air Force subsequently used funding that it planned to invest in PIPs to cover the cost growth in production lots 1 and 2. As a result, there has not been as much funding available for investment in these PIPs as planned. Figure 1 shows that funding was reduced $61 million in fiscal year 2001 and $26 million in fiscal year 2002. It is unlikely that the Air Force will achieve the estimated $3.7 billion in cost growth offsets from the implementation of these PIPs if investment continues to be less than planned. Figure 2 shows the remaining planned investment in PIPs through fiscal year 2006 and the $3.7 billion in estimated cost growth that can potentially be offset through fiscal year 2010 if the Air Force invests as planned in these PIPs. In the past, Congress has been concerned about the Air Force’s practice of requesting fiscal year funding for these PIPs but then using part of that funding for F/A-22 airframe cost increases. Recently, Congress directed the Air Force to submit a request if it plans to use PIP funds for an alternate purpose. We found indications that, in the future, F/A-22 production costs are likely to increase more than the latest $5.4 billion in cost growth recently estimated by the Air Force and OSD. First, the current OSD production estimate does not include all costs. Second, schedule delays in developmental testing could delay the start of a multiyear contract designed to help control production costs. Third, as a result of schedule delays that have already occurred, the Air Force has already delayed the awarding of this contract to fiscal year 2006. As a consequence, the aircraft planned for fiscal year 2005 are not currently included in any agreements with the contractor that are designed to help control production costs. Last, we found several risk factors that may increase future production costs, including the dependency of certain cost reduction plans on congressional action and a reduction in funding for support costs. OSD’s latest cost estimate does not include costs identified by the Air Force during the development of the Air Force’s current F/A-22 acquisition plan. The Air Force developed this acquisition plan after OSD completed its estimate. Table 2 shows some areas of additional costs that the Air Force believes the program will incur. According to an OSD official, these additional costs should be considered in any future OSD production cost estimate, which would increase OSD’s estimate by $1.29 billion. If the F/A-22’s developmental testing program experiences additional delays, there is a greater risk that operational testing, full-rate production, and multiyear procurement will be delayed as a result. Delays in production and multiyear procurement would likely increase production costs. The Air Force has not addressed ongoing problems with the developmental testing and therefore remains at high risk for further schedule delays. For example, in March 2002, we reported that the Air Force’s plan to complete the developmental airframe testing necessary for the start of operational testing was at high risk because (1) the planned number of test objectives per flight-hour was not being achieved and (2) most of the planned flight-test program was essentially being performed by only one test aircraft rather than the three originally planned. Air Force officials told us they understood that completing the tests as scheduled with only one development test aircraft was high risk. As a result of this strategy, in late 2001, the Air Force delayed the F/A-22’s schedule, including the start of a multiyear contract designed to save production costs. The cost of the fiscal year 2005 production lot could increase because it is currently not included in plans to help control production costs. In late 1996, as part of a major program review, the Air Force and major F/A-22 contractors entered into a Target Price Curve agreement designed to help reduce production costs and ensure production affordability. The agreement established production cost goals for the first five production lots (fiscal years 1999-2003) and provided the contractors with incentives if they achieved these cost goals. Previously, the Air Force planned to transition directly to multiyear procurement starting with the next production lot. However, since the Air Force delayed the start of multiyear procurement from fiscal year 2004 to fiscal 2006, fiscal 2005 is now not covered by either the agreement with the contractor or the planned multiyear procurement contract. Therefore, there is less assurance that the cost of the fiscal year 2005 production lot will match the current estimate for this production lot. If a method to help control costs is not implemented for the fiscal year 2005 production lot, the cost of this production lot could increase more than expected. We found several additional risk factors that may increase production costs in the future. As we have also previously reported, the Air Force is depending on both multiyear procurement and the Joint Strike Fighter initiatives to achieve offsets to estimated cost growth. Multiyear procurement, because of the cost reductions available through long-term commitments such as a 5-year contract, make it possible for the contractors and subcontractors to charge lower prices for the aircraft being procured. Joint Strike Fighter-related savings are planned because the Air Force plans to use many of the same contractors and subcontractors as with the Joint Strike Fighter in the F/A-22 program, thereby lowering overhead rates and increasing buying power. Even though the Air Force is depending on both the multiyear procurement and Joint Strike Fighter initiatives to achieve offsets to estimated cost growth, approval to proceed with multiyear procurement is determined from the availability of funding. Thus, if entry into a multiyear procurement contract does not occur as planned, offsets from the implementation of multiyear procurement cannot be achieved. Similarly, the success of the Joint Strike Fighter cost reduction plan is dependent on the schedule of the Joint Strike Fighter program and the quantity of the aircraft procured, which are determined by Congress and OSD. In an earlier report, we cautioned that if the Joint Strike Fighter program were not approved or were delayed, then the F/A-22 production program would not achieve the estimated cost reductions. Furthermore, the Air Force reduced estimated funding for F/A-22 support costs by $1.8 billion in its latest production cost estimate. Support costs are for such items as spare components for the aircraft and engines, spare engines, and equipment used to support and maintain aircraft. F/A-22 program officials explained that the latest support costs estimate is a detailed, requirements-based estimate that is more accurate than previous estimates, but they could not provide us with the detailed rationale for this new estimate. At the same time, we also observed that the Air Force added about $1.8 billion to the estimated production costs associated with the aircraft and engine. If it is determined the F/A-22 program will require the same level of support cost funding identified by the Defense Acquisition Board’s review, the production cost estimate will increase. DOD has not fully informed Congress about specifics related to the total cost of the F/A-22 production program or the quantity of aircraft that can be purchased within the cost limitation. DOD uses selected acquisition reports and the President’s budget submissions to inform Congress about weapon systems programs. Since 1999, neither the F/A-22 selected acquisition reports nor the President’s annual budget submissions to Congress have included details about the amount of cost reduction plans identified to offset cost growth. More importantly, these documents have not included the potential cost of the F/A-22 production program if cost reduction plans do not offset cost growth as planned. From 1996 to 1998, selected acquisition reports did inform Congress about the potential cost of production if cost reduction plans did not offset cost growth as planned. If cost growth is not offset as planned, the cost of F/A-22 production could be several billion dollars higher than currently estimated. Furthermore, recent documentation, including the latest selected acquisition report (December 2001) and Fiscal Year 2003 President’s Budget submission have also not provided Congress with information about the quantity of aircraft DOD believes can be procured under the existing production cost limitation. Even though the production cost limitation remains, as adjusted, at $37.5 billion, the official documentation provided to Congress to date has not provided the number of aircraft that can be purchased for this amount. Even at the higher cost estimate of $43 billion, OSD and the Air Force have not been able to agree on the aircraft quantity that can be purchased. In July 2001, we projected that the Air Force would have to buy 85 fewer F/A-22s rather than the 333 that it planned to buy to stay within the cost limit. Despite the success of early cost reduction plans, we identified estimated cost growth beyond the amounts recognized by the Air Force and DOD. Therefore, it is important for the Air Force to take advantage of every opportunity to offset cost growth. PIPs can be an important mechanism for offsetting this cost growth. However, the Air Force is not investing funding as planned in F/A-22 PIPs designed to offset estimated cost growth. The failure to invest in PIPs at the planned level will likely not allow estimated cost growth to be offset as planned and therefore may affect the quantity of aircraft that can be acquired. The F/A-22 production program has experienced a number of schedule delays and problems that have increased the estimated costs of a program that already requires a significant investment. DOD has not fully informed Congress about the amount of cost reduction plans identified to offset cost growth, the potential cost of production if cost reduction plans are not as effective as planned, or the quantity of aircraft that can be produced within the production cost limit. Congress would be able to utilize this information to help exercise proper program oversight. For the Air Force to achieve planned offsets to estimated cost growth, we recommend that the Secretary of the Air Force make the funding of PIPs at the planned level a priority. To ensure proper congressional oversight of the F/A-22 program, we also recommend that the Secretary of Defense provide Congress with documentation showing that funding for PIPs is being invested at the planned level each fiscal year, and if not, explaining the reasons why and the potential consequences of not fully investing and potentially not offsetting cost growth as planned; reflecting the potential cost of F/A-22 production if cost reduction plans do not offset cost growth as planned; and reflecting the quantity of aircraft DOD believes can be procured with the existing production cost limit. In written comments on a draft of this report, DOD stated that it did not concur with either of our recommendations. Regarding our first recommendation on making investments in PIPs a priority, DOD said that while it believes that PIP investments in general are a good idea, the Department intends to implement PIPs on a case-by-case basis, using expected return-on-investment criteria. DOD also commented that our report does not provide evidence that investments in PIPs reduce costs. Our recommendation that the Air Force make the funding of PIPs at the planned level a priority is based on evidence from both the Air Force and OSD that investment in PIPs at the planned level will generate a significant return-on-investment. In addition, during the course of our review, Air Force officials told us they planned to make up for not fully investing in PIPs during the last 2 fiscal years by investing more in subsequent years in order to achieve the planned savings. The Air Force’s plan appears to recognize that it has moved beyond a case-by-case approach. Our recommendation would support such a plan. Finally, the reluctance to embrace PIPs in DOD’s comments appears to be contrary to the position taken within the Department. The potential benefits of investing in PIPs continue to be highlighted in high-level F/A-22 meetings and correspondence to Congress. A September 2001 letter to Congress from the Under Secretary of Defense for Acquisition, Technology and Logistics estimates that the quantity of F/A-22 aircraft will need to be reduced, but more aircraft can be procured if cost reduction plans (which include PIPs) prove more successful than OSD’s estimates. We believe our recommendation to make the funding of PIPs at the planned level a priority puts DOD in a better position to enhance the affordability of the F/A-22. Conversely, by not funding PIPs at the planned level, DOD may lose opportunities to create greater production efficiencies and as a result, have to acquire fewer aircraft. Regarding our second recommendation related to providing documentation to Congress on cost reduction plans, the implications of not investing in PIPs as planned, and the aircraft quantities that can be acquired within the existing production cost limit, DOD stated that our recommendation is inconsistent with its decision to use a “buy-to-budget” approach for the F/A-22 (buying the highest quantity of aircraft possible each year on the basis of appropriated funding each year). DOD also stated that providing this information to Congress would not provide a reliable projection of the number of aircraft possible because (1) there are other factors that affect cost and (2) the projected savings are uncertain and may not materialize as the estimator expects. We continue to believe that the Secretary of Defense should provide Congress with this documentation. As we have discussed in this and several earlier reports, we agree that there are indeed many factors that can cause F/A-22 production costs to rise. And, as we have also noted, projected offsets generated by PIPs and other cost reduction plans are uncertain and may not all materialize, even if investments are made as planned. Shifts in these realities are frequent and create a constantly changing picture of F/A-22 production costs, offsets, and aircraft quantities. This is particularly the case when PIP investments are not made as planned. Hence, it is important that updated and accurate information be regularly and routinely made available to Congress as the picture changes. DOD’s argument that it is implementing a “buy-to-budget” approach makes our recommendation more compelling because aircraft quantities planned each fiscal year can change in the few months between when fiscal year funding is appropriated and when a production contract is negotiated with the prime contractor and awarded. Providing visibility to the projection of how many aircraft can be acquired within the cost limitation would enhance program oversight. DOD has several extant reporting options that can be used to provide this information. A new report is not required. For example, DOD could return to its former practice of using annual selected acquisition reports to inform Congress about the potential cost of production if cost reduction plans do not offset cost growth as planned. This information was included in these reports from 1996 to 1998. In addition, the President’s Budget submission could be used as a vehicle to provide Congress with updated information about the quantity of aircraft DOD believes can be acquired under the existing production cost limitation. Finally, requests to reprogram PIP investment funds could be expanded to include this information along with justification for PIP reprogramming. To identify the F/A-22 production cost growth, we examined documents related to the Joint Estimating Team’s review completed in January 1997 and received clarification on some review conclusions from the F/A-22 program office. We also reviewed documentation and discussed with program officials the results of the 2001 F/A-22 Defense Acquisition Board’s review that estimated $5.4 billion more in production cost growth. To evaluate the planned effectiveness of cost reduction plans designed to offset production cost growth, we assessed the reliability of a contractor’s and the Air Force’s database on cost reduction plans to ensure that the data were complete, sufficient, and relevant to our work. We reviewed information from this database on implemented and not yet implemented cost reduction plans. We compared estimated cost reduction plan offsets from fiscal years 2000 and 2002 to determine current versus planned estimated offsets for F/A-22 production lots. We also analyzed cost information from the Air Force to determine the amount of planned and actual funding invested in PIPs designed to offset estimated cost growth by improving production processes. To identify areas where additional production cost growth has occurred and may occur, we reviewed several aspects of the F/A-22 program that were likely to contribute to future cost growth. We examined previous and current OSD and Air Force production cost estimates, expected delays in the F/A-22 program’s completion of operational testing, aircraft unit price estimates and controls, and funding for support costs. To evaluate the degree to which DOD has informed Congress about the potential cost of F/A-22 production, we examined the content of recent official documentation (selected acquisition reports and President’s budgets) provided to Congress and compared them with required content and content that would be expected considering the congressionally imposed F/A-22 production cost limitation. In performing our work, we obtained information or interviewed officials from the Office of the Secretary of Defense, Washington D.C.; the F/A-22 Program Office, Wright-Patterson Air Force Base, Ohio; and the Defense Contract Management Agency, Marietta, Georgia. We performed our work from March 2002 through February 2003 in accordance with generally accepted government auditing standards. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution of it until 30 days from the date of this report. At that time, we will send copies to interested congressional committees; the Secretary of Defense; the Secretary of the Air Force; and the Director, Office of Management and Budget. We will also make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. Please contact me at (202) 512-4841 or Catherine Baltzell at (202) 512-8001 if you or your staff have any questions concerning this report. Major contributors to this report are listed in appendix VII. F/A-22 cost reduction plans are categorized as either “implemented” or “not yet implemented.” The Air Force and contractors’ criteria for determining if a cost reduction plan is implemented include (1) whether the contractor has submitted a firm-fixed price proposal that recognizes the impact of the cost reduction, (2) whether the impact of the reduction has been reflected in a current contract price or negotiated in an agreement, or (3) whether the contractor has reduced the number of hours allocated to a task. Cost reduction plans are categorized as “not yet implemented” if the plans are well defined but none of the criteria listed above are met. Based on a plan to procure 438 aircraft. Cost growth as a percentage $4.60 0.95 0.50 0.25 0.20 69 14 8 4 3 0.14 $6.64 (0.80) Catherine Baltzell, Marvin Bonner, Edward Browning, Gary Middleton, Sameena Nooruddin, Robert Pelletier, and Don M. Springman made key contributions to this report. Tactical Aircraft: F-22 Delays Indicate Initial Production Rates Should Be Lower to Reduce Risks. GAO-02-298. Washington, D.C.: March 5, 2002. Tactical Aircraft: Continuing Difficulty Keeping F-22 Production Costs within the Congressional Limitation. GAO-01-782. Washington, D.C.: July 16, 2001. Tactical Aircraft: F-22 Development and Testing Delays Indicate Need for Low-Rate Production. GAO-01-310. Washington, D.C.: March 15, 2001. Defense Acquisitions: Recent F-22 Production Cost Estimates Exceeded Congressional Limitation. GAO/NSIAD-00-178. Washington, D.C.: August 15, 2000. Defense Acquisitions: Use of Cost Reduction Plans in Estimating F-22 Total Production Costs. GAO/T-NSIAD-00-200. Washington, D.C.: June 15, 2000. Budget Issues: Budgetary Implications of Selected GAO Work for Fiscal Year 2001. GAO/OCG-00-8. Washington, D.C.: March 31, 2000. F-22 Aircraft: Development Cost Goal Achievable If Major Problems Are Avoided. GAO/NSIAD-00-68. Washington, D.C.: March 14, 2000. Defense Acquisitions: Progress in Meeting F-22 Cost and Schedule Goals. GAO/T-NSIAD-00-58. Washington, D.C.: December 7, 1999. Fiscal Year 2000 Budget: DOD’s Procurement and RDT&E Programs. GAO/NSIAD-99-233R. Washington D.C.: September 23, 1999. Budget Issues: Budgetary Implications of Selected GAO Work for Fiscal Year 2000. GAO/OCG-99-26. Washington, D.C.: April 16, 1999. Defense Acquisitions: Progress of the F-22 and F/A-18E/F Engineering and Manufacturing Development Programs. GAO/T-NSIAD-99-113. Washington, D.C.: March 17, 1999. F-22 Aircraft: Issues in Achieving Engineering and Manufacturing Development Goals. GAO/NSIAD-99-55. Washington, D.C.: March 15, 1999. F-22 Aircraft: Progress of the Engineering and Manufacturing Development Program. GAO/T-NSIAD-98-137. Washington D.C.: March 25, 1998. F-22 Aircraft: Progress in Achieving Engineering and Manufacturing Development Goals. GAO/NSIAD-98-67. Washington, D.C.: March 10, 1998. Tactical Aircraft: Restructuring of the Air Force F-22 Fighter Program. GAO/NSIAD-97-156. Washington, D.C.: June 4, 1997. Defense Aircraft Investments: Major Program Commitments Based on Optimistic Budget Projections. GAO/T-NSIAD-97-103. Washington, D.C.: March 5, 1997. F-22 Restructuring. GAO/NSIAD-97-100BR. Washington, D.C.: February 28, 1997. Tactical Aircraft: Concurrency in Development and Production of F-22 Aircraft Should Be Reduced. GAO/NSIAD-95-59. Washington, D.C.: April 19, 1995. Tactical Aircraft: F-15 Replacement Issues. GAO/T-NSIAD-94-176. Washington, D.C.: May 5, 1994. Tactical Aircraft: F-15 Replacement Is Premature as Currently Planned. GAO/NSIAD-94-118. Washington, D.C.: March 25, 1994.

In 1991, the Air Force began developing the F/A-22 aircraft with advanced features to make it less detectable to adversaries and capable of high speeds for long distances. After a history of program cost increases, Congress limited the cost of F/A-22 production to $37.5 billion in 1997. Congress has remained interested in the potential cost of production. As requested, we (1) identified the latest production cost estimate and assessed the planned offsets from cost reduction plans, (2) identified areas where additional cost growth is likely to occur, and (3) determined the extent that DOD has informed Congress about production costs. The Department of Defense (DOD) has identified about $18 billion in estimated production cost growth over the last 6 years. Even though the Air Force has designed cost reduction plans to offset a significant amount of this estimated cost growth, DOD still estimates that the cost of production will exceed the cost limit established by Congress in 1997. Furthermore, the Air Force has not fully funded certain cost reduction plans called production improvement programs (PIPs), and as a result, these PIPs may not achieve their estimated $3.7 billion in offsets to cost growth. In addition to the cost growth estimated by DOD, GAO identified areas where, in the future, F/A-22 production cost growth is likely to occur. First, the Office of the Secretary of Defense's current production cost estimate does not include about $1.3 billion in costs that should be considered in future cost estimates. Second, schedule delays in developmental testing could delay the start of a multiyear contract designed to control costs. These delays could also result in additional costs owing to the expiration of an Air Force agreement with the contractor designed to help control production costs in fiscal year 2005. Last, other risk factors may increase future production costs, including the dependency of certain cost reduction plans on the availability of funding and a reduction in funding for support costs. DOD has not fully informed Congress (1) about what the total cost of the production program could be if cost reduction plans do not offset cost growth as planned or (2) about the aircraft quantity that can be procured within the production cost limit. If the cost limit is maintained and estimated production costs continue to rise, the Air Force will likely have to procure fewer F/A-22s.

DOD recognizes that it is neither practical nor feasible to protect its entire infrastructure against every possible threat and, similar to DHS, it is pursuing a risk-management approach to prioritize resource and operational requirements. Risk management is a systematic, analytical process to determine the likelihood that a threat will harm assets, and then to identify actions to reduce risk and mitigate the consequences of the threat. While risk generally cannot be eliminated, enhancing protection from threats or taking actions—such as establishing backup systems or hardening infrastructure—to reduce the effect of an incident can serve to significantly reduce risk. DOD’s risk-management approach is based on assessing threats, vulnerabilities, criticalities, and the ability to respond to incidents. Threat assessments identify and evaluate potential threats on the basis of capabilities, intentions, and past activities before they materialize. Vulnerability assessments identify weaknesses that may be exploited by identified threats and suggest options that address those weaknesses. For example, a vulnerability assessment might reveal weaknesses in unprotected infrastructure, such as satellites, bridges, and personnel records. Criticality assessments evaluate and prioritize assets on the basis of their importance to mission success. For example, certain power plants, computer networks, or population centers might be identified as important to the operation of a mission-critical seaport. These assessments help prioritize limited resources while reducing the potential for expending resources on lower-priority assets. DOD’s risk-management approach also includes an assessment of the ability to respond to, and recover from, an incident. The amount of non-DOD infrastructure that DOD relies on to carry out missions has not been identified; however, it is immense. To date, DHS has identified about 80,000 items of non-DOD infrastructure, some of which is also critical to DOD. Additionally, according to the Office of the Deputy Under Secretary of Defense for Installations and Environment, DOD owns more than 3,700 sites with more than half a million real property assets worldwide that could also qualify as critical infrastructure. The methodology DOD uses to identify critical infrastructure involves linking DOD missions to supporting critical infrastructure. Figure 2 shows three representative types of DOD-owned and non-DOD-owned critical infrastructure. In 1998, the Office of the Assistant Secretary of Defense for Command, Control, Communications, and Intelligence was responsible for DOD’s critical infrastructure protection efforts; however, in September 2003, the Deputy Secretary of Defense moved this program to the newly established Office of the Assistant Secretary of Defense for Homeland Defense. DOD’s critical infrastructure efforts were formalized in August 2005 with the issuance of DOD Directive 3020.40, which established DCIP. On December 13, 2006, this office was renamed the Office of the Assistant Secretary of Defense for Homeland Defense and Americas’ Security Affairs. Shortly after the office became responsible for DOD’s critical infrastructure protection efforts in October 2003, ASD(HD&ASA) established the Defense Program Office for Mission Assurance in Dahlgren, Virginia, to manage the day-to-day activities of DCIP. The Program Office—now a Mission Assurance Division—was responsible for coordinating DCIP efforts across DOD components and sector lead agents, developing training and exercise programs, overseeing the development of analytical tools and standards to permit DOD-wide analyses, and developing a comprehensive system to track and evaluate critical infrastructure. DOD organizations that have significant DCIP roles and responsibilities are shown in figure 3. The COCOMs, in collaboration with the Joint Staff, identify and prioritize DOD missions that are the basis for determining infrastructure criticality. The military services, as the principal owners of DOD infrastructure, identify and link infrastructure to specific COCOM mission requirements. Defense sector lead agents address the interdependencies among infrastructure that cross organizational boundaries, and evaluate the cascading effects of degraded or lost infrastructure on other infrastructure assets. Further, DOD officials told us that DTRA performs infrastructure vulnerability assessments for the Joint Staff in support of DCIP to determine single points of failure from all hazards. DOD has taken some important steps to implement DCIP; however, it has not developed a comprehensive management plan to guide its efforts. Although an ASD(HD&ASA) official told us they are preparing an outline for a plan to implement DCIP, it is unclear the extent to which such a plan will address key elements associated with sound management practices, including issuing guidance, coordinating program stakeholders’ efforts, and identifying resource requirements. DOD has been slow finalizing DCIP guidance and policies. As of May 2007, most of DOD’s DCIP guidance and policies were either newly issued or still in draft, which has resulted in DOD’s components pursuing varying approaches to implement DCIP. DOD has taken steps to improve information sharing and coordination within and outside of DOD. Finally, through DOD’s budget process, DCIP has received over $160 million from fiscal years 2004 to 2007. Of this amount, the components and sector lead agents have received $68.6 million, of which about 21 percent is from supplemental appropriations. Our prior work has shown that supplemental funding is not an effective means for decision makers to effectively and efficiently plan for future years resource needs, weigh priorities, and assess budget trade-offs. Until DOD completes a comprehensive management plan to implement DCIP, which includes issuing remaining DCIP guidance and fully identifying funding requirements, its ability to implement DCIP will be challenged. While our prior work has shown that issuing timely guidance is a key element of sound management, as of May 2007, the majority of DCIP guidance and policies were either newly issued or still in draft form, more than 3½ years after the Deputy Secretary of Defense assigned DCIP to ASD(HD&ASA) in September 2003 (see table 1). In the absence of finalized guidance and policies, DOD components have been pursuing varying approaches to implement their critical infrastructure programs, a condition that has not changed markedly with the issuance of several guidance documents in the past year. According to officials responsible for the critical infrastructure programs from several of the DOD components, they were either unaware that the guidance had been finalized or had decided to continue the approach they had previously adopted. Although DOD issued a DCIP directive in August 2005, ASD(HD&ASA) lacks a chartering directive that, among other things, clearly defines important roles, responsibilities, and relationships with other DOD organizations and missions—including the relationship between ASD(HD&ASA) and the Assistant Secretary of Defense for Special Operations and Low-Intensity Conflict and Interdependent Capabilities. At present, responsibility for antiterrorism guidance resides with the Assistant Secretary of Defense for Special Operations and Low-Intensity Conflict and Interdependent Capabilities. A memorandum entitled Implementation Guidance Regarding the Office of the Assistant Secretary of Defense for Homeland Defense issued by the Deputy Secretary of Defense in March 2003 required the Director of Administration and Management within the Office of the Secretary of Defense to develop and coordinate within 45 days a chartering DOD Directive that would define, among other things, the relationship between ASD(HD&ASA) and the Assistant Secretary of Defense for Special Operations and Low-Intensity Conflict and Interdependent Capabilities. However, more than 4 years later, this chartering DOD directive still has not been accomplished. Currently, DCIP implementation is diffused among program stakeholders, such as the COCOMs and the military services. As a consequence, some components, such as the U.S. Northern Command and U.S. Special Operations Command, leveraged DOD’s antiterrorism guidance to develop critical infrastructure programs, while other components, such as the U.S. Strategic Command and U.S. European Command, have kept the two programs separate. Until DOD addresses the need for a chartering directive to properly identify the relationship between DCIP and the antiterrorism program, and sets timelines for finalizing its remaining guidance, it cannot be assured that components and sector lead agents identify, prioritize, and assess their critical infrastructure in a consistent manner. This lack of consistency could impair DOD’s ability to perform risk-based decision making across component lines over the long term. Existing DCIP guidance emphasizes information sharing and collaboration with relevant government and private-sector entities. While DOD has taken steps to facilitate information sharing and coordination within the department, as well as with other federal agencies and private sector companies, we believe additional measures could be taken, such as greater cooperation with federal-level counterparts on the identification, prioritization, and assessment of critical infrastructure. Since 2003, ASD(HD&ASA) has established and sponsored several information sharing and coordination forums, such as the Defense Infrastructure Sector Council and Critical Infrastructure Program Integration Staff. The Defense Infrastructure Sector Council provides a recurring forum for DCIP sector lead agents to share information, identify common areas of interest, and leverage the individual activities of each sector to eliminate duplication. The Critical Infrastructure Program Integration Staff is comprised of representatives from more than 30 DOD organizations. Additionally, ASD(HD&ASA) maintains an Internet site that is used to post relevant information, such as policies, available training, and announcement of meetings and conferences. ASD(HD&ASA) also is a member of several critical infrastructure forums whose membership extends beyond DOD, such as the Homeland Infrastructure Foundation Level Database Working Group, and several Critical Infrastructure Partnership Advisory Council Committees, including those pertaining to communications, electricity, and dams. In another effort to coordinate DOD components’ and sector lead agents’ critical infrastructure protection practices, DOD released, in September 2006, its DCIP Geospatial Data Strategy, which lays out a standardized approach to depict geographically critical infrastructure data. Both DHS and DOD officials acknowledged the potential benefits of increasing collaborative efforts, particularly with respect to critical infrastructure identification, tracking, and assessing. To promote clear and streamlined communication, ASD(HD&ASA) has directed DOD components and sector lead agents to channel their interactions with DHS through them. However, with the exception of the Health Affairs and Financial Services defense sectors, there has been little to no coordination between the defense sectors and their corresponding federal-level critical infrastructure sector counterparts due to the immaturity of the program. Table 2 shows the defense-level sectors that are comparable to those at the federal level. DOD components are collecting different data to track and monitor their critical infrastructure to meet the needs of DCIP as well as their own, which could impede information sharing and analysis over time, and hinder DOD’s ability to identify and prioritize critical infrastructure across DOD components and sector lead agents. ASD(HD&ASA) guidance on how DOD components and sector lead agents should track and monitor their critical infrastructure is in various stages of development and review. For example, in May 2006, ASD(HD&ASA) issued a draft version of the DCIP Data Collection Essential Elements of Information Data Sets requiring DOD components and sector lead agents to collect a common set of data on their critical infrastructure. However, officials from several of the COCOMs and defense sectors told us that they have not incorporated the DCIP Data Collection Essential Elements of Information Data Sets into their data collection efforts because the guidance has not been finalized. These officials further stated that they are following departmental guidance not specific to DCIP that pertains to database interoperability and data sharing. During fiscal year 2006, ASD(HD&ASA) tasked the Mission Assurance Division to develop the capability to geospatially display DOD components’ and sector lead agents’ critical infrastructure and interdependencies. The Mission Assurance Division has received and modeled critical infrastructure data from several defense sector lead agents, but according to division officials, the combination of funding constraints and the components and sector lead agents independently acquiring technical support for their individual critical infrastructure programs, has limited its utility. In an effort to maximize the potential information DOD could receive about critical infrastructure it does not own, DOD officials told us that they plan to obtain Protected Critical Infrastructure Information (PCII) accreditation from DHS. The PCII program was established by DHS pursuant to the Critical Infrastructure Information Act of 2002. The act provides that critical infrastructure information that is voluntarily submitted to DHS for use by DHS regarding the security of critical infrastructure and protected systems, analysis, warning, interdependency study, recovery, reconstitution, or other informational purpose, when accompanied with an express statement, shall receive various protections, including exemption from disclosure under the Freedom of Information Act. If such information is validated by DHS as PCII, then the information can only be shared with authorized users. Before accessing and storing PCII, organizations or entities must be accredited and have a PCII officer. Authorized users can request access to PCII on a need-to-know basis, but users outside of DHS do not have the authority to store PCII until their agency is accredited. However, the lack of accreditation does not otherwise prevent entities from sharing information directly with DOD. For example, in the aftermath of September 11, 2001, the Association of American Railroads began prioritizing railroad assets and vulnerabilities— information that it shares with DOD—on the more than 30,000 miles of commercial rail line used to transport defense critical assets. DOD officials told us that DOD has not yet fully evaluated the costs and benefits of accreditation for its purposes. We noted in our April 2006 report that nonfederal entities continued to be reluctant to provide their sensitive information to DHS because they were not certain that their information will be fully protected, used for future legal or regulatory action, or inadvertently released. Since our April report, DHS published on September 1, 2006, its final rule implementing the act, but we have not examined whether nonfederal entities are more willing to provide sensitive information to DHS under the act at this time, or DOD’s cost to apply for, receive, and maintain accreditation. It is unclear to us, at this time, the extent to which obtaining accreditation would be beneficial to DOD when weighed against potential costs. DCIP has received about $160 million from fiscal years 2004 to 2007, through DOD’s budget process. Of this amount, ASD(HD&ASA) received approximately $86.8 million, while the Joint Staff received approximately $5.3 million. The DOD components and sector lead agents, which are responsible for identifying critical infrastructure, received $68.5 million during the same 4-year period, of which $14.3 million (about 21 percent of the component and sector lead agents’ combined funding) has come from supplemental appropriations. Figures 4 and 5 show how much DCIP funding was received by the components and sector lead agents during fiscal years 2004 to 2007. The extent to which individual components and sector lead agents relied on supplemental funding for their critical infrastructure programs varied by fiscal year. In fiscal year 2005, for example, both the U.S. Special Operations Command and the Health Affairs defense sector did not receive any programmed funding and relied exclusively on supplemental appropriations. The Defense Intelligence Agency, the lead agent for the Intelligence, Surveillance, and Reconnaissance defense sector, received 78 percent of its fiscal year 2005 critical infrastructure funding from supplemental appropriations. Likewise, the U.S. Northern Command received almost three-quarters (72 percent) of its critical infrastructure funding from supplemental appropriations in fiscal year 2006. Management control standards contained in the Standards for Internal Control in the Federal Government and sound management practices emphasize the importance of effective and efficient resource use. Relying on supplemental funding to varying degrees for their DCIP budget prevents the components and sector lead agents from effectively planning future years’ resource needs, weighing priorities, and assessing budget trade-offs. DCIP funding has been centralized in ASD(HD&ASA) since fiscal year 2004; however, beginning in fiscal year 2008, the military departments will be required to fund service critical infrastructure programs as well as the nine COCOM critical infrastructure programs. According to DOD Directive 3020.40, the military departments and COCOMs are required to provide resources for programs supporting DCIP. This responsibility is reiterated and amplified in a memorandum from the Principal Deputy Assistant Secretary of Defense for Homeland Defense that instructs the military departments and the COCOMs to include DCIP funding in their fiscal year 2008 to 2013 budget submissions. ASD(HD&ASA) will continue to fund defense sector critical infrastructure programs for fiscal years 2008 and 2009, and ASD(HD&ASA) officials stated that they will work with the defense sector lead agents to obtain funding through the lead agents’ regular budget process, beginning in fiscal year 2010. Including DCIP in the lead agents’ baseline budgets should reduce reliance on supplemental appropriations to implement critical infrastructure responsibilities. Overall DCIP funding received (fiscal years 2004 to 2007), and requested (fiscal years 2008 to 2013) is shown in figure 6. If DCIP is funded at requested levels in future years, then it will represent a substantial increase over current actual funding levels. However, in previous years, DCIP consistently has been funded at less than the requested amounts. For example, in fiscal year 2005, the military services collectively requested approximately $8 million in DCIP funding from ASD(HD&ASA) and received $2.1 million. That year, the military services also received an additional $2.3 million in supplemental appropriations, raising their total funding in fiscal year 2005 to $4.4 million, which is approximately 55 percent of what was requested. Even if DCIP funding is substantially increased, without a comprehensive management plan in place, it is not clear that the funds would be allocated to priority needs. DOD estimates that it has identified about 25 percent of the critical infrastructure it owns, and expects to finish identifying the remaining 75 percent by the end of fiscal year 2009. DOD has identified considerably less of its critical infrastructure owned by non-DOD entities, and has not set a target date for its completion. A principal reason why DOD has not identified a greater amount of its critical infrastructure is the lack of timely DCIP guidance and policies, which has resulted in DOD’s components pursuing varying approaches in identifying their critical infrastructure. DOD has been performing a limited number of vulnerability assessments on DOD-owned infrastructure; however, until DOD identifies and prioritizes all of the critical infrastructure it owns, results have questionable value for deciding where to target funding investments. Currently, DOD includes the vulnerability assessment of DOD-owned infrastructure as a module to an existing assessment. However, it has not formally adopted this practice DOD-wide, which would reduce the burden on installation personnel and asset owners. Moreover, DOD does not have a mechanism to flag domestic mission-critical infrastructure for DHS to consider including among its assessments of the nation’s critical infrastructure, and has delayed coordinating the assessments of non-DOD critical infrastructure located abroad. DOD has not identified funding to remediate vulnerabilities identified through the assessment process. DOD estimates that it has identified about 25 percent of the critical infrastructure it owns, and ASD(HD&ASA) officials anticipate identifying all DOD-owned critical infrastructure (estimated to be about 15 percent of the total) by the fiscal year 2008–2009 time frame. DOD has identified considerably less critical infrastructure that it does not own (estimated to be about 85 percent of the total), but that it relies upon to perform its missions (see fig. 7). Without knowing how much non-DOD-owned infrastructure is mission critical, ASD(HD&ASA) officials were unable to estimate how much of the non-DOD infrastructure has already been identified or a completion date. DOD has determined that a small portion of the non-DOD infrastructure— about 200 assets—that belongs to the defense industrial base sector are mission critical. The Mission Assurance Division developed a database to track and geospatially display defense critical infrastructure both within the United States and overseas, and its associated interdependencies. According to Mission Assurance Division officials, the willingness of DOD components to share their critical infrastructure information has varied. For example, division officials told us that the defense sectors have been more forthcoming than either the military services or the COCOMs. Consequently, the database provides an incomplete view of defense critical infrastructure, which significantly reduces DOD’s ability to analyze the importance of infrastructure across the components and sector lead agents. ASD(HD&ASA) officials are aware that several of the DOD components and sector lead agents have developed databases to track their specific infrastructure. For example, the Air Force, Marine Corps, Health Affairs sector, Space sector and Personnel sector have each developed their own databases. According to ASD(HD&ASA) officials, they are focusing on ensuring compatibility among the databases rather than prescribing a central database. Until DOD identifies the remaining portion of its critical infrastructure, including the portion owned by non- DOD entities, it cannot accurately prioritize and assess the risks associated with that infrastructure. Table 3 shows the amount of infrastructure assets—rounded to the nearest hundred—that the DOD components have provisionally identified as critical as of December 2006. DOD officials cautioned that not all of this information has been validated and is subject to change. For example, some infrastructure may be counted more than once due to components performing multiple missions or being assigned dual roles. The numbers in table 3 are presented to provide an order of magnitude. According to the Standards for Internal Control in the Federal Government, appropriate policies and procedures should exist with respect to an agency’s planning and implementation activities. The length of time DOD has taken to issue DCIP guidance and policies has resulted in components pursuing varying approaches in identifying and prioritizing critical infrastructure, approaches that may not be complementary. For example, Navy officials told us that, prior to 2004, they were basing infrastructure criticality on its importance to Operation Enduring Freedom, whereas Army officials indicated that they are using wartime planning scenarios based on the 2006 Quadrennial Defense Review to determine criticality. The COCOMs and the Joint Staff are basing infrastructure criticality on its importance in accomplishing individual COCOM mission requirements—an idea proposed by the Mission Assurance Division. In 2003, the Mission Assurance Division proposed linking infrastructure criticality with COCOM mission requirements, and Joint Staff officials stated that a preliminary list has been formulated and will undergo further review in 2007. Furthermore, defense sector lead agents, such as Financial Services and Personnel, are identifying all of their infrastructure regardless of COCOM mission requirements. These variations in approaches used to determine criticality exist because DOD’s published policy, the DCIP Criticality Process Guidance Document, which directs the components and sector lead agents to use one set of criteria—COCOM mission requirements—was not finalized until December 2006. DOD has begun conducting a limited number of infrastructure vulnerability assessments on the infrastructure it owns. Between calendar years 2004 and 2007, DTRA will have conducted approximately 361 antiterrorism vulnerability assessments, 45 (about 12 percent) of which will include an assessment of critical infrastructure. Which installations receive antiterrorism vulnerability assessments with a module that focuses on critical infrastructure is based on perceived infrastructure criticality, as determined by the Joint Staff in coordination with the COCOMs, and to a lesser extent the military services. However, we believe DOD cannot effectively target infrastructure vulnerability assessments without first identifying and prioritizing its mission-critical infrastructure. Depending on the amount of infrastructure that DOD deems critical, it may not be able to perform an on-site assessment of every DOD asset. To address this limitation, ASD(HD&ASA) officials told us that they plan to implement a self-assessment program that the military services—the infrastructure owners—can conduct in lieu of or in between the scheduled vulnerability assessments. DOD is in the process of developing a vulnerability self- assessment handbook that would provide guidance on how to conduct these assessments but, as of May 2007, a release date had not been set. To reduce the burden of multiple assessments on installation personnel and asset owners, in 2005, DOD incorporated an all-hazards infrastructure assessment module into its existing antiterrorism vulnerability assessments. Including the vulnerability assessment of DOD infrastructure in an established assessment program, such as the one that exists for antiterrorism, has not been formally adopted as a departmentwide practice. Unless this practice is adopted, it is possible that infrastructure assessments could be conducted independently, thereby increasing the burden on installation personnel and asset owners that the modular approach alleviates. Beginning in calendar year 2006, the Air Force piloted its own critical infrastructure assessments at those Air Force installations not receiving DTRA-led vulnerability assessments. The Air Force completed two of these pilot critical infrastructure assessments in 2006, and has nine additional assessments planned in 2007. Unlike the DTRA-led assessments, the Air Force pilot assessments are based on risk rather than vulnerabilities. We did not examine the quality or the sources of the threat, asset criticality, and vulnerability data that the Air Force is using to perform its risk assessments. We did not evaluate the effectiveness of either the DTRA-led or Air Force assessments as part of our review. DOD is not in a position to address domestic, non-DOD, mission-critical infrastructure, with the exception of defense industrial base assets and transportation infrastructure supporting seaports and airports, much less perform vulnerability assessments on them. DHS conducts on-site vulnerability assessments of domestic non-DOD-owned critical infrastructure and has developed a model that enables owners of private- sector critical infrastructure to perform vulnerability self-assessments. DOD currently does not have a mechanism to flag mission-critical infrastructure for DHS to consider including among its assessments of the nation’s critical infrastructure. For example, if DOD knew that DHS was planning to conduct a vulnerability assessment of critical infrastructure in the Atlanta, Georgia, area, it could flag for DHS’s consideration privately- owned infrastructure that DOD deemed critical—such as an electrical substation or a railroad junction. Officials from both agencies expressed an interest in coordinating vulnerability assessments of non-DOD-owned critical infrastructure. DOD has delayed coordinating the assessments of non-DOD-owned infrastructure located abroad because it has decided to focus on identifying infrastructure that it owns. For example, U.S. European Command and U.S. Central Command officials stated that they are concentrating on identifying critical infrastructure located on their installations. In addition, DTRA officials pointed out that gaining access to relevant information about foreign-owned infrastructure is more challenging than for infrastructure owned domestically. Future DCIP funding requests may be understated because current funding levels, including supplemental appropriations, do not include the resources that may be needed to remediate vulnerabilities. Our prior work has shown the importance of identifying all program costs to enable decision makers to weigh competing priorities. According to critical infrastructure officials from several DOD components and sector lead agents, there is insufficient funding to remediate vulnerabilities identified through the assessment process. Remediation in the form of added protective measures, backup systems, hardening infrastructure against perceived threats, and building redundancy could be costly. As a point of reference, the Joint Staff spent $233.7 million from fiscal years 2004 through 2007 to correct high-priority antiterrorism vulnerabilities—more than the $160 million spent on all DCIP activities over this same period. Additionally, these antiterrorism remediation expenditures were for DOD- owned assets only and do not reflect costs to remediate vulnerabilities to infrastructure not owned by DOD. In 2000, the Congress directed the Secretary of Defense to establish a loan guarantee program that makes a maximum of $10 million loan principal guarantee available each fiscal year for qualified commercial firms to improve the protection of their critical infrastructure at their facilities or refinance improvements previously made. Once DOD identifies the critical infrastructure it relies on but does not own and its associated vulnerabilities, this program could potentially be utilized to help qualified commercial firms obtain funding for remediation. DOD depends on critical infrastructure to project, support, and sustain its forces and operations worldwide, but its lack of a comprehensive management plan to guide its efforts that addresses guidance, coordination of program stakeholders’ efforts, and resource requirements, has prevented the department from effectively implementing an efficient critical infrastructure program. ASD(HD&ASA) has overseen DCIP since September 2003; however, because key DCIP guidance has either recently been issued or remains in draft more than 3½ years later, DOD components have been pursuing different approaches to fulfill their DCIP missions—approaches that are not optimally coordinated and may conflict with each other or their federal-level counterparts. Moreover, because the relationship between the Directorates for HD&ASA and Special Operations and Low-Intensity Conflict and Interdependent Capabilities regarding the DCIP and antiterrorism missions remains undefined, some components are relying on antiterrorism guidance to implement their critical infrastructure programs while others take different approaches. Furthermore, some DCIP funding for the components and sector lead agents has come from supplemental appropriations, which, as we have reported previously, is not a reliable means for decision makers to effectively and efficiently assess resource needs. Until DOD develops a comprehensive management plan for DCIP—that includes timelines for finalizing remaining guidance and actions to improve information sharing, its ability to implement DCIP will be challenged. In addition, until DOD identifies and prioritizes what infrastructure is critical, the utility of vulnerability assessments is limited in targeting funding and investments and could lead to an inefficient use of DOD resources. Combining the infrastructure vulnerability assessment with an existing assessment, as DOD is currently doing on infrastructure that it owns, has the added advantage of reducing the burden of multiple assessments on installation personnel and asset owners. However, because DOD has not formally adopted this modular approach as a DOD- wide practice, the possibility exists that infrastructure vulnerability could be assessed separately. Still, to date, no DCIP funds have been spent on reducing vulnerabilities to infrastructure. Remediation of risk identified in the assessment process could be costly—possibly more than doubling current identified funding requirements. Finally, by not coordinating with DHS on vulnerability assessments of non-DOD domestic infrastructure, DOD is missing an opportunity to increase awareness of matters affecting the availability of assets that it relies on but does not control. When DOD components and sector lead agents consistently identify, prioritize, and assess their critical infrastructure, as well as including the remediation of vulnerabilities in their funding requirements, DOD’s ability to perform risk- based decision making and target funding to priority needs will be improved. To guide DCIP implementation, we recommend that the Secretary of Defense direct ASD(HD&ASA) to develop and implement a comprehensive management plan that addresses guidance, coordination of stakeholders’ efforts, and resources needed to implement DCIP. Such a plan should include establishing timelines for finalizing the DCIP Data Collection Essential Elements of Information Data Sets to enhance the likelihood that DOD components and sector lead agents will take a consistent approach in implementing DCIP. To implement the intent of the Deputy Secretary of Defense’s memorandum Implementation Guidance Regarding the Office of the Assistant Secretary of Defense for Homeland Defense dated March 25, 2003, we recommend that the Secretary of Defense direct the Director of Administration and Management to issue a chartering directive to, among other things, define the relationship between the Directorates for HD&ASA and Special Operations and Low-Intensity Conflict and Interdependent Capabilities. As part of this comprehensive management plan, to increase the likelihood that the defense sector lead agents are able to make effective budgetary decisions, we recommend that the Secretary of Defense direct ASD(HD&ASA) to assist the defense sector lead agents in identifying, prioritizing, and including DCIP funding requirements through the regular budgeting process beginning in fiscal year 2010. In addition, as part of developing a comprehensive management plan for DCIP, we recommend that the Secretary of Defense direct ASD(HD&ASA), in coordination with the DOD components and sector lead agents, to determine funding levels and sources needed to avoid reliance on supplemental appropriations and identify funding for DCIP remediation. We further recommend that the Secretary of Defense direct ASD(HD&ASA) to take the following actions to increase the utility of vulnerability assessments: Complete the identification and prioritization of critical infrastructure before increasing the number of infrastructure vulnerability assessments performed. Adopt the practice of combining the defense critical infrastructure vulnerability assessment module with an existing assessment as the DOD-wide practice. Issue guidance and criteria for performing infrastructure vulnerability self-assessments. Identify and prioritize domestic non-DOD-owned critical infrastructure for DHS to consider including among its assessments of the nation’s critical infrastructure. In written comments on a draft of this report, DOD concurred with all of our recommendations. DOD also provided us with technical comments, which we incorporated in the report, as appropriate. DOD’s comments are reprinted in appendix II. DHS also was provided with an opportunity to comment on a draft of this report, but informed us that it had no comments. In its written comments, DOD stated that it expects to issue its DCIP management plan by September 2007 and a chartering directive for ASD(HD&ASA) by July 2007—guidance that we believe will contribute to a more efficient and effective critical infrastructure program. Although DOD did not describe the contents of the management plan, we encourage the department to address points raised in our report—guidance, coordination of stakeholders’ efforts, and resource requirements. DOD concurred with our recommendations pertaining to infrastructure vulnerability assessments. Specifically, it agreed to identify and prioritize all DOD-owned critical infrastructure before increasing the number of assessments; to codify the practice of combining the infrastructure assessment with an existing vulnerability assessment, thereby reducing the burden of multiple assessments on installation personnel and asset owners; and to issue self-assessment guidance and criteria. In its comments, DOD stated that vulnerability assessments are a valid tool to address risk and support risk management decisions, and that delaying these assessments until all assets are identified—projected in fiscal year 2009—is unadvisable. While we agree that infrastructure vulnerability assessments can reveal exploitable weaknesses, without evaluating the capabilities, intentions, or probability of occurrence of human and natural threats, as well as the importance of a particular asset to accomplishing the mission, reducing vulnerabilities may result in little, if any, risk reduction. We agree with the department that it should continue to perform infrastructure vulnerability assessments, but believe that increasing the number of assessments performed above current levels will have limited value without considering threat and asset criticality. With respect to our recommendation on vulnerability self-assessments, DOD’s expectation that installation personnel and asset owners have the expertise and resources to apply standards and criteria that mirror what DTRA is using to perform its DCIP vulnerability assessments may be unrealistic. We believe that DOD’s earlier approach of preparing a self- assessment handbook tailored to meet a range of installation and asset requirements and capabilities will likely result in more and higher-quality self-assessments. DOD also agreed with our recommendation to identify and prioritize non-DOD-owned domestic infrastructure for DHS to consider including among its assessments of the nation’s critical infrastructure. We expect that this action will increase DOD’s awareness of vulnerabilities associated with infrastructure that it relies on but does not control. As agreed with your offices, we are sending copies of this report to the Chairman and Ranking Member of the Senate and House Committees on Appropriations, Senate and House Committees on Armed Services, and other interested congressional parties. We also are sending copies of this report to the Secretary of Defense; the Secretary of Homeland Security; the Director, Office of Management and Budget; and the Chairman of the Joint Chiefs of Staff. We will also make copies available to others upon request. In addition, this report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-5431 or by e-mail at dagostinod@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix III. To conduct our review of the Department of Defense’s (DOD) Defense Critical Infrastructure Program (DCIP), we obtained relevant documentation and interviewed officials from the following DOD organizations: Office of the Secretary of Defense Under Secretary of Defense for Personnel and Readiness, Under Secretary of Defense for Acquisition, Technology, and Logistics, Office of the Deputy Under Secretary of Defense for Industrial Policy; Under Secretary of Defense for Intelligence, Counterintelligence & Security, Physical Security Programs; DOD Counterintelligence Field Activity, Critical Infrastructure Protection Program Management Directorate; Under Secretary of Defense (Comptroller)/Chief Financial Officer; Deputy Under Secretary of Defense for Installations and Environment, Business Enterprise Integration Directorate; Assistant Secretary of Defense for Homeland Defense and Americas’ Security Affairs (ASD), Critical Infrastructure Protection Office; Assistant Secretary of Defense for Special Operations and Low- Intensity Conflict and Interdependent Capabilities, Antiterrorism Policy; Assistant Secretary of Defense for International Security Policy, Deputy Assistant Secretary of Defense for Forces Policy, Office of Space Policy; Assistant Secretary of Defense for Health Affairs, Force Health Protection & Readiness; and Assistant Secretary of Defense for Networks and Information Integration, Information Management & Technology Directorate; Joint Staff, Directorate for Operations, Antiterrorism and Homeland Defense Threat Reduction Agency (DTRA), Combat Support Department of the Army, Asymmetric Warfare Office, Critical Infrastructure Risk Management Branch; Office of the Chief Information Officer; Mission Assurance Division, Naval Surface Warfare Center, Dahlgren Division, Dahlgren, Virginia; Department of the Air Force, Air, Space and Information Operations, Plans, and Requirements, Homeland Defense Division; and Headquarters, U.S. Marine Corps, Security Division, Critical Headquarters, U.S. Central Command, Defense Critical Infrastructure Program Office, MacDill Air Force Base, Florida; Headquarters, U.S. European Command, Critical Infrastructure Protection Program Office, Patch Barracks, Germany; Headquarters, U.S. Joint Forces Command, Critical Infrastructure Protection Office, Norfolk, Virginia; Headquarters, U.S. Northern Command, Force Protection/Mission Assurance Division, Peterson Air Force Base, Colorado; Headquarters, U.S. Pacific Command, Critical Infrastructure Protection Plans & Policy Office, Camp H.M. Smith, Hawaii; Headquarters, U.S. Southern Command, Joint Operations Support Division, Miami, Florida; Headquarters, U.S. Special Operations Command, Mission Assurance Division, MacDill Air Force Base, Florida; Headquarters, U.S. Strategic Command, Mission Assurance Division, Offutt Air Force Base, Nebraska; and Headquarters, U.S. Transportation Command, Critical Infrastructure Program, Scott Air Force Base, Illinois; Headquarters, Defense Intelligence Agency, Office for Critical Infrastructure Protection & Homeland Security/Defense; Headquarters, Defense Information Systems Agency, Critical Headquarters, Defense Finance and Accounting Service, Critical Infrastructure Protection Program Office, Indianapolis, Indiana; Headquarters, Defense Logistics Agency, Logistics Sector Critical Headquarters, U.S. Army Corps of Engineers, Directorate of Military Under Secretary of Defense for Personnel and Readiness, Assistant Secretary of Defense for Health Affairs, Directorate of Force Health Protection & Readiness; Headquarters, U.S. Transportation Command, Critical Infrastructure Program, Operations Directorate, Scott Air Force Base, Illinois; and Headquarters, U.S. Strategic Command, Mission Assurance Division, Offutt Air Force Base, Nebraska. To evaluate the extent to which DOD has developed a comprehensive management plan to implement DCIP, we reviewed and analyzed policies, assurance plans, strategies, handbooks, directives, and instructions, and met with officials from each of the military services, combatant commands (COCOM) (hereafter referred to as “DOD components”), and the defense sector lead agents, as well as the Joint Staff. We compared DOD’s current approach to issuing guidance, stakeholder coordination, and resource requirements to management control standards contained in the Standards for Internal Control in the Federal Government. We also attended the August 2006 DCIP tabletop exercise sponsored by the Defense Intelligence Agency, and the October 2006 Homeland Infrastructure Foundation Level Database Working Group meeting. We met with representatives from ASD(HD&ASA), the Joint Staff, and several offices within the Office of the Secretary of Defense assigned DCIP responsibilities in DOD Directive 3020.40, Defense Critical Infrastructure Protection (DCIP), as well as the Office of the Assistant Secretary of Defense for Special Operations and Low-Intensity Conflict and Interdependent Capabilities. Further, we met with officials from the Department of Homeland Security’s (DHS) Office of Infrastructure Protection to discuss mechanisms to coordinate and share critical infrastructure information with DOD. To determine DCIP funding levels for fiscal years 2004 through 2013, we met with officials from ASD(HD&ASA) and each of the DOD components and sector lead agents, and analyzed actual and projected funding data. We also met with an official from the Office of the Under Secretary of Defense (Comptroller)/Chief Financial Officer familiar with DCIP. Additionally, we obtained information from the Joint Staff on funds expended to remediate high-priority antiterrorism vulnerabilities to illustrate the potential cost of critical infrastructure remediation. We found the data provided by DOD to be sufficiently reliable for representing the nature and extent of DCIP funding. To evaluate the extent to which DOD has identified, prioritized, and assessed its critical infrastructure, we met with officials and obtained relevant documentation from each of the DOD components, sector lead agents, ASD(HD&ASA), the Joint Staff, and the Mission Assurance Division. We examined various data collection instruments and databases DOD components and sector lead agents are using to catalog, track, and map infrastructure, including the Mission Assurance Division’s database, the Air Force’s Critical Asset Management System, the Health Affairs defense sector’s Primary Health Assets Staging Tool, the Personnel defense sector’s Characterization and Dependency Analysis Tool, and the Space defense sector’s Strategic Mission Assurance Data System. We also received a demonstration of DHS’s National Asset Database, which catalogs the nation’s infrastructure. We did not verify the accuracy of infrastructure provisionally identified as critical by the DOD components and sector lead agents because the data is incomplete and, has not been validated by the department. Further, we did not verify the interoperability of these databases because it was outside the scope of our review. We met with DTRA officials to obtain information on the scope, conduct, and results of infrastructure vulnerability assessments. We also met with Air Force officials to discuss their infrastructure risk assessments. We did not evaluate the effectiveness of either the DTRA-led or Air Force assessments as part of our review. Finally, to become familiar with prior work relevant to defense critical infrastructure, we met in Arlington, Virginia, with officials from the George Mason University School of Law’s Critical Infrastructure Protection Program and in Washington, D.C., with the Congressional Research Service (Resources, Science, and Industry Division and Foreign Affairs, Defense, and Trade Division). We conducted our review from June 2006 through May 2007 in accordance with generally accepted government auditing standards. Mark A. Pross, Assistant Director; Burns D. Chamberlain; Alissa Czyz; Michael Gilmore; Cody Goebel; James Krustapentus; Kate Lenane; Thomas C. Murphy; Maria-Alaina Rambus; Terry Richardson; Jamie A. Roberts; Marc Schwartz; and Tim Wilson made key contributions to this report.

The Department of Defense (DOD) relies on a network of DOD and non-DOD infrastructure assets in the United States and abroad so critical that its unavailability could hinder DOD's ability to project, support, and sustain its forces and operations worldwide. DOD established the Defense Critical Infrastructure Program (DCIP) to identify and assure the availability of mission-critical infrastructure. GAO was asked to evaluate the extent to which DOD has (1) developed a comprehensive management plan to implement DCIP and (2) identified, prioritized, and assessed its critical infrastructure. GAO analyzed relevant DCIP documents and guidance and met with officials from more than 30 DOD organizations that have DCIP responsibilities, and with Department of Homeland Security (DHS) officials involved in protecting critical infrastructure. While DOD has taken important steps to implement DCIP, it has not developed a comprehensive management plan to guide its efforts. GAO's prior work has shown the importance of developing a plan that incorporates sound management practices, such as issuing guidance, coordinating stakeholders' efforts, and identifying resource requirements and sources. Most of DOD's DCIP guidance and policies are either newly issued or in draft form, leading some DOD components to rely on other, better-defined programs, such as the antiterrorism program, to implement DCIP. Although DOD issued a DCIP directive in August 2005, the lead office responsible for DCIP lacks a chartering directive that defines important roles, responsibilities, and relationships with other DOD organizations and missions. DOD has created several information sharing and coordination mechanisms; however, additional measures could be taken. Also, DOD's reliance on supplemental appropriations to fund DCIP makes it difficult to effectively plan future resource needs. Until DOD completes a comprehensive DCIP management plan, its ability to implement DCIP will be challenged. DOD estimates that it has identified about 25 percent of the critical infrastructure it owns, and expects to identify the remaining 75 percent by the end of fiscal year 2009. In contrast, DOD has identified significantly less of the critical infrastructure that it does not own, and does not have a target date for its completion. Among the non-DOD-owned critical infrastructure that has been identified are some 200 assets belonging to private sector companies that comprise the defense industrial base--the focus of another report we plan to issue later this year. DOD estimates that about 85 percent of its mission-critical infrastructure assets are owned by non-DOD entities, such as the private sector; state, local, and tribal governments; and foreign governments. DOD has conducted vulnerability assessments on some DOD-owned infrastructure. While these assessments can provide useful information about specific assets, until DOD identifies and prioritizes all of the critical infrastructure it owns, assessment results have limited value for deciding where to target funding investments. For the most part, DOD cannot assess assets it does not own, and DOD has not coordinated with DHS to include them among DHS's assessments of the nation's critical infrastructure. DOD has delayed coordinating the assessment of non-DOD-owned infrastructure located abroad while it focuses on identifying the critical infrastructure that it does own. Regarding current and future DCIP funding levels, they do not include the cost to remediate vulnerabilities that are identified through the assessments. When DOD identifies, prioritizes, and assesses its critical infrastructure, and includes remediation in its funding requirements, its ability to perform risk-based decision making and target funding to priority needs will be improved.

As part of a multilayered defense strategy, MTSA required vessels and port facilities to have security plans in place by July 1, 2004, including provisions establishing and controlling access to secure areas of vessels and ports. Given that ports are not only centers for passenger traffic and import and export of cargo, but also sites for oil refineries, power plants, factories, and other facilities important to the nation’s economy, securing sensitive sites of ports and vessels against access from unauthorized persons is critical. But because ports are often large and diverse places, controlling access can be difficult. To facilitate access control, MTSA required the DHS Secretary to issue a biometric identification card to individuals who required unescorted access to secure areas of port facilities or to vessels. These secure areas are to be defined by port facilities and vessels in designated security plans they were to submit to the United States Coast Guard (USCG) in July 2004. About 1 year before the passage of MTSA in 2002, work on a biometric identification card began at the Department of Transportation (DOT), partly in response to provisions in the Aviation and Transportation Security Act and the USA PATRIOT Act that relate to access control in transportation sectors. TSA—then a part of DOT—began to develop a transportation worker identification credential (TWIC) as an identity authentication tool that would ensure individuals with such an identification card had undergone an assessment verifying that they do not pose a terrorism security risk. The credential was designed by TSA to be a universally recognized identification card accepted across all modes of the national transportation system, including airports, seaports, and railroad terminals, for transportation workers requiring unescorted physical access to secure areas in this system. The credential is also to be used to help secure access to computers, networks, and applications. As shown in figure 1, ports or facilities could use an identification credential that stored a biometric, such as a fingerprint, to verify a worker’s identity and, through a comparison with data in a local facility database, determine the worker’s authority to enter a secure area. During early planning stages in 2003 and while still a part of DOT, TSA decided that the most feasible approach to issue a worker identification card would be a cost-sharing partnership between the federal government and local entities, with the federal government providing the biometric card and a database to confirm a worker’s identity and local entities providing the equipment to read the identity credential and to control access to a port’s secure areas. In 2003, TSA projected that it would test a prototype of such a card system within the year and issue the first of the cards in August 2004. In March 2003, as part of a governmentwide reorganization, TSA became a part of DHS and was charged with implementing MTSA’s requirement for a maritime worker identification card. TSA decided to use the prototype card system to issue the maritime identification card required under MTSA. At that time, TSA was preparing to test a prototype card system; later, DHS policy officials directed the agency to explore additional options for issuing the identification card required by MTSA. As a result, in addition to testing its prototype card system, TSA is exploring the cost- effectiveness of two other program alternatives: (1) a federal approach: a program wholly designed, financed, and managed by the federal government and (2) a decentralized approach: a program requiring ports and port facilities to design, finance, and manage programs to issue identification cards. According to TSA documents, each approach is to meet federally established standards for technical performance and interoperability across different transportation modes (such as air, surface, or rail). Appropriations committee conference reports, for fiscal years 2003 and 2004, directed up to $85 million of appropriated funds for the development and testing of a maritime worker identification card system prototype. With respect to fiscal year 2005 appropriations, $15 million was directed for the card program. The fiscal year 2005 funding was decreased from the $65 million as proposed by the House and the $53 million as proposed by the Senate because of delays in prototyping and evaluating the card system, according to the conference committee report. Several forms of guidance and established best practices apply to the acquisition and management of a major information technology system such as the maritime worker identification card program. For major information technology investments, DHS provided capital planning and investment control guidance as early as May 2003 that established four levels of investments, the top three of which are subject to review by department-level boards, including the Investment Review Board (IRB) and the Enterprise Architecture Board. The guidance also laid out a process for selecting, controlling, and managing investments. For example, DHS guidance suggests that as part of the control process, the agency should consider alternative means of achieving program objectives, such as different methods of providing services and different degrees of federal involvement. The guidance recommends that an alternatives analysis—a comparison of various approaches that demonstrates one approach is more cost-effective than others—should be conducted and a preferred alternative selected on the basis of that analysis. For projects like the maritime worker identification card program, whose costs and benefits extend 3 or more years, OMB also instructs federal agencies, including TSA, to complete an alternative analysis as well as a cost-benefit analysis. This analysis is to include intangible and tangible benefits and costs and willingness to pay for those benefits. In addition to DHS and OMB guidance, established industry best practices identify project management and planning best practices for major information technology system acquisition, including the development of a comprehensive plan to guide the project as detailed later in this report. Three main factors, all of which resulted in delays for testing the prototype card system, caused the agency to miss its initial August 2004 target date for issuing maritime worker identification cards. First, program officials said that although they received permission from TSA and DHS information technology officials to test a card system prototype, TSA officials had difficulty obtaining a response from DHS policy officials, contributing to the schedule slippage. Program officials said that although DHS officials reviewed the proposed card system during late 2003, senior officials provided no formal direction to program staff. Senior DHS officials said that while they were consistently briefed throughout the development of the worker identification card system, they did not provide formal direction regarding the prototype test because other important statutory and security requirements required their attention. For example, the creation and consolidation of DHS and the planning and execution of measures to close security gaps in the international aviation arena led to competition for executive-level attention and agency resources. DHS policy officials subsequently approved the test of a card system prototype. Second, while providing this approval, DHS officials also directed TSA, as part of the prototype test, to conduct a cost-benefit analysis and to evaluate the feasibility of other program alternatives for providing a card. TSA had completed these analyses earlier in the project, but DHS officials said they did not provide sufficiently detailed information on the costs and benefits of the various program alternatives. TSA officials said that because of the urgency to establish an identification card program after the terrorist attacks of September 11, 2001, the earlier cost-benefit and alternatives analyses were not completely documented as typically required by OMB regulations and DHS guidance. Working with DHS and OMB officials to identify additional information needed for a cost-benefit analysis and alternatives analysis required additional time, further delaying the prototype test. Third, TSA officials said that before testing the card system prototype, in response to direction from congressional committees, TSA conducted additional tests of various card technologies. Officials assessed the capabilities of various card technologies, such as their reliability, to determine which technology was most appropriate for controlling access in seaports. This technology assessment required 7 months to complete, more time than anticipated, delaying the prototype test. This analysis is typical of good program management and planning and, while it may have delayed the original schedule, the purpose of such assessments is to prevent delays in the future. DHS has not determined when it may begin issuing cards under any of the three proposed program alternatives—the federal, decentralized, or TWIC programs. Because of the delays in the program, some port facilities have made temporary security improvements while waiting for TSA’s maritime worker identification card system. Others, recognizing an immediate need to enhance access control systems, are proceeding with plans for local or regional identification cards that may require additional investment in order to make them compatible with TSA’s system. For example, the state of Georgia is implementing a state-based maritime worker identification card, and ports along the eastern seaboard are pursuing plans for a regional identification card. TSA officials indicated that in the near future, as they move forward with developing and operating a maritime worker identification card program, they face a number of challenges, including resolving issues with stakeholders, such as how to share costs of the program, determining the fee for the maritime worker identification card, obtaining funding for the next phase of the program. Further, in the coming months, regardless of which approach the DHS chooses—the federal, decentralized, or TWIC approach—TSA will also face challenges completing key program policies, regulatory processes, and other work as indicated in table 1. While TSA officials acknowledged the importance of completing key program policies, for example, establishing the eligibility requirements a worker must meet before receiving a card and processes for adjudicating appeals and requests for waivers from workers denied a card, officials also said that this work had not yet been completed. A senior TSA official and DHS officials said they plan to base these policies and regulations for the maritime worker identification card on those TSA is currently completing for the hazardous materials endorsement for commercial truck drivers. According to a senior TSA official who was in charge of the card program, TSA placed a higher priority on completing regulations for the hazardous materials endorsement than completing those for the maritime worker identification card. TSA has other work to complete in addition to these policies and regulations. TSA officials said OMB recently directed them and DHS officials to develop the TWIC program card in a way that allows its processes and procedures to also be used for other DHS credentialing programs. To develop such a system, DHS expects TSA to standardize, to some degree, eligibility requirements for the maritime worker identification card with those for surface and aviation workers, a task that will be challenging, according to officials. In the near future, TSA will need to produce other work, for instance, it has initiated but not yet finalized cost estimates for the card program and a cost-benefit analysis, which is a necessary part of a regulatory impact analysis required by OMB regulations. Our analysis, however, indicates that TSA faces another significant challenge besides the ones it has identified. This challenge is that TSA is attempting to proceed with the program without following certain industry-established best practices for project planning and management. Two key components of these practices are missing. The first is a comprehensive plan that identifies work to be completed, milestones for completing this work, and project budgets for the project’s remaining life. The second is detailed plans for specific and important components of the project—particularly mitigating risks and assessing alternative approaches—that would support the overall project plan. Failure to develop these plans holds significant potential to adversely affect the card program, putting it at higher risk of cost overruns, missed deadlines, and underperformance. Over the years, we have analyzed information technology systems across a broad range of federal programs and agencies, and these analyses have repeatedly shown that without adequate planning, the risks increase for cost overruns, schedule slippages, and systems that are not effective or usable. According to industry best practices for managing information technology projects like the maritime worker identification card, program managers should develop a comprehensive project plan that governs and defines all aspects of the project, tying them together in a logical manner. A documented comprehensive project plan is necessary to achieve the mutual understanding, commitment, and performance of individuals, groups, and organizations that must execute or support the plans. A comprehensive project plan identifies work to be completed, milestones for completing this work, and project budgets as well as identifying other specific, detailed plans that are to be completed to support the comprehensive project plan. The comprehensive plan, in turn, needs to be supplemented by specific, detailed plans that support the plan where necessary. Such plans might be needed to address such matters as the program’s budget and schedule, data to be analyzed, risk management and mitigation, staffing. For example, a risk mitigation plan would be important in situations where potential problems exist. One purpose of risk management is to identify potential problems before they occur; a risk mitigation plan specifies risk mitigation strategies and when they should be invoked to mitigate adverse outcomes. Effective risk management includes early and aggressive identification of risks because it is typically easier, less costly, and less disruptive to make changes and correct work efforts during the earlier phases of the project. In addition, plans for activities such as cost-benefit and alternatives analyses should be developed to help facilitate data collection and analysis. These types of plans typically describe, among other things, the data to be collected, the source of these data, and how the data will be analyzed. Such plans are important to guide needed data analysis as well as prevent unnecessary data collection, which can be costly. For this program, both risk mitigation and data analysis are key, because the program runs significant risks with regard to ensuring cooperation of stakeholders, and because TSA still faces considerable analytical work in deciding which approach to adopt. According to TSA officials, the agency lacks an approved, comprehensive project plan to guide the remaining phases of the project, which include the testing of a maritime worker identification card system prototype and issuance of the cards. While it has initiated some project planning, according to officials, the agency has not completed a comprehensive project plan, which is to identify work to be completed, milestones for completing this work, and project budgets as well as identifying other specific, detailed plans that are to be completed. Officials said that with contractor support they intended to develop a plan to manage the prototype test. However, officials did not intend to develop a plan for the remainder of the project until key policy decisions had been made, such as what type of card program will be selected to issue the cards. Once key policies are determined, TSA may move forward with a comprehensive plan. As a consequence of not having such a plan in place, officials have not documented work to be completed, milestones for completing it, or accountability for ensuring that the work is done. Without a comprehensive project plan and agreement to follow the plan from the appropriate DHS and TSA officials, TSA program staff may have difficulty managing future work, putting the program at higher risk of additional delays and cost overruns. Officials did not provide a timeframe for completing such a project plan. According to TSA planning documents and discussions with officials, TSA lacks a risk management plan that specifies strategies for mitigating known risks which could limit TSA’s ability to manage these risks. For instance, TSA documents identified failure to sustain the support of external stakeholders, such as labor unions for port workers, as a program risk and indicated a mitigation strategy was needed to address this risk. But, TSA has not developed such a strategy to address this specific risk. TSA documents also indicated that involving stakeholders in decision making could help mitigate program risks associated with defining the eligibility requirements for the card. However, TSA has not planned for stakeholder involvement in decision-making. Several stakeholders at ports and port facilities told us that while TSA solicited their input on some issues, TSA did not respond to their input or involve them in making decisions regarding eligibility requirements for the card. In particular, some stakeholders said they had not been included in discussions about which felony convictions should disqualify a worker from receiving a card, even though they had expected and requested that DHS and TSA involve them in these decisions. One port security director said TSA promised the port a “large role” in determining the eligibility requirements which has not materialized, and others said that in the absence of TSA defining the eligibility requirements for the card, they recently drafted and sent proposed eligibility requirements to TSA. TSA officials said they have an extensive outreach program to inform external stakeholders about the program, for instance, by frequently attending industry conferences and maritime association meetings. Obtaining stakeholder involvement is important because achieving program goals hinges on the federal government’s ability to form effective partnerships among many public and private stakeholders. If such partnerships are not in place—and equally important, if they do not work effectively—TSA may not be able to test and deliver a program that performs as expected. For example, TSA currently relies on facilities and workers to voluntarily participate in tests of the prototype card system. Without this and other support provided by stakeholders, the prototype card system could not be tested as planned. Planning for stakeholder involvement is also important because in the future other groups or organizations, for instance, other federal agencies or states, may be charged with developing biometric identification card programs and emerge as important external stakeholders for the maritime worker identification card program. According to best practices, in order to ensure that the appropriate data are collected to support analyses on which program decisions are made, managers should develop a plan that describes data to be collected, the source of these data, and how the data will be analyzed. During the test of the prototype card system, officials said they are to collect data on the feasibility of the federal and decentralized approaches in order to conduct an alternatives analysis—a comparison of the three possible approaches that demonstrates one approach is more cost-effective than the others. TSA officials acknowledge they have not yet completed a plan; however, they said they intend to do so with contractor support. On the basis of interviews with a number of officials and review of documents, we determined TSA has not identified who would be responsible for collecting the data; the sources for the data, and how it will be analyzed. These details are needed to ensure production of a good result. Completing the cost-benefit and alternatives analyses is important because not only do OMB regulations and DHS guidance instruct agencies to complete them, but DHS officials said the alternatives analysis would guide their decision regarding which approach is the most cost-effective way to provide the card. Without a plan to guide this activity, TSA may not perform the necessary analysis to inform sound decision making, possibly causing further delays. With the passage of MTSA, Congress established a framework for homeland security that relies on a multilayered defense strategy to enhance port security. Improving access control by providing ports a maritime worker identification card is an important part of this strategy. Each delay in TSA’s program to develop the card postpones enhancements to port security and complicates port stakeholders’ efforts to make wise investment decisions regarding security infrastructure. Despite delays and the difficulties of a major governmentwide reorganization, DHS and TSA have made some progress in developing a maritime worker identification card. Nevertheless, without developing a comprehensive project plan and its component parts—an established industry best practice for project planning and management—TSA is placing the program’s schedule and performance at higher risk. More delays could occur, for example, unless DHS and TSA agree on a comprehensive project plan to guide the remainder of the project, identify work that TSA and DHS officials must complete, and set deadlines for completing it. Without adequate risk mitigation plans, TSA may not be able to resolve problems that could adversely affect the card program objectives, such as insufficient stakeholder support to successfully develop, test, and implement the card program. Further, without a plan to guide the cost-benefit and alternatives analyses, TSA increases the risk that it may fail to sufficiently analyze the feasibility of various approaches to issue the card, an analysis needed by DHS policy officials to make informed decisions about the program, putting the program at risk for further delays. To help ensure that TSA meets the challenges it is facing in developing and operating its maritime worker identification card program, we are recommending that the Secretary of Homeland Security direct the TSA Administrator to employ industry best practices for project planning and management, by taking the following two actions: Develop a comprehensive project plan for managing the remaining life of the project. Develop specific, detailed plans for risk mitigation and cost-benefit and alternatives analyses. We provided a draft of this report to DHS and TSA for their review and comment. DHS and TSA generally concurred with the findings and recommendations that we made in our report and provided technical comments that we incorporated where appropriate. DHS and TSA also provided written comments on a draft of this report (see app. I). In its comments, DHS noted actions that it has recently taken or plans to take to address concerns we raised regarding outstanding regulatory and policy issues. Although DHS and TSA concurred with our recommendations, in their comments, they contend that project plans and program management controls are currently in place to manage their test of the TWIC prototype. However, at the time of our review, the project planning documents identified by DHS and TSA in their comments were incomplete, lacked the necessary approvals from appropriate officials, or were not provided during our audit. Furthermore, project plans and other management controls have not been developed for the remaining life of the project. We are sending copies of this report to other interested Members of Congress. We are also sending copies to the Secretary of Homeland Security. We will make copies available to others upon request. In addition, the report will be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (415) 904-2200 or at wrightsonm@gao.gov. Other major contributors to this report included Jonathan Bachman, Chuck Bausell, Tom Beall, Steve Calvo, Ellen Chu, Matt Coco, Lester Diamond, Geoffrey Hamilton, Rich Hung, Lori Kmetz, Anne Laffoon, Jeff Larson, David Powner, Tomas Ramirez, and Stan Stenerson.

As part of a multilayered effort to strengthen port security, the Maritime Transportation Security Act (MTSA) of 2002 calls for the Department of Homeland Security (DHS) to issue a worker identification card that uses biological metrics, such as fingerprints, to control access to secure areas of ports or ships. Charged with the responsibility for developing this card, the Transportation Security Administration (TSA), within DHS, initially planned to issue a Transportation Worker Identification Credential in August 2004 to about 6 million maritime workers. GAO assessed what factors limited TSA's ability to meet its August 2004 target date for issuing cards and what challenges remain for TSA to implement the card. Three main factors, all of which resulted in delays for testing a prototype of the maritime worker identification card system, caused the agency to miss its initial August 2004 target date for issuing the cards: (1) officials had difficulty obtaining timely approval to proceed with the prototype test from DHS, (2) extra time was required to identify data to be collected for a cost-benefit analysis, and (3) additional work to assess card technologies was required. DHS has not determined when it may begin issuing cards. In the future, TSA will face difficult challenges as it moves forward with developing and operating the card program, for example, developing regulations that identify eligibility requirements for the card. An additional challenge--and one that holds potential to adversely affect the entire program--is that TSA does not yet have a comprehensive plan in place for managing the project. Failure to develop such a plan places the card program at higher risk of cost overruns, missed deadlines, and underperformance. Following established, industry best practices for project planning and management could help TSA address these challenges. Best practices suggest managers develop a comprehensive project plan and other, detailed component plans. However, while TSA has initiated some project planning, the agency lacks an approved comprehensive project plan to govern the life of the project and has not yet developed other, detailed component plans for risk mitigation or the cost-benefit and alternatives analyses.

VA pays monthly compensation to veterans with service-connected disabilities (i.e., injuries or diseases incurred or aggravated while on active military duty) according to the severity of the disability. It pays additional compensation for certain dependent spouses, children, and parents of veterans. The department processes service-connected disability compensation benefits for about 3 million beneficiaries each month. It reported completing the processing of approximately 1.3 million disability rating claims for veterans and beneficiaries in fiscal year 2014— about 150,000 more than in fiscal year 2013. As of mid-June 2015, the department reported a backlog of approximately 137,000 disability claims, down from its peak of about 611,000 disability claims in March 2013. VA has reported that the accuracy of claims decisions rose from a level of 83 percent in 2011 to 90 percent in 2014. However, even as VBA reported providing historic numbers of veterans with decisions on their claims in 2014, the number and complexity of claims continue to increase. For example, VBA reported for 2014 that its employees made decisions on 5.5 million medical issues within claims—a 101 percent increase in the number of issues since 2009. Throughout the disability compensation claims process, VBA staff have various roles and responsibilities. Claims assistants are primarily responsible for establishing the electronic claims folders to determine whether the dispositions of the claims and control actions have been appropriately identified. Veteran service representatives are responsible for providing veterans with explanations regarding the disability compensation benefits programs and entitlement criteria. They also are to conduct interviews, gather relevant evidence, adjudicate claims, authorize payments, and input the data necessary to generate the awards and notification letters to veterans describing the decisions and the reasons for them. Rating veterans service representatives are to make claims rating decisions and analyze claims by applying VBA’s schedule for rating disabilities (rating schedule) against claims submissions, and preparing rating decisions and the supporting justifications. They also are to inform the veteran service representative who then notifies the claimant of the decision and the reasons for the decision. Supervisory veteran service representatives are to ensure that the quality and timeliness of service provided by VBA meets performance indicator goals. They are also responsible for the cost-effective use of resources to accomplish assigned outcomes. Decision review officers are to examine claims decisions and perform an array of duties to resolve issues raised by veterans and their representatives. They may conduct a new review or complete a review of a claim without deference to the original decision; they also can revise that decision without new evidence or clear and obvious evidence of errors in the original evaluation. The disability compensation claims process starts when a veteran (or other designated individual) submits a claim to VA, in paper or electronic form. If submitted electronically, a claim folder is created automatically. When a paper claim is submitted, a claims assistant creates the electronic folder. Specifically, when a regional office receives a new paper claim, the receipt date is recorded electronically and the paper files (e.g., medical records and other supporting documents) are shipped to one of four document conversion locations so that the supporting documents can be scanned and converted into a digital image. In the processing of both electronic and paper claims, a veteran service representative is to review the information supporting the claim and help identify any additional evidence that is needed to evaluate the claim, such as the veteran’s military service records, medical examinations, and treatment records from medical facilities and private medical service providers. Also, if necessary to provide support to substantiate the claim, the department is to perform a medical examination on the veteran. Once all of the supporting evidence has been gathered, a rating veterans service representative evaluates the claim and determines whether the veteran is eligible for benefits. If so, the rating veterans service representative assigns a disability rating (expressed as a percentage). A veteran who submits a claim with multiple disabilities receives a single composite rating. If the veteran is due to receive compensation, an award is prepared and the veteran is notified of the decision. A veteran can reopen a claim for additional disability benefits if, for example, the veteran experiences a new or worsening service-connected disability. If the veteran disagrees with the regional office’s decision on the additional claim, he or she may submit a written notice of disagreement to the regional office to appeal the decision, and may request to have the appeal processed at the regional office by a decision review officer or through the Board of Veterans’ Appeals. Figure 1 presents a simplified view of VA’s disability compensation claims process. VBA began the transformation of its paper-intensive claims process to a paperless environment in March 2009. This effort became formally established as the Veterans Benefits Management System program in May 2010. VBA’s initial plans for VBMS emphasized the development of a paperless claims platform to fully support the processing of disability compensation and pension benefits, as well as appeals. The program’s importance was further elevated in January 2013 when the Secretary of Veterans Affairs identified the system as the enabling technology to meet the goal to eliminate the disability claims backlog in fiscal year 2015 and improve the efficiency and accuracy of all compensation claims processing to 125 days and 98 percent, respectively. In a March 2013 Senate Veterans Affairs Committee hearing, VA’s Under Secretary for Benefits stated that VBMS development was expected to be completed in 2015. The program’s primary focus was to convert existing paper-based claims folders into electronic claim folders (eFolders) to allow VBA employees to access claims information and evidence in an electronic format. Beyond the establishment of eFolders, VBMS is also intended to streamline the entire disability claims process, from establishment through award, by automating rating decision recommendations, award and notification processes, and communications between VBA and the veteran throughout the claims life cycle; assist in eliminating the claims backlog and serve as the enabling technology for quicker, more accurate, and integrated claims processing in the future; and replace many of the key outdated legacy systems—which are still in use today—for managing the claims process, including: Share—used to establish claims; it records and updates basic information about veterans and dependents. Modern Award Processing–Development —used to manage the claims development process, including the collection of data to support the claims and the tracking of claims. Rating Board Automation 2000—provides information about laws and regulations pertaining to disabilities, which are used by rating specialists in evaluating and rating disability claims. Award—used to prepare and calculate the benefit award based on the rating specialist’s determination of the claimant’s percentage of disability. It is also used to authorize the claim for payment. VBMS is to consist of three modules: VBMS-Core is intended to provide the foundation for document processing and storage during the claims development process, including establishing claims; viewing and storing electronic documents in the eFolder; and tracking evidence requested from beneficiaries. The eFolder serves as a digital repository for all documents related to a claim, such as the veteran’s military service records, medical examinations, and treatment records from VA and Department of Defense medical facilities, and from private medical service providers. Unlike with paper files, this evidence can be reviewed simultaneously by multiple VBA claims processors at any location. VBMS-Rating is to provide raters with Web-accessible tools, including rules-based rating calculators and the capability for automated decision recommendations. For example, the hearing loss calculator is to automate decisions using objective audiology data and rules- based functionality to provide the rater with a suggested rating decision. In addition, VBMS-Rating is expected to include stand-alone evaluation builders—essentially interactive disability rating schedules—for all parts of the human body. With this tool, the rater uses a series of check boxes to identify the veteran’s symptoms, and the evaluation builder identifies the proper diagnostic code and the level of compensation based on those symptoms. VBMS-Awards is to provide an automated award and notification process to improve award accuracy and reduce rework associated with manual development of awards. VBMS-Awards is intended to automate and standardize communications between VBA and the veteran at the final stages of the claims process. VBA is using an agile software development methodology to develop, test, and deliver VBMS functionality to its users. An agile approach allows subject matter experts to validate requirements, processes, and system functionality in increments, and to deliver the functionality to users in shorter cycles. To help guide its system development efforts, the VBMS PMO has developed both a strategic road map that identifies the program’s high-level objectives, timeline and intended outcomes, as well as a regularly revised tactical road map that describes prospective capabilities the PMO expects to develop and deploy for each system release. Consistent with VA’s policy for incremental development, VBMS is to be developed and implemented in a series of releases that are to occur every 6 months. As shown in table 1, VBA plans to develop VBMS over multiple years, with each year generally corresponding to a new system generation. From 2010 through January 2013, VBA planned, developed, and deployed a foundational, Web-based version of VBMS to five pilot sites. This phase included development of the eFolder capability and the ability to establish, develop, and rate disability compensation claims in VBMS, as well as a user interface with search capabilities. Generation 1 development was completed in January 2013; at that time, the system was implemented at 18 regional offices. From February 2013 through September 2013, the PMO continued to add functionality to the system, including enhancement of data exchange capabilities, correspondence tools, and rating functionality. By June 2013, VBA had completed national rollout of the initial version of the system to all 56 regional offices. The system was also made accessible to VA’s Appeals Management Center, the Board of Veterans’ Appeals, VBA National Call Centers, veterans service organizations, and all VA medical centers that complete compensation exams. Subsequent to the nationwide rollout of VBMS in June 2013, VBA has continued incremental system development and enhancement of VBMS. The VBMS program is dependent on multiple organizations within the department to meet its goals. It is jointly led by two program managers: the VBMS PMO Director and a Program Manager in VA’s Office of Information and Technology (OI&T). Specifically, the VBMS PMO, which resides in VBA’s Office of Strategic Planning, is responsible for all aspects of the program’s management, including the coordination and direction of contract staff and VBA partners in integrating the system’s components and managing program-level dependencies, risks, and issues. The PMO has responsibility for gathering and delivering system requirements, performing testing, and providing training. The Director oversees all activities of the PMO and reports to the Under Secretary for Benefits. The OI&T program manager oversees the development and implementation of VBMS and reports to the VA Chief Information Officer. OI&T entered into an interagency agreement with the Department of the Navy’s Space and Naval Warfare Systems Command (SPAWAR) Systems Center Atlantic to lead the development of VBMS. SPAWAR manages multiple contractors to develop the system and is providing technical, information assurance, program management, testing, and data integration services to support application development. VA and SPAWAR work together to manage and develop the system. Specifically, VBA subject matter experts and OI&T technical representatives are part of the system development teams. Further, VBA’s Office of Business Process Integration is to ensure that strategic needs and requirements for business and data systems are properly documented, integrated, and communicated. The office provides internal coordination across the VBA lines of business (compensation, pension, education, loan guaranty, etc.) and helps communicate the system’s requirements to the OI&T at the department level. The Office of Business Process Integration also manages VBA’s legacy claims processing systems and their sustainment. As such, it is working with the OI&T on long-term planning for the systems, which may include their decommissioning. In addition to these offices, VBMS is governed by the Transformation Joint Executive Board. Jointly chaired by the VA CIO and the Under Secretary for Benefits, this board is responsible for discussing and addressing the program’s risks and other issues. In addition, VBMS is to follow the OI&T’s Project Management Accountability System, which is the department’s process for managing IT projects. As part of this process, large IT programs are broken down into multiple projects that are typically planned, developed, and implemented in 2-year cycles. Each project is broken into several increments, with each increment typically lasting 6 months or less. The PMO is responsible for presenting a life-cycle cost estimate for the 2-year project as part of each increment’s milestone review. Resources, including staff and funding, are released once the increment has been approved. In 2008, VBA developed an initial, high-level life-cycle cost estimate of $560 million for VBMS that reflected system development costs through fiscal year 2012. The PMO revised the cost estimate in 2011, to include costs for system development, sustainment, and general operating expenses. This estimate showed that the department expected to spend $934.8 million on VBMS through its life cycle. In July 2012, we reported on the reliability of the cost estimate for the VBMS program and noted that, while the 2011 cost estimate for the system partially reflected key practices for developing a comprehensive and well-documented estimate, it did not reflect key practices for developing an accurate and credible estimate. We recommended that any future life-cycle cost estimates for the VBMS program address the detailed weaknesses that we identified using cost-estimating best practices. VA’s Chief of Staff stated that the department concurred with our recommendation and had efforts under way to improve its cost- estimating capabilities. Officials from the PMO subsequently provided us information showing that from fiscal years 2009 through 2015, the VBMS program had received funding in the amount of approximately $1 billion— about $502 million for system development, $308 million for IT sustainment, and $194 million for general operating expenses. VBA has developed and implemented capabilities for VBMS to support disability claims processing. However, completion of additional functionality to fully support processing disability claims will be delayed beyond fiscal year 2015. Further, VBA’s plans for developing and implementing capabilities to support the processing of pension benefits and appeals are uncertain. As VBA continues to develop and implement the system, three areas could benefit from increased management attention. First, the PMO does not have a reliable estimate of the cost for completing the system. Second, although VBA has improved VBMS’s availability to users, it has not established goals for system response times. Third, while the program has actively managed system defects, a recent system release included multiple unresolved defects that adversely impacted performance and users’ experiences. Since completing the implementation of VBMS at all regional offices in June 2013, VBA has continued to make important progress toward developing and implementing additional system functionality and enhancements that support electronic processing of disability compensation claims. As a result, 95 percent of records related to veterans’ disability claims are electronic and reside in VBMS. However, although the Under Secretary for Benefits stated in March 2013 that the system’s development was expected to be completed in 2015, implementation of functionality to fully support electronic claims processing has been delayed beyond 2015. Additionally, federal guidance and IT project management principles stress the importance of continuous planning throughout the life of a program to serve as a basis for managing trade-offs between cost, schedule, and scope. During fiscal year 2014, VBA’s progress on the system included developing and implementing capabilities that addressed the strategic objectives for generation 3 of the strategic road map. For example: Consistent with the objective to provide system users with more complex automation capabilities, the PMO released an automation feature that populates the VBMS rating calculator with a veteran’s medical information for 37 of the 71 disability benefits questionnaires that medical providers can use to submit medical information in support of a disability claim. Regarding the strategic objective to reduce dependency on legacy systems, in April 2014, the PMO delivered the first iteration of the VBMS-Awards module to all regional offices, providing claims processors the ability to rely less on legacy systems, prepare an award based on information from VBMS’s rating module, and generate associated notification letters for veterans. The functionality released during fiscal year 2014 also included the integration of VBMS with VA’s legacy electronic records store, “Virtual VA,” and the integration of additional correspondence letters. To address the strategic objective to deliver a capability to accept veterans’ electronic service treatment records, in December 2013, the PMO implemented new functionality in VBMS to enable electronic request and receipt of service treatment records from the Department of Defense for veterans who had separated or been discharged from military service after January 1, 2014. VBA has continued its progress toward developing and implementing VBMS during fiscal year 2015 with efforts to address generation 4 of the strategic road map. Specifically, to address the strategic objective calling for further reduced reliance on legacy systems, the PMO has made enhancements to the system’s correspondence capabilities and letter templates, and added several new tools to the rating module. Additionally, the department has established plans and a schedule for retiring one legacy system. According to an analysis by VBA’s Office of Business Process Integration, fewer than 5 percent of disability cases are rated using the legacy rating system, Rating Board Automation 2000, and the functionality available in VBMS-Rating has largely eliminated the need for claims processors to access the legacy rating system. In February 2015, VBA conducted a 5-week pilot at four regional offices to determine whether VBA’s Office of Business Process Integration can proceed with plans to retire the legacy rating application. According to VBMS program officials, the Office of Business Process Integration expects to retire the legacy rating system by the end of September 2015. According to the PMO Director, the office has also been working to capture process improvements and use lessons learned from earlier system development cycles to address the generation 4 strategic objective to implement improved system development processes. Even with the progress VBA has made toward developing and implementing VBMS, the timeline for initial deployment of a national workload management capability has been delayed beyond the originally planned date of September 2014 to October 2015, with additional deployment to occur throughout fiscal year 2016. VBMS generation 4 development efforts have included addressing the strategic objective that calls for delivery of a national workload management capability. This effort has entailed developing the technology and business processes needed to support the national work queue, which is intended to handle new disability claims in a centralized queue and assign claims to the next regional office with available capacity. The PMO began work for the national work queue in June 2014. The office had intended to deploy the first phase of the work queue functionality to users in September 2014. However, in late May 2015, the PMO Director informed us that VBA had decided to delay the initial rollout of the work queue to October 2015 so that the department can fully focus on meeting its goal to eliminate the claims backlog by the end of September 2015. Following the initial rollout, the PMO intends to implement the work queue at all regional offices through fiscal year 2016. VBMS program documentation also identifies additional work to be performed after fiscal year 2015 to fully automate disability claims processing. Specifically, the PMO has identified the need to automate steps associated with a veteran’s request for an increase in disability benefits, such as when an existing medical condition worsens. In addition, according to the Director, the PMO intends to develop a capability to automatically associate veterans’ correspondence when a new piece of evidence to support a claim has been received electronically or scanned into VBMS. The PMO also plans to integrate VBMS with VA’s Integrated Disability Evaluation System, which contains the results of veterans’ disability medical examinations, as well as with external systems that contain military service treatment records for veterans, including those at the National Personnel Records Center. Further, VBA has not yet developed and implemented end-to-end pension processing capabilities in VBMS. Without such capabilities, VBA continues to rely on three legacy systems to process pension claims. Specifically, according to program officials, both the Modern Award Processing–Development and Award legacy systems contain functionality related to processing pensions and will need to remain operational until VBMS can process these claims. In addition, they said that the Share legacy system contains functionality that is still needed throughout the claims process. VBMS program officials stated that additional system analysis is needed before they can develop plans for the retirement of Share. Program documentation indicates that the first phase of pension-related functionality is expected to be introduced in December 2015. However, VBA has not yet developed plans and schedules for retiring the Modern Award Processing-Development, Award, and Share systems and fully developing and implementing the functionality of these legacy systems in VBMS. VBA’s progress toward developing and implementing appeals processing capabilities in VBMS has also been limited. Specifically, although the information in a veteran’s eFolder is available to appeals staff for review, the appeals process for disability claims is not managed using the new system. According to VA’s fiscal year 2016 budget submission, the department is pursuing a separate effort to manage end-to-end appeals modernization, and has requested $19.1 million in fiscal year 2016 funds to develop a system that will provide functionality not available in VBMS or other VA systems. The PMO Director stated that VBA is currently analyzing commercial IT solutions that can meet the business requirements for appeals, such as providing document navigation capabilities. Nevertheless, the Director added that VBMS is expected to be part of the appeals modernization solution because components of the system, such as the VBMS eFolder and certain workload management functionality, are planned to continue supporting appeals management. According to the PMO Director, the fact that VBMS requires additional development beyond 2015 does not reflect a delay in completing the system’s development. Instead, the additional time is a consequence of decisions to enlarge the VBMS program’s scope over time. This official added that VBMS’s original purpose was to serve primarily as an electronic document repository, and they have met this goal. The PMO Director further stated that, as the program’s mission has expanded to support the department’s efforts to eliminate the disability claims backlog, the PMO has had to re-prioritize, add, and defer system requirements to accommodate broader departmental decisions and, in some cases, regulatory changes. For example, the PMO was tasked with developing functionality in VBMS to meet regulatory requirements for processing disability claims using mandatory forms. Officials from the VBMS PMO explained that they were made aware of this requirement well after system planning for the March 2015 release had been completed, and it introduced significant complexity to their development work. Finally, VBA included in its strategic road map the strategic objectives that are to be addressed in generation 5 of the system, which is planned for fiscal year 2016. Further, officials from the VBMS PMO stated that they intend to develop tactical plans that identify the expected capabilities to be provided in the generation 5 releases. Nevertheless, due to the department’s incremental approach to developing and implementing VBMS, VBA has not yet produced a plan that identifies when the system will be completed and can be expected to fully support disability and pension claims processing and appeals. Thus, it will be difficult for VA to hold its managers accountable for meeting its time frame and for demonstrating progress. Consistent with our guidance on estimating program costs, an important aspect of planning for IT projects such as VBMS involves developing a reliable cost estimate to help managers evaluate a program’s affordability and performance against its plans, and provide estimates of the funding required to efficiently execute a program. Without this information, programs are at risk of experiencing cost overruns, missed deadlines, and performance shortfalls. Additionally, federal guidance and IT project management principles stress the importance of continuous planning throughout the life of a program to serve as a basis for managing trade- offs between cost, schedule, and scope. In July 2012, we identified several weaknesses in VA’s policies related to cost estimating and also found that the department’s 2011 cost estimate for VBMS did not reflect key practices for developing an accurate and credible estimate. At that time, we recommended that VA modify its policies governing cost estimating to establish, among other things, a requirement to prepare a full life-cycle cost estimate for a program (as opposed to preparing an estimate for each program increment) and a requirement for programs to prepare cost estimates using best practices. In February 2013, VA’s Office of Corporate Analysis and Evaluation released a cost estimating process guide, which states that VA program offices are required to develop a life-cycle cost estimate to support senior leaders’ decision making and the department’s planning and budgeting processes. Further, this guide states that, at the request of VA leadership or a program office, VA’s Office of Corporate Analysis and Evaluation is to be available to develop independent cost estimates and review existing program cost estimates to ensure they have been developed in accordance with our cost-estimating guidance. In 2011, VBA submitted a life-cycle cost estimate for VBMS of $934.8 million to the Office of Management and Budget. This estimate was intended to capture costs for the system’s development, deployment, sustainment, and general operating expenses through the end of fiscal year 2018. However, as of July 2015, the program’s actual costs exceeded the 2011 life-cycle cost estimate. Specifically, VBMS has received approximately $1 billion in funding through the end of fiscal year 2015 and the department has requested an additional $290 million for the program in fiscal year 2016. The PMO has not reliably updated VBMS’s life-cycle cost estimate to reflect the program’s expanded scope and timelines for completing the system. This is largely attributable to the fact that the PMO has developed cost estimates for 2-year project cycles that are used for VBMS milestone reviews under OI&T’s Project Management Accountability System. When asked how the PMO arrived at the cost estimates reported in the milestone reviews, officials stated that they develop rough order of magnitude estimates for each business need based on expert knowledge of the system, past development and engineering experience, and lessons learned. However, while this approach may have provided adequate information for VBA to prioritize VBMS system requirements to address in the next release, it has not produced estimates that could serve as a basis for identifying VBMS’s funding needs. Because it is typically derived from limited data and in a short time, a rough order of magnitude analysis is not equivalent to a budget-quality cost estimate and may limit an agency’s ability to identify the funding necessary to efficiently execute a program. In addition, the PMO’s annual operating plan, which is generally limited to high-level information about the program’s organization, priorities, staffing, milestones, and performance measures for fiscal year 2015, also shows estimated costs totaling $512 million for VBMS development from fiscal years 2017 through 2020. However, according to the PMO Director, this estimate was also developed using rough order of magnitude analysis. Further, the estimate does not provide reliable information on life-cycle costs because it does not include estimated IT sustainment and general operating expenses. Even though the PMO has developed rough order of magnitude cost estimates for VBMS, these estimates have not been sufficiently reliable to effectively identify the program’s funding needs. Instead, during the last 3 fiscal years, the PMO Director has had to request an additional $118 million in IT development funds to meet program demands and to ensure support for ongoing development contracts. Specifically, in May 2013, VA requested $13.3 million to support additional work on VBMS. Then, during fiscal year 2014, VA reprogrammed $73 million of unobligated IT sustainment funds to develop functionality to transfer service treatment records from DOD to VA, and to support development of VBMS-Core functionality. In December 2014, the PMO identified the need for additional fiscal year 2015 funds for ongoing system development contracts for VBMS-Core and VBMS-Awards, and, in late April 2015, VA leadership submitted a letter to Congress requesting permission to reprogram $31.7 million to support work on these contracts, the National Work Queue, and other VBMS efforts. According to the PMO Director, the need to request additional funding does not represent additional risk to the program, but is the result of VBMS’s success. The Director further noted that, as the PMO has identified opportunities to increase functionality to improve the electronic claims process, their funding needs have also increased. We recognize that as new capabilities are deployed, additional requirements may surface. Nevertheless, evolution of the VBMS program illustrates the importance of continuous planning, including cost estimating, so that trade-offs between cost, schedule, and scope can be effectively managed. Further, without a reliable estimate of the total costs associated with completing work on VBMS, stakeholders will have a limited view of VBMS’s future resource needs and the program is at risk of not being able to secure appropriate funding to fully develop and implement the system. GAO and federal IT guidance recognize the importance of defining program goals and related performance targets and using such targets to assess progress in achieving the goals. System performance and response times have a large impact on whether staff successfully complete work tasks. If systems are not responding at agreed-upon levels for availability and performance, it can be difficult to ensure that staff will complete tasks in a timely manner. This is especially important in the VBA claims processing environment, where staff are evaluated on their ability to process claims in a timely manner. VBA’s availability goal for VBMS is 95 percent per month, operating on a 24 hour-a-day, 6-day-a-week schedule. According to the OI&T Program Manager for VBMS, VBA defines availability as the extent to which the system is operational, functional, and usable for completing business requirements (e.g., processing claims). When an unscheduled system outage occurs, the office identifies it as being in one of three categories: critical, impacting all VBMS users across VA; moderate, impacting VBMS users at a local or individual level; and serious, impacting VBMS users in a certain regional area. The Program Manager stated that the department does not consider the system to be unavailable if users are unable to perform their duties due to non-system issues (such as local network issues or the unavailability of systems that feed data into VBMS). VBA has reported that, since its initial rollout in January 2013, the system has exceeded the 95 percent goal for availability. Specifically, VBA reported that the system was available at a rate of 98.9 percent in fiscal year 2013 and 99.3 percent in fiscal year 2014. Through May of fiscal year 2015, it was available for 99.98 percent of the time. Nonetheless, while the department reported exceeding its availability goals for VBMS, it has experienced periods of system unavailability, many times at the critical level affecting all users, as reflected in figure 2. Specifically, since January 2013, the department has reported 57 VBMS outages that have totaled about 117 hours of system unavailability. VBA experienced about 18 hours of VBMS outages in January 2014, which were almost entirely at the critical level and affected all users. To the department’s credit, it reported experiencing only 2 system outages since July 2014—a 30-minute critical outage in December 2014 and a 23- minute critical outage in May 2015. In addition to system availability, VBA monitors system response times for each of the VBMS modules using an application that measures the amount of time taken for each transaction. The PMO defines response time as the time that elapses from when a user executes a transaction (i.e., clicks a link or selects “Enter”) to when the resulting page fully loads on the user’s screen. The PMO uses average page response times to measure VBMS system performance, with lower response times indicating optimal system performance and higher response times indicating performance issues. From September 2013 through April 2015, VBA reported a decrease in average response times for VBMS-Core and VBMS-Rating (see table 2). VBA attributed the decrease in response times to continuous engineering improvements to system performance. Program officials also explained that the difference in response times between modules was due to the type of information that is being pulled into each module from various other VBA systems. For example, both VBMS-Core and VBMS-Rating require information from the VBA corporate database, but VBMS-Core is populated with data from multiple VBA systems in addition to the corporate database. Program officials told us there are no specific goals for mean transaction response times because they feel that there are adequate tools in place to monitor system performance and provide alerts if there are response time issues. For example, VBMS performance is monitored in real time by dedicated staff at a contractor’s facility and users have access to a live chat feature where they can provide feedback on any issues they are experiencing with the system. The VBMS help desk provides another avenue for users to provide feedback on the system’s performance. Officials also noted that, because transaction response times have decreased, which can be indicative of an improvement to system performance, they are focusing their resources on adding additional functionality instead of trying to get the system to achieve a specific average transaction response time. While VBA’s monitoring of VBMS’s performance is commendable and the system’s performance and response time have improved over time, the system is still in development and there is no guarantee that performance will remain at current levels as the system evolves. Performance targets and goals for VBMS response times would provide users with an expectation of the system response times they should anticipate, and management with an indication of how well the system is performing relative to performance goals. A key element of successful system testing is appropriately identifying and handling defects that are discovered during testing. Outstanding defects can delay the release of functionality to end users, denying them the benefit of features. Key aspects of a sound defect management process include the planning, identification and classification, tracking, and resolution of defects. Leading industry and government organizations consider defect management and resolution to be among the primary goals of testing. The VBMS program has defect management policies in place and it is actively performing defect management activities. Specifically, in October 2012, the department developed a VBMS Program Management and Technical Support Defect Management Plan, Version 1.0, which describes the program’s defect management process. The plan was updated in March 2015 and describes, among other things, the process for identifying, classifying, tracking, and resolving VBMS defects. For example, it provides criteria for assigning four different severity levels for defects—critical, high, medium, and low. According to the plan, critical severity defects are characterized by complete system or subsystem failure, complete loss of functionality, and compromised security or confidentiality. Critical defects also have extensive user impact and workarounds do not exist. High severity defects can have major user impact, leading to significant loss of system functionality. Medium severity defects can have moderate user impact and lead to moderate loss of functionality. Low severity defects lead to minor loss of functionality with no workaround necessary. For high and medium severity defects, workarounds could exist. According to the PMO, high, medium, and low defects do not need to be resolved prior to a system release. To monitor and track defects, the PMO uses an automated tool to maintain the VBMS defect repository. This tool is used to produce a daily defect management report that is shared with VBMS leadership. The purpose of the daily defect management report is to provide the current status of all open defects identified in testing of a forthcoming VBMS release or identified during production of a previous release. According to the defect management plan, defects can be resolved in a number of different ways. Once a defect is fixed, tested, and has passed testing, it is considered done or resolved. Defects that cannot be attributed to an existing requirement are reclassified as a system enhancement and considered resolved, as they do not affect a current system release requirement. A defect is also considered resolved if it is determined to work as designed, duplicate another defect, or if it is no longer evident in the system. From March 2014 through March 2015, the total number of VBMS defects declined as release dates approached for releases 7.0, 7.1, 8.0, and 8.1. Additionally, to the department’s credit, no critical defects remained at the time of each of these releases. Specifically, prior to VBMS Release 7.0, the total number of defects identified peaked at 226, while 97 remained at the time of release. Prior to release 7.1, the total number of defects identified peaked at 330, with 169 remaining at the time of release. Prior to release 8.0, the total number of defects identified peaked at 330, with 161 remaining at the time of release. Prior to release 8.1, the total number of defects remaining peaked at 421, with 254 remaining at the time of release. Figure 3 shows the trend in the number of open total defects for this time period. Even with the department’s efforts to resolve defects prior to a VBMS release, defects that affected system functionality remained open at the time of the releases. Specifically, of the 254 open defects at the time of VBMS release 8.1, 76 were high severity, 99 were medium severity, and 79 were low severity. Examples of defects that remained open at the time of VBMS release 8.1 are described in table 3. According to the PMO, these defects were communicated to users and an appropriate workaround for each was established. Nevertheless, even with workarounds, high- and medium-severity open defects, which by definition impact system functionality, degraded users’ experiences with the system. Continuing to deploy system releases with defects that impact system functionality increases the risk that these defects will diminish users’ ability to process disability claims in an efficient manner. While VBA has several methods to obtain VBMS users’ feedback, it has neither established goals to define user satisfaction, nor conducted a survey of claims processing employees to obtain a comprehensive picture of overall customer satisfaction. Our survey of VBMS users found that a majority reported satisfaction with the system, but decision review officers were considerably less satisfied. Although the results of our survey provide VBA with useful data about users’ satisfaction with the system, the absence of user satisfaction goals limits the utility of survey results. GAO and federal IT guidance recognize the importance of defining program goals and related performance targets and using such targets to assess progress in achieving the goals. Also, leading practices identify continuous customer feedback as a crucial element of IT project success, from project conception through sustainment. Particularly for IT projects like VBMS, where development activities are iterative, customer (i.e., end user) perspectives and insights can be solicited through various methods—user acceptance testing, interviews, complaint programs, and satisfaction surveys—to validate or raise questions about the project’s implementation. Further, leading practices emphasize that periodic customer satisfaction data be proactively used to improve performance and demonstrate the level of satisfaction the project is delivering. The Office of Management and Budget has developed standards and guidelines in survey research that are generally consistent with best practices and call for statistically valid data collection efforts to be used in fulfilling agencies’ customer service data collection. These leading practices also stress the importance of centrally integrating all customer feedback data in order to have more complete diagnostic information to guide improvement efforts. VA has used a variety of methods for obtaining customer or end user feedback on the performance of VBMS. For example, the department solicits end user involvement and feedback in the iterative system development process based on user acceptance criteria. According to the Senior Project Manager for VBMS Development within OI&T, at the end of each development cycle and before a new version of VBMS is deployed, end users are involved in user acceptance testing and a final customer acceptance meeting. VA also provides training to a subset of end users—known as “superusers”—on the updated functionality introduced in a new version of VBMS. These superusers are then expected to train the remaining users in the field on the new version’s features. The department tracks the overall satisfaction level with training received after each VBMS major release. However, this tracking is limited to superusers’ satisfaction with the training, rather than their satisfaction with the system. VA also solicits customer feedback about the system through interviews. For example, the PMO Director stated that the Under Secretary for Benefits hosts a weekly phone call with bargaining unit employees as a “pulse check” on VBA transformation activities, including VBMS. According to this official, the VBA Office of Field Operations also offers an instant messaging chat service to all regional office employees to solicit feedback about the latest VBMS functionality deployment. Another method in which the department obtains customer input is through a formal feedback process. For example, according to the PMO Director, VA provides national service desk support to assist users in troubleshooting system issues and identifying system defects. In addition, VBMS applications include a built-in feature that enables users to provide feedback to the PMO on problems with the system. According to the Director, the feedback received by the office also helps to identify user training issues. Nevertheless, while VA has taken steps to obtain feedback on the performance and implementation of VBMS, it has not established goals to define user satisfaction that can be used as a basis for gauging the success of its efforts to promote satisfaction with the system. Further, although the efforts that have been taken to solicit users’ feedback provide VBA with useful insights about particular problems, data are not centrally compiled or sufficient for supporting overall conclusions about whether customers are satisfied. In addition, VBA has not employed a customer satisfaction survey of claims processing employees who use the system on a daily basis to process disability claims. Such a survey could provide a more comprehensive picture of overall customer satisfaction and help identify areas where VBMS development and implementation efforts might need additional attention. According to the PMO Director, VBA had not used a survey to solicit feedback because of concern that such a mechanism may have negatively impacted the efficiency of claims processors in completing disability compensation claims on behalf of veterans. Further, the director believed that the office had the benefit of receiving ongoing end user input on VBMS by virtue of the intensive testing cycles, as well as several of the other mechanisms noted previously by which end users have provided ongoing feedback. Nevertheless, without establishing user satisfaction goals and collecting the comprehensive data that a statistically valid survey can provide, the PMO limits its ability to obtain a comprehensive understanding of VBMS users’ levels of satisfaction with the system. Thus, VBA could miss opportunities to improve the efficiency of its claims process by increasing satisfaction with VBMS. In response to the statistical survey that we administered, a majority of VBMS users (i.e., VBA claims processors) were satisfied with the system that had been implemented at the time of the survey. These users (which represent claims assistants, veteran service representatives, supervisory veteran service representatives, rating veterans service representatives, decision review officers, and others) were satisfied with the three modules of VBMS. Specifically, an estimated 59 percent of the population of claims processors were satisfied with VBMS-Core; an estimated 63 percent were satisfied with the Rating module, and an estimated 67 percent were satisfied with the Awards module. Figure 4 depicts the estimated percentage of claims processors who were satisfied with VBMS. Although a majority of users were satisfied with the three modules, decision review officers were considerably less satisfied than other users with VBMS-Core and VBMS-Rating. Specifically, for VBMS-Core, an estimated 27 percent of decision review officers were satisfied compared to an estimated 59 percent of all roles of claims processors (including decision review officers) who were satisfied. In addition, for VBMS-Rating, an estimated 38 percent of decision review officers were satisfied, compared to an estimated 63 percent of all roles of claims processors. Figure 5 depicts the estimated satisfaction levels of decision review officers in comparison to all roles. Decision review officers were considerably less satisfied with VBMS in comparison to all roles of claims processors in additional areas. For example, an estimated 26 percent of decision review officers viewed VBMS-Core as an improvement over the previous legacy system or systems for establishing claims and storing and reviewing electronic documents related to a claim in an eFolder. In contrast, an estimated 58 percent of all users (including decision review officers) viewed the Core module as an improvement. In addition, an estimated 26 percent of decision review officers viewed VBMS-Rating as an improvement over the previous systems with respect to providing Web-accessible tools, including rules-based rating calculators, to assist in making claims rating decisions. In contrast, an estimated 55 percent of all roles of claims processors viewed the Rating module as an improvement. For VBMS- Awards, an estimated 61 percent of all roles viewed this module as an improvement over the previous systems to automate the award and notification process. Figure 6 depicts the estimated percentage of decision review officers, in comparison to all claims processors, who viewed VBMS as an improvement over legacy systems. Similarly, for the three modules, a majority of users (including decision review officers) would have chosen VBMS over the legacy system or systems. However, decision review officers indicated that they were less likely to have chosen VBMS-Core and VBMS-Rating over legacy systems. Specifically, an estimated 27 percent of decision review officers would have chosen VBMS-Core compared to an estimated 60 percent of all roles of claims processors. In addition, an estimated 27 percent of decision review officers would have chosen VBMS-Rating compared to 61 percent of all roles who would have chosen the system over the legacy system or systems. For VBMS-Awards, an estimated 67 percent of all roles would have chosen this module over the previous systems. Figure 7 depicts the estimated percentage of decision review officers, in comparison to all claims processors, who would have chosen VBMS instead of the legacy systems to process claims. Decision review officers perform an array of duties to resolve claims issues raised by veterans and their representatives. They may also conduct a new review or complete a review of a claim without deference to the original decision and in doing so, must click through all documents included in the e-Folder. Survey comments from decision review officers stated, for example, that reviews in the VBMS paperless environment take longer because of the length of time spent loading, scrolling, and viewing each document (particularly if the documents are large, such as a service medical record file). Additionally, multiple decision review officers commented that it is easier and faster to review documents in a paper file. Although such comments provide illustrative examples of individual decision review officer’s views and are not representative, according to the PMO Director, decision review officers’ relative dissatisfaction is not surprising because the system does not yet include functionality that supports their work, which primarily relates to appeals processing. Even though VA has made progress toward completing the development and implementation of VBMS, there is more work to be done and VBA can improve management of its ongoing efforts. While 95 percent of records related to veterans’ disability claims are electronic and reside in VBMS, additional capabilities to fully process disability claims will be delayed beyond fiscal year 2015, which is when completion was originally planned. Further, VA’s incremental approach to developing and implementing VBMS has not yet produced a plan that identifies when the system will be completed and can be expected to fully support disability compensation and pension claims processing and appeals. Thus, it will be difficult for VA to hold its managers accountable for meeting its time frame and for demonstrating progress. VBA’s management of the system highlights areas that could benefit from improvement as development and implementation of the system continue. Specifically, without a reliable estimate of the total costs associated with completing work on VBMS, the department’s stakeholders have only a limited view of future resource needs, and the program risks not having sufficient funding to complete development and implementation of the system. Additionally, established goals for system response times would provide users with an expectation of the response times they can anticipate, and management with an indication of how well the system is performing relative to performance goals. Furthermore, continuing to deploy system releases with large numbers of defects that reduce system functionality could adversely affect users’ ability to process disability claims in an efficient manner. Without user satisfaction goals and the data a customer satisfaction survey could yield, VA may miss opportunities to collect important data on how users view the system’s performance, and ultimately, to improve the system. Our survey of VBMS users found that a majority of them were satisfied with the system, but decision review officers were considerably less satisfied. Although the results of our survey provide VBA with useful data about users’ satisfaction with VBMS, the absence of user satisfaction goals limits the utility of survey results. Specifically, without having established goals to define user satisfaction, VBA does not have a basis for gauging the success of its efforts to promote satisfaction with the system or for identifying areas where its efforts to complete development and implementation of the system might need attention. We recommend that the Secretary of Veterans Affairs direct the Under Secretary for Benefits and the Chief Information Officer to take the following five actions to improve VA’s efforts to effectively complete the development and implementation of VBMS: Develop an updated plan for VBMS that includes (1) a schedule for when VBA intends to complete development and implementation of the system, including capabilities that fully support disability claims, pension claims, and appeals processing and (2) the estimated cost to complete development and implementation of the system. Establish goals for system response time and use the goals as a basis for periodically reporting actual system performance. Reduce the incidence of high- and medium-priority level defects that are present at the time of future VBMS releases. Develop and administer a statistically valid survey of VBMS users to determine the effectiveness of steps taken to make improvements in users’ satisfaction. Establish goals that define customer satisfaction with VBMS and report on actual performance toward achieving the goals based on the results of GAO’s survey of VBMS users and any future surveys VA conducts. We received written comments on a draft of this report (reprinted in appendix III). In addition, VA provided technical comments, which we incorporated, as appropriate. In its comments, VA generally agreed with our conclusions. The department also concurred with our recommendations and described actions it is planning to take in response to four of our five recommendations. Specifically, VA concurred with our recommendation calling for an updated plan for VBMS. Although it described the importance of VBMS to ensuring timely delivery of benefits to veterans and recognized the need for continued investment in the system, the department did not, however, identify actions to develop a VBMS plan that includes a schedule for when VBA intends to complete development and implementation of the system and the estimated cost of doing so. We believe development of such a plan is an important action to help ensure effective development and implementation of VBMS and to hold managers accountable. With regard to the remaining four recommendations, VA described actions it is planning to take in response to each. For example, with regard to our recommendation to establish goals for system response time, the department stated that the VBMS program is participating in a review of service-level agreements to establish metrics for the system’s performance. Additionally, regarding our recommendation that it reduce the incidence of high and medium priority level defects that are present in future VBMS releases, the department reiterated its plans and procedures for decreasing the defects in each release. With respect to our recommendation to conduct a survey of VBMS users, the department stated that the VBMS PMO is working with the Office of Field Operations and labor partners regarding the distribution of a survey to measure users’ satisfaction and expects to release a survey in March 2016. Consistent with our final recommendation, the department is also planning to establish customer satisfaction goals and report on actual performance toward achieving the goals after all survey results are received and analyzed with a target of July 2016 for completing this action. If VA develops the updated plan, including schedule and cost, for VBMS as we recommended and follows through on the actions it described in response to our remaining recommendations, the department will be better positioned to effectively complete the development and implementation of VBMS and to more effectively provide benefits and services to our nation’s veterans. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to appropriate congressional committees, the Secretary of Veterans Affairs, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have questions about this report, please contact me at (202) 512-6304 or melvinv@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix IV. The objectives of this study were to (1) assess the Department of Veterans Affairs’ (VA) progress toward developing and implementing the Veterans Benefits Management System (VBMS) and (2) determine to what extent users report satisfaction with the system. To assess the department’s progress toward developing and implementing VBMS, we obtained and reviewed program plans and other documentation that articulated the system’s goals and expected benefits. These included the VA Strategic Plan to Eliminate the Compensation Claims Backlog (2013), VBMS Strategic Roadmap, VBMS Tactical Roadmap, VBMS Operating Plans, VBA’s 2011 life-cycle cost estimate for the program, documentation supporting VBMS’s progression through VA’s milestone review process, VA’s annual budget submissions, and contracts and related contractor documentation. We compared program plans and other documentation articulating the system’s goals, expected benefits, and the system functionality expected to be delivered with documents showing VBMS’s progress, such as monthly program management reviews and cost and performance reports. We also compared program plans to VA’s policies for incremental system development, as well as federal guidance and IT project management principles on program planning and cost estimating. In addition, we obtained reports documenting the system’s availability and performance and analyzed information in the reports to determine trends in system availability and response time. We compared system performance data with GAO and federal IT guidance for defining program goals and related performance targets that can be used to assess progress in achieving the goals. We assessed the reliability of the data by reviewing it for obvious errors and missing data; corroborating the data with related documentation; and interviewing responsible officials about their use of an automated tool that monitors system performance. We determined the data to be sufficiently reliable for the purposes of this report. We reviewed the VBMS Defect Management Plan and compared the plan with key principles of sound defect management. We also compared the processes described in the plan with actions taken to manage defects identified for VBMS. We conducted analysis of critical defects identified for VBMS releases 7.0, 7.1, 8.0, and 8.1 to determine whether they remained open at the time of system release. Further, we reviewed the department’s methods for soliciting end user feedback on the performance of the system (e.g., VBMS training strategy, operational management reviews, cost and performance reports, and end of month reports). We compared these methods to leading practices for obtaining customer feedback and collecting customer service data to improve performance and demonstrate customer satisfaction of IT projects. To determine the extent that users report satisfaction with the system, we conducted a Web-based survey of a nationally representative stratified random sample of disability compensation claims processors. The survey was administered during the time frame of September through November 2014. We developed survey questions with input from officials in the VBMS PMO and VA’s Office of Information and Technology. We pretested versions of the draft survey and observed claims processors to gain initial insight regarding their use of the system in VBA’s Baltimore, Maryland, and Philadelphia, Pennsylvania, regional offices. We selected these offices based on their large size and location. We revised the draft survey based on comments received during the pretests. Once finalized, the claims processors included in the sample were sent an e-mail that asked them to complete the survey, which was available to them over the Web from September 30, 2014, to November 19, 2014. Following the initial request, we e-mailed weekly follow-up requests to any nonrespondents. Our eligible population for this survey consisted of all 10,622 disability compensation claims processors employed by the agency as of July 31, 2014. To determine whether VBMS experiences differed based on position and office workload, we designed a stratified random sample of claims processors, with strata defined by position and office workload. VBA provided us lists of these claims processors with information that indicated their position and office. We determined average office workload based on information published in VBA’s public weekly workload reports. We divided the list of claims processors into 10 strata based on their position at both smaller and larger workload regional offices. Table 4 shows the size of each eligible population, the size of the sample drawn from each eligible population, and the number of claims processors in each stratum that responded to the survey. We drew an independent random sample from each of these strata to enable us to project survey results to all VBA claims processors, in general, and to claims processors in each of the positions listed above, within small and large workload offices. Of our total sample of 3,475 claims processors, 2,098 responded to the survey for an overall response rate of 60 percent. To produce estimates regarding the experiences and views of claims processors from the survey responses of those included in our sample, we analyzed the data with methods that are appropriate for a stratified random sample using analysis weights. We weighted each response from claims processors in each stratum to statistically account for all members of that stratum. Because estimates are based on responses from a sample, each estimate we report has a measurable precision or sampling error. The sampling error or margin of error surrounding an estimate is expressed as a number of percentage points higher or lower than that estimate and the entire range that the sampling error covers is referred to as a confidence interval. Confidence intervals are calculated based on a certain confidence level. Confidence intervals for estimates we report from this survey are based on a confidence level of 95 percent and are calculated using methods appropriate for a stratified random sample. Confidence intervals for percentage estimates in this report are never wider than plus or minus 5 percentage points. At a 95 percent confidence level, this means that in about 95 out of 100 instances, the sampling procedures we used would be expected to produce a confidence interval (in this case, a 10 percentage point range) containing the true population value we estimate. In addition to sampling error, estimates based on survey results are subject to what is referred to as nonsampling error that can result from, among other things, poorly designed survey questions and mistakes in data entry or analysis or nonresponse. We took a number of steps in developing the survey and in entering and analyzing the data to minimize nonsampling error. For example, a social science survey specialist collaborated with our subject matter experts in designing the survey. In addition, it was reviewed by VBA officials and, as previously noted, was pretested with a number of different types of claims processors in two locations. Also, when we analyzed the data, an independent analyst verified all computer programs. Because this was a Web-based survey, respondents’ answers to survey questions were automatically entered into an electronic file, eliminating the need to separately key responses into a data file, further minimizing the potential for nonsampling error. We also took steps to mitigate potential nonresponse error. For example, we used follow-up e-mails to remind users to complete the survey in order to reduce nonresponse. Further, we adjusted for characteristics that were associated with survey response propensity using standard weighting class adjustments defined by sampling strata. We assumed that nonresponse adjusted data were missing at random, and therefore concluded the respondent analyses using the nonresponse adjusted weights were unbiased for the population of VBMS users sampled in the survey. We supplemented our analyses with interviews of VA officials that had knowledge of the VBMS program, including officials in VA’s Office of Information and Technology and the Veterans Benefits Administration’s VBMS PMO and Office of Field Operations. We conducted this performance audit from March 2014 through August 2015 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. The questions we asked in our survey of Veterans Benefits Administration (VBA) claims processors are shown here. Our survey was comprised of closed- and open-ended questions. In this appendix, we include all survey questions and aggregate results of responses to the closed-ended questions. For a more detailed discussion of our survey methodology, see appendix I. 1. What is your current role in processing veterans’ disability compensation claims? If other, describe your role. (Open-ended response.) 2. What responsibility (or responsibilities) do you have in this role (Check all that apply.) If other, describe your responsibility. (Open-ended response.) 3. How many years have you worked within the VBA? 4. The Department of Veterans Affairs’ (VA) records show that you currently work at the (office name prepopulated) office. Is this correct? If no, enter the name of the office where you currently work. (If you work in more than one office, enter the office where you spend most of your time.) (Open-ended response.) 5. How long have you been using the Veterans Benefits Management System (VBMS)? Response Less than 6 months 95 percent confidence interval - lower bound 6.9 95 percent confidence interval - upper bound 9.8 6 months to less than 1 year 1 year to less than 2 years 2 years to less than 3 years Not applicable, I do not use VBMS (Skip to question 17.) 6. In an average work week, how much time, if at all, do you spend processing the following types of claims in VBMS? (Choose a time period for each item listed.) Response 6a. Fully electronic claims 95 percent confidence interval - lower bound 8.0 95 percent confidence interval - upper bound 10.6 8 to less than 16 hours 16 to less than 24 hours 24 to less than 32 hours 32 to less than 40 hours All week (40 or more hours) I do not process these 6b. Fully paper- based claims 8 to less than 16 hours 16 to less than 24 hours 24 to less than 32 hours 32 to less than 40 hours All week (40 or more hours) I do not process these 6c. Hybrid claims (that is, claims that consist of both paper and electronic records) 8 to less than 16 hours 16 to less than 24 hours 8.7 24 to less than 32 hours 95 percent confidence interval - lower bound 2.9 95 percent confidence interval - upper bound 4.9 32 to less than 40 hours All week (40 or more hours) 7. In addition to using VBMS, how much do you depend on each of the following systems or applications to process disability claims? (Choose a category for each system or application listed.) None or do not use 95 percent confidence interval - lower bound 49.0 95 percent confidence interval - upper bound 52.9 7b. VETSNET: Modern Award Processing- Development (MAP-D) None or do not use 7c. VETSNET: Rating Board Automation 2000 (RBA 2000) Section 2—Veterans Benefits Management System-Core (VBMS-Core) 8. In the course of your work, do you use VBMS-Core? Complete questions 8.1 and 8.2 for VBMS-Core. In question 8.3, you will have a chance to comment on any of your answers after responding to the questions. 8.1. Do you agree or disagree with the following statements based on your current experience using VBMS-Core? Select one answer in each row. Not applicable or no basis to judge 8.1c. VBMS-Core helps me be more productive compared to the previous system(s) Not applicable or no basis to judge 8.1d. VBMS-Core helps me be more efficient compared to the previous system(s) Not applicable or no basis to judge 8.1f. VBMS-Core is easier to use compared to the previous system(s) Not applicable or no basis to judge 8.1n. VBMS-Core performance is timely (e.g., minimal time to load pages and respond to commands) Response 8.1o. I believe VBMS-Core helps to reduce claims processing times compared to the previous system(s) Not applicable or no basis to judge 8.1r. VBMS-Core is an improvement over the previous system(s) 8.2. While using VBMS-Core how often, if at all, do you experience each of the following? Select one answer in each row. Response 8.2a. The scanned documents displayed on the screen within VBMS- Core are clearly legible and correctly oriented (i.e., not upside-down) 8.3. If you have any comments regarding your answers to questions 8.1 and/or 8.2 concerning VBMS-Core, share them here. (Open-ended response.) Section 3—Veterans Benefits Management System-Rating (VBMS-R) 9. In the course of your work, do you use VBMS-R? Complete questions 9.1 and 9.2 for VBMS-R. In question 9.3, you will have a chance to comment on any of your answers after responding to the questions. 9.1. Do you agree or disagree with the following statements based on your current experience using VBMS-R? Select one answer in each row. Not applicable or no basis to judge 9.1c. VBMS-R helps me be more productive compared to the previous system(s) Not applicable or no basis to judge 9.1d. VBMS-R helps me be more efficient compared to the previous system(s) Not applicable or no basis to judge 9.1f. VBMS-R is easier to use compared to the previous system(s) Not applicable or no basis to judge 9.1n. VBMS-R performance is timely (e.g., minimal time to load pages and respond to commands) Not applicable or no basis to judge 9.1o. I believe VBMS-R helps to reduce claims processing times compared to the previous system(s) Not applicable or no basis to judge 9.1r. VBMS-R is an improvement over the previous system(s) 9.2. While using VBMS-R, how often, if at all, do you experience each of the following? Select one answer in each row. 9.3. If you have any comments regarding your answers to questions 9.1 and/or 9.2 concerning VBMS-R, share them here. (Open-ended response.) Section 4—Veterans Benefits Management System-Awards (VBMS-A) 10. In the course of your work, do you use VBMS-A? Complete questions 10.1 and 10.2 for VBMS-A. In question 10.3, you will have a chance to comment on any of your answers after responding to the questions. 10.1. Do you agree or disagree with the following statements based on your current experience using VBMS-A? Select one answer in each row. Not applicable or no basis to judge 10.1c. VBMS-A helps me be more productive compared to the previous system(s) Not applicable or no basis to judge 10.1d. VBMS-A helps me be more efficient compared to the previous system(s) Not applicable or no basis to judge 10.1f. VBMS-A is easier to use compared to the previous system(s) Not applicable or no basis to judge 10.1n. VBMS-A performance is timely (e.g., minimal time to load pages and respond to commands) Not applicable or no basis to judge 10.1o. I believe VBMS-A helps to reduce claims processing times compared to the previous system(s) Not applicable or no basis to judge 10.1r. VBMS-A is an improvement over the previous system(s) 10.2. While using VBMS-A how often, if at all, do you experience each of the following? Select one answer in each row. Response 10.2a. VBMS-A gives error messages that clearly tell me how to fix problems. 10.3. If you have any comments regarding your answers to questions 10.1 and/or 10.2 concerning VBMS-A, please share them here (Open-ended response.) Section 5—VBMS Benefits and Challenges 11. How much, if at all, has VBMS improved your ability to do the following? Select one answer in each row. Not applicable or no basis to judge 11d. Access claims documentation on demand (e.g., the ability to electronically view a Veteran’s claim file at multiple regional offices at the same time) Not applicable or no basis to judge 11f. Other improvement(s) If other improvement(s), describe. (Open-ended response.) 12. Which of the following would you consider to be the single greatest improvement in using VBMS? Make more accurate rating decisions Use automated tools (e.g., standardized correspondence, rating application evaluation builder, rules-based calculators, etc.) for quicker/more accurate processing Access claims documentation on demand (e.g., the ability to electronically view a Veteran’s claim file at multiple regional offices at the same time) If other improvement, please describe the other VBMS improvement you consider to be the greatest. (Open-ended response.) 13. How much of a challenge, if at all, has VBMS been to you in the following areas? Select one answer in each row. Response 13a. System performance (e.g., slow response times) Not applicable or no basis to judge 13b. System access (e.g., system outages) Response 13c. The need to rely on previous systems in addition to VBMS (e.g., VETSNET systems) Not applicable or no basis to judge 13d. Knowing whether the evidence required to process claims is complete in VBMS (e.g., not knowing whether all documents have been scanned and added to the veteran’s file) Not applicable or no basis to judge 13e. Other challenge(s) If other challenges(s), describe. (Open-ended response.) 14. Which of the following would you consider to be the single greatest challenge in using VBMS? Response System performance (e.g., slow response times) System access (e.g., system outages) The need to rely on previous systems in addition to VBMS (e.g., VETSNET systems) Knowing whether the evidence required to process claims is complete in VBMS (e.g., not knowing whether all documents have been scanned and added to the veteran’s file) If other challenge, describe the other VBMS challenge you consider to be the greatest. (Open-ended response.) 15. If you could pick one change to be included in the next release/update of VBMS, what would it be and why? (Open-ended response.) 16. Please share any comments or suggestions for improvement you have about the VBMS system. (Open-ended response.) In addition to the contact named above, Mark T. Bird (Assistant Director), Chris Businsky, Virginia Chanley, Heather A. Collins, Kavita Daitnarayan, Kelly Dodson, Nancy Glover, Stuart Kaufman, Michael Little, Jamelyn Payan, Scott Pettis, Martin L. Skorczynski, Roger M. Smith, Eric Trout, Sonya Vartivarian, and Kevin Walsh made key contributions to this report.

VBA pays disability benefits for disabling conditions incurred or aggravated while in military service, while pension benefits are for low-income veterans who are either elderly or have disabilities unrelated to military service. In fiscal year 2014, the department paid about $58 billion in disability compensation and about $5 billion in pension claims. The disability claims process has been the subject of attention by Congress and others, due in part, to long waits for processing claims and a large backlog of claims. To process disability and pension claims more efficiently, VA began implementation of an electronic, paperless system in 2009. GAO was asked to study VBMS. Specifically, GAO (1) assessed VA's progress toward completing the development and implementation of VBMS and (2) determined to what extent users report satisfaction with the system. To do so, GAO reviewed relevant program documentation, administered a survey to a stratified random sample of about 3,500 users, and interviewed appropriate VA officials. The Veterans Benefits Administration (VBA) within the Department of Veterans Affairs (VA) has made progress in developing and implementing the Veterans Benefits Management System (VBMS), with deployment of the system to all of its regional offices as of June 2013. While 95 percent of records related to veterans' disability claims are electronic and reside in the system, additional capabilities have not yet been completed, such as automation of the steps associated with a veteran's request for an increase in benefits. Further, VBA has not yet developed and implemented pension processing capabilities in VBMS, nor has it articulated when the system will support appeals processing. The VBMS program reported receiving funding of about $1 billion from fiscal years 2009 to 2015, at which time system completion was originally planned. Although development of the system is expected to continue beyond 2015, the incremental approach VA is using to develop and implement VBMS has not yet produced a plan that identifies when the system will be completed and can be expected to fully support disability and pension claims processing and appeals. Thus, it will be difficult for VA to hold its managers accountable for meeting its time frame and for demonstrating progress. As VA continues its efforts to complete development and implementation of the system, three areas could benefit from increased management attention. Cost estimating: The program office does not have a reliable estimate of the cost for completing the system. Without such an estimate, VA management and the department's stakeholders have a limited view of the system's future resource needs, and the program risks not having sufficient funding to complete development and implementation of the system. System availability: Although VBA has improved its performance for ensuring the system is available to users, it has not established system response time goals. Without such goals, users do not have an expectation of the system response times they can anticipate and management does not have an indication of how well the system is performing relative to performance goals. System defects: While the program has actively managed system defects, a recent system release included unresolved defects that impacted system performance and users' experiences. Continuing to deploy releases with large numbers of defects that reduce system functionality could adversely affect users' ability to process disability claims in an efficient manner. While VBA has employed various methods to obtain VBMS users' feedback, it has neither established goals to define user satisfaction, nor conducted a survey of claims processing employees to obtain a more comprehensive picture of overall customer satisfaction. GAO's survey of VBMS users estimated that a majority report satisfaction with the system, but that one group of users who are responsible for examining claims decisions was considerably less satisfied. Although the results of GAO's survey provide VBA with useful data about users' satisfaction with the system, the absence of user satisfaction goals limits the utility of survey results. Specifically, without having established goals to define user satisfaction, VBA does not have a basis for gauging the success of its efforts to promote satisfaction with the system. GAO recommends that VA develop a plan for completing VBMS, establish goals for system response time, minimize the incidence of high and medium priority system defects for future VBMS releases, assess user satisfaction, and establish satisfaction goals to promote improvement. VA concurred with the recommendations and described actions it is planning to take in response, except for the first recommendation. GAO continues to believe development of a plan for completing the system is important.

Although people across the country were exposed through the media to the emotional trauma of the WTC attack, the residents, office workers, and others living, working, or attending school in the WTC area and the WTC responders not only experienced the traumatic event in person but also were exposed to a complex mixture of potentially toxic contaminants in the air and on the ground, such as pulverized concrete, fibrous glass, particulate matter, and asbestos. Almost 3,000 people, including some who were trapped above the impact zone and others who entered the buildings to assist in the evacuation, were killed in the attack. The majority of the estimated 16,400 to 18,800 people who were in the WTC complex that morning were able to evacuate, however, with minor or no injuries. An estimated 40,000 responders were at or in the vicinity of the WTC site or the Staten Island Fresh Kills landfill, participating in rescue, recovery, and cleanup efforts; conducting environmental and occupational health assessments; providing crisis counseling and other treatment; providing security; and assisting with the criminal investigation. The responders included personnel from many agencies at the federal, state, and local levels, as well as from organizations in the private sector, and various other workers and volunteers. The agencies and organizations include HHS’s Agency for Toxic Substances and Disease Registry (ATSDR), HHS’s Centers for Disease Control and Prevention (CDC), the Department of Energy, EPA, DOJ’s Federal Bureau of Investigation (FBI), DHS’s Federal Emergency Management Agency (FEMA), HHS’s National Institute for Occupational Safety and Health (NIOSH), HHS’s National Institute of Environmental Health Sciences (NIEHS), the Department of the Interior’s National Park Service, DOL’s Occupational Safety and Health Administration (OSHA), HHS’s Public Health Service Commissioned Corps, HHS’s Substance Abuse and Mental Health Services Administration (SAMHSA), DOD’s U.S. Coast Guard, DOJ’s U.S. Marshals Service, the New York State Department of Environmental Conservation, the New York State Emergency Management Office, the New York State National Guard, the New York State Office of Mental Health, the New York State Department of Health, the Metropolitan Transportation Authority’s New York City Transit, FDNY and emergency medical services (EMS), the New York City Department of Health and Mental Hygiene, the New York City Police Department (NYPD), the New York City Department of Design and Construction, the New York City Department of Environmental Protection, the New York City Department of Sanitation, the New York City Office of Emergency Management, the American Red Cross, and the Salvation Army. Recognizing a need to monitor and understand the full health effects of the WTC collapse, officials from various organizations secured federal funding to establish programs to monitor the health of affected people. FDNY sought federal support in order to provide comprehensive medical evaluations to its firefighters, and established its WTC Medical Monitoring Program (referred to here as the FDNY program). The Mount Sinai Clinical Center for Occupational and Environmental Medicine also sought federal support in the weeks following the attack to develop its WTC Worker and Volunteer Medical Monitoring Program (referred to here as the Mount Sinai program). Through its Federal Occupational Health (FOH) services, HHS initiated a WTC responder screening program for federal workers (referred to here as the FOH program) involved in WTC rescue, recovery, and cleanup activities. Similarly, the New York State Department of Health established the medical monitoring program for New York State responders (referred to here as the NYS program) engaged in emergency activities related to the September 11 attack. In addition, two registries were established to compile lists of exposed persons and collect information through interviews and surveys in order to provide a basis for understanding the health effects of the attack. The New York City Department of Health and Mental Hygiene contacted ATSDR in February 2002 to develop the WTC Health Registry. ATSDR provided technical assistance to the New York City Department of Health and Mental Hygiene and worked with FEMA to obtain funds for the WTC Health Registry for responders and people living or attending school in the vicinity of the WTC site, or working or present in the vicinity on September 11. Separately, Johns Hopkins received a grant from NIEHS to create another registry (referred to here as the Johns Hopkins registry) of WTC site workers who were involved in cleanup efforts. A wide variety of physical and mental health effects have been observed and reported across a wide range of people in the aftermath of the September 11 attacks. The health effects include various injuries, respiratory conditions, reproductive health effects, and mental health effects. Unlike the physical health effects, the mental health effects of the September 11 attacks were not limited to responders and people who were in the WTC area but were also experienced by people across the nation. Because most of the information about mental health effects comes from questionnaire or survey data, what is reported in most cases are symptoms associated or consistent with a disorder, such as PTSD, rather than a clinical diagnosis of a disorder. The most commonly reported mental health effects were symptoms associated with PTSD, depression, stress, and anxiety, as well as behavioral effects such as increases in substance use and difficulties coping with daily responsibilities. Although the total number of people injured during the WTC attack is unknown, data on hospital visits show that thousands of people were treated in its immediate aftermath for injuries, including inhalation injuries, musculoskeletal injuries, burns, and eye injuries. Unpublished data collected by the Greater New York Hospital Association from September 11 through September 28, 2001, showed 6,232 emergency room visits and 477 hospitalizations related to the attack in 103 hospitals in New York State and 1,018 emergency room visits and 84 hospitalizations related to the attack in nearby New Jersey hospitals. These numbers do not include injured people who may have been treated in more distant New York State, New Jersey, and Connecticut hospitals, in triage stations, or by private physicians, and those who did not seek professional treatment. More detailed information on injuries is available from the four hospitals closest to the WTC and a fifth hospital that served as a burn referral center. According to the New York City Department of Health and Mental Hygiene, between September 11 and September 13, 2001, these hospitals treated 790 people, 2 of whom later died, for injuries related to the attack (CDC, 2002c). The most common of these injuries were musculoskeletal injuries—such as fractures, sprains, and crush injuries—and inhalation injuries. The majority of people with injuries were treated and released, although about 18 percent required hospitalization. In addition, thousands of responders were treated for injuries, a small proportion of which were classified as serious, during the 10-month cleanup period. The disaster site was considered to be extremely dangerous, yet no additional life was lost after September 11. Using data from five Disaster Medical Assistance Teams (DMAT) temporary medical facilities and the four hospitals closest to the WTC site, researchers documented 5,222 visits by rescue workers to DMAT facilities and emergency rooms in the first month of the cleanup period (Berrios-Torres et al., 2003). During this month, musculoskeletal injuries were the leading cause of rescue worker visits and hospitalizations. Other injuries included burns and eye injuries. According to OSHA, despite logging more than 3.7 million work hours over the 10-month cleanup period, WTC site workers reported only 57 injuries that OSHA classified as serious because they resulted in lost workdays, yielding a lost workday injury rate of 3.1 injuries per 100 workers per year. This rate is lower than that seen in the type of construction deemed by OSHA to be the most similar to the WTC cleanup, specialty construction, which has a lost workday injury rate of 4.3. A range of respiratory health effects, including a new syndrome called WTC cough and chronic diseases such as asthma, were observed among people exposed to the WTC collapse and its aftermath. Many of the programs examining respiratory health effects are ongoing and have published only preliminary results. Nevertheless, the studies present a consistent collection of conditions among those people who were involved in rescue, recovery, and cleanup as well as those who lived and worked in the WTC vicinity. The most commonly reported conditions include cough, wheezing, shortness of breath, sinusitis, and asthma. Many of the findings on respiratory effects published to date have focused on firefighters, and FDNY medical staff first described WTC cough, which consists of persistent cough accompanied by severe respiratory symptoms, often in conjunction with sinusitis, asthma, and gastroesophageal reflux disease (GERD). Several studies report on other WTC responders, such as the police, ironworkers, and cleanup workers, and a few studies report on the respiratory effects among people living and working in lower Manhattan. Almost all of the FDNY firefighters who had responded to the attack experienced respiratory effects, and hundreds had to end their firefighting careers due to WTC-related respiratory illness. Within 48 hours of the attack, FDNY found that about 90 percent of its 10,116 firefighters and EMS workers who were evaluated at the WTC site reported an acute cough. The FDNY Bureau of Health Services also noted wheezing, sinusitis, sore throats, asthma, and GERD among firefighters who had been on the scene. During the first 6 months after the attack, FDNY observed that of the 9,914 firefighters who were present at the WTC site within 7 days of the collapse, 332 firefighters had WTC cough (Prezant et al., 2002). Eighty-seven percent of the firefighters with WTC cough reported symptoms of GERD. According to the FDNY Bureau of Health Services, symptoms of GERD are typically reported by less than 25 percent of patients with chronic cough. Some FDNY firefighters exhibited WTC cough that was severe enough for them to require at least 4 weeks of medical leave. Despite treatment of all symptoms, 173 of the 332 firefighters and one EMS technician with WTC cough showed only partial improvement. FDNY also found that the risk of reactive airway dysfunction syndrome, or irritant-induced asthma, and WTC cough was associated with intensity of the exposure, defined as the time of arrival at the site (Banauch et al., 2003). In addition, FDNY reports that one firefighter who worked 16-hour days for 13 days and did not use respiratory protection during the first 7 to 10 days was diagnosed with a rare form of pneumonia that results from acute high dust exposure (Rom et al., 2002). According to an official from the FDNY Bureau of Health Services, because one of the criteria for being a firefighter is having no respiratory illness, about 380 firefighters were no longer able to serve as firefighters as of March 2004 as a consequence of respiratory illnesses they developed after WTC exposure. Studies and screenings conducted among other responders—carpenters, cleanup workers, federal civilian employees, heavy equipment operators, ironworkers, mechanics, National Guard members, police officers, telecommunications technicians, truck drivers, and U.S. Army military personnel—have found respiratory health effects similar to those seen in FDNY firefighters. Some of the responders with existing respiratory conditions reported that symptoms worsened, and others reported that they developed new respiratory symptoms on or after September 11. The most commonly reported symptom was cough. For example, about 63 percent of officers from NYPD’s Emergency Services Unit who were evaluated about 1 to 4 months after September 11 reported having a cough (Salzman et al., 2004). Other symptoms observed among responders included chest tightness, nasal congestion, shortness of breath, sore throat, and wheezing. Unpublished results from respiratory health assessments of WTC site workers—including truck drivers, heavy equipment operators, mechanics, laborers, and carpenters—conducted by Johns Hopkins in December 2001 show that among those who reported no previous history of lower respiratory symptoms, 34 percent reported developing a cough and 19 percent reported wheezing. While some responders reported that symptoms improved or resolved a few months after the attack, others reported that they continued to experience symptoms. For example, initial results from screenings of 250 participants in Mount Sinai’s monitoring program show that 46 percent of these responders were still experiencing at least one pulmonary symptom and 52 percent were still experiencing an ear, nose, or throat symptom 9 months after the attack (Herbert and Levin, 2003). Surveys conducted among people living or working in lower Manhattan show that these people experienced respiratory health effects similar to those experienced by responders, such as nose or throat irritation and cough. For example, a door-to-door survey conducted by the New York City Department of Health and Mental Hygiene in three residential areas in lower Manhattan between October 25 and November 2, 2001, showed that the most frequently reported symptoms were nose or throat irritation (about 66 percent) and cough (about 47 percent) (CDC, 2002a). A NIOSH survey of federal employees working near the WTC site found that 56 percent of respondents reported having a cough (Trout et al., 2002). Other symptoms observed among those living or working in lower Manhattan include chest tightness, head or sinus congestion, shortness of breath, and wheezing. Some people reported that the WTC collapse and its aftermath exacerbated existing respiratory conditions, such as asthma, and others reported symptoms that developed after September 11, 2001. For example, a review of medical charts of children with existing asthma from a lower Manhattan clinic found that after September 11 there was a significant increase in asthma-related clinic visits among children who lived within 5 miles of the WTC site (Szema et al., 2004). Unpublished preliminary findings from a New York State Department of Health survey of NYC residents found that almost three-fourths of respondents living near the WTC site experienced new upper respiratory symptoms after September 11. For all measures of reproductive health studied except birth weight for gestational age, no differences were found between infants born to women who were in or near the WTC on September 11 and infants of those who were not. The Mount Sinai School of Medicine conducted a study of the 187 pregnant women who were either in or near the WTC on September 11. This study found no significant differences in average gestational duration at birth or average birth weight between infants of the women who were in or near the WTC on September 11 during their pregnancy and infants of the 2,367 women in the study’s comparison group, who were not (Berkowitz et al., 2003). Additionally, no significant differences in frequency of preterm births (less than 37 weeks of gestation) or in incidence of low birth weight (less than 2,500 grams) were observed. Nor was an association observed between symptoms of posttraumatic stress in the mother and frequency of preterm birth, low birth weight, or small-for-gestational-age infants. However, 8.2 percent of infants born to women who were in or near the WTC on September 11 were born with a birth weight below the tenth percentile for gestational age, compared to 3.8 percent of infants born to women in the study’s control group. This difference was still statistically significant after variables such as maternal age, race/ethnicity, sex of the infant, and maternal smoking history were taken into account. Because small-for- gestational-age infants are at risk for developmental problems, the Mount Sinai program includes a follow-up study in which researchers plan to obtain physical measurements of growth and perform assessments of early cognitive development. In the weeks and months after the WTC attack, people living, working, or attending school in NYC and responders involved in the rescue, recovery, and cleanup reported symptoms associated with PTSD, as did people across the nation. PTSD is an often debilitating and potentially chronic disorder that can develop after experiencing or witnessing a traumatic event. It includes such symptoms as difficulty sleeping, irritability or anger, detachment or estrangement, poor concentration, distressing dreams, intrusive memories and images, and avoidance of reminders of the trauma. People living or working near the WTC site reported a higher rate of symptoms associated with PTSD than did those living or working farther from the site. For example, researchers found that about 7.5 percent of Manhattan residents reported symptoms consistent with PTSD 5 to 8 weeks after the attack, with 20 percent of those living in close proximity to the WTC reporting symptoms (Galea et al., 2002a). Similarly, NIOSH surveys found that reports of symptoms consistent with PTSD were significantly higher among school staff in the WTC vicinity than among school staff working at least 6 miles from the WTC site (CDC, 2002a). Some groups of people, such as children and responders, were found to have experienced traumatic reactions to the attack. For example, a citywide survey of a representative sample of NYC fourth to twelfth graders 6 months after the attack found that over 10 percent reported having symptoms consistent with PTSD. The researchers who conducted this survey noted that these symptoms were five times more prevalent than pre-September 11 rates reported for other communities (Hoven et al., 2002). Responders, many of whom lost colleagues, were also affected. Initial findings from the Mount Sinai program show that about 22 percent of a sample of 250 WTC responders reported symptoms consistent with PTSD (Herbert and Levin, 2003). People across the nation also reported symptoms associated with PTSD. A nationwide survey comparing reactions in NYC to those across the country using a nationally representative sample of U.S. adults found that the prevalence of symptoms associated with PTSD 1 to 2 months after the attack was significantly higher in the NYC metropolitan area (11.2 percent) than in other major metropolitan areas (3.6 percent) and the rest of the country (4 percent) (Schlenger et al., 2002). Another nationally representative sample in a nationwide survey of U.S. adults shows that 17 percent of the U.S. population outside of NYC reported symptoms associated with PTSD 2 months after the attack (Silver et al., 2002). Although no baseline data are available on the prevalence of symptoms related to PTSD, typically about 3.6 percent of U.S. adults have a psychiatric diagnosis of PTSD during the course of a year. People living, working, and attending school in NYC and WTC responders, as well as people across the nation, reported symptoms associated with depression, stress, and anxiety. For example, in NYC, researchers found that about 9.7 percent of Manhattan residents surveyed 5 to 8 weeks after the attack reported symptoms consistent with depression (Galea et al., 2002a). Nine hospitals in NYC reported that from September 11 to September 24, 2001, the predominant symptoms related to the WTC attack were those associated with anxiety, stress, and depression (Greater New York Hospital Association, 2001). Data from these hospitals show that anxiety declined over the month following the attack but increased again around the time that the first case of anthrax in NYC was announced in mid-October 2001. A NIOSH survey of people working in schools near the WTC site also reported symptoms of depression (CDC, 2002a). Among the responders, initial screenings from the Mount Sinai program show that nearly 37 percent of 250 program participants reported symptoms associated with anxiety, insomnia, and depression (Herbert and Levin, 2003). In addition, a nationwide survey conducted 3 to 5 days after the attack in a nationally representative sample of U.S. adults found that 44 percent of those surveyed reported one or more substantial symptoms of stress, including having difficulty concentrating, feeling irritable, feeling upset when something reminds the person of the attack, having disturbing thoughts or dreams, and having trouble sleeping (Schuster et al., 2001). The behavioral effects in the aftermath of the WTC attack included increased use of substances such as alcohol, tobacco, and marijuana. Increased use of alcohol and tobacco was identified through surveys of the general population conducted by the states of Connecticut, New Jersey, and New York in the 3 months following the attack (CDC, 2002b). In Manhattan, researchers found that almost 29 percent of people who responded to a survey administered 5 to 8 weeks after September 11 reported increased use of cigarettes, alcohol, or marijuana after the attack (Vlahov et al., 2002). According to these researchers, this increase in substance use was still evident 6 months after September 11 (Vlahov et al., 2004a,b). The behavioral effects also included difficulty coping with daily responsibilities. Some NYC children and adolescents, family members, and other adults, including members of the response community, are still having difficulty coping 3 years after September 11. For example, an ongoing SAMHSA-supported youth mental health program in NYC is treating 220 children and adolescents who are having problems coping, such as having difficulties functioning in school. In addition, researchers affiliated with the New York University School of Medicine’s Child Study Center’s bereavement program for families of uniformed personnel killed in responding to the WTC attacks noted that the psychological and emotional reactions of children and adolescents directly affected by the attacks have diminished somewhat over time but that some children continue to be affected by the emotional state and coping difficulties of their parents. Of particular concern to these researchers are the widowed mothers, who are experiencing sustained distress at twice the level typically found in the general population and are having difficulty coping with their daily responsibilities, such as single parenthood, almost 3 years later. Some responders, such as members of FDNY, also report having difficulty coping in the aftermath of September 11. The programs established to monitor and understand the health effects of the attack vary in terms of which people are eligible to participate, methods for collecting information about the health effects, options for treatment referral, and number of years people will be monitored. (See table 1.) FEMA provided funding for most of these programs through interagency agreements with HHS. These programs are not centrally coordinated, but some of them are collaborating with each other. The six programs that have been created to monitor people who were exposed to the WTC attack and its aftermath vary in terms of populations eligible to participate. Although five of the programs focus on various responder populations, the largest program—the WTC Health Registry—is open not only to responders but also to people living or attending school in the vicinity of the WTC site, or working or present in the vicinity on September 11. Specifically, people eligible for participation in the WTC Health Registry include anyone who was in a building, on the street, or on the subway south of Chambers Street on September 11; residents and staff of or students enrolled in schools (prekindergarten through twelfth grade) or day care centers south of Canal Street on September 11; and those involved in rescue, recovery, cleanup, or other activities at the WTC site and/or WTC recovery operations on Staten Island anytime between September 11, 2001, and June 30, 2002. (See figure 1.) An estimated 250,000 to 400,000 people are eligible for the WTC Health Registry; however, the registry was planned with the expectation that 100,000 to 200,000 people would enroll. Together the FDNY program and the Mount Sinai program cover more than half of the estimated 40,000 WTC responders. The FDNY program is open to all 11,000 FDNY firefighters and all 3,500 FDNY EMS technicians, including firefighters and technicians who were not exposed. Some 12,000 other responders are eligible to participate in the Mount Sinai program. Responders who were government employees are eligible for participation in programs such as the FOH program, which is open to the estimated 10,000 federal workers who responded to the WTC attacks, and the NYS program, which was open to about 9,800 New York State employees and New York National Guard personnel who were directed to respond to the WTC disaster. In addition, approximately 12,000 members from three NYC unions and the NYC Department of Sanitation, whether they were responders or not, were eligible to participate in the Johns Hopkins registry. R e c t o r Concerns have been raised by community and labor representatives regarding the eligibility requirements for some of these programs, and while changes have been made to accommodate some of these concerns, others remain unresolved, particularly with respect to the WTC Health Registry. For example, the eligibility criteria for participation in the Mount Sinai program were initially more restrictive, covering responders who had been at the site at least 24 hours between September 11 and 14, 2001. After discussions with labor representatives and CDC officials, the program expanded its eligibility criteria to include additional responders who may not have been there on those days but were there later in September. In contrast, community and labor representatives have been unsuccessful in their attempts to expand the eligibility criteria of the WTC Health Registry. These representatives have noted that the geographic boundaries used by the registry exclude office workers below Chambers Street who were not at work on September 11 but returned to work in the following weeks; office workers, including several groups of city employees, working between Chambers and Canal Streets; and Brooklyn residents who may have been exposed to the cloud of dust and smoke. Registry officials told us that they understand the desire to be included but they believe coverage is adequate to provide a basis for understanding the health effects of the WTC attack. The monitoring programs vary in their methods for identifying those who may require treatment, and although none of these programs are funded to provide treatment, they provide varying options for treatment referral. Some programs refer participants to affiliated treatment programs, whereas others provide information on where participants can seek care. The FDNY program offers a comprehensive medical evaluation that includes collection of blood and urine for analysis, a pulmonary function test, a chest X-ray, a renal toxicity evaluation, a cardiogram, a hepatitis C test, and hearing and vision tests, as well as self-administered questionnaires on exposures and physical and mental health. Funds for the monitoring program do not cover treatment services. However, FDNY members who require treatment after being screened can obtain treatment and counseling services from the FDNY Bureau of Health Services and the FDNY Counseling Services Unit as a benefit of their employment. Similarly, under the Mount Sinai program, people receive a comprehensive physical examination that includes blood and urine analysis, a chest X-ray, a pulmonary function test, and complete self-administered as well as nurse-administered questionnaires on exposure, clinical history, and mental health. If a person requires follow-up medical care or mental health services but is unable to pay for the services, he or she can be referred for care to other Mount Sinai programs such as the Health for Heroes program, which is supported through philanthropic donations. The FOH and NYS programs also consist of medical evaluations of participants and self-administered health and exposure questionnaires. The FOH program conducted about 400 medical evaluations of federal workers. These evaluations included a physical examination, a pulmonary function test, a chest X-ray, and blood tests. Under the NYS program, the New York State Department of Civil Service Employee Health Service clinics or affiliated clinics conducted medical evaluations that included a physical examination and a pulmonary evaluation of almost 1,700 state workers. The questionnaires for both programs are more limited than the FDNY or Mount Sinai questionnaires; for example, they have fewer mental health questions. Under the FOH and NYS programs, workers who require care have been told to follow up with their primary care physicians under their own insurance. Unlike most of the other monitoring programs, the WTC Health Registry and the Johns Hopkins registry do not include a medical evaluation, and neither effort is affiliated with a treatment facility or program. Instead, the programs collect information from participants solely through questionnaires and provide information on where participants can seek care. The WTC Health Registry questionnaire is generally administered over the telephone. The program provides all participants with a resource guide of occupational, respiratory, environmental, and mental health facilities in New York State, New Jersey, and Connecticut where people can seek treatment. Some of the services provided by these facilities require health insurance, whereas others are free of charge. If in the course of a telephone questionnaire, a person’s responses to the mental health questions suggest that he or she may need to speak with a mental health professional, the person is given the option of being connected directly to a LIFENET counselor. The LIFENET counselor provides the person with information on where to go and whom to call for help with problems related to the WTC disaster. For the Johns Hopkins registry, the participants complete a mail-in questionnaire on physical and mental health. Responders who report mental health symptoms and agree to be recontacted may receive follow-up calls to refer them to mental health services. The referral process is facilitated by Columbia University’s Resiliency Program, which provides free, short-term mental health services to affected people. The Johns Hopkins registry also provides participants with brochures about health services and programs they may find useful, including information about the Mount Sinai program. The duration of the monitoring programs may not be long enough to fully capture critical information on health effects. Under current plans, HHS funding for the programs will not extend beyond 2009. For example, ATSDR entered into a cooperative agreement with the New York City Department of Health and Mental Hygiene in fiscal year 2003 with the intent to continue support of the WTC Health Registry for 5 years of its planned 20-year duration. Similarly, NIOSH awarded 5-year grants in July 2004 to continue the FDNY and Mount Sinai programs, which had begun in 2001 and 2002, respectively. Health experts involved in the monitoring programs, however, cite the need for long-term monitoring of affected groups because some possible health effects, such as cancer, do not appear until several decades after a person has been exposed to a harmful agent. They also emphasize that monitoring is important for identifying and assessing the occurrence of newly identified conditions, such as WTC cough, and chronic conditions, such as asthma. Although the monitoring programs began as separate efforts, some of the programs are collaborating with each other. In addition, there are other kinds of collaborative efforts, including those in which programs receive advice from various outside partners. The WTC Responder Health Consortium is an example of collaboration between monitoring programs. It was established by NIOSH in March 2004 to coordinate the existing health monitoring of WTC responders initiated by the FDNY and Mount Sinai programs and to facilitate data sharing. It awarded $81 million in 5-year grants to six institutions to become clinical centers for WTC health monitoring. FDNY and Mount Sinai serve as coordinating centers under the consortium, and the other four institutions are coordinated with Mount Sinai. Together, these institutions will provide follow-up health evaluations to a total of about 12,000 NYC firefighters and EMS technicians and up to 12,000 other WTC responders. Collaboration efforts have also been fostered between the monitoring programs and outside partners and researchers. For example, the WTC Registry has a Scientific Advisory Group that includes representatives from the Mount Sinai School of Medicine, FDNY, the Johns Hopkins University, Columbia University, Hunter College, New York Academy of Medicine, New York University, the New York State Department of Health, and the New Jersey Department of Health. The group has assisted the New York City Department of Health and Mental Hygiene and ASTDR in development of the WTC Registry protocol, selection of the eligible population, and analysis methods. It has been meeting with WTC officials quarterly since early 2002 to advise on such issues as data collection, study options, and guidelines for research studies to be done using the registry. In addition, EPA convened an expert review panel in March 2004 to obtain greater input on ongoing efforts to monitor the health effects of workers and residents affected by the WTC collapse. The panel consists of representatives from federal and NYC agencies involved in air monitoring; from WTC health effects monitoring programs; and from academic institutions and the affected community. The goals of the panel include identification of unmet public health needs, gaps in exposure data, gaps in efforts to understand the health effects of the WTC attack, and ways in which the WTC Health Registry could be enhanced to allow better tracking of workers and residents. A multitude of physical and mental health effects have been reported in the years since the terrorist attack on the World Trade Center on September 11, 2001, but the full health impact of the attack is unknown. Concern about potential long-term effects on people affected by the attack remains. The monitoring programs may not be in operation long enough to adequately capture information about new conditions, chronic conditions, and diseases whose onset may occur decades after exposure to a harmful agent, such as many cancers. Nevertheless, these programs are providing a more complete picture of the health impact of such events, and as they proceed they are also providing the opportunity to identify people needing treatment. We provided a draft of this testimony to DHS, DOL, EPA, and HHS. HHS provided written comments, in which it noted that the testimony does not include significant discussion on the ways in which mental health symptoms have changed over time. We relied primarily on data from published, peer-reviewed articles and government reports, and some of the researchers we spoke with emphasized that their studies are ongoing and they expect to publish further results. In the absence of these results, the evidence we examined did not support a full discussion of changes in mental or physical health effects over time. HHS and the other agencies also provided technical comments, which we incorporated as appropriate. Mr. Chairman, this completes my prepared statement. I would be happy to respond to any questions you or other members of the Subcommittee may have at this time. For further information about this testimony, please contact Janet Heinrich at (202) 512-7119. Michele Orza, Angela Choy, Alice London, Nkeruka Okonmah, and Roseanne Price made key contributions to this statement. Banauch, G.I., et al. “Persistent Hyperreactivity and Reactive Airway Dysfunction in Firefighters at the World Trade Center.” American Journal of Respiratory and Critical Care Medicine, 168, no. 1 (2003): 54-62. Berkowitz, G.S., et al. “The World Trade Center Disaster and Intrauterine Growth Restriction.” (Letter) Journal of the American Medical Association, 290, no. 5 (2003): 595-96. Berrios-Torres, S.I., et al. “World Trade Center Rescue Worker Injury and Illness Surveillance, New York, 2001.” American Journal of Preventive Medicine, 25, no. 2 (2003): 79-87. Boscarino, J.A., et al. “Mental Health Service and Medication Use in New York City after the September 11, 2001, Terrorist Attack.” Psychiatric Services, 55, no. 3 (2004): 274-83. Boscarino, J.A., et al. “Psychiatric Medication Use among Manhattan Residents Following the World Trade Center Disaster.” Journal of Traumatic Stress, 16, no. 3 (2003): 301-6. Centers for Disease Control and Prevention. “Injuries and Illnesses among New York City Fire Department Rescue Workers after Responding to the World Trade Center Attacks.” Special Issue, Morbidity and Mortality Weekly Report, 51 (2002a): 1-13. Centers for Disease Control and Prevention. “Psychological and Emotional Effects of the September 11 Attacks on the World Trade Center— Connecticut, New Jersey, and New York, 2001.” Morbidity and Mortality Weekly Report, 51, no. 35 (2002b): 784-85. Centers for Disease Control and Prevention. “Rapid Assessment of Injuries among Survivors of the Terrorist Attack on the World Trade Center—New York City, September 2001.” Morbidity and Mortality Weekly Report, 51, no. 1 (2002c): 1-5. Chi, J.S., et al. “Hospital Admissions for Cardiac Events in New York City after September 11, 2001.” American Journal of Cardiology, 92, no. 1 (2003): 61-63. Cushman, J.G., et al. “Two New York City Hospitals’ Surgical Response to the September 11, 2001, Terrorist Attack in New York City.” Journal of Trauma, 54, no. 1 (2003): 147-55. DeLisi, L.E., et al. “A Survey of New Yorkers after the Sept. 11, 2001, Terrorist Attacks.” American Journal of Psychiatry, 160, no. 4 (2003): 780-83. Deren, S., et al. “Impact of the September 11th Attacks in New York City on Drug Users: A Preliminary Assessment.” Journal of Urban Health: Bulletin of the New York Academy of Medicine, 79, no. 3 (2002): 409-12. Edelman, P., et al. “Biomonitoring of Chemical Exposure among New York City Firefighters Responding to the World Trade Center Fire and Collapse.” Environmental Health Perspectives, 111, no. 16 (2003): 1906-11. Fairbrother, G., et al. “Posttraumatic Stress Reactions in New York City Children after the September 11, 2001, Terrorist Attacks.” Ambulatory Pediatrics, 3, no. 6. (2003): 304-11. Feldman, D.M., et al. “Symptoms, Respirator Use, and Pulmonary Function Changes among New York City Firefighters Responding to the World Trade Center Disaster.” Chest, 125, no. 4 (2004): 1256-64. Galea, S., et al. “Posttraumatic Stress Disorder in Manhattan, New York City, after the September 11th Terrorist Attacks.” Journal of Urban Health: Bulletin of the New York Academy of Medicine, 79, no. 3 (2002a): 340-53. Galea, S., et al. “Psychological Sequelae of the September 11 Terrorist Attacks in New York City.” New England Journal of Medicine, 346, no. 13 (2002b): 982-87. Galea, S., et al. “Trends of Probable Post-Traumatic Stress Disorder in New York City after the September 11 Terrorist Attacks.” American Journal of Epidemiology, 158, no. 6 (2003): 514-24. Greater New York Hospital Association. GNYHA Survey on the Utilization of Hospital-Based Psychiatric Services Following the Attack on the World Trade Center: A Final Report. New York, 2001. Herbert, R., and S. Levin. World Trade Center Worker and Volunteer Medical Screening Program. Report of Initial Findings to the National Institute for Occupational Safety and Health of the Centers for Disease Control and Prevention. New York, 2003. Hoven, C.W., et al. “Mental Health of New York City Public School Children after 9/11: An Epidemiologic Investigation.” In S. Coates et al. (eds.), September 11: Trauma and Human Bonds. Hillsdale, NJ: Analytic Press, 2003. Landrigan, P.J., et al. “Health and Environmental Consequences of the World Trade Center Disaster.” Environmental Health Perspectives, 112, no. 6 (2004): 731-39. Levin, S., et al. “Health Effects of World Trade Center Site Workers.” American Journal of Industrial Medicine, 42, no. 6 (2002): 545-47. Levy, B.S., and V.W. Sidel (eds.). Terrorism and Public Health. New York: Oxford University Press, 2003. Lippy, B.E. “Safety and Health of Heavy Equipment Operators at Ground Zero.” American Journal of Industrial Medicine, 42, no. 6 (2002): 539-42. Malievskaya, E., et al. “Assessing the Health of Immigrant Workers near Ground Zero: Preliminary Results of the World Trade Center Day Laborer Medical Monitoring Project.” American Journal of Industrial Medicine, 42, no. 6 (2002): 548-49. Martyny, J., et al. “Respiratory Protection.” New England Journal of Medicine, 347, no. 11 (2002): 824-30. Niles, B.L., et al. “Posttraumatic Stress Disorder Symptomatology in Vietnam Veterans before and after September 11.” Journal of Nervous and Mental Disease, 191, no. 10 (2003): 682-84. Prezant, D.J., et al. “Cough and Bronchial Responsiveness in Firefighters at the World Trade Center Site.” New England Journal of Medicine, 347, no. 11 (2002): 806-15. Rom, W.N., et al. “Acute Eosinophilic Pneumonia in a New York City Firefighter Exposed to World Trade Center Dust.” American Journal of Respiratory and Critical Care Medicine, 166, no. 6 (2002): 797-800. Rosenheck, R.A., and A. Fontana. “Post-September 11 Admission Symptoms and Treatment Response among Veterans with Posttraumatic Stress Disorder.” Psychiatric Services, 54, no. 12 (2003): 1610-17. Salzman, S.H., et al. “Early Respiratory Abnormalities in Emergency Services Police Officers at the World Trade Center Site.” Journal of Occupational and Environmental Medicine, 46, no. 2 (2004): 113-22. Schlenger, W.E., et al. “Psychological Reactions to Terrorist Attacks: Findings from the National Study of Americans’ Reactions to September 11.” Journal of the American Medical Association, 288, no. 5 (2002): 581-88. Schuster, M.A., et al. “A National Survey of Stress Reactions after the September 11, 2001, Terrorist Attacks.” New England Journal of Medicine, 345, no. 20 (2001): 1507-12. Silver, R.C., et al. “Nationwide Longitudinal Study of Psychological Responses to September 11.” Journal of the American Medical Association, 288, no. 10 (2002): 1235-44. Simeon, D., et al. “Peritraumatic Reactions Associated with the World Trade Center Disaster.” American Journal of Psychiatry, 160, no. 9 (2003): 1702-5. Skloot, G., et al. “Respiratory Symptoms and Physiologic Assessment of Ironworkers at the World Trade Center Disaster Site.” Chest, 125, no. 4 (2004): 1248-55. Spadafora, R. “Firefighter Safety and Health Issues at the World Trade Center Site.” American Journal of Industrial Medicine, 42, no. 6 (2002): 532-38. Szema, A.M., et al. “Clinical Deterioration in Pediatric Asthmatic Patients after September 11, 2001.” Journal of Allergy and Clinical Immunology, 113, no. 3 (2004): 420-26. Thiel de Bocanegra, H., and E. Brickman. “Mental Health Impact of the World Trade Center Attacks on Displaced Chinese Workers.” Journal of Traumatic Stress, 17, no. 1 (2004): 55-62. Trout, D., et al. “Health Effects and Occupational Exposures among Office Workers near the World Trade Center Disaster Site.” Journal of Occupational and Environmental Medicine, 44, no. 7 (2002): 601-5. Vlahov, D., et al. “Increased Use of Cigarettes, Alcohol, and Marijuana among Manhattan, New York, Residents after the September 11th Terrorist Attacks.” American Journal of Epidemiology, 155, no. 11 (2002): 988-96. Vlahov, D., et al. “Consumption of Cigarettes, Alcohol, and Marijuana among New York City Residents Six Months after the September 11 Terrorist Attacks.” American Journal of Drug and Alcohol Abuse, 30, no. 2 (2004a): 385-407. Vlahov, D., et al. “Sustained Increased Consumption of Cigarettes, Alcohol, and Marijuana among Manhattan Residents after September 11, 2001.” American Journal of Public Health, 94, no. 2 (2004b): 253-4. Wallingford, K.M., and E.M. Snyder. “Occupational Exposures during the World Trade Center Disaster Response.” Toxicology and Industrial Health, 17, no. 5-10 (2001): 247-53. Warren, T., et al. “Factors Influencing Experienced Distress and Attitude toward Trauma by Emergency Medicine Practitioners.” Journal of Clinical Psychology in Medical Settings, 10, no. 4 (2003): 293-96. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

When the World Trade Center (WTC) buildings collapsed on September 11, 2001, nearly 3,000 people died and an estimated 250,000 to 400,000 people who were visiting, living, working, and attending school nearby, or responding to the attack, were exposed to a mixture of dust, debris, smoke, and various chemicals. In the months to follow, thousands of people who returned to the area to live and work, as well as responders who were involved in the search for remains and site cleanup, were also exposed. In addition, people in New York City and across the country were exposed to the emotional trauma of a terrorist attack on American soil. Concerns have been raised about the short- and long-term physical and mental health effects of the attack. Various government agencies and private organizations established efforts to monitor and understand these health effects. GAO was asked to describe the health effects that have been observed in the aftermath of the WTC attack and the efforts that are in place to monitor and understand those health effects. GAO searched bibliographic databases such as Medline to determine the pertinent scientific literature, reviewed that literature, and interviewed and reviewed documents from government officials, health professionals, and officials of labor groups. In the aftermath of the September 11 attack on the World Trade Center, a wide variety of physical and mental health effects have been reported in the scientific literature. The primary health effects include various injuries, respiratory conditions, and mental health effects. In the immediate aftermath of the attack, the primary injuries were inhalation and musculoskeletal injuries. During the 10-month cleanup period, despite the dangerous work site, responders reported few injuries that resulted in lost workdays. A range of respiratory conditions have also been reported, including wheezing, shortness of breath, sinusitis, asthma, and a new syndrome called WTC cough, which consists of persistent cough accompanied by severe respiratory symptoms. Almost all the firefighters who responded to the attack experienced respiratory effects, and hundreds had to end their firefighting careers due to WTC-related respiratory illness. Unlike the physical health effects, the mental health effects were not limited to people in the WTC area but were also experienced nationwide. Because most of the information about mental health effects comes from questionnaire or survey data, what is reported in most cases are symptoms associated with a psychiatric disorder, rather than a clinical diagnosis of disorder. The most commonly reported mental health effects include symptoms associated with depression, stress, anxiety, and posttraumatic stress disorder (PTSD)--a disorder that can develop after experiencing or witnessing a traumatic event and includes such symptoms as intrusive memories and distressing dreams--as well as behavioral effects such as increased use of alcohol and tobacco and difficulty coping with daily responsibilities. Six programs were established to monitor and understand the health effects of the attack, and these programs vary in terms of which people are eligible to participate, methods for collecting information about the health effects, options for treatment referral, and number of years people will be monitored. Although five of the programs focus on various responder populations, the largest program--the WTC Health Registry--is open not only to responders but also to people living or attending school in the vicinity of the WTC site, or working or present in the vicinity on September 11. The monitoring programs vary in their methods for identifying those who may require treatment, and although none of these programs are funded to provide treatment, they provide varying options for treatment referral. Under current plans, HHS funding for the programs will not extend beyond 2009. Some long-term health effects, such as lung cancer, may not appear until several decades after a person has been exposed to a harmful agent. GAO provided a draft of this testimony to DHS, EPA, HHS, and the Department of Labor. In its written comments, HHS noted that the testimony does not include significant discussion of ways in which mental health symptoms have changed over time. The evidence GAO examined did not support a full discussion of changes in mental or physical health effects over time.

Each year, CMS evaluates approximately 3,000 acute care hospitals participating in HVBP on their performance in prior years on a series of quality and efficiency measures. Prior to the HVBP program, hospitals received slightly higher Medicare payments for submitting data on measures within CMS’s public Inpatient Quality Reporting (IQR). Beginning in fiscal year 2013, the HVBP program provided new bonuses and penalties that were based on each hospital’s performance on a subset of these measures. Each individual hospital’s performance is calculated for each measure within a domain using a baseline period and a performance period, both of which are in prior years. For each of the HVBP measures, CMS considers both the results of a hospital’s absolute performance— awarding achievement points if performance on a measure was at or above the median for all participating hospitals—and improvements in its performance over time—awarding improvement points if current performance had improved. CMS uses the higher of these points as the hospital’s score on each measure. Related measures are grouped into specific performance categories, called domains. The domain scores are weighted to develop a total performance score for each hospital. The measures that constitute each domain, the number of domains, and the weighting of the domain scores have changed over the years of the program (see table 1). In fiscal year 2013, HVBP had two quality domains—clinical processes and patient experience; by 2017, two additional quality domains—patient outcomes and safety—and one efficiency domain were added to the program. By law, the HVBP program is budget neutral, which means that the total amount of payment increases, or bonuses, awarded to hospitals deemed to provide higher quality of care must equal the total amount of payment reductions, or penalties, applied to hospitals deemed to provide lower quality of care. To fund the HVBP program, CMS first applies an initial fixed percentage reduction to the amount of each hospital’s Medicare reimbursements for its patients that fiscal year. The initial percentage reduction was 1 percent in fiscal year 2013 and has grown by 0.25 percent each year to the maximum of 2 percent for fiscal year 2017 and beyond, as specified in PPACA. CMS determines each hospital’s payment adjustment based on the hospital’s total performance score relative to all participating hospitals. Hospitals with payment adjustments that exceed the initial reduction receive a net increase, or bonus. Hospitals with a payment adjustment less than the initial reduction have a net decrease, or a penalty. (For two hypothetical examples using the initial percentage reduction for fiscal year 2017, see fig. 1.) These payment adjustments are applied to the inpatient Medicare payment for each discharged patient throughout the upcoming fiscal year. In October 2015, we reported on certain HVBP performance measures prior to and after the implementation of the HVBP program. We found that trends in performance for many of these measures were unchanged since the implementation of the HVBP program. This report included information from interviews with officials from selected hospitals who noted that the HVBP program reinforced ongoing quality improvement efforts but did not lead to major changes in focus. Hospital officials also indicated that there were patient population and community barriers to their quality improvement efforts. In a related report on the HVBP program, HHS noted challenges that rural hospitals face that affect their performance on quality measures and the reliability of their outcome measurements, including lower occupancy rates, higher percentages of uncompensated care, and lower operating margins than urban hospitals. Safety net hospitals generally had lower median quality domain scores in comparison to all hospitals, while small rural and small urban hospitals generally scored higher on quality and efficiency domains during fiscal years 2013 through 2017. Median scores for each of the separate quality domains—clinical processes, patient experience, patient outcomes, and safety—were consistently lower for safety net hospitals and were generally higher for small rural and small urban hospitals than for hospitals overall during fiscal years 2013 through 2017. Specifically, for the four quality domains, we found the following: Clinical processes: The clinical processes median domain scores— which summarize measures for preventive or routine care—were lower for safety net hospitals and generally higher for small urban hospitals than for all hospitals during fiscal years 2013 through 2017. Median clinical processes scores for small rural hospitals were generally lower—between 4 and 9 percent—than for hospitals overall in fiscal years 2013 through 2015 (see fig. 2). Patient experience: Small hospitals consistently had higher patient experience scores—which consist of measures for communication and responsiveness—than hospitals overall, while safety net hospitals had the lowest scores of any of the hospital types (see fig. 3). Patient outcomes: Median scores for the patient outcomes domain— which comprises measures for mortality rates and other results and was added in fiscal year 2014—were generally lowest for small rural hospitals in each year of our analysis, except for fiscal year 2016, when compared to hospitals overall (see fig. 4). Safety net hospitals and small urban hospitals—with the exception of fiscal year 2016— also did not perform as well as all hospitals in the years of our analysis. Safety: Safety scores—which were added in fiscal year 2017 and include measures for infection rates and other complications—were lowest for safety net hospitals and higher for small rural and small urban hospitals than the median scores for hospitals overall. The median score for the safety net hospitals was about 11 percent lower than the median score for all hospitals. Small rural hospitals had the highest median score and small urban hospitals also had a higher median score than hospitals overall. However, 21 percent of all hospitals were missing scores for this new domain in fiscal year 2017. Trends for the efficiency domain, which contains the single cost measure—Medicare spending per beneficiary—were similar to the quality domains in that small hospitals tended to perform better than safety net hospitals and better than hospitals overall from fiscal year 2015, when the domain was added, through fiscal year 2017 (see fig. 5). Safety net hospitals have had the same median efficiency scores as for hospitals overall during the 3 years it has been included in the program. However, over 40 percent of all hospitals had an efficiency score of 0 during these years due to CMS’s methodology for calculating scores. This methodology resulted in a low median score of 10 for all hospitals, though many hospitals had considerably higher efficiency scores. Hospitals’ total performance scores were consistent with the trends in the quality and efficiency domain scores. Specifically, when compared to all hospitals, total performance scores were lowest for safety net hospitals and generally highest for small urban hospitals during fiscal years 2013 through 2017 (see fig. 6). Median payment adjustments generally have varied for all hospitals, and small rural and small urban hospitals, since the program began; however, in most years, the median payment adjustment for safety net hospitals has been a penalty—that is, a negative payment adjustment. In contrast, the small hospitals, as well as hospitals overall, generally had positive payment adjustments, indicating a bonus, with the exception of fiscal year 2014. Small urban hospitals consistently received higher payment adjustments than all hospitals—between 0.03 and 0.36 percentage points higher—every fiscal year. (See table 2.) The majority of all hospitals received a bonus or a penalty of less than 0.5 percent each year of the program (see fig. 7). However, over time, an increasing percentage of hospitals received bonuses of more than 0.5 percent, and by fiscal year 2016, more than one-quarter of all participating hospitals received a bonus of more than 0.5 percent. Compared to all hospitals, a higher percentage of small rural and small urban hospitals received bonuses of more than 0.5 percent, and this disparity has grown as the program continues. An increasing percentage of hospitals have also received penalties of greater than 0.5 percent over time, and safety net hospitals consistently had the highest percentage of penalties of 0.5 percent or more when compared to all hospitals, small rural hospitals, and small urban hospitals. In part, the size of the bonuses and penalties, in dollar terms, has been increasing due to the increase in the initial reduction from 1 percent in fiscal year 2013 to 2 percent in fiscal year 2017 (see table 3). In addition, as more hospitals receive bonuses in excess of 0.5 percent, the difference between the bonuses and penalties has been increasing. For example, in fiscal year 2013, the median bonus and penalty for all hospitals was nearly identical. Over the years, the median bonus has more than doubled, but the median penalty has nearly tripled. For most hospitals, the annual bonus or penalty is less than $100,000, and by the end of the fiscal year 2017, over $690 million will have been redistributed from hospitals that received penalties to hospitals that received bonuses. Safety net hospitals received a smaller percentage of the bonuses and paid a greater share of the penalties than small rural and small urban hospitals. For example, safety net hospitals have received about 5 percent of the bonus dollars and paid approximately 10 percent of the penalty dollars each year. In contrast, small rural and urban hospitals have received an average of about 9 and 12 percent of the bonus dollars, respectively, and both groups of these small hospitals paid about 5 percent or less of the penalties dollars during fiscal years 2013 through 2017. Since the efficiency score was added to the HVBP program in fiscal year 2015, about 20 percent of the hospitals that received bonuses each year had weighted composite quality scores below the median for all hospitals in fiscal years 2015 through 2017 (see table 4). For each fiscal year, a higher percentage of safety net and small rural hospitals received bonuses (between 26 and 36 percent) when compared to all hospitals, despite having quality scores below the median score for all hospitals. The median payment adjustments for the hospitals that received a bonus with lower quality scores were less than median bonuses overall. For example, in fiscal year 2015, the median bonus for all hospitals was 0.32 percent, and the median bonus for the hospitals that received a bonus with composite quality scores below the median was 0.17 percent. Hospitals that received a bonus despite having composite quality scores below the median for all hospitals had sufficiently high efficiency scores to achieve total performance scores that made them eligible for bonuses. Across all hospital types and years, the median efficiency scores for these hospitals ranged from 1.50 and 6.00 times higher than the median efficiency scores for hospitals overall. For example, in fiscal year 2017, the overall median efficiency score for small rural hospitals was 30.00. In contrast, the median efficiency score for small rural hospitals that received a bonus with a composite quality score below the all-hospital median was more than twice as high at 70.00. Table 5 compares two actual hospitals—both of which received a bonus—with similar total performance scores but different composite quality scores. Hospital A outperformed Hospital B in every quality domain except safety and received a composite quality score of 40.00, well above the median of 29.03. While both hospitals had an efficiency score above the median of 10.00, Hospital B’s high efficiency score results in a total performance score above that of the higher quality Hospital A. According to CMS documentation, the agency developed the weighting formula to ensure that the Medicare spending per beneficiary measure— the sole measure in the efficiency domain—would make up only a portion of the total performance score and that the remainder would be based on hospitals’ performance on the other measures. The same documentation stated that the distinct measure of cost, independent of quality, would enable the agency to identify—and subsequently reward through payment adjustments—hospitals involved in the provision of high- quality care at a lower cost to Medicare. However, CMS’s formula for weighting the domain scores to determine a total performance score has created a system that, in some cases, rewards lower quality hospitals that provide care at a lower cost. In a November 2016 report to Congress, CMS indicated that it was aware of reports that the added efficiency metric resulted in some lower quality hospitals receiving bonus HVBP payments in 2015. However, in the report CMS reiterated that its scoring methodology—the weighting of quality domains at 75 percent and the efficiency domain at 25—provided balanced consideration for quality and efficiency and would ensure that high-quality hospitals were being rewarded. Our work shows that CMS has not achieved this balanced consideration as it intended, thereby rewarding some lower quality hospitals due to their high efficiency scores. CMS did not require a complete set of domain scores to participate in the HVBP program after 2015, but instead proportionately redistributed the missing scores’ domain weights to the other domains, including efficiency. As a result, the efficiency score can carry even more than its assigned weight, and hospitals with missing domain scores had efficiency scores that were weighted higher than those of the other participating hospitals. This amplified the contribution of the efficiency domain to hospitals’ total performance scores. The assigned weight for the efficiency score was 20 percent in fiscal year 2015 and 25 percent in fiscal years 2016 and 2017. However, due to the proportional redistribution, a hospital’s efficiency score could be weighted between 25 and 50 percent—rather than the original 20 percent—in fiscal year 2015 and between 26 and 71 percent—rather than the original 25 percent—in fiscal years 2016 and 2017, depending on how many and which domains were missing. Table 6 illustrates the impact of redistributed domain weights on hospitals in fiscal year 2017. Hospital A, the same hospital noted in table 5, is considered a higher quality hospital, with a composite quality score well above the median of 29.03 for all hospitals in 2017. Three other actual hospitals—hospitals C through E—show how the proportional redistribution of weights can dramatically increase the effect that a hospital’s efficiency score can have on its total performance score. Hospital C is missing two domains, together worth 45 percent of the total performance score. The 45 percent is then proportionally redistributed to the other domains so that the clinical processes domain weight increases from 5.00 percent to 9.10 percent and the weights of the patient experience and efficiency domains each increase from 25 percent to 45.45 percent. We also found that hospitals with missing domain scores were more likely to receive a bonus than hospitals with all domain scores. Specifically, in fiscal year 2017, 68 percent of hospitals with missing domain scores received a bonus, compared to 50 percent of hospitals with all domain scores. Of the approximately 20 percent of hospitals that received a bonus with a quality score below the median described earlier, many were also missing domain scores. For example, in fiscal year 2017, 182 of the 345 lower quality hospitals that received a bonus (53 percent) were missing at least one quality domain score. Hospitals with missing domain scores had bonuses that grew to exceed the median bonus payment adjustment for all hospitals. In fiscal 2015, the median bonus adjustment for all hospitals was 0.32 percent. For lower quality hospitals with missing domain scores, the median bonus adjustment that year was slightly lower at 0.31 percent. However, by fiscal year 2017, lower quality hospitals with missing domain scores that received bonuses had a bonus adjustment of 0.74 percent, considerably higher than the median bonus adjustment of 0.54 percent for hospitals overall. CMS decided to proportionally redistribute missing domain scores in order to maintain the relative weights of each remaining domain and reliably score hospitals on their performance. However, the issues we identified with the weighting formula—in that it results in some lower quality hospitals receiving bonuses—are exacerbated for hospitals with missing domain scores. As a result, hospitals with missing domain scores are more likely to get a bonus, and, in some cases, those bonuses are greater than median bonuses overall. Additionally, while CMS intended to keep the efficiency metric independent of quality, the effective weight of the efficiency measure depends on the extent to which hospitals report quality measures. As a result, the balance the agency tried to achieve in the total performance score—allocating 75 percent of the score to the quality domains and 25 percent of the score to the efficiency domain—is no longer achieved. The aim of the HVBP program is to improve hospital quality and efficiency by providing incentives for hospitals to improve their quality of care and to become more cost efficient. Throughout the 5 years of the program, CMS has made modifications to meet these goals by changing quality performance domains and domain weighting from year to year. With the addition of the efficiency domain in fiscal year 2015, CMS signaled the importance of hospitals’ providing care at a lower cost to Medicare, and, in its weighting formula, the agency tried to find balanced consideration for quality and cost. Rather than achieving this balance—which would have allowed the agency to identify and reward higher quality and lower cost hospitals—CMS’s weighting formula has resulted in bonuses for some lower quality hospitals, solely due to their cost efficiency. Because the program is budget neutral, bonuses for lower quality hospitals may result in smaller bonuses for hospitals that are performing well across all domains. The issue is especially stark for between 10 and 25 percent of the hospitals that were missing domain scores in fiscal years 2015 through 2017, which has also contributed to the awarding of bonuses to lower quality hospitals. If CMS continues to use the current formula, it will continue to reward hospitals that do not score well on quality and efficiency metrics. To ensure that the HVBP program accomplishes its goal to balance quality and efficiency and to ensure that it minimizes the payment of bonuses to hospitals with lower quality scores, we recommend that the Administrator of CMS take the following two actions: Revise the formula for the calculation of hospitals’ total performance score or take other actions so that the efficiency score does not have a disproportionate effect on the total performance score. Revise the practice of proportional redistribution used to correct for missing domain scores so that it no longer facilitates the awarding of bonuses to hospitals with lower quality scores. We provided a draft of this report to HHS for comment, and its written comments are reprinted in appendix III. The department indicated that it would examine the formula used for calculating hospitals’ total performance scores and would explore alternatives to the practice of proportional redistribution. While HHS stated it would consider revisions to these practices, it indicated that any changes to the weights of the domains, or the distribution of weights for missing domains, would be evaluated for potential negative impacts and would be subject to notice and comment rulemaking. HHS also provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of Health and Human Services, the CMS Administrator, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-7114 or cosgrovej@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Table 7 lists the Inpatient Quality Reporting program measures that the Centers for Medicare & Medicaid Services (CMS) used to analyze hospitals’ performance in the Hospital Value-based Purchasing program during fiscal years 2013 through 2017. This table identifies the domain associated with each measure, which measures were used to calculate domain scores each year, the measure code, and a description of each measure. In addition to the contact named above, Martin T. Gahart (Assistant Director), Erin C. Henderson (Analyst-in-Charge), Zhi Boon, Kye Briesath, and Elizabeth Morrison made key contributions to this report. Also contributing were Muriel Brown and Jacquelyn Hamilton. Medicare Value-based Payment Models: Participation Challenges and Available Assistance for Small and Rural Practices. GAO-17-55. Washington, D.C.: December 9, 2016. Health Care Quality: HHS Should Set Priorities and Comprehensively Plan Its Efforts to Better Align Health Quality Measures. GAO-17-5. Washington, D.C.: October 13, 2016. Patient Safety: Hospitals Face Challenges Implementing Evidence-based Practices. GAO-16-308. Washington, D.C.: February 25, 2016. Hospital Value-based Purchasing: Initial Results Show Modest Effects on Medicare Payments and No Apparent Change in Quality-of-Care Trends. GAO-16-9. Washington, D.C.: October 1, 2015. Health Care Transparency: Actions Needed to Improve Cost and Quality Information for Consumers. GAO-15-11. Washington, D.C.: October 20, 2014. Electronic Health Record Programs: Participation Has Increased, but Action Needed to Achieve Goals, Including Improved Quality of Care. GAO-14-207. Washington, D.C.: March 6, 2014. Health Care Quality Measurement: HHS Should Address Contractor Performance and Plan for Needed Measures. GAO-12-136. Washington, D.C.: January 13, 2012. Hospital Quality Data: Issues and Challenges Related to How Hospitals Submit Data and How CMS Ensures Data Reliability. GAO-08-555T. Washington, D.C.: March 6, 2008.

The HVBP program, enacted as part of the Patient Protection and Affordable Care Act (PPACA), evaluates hospital performance on quality and efficiency (Medicare spending per beneficiary) measures. Based on those results, CMS adjusts Medicare payments, leading to bonuses or penalties for hospitals. The first HVBP payment adjustments started in fiscal year 2013. PPACA included a provision for GAO to assess the HVBP program's impact on Medicare quality and efficiency, including the effects on safety net, small rural, and small urban hospitals. This report addresses (1) hospitals' performance in quality and efficiency categories; (2) how hospitals' payment adjustments have changed over time; and (3) the effect, if any, of efficiency scores on payment adjustments. GAO analyzed CMS documentation and data on performance scores and payment adjustments in each year for all hospitals participating in fiscal years 2013 through 2017. GAO also analyzed results for safety net, small rural, and small urban hospitals and interviewed CMS officials. The Hospital Value-based Purchasing (HVBP) program aims to improve quality of care and efficiency by creating financial incentives for about 3,000 participating hospitals. From fiscal years 2013 through 2017, performance on quality and efficiency measures varied by hospital type. Safety net hospitals—those that serve a high proportion of low-income patients—generally scored lower in quality compared to all participating hospitals. In contrast, small rural and small urban hospitals—those with 100 or fewer acute care beds—scored higher on efficiency compared to all hospitals. Payment adjustments—bonuses or penalties, announced prior to each fiscal year—have varied over time for all hospitals. In four out of the five years of GAO's analysis, small rural and small urban hospitals were more likely to receive a bonus compared to all participating hospitals, while safety net hospitals were more likely to receive a penalty. While a majority of all hospitals received a bonus or a penalty of less than 0.5 percent each year, the percentage of hospitals receiving a bonus greater than 0.5 percent increased from 4 percent to 29 percent from fiscal year 2013 to 2017. In dollar terms, most hospitals had a bonus or penalty of less than $100,000 in fiscal year 2017. Some hospitals with high efficiency scores received bonuses, despite having relatively low quality scores, which contradicts the Centers for Medicare & Medicaid Service's (CMS) stated intention to reward hospitals providing high-quality care at a lower cost. Further, among hospitals that were missing one or more quality scores, the efficiency score had a greater effect on the total performance score because of the methodology used by CMS. This methodology compensated for the missing scores by increasing the weights of all of the non-missing scores. Consequently, hospitals with missing scores were more likely to receive bonuses than hospitals with complete scores. Bonus or Penalty Status of Hospitals Participating in the Hospital Value-based Purchasing Program, Fiscal Years 2015 through 2017 So that lower quality hospitals do not receive bonuses, GAO recommends that CMS revise (1) the methodology used to calculate total performance scores and (2) its method of accounting for missing quality scores. In its written comments, HHS indicated that it would consider revising these two methodologies.

“Every telecommunication user suffers from telabuse. The only question is how much each loses each month.” “There are two kinds of customers: those who have been victims of toll fraud, and those who will be.” The abuse and theft of telecommunications services is one of the fastest growing crimes in the United States. According to Telecommunications Advisors, Incorporated (TAI), a consulting firm that has done extensive research on telephone fraud and abuse problems, these crimes cost industry and government an estimated $9 billion each year. Telephone abuse or “telabuse” is the misuse or waste of telephone resources by employees from within an organization or by their relatives or acquaintances. This typically involves personal long-distance calls made by employees at an organization’s expense. Toll fraud is the theft of an organization’s long-distance services by individuals from the outside. This can involve fraud committed by experienced telephone hackers who are able to penetrate an organization’s voice message systems and private networks. It also involves the fraudulent use of telephone company calling cards and cellular telephones by hackers or others who steal calling card numbers and cellular telephone services. A security manager for one major telephone company attributed a large part of the ever-increasing levels of toll fraud and telabuse to user neglect and inattention. According to the security manager, the attention and consideration industry and government organizations give to telephone equipment and services often stop after the initial purchase, leaving them more vulnerable to the risk of telephone abuse and fraud. Furthermore, it is commonplace, in industry and government, that bills for telecommunications services are often not reviewed to ensure that charges incurred are appropriate and justified. For most organizations, controls over telephone services, as expected, are secondary to many of the more pressing daily business functions. According to TAI, organizations often treat their telephone bills like other bills, such as rent and electric, and simply pay them without examination. Cases reported by TAI have also shown that failure to establish adequate controls over the use of telecommunications resources can have costly consequences. For example, one toll fraud incident at the Drug Enforcement Administration (DEA) reportedly cost the government over $2 million because DEA, which did not monitor telephone activity and review billing records, did not detect fraudulent calls by telephone hackers for 18 months. In another case, not adequately monitoring telephone calling activity at New York City’s Human Resources Administration offices cost the city over $200,000 for thousands of employee calls to party lines and other personal services over several years. These problems are not confined to government. For example, a private company in Texas was billed $25,000 in 1 month for improper calls and, by instituting minor controls over employee use of telephones, a major utility company in the southeast was able to reduce its telephone bill by over $60,000 per year. USDA and its 29 component agencies spend over $100 million on telecommunications annually, including more than $50 million for commercial telecommunications services obtained from over 1,500 telephone companies. These companies provide local telephone service as well as international and domestic long-distance services that are not available on the FTS 2000 network. USDA headquarters offices and other USDA agency and staff offices within the Washington, D.C., metropolitan area pay for over 24,000 separate telephone lines each month. The Federal Information Resources Management Regulation governs use of telecommunications services for all government agencies and states that telephone calls paid for by the government shall be used to conduct official business only. Unauthorized calls, which are calls that are not necessary in the interest of the government, are prohibited, and agencies are required to collect for any unauthorized calls if it is cost-effective to do so. USDA’s telecommunications policy (DR-3300-1) requires that USDA agencies ensure that government-provided telephones are used only for official business and for calls the agency considers necessary. Under DR-3300-1, the Office of Information Resources Management (OIRM) is responsible for establishing policy and procedures for the management and cost control of telecommunication systems and each component agency and staff office is responsible for ensuring compliance with departmental policy and that government telephones are used for authorized purposes only in accordance with this policy. USDA’s Chief Financial Officer (CFO) is responsible for overseeing all financial management activities relating to the programs and operations of the Department, including managing USDA’s National Finance Center (NFC). To assess USDA’s controls over telephone use, we examined USDA’s policies and procedures governing the use of government telephones. We also obtained and reviewed commercial telephone billing records for USDA agency offices in the Washington, D.C., metropolitan area for 4 months in fiscal year 1995, which totaled about $580,000 or 1 percent of the $50 million USDA spends annually for commercial telecommunications costs. The 4 months were selected from early, mid, and late parts of the fiscal year to adjust for any seasonal variations in calling patterns, and we reviewed billing records of all collect calls accepted by the Department during the 4 months as well as selected long-distance calls made during the month of August 1995. In cases where billing records disclosed instances of telephone abuse, we discussed these cases with USDA officials and telephone company representatives and provided billing records of the calls to USDA officials for appropriate action. We also discussed cases involving collect calls from prisons with correctional facility personnel and USDA officials. In addition, we reviewed telephone company records and USDA documentation pertaining to a March 1995 hacker case at the Department and examined USDA actions taken in response to this attack. Appendix I provides further details on our scope and methodology. We conducted our review from October 1995 through February 1996 in accordance with generally accepted government auditing standards. We discussed the facts in our report with USDA officials, including the Assistant Secretary and the Deputy Assistant Secretary for Administration, the acting Chief Financial Officer, the Director of USDA’s Office of Information Resources Management, and the Assistant Inspector General for Investigations and have incorporated their comments where appropriate. We also provided a draft of the report to USDA for comment. USDA’s comments are discussed in the report and are included in full in appendix II. Our review of four monthly telephone bills for USDA agencies and offices in the Washington, D.C., metropolitan area found that 652 collect calls, or about 50 percent of all collect calls accepted and paid for by USDA during this 4-month period, were from individuals at 18 correctional institutions. In these cases, USDA paid about $2,600 for collect calls accepted from correctional centers in addition to unknown charges for subsequent calls placed by USDA on behalf of individuals at these centers. Because these subsequent calls cannot be easily differentiated from other calls on telephone bills, it is difficult to determine the extent to which this occurred and the total charges that resulted from all collect calls. Additionally, our review of just a few calls from the thousands of long-distance calls made by USDA agencies and offices in the Washington, D.C., metropolitan area each month found several other cases of telephone abuse involving personal long-distance calls outside the country to adult entertainment services and companies advertising jobs. USDA has been aware of cases of collect calling abuse since at least 1994, but has not taken adequate action to stop it. Although USDA policy does not specifically address collect calls placed from a nongovernment number to a government number, it states that USDA should ensure that government telephones are used only for authorized purposes. However, as discussed later in this report, USDA generally does not review its telephone bills to make such determinations. Individuals in at least 20 different USDA agencies or offices in the Washington, D.C., metropolitan area have accepted, at USDA’s expense, collect calls from individuals at federal, state, and county correctional institutions. This problem is exacerbated because individuals who accept these collect calls can use USDA telephones to place long-distance calls for the callers and transfer them to these calls. However, it is difficult to determine to what extent this has occurred or the total cost involved because charges for these additional calls cannot be easily identified. According to telephone company representatives, charges for these long-distance calls may appear on any one of many separate carriers’ bills and, because the termination point of the call is unknown, it is difficult to identify these calls on bills. As discussed later, cases of telephone abuse in 1994 investigated by the Office of the Inspector General (OIG) found that collect calls from inmates at correctional centers were transferred to other calls. Table 1 shows the number of collect calls made from correctional centers to USDA agencies and offices located in the Washington, D.C., metropolitan area. However, this may not represent all the collect calls made to USDA from correctional centers during this 4-month period because USDA could not provide us with complete billing records for these periods. An inmate uses a telephone in a correctional facility to place a collect call to a government agency or private company office telephone number. Upon answering the telephone, an individual at that agency or office hears a recorded message giving the inmate’s name and the name of the correctional facility. The individual is then asked whether he/she will accept the charge for the collect call. An individual, who is cooperating with the inmate, will accept the unauthorized call. In many cases, after accepting the call, the cooperating individual will in turn make other long-distance calls for the inmate, which are also charged to the agency or office. The facility operations manager stated that he is often contacted by individuals, government agencies, and private companies, who detect abuse on their telephones and arrange to have certain telephone numbers blocked. Blocking the numbers prevents inmates from placing calls to these telephones. The operations manager added that, with large organizations such as USDA, this may not be a viable solution to the problem because there are often many different agency telephones involved. Although the operations manager told us he has never been contacted by USDA about any inmate collect calls, he stated that at least four other federal agencies over the past 6 years have contacted him about this problem. According to the operations manager, agencies have had success stopping some abuse by blocking agency telephone numbers and taking punitive action against employees responsible for this abuse. While we were able to identify the cost of the collect calls (as shown in the table), charges for calls that are transferred were not identified. However, because many of these collect calls could have been transferred to long-distance lines, thousands more could have been added to USDA’s telephone bills. In the past, USDA identified abuses involving collect calls from inmates similar to the cases we found. However, the Department did not take adequate action to stop the problem. According to USDA documentation, in August 1994, an OIRM telecommunications specialist uncovered cases of telabuse at the Department dating back to 1993, which involved collect calls from inmates at the Federal Corrections Center in Lorton, Virginia. This individual found these cases while examining monthly telephone charges on commercial carrier telephone bills. He told us he had made a special request for the billing records to review telephone charges for a contractor working on-site at USDA headquarters offices. Generally, as we discuss later in this report, USDA officials do not review commercial carrier telephone bills. OIRM referred the matter to the Department’s OIG in August 1994. The OIG conducted a preliminary inquiry and determined from billing records prior to December 1994 that employees working in several USDA agency offices in the Washington, D.C., metropolitan area were improperly accepting collect calls placed from six correctional institutions. These institutions are located in Lorton and Oakwood, Virginia; Waldorf and La Plata, Maryland; and New Bern and Bayboro, North Carolina. The OIG also found that, after accepting collect calls from inmates, USDA employees made other unauthorized long-distance calls for the inmates and transferred the inmates to those calls. Costs for these calls were also charged to the Department. According to the OIG, it determined that individuals in multiple USDA agencies and offices had accepted more than $4,500 in collect calls from inmates. In one case, the OIG identified a contractor employee who had been accepting collect calls from a correctional facility while working in the OIG’s Washington, D.C., office. This individual, who had left USDA at the time of the inquiry, reimbursed the Department $177 for these unauthorized collect calls. The OIG referred all the remaining open cases to OIRM for action. Specifically, in a May 1995 letter to the Director of the OIRM’s Washington Service Center, the Deputy Assistant Inspector General for Investigations turned these cases over to OIRM for “handling and further distribution, as this type of misconduct matter is appropriately handled by the personnel investigators within each of the affected agencies.” However, the Director could not explain why no further action on these specific cases was taken. The OIG also tried to have some of the collect calls it identified from Lorton blocked, but USDA records indicate that collect calls continued because the carrier did not keep these blocks in place. As a result, the carrier agreed to reimburse USDA for collect calls from Lorton identified during the period investigated by the OIG. USDA received credit for some collect calls from Lorton. However, at the time of our review, USDA had not followed up to determine whether the Department received reimbursement for all the collect calls from Lorton because no one had reviewed the bills to match records of the calls with the credits being given. Moreover, OIRM took no action to determine whether there were other collect calling abuses at the Department. Consequently, as shown by our review, collect calls from correctional centers to the Department have persisted. In fact, our review found at least seven cases where the same office telephones identified by the OIG in 1994 were still being used to accept collect calls. In reviewing USDA’s Washington, D.C., telephone bills, we noted that agencies and offices spend over $30,000 per month for long-distance calls. We selected a few records from August 1995 bills for detailed examination of long-distance calls and identified several cases where unauthorized calls were made to the Dominican Republic. Some of these calls involved connections to adult entertainment lines, such as sex and party lines, and to companies advertising jobs. In one case, for example, USDA paid over $33 for four calls made from one office to a party line “chat” service in the Dominican Republic. In another case, the Department also paid for international calls made from several agency offices to job advertisement lines where home-based business and other employment opportunities are discussed. In addition, we found one case where a sex entertainment line in the Dominican Republic was called at USDA’s expense. In large part, these problems exist at USDA because bills for the tens of millions of dollars in commercial telephone services paid annually by USDA are generally not reviewed to monitor calling activity. USDA pays over 23,000 bills each month for commercial telephone services obtained from over 1,500 private vendors across the country. This includes the bills from telephone companies that provide commercial telephone and long-distance services to USDA offices in the Washington, D.C., metropolitan area. Vendors send these bills directly to USDA’s NFC in New Orleans, Louisiana, where they are processed and paid. USDA policy requires agencies and staff offices to ensure that government telephones are used for authorized purposes only. Specifically, the policy states that the use of government telephones shall be limited to the conduct of official business and other authorized uses, which can also include such things as making a brief daily telephone call to a spouse and children within a local commuting area. However, as we recently reported, agency managers rarely review telephone bills. Consequently, agency managers lack the information they need to determine whether telephones and long-distance services are used properly in accordance with departmental policy. In our prior reports, we also found that USDA wasted tens of thousands of dollars because it had not established adequate procedures for reviewing bills to verify the appropriateness of telephone charges by private vendors. To help ensure that controls are appropriate and cost-effective, agencies need to consider the extent and cost of controls relative to the importance and risk associated with a given program. Because USDA rarely reviews its telephone bills, we reported in September 1995 that the Department had paid tens of thousands of dollars each year to lease telephone equipment, such as rotary telephones and outdated modems that were either no longer used or could not be located. In addition, USDA wasted thousands more paying for telephone services for field offices that had been closed more than a year. USDA has begun to take positive steps in response to our previous reports to improve controls over payments for commercial telephone services. Specifically, USDA stopped payments for leased telecommunications equipment it no longer uses and is seeking reimbursement from carriers for overcharges. In addition, in October 1995, USDA formed a task force to investigate and develop action plans to correct telecommunications management deficiencies at the Department. In December 1995, the task force agreed with GAO’s findings and reported that “the process of planning, acquiring, ordering, billing, invoicing, inventory control, payments, and management of telecommunications services and equipment is chaotic at best and totally out of control at the very least.” Therefore, the task force recommended that USDA’s current paper-based billing system be reengineered and subsequently automated so that billing data can be cost-effectively verified. On March 1, 1996, USDA’s acting Chief Financial Officer told us that the Department agreed to implement the task force’s recommendations which the Department estimates will take about 2 years to complete. While this action is encouraging, USDA has not specifically responded to our September 22, 1995, report recommending, among other things, that the Secretary of Agriculture report the Department’s management of telecommunications as a material internal control weakness under the Federal Managers’ Financial Integrity Act (FMFIA) and that this weakness remain outstanding until USDA fully complies with federal regulations for managing telecommunications and institutes effective management controls. USDA’s fiscal year 1995 FMFIA report did not identify the Department’s management of telecommunications as a material internal control weakness, and on March 1, 1996, USDA’s acting Chief Financial Officer and Assistant Secretary for Administration told us that the Department had not determined whether to report telecommunications management as an material internal control weakness for fiscal year 1996. Since USDA has not yet established adequate and cost-effective controls for ensuring that its telephones are used properly, it is putting itself at continuing risk of telephone abuse and fraud. Moreover, because USDA does not know how widespread telephone abuse is at the Department or the total cost, it is not in a position to develop a plan defining cost-effective controls to mitigate the risk of telephone abuse and fraud or take appropriate action to address abuses that have occurred. We also found indications that USDA is vulnerable to other types of telephone fraud, waste, and abuse because bills are not reviewed. Billing records show that USDA agencies and offices in the Washington, D.C., metropolitan area pay tens of thousands of dollars each month for international calls. However, because these bills are generally not reviewed, USDA does not know whether these calls are authorized and it cannot detect instances where telephone fraud and abuse may have occurred. USDA is at risk of further waste and abuse by employees who use telephone company credit cards, instead of FTS 2000 Federal Calling Cards, to charge thousands of dollars in long-distance calls each month which are paid by USDA. These cards, which have been issued to USDA offices by commercial carriers, are not approved for use by the Department. USDA’s telecommunications policy DR 3100-1 states that the only telephone credit card approved for use by USDA employees is the FTS 2000 Federal Calling Card. Even though this policy has been in place over 2 years, some USDA employees have continued to use telephone company credit cards to charge their long-distance calls. Consequently, employees may be using these cards to charge long-distance calls at commercial rates, which are, according to USDA, as much as three times higher than FTS 2000 rates. Moreover, USDA does not know whether calls charged to telephone company credit cards are authorized because, like other commercial telephone bills, credit card bills are generally paid by the Department without being reviewed. Also, USDA does not know whether there have been any cases of telephone fraud involving telephone company credit cards by individuals outside USDA because the Department has no inventory of these cards and it performs no periodic checks to ensure proper accountability over their use. Therefore, USDA cannot tell whether any of these cards have been lost or stolen. Although USDA officials were unable to tell us how many employees have telephone company calling cards, one official told us hundreds of agency staff have been using them regularly to charge long-distance services. USDA also does not know the extent to which it has been the victim of toll fraud committed by outside hackers. In this regard, USDA has had at least one instance where hackers broke into USDA’s telephone system and, according to USDA records, made an estimated $40,000 to $50,000 in international long-distance calls over one weekend in March 1995. In this case, the hacker penetrated the Department’s telephone system by successfully exploiting vulnerabilities in a USDA contractor’s voice mail system. USDA only became aware of this incident after it was identified by a long-distance carrier and brought to the Department’s attention. To make matters worse, USDA did not seek reimbursement for any of the fraudulent calls it paid for from the voice mail contractor even though the contractor acknowledged that it was to blame for the vulnerabilities in the voice mail system. The extent of USDA’s telephone abuse and fraud problem is unknown and could be costing the Department thousands of dollars each month. Like other problems we identified in earlier reports, the Department lacks adequate management controls over the $50 million it spends each year for commercial telecommunications services. To its credit, the Department has begun to take positive steps toward addressing some of its telecommunications management weaknesses by planning an effort to reengineer telecommunications management and making billing data more accessible to agency managers for review. If successful, this effort, which will take about 2 years to implement according to USDA, should also help deter telephone abuse and fraud. However, without taking interim steps to determine its vulnerability to telephone abuse and fraud, identify cost-effective ways to enforce current policies and procedures, and investigate and take action on past abuses, the Department is at risk of continued losses to telephone abuse and fraud. We recommend that the Secretary direct the Assistant Secretary for Administration and the Chief Financial Officer, in cooperation with the Under Secretaries and the Office of Inspector General, to determine the risk of and vulnerability to telephone fraud, waste, and abuse departmentwide, develop an appropriate plan with cost-effective controls to mitigate these risks, and expeditiously implement this plan. In developing this plan, among other things the Department should consider determining whether there is a need to continue to accept collect calls and, if deemed necessary, evaluate the viability and cost-effectiveness of alternatives to collect calls such as offering toll free numbers. In the interim, the Department should identify and implement actions necessary, but at the same time cost-effective, to minimize USDA’s exposure to telephone abuse. Alternatives that the Department might consider could include blocking collect calls to the Department, notifying commercial carriers to cancel all telephone company credit cards issued to USDA personnel, and/or identifying methods that other large organizations employ to combat telephone abuse and fraud. The Secretary should also direct the Assistant Secretary for Administration and the Chief Financial Officer, in cooperation with the Under Secretaries and the Office of Inspector General, to take appropriate disciplinary actions against employees involved in the telabuse cases we identified to ensure that these abuses are stopped immediately and recover losses where it is cost-effective to do so. We also recommend that the Secretary direct that billing records be reviewed to identify all long-distance and other service charges associated with the March 1995 hacker incident and expeditiously seek restitution for these amounts from the contractor responsible for the defective voice mail equipment that led to these charges. We further recommend that the Secretary of Agriculture, in accordance with 31 U.S.C. 720, provide a written statement on actions taken on recommendations contained in our prior report, USDA Telecommunications: Better Management and Network Planning Could Save Millions (GAO/AIMD-95-203), to the Senate Committee on Governmental Affairs and the House Committee on Government Reform and Oversight. A written statement also must be sent to the House and Senate Committees on Appropriations. In addition, the Secretary should provide the appropriate congressional oversight committees with a report on the Department’s planned actions to correct its telecommunications management weaknesses and mitigate the risk of telephone fraud, waste, and abuse. USDA’s Assistant Secretary for Administration provided written comments on April 1, 1996, on a draft of this report. USDA’s comments are summarized below and reproduced in appendix II. The Assistant Secretary agreed with the need to strengthen telecommunications management controls in the Department and to prevent fraud and abuse of telecommunications services. Regarding cases of abuse that we identified involving collect calls from correctional centers, the Assistant Secretary stated that on March 13, 1996, the deputy administrators for management within each USDA agency in the Washington, D.C., area were briefed on telephone abuse and provided copies of the past 6 months’ commercial telephone bills to review. In addition, the Assistant Secretary stated that the deputy administrators were also given telephone numbers to investigate and were instructed to pursue disciplinary action against employees who are found to have abused the use of USDA telephones. The Assistant Secretary also stated that USDA is exploring the potential for blocking all third party and collect calls in the Washington, D.C., metropolitan area and nationwide, and replacing these services with the expanded use of “1-800” service and FTS 2000 telephone credit cards as a way of reducing telephone abuse and fraud. In addition, the Assistant Secretary stated that USDA will, as we recommended, seek reimbursement for the cost of all calls paid for by the Department during the March 1995 telephone hacker incident. The Assistant Secretary also agreed that telephone abuse and fraud at USDA is indicative of systemic weaknesses in the Department’s existing processes for billing and paying for telecommunications services. In this regard, the Assistant Secretary stated that a team is now being assembled to begin work on implementing the telecommunications task force’s recommendation which we discussed. We are encouraged by the actions described by the Assistant Secretary for Administration to prevent fraudulent use of government telephones at USDA. While the Assistant Secretary did not respond to our specific recommendations, it is important for the Department to address actions it plans to take on each recommendation as it moves ahead in preventing telephone fraud and abuse. It is especially important for the Department to implement our first recommendation that the Secretary direct the Assistant Secretary for Administration and the CFO, in cooperation with the Under Secretaries and the OIG, to determine the risk of and vulnerability to telephone fraud, waste, and abuse departmentwide, develop an appropriate plan with cost-effective controls to mitigate these risks, and expeditiously implement this plan. We are sending copies of this report to the Secretary of Agriculture; the Chairmen and Ranking Minority Members of the Senate Committee on Governmental Affairs, the Senate and House Committees on Appropriations, the House Committee on Agriculture, and the House Committee on Government Reform and Oversight; the Director of the Office of Management and Budget; the Administrator of the General Services Administration; and other interested parties. Copies will also be made available to others upon request. Please contact me or Steve Schwartz at (202) 512-6240 if you or your staff have any questions concerning the report. To assess USDA’s controls over telephone use, we obtained and reviewed commercial telephone billing records representing 4 monthly billing periods during fiscal year 1995 for USDA agency offices in the Washington, D.C., metropolitan area. Our 4-month sample of commercial telephone billing records totaled about $580,000 or 1 percent of the $50 million USDA spends annually for commercial telecommunications costs. The four months in our review—December 1994, March 1995, July 1995, and August 1995—were selected from early, mid, and late parts of the fiscal year to adjust for any seasonal variations in calling patterns. We reviewed billing records of all collect calls accepted by the Department during these 4 months as well as selected long-distance calls made during the month of August 1995. We were unable to review all of USDA’s telephone bills for the periods covered by our review because the Department did not provide all the bills to us by the end of our audit work in February 1996. According to the official at USDA’s National Finance Center responsible for handling our request for bills, there were delays because complex computer runs were necessary to identify commercial billing accounts associated with all the 24,000 separate telephone lines in the Washington, D.C., metropolitan area and because a manual process is used at the National Finance Center for tracking down each paper bill. USDA subsequently provided additional bills in March 1996, but since this information was submitted after we had completed our audit work, it was not included in our report. To confirm that the cases we identified involved telephone abuse, we also discussed them with officials in USDA’s Office of Information Resources Management as well as telephone company representatives and we provided records for these calls to USDA officials for appropriate action. To obtain detailed information on cases of collect calling abuse we found, we interviewed correctional facility personnel about cases involving collect calls from correctional centers and discussed these cases with officials in USDA’s Office of Inspector General and Office of Information Resources Management. In addition, we reviewed telephone company records and USDA documentation pertaining to a March 1995 hacker case at the Department and interviewed telephone company representatives, voice mail vendor staff, and USDA officials involved in the incident. To identify USDA’s procedures for processing commercial telephone bills for its offices in the Washington, D.C., metropolitan area, we interviewed officials from USDA Office of Information Resources Management and General Services Administration. We also reviewed industry publications and reports on telabuse and toll fraud. We examined the Department’s policies and procedures for managing the use of government telephones and commercial telephone and long-distance services and USDA plans for improving telecommunications management controls. We performed our audit work from October 1995 through February 1996, in accordance with generally accepted government auditing standards. Our work was primarily done at USDA headquarters in Washington, D.C. We also conducted work at the General Services Administration in Washington, D.C.; Prince George’s County Correctional Center in Upper Marlboro, Maryland; and the Animal and Plant Health Inspection Service in College Park, Maryland. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO reviewed the Department of Agriculture's (USDA) use of its telecommunications resources, focusing on: (1) whether USDA ensures that the commercial telephone and long-distance services it pays for are used in accordance with federal regulations and departmental policy; and (2) USDA efforts to address recommendations from a previous GAO report. GAO found that: (1) during the 4-month period reviewed, collect calls from federal, state, and county correctional institutions were accepted at 20 USDA offices in the District of Columbia area and may have been transferred to USDA long-distance lines; (2) these collect calls amounted to 50 percent of the collect calls accepted at these USDA offices and cost about $2,600; (3) despite the discovery of inappropriately accepted collect calls as early as 1993, USDA did not initiate adequate measures to stop the abuses; (4) USDA offices in the District area made unauthorized long-distance calls, including international calls to adult entertainment lines and companies advertising jobs; (5) such fraud and abuse exist because USDA does not review its commercial telephone bills; (6) despite some positive actions to control fraud and abuse, USDA has not responded to GAO recommendations concerning the inappropriate use of its telecommunications resources; and (7) USDA is vulnerable to more fraud, waste, and abuse because it does not review its telephone and telephone credit card bills.

The dramatic decline in the U.S. housing market that began in 2006 precipitated a decline in the price of mortgage-related assets, particularly mortgage assets based on subprime loans, in 2007. Some institutions found themselves so exposed that they were threatened with failure, and some failed because they were unable to raise capital or obtain liquidity as the value of their portfolios declined. Other institutions, ranging from government-sponsored enterprises such as Fannie Mae and Freddie Mac to large securities firms, were left holding “toxic” mortgages or mortgage- related assets that became increasingly difficult to value, were illiquid, and potentially had little worth. Moreover, investors not only stopped buying private-label securities backed by mortgages but also became reluctant to buy securities backed by other types of assets. Because of uncertainty about the liquidity and solvency of financial entities, the prices banks charged each other for funds rose dramatically, and interbank lending conditions deteriorated sharply. The resulting liquidity and credit crunch made the financing on which businesses and individuals depend increasingly difficult to obtain. By late summer of 2008, the ramifications of the financial crisis ranged from the continued failure of financial institutions to increased losses of individual savings and corporate investments and further tightening of credit that would exacerbate the emerging global economic slowdown. Treasury and federal financial regulators play a role in regulating and monitoring the financial system. Historically, Treasury’s mission has been to act as steward of U.S. economic and financial systems. Among its many activities, Treasury has taken a leading role in addressing underlying issues such as those precipitating the recent financial crisis. The key federal banking regulators include the following: The Federal Reserve, an independent agency that is responsible for conducting the nation’s monetary policy by influencing the monetary and credit conditions in the economy in pursuit of maximum employment, stable prices, and moderate long-term interest rates; supervising and regulating bank holding companies and state-chartered banks that are members of the Federal Reserve System; and maintaining the stability of the financial system and containing systemic risk that may arise in financial markets through its role as lender of last resort; FDIC, an independent agency created to help maintain stability and public confidence in the nation’s financial system by insuring deposits, examining and supervising insured state-chartered banks that are not members of the Federal Reserve System, and resolving failed or failing banks; OCC, which charters and supervises national banks; and the Office of Thrift Supervision, which supervises savings associations (thrifts) and savings and loan holding companies. In 1991, Congress enacted FDICIA in response to the savings and loan crisis. FDICIA enacted a number of reforms, including some designed to address criticisms that federal regulators had not taken prompt and forceful actions to minimize or prevent losses to the deposit insurance funds caused by bank and thrift failures. Among other things, FDICIA amended the FDI Act by establishing a rule requiring FDIC to follow the least costly approach when resolving an insured depository institution. Specifically, under the least cost rule, FDIC must resolve a troubled insured depository institution using the method expected to have the least cost to the deposit insurance fund and cannot use the fund to protect uninsured depositors and creditors who are not insured depositors if such protection would increase losses to the fund. To make a least-cost determination, FDIC must (1) consider and evaluate all possible resolution alternatives by computing and comparing their costs on a present-value basis, and (2) select the least costly alternative on the basis of the evaluation. Under the least-cost requirements, FDIC generally has resolved failed or failing banks using three basic methods, which do not constitute open bank assistance. These are: (1) directly paying depositors the insured amount of their deposits and disposing of the failed bank’s assets (deposit payoff and asset liquidation); (2) selling only the bank’s insured deposits and certain other liabilities, and some of its assets, to an acquirer (insured deposit transfer); and (3) selling some or all of the failed bank’s deposits, certain other liabilities, and some or all of its assets to an acquirer (purchase and assumption). According to FDIC officials, they have most commonly used purchase and assumption, as it is often the least costly and disruptive alternative. FDICIA also amended the FDI Act to create an exception to the least-cost requirements, known as the systemic risk exception, that allows FDIC assistance without complying with the least cost rule if compliance would have “serious adverse effects on economic conditions and financial stability”—that is, would cause systemic risk—and if such assistance would “avoid or mitigate such adverse effects.” FDIC may act under the exception only under the process specified in the statute. The FDIC Board of Directors and the Board of Governors of the Federal Reserve System each must recommend use of the exception by a vote of not less than two- thirds of their respective members and deliver a written recommendation to the Secretary of the Treasury. Based on a review of the FDIC and Federal Reserve recommendations, the Secretary of the Treasury, in consultation with the President, may make a systemic risk determination authorizing FDIC to take action or provide assistance that does not meet the least-cost requirements. For example, under a systemic risk determination, FDIC is not bound to identify and follow the least-cost resolution strategy and may provide assistance (such as debt or deposit guarantees) that protects uninsured depositors and creditors, who otherwise might suffer losses under a least-cost method such as a purchase and assumption or depositor payoff. Until recently, the systemic risk exception required FDIC to recover any resulting losses to the insurance fund by levying one or more emergency special assessments on insured depository institutions. Congress amended this requirement in May 2009 to also authorize assessments on bank holding companies, and savings and loan holding companies. Finally, the systemic risk exception includes requirements that serve to ensure accountability for regulators’ use of this provision. The Secretary of the Treasury must notify Congress in writing of any systemic risk determination and must document each determination and retain the documentation for GAO review, and GAO must report its findings to Congress. On five occasions, collaboration among high-level officials at Treasury, FDIC, and the Federal Reserve resulted in the announcement of emergency actions that would require a systemic risk exception. FDIC and the Federal Reserve provided written recommendations to the Secretary of the Treasury for all five announced actions, but the Secretary has made a determination on only three of these announcements. Treasury made the first two determinations concurrent with the initial announcements, and the third determination was made nearly 2 months after the announcement of action. Treasury has not made a determination on the remaining two announced actions which have not been implemented to date. Such announcements can affect market expectations and contribute to moral hazard, but the announcements alone—absent a Treasury determination— do not trigger requirements established by Congress for documentation and communication of the agencies’ use of the systemic risk exception. Such requirements serve to ensure transparency and accountability related to the application of the systemic risk exception. On five occasions between late 2008 and early 2009, regulators announced potential emergency actions that would require a systemic risk determination before they could be implemented. In each case, a liquidity crisis—either at a single institution or across the banking industry— triggered discussions among FDIC, Federal Reserve, and Treasury officials about whether to invoke the systemic risk exception. According to regulators, these discussions generally occurred among high-level officials at the three agencies over a period of a few days, through e-mail, memorandums, telephone calls, and emergency meetings. The regulators shared and analyzed information, such as data describing the liquidity pressures facing financial institutions, to help them understand the financial condition of the troubled institutions and the potential systemic implications of complying with the least-cost resolution requirements. In the second section of this report, we discuss in detail publicly available information about the financial condition of the institutions that received emergency assistance, the basis for each decision to invoke the systemic risk exception, and the actions that FDIC took under the provision. Following these collaborations, FDIC and Federal Reserve staff submitted documentation of their analyses and recommendations to support invoking the systemic risk exception to their respective Boards. In each of the five cases, FDIC’s Board of Directors and Federal Reserve Board members voted in favor of recommending a systemic risk determination, and FDIC and the Federal Reserve provided written recommendations to the Secretary of the Treasury (see fig. 1). On each occasion, the regulators issued public statements announcing planned FDIC actions that would require a systemic risk determination for implementation. The Secretary of the Treasury made a determination in response to three of the five recommendations (Wachovia, TLGP, and Citigroup); therefore, we reviewed, as provided by the mandate, documentation related to these three cases. Treasury documents that we reviewed indicate that the Secretary of the Treasury signed and approved the determinations (as required by the FDI Act) and authorized FDIC to take planned action after having reviewed the FDIC and the Federal Reserve’s written recommendations and consulted with the President. Also as required by the FDI Act, Treasury sent letters to Congress to notify the relevant committees of all three determinations. In all five cases, planned emergency actions were announced by regulators, but Treasury did not make an immediate determination for three of these announcements and still has not made a determination to date in two of them. In two cases, Wachovia and TLGP, Treasury made a determination before regulators finalized the terms of the assistance. According to Treasury, FDIC, and Federal Reserve officials, they publicly announced emergency assistance prior to a Treasury determination in these cases to reassure the markets that the government was committed to supporting financial market stability. In the Citigroup case, the public announcement preceded Treasury’s determination by about 2 months. Specifically, on November 23, 2008, Treasury, FDIC, and the Federal Reserve jointly announced an agreement-in-principle to assist Citigroup. FDIC and the Federal Reserve delivered written recommendations by early December 2008 and Treasury signed the determination in January 2009 when the finalized agreement was executed. Since the Citigroup determination, Treasury has not made determinations following two announcements of emergency actions and those announced initiatives have not been implemented. On January 16, 2009, FDIC announced an agreement-in-principle with Bank of America to share losses on a fixed pool of Bank of America assets. Although FDIC and the Federal Reserve provided written recommendations in support of a determination, according to Treasury and FDIC officials, the Secretary of the Treasury did not make a determination at the time because the terms of the agreement had not been finalized. In May 2009, Bank of America requested a termination of the term sheet for the announced guarantee of up to $118 billion in assets by the U.S. government and in September 2009, the parties to the agreement-in-principle executed a termination agreement in which Bank of America agreed to pay $425 million to Treasury, the Federal Reserve, and FDIC. Similarly, on March 23, 2009, FDIC and Treasury announced the creation of the PPIP’s LLP, but Treasury has not yet made a determination. According to a Treasury official with whom we spoke, Treasury has delayed making a determination while regulators considered how to structure the program. While important to stabilizing markets, the public announcement of planned actions can serve as a de facto determination by implying that Treasury has made a systemic risk determination. An announcement alone could have given rise to some of the benefits of a systemic risk determination, while similarly generating the potential for negative incentives such as moral hazard. For example, although FDIC did not provide assistance to Bank of America, the announcement of the planned Bank of America guarantees signaled regulators’ willingness to provide such assistance and may have achieved to some degree the intended effect of increasing market confidence in Bank of America. The agreement requiring Bank of America to pay a $425 million termination fee recognized that although the parties never entered into a definitive documentation of the transaction, Bank of America received value from the announced term sheet, including benefits in terms of market confidence in the institution. Although the effects of announcements and determinations can be similar, determinations must be conducted under procedural and documentation requirements that do not apply to announcements. Under the determination process, Treasury must consider recommendations from FDIC and the Federal Reserve, consult with the President before making a determination, and document its reasons for making a determination and retain the documentation for later review. Treasury must also notify Congress in writing of each systemic risk determination. None of these requirements applies when a determination is not made. We acknowledge that Treasury is not required to make a determination within a set period and recognize the need for some flexibility during crisis situations. However, absent a determination, the agency is not required to follow the formal process put in place by Congress to ensure transparency and accountability in the application of the systemic risk exception. Therefore, when a determination is not made along with the announced actions, Congress cannot be assured that Treasury’s reasoning would be open to the same scrutiny required in connection with a formal systemic risk determination because Treasury does not have to act upon the FDI Act’s documentation and accountability measures. For instance, Congress cannot be assured that the documentation required to support a determination will be or has been generated, even when the announcement by the agencies can have some of the same effects a systemic risk determination would have. Furthermore, uncertainty in de facto determination situations can arise because Treasury is not required to communicate that it will not make a systemic risk determination for an announced action. For example, since the announcement proposing the creation of the PPIP’s LLP in March 2009, it has not been clear whether Treasury intends to make a systemic risk determination, raising questions about whether Treasury will make a determination to authorize the program. The Secretary of the Treasury’s three systemic risk determinations authorized FDIC guarantees that FDIC, the Federal Reserve, and Treasury determined were needed to avoid or mitigate further serious adverse effects on already deteriorating financial and economic conditions. Treasury invoked the exception so that FDIC could provide assistance to Wachovia and its insured institution subsidiaries, the banking industry as a whole (through TLGP), and Citigroup and its insured institution subsidiaries. In describing the basis for the first systemic risk determination in September 2008, Treasury, FDIC, and the Federal Reserve noted that mounting problems at Wachovia could have led to a failure of the firm, which in turn could have exacerbated the disruption in the financial markets. At the time, the failures and near-failures of several large institutions had increased stress in key funding markets. As noted earlier, by late summer 2008, the potential ramifications of the financial crisis included the continued failure of financial institutions and further tightening of credit that would exacerbate the emerging global economic slowdown that was beginning to take shape. In this environment, many financial institutions, including Wachovia, were facing difficulties in raising capital and meeting their funding obligations. In its recommendation, FDIC said that the rapidly deteriorating financial condition of Wachovia Bank, N.A.—Wachovia’s largest bank subsidiary— was due largely to its portfolio of payment-option adjustable-rate mortgage (ARM) products, commercial real-estate portfolio, and weakened liquidity position. Over the first half of 2008, Wachovia had suffered more than $9 billion in losses due in part to mortgage-related asset losses and investors increasingly had become concerned about the firm’s prospects, given the worsening outlook for home prices and mortgage credit quality. In addition, during the week preceding Treasury’s determination, Wachovia’s stock price declined precipitously and the spreads on credit default swaps that provide protection against losses on Wachovia’s debt widened, indicating that investors considered a Wachovia default increasingly likely. FDIC consulted with Treasury and the Federal Reserve in conducting an analysis of Wachovia’s liquidity and determined that Wachovia would soon be unable to meet its funding obligations as a result of strains on its liquidity, particularly from projected outflows of deposits and retail brokerage accounts. In considering actions to avert a Wachovia failure, Treasury determined that a least-cost resolution of Wachovia’s bank and thrift subsidiaries, without protecting creditors and uninsured depositors, could—in light of conditions in the financial markets and the economy at the time—weaken confidence and exacerbate liquidity strains in the banking system. FDIC could have effected a least-cost resolution of Wachovia Bank, N.A. through a depositor payoff or purchase and assumption transaction following appointment of FDIC as the receiver of the bank’s assets. FDIC and the Federal Reserve projected that either of these least-cost resolution options would have resulted in no cost to the deposit insurance fund, but that either option likely would have imposed significant losses on subordinated debtholders and possibly senior note holders. In addition, Treasury, the Federal Reserve, and FDIC expected these resolution options to impose losses on foreign depositors, a significant funding source for several large U.S. institutions. Their concerns over the possible significant losses to creditors holding Wachovia subordinated debt and senior debt were reinforced by the recent failure of Washington Mutual, a large thrift holding company. According to Treasury’s determination, under the least- cost resolution of Washington Mutual, senior and subordinated debtholders of the holding company and its insured depository subsidiaries suffered large losses. Treasury, FDIC, and the Federal Reserve expressed concern that imposing similarly large losses on Wachovia’s creditors and foreign depositors could intensify liquidity pressures on other U.S. banks, which were vulnerable to a loss of confidence by creditors and uninsured depositors (including foreign depositors), given the stresses already present in the financial markets at that time. According to FDIC and Federal Reserve documents, Wachovia’s sudden failure would have led to investor concern about direct exposures of other financial institutions to Wachovia. Furthermore, a Wachovia failure also could have led investors and other market participants to doubt the financial strength of other institutions that might be seen as similarly situated. In particular, the agencies noted that a Wachovia failure could intensify pressures on other large banking organizations that, like Wachovia, reported they were well capitalized but continued to face investor concerns about deteriorating asset quality. At the time of the Wachovia determination, the Emergency Economic Stabilization Act had not yet been passed and, thus, the authorities under that law to create the Troubled Asset Relief Program (TARP) were not available to help mitigate these effects. Furthermore, a least-cost resolution of Wachovia, N.A. could have negatively affected the broader economy, because with banks experiencing reduced liquidity and increased funding costs, they would be less willing to lend to businesses and households. In recommending a systemic risk determination, the Federal Reserve and FDIC described the extent of Wachovia’s interdependencies and the potential for disruptions to markets in which it played a significant role. The Federal Reserve listed the top financial entities exposed to Wachovia, noting that mutual funds were prominent among these counterparties. In addition, FDIC expressed concern that a Wachovia failure could result in losses for mutual funds holding its commercial paper, accelerating runs on those and other mutual funds. The Federal Reserve also noted that Wachovia was a major participant in the full range of major domestic and international clearing and settlement systems and that a least-cost resolution would likely have raised some payment and settlement concerns. Treasury, FDIC, and the Federal Reserve concluded that FDIC assistance under the systemic risk exception could avert the potential systemic consequences of a least-cost resolution of Wachovia’s bank and thrift subsidiaries. In particular, they determined that authorizing FDIC guarantees to protect against losses to Wachovia’s uninsured creditors would avoid or mitigate the potential for serious adverse effects on the financial system and the economy by facilitating the acquisition of Wachovia’s banking operations by Citigroup. On September 29, 2008, pursuant to Treasury’s systemic risk determination, FDIC announced that it had agreed to provide protection against large losses on a fixed pool of Wachovia assets to facilitate the orderly sale of Wachovia’s banking operations to Citigroup and avert an imminent failure that might exacerbate the serious strains then affecting the financial markets, financial institutions, and the economy. On September 28, 2008, Citigroup and Wells Fargo both submitted bids to FDIC to acquire Wachovia’s banking operations with FDIC open bank assistance in the form of loss sharing on Wachovia assets. The Citigroup and Wells Fargo bids differed in terms of the amount of losses each proposed to absorb and the result of the bidding process held by FDIC was the acceptance of Citigroup’s bid. After agreeing with FDIC to a loss- sharing agreement on selected Wachovia assets, Citigroup announced that it would acquire Wachovia’s banking operations for $2.2 billion and assume the related liabilities, including senior and subordinated debt obligations and all of Wachovia’s uninsured deposits. Under the agreement, Citigroup agreed to absorb the first $42 billion of losses on a $312 billion pool of loans and FDIC agreed to assume losses beyond that. To compensate FDIC for its assumption of this risk, Citigroup agreed to grant FDIC $12 billion in preferred stock and warrants. A few days after the announcement of the proposed Citigroup acquisition, Wachovia announced that it would instead merge with Wells Fargo in a transaction that would include all of Wachovia’s operations and, in contrast to the bids submitted days earlier by Citigroup and Wells Fargo, require no FDIC assistance. As a result, the FDIC loss-sharing agreement on Wachovia assets was not implemented and no assistance was provided under the systemic risk exception. Although the loss-sharing agreement never took effect, the announcement of the Citigroup acquisition and loss-sharing agreement may have helped to avert a Wachovia failure with potential systemic consequences. While acknowledging that isolating the impact of FDIC’s assistance from other factors is difficult, Treasury and FDIC officials with whom we spoke said that one measure of the success of the loss-sharing agreement was that Wachovia was able to remain open and meet its funding obligations on Monday, September 29, 2008. In particular, the determination and the announcement of Citigroup’s assumption of debt and deposit liabilities of Wachovia and its insured bank and thrift subsidiaries may have helped to allay the concerns of creditors and depositors that might otherwise have withdrawn liquidity support. As Wachovia did not fail, the extent to which a Wachovia failure would have had adverse effects on financial stability is not known. In describing the basis for the second systemic risk determination, which authorized TLGP, Treasury, FDIC, and the Federal Reserve said that disruptions in credit markets posed a threat to the ability of many institutions to fund themselves and lend to consumers and businesses. In a memorandum provided to Treasury, FDIC noted that the reluctance of banks and investment managers to lend to other banks and their holding companies made finding replacement funding at a reasonable cost difficult for these financial institutions. The TED spread—a key indicator of credit risk that gauges the willingness of banks to lend to other banks—peaked at more than 400 basis points in October 2008, likely indicating an increase in both perceived risk and in risk aversion among investors (see fig. 2). In addition to disruptions in interbank lending, financial institutions also faced difficulties raising funds through commercial paper and asset- backed securitization markets. The resulting credit crunch made the financing on which businesses and individuals depend increasingly difficult to obtain. In addition, FDIC was concerned that large outflows of uninsured deposits could strain many banks’ liquidity. According to FDIC officials with whom we spoke, they were not tracking outflows of these deposits, but relied on anecdotal reports from institutions and the regulators serving as their primary supervisors. In light of the liquidity strains many institutions faced, Treasury, FDIC, and the Federal Reserve determined that resolving institutions on a bank-by- bank basis in compliance with least-cost requirements would result in adverse impacts on financial stability and the broader economy. In its recommendation letter, FDIC concluded that the threat to the market for bank debt was a systemic problem that threatened the stability of a significant number of institutions, thereby increasing the potential for failures of these institutions and losses to the Deposit Insurance Fund. The Federal Reserve reasoned, among other things, that the failures and least- cost resolutions of a number of institutions could impose unexpected losses on investors and further undermine confidence in the banking system, which already was under extreme stress. Treasury concurred with FDIC and the Federal Reserve in determining that relying on the least-cost resolution process would not sufficiently address the systemic threat to bank funding and the broader economy. Treasury concluded that FDIC actions under a systemic risk exception would avoid or mitigate adverse effects that would have resulted if assistance were provided subject to the least cost rules. Specifically, Treasury, FDIC, and the Federal Reserve advised that certain FDIC debt and deposit guarantees—otherwise subject to the prohibition against use of the Deposit Insurance Fund to protect uninsured depositors and creditors who are not insured depositors—could address risk aversion among institutions and investors that had become reluctant to provide liquidity to financial institutions and their holding companies. In a memorandum describing the basis for TLGP determination, Treasury explained the need for emergency actions in the context of a recent agreement among the United States and its G7 colleagues to implement a comprehensive action plan to provide liquidity to markets and prevent the failure of any systemically important institution, among other objectives. To implement the G7 plan, several countries had already announced programs to guarantee retail deposits and new debt issued by financial institutions. Treasury noted that if the United States did not take similar actions, global market participants might turn to institutions and markets in countries where the perceived protections were the greatest. Some have noted that under a possible reading of the exception, the statute may authorize assistance only to particular institutions, based on those institutions’ specific problems, not, as was done in creating TLGP, systemic risk assistance based on problems affecting the banking industry as a whole. Treasury, FDIC, and the Federal Reserve considered this and other legal issues in recommending and making TLGP determination. The agencies believe the statute could have been drafted more clearly and that it can be interpreted in different ways. They concluded, however, that under a permissible interpretation, assistance may be based on industry- wide concerns. They also concluded that a systemic risk determination waives all of the normal statutory restrictions on FDIC assistance and then creates new authority to provide assistance, both as to the types of aid that may be provided and the entities that may receive it. Under this reading, the agencies believe the statutory criteria were met in the case of TLGP and that the assistance was authorized. We examined these issues as part of our review of the basis of the systemic risk determinations made to date. As detailed in appendix II, the recent financial crisis is the first time the agencies have relied on the systemic risk exception since its enactment in 1991, and no court to date has ruled on when or how it may be used. We found there is some support for the agencies’ position that the exception authorizes assistance of some type under TLGP facts, as well as for their position that the exception permits assistance to the entities covered by this program. There are a number of questions concerning these interpretations, however. In the agencies’ view, for example, some of the statutory provisions are ambiguous. What is clear, however, that the systemic risk exception overrides important statutory restrictions designed to minimize costs to the Deposit Insurance Fund, and in the case of TLGP, that the agencies used it to create a broad-based program of direct FDIC assistance to institutions that had never before received such relief—”healthy” banks, bank holding companies, and other bank affiliates. Because application of the systemic risk exception raises novel legal and policy issues of significant public interest and importance, and because of the need for clear direction to the agencies in a time of financial crisis, the requirements and the assistance authorized under the systemic risk exception may require clarification by Congress. In October 2008, FDIC created TLGP to complement the TARP Capital Purchase Program and the Federal Reserve’s Commercial Paper Funding Facility and other liquidity facilities in restoring confidence in financial institutions and repairing their capacity to meet the credit needs of American households and businesses. TLGP’s Debt Guarantee Program (DGP) was designed to improve liquidity in term-funding markets by guaranteeing certain newly issued senior unsecured debt of financial institutions and their holding companies. According to FDIC officials, by guaranteeing payment of these debt obligations, DGP was intended to address the difficulty that creditworthy institutions were facing in replacing maturing debt because of risk aversion in the markets. TLGP’s Transaction Account Guarantee Program (TAGP) also was created to stabilize an important source of liquidity for many financial institutions. TAGP temporarily extended an unlimited deposit guarantee to certain non- interest-bearing transaction accounts to assure holders of the safety of these deposits and limit further outflows. By facilitating access to borrowed funds at lower rates, Treasury, FDIC, and the Federal Reserve expected TLGP to free up funding for banks to make loans to creditworthy businesses and consumers. Furthermore, by promoting stable funding sources for financial institutions, they intended TLGP to help avert bank and thrift failures that would impose costs on the insurance fund and taxpayers and potentially contribute to a worsening of the crisis. FDIC structured TLGP requirements to provide needed assistance to insured banks while avoiding costs to the deposit insurance fund. According to FDIC officials, in designing TLGP, FDIC sought to achieve broad participation to avoid the perception that only weak institutions participated and to help ensure collection of fees needed to cover potential losses. Initially, all eligible institutions, which included insured depository institutions, their holding companies, and qualified affiliates, were enrolled in TLGP for 30 days at no cost and only those that participated in DGP or TAGP (or both) after the opt-out date became subject to fee assessments. Table 1 provides additional details related to TLGP features and requirements. As of March 31, 2009, among depository institutions with assets over $10 billion, 92 percent and 94 percent had opted into DGP and TAGP, respectively. According to one regulatory official, this high participation rate indicated that many large institutions judged the benefits of the program to outweigh fee and other costs. However, while seeking to encourage broad participation, TLGP was not intended to prop up nonviable institutions, according to FDIC officials. The TLGP rule allowed FDIC to prospectively cancel eligibility for DGP if an institution had weak supervisory ratings. According to FDIC officials, some financial institutions were privately notified by their regulatory supervisors that they were not eligible to issue TLGP-guaranteed debt. In addition, FDIC required all parts of a holding company to make the same decision about TLGP participation to prevent an entity from issuing guaranteed debt through its weakest subsidiary. As of December 31, 2009, FDIC had collected $11.0 billion in TLGP fees and surcharges and incurred claims of $6.6 billion on TLGP guarantees. All of the claims to date, except for one $2 million claim under DGP, have come from TAGP, under which FDIC has collected $639 million in fees. Since the creation of TLGP, bank failures have been concentrated among small banks (assets under $10 billion), which as a group have not been significant participants in DGP. Although the high number of small bank failures has resulted in higher-than-expected costs under TAGP, FDIC officials still expect total TLGP fees collected to exceed the costs of the program. At the end of the program, FDIC will be permitted to account for any excess TLGP fees as income to the deposit insurance fund. If a supportable and documented analysis demonstrates that TLGP assets exceed projected losses, FDIC may recognize income to the deposit insurance fund prior to the end of the program. As noted earlier, in the event that TLGP results in losses to the deposit insurance fund, FDIC would be required to recover these losses through one or more special assessments. Since TLGP was creted in Octoer 200, FDIC has extended oth component of the progrm. In Mrch 2009, FDIC extended the finl dte for new det issuance nder DGP from Jne 30, 2009, to Octoer 31, 2009, nd in Aust 2009, extended the TAGP for 6 month, throgh Jne 30, 2010. FDIC offici with whom we poke said they consulted with other regtor in determining th epte temic rik determintion was not reqired for thee extenion. Thee offici noted tht economic conditiond improved t the time of the extenion, but hd not yet retrned to precri condition. FDIC noted the need to ensure n orderly phase-ot of TLGP assnce nd otlined certin higher fee reqirement for intittion chooing to contine prticiption past the originl end dte. Art of the DGP extenion, FDIC eablihed new surchrge eginning on April 1, 2009, for certin det issued prior to the originl Jne 2009 dedline nd for ll det issued nder the extenion. In extending TAGP, FDIC nnonced tht eligile intittion not opting ot of the 6-month extenion wold subject to higher fee based on the intittion’ rik ctegory as determined y regtor assssment. effective in improving short-term and intermediate-term funding markets. In addition, FDIC officials with whom we spoke said that although they do not track outflows of deposits of transaction accounts covered under TAGP, several institutions have told them that TAGP was helpful in stemming such outflows. DGP conclded on Octoer 31, 2009, for mot entitierticipting in the progrm. To frther ensure n orderly phase-ot of the progrm, FDIC eablihed limited emergency guantee fcility throgh which eligile entitie (pon ppliction nd FDIC pprovl) cold issue guanteed det throgh April 30, 2010, subject to minimnnualized assssment of 300 bas point. In April 2010, FDIC’ Bord of Director pproved n interim rle to extend the TAGP ntil Decemer 31, 2010, nd give the Bord dicretion to extend the progrm to the end of 2011, if necessary. Moreover, some market observers have commented that FDIC’s assumption of risk through the debt guarantees enabled many institutions to obtain needed funding at significantly lower costs. Eligible financial institutions and their holding companies raised more than $600 billion under DGP, which concluded on October 31, 2009, for most participating entities. Notably, several large financial holding companies each issued tens of billions of dollars of TLGP-guaranteed debt and most did not issue senior unsecured debt outside DGP before April 2009. Although determining the extent to which FDIC guarantees lowered debt costs is difficult, a U.S. government guarantee significantly reduces the risk of loss and accordingly, and would be expected to substantially reduce the interest rate lenders charge for TLGP-guaranteed funds. By comparing yields on TLGP-guaranteed debt to yields on similar debt issued without FDIC guarantees, some market observers have estimated that FDIC guarantees lowered the cost of certain debt issues by more than 140 basis points. To the extent that TLGP helped banking organizations to raise funds during a very difficult period and to do so at substantially lower cost than would otherwise be available, it may have helped improve confidence in institutions and their ability to lend. However, some market observers have expressed concern that the large volume of issuance under TLGP could create difficulties associated with rolling over this debt in a few years when much of this debt matures in a short time frame. According to one financial analyst with whom we spoke, potential difficulties associated with rolling over this debt could be mitigated by any improvements in other funding markets, such as asset-backed securitization markets. In describing the basis for the third systemic risk determination, Treasury, FDIC, and the Federal Reserve cited concerns similar to those discussed in connection with the Wachovia determination. During November 2008, severe economic conditions persisted despite new Federal Reserve liquidity programs and the announcements of the Treasury’s Capital Purchase Program and FDIC’s TLGP. Similar to Wachovia, Citigroup had suffered substantial losses on mortgage-related assets and faced increasing pressures on its liquidity as investor confidence in the firm’s prospects and the outlook for the economy declined. On Friday, November 21, 2008, Citigroup’s stock price fell below $4, down from over $14 earlier that month. In their memoranda supporting their recommendations for a systemic risk determination, FDIC and the Federal Reserve expressed concern that Citigroup soon would be unable to meet its funding obligations and expected deposit outflows. FDIC concluded that the government funding support otherwise available to Citigroup through the Federal Reserve’s lending programs such as the Commercial Paper Funding Facility and the Primary Dealer Credit Facility and TLGP provided the firm with short-term funding relief but would not be sufficient to help Citigroup withstand the large deposit outflows regulators expected if confidence in the firm continued to deteriorate. As was the case with Wachovia, Treasury, FDIC, and the Federal Reserve were concerned that the failure of a firm of Citigroup’s size and interconnectedness would have systemic implications. They determined that resolving the company’s insured institutions under the least-cost requirements likely would have imposed significant losses on Citigroup’s creditors and on uninsured depositors, thus threatening to further undermine confidence in the banking system. According to Treasury, a least-cost resolution would have led to investor concern about the direct exposures of other financial firms to Citigroup and the willingness of U.S. policymakers to support systemically important institutions, despite Treasury’s recent investments in Citigroup and other major U.S. banking institutions. In its recommendation to Treasury, the Federal Reserve listed the banking organizations with the largest direct exposures to Citigroup and estimated that the most exposed institution could suffer a loss equal to about 2.6 percent of its Tier 1 regulatory capital. Furthermore, Treasury, FDIC, and the Federal Reserve were concerned that a failure of Citigroup, which reported that it was well-capitalized (as did Wachovia at the time of the first systemic risk determination), could lead investors to reassess the riskiness of U.S. commercial banks more broadly. In comparison to Wachovia, Citigroup had a much larger international presence, including more than $500 billion of foreign deposits—compared to approximately $30 billion for Wachovia. Given Citigroup’s substantial international presence, imposing losses on uninsured foreign depositors under a least-cost framework could have intensified global liquidity pressures and increased funding pressures on other institutions with significant amounts of foreign deposits. For example, this could have caused investors to raise sharply their assessment of risks of investing in U.S. banking organizations, making raising capital and other funding more difficult. In addition to the potential serious adverse effects on credit markets, Treasury, FDIC, and the Federal Reserve expressed concern that a Citigroup failure could disrupt other markets in which Citigroup was a major participant. Citigroup participated in a large number of payment, settlement, and counterparty arrangements within and outside the United States. The Federal Reserve expressed concern that Citigroup’s inability to fulfill its obligations in these markets and systems could lead to widespread disruptions in payment and settlement systems worldwide, with important spillover effects back to U.S. institutions and other markets. Citigroup was a major player in a wide range of derivatives markets, both as a counterparty for over-the-counter trades and as a broker and clearing firm for trades on exchanges. If Citigroup had failed, many of the firm’s counterparties might have faced difficulties replacing existing contracts with Citigroup, particularly given concerns about counterparty credit risk at the time. Treasury, FDIC, and the Federal Reserve determined that FDIC assistance under the systemic risk exception, which would complement other U.S. federal assistance and TARP programs, would promote confidence in Citigroup. Specifically, they determined that if the systemic risk exception were invoked, FDIC could provide guarantees that would help protect Citigroup from outsize losses on certain assets and thus reduce investor uncertainty regarding the potential for additional losses to weaken Citigroup. In addition, such actions could help to reassure depositors and investors that the U.S. government would take necessary actions to stabilize systemically important U.S. banking institutions. On November 23, 2008, Treasury, FDIC, and the Federal Reserve announced a package of assistance to Citigroup, including a loss-sharing agreement on a fixed pool of Citigroup’s assets, to help restore confidence in the firm and maintain financial stability. By providing protection against large losses on these assets, regulators hoped to promote confidence among creditors and depositors providing liquidity to the firm to avert a least-cost resolution with potential systemic risk consequences. In particular, the loss-sharing agreement limited the potential losses Citigroup might suffer on a fixed pool of approximately $300 billion of loans and securities backed by residential and commercial real estate and other such assets. Under the final agreement executed on January 15, 2009, Citigroup agreed to absorb the first $39.5 billion in losses plus 10 percent of any remaining losses incurred. Ninety percent of covered asset losses exceeding $39.5 billion would be borne by Treasury and FDIC, with maximum guarantee payments capped at $5 billion and $10 billion, respectively. In addition, if all of these loss protections were exhausted, the Federal Reserve Bank of New York committed to allow Citigroup to obtain a nonrecourse loan equal to the aggregate value of the remaining covered asset pool, subject to a continuing 10 percent loss-sharing obligation of Citigroup. Citigroup issued FDIC and Treasury approximately $3 billion and $4 billion of preferred stock, respectively, for bearing the risk associated with the guarantees. The Federal Reserve loan, if extended, would have borne interest at the overnight index swap rate plus 300 basis points. Citigroup also received a $20 billion capital infusion from the TARP’s Targeted Investment Program, in addition to the initial $25 billion capital infusion received from TARP’s Capital Purchase Program in October 2008. In addition, the agreement subjected Citigroup to specific limitations on executive compensation and dividends during the loss share period. Isolating the impact of FDIC assistance to Citigroup is difficult, but according to Treasury and FDIC, the package of assistance provided by regulators may have helped to allow Citigroup to continue operating by encouraging private sector sources to continue to provide liquidity to Citigroup during the crisis. According to one FDIC official, one measure of success was that Citigroup could do business in Asia the business day following the announcement. With the package of assistance, regulators hoped to improve the confidence of creditors and certain depositors, facilitating Citigroup’s funding. Changes in the market’s pricing of Citigroup’s stock and its default risk, as measured by credit default swap spreads, indicate that the November 23, 2008, announcement boosted market confidence in the firm, at least temporarily. From a closing price of $3.77 on Friday, November 21, 2008, Citigroup’s common stock price rose 58 percent on Monday, November 24 and more than doubled by the end of the week. However, market confidence in Citigroup fell sharply again in early 2009 before the company’s stock price recovered and stabilized in spring 2009. On December 23, 2009, Citigroup announced that it had reached an agreement with FDIC, the Federal Reserve Bank of New York, and Treasury to terminate the loss-sharing and residual financing agreement. As part of the termination agreement, Citigroup agreed to pay a $50 million termination fee to the Federal Reserve. As of September 30, 2009, Citigroup reported that it had recognized $5.3 billion of losses on the pool of assets covered by the loss-sharing agreement. These losses did not reach the thresholds that would trigger payments by Treasury or FDIC. In July 2009, the Special Inspector General for the Troubled Asset Relief Program (SIGTARP) announced plans to audit the asset guarantees provided to Citigroup. According to SIGTARP, this audit is to examine why the guarantees were provided, how the guaranteed assets were structured, and whether Citigroup’s risk controls were adequate to prevent government losses. At the close of this review, the SIGTARP review was ongoing. While the systemic risk determinations and associated federal assistance may have helped to contain the crisis by mitigating potential systemic adverse effects, they may have induced moral hazard—encouraging market participants to expect similar emergency actions in future crises, thereby weakening their incentives to properly manage risks and also creating the perception that some firms are too big to fail. Federal assistance required for large and important institutions, such as non-bank holding companies, whose activities could affect the financial system, also highlighted gaps in the current regulatory regime, including inconsistent supervision and regulatory standards and lack of resolution authority for these institutions. Regulators, the Administration and Congress currently are considering financial regulatory reform proposals that could help address these concerns. Reforms that would enhance the supervision of financial institutions—particularly large financial holding companies— whose market discipline is likely to have been weakened by the recent exercises of the systemic risk exceptions are essential. While the federal assistance authorized by the systemic risk determinations may have helped to contain the financial crisis by mitigating potential adverse systemic effects that would have resulted from traditional FDIC assistance, they may have exacerbated moral hazard, particularly for large financial institutions. According to regulators and some economists, the expansion of deposit insurance under TAGP, in which most insured depository institutions of all sizes participated, could weaken incentives for newly protected, larger depositors to monitor their banks, and in turn banks may be more able to engage in riskier activities. According to some economists, the higher the deposit insurance guarantee, the greater the risk of moral hazard. In principle, deposit insurance helps prevent bank runs by small depositors, while lack of insurance encourages (presumably better informed) large depositors to protect their deposits by exerting discipline on risk taking by banks. Unlike the broad participation in TAGP, the majority of institutions that participate in DGP are large financial institutions. In addition, according to FDIC data, most of the senior unsecured debt under DGP has been issued by the largest U.S. financial institutions. Market observers with whom we spoke said that small banks did not participate in DGP as generally they primarily rely on deposits for funding. The bank debt guarantees, according to some economists, allow large financial institutions to issue debt without regard for differences in their risk profiles and can weaken the incentives for creditors to monitor bank performance and exert discipline against excessive risk taking for these institutions. In general, some economists said that to help mitigate moral hazard, it is important to specify when the extra deposit insurance and debt guarantee programs will end. Further, while recognizing that uncertainty about the duration of the crisis makes it difficult to specify timetables for phasing out guarantees, some economists said it is important to provide a credible “exit strategy” to prevent further disruption in the financial markets when withdrawing government guarantees. In addition, some economists noted that while government guarantees can be withdrawn once the crisis abates, a general perception may persist that a government guarantee always will be made available during a crisis—thus perpetuating the risk for moral hazard. Similarly, while the assistance to open banks authorized by the systemic risk determinations may have helped to contain the crisis by stabilizing the large and other financial institutions and mitigating potential systemic adverse effects, it also may have exacerbated moral hazard. According to regulators and market observers, assistance to open banks may weaken the incentives of large uninsured depositors, creditors, and investors to discipline large complex financial institutions deemed too big to fail. Federal Reserve Chairman Bernanke stated in March 2009 to the Council on Foreign Relations that the belief of market participants that a particular institution is considered too big to fail has many undesirable effects. He explained such perceptions reduce market discipline, encourage excessive risk taking by the firm, and provide artificial incentives for firms to grow. He also noted these beliefs can create an unlevel playing field, in which smaller firms may not be regarded as having implicit government support. Similarly, others have noted how such perceptions may encourage risk taking—for example, that large financial institutions are given access to the credit markets at favorable terms without consideration of the institutions’ risk profile because creditors and investors believe their credit exposure is reduced since they believe the government will not allow these firms to fail. Although regulators’ use of the systemic risk exception may weaken incentives of institutions to properly manage risk, the financial regulatory framework could serve an important role in restricting the extent to which they engage in excessive risk-taking activities as a result of weakened market discipline. Responding to the recent financial crisis, recent actions by the Federal Reserve as well as proposed regulatory reform and new FDIC resolution authority could help address concerns raised about the potential conduct ( to monitor and control risks) of institutions receiving federal assistance or subject to the systemic risk determinations. In an effort to mitigate moral hazard and weakened market discipline for large complex financial institutions including those that received federal assistance, regulators, the administration, and Congress are considering regulatory reforms to enhance supervision of these institutions. These institutions not only include large banks but also nonbank institutions. In the recent crisis, according to a testimony by FDIC Chairman Bair, bank holding companies and large nonbank affiliates have come to depend on the banks within their organizations as a source of funding. Bank holding companies must, under Federal Reserve regulations, serve as the source of strength for their insured institution subsidiaries. Subject to the limits of sections 23A and 23B of the Federal Reserve Act, however, bank holding companies and their nonbank affiliates may rely on the depository institution for funding. Also, according to regulators, institutions that were not bank holding companies (such as large thrift holding companies, investment banks, and insurance organizations) were responsible for a disproportionate share of the financial stress in the markets in the past 2 years and the lack of a consistent and coherent regulatory regime for these institutions helped mask problems until they were systemic and gaps in the regulatory regime constrained the government’s ability to deal with them once they emerged. Legislation has been proposed to create enhanced supervision and regulation for any systemically important financial institution, regardless of whether the institution owns an insured depository institution. The proposals would establish a council chaired by the Secretary of the Treasury with voting members comprising the chairs of the federal financial regulators which would oversee systemic risk and help identify systemically important companies. An institution could be designated “systemically important” if material financial distress at the firm could threaten financial stability or the economy. Systemically important institutions would be regulated by the Federal Reserve under enhanced supervisory and regulatory standards and stricter prudential standards. Regulators and market observers generally agree that these systemically important financial institutions should be subject to progressively tougher regulatory standards to hold adequate capital and liquidity buffers to reflect the heightened risk they pose to the financial system. They also generally agree that systemically important firms should face additional capital charges based both on their size and complexity. Such capital charges (and perhaps also restrictions on leverage and the imposition of risk-based insurance premiums on systemically important or weak insured depository institutions and risky activities) could help ensure that institutions bear the costs of growth and complexity that raise systemic concerns. Regulators and market observers believe that imposing systemic risk regulation and its associated safeguards will strengthen the ability of these firms to operate in stressed environments while the associated costs can provide incentives to firms to voluntarily take actions to reduce the risks they pose to the financial system. Under legislative proposals, these institutions also would be subject to a prompt corrective action regime that would require the firm or its supervisors to take corrective actions as the institutions’ regulatory capital level or other measures of financial strength declined, similar to the existing prompt corrective action regime for insured depository institutions under the FDI Act. Regulators also are considering regulatory reforms to improve the overall risk management practices of systemically important institutions. The Federal Reserve has proposed standards for compensation practices across all banking organizations it supervises to encourage prudent risk taking by creating incentives focusing on long-term rather than short-term performance. Regulators noted that compensation practices that create incentives for short-term gains may overwhelm the checks and balances meant to mitigate excessive risk taking. In its proposal for financial regulatory reform, Treasury recommended that systemically important financial institutions be expected to put in place risk management practices commensurate with the risk, complexity, and scope of their operations and be able to identify firmwide risk concentrations (such as credit, business lines, liquidity) and establish appropriate limits and controls around these concentrations. Also, under Treasury’s proposal, to measure and monitor risk concentrations, these institutions would be expected to be able to identify and report aggregate exposures quickly on a firmwide basis. Regulators also have indicated the need for measures to improve their oversight of risk management practices by these institutions. In our prior work on regulatory oversight of risk management at selected large institutions, we found that oversight of institutions’ risk management systems before the crisis illustrated some limitations of the current regulatory system. For example, regulators were not looking across groups of institutions to effectively identify risks to overall financial stability. In addition, primary, functional, and holding company regulators faced challenges aggregating certain risk exposures within large, complex financial institutions. According to testimony by a Federal Reserve official, the recent crisis highlighted the need for a more comprehensive and integrated assessment of activities throughout bank holding companies—a departure from the customary premise of functional regulation that risks within a diversified organization can be managed properly through supervision focused on individual subsidiaries within the firm. Accordingly, the bank supervisors, led by the Federal Reserve, recently completed the Supervisory Capital Assessment Program, which reflects some of the anticipated changes in the Federal Reserve’s approach to supervising the largest banking organizations. The Supervisory Capital Assessment Program involved aggregate analyses of the 19 largest bank holding companies, which according to Federal Reserve testimony, accounted for a majority of the assets and loans within the financial system. Bank supervisors evaluated on a consistent basis the expected performance of these firms under baseline and more-adverse-than- expected scenarios, drawing on individual firm information and using independently estimated outcomes. In addition, according to the agency’s officials, the Federal Reserve is creating an enhanced quantitative surveillance program for the largest and most complex firms, that will use supervisory information, firm-specific data analysis, and market-based indicators to identify emerging systemic risks as well as risks to specific firms. Some economists argue that a formal designation of systemically important institutions would have significant, negative competitive consequences for other firms and could encourage designated firms to take excessive risk because they would be perceived to be too big to fail. Instead some argue that a market stability regulator should be authorized to oversee all types of financial markets and all financial services firms, whether otherwise regulated or unregulated. Market observers also point out factors that complicate such determinations and make maintaining an accurate list of such institutions difficult. Aside from asset size and degree of leverage, they include degree of interconnectivity to other financial institutions, risks of activities in which they engage, nature of compensation practices, and degree of concentration of financial assets and activities. Moreover, maintaining a list would require regular monitoring in order to ensure the list was kept up to date, and some risky institutions would likely go unidentified, at least for a time. Such designation also would likely depend on factors outside the firm, such as economic and financial conditions. However, supervisors would presumably be doing much of the monitoring activity regardless of the existence of a public list, and they would have to establish standards, including assumptions regarding the economic and financial circumstances assumed when making such designations. It is important for Congress and regulators to subject systemically important institutions to stricter regulatory requirements and oversight in order to restrict excessive risk-taking activities as a result of weakened market discipline particularly after the use of federal assistance during the crisis to stabilize such institutions. The recent crisis also highlighted how a lack of a resolution authority for failing bank holding companies including those subject to the systemic risk determinations as well as nonbank financial firms such as Bear Stearns, Lehman Brothers, and American International Group, Inc. (AIG), complicated federal government responses. For example, regulators invoked the systemic risk exception to assist bank holding companies and savings and loan holding companies to prevent systemic disruptions in the financial markets and provided emergency funding to AIG, and in doing so potentially contributed to a weakening of incentives at these institutions and similarly situated large financial institutions to properly manage risks. According to regulators, the lack of a resolution authority for systemically important institutions also contributes to a belief by market participants that the government will not allow these institutions to fail and thereby weakens market discipline. Proposals for consideration by Congress include providing federal resolution authority for large financial holding companies deemed systemically important. One purpose of this authority would be to encourage greater market discipline and limit moral hazard by forcing market participants to realize the full costs of their decisions. In order to achieve its intended purpose, the use of a new resolution authority must be perceived by the market to be credible. The authority would need to provide for a regime to resolve systemically important institutions in an orderly manner when the stability of the financial system is threatened. As noted in our prior work, a regulatory system should have adequate safeguards that allow financial institution failures to occur while limiting taxpayers’ exposure to financial risk while minimizing moral hazard. Regulators and market observers generally agree that a credible resolution authority to resolve a distressed systemically important institution in an orderly manner would help to ensure that no bank or financial firm would be too big to fail. Such authority would encourage market discipline if it were to provide for the orderly allocation of losses to risk takers such as shareholders and unsecured creditors, and allow for the replacement of senior management. It also should help to maintain the liquidity and key activities of the organization so that the entity could be resolved in an orderly fashion without disrupting the functioning of the financial system. Unlike the statutory powers that exist for resolving insured depository institutions, the current bankruptcy framework available to resolve large complex nonbank financial entities and financial holding companies was not designed to protect the stability of the financial system. Without a mechanism to allow for an orderly resolution for a failure of a systemically important institution, failures of such firms could lead to a wider panic as indicated by the problems experienced after the failure of such large financial companies as Lehman Brothers, and near failures of Bear Stearns and AIG. Proposed new authority for resolution of a systemically important failing institution would provide for a receiver to resolve the institution in an orderly way. Government assistance such as loans, guarantees, or asset purchases would be available only if the institution is in government receivership. The receiver would have authority to operate the institution, enforce or repudiate its contracts, and pay its claims as well as remove senior management. In addition, shareholders and creditors to the firm would absorb first losses in the resolution. However, imposing losses on unsecured debt investors of large, interconnected, and systemically important firms might be inconsistent with maintaining financial stability during a crisis. In particular, faced with the potential failures of Wachovia and Citigroup, Treasury, FDIC, and the Federal Reserve concluded that the exercise of authority under the systemic risk exception was necessary because the failure of these firms would have imposed large losses on creditors and threatened to undermine confidence in the banking system. An effective resolution authority must properly balance the need to encourage market discipline with the need to maintain financial stability, in particular in a crisis scenario. One market observer argued that no such losses would be taken immediately by creditors because the objective of the resolution authority is to prevent a disorderly failure in which such creditors suffer immediate losses. Therefore an appropriate degree of flexibility would mean that some of an institution’s creditors might be protected, at least to some extent, against losses where doing so would be necessary to protect the stability of the financial system. Other features of the resolution authority would continue to promote market discipline even if some credit obligations were honored, because shareholders and senior management would still suffer losses. A regulatory official stated that the intertwining of functions among an institution’s affiliates can present significant issues when winding down the institution and recommended requirements that mandate greater functional autonomy of holding company affiliates. In addition, some economists and market observers also have recommended that regulators break up large institutions in resolution to limit a continuation of too-big-to-fail problems. That is, when a regulator assumes control of a troubled important financial institution, it should make reasonable efforts to break up the institution before returning it to private hands or to avoid selling it to another institution when the result would create a new systemically important institution. It is important for Congress and regulators to establish a credible resolution process to allow for an orderly resolution of a failed systemically important institution thereby helping to ensure that no bank or financial firm would be too big to fail. The recent financial crisis underscored how quickly liquidity can deteriorate at a financial institution. As a result, regulators’ deliberations about whether to invoke the systemic risk exception often occurred under severe time constraints. Treasury, FDIC, and the Federal Reserve collaborated prior to making announcements intended to reassure the markets, but the lack of a determination after two of these announcements of planned FDIC assistance under the systemic risk exception heightened the risk that such actions will be undertaken without appropriate transparency and accountability. Specifically, such an announcement signals regulators’ willingness to provide assistance and may give rise to moral hazard. However, in cases where Treasury does not make a determination, FDI Act requirements for communication and documentation do not apply. Therefore, when a determination is not made along with the announced actions, Congress cannot be assured that Treasury’s reasoning would be open to the same scrutiny required in connection with a formal systemic risk determination. Furthermore, uncertainty in these situations can arise because there is no requirement for Treasury to communicate that it will not make a systemic risk determination for an announced action. Our review of Treasury’s systemic risk determinations highlights that the announced FDIC actions were made to reduce strains on the deteriorating markets and to promote confidence and stability in the banking system. Regarding the systemic risk determinations, the regulators concluded that resolving the depository institutions at issue under the traditional least- cost approach would have worsened adverse conditions in the economy and in the financial system. While it is difficult to isolate the impact of those actions from other government assistance, the actions seem to have reassured investors and depositors at the particular banks and encouraged them to continue to provide liquidity, thereby allowing the banks to keep operating. In the case of TLGP, some regulators and market observers have attributed short-term benefits to FDIC guarantees on certain debt obligations, citing improved cost and availability of credit for many institutions. However, with respect to TLGP determination, some have noted that under a possible reading of the systemic risk exception, the statute may authorize assistance only to particular institutions based on those institutions’ specific problems, not systemic risk assistance based on problems affecting the banking industry as a whole. Treasury, FDIC, and the Federal Reserve considered this and other legal issues in recommending and making TLGP determination. The agencies believe the statute could have been drafted more clearly and that it can be interpreted in different ways. They concluded, however, that under a permissible interpretation, assistance may be based on industry-wide concerns. They also concluded that a systemic risk determination waives all of the normal statutory restrictions on FDIC assistance and then creates new authority to provide assistance, both as to the types of aid that may be provided and the entities that may receive it. Under this reading, the agencies believe the statutory criteria were met in the case of TLGP and that the assistance was authorized. We examined these issues as part of our review of the basis of the systemic risk determinations made to date. We found there is some support for the agencies’ position that the exception authorized systemic risk assistance of some type under TLGP facts, as well as for their position that the exception permits assistance to the entities covered by this program. There are a number of questions concerning these interpretations, however. For example, the agencies agree that some of the statutory provisions are ambiguous. Because application of the systemic risk exception raises novel legal and policy issues of significant public interest and importance, and because of the need for clear direction to the agencies in a time of financial crisis, the requirements and assistance authorized under the systemic risk exception may require clarification by Congress. Systemic risk assistance also raises long-term concerns about moral hazard and weakened market discipline, particularly for large complex financial institutions. This involves a trade-off between the short-term benefits to markets, the economy, and business and households of federal action and the long-term effects of any federal action on market discipline. While the financial regulatory framework can serve an important role in restricting excessive risk-taking activities as a result of weakened market discipline, the financial crisis revealed limitations in this framework. In particular, these limitations include inconsistent oversight of large financial holding companies (bank versus nonbank). Another limitation was a weakness in the risk management practices of these companies. Legislators and regulators currently are considering regulatory proposals to subject systemically important institutions, including those whose market discipline is likely to have been weakened by the recent exercises of the systemic risk exception, to stricter regulatory standards such as higher capital and stronger liquidity and risk management requirements. Furthermore, according to regulators, the lack of resolution authority for systemically important institutions contributes to a belief by market participants that the government will not allow these institutions to fail and thereby weakens incentives for market participants to monitor the risks posed by these institutions. Legislation has been proposed to expand resolution authority to large financial holding companies deemed systemically important that is intended to impose greater market discipline and limit moral hazard by forcing market participants to face significant costs from their risk-taking decisions. It is important for the use of a new resolution authority to be perceived by the market to be credible for it to help achieve the intended effects. To help ensure transparency and accountability in situations where FDIC, the Federal Reserve, and Treasury publicly announce intended emergency actions but Treasury does not make a systemic risk determination required to implement them, Congress should consider requiring Treasury to document and communicate to Congress the reasoning behind delaying or not making a determination. Recent application of the systemic risk exception raises novel legal and policy issues, including whether the exception may be invoked based only on the problems of particular institutions or also based on problems of the banking industry as a whole, and whether and under what circumstances assistance can be provided to “healthy” institutions, bank holding companies, and other bank affiliates. Because these issues are of significant public interest and importance, as Congress debates the modernization and reform of the financial regulatory system, Congress should consider enacting legislation clarifying the requirements and assistance authorized under the systemic risk exception. Enacting more explicit legislation would provide legal clarity to the banking industry and financial community at large, as well as helping to ensure ultimate accountability to taxpayers. As Congress contemplates reforming the financial regulatory system, Congress should ensure that systemically important institutions receive greater regulatory oversight. This could include such things as more consistent and enhanced supervision of systemically important institutions and other regulatory measures, such as higher capital requirements and stronger liquidity and risk management requirements and a resolution authority for systemically important institutions to mitigate risks to financial stability. We provided a draft of this report to the Federal Reserve, FDIC and Treasury for their review and comment. The Federal Reserve and Treasury provided us with written comments. These comments are summarized below and reprinted in appendixes III and IV, respectively. FDIC did not provide written comments. We also received technical comments from the Federal Reserve, FDIC, and Treasury that we have incorporated in the report where appropriate. In its comments, the Federal Reserve agreed with our findings that while the agencies’ actions taken under the systemic risk exception were important components of the response by the government to the financial crisis, these actions have the potential to increase moral hazard and reinforce perceptions that some firms are too big to fail. In order to mitigate too big to fail and risks to financial stability, the Federal Reserve stated that it agrees with our matter for Congressional consideration that all systemically important financial institutions be subject to stronger regulatory and supervisory oversight and that a resolution system be put in place that would allow the government to manage the failure of these firms in an orderly manner. Treasury also commented that it agreed with our findings and our matter for Congressional consideration for greater regulatory oversight of the largest, most interconnected financial firms and resolution authority to wind down failing nonbank financial firms in a manner that mitigates the risks that their failure would pose to financial stability and the economy. We are sending copies of this report to the Chairman of the Board of Governors of the Federal Reserve System; the Chairman of FDIC; the Secretary of the Treasury; and other interested parties. In addition, the report will be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staff have any questions regarding this report, please contact me at (202) 512-8678 or williamso@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix V. To describe the steps taken by the Federal Deposit Insurance Corporation (FDIC) and the Board of Governors of the Federal Reserve System (Federal Reserve) to make the recommendations and the Department of the Treasury (Treasury) to make determinations in some cases, we reviewed documentation of recommendations that FDIC and the Federal Reserve made for Wachovia, the Temporary Liquidity Guarantee Program (TLGP), Citigroup, Bank of America, and the Public-Private Investment Programs proposed Legacy Loans Program (LLP) as well as documentation of Treasury’s determination for Wachovia, TLGP, and Citigroup. We also reviewed press releases by the agencies announcing the respective intended actions. In addition, to gain an understanding of how the agencies collaborated prior to the announcements of emergency actions, we interviewed officials from Treasury, FDIC, and the Federal Reserve. We also spoke with these officials about the status of emergency actions that were announced, but did not result in a systemic risk determination by Treasury. Finally, we reviewed the Federal Deposit Insurance Act (FDI Act) requirements for transparency and accountability with respect to the use of the systemic risk provision and analyzed the implications of announcements that are not followed by a Treasury determination that would trigger these requirements. To describe the basis for each determination and the purpose of actions taken pursuant to each determination we reviewed and analyzed documentation of Treasury’s systemic risk determinations and the supporting recommendations that FDIC and the Federal Reserve made for Wachovia, TLGP, and Citigroup. We interviewed officials from Treasury, FDIC, the Federal Reserve, and the Office of the Comptroller of the Currency (OCC) to gain an understanding of the basis and authority for each determination and the purpose of the actions taken under each determination. We also interviewed three economists, one banking industry association, and a banking analyst. In addition, we collected and analyzed various data to illustrate financial and economic conditions at the time of each determination and the actions taken pursuant to each determination. We examined whether the legal requirements for making the systemic risk determination with respect to TLGP were met and whether the assistance provided under that program was authorized under the systemic risk exception. For this legal analysis, we reviewed and analyzed the FDI Act, its legislative history including the Federal Deposit Insurance Corporation Improvement Act (FDICIA), and other relevant legislation. We reviewed FDIC regulations and policy statements as well as written background material prepared by the agencies. We obtained the legal views of Treasury, FDIC, and the Federal Reserve on the agencies’ legal authority to establish TLGP, and also obtained the views of banking law specialists in private practice and academia on these issues. In describing the likely effects of each determination on the incentives and conduct of insured depository institutions and uninsured depositors, as well as assessing proposals to mitigate moral hazard created by such federal assistance, we reviewed and analyzed the research reports of one credit rating agency, Congressional testimonies of regulators and market observers, proposed legislation, and academic studies. We interviewed officials from Treasury, FDIC, the Federal Reserve, and OCC as well as one academic, and three market observers to gain an understanding of how each determination and action impact the incentives and conduct of insured depository institution and uninsured depositors. Finally, we reviewed prior GAO work on the financial regulatory system. We conducted this performance audit from October 2008 to April 2010 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Analysis of Legal Authority for the Temporary Liquidity Guarantee Program (TLGP) Introduction and Summary of Conclusions As part of our review of the basis of the systemic risk determinations made to date under the Federal Deposit Insurance Act’s (“FDI Act”) systemic risk exception, we examined whether the legal requirements for making such determinations were met with respect to the Temporary Liquidity Guarantee Program (“TLGP”) and whether the assistance provided under that program was authorized under the exception. We note that the recent financial crisis is the first time that Treasury, FDIC, and the Federal Reserve (“the agencies”) have relied on the exception since its enactment as part of the Federal Deposit Insurance Corporation Improvement Act (“FDICIA”) in 1991, and that no court to date has ruled on when or how the exception may be used. We also acknowledge the volatile economic circumstances under which the agencies created TLGP. The agencies believe that while the statute could have been drafted more clearly and that it can be interpreted in different ways, under a permissible interpretation, a systemic risk determination may be based on adverse circumstances affecting the banking industry as a whole—the situation that prompted creation of TLGP—as well as on adverse circumstances of one or more particular banking institutions. The agencies also believe a systemic risk determination waives all of the normal statutory restrictions on FDIC assistance, as well as creating new authority to provide assistance, both as to the types of aid that may be provided and the entities that may receive it. Under this reading, the agencies believe that the statutory criteria were met in the case of TLGP and that the assistance was authorized. We agree there is some support for the agencies’ position that the statute authorizes systemic risk assistance of some type under TLGP facts, as well as for their position that the exception permits assistance to the entities covered by TLGP. There are a number of questions concerning these interpretations, however. In the agencies’ view, for example, some of the statutory provisions are ambiguous. What is clear, however, is that the systemic risk exception overrides important statutory restrictions designed to minimize costs to the Deposit Insurance Fund, and, in the case of TLGP, that the agencies used it to create a broad-based program of direct FDIC assistance to institutions that had never before received such relief—”healthy” banks, bank holding companies, and other bank affiliates. Because these novel legal issues are matters of significant public interest and importance, and because of the need for clear direction to the agencies in a time of financial crisis, we recommend that Congress consider enacting legislation clarifying the requirements and the assistance authorized under the exception. Background on FDIC’s Statutory Authority to Use the Deposit Insurance Fund As described in greater detail in this report, TLGP provided direct assistance, backed by FDIC’s Deposit Insurance Fund, both to insured depository institutions and to their holding companies and other bank affiliates. The FDI Act normally permits Deposit Insurance Fund- supported assistance only to insured depository institutions, however, and allows assistance to operating (“open”) insured depository institutions (so- called “open bank assistance”) only in three situations, and then only by certain means. As specified in section 13(c) of the FDI Act: “(c) Assistance to insured depository institutions “(1) is authorized . . . to make loans to, to make deposits in, to purchase the assets or securities of, to assume the liabilities of, or to make contributions to, any insured depository institution— “(A) if such action is taken to prevent the default of such “(B) if, with respect to an insured bank in default, such action is taken to restore such insured bank to normal operation; or “(C) if, when severe financial conditions exist which threaten the stability of a significant number of insured depository institutions or of insured depository institutions possessing significant financial resources, such action is taken in order to lessen the risk to the posed by such insured depository institution under such threat of instability.” 12 U.S.C. § 1823(c)(1)(A)-(C). The FDI Act contains a number of additional restrictions on when and how the FDIC may use Deposit Insurance Fund monies: First, before FDIC may provide assistance to an open bank, FDI Act section 13(c)(8) normally requires it to make a formal determination that the bank is in “troubled condition” under specific undercapitalization and other criteria and that the bank meets other requirements. FDIC must publish notice of any such determination in the Federal Register. Second, if FDIC decides to provide assistance, the assistance normally must meet so-called “least-cost requirements” in FDI Act section 13(c)(4). Section 13(c)(4)(A) requires FDIC to determine, using financial data about a specific institution, that the proposed assistance is necessary to meet FDIC’s deposit-insurance obligations with respect to the institution’s insured deposits and is the least costly of all possible methods of meeting those obligations. Section 13(c)(4)(E) prohibits FDIC from providing assistance to creditors or non-insured depositors of the institution if doing so would increase losses to the Fund beyond those that otherwise might result from protecting insured depositors. Third, FDI Act section 11(a)(4)(C) normally prohibits FDIC from using the Fund to benefit affiliates or shareholders of an assisted depository institution in any way, regardless of whether such assistance would cause a loss to the Fund. The agencies believe, however, that if Treasury makes an emergency determination under the systemic risk exception, this waives all of the foregoing requirements and also creates new authority to provide any type of assistance to any type of entity, as long as the assistance is deemed necessary to avoid or mitigate systemic risk. Specifically, the words of the statute require Treasury to determine: (i) that “compliance with subparagraphs (A) and (E) with respect to an insured depository institution would have serious adverse effects on economic conditions or financial stability”; and (ii) that “any action or assistance under . . . would avoid or mitigate” such effects. If Treasury makes this determination, the FDIC may then “take other action or provide assistance under this section as necessary to avoid or mitigate such effects.” The statute imposes significant deliberative and consultative requirements on the process for making such a determination: it must be made by the Secretary of the Treasury; the Secretary must receive written recommendations from both the FDIC Board of Directors and the Federal Reserve Board of Governors, each made pursuant to at least a two-thirds vote; and the Secretary must consult with the President. The Agencies’ Reliance on the Systemic Risk Exception to Create TLGP The agencies agree that without Treasury’s systemic risk determination for TLGP in October 2008, the above statutory restrictions would have prohibited FDIC assistance to most of TLGP recipients, either because the entities would not have met the statutory “troubled condition criteria” for open banks (in the case of many TLGP participants) or because they were bank holding companies or other affiliates of insured depository institutions for which FDIC assistance normally is unavailable. The agencies clearly followed the requisite process in issuing the determination: FDIC and the Federal Reserve both submitted unanimous written recommendations in favor of Treasury making a systemic risk determination; the Secretary consulted with the President; and the Secretary signed a formal determination on October 14, 2008. Acknowledging that the systemic risk exception can be interpreted in different ways, the agencies believe they also met the statute’s substantive requirements under a permissible interpretation of the statute. We discuss below two key legal issues that the agencies considered in making their recommendations and determination. 1. Authority to Provide Assistance Based on Problems of the Banking Industry As a Whole In invoking the exception for the other two systemic risk determinations made to date—with respect to Wachovia in September 2008 and Citigroup in January 2009—the agencies concluded, based on the facts of those specific institutions, that providing least-cost assistance to those entities’ insured institutions would have had “serious adverse effects on economic conditions or financial stability”—that is, would have caused systemic risk. The agencies applied the exception differently for TLGP determination: they made what they characterized as a “generic systemic risk determination” made “generically for all institutions,” that is, a determination made with respect to “the U.S. banking system in general” and “insured depository institutions in general.” According to the agencies, they took this approach because the problem at hand was not limited to weakness in one or more individual institutions, but was a banking system problem, an overall scarcity of liquidity caused by lack of interaction among institutions. The agencies therefore concluded that providing assistance on a bank-by-bank basis would not have relieved the existing instability in the industry and that waiting to provide bank-by- bank relief after individual banks had begun to fail would not have mitigated further systemic risk. The agencies also believed that a bank- specific, wait-for-failure approach would have been more costly than TLGP assistance provided. The agencies believe these facts supported the required statutory finding that “compliance with with respect to an insured depository institution” would cause systemic risk, in that they showed that having to apply the least-cost requirements on a bank-by-bank basis (“compliance with the . . . requirements with respect to an . . . institution”) would have caused systemic risk. The agencies also believe the statute permits a generic rather than a bank-specific determination because under general statutory construction and grammar rules reflected in the Dictionary Act, 1 U.S.C. § 1, the required finding regarding compliance “with respect to an institution” can be read as “compliance with respect to one or more institutions” unless statutory context indicates otherwise. Some have noted that a possible reading of the exception authorizes assistance only to particular institutions, based on those institutions’ specific problems, not, as was done in creating TLGP, systemic risk assistance to all institutions based on problems affecting the banking industry in general. In our view, the language, context, and history of the exception do not clearly restrict its use to assistance to specific institutions. The statute does not prescribe a detailed method by which Treasury must determine whether “compliance . . . would cause systemic risk,” and we agree with the agencies that “compliance . . . with respect to an institution” means the determination can be based on the circumstances of more than one bank—that is, “an institution” can mean “one or more institutions.” As to whether the statute permits a determination to be made generically based on industry-wide problems at insured depository institutions apart from the health of any particular institution, nothing in the legislative history of the exception explicitly refutes the agencies’ position that the statute permits such a generic determination. The debate leading to enactment of FDICIA centered on FDIC’s role in resolving “too big to fail” institutions whose collapse might pose a risk to the entire financial system, rather than on banking system-wide problems already posing serious risk. However, nothing indicates Congress intended to preclude use of the exception when, as with the facts leading to TLGP, an adverse systemic condition is itself the cause of imminent bank failures and the agencies determine that individual least-cost resolutions would not adequately address the condition and in fact would worsen it. While Congress’ intent to further restrict the FDIC’s authority to provide open bank assistance is clear from the significant new limitations it imposed in FDICIA, Congress’ simultaneous enactment of the systemic risk exception indicated a parallel objective to avoid wholesale systemic failure. In light of these objectives and the language of the statute, we believe there is some support for the agencies’ position that the law does not require an institution-specific evaluation where it would result in systemic risk. Unexcelled Chemical Corp. v. United States, 345 U.S. 59 (1953) (laws written in comprehensive terms apply to unanticipated circumstances if they reasonably fall within the scope of the statutory language); see generally GAO-08-606R (March 31, 2008) at 13-18. In sum, given Treasury’s factual determination that systemic risk would have resulted from application of the least-cost requirements to the circumstances leading to creation of TLGP, we believe there is some support for the agencies’ legal position that systemic risk assistance of some type was authorized. Whether the particular TLGP assistance provided was within the scope authorized by the systemic risk exception was a second issue the agencies considered, and which we now address. 2. Authority to Provide Assistance to Non-”Troubled” Banks, Bank Holding Companies, and Other Bank Affiliates In addition to considering whether the banking industry-wide liquidity crisis could be mitigated by providing systemic risk relief, the agencies considered whether the statute authorized relief for all of the entities they believed should receive assistance. The agencies addressed whether the language of the statute—authorizing FDIC, in the event of a systemic risk determination, to “take other action or provide assistance under this section as necessary to avoid or mitigate” systemic risk—waived the statute’s other restrictions, and they concluded that it both waived the restrictions and then gave FDIC new authority to provide assistance even beyond that otherwise authorized by the FDI Act, as long as the assistance was “necessary to avoid or mitigate” systemic risk. Under this interpretation, the agencies believe FDIC had authority to provide TLGP assistance directly to bank holding companies and other bank affiliates, as well as to insured depository institutions that FDIC had not determined met the statutory troubled-condition criteria (and would not have met them in most cases because most of the institutions were “healthy” under the statutory standards). The agencies base their interpretation in part on Congress’ use of the disjunctive term “or” in authorizing “other action or . . . assistance under this section,” noting that “or” generally indicates an intention to differentiate between two phrases. They also rely on a statutory construction principle known as the “grammatical rule of the last antecedent,” where a limiting phrase—here, “under this section”— generally should be read as modifying only the words immediately preceding it—here, “assistance,” not “other action.” In the agencies’ view, a systemic risk determination creates two distinct options for assistance: (1) “other action,” that is, “action” “other” than assistance allowed by FDI Act section 13; and (2) assistance allowed by section 13. “Other action” is not subject to the restrictions on section 13 assistance, in the agencies’ view, because by definition it is not section 13 assistance, while “assistance under this section” remains subject to those restrictions unless explicitly waived by the systemic risk exception, as in the case of the least- cost requirements. Under this interpretation, TLGP’s aid to all open healthy (non-”troubled”) banks, considered as a whole, was authorized because it constituted “other action” not subject to the section 13(c)(8) ban on relief to healthy open banks, rather than “assistance under this section” which would have prohibited relief to the same institutions if considered individually. Likewise, according to the agencies, TLGP’s direct assistance to bank holding companies and other bank affiliates constituted “other action” rather than “assistance under this section.” The agencies’ reading of the statute raises several issues. First, while rules of grammar and statutory construction can provide general guidance about what Congress intended, the actual context and structure of the statute are of equal if not paramount importance. Here, Congress’ interchangeable use of the terms “action” and “assistance” throughout section 13 suggests it did not intend to differentiate between those terms when it used them in section 13(c)(4)(G), the systemic risk exception. For example, although Congress entitled section 13(c), the general FDI Act provision authorizing FDIC aid to open insured institutions, as “Assistance to insured depository institutions,” it then used the term “action” to identify each circumstance in which such assistance is authorized. See, e.g., 12 U.S.C. §§ 1823(c)(1)(A)-(C), 1823(c)(2), 1823(c)(4)(E). This suggests Congress created only one basic option for systemic risk relief: action or assistance, authorized by section 13 and related restrictions, except restrictions expressly waived by the systemic risk exception. The sequence of the terms “other action” and “assistance” in the statute supports this reading, because if Congress intended to create two types of relief—one subject to the section 13 restrictions and the other subject to no restrictions—arguably it would have reversed the order and authorized “assistance under this section or other action” rather than “other action or assistance under this section.” Second, the fact that the systemic risk exception explicitly waives the least-cost requirements, by two specific references to “subparagraphs (A) and (E),” but waives none of the other statutory requirements, also supports the one-option interpretation because it suggests Congress did not intend its authorization of “other action” to override other statutory restrictions. In this regard, FDIC has long recognized, since promulgation of its revised Open Bank Assistance Policy in 1992, that the section 13(c)(8) restrictions against assistance to healthy open banks apply even when there is a systemic risk determination and that these restrictions must be met prior to providing systemic risk assistance. FDIC thus applied the restrictions as part of its recommendation for the Citigroup systemic risk determination in January 2009, and determined that the open insured depository institutions there were “troubled,” thus qualifying for open bank—and systemic risk—assistance. In FDIC’s view, its position that the Citigroup systemic risk determination did not waive (c)(8) for open depository institutions is consistent with its position that TLGP determination did waive (c)(8) for open depository institutions, because the former constituted “assistance under this section” relief while the latter constituted “other action” relief. FDIC’s 1992 Open Bank Assistance Policy did not address this aspect of the systemic risk exception, FDIC told us, because until TLGP, no one had considered t possibility of systemic risk stemming from industry-wide conditions rathehe r than bank-specific conditions. We recognize that FDIC’s interpretation has evolved in response to new circumstances, but we believe its current and arguably inconsistent “other action” interpretation is subject to questionfor the reasons noted above. Third, the practical effect of a systemic risk determination under the agencies’ reading is to authorize any type of assistance to any type of entity, provided the aid is deemed necessary to avoid or mitigate systemic risk. This is because if relief does not meet the restrictions imposed on “assistance under this section,” the identical relief is by definition authorized as “other action.” If Congress had intended to give FDIC such broad new authority, however, it could have simply said so, authorizing FDIC to “take action” in the event of systemic risk. Instead, Congress added qualifying language apparently intended to limit FDIC’s options, only authorizing it to “take other action or provide assistance under this section.” Finally, the overall legislative history of FDICIA also suggests Congress did not intend the exception to provide the breadth of new authority claimed by the agencies. FDICIA was aimed in part at curbing what Congress believed had been excessive costs of FDIC bank assistance that increased the exposure of the Deposit Insurance Fund. Congress therefore imposed new restrictions intended to raise, not lower, the bar for FDIC relief. Limits were added, for example, on which entities could receive assistance (e.g., only open banks in “troubled condition”) and how much assistance could be provided (least-cost). Congress also imposed so-called prompt corrective action mandates on the banking regulators, requiring them to take increasingly severe actions as an institution’s capital deteriorates. Additionally, like its predecessor exception, the systemic risk exception was enacted as part of a provision imposing cost-based limits on FDIC assistance—the least-cost requirements—rather than as a separate provision granting new authority. In light of FDICIA’s overarching remedial purposes, it is questionable that Congress would have intended to simultaneously provide FDIC with new and substantially broader authority than the agency had been given since its creation in 1933, and would have done so by means of an implication in a narrowed exception to a cost restriction. Commissioner v. Clark, 489 U.S. 726, 738- 39 (1989)(“Given that Congress has enacted a general rule . . ., we should not eviscerate that legislative judgment through an expansive reading of a somewhat ambiguous exception.”); Whitman v. American Trucking Ass’n, 531 U.S. 457, 468 (2001)(“Congress, we have held, does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions—it does not, one might say, hide elephants in mouseholes.”) (citations omitted). In response to these issues, the agencies make two additional points. First, they suggest that any uncertainty regarding whether “other action” authorized TLGP assistance to bank holding companies was resolved by 2009 amendments to the systemic risk exception. At the time TLGP was created, the exception required FDIC to recover any losses to the Deposit Insurance Fund caused by its systemic risk assistance, but authorized recovery only from insured depository institutions. In response to concerns by FDIC and banking industry representatives that bank holding companies should also bear some of TLGP costs because they had received substantial assistance under the program, Congress modified the provision in May 2009 to permit assessments against bank holding companies as well as depository institutions. Pub. L. No. 111-22, sec. 204(d), codified at 12 U.S.C. § 1823(c)(4)(G)(ii). The agencies believe this confirms the FDIC’s authority to provide assistance to bank holding companies under TLGP because Congress did not simultaneously amend the exception to explicitly prohibit such assistance going forward. We agree the amendment provides some support for the agencies’ position under a general tenet of statutory construction that congressional awareness of an agency’s practice in implementing a statute, without striking down that practice, indicates congressional acquiescence in the agency’s interpretation. Second, the agencies maintain that their interpretation of any ambiguous aspects of the systemic risk exception warrants substantial deference under the Supreme Court’s decision in Chevron U.S.A. v. Natural Resources Defense Council, 467 U.S. 837 (1984), and related cases. Under Chevron, when the meaning of a statute is unclear, either because the statute is silent on an issue or the language is ambiguous, an interpretation by an agency charged with the statute’s administration warrants substantial deference provided the interpretation is reasonable, even if it is not the only interpretation or the best interpretation. Whether and to what extent deference is warranted depends on factors including the agency’s specialized expertise in implementing the statute, whether the agency’s interpretation has been subjected to public scrutiny through public notice- and-comment rulemaking, and whether its interpretation is consistent with its previous pronouncements. United States v. Mead Corp., 533 U.S. 218, 227-29 (2001) (citations omitted). Under Mead, Chevron deference is warranted where the interpretation is made as part of an agency rulemaking or other agency action that Congress intended to carry the force of law, and, even if Chevron deference is not warranted, lesser deference is warranted under Skidmore v. Swift, 323 U.S. 134 (1944), if the agency’s interpretation is “persuasive” based on factors such as the thoroughness and validity of the agency’s reasoning, the consistency of its interpretation over time, and the formality of its action. We believe these deference principles have some force as applied to the systemic risk exception and TLGP. Congress did not explicitly address whether FDIC may provide systemic risk relief directly to bank holding companies or healthy open banks, and a court arguably could find that the statute’s authorization of “other action or assistance under this section” is ambiguous. If it did, we believe the agencies’ reading might merit at least some degree of deference. Congress charged these three financial regulatory agencies with implementing the systemic risk exception, and charged FDIC with implementing other provisions of the FDI Act related to this exception. The agencies interpreted the exception to authorize assistance to holding companies, other bank affiliates, and non-troubled banks as part of the systemic risk determination, and FDIC exercised its general rulemaking authority to issue regulations establishing TLGP, including regulations providing for assistance to these entities. According to the agencies, their interpretation of what “other action or . . . assistance” authorizes was necessarily part of the rulemaking because critical aspects of the program—assistance to “healthy” banks, and to bank holding companies and other bank affiliates—were premised upon this interpretation and would otherwise have been prohibited. Further, FDIC’s rulemaking preambles asserted that TLGP was authorized by Treasury’s systemic risk determination. The fact that the regulations and preambles did not solicit public comment on the underlying legal interpretations—and in fact did not indicate what the interpretations were—did not disqualify them from Chevron deference, according to the agencies, because under other Supreme Court precedent, an agency’s interpretation of a statute may warrant deference if the interpretation was the only logical basis for a rulemaking, even if the agency does not disclose its interpretation. Finally, we note that the very process Congress established for issuance of systemic risk determinations reflects great congressional respect for the agencies’ judgment and expertise, if not a strict basis for legal deference to their interpretation of the statute. We nonetheless believe the arguments for deference to the agencies’ interpretation are undercut by the statutory interpretation concerns discussed above, which raise questions about the persuasiveness of the agencies’ arguments, and by the different and arguably inconsistent positions taken by FDIC regarding whether the systemic risk exception waives the prohibition against assistance to “healthy” institutions. We believe there is some support for the agencies’ position that the systemic risk exception authorizes assistance of some type under TLGP facts, as well as for their position that the exception permits assistance to the entities covered by this program. There are a number of questions concerning these interpretations, however. Because application of the systemic risk exception raises novel legal and policy issues of significant public interest and importance, and because of the need for clear direction to the agencies in a time of financial crisis, we recommend that Congress consider enacting legislation clarifying the requirements and assistance authorized under the exception. Congress is now debating modernization and reform of the financial regulatory system, including regulation that addresses systemic risk, and this may provide an opportunity for such congressional consideration. Enacting more explicit legislation will provide legal clarity to the agencies, the banking industry, and the financial community at large, and will help to ensure greater transparency and accountability to the taxpaying public. Federal Deposit Insurance Act Section 13(c)(4)(G), Title 12, United States Code, Section 1823(c)(4)(G) § 1823. Corporation monies * * * (c) Assistance to insured depository institutions * * * (4) Least-cost resolution required * * * (G) Systemic risk (i) Emergency determination by Secretary of the Treasury Notwithstanding subparagraphs (A) and (E) , if, upon the written recommendation of the Board of Directors (upon a vote of not less than two-thirds of the members . . .) and the Board of Governors of the Federal Reserve System (upon a vote of not less than two-thirds of the members . . . ), the Secretary of the Treasury (in consultation with the President) determines that— (I) the Corporation’s compliance with subparagraphs (A) and (E) with respect to an insured depository institution would have serious adverse effects on economic conditions or financial stability; and (II) any action or assistance under this subparagraph would avoid or mitigate such adverse effects, the Corporation may take other action or provide assistance under this section as necessary to avoid or mitigate such effects. (ii) Repayment of loss (I) The Corporation shall recover the loss to the Deposit Insurance Fund arising from any action taken or assistance provided with respect to an insured depository institution under clause (i) from 1 or more special assessments on insured depository institutions, depository institution holding companies (with the concurrence of the Secretary of the Treasury with respect to holding companies), or both, as the Corporation determines to be appropriate. (II) Treatment of depository institution holding companies For purposes of this clause, sections 1817(c)(2) and 1828(h) of this title shall apply to depository institution holding companies as if they were insured depository institutions. The Corporation shall prescribe such regulations as it deems necessary to implement this clause. In prescribing such regulations, defining terms, and setting the appropriate assessment rate or rates, the Corporation shall establish rates sufficient to cover the losses incurred as a result of the actions of the Corporation under clause (i) and shall consider: the types of entities that benefit from any action taken or assistance provided under this subparagraph; economic conditions, the effects on the industry, and such other factors as the Corporation deems appropriate and relevant to the action taken or the assistance provided. Any funds so collected that exceed actual losses shall be placed in the Deposit Insurance Fund. (iii) Documentation required The Secretary of the Treasury shall— (I) document any determination under clause (i); and (II) retain the documentation for review under clause (iv). (iv) GAO review The Comptroller General of the United States shall review and report to the Congress on any determination under clause (i), including— (I) the basis for the determination; (II) the purpose for which any action was taken pursuant to such clause; and (III) the likely effect of the determination and such action on the incentives and conduct of insured depository institutions and uninsured depositors. (v) Notice (I) The Secretary of the Treasury shall provide written notice of any determination under clause (i) to the Committee on Banking, Housing, and Urban Affairs of the Senate and the Committee on Banking, Finance and Urban Affairs of the House of Representatives. (II) Description of basis of determination The notice under subclause (I) shall include a description of the basis for any determination under clause (i). In addition to the contacts named above, Karen Tremba (Assistant Director), Rachel DeMarcus, John Fisher, Kristopher Hartley, Michael Hoffman, Marc Molino, Akiko Ohnuma, Barbara Roesmann, Carla Rojas, Susan Sawtelle, and Paul Thompson made key contributions to this report.

In 2008 and 2009, the Federal Deposit Insurance Corporation (FDIC) provided emergency assistance that required the Secretary of the Department of the Treasury (Treasury) to make a determination of systemic risk under the systemic risk exception of the Federal Deposit Insurance Act (FDI Act). The FDI Act requires GAO to review each determination made. For the three determinations made to date, this report examines (1) steps taken by FDIC, the Board of Governors of the Federal Reserve System (Federal Reserve), and Treasury to invoke the exception; (2) the basis of the determination and the purpose of resulting actions; and (3) the likely effects of the determination on the incentives and conduct of insured depository institutions and uninsured depositors. To do this work, GAO reviewed agency documentation, relevant laws, and academic studies; and interviewed regulators and market participants. Treasury, FDIC, and the Federal Reserve collaborated before the announcement of five potential emergency actions that would require a systemic risk determination. In each case, FDIC and the Federal Reserve recommended such actions to Treasury, but Treasury made a determination on only three of the announced actions. Although two recommendations have not resulted in FDIC actions to date, their announcement alone could have created the intended effect of increasing confidence in institutions, while similarly generating negative effects such as moral hazard. However, because announcements without a determination do not trigger FDI Act requirements for documentation and communication, such as Treasury consultation with the President and notification to Congress, such de facto determinations heightened the risk that the decisions were made without the level of transparency and accountability intended by Congress. Further, uncertainties can arise because there is no requirement for Treasury to communicate that it will not be invoking a systemic risk determination for an announced action. Two of Treasury's systemic risk determinations--for Wachovia and Citigroup--were made to avert the failure of an institution that regulators determined could exacerbate liquidity strains in the banking system. A third determination was made to address disruptions to bank funding affecting all banks. Under this latter determination, FDIC established the Temporary Liquidity Guarantee Program (TLGP), which guaranteed certain debt issued through October 31, 2009, and certain uninsured deposits of participating institutions through December 31, 2010, to restore confidence and liquidity in the banking system. While there is some support for the agencies' position that the statute authorizes systemic risk assistance of some type under TLGP facts and that it permits assistance to the entities covered by the program, there are questions about these interpretations, under which FDIC created a broad-based program of direct assistance to institutions that had never before received such relief--"healthy" banks, bank holding companies, and other bank affiliates. Because these issues are matters of significant public interest and importance, the statutory requirements may require clarification. Regulators' use of the systemic risk exception may weaken market participants' incentives to properly manage risk if they come to expect similar emergency actions in the future. The financial crisis revealed limits in the current regulatory framework to restrict excessive risk taking by financial institutions whose market discipline is likely to have been weakened by the recent use of the systemic risk exception. Congress and regulators are considering reforms to the current regulatory structure. It is important that such reforms subject systemically important financial institutions to stricter regulatory oversight. Further, legislation has been proposed for an orderly resolution of financial institutions not currently covered by the FDI Act. A credible resolution regime could help impose greater market discipline by forcing participants to face significant costs from their decisions and preclude a too-big-to-fail dilemma.

CBP is the lead federal agency charged with keeping terrorists, criminals, and inadmissible aliens out of the country while facilitating the flow of legitimate travel and commerce at the nation’s borders. CBP has three main components that have border security responsibilities. First, CBP’s Office of Field Operations is responsible for processing the flow of people and goods that enter the country through air, land, and sea ports of entry where CBP officers inspect travelers and goods to determine whether they may be legally admitted into the country. Second, CBP’s Border Patrol works to prevent the illegal entry of persons and contraband into the United States between the ports of entry. The Border Patrol is responsible for controlling nearly 7,000 miles of the nation’s land borders between ports of entry and 95,000 miles of maritime border in partnership with the United States Coast Guard. Third, CBP’s Office of Air and Marine helps to protect the nation’s critical infrastructure through the coordinated use of an integrated force of air and marine resources and provides mission support to the other CBP components. For fiscal year 2007, CBP had a $9.3 billion budget, of which $2.5 billion was for border security and trade facilitation at ports of entry. In carrying out its responsibilities, CBP operates 326 official ports of entry, composed of airports, seaports, and designated land ports of entry along the northern and southern borders. Ports of entry vary considerably in size and volume, including diverse locations such as major airports like New York’s John F. Kennedy (JFK) International Airport, and the busiest land crossing in the United States at San Ysidro, California, which processes over 17 million vehicles a year (see fig. 1); small ports in remote rural locations along the Canadian border that process only a few thousand vehicles every year; and seaports like the Port of Miami where cruise ships transport more than 3 million travelers into and out of the country each year. Most ports of entry are land border crossings located along the northern border with Canada or the southern border with Mexico. The four largest land border ports of entry by traveler volume are at San Ysidro, Calexico, and Otay Mesa in California, and the Bridge of Americas in El Paso, Texas. In total, these four ports process about 27 percent of all travelers who enter the country by land. CBP annually processes over 400 million passenger and pedestrian entries, 20 million containers, and 130 million conveyances through ports of entry. In fiscal year 2005, the most recent year for which traveler data are available by mode of entry, land border crossings were by far the busiest for processing people, with about three out of four entries into the country occurring through a land port of entry (see fig. 2). The process for inspecting travelers at airports is significantly different than the process at land ports of entry. Prior to departure from foreign airports, airline carriers electronically submit passenger manifest information to CBP. CBP officers cross-check passengers against a wide range of law enforcement databases before travelers enter the country. Upon arrival in the United States, international airline passengers are first subject to immigration inspections that check visas, passports, and biometric data (see fig. 3). Generally, international passengers arriving by air must present a U.S. passport, permanent resident card, foreign passport, or a foreign passport containing a visa issued by the Department of State. CBP officers may also inspect the luggage of travelers. CBP faces a much greater challenge to identify and screen individuals at land ports of entry, in part because of the lack of advance traveler information and the high volume of traffic at many locations. Unlike travelers who enter the country at airports, travelers entering through land ports of entry can arrive at virtually any time and may present thousands of different forms of documentation, ranging from oral declarations of U.S. or Canadian citizenship, driver’s licenses, birth certificates, passports, visas, permanent resident cards, or U.S. military identity cards. Travelers entering the country by bus or rail must provide documentation and may be subject to further inspection. CBP has implemented measures to help provide advance information on passengers arriving at land ports of entry, including trusted traveler programs that register frequent, low-risk travelers for expedited entry, and license plate readers that match license plate numbers against law enforcement databases. The Immigration and Nationality Act, implementing regulations, and CBP policies and procedures for traveler inspection at all ports of entry require officers to establish, at a minimum, the nationality of individuals and whether they are eligible to enter the country. The first requirement is for the CBP officer to determine if the person is a U.S. citizen or an alien, and if an alien, establish whether the person meets the criteria for admission into the country. Current documentation requirements for entry into the country vary depending on the nationality of the traveler and the mode of entry. For example, U.S. citizens arriving at land ports of entry currently may seek to establish citizenship to a CBP officer through an oral declaration of citizenship. In general, nonimmigrant aliens arriving at land and air ports of entry must present a valid, unexpired passport as well as, depending on country of origin and intended length of stay in the United States, a valid, unexpired visa issued by a U.S. embassy or consulate for entry into the country. As most travelers attempting to enter the country through ports of entry have a legal basis for doing so, a streamlined screening procedure referred to as a primary inspection is used to process those individuals who can be readily identified as admissible. Persons whose admissibility cannot be readily determined and persons selected as part of a random selection process are subjected to a more detailed review called a secondary inspection. This involves a closer inspection of travel documents and possessions, additional questioning by CBP officers, and cross references through multiple law enforcement databases to verify the traveler’s identity, background, purpose for entering the country, and other corroborating information. At the end of this process, the individual may be admitted, refused entry and returned to the country of origin, or detained while admissibility is subject to further review. As part of the original reorganization plan for border security, DHS found that having border security and inspections performed by three separate legacy agencies with different priorities, conflicting policies, and varying leadership structures had led to inconsistent inspections and gaps in the sharing of information between these agencies. As part of its actions to address these concerns, in March 2003, DHS created CBP by merging employees from the three legacy agencies previously responsible for border security. Among other considerations, DHS formed CBP to establish a unified command structure that was intended to reduce duplication of efforts while improving the sharing of information. For operations at ports of entry, in September 2003 CBP issued its plan for consolidating the inspection functions formerly performed by separate inspectors from the three legacy agencies. The plan, referred to as “One Face at the Border,” called for unifying and integrating the legacy inspectors into two new positions—a CBP officer and a CBP agricultural specialist. The new CBP officer would serve as the frontline officer responsible for carrying out the priority anti-terrorism mission as well as the traditional customs and immigration inspection functions, while also identifying and referring goods in need of a more extensive agricultural inspection to the agricultural specialist. CBP anticipated that having a well-trained and well-integrated workforce that could carry out the complete range of inspection functions involving the processing of individuals and goods would allow it to utilize its inspection resources more effectively and enable it to better target potentially high-risk travelers. Together, CBP envisioned the result to be more effective inspections and enhanced security at ports of entry while also accelerating the processing of legitimate trade and travel. While it has been about 4 years since the formation of DHS and CBP, our prior work on mergers and acquisitions found that it generally takes 5 to 7 years to successfully complete such a transformation. For example, GAO designated DHS’s overall transformation as a high-risk area in 2003 based on three factors. First, DHS faced a formidable task in implementing a transformation process that would effectively combine 22 disparate agencies with an estimated 170,000 employees into one department. Second, many of these agencies were facing their own challenges in management areas such as strategic human capital, information technology, and financial management; thus, DHS inherited a host of operational and management challenges from the beginning. Third, DHS’s national security mission is critically important and failure to effectively address its management challenges and program risks could have serious consequences for national security as well as have major economic impacts. CBP, as part of DHS, faces many similar challenges in its efforts to unify three agencies into one and in transforming the role of its officers. For example, with over 40,000 employees, CBP represented the largest merger of people and functions within DHS. Additionally, our prior work on the Immigration and Naturalization Service and the U.S. Customs Service, two of the primary agencies involved in the merger, showed that these agencies experienced many management challenges before their merger into CBP. Finally, like DHS, CBP’ has a primary mission of preventing terrorist attacks that is critical to national security. CBP has had some success in identifying inadmissible aliens and other violators. In fiscal year 2006, CBP successfully caught tens of thousands of violators and it made security improvements at its ports of entry, such as installing new cargo inspection technology. Nevertheless, the agency faces major challenges in overcoming weaknesses in both traveler inspections and physical infrastructure. In regards to traveler inspections, at our request, CBP officials showed us a videotape that identified numerous examples of officers not establishing the nationality of individuals and their eligibility for entering the country as required by law. CBP took action in the summer of 2006 to address the problems by holding high- level management meetings and delivering training on traveler inspections to its officers. However, we later found that CBP’s initial set of corrective actions did not always address the problems and we found similar problems as those on the videotape. CBP issued new policies and procedures to overcome these inspection weaknesses at its land ports of entry including requiring field office directors to conduct assessments to ensure compliance with these new inspection procedures. However, the policies and procedures do not require that field office directors share their assessment results with CBP headquarters management, which may hinder its ability to use the information to overcome weaknesses in traveler inspections and to identify best practices that may occur during implementation of its new policies and procedures. CBP faces a challenge in addressing physical infrastructure weaknesses at land ports of entry in a timely way because some ports are owned by other governmental or private entities, potentially adding to the time needed to agree on infrastructure changes and put them in place. CBP has identified and interdicted thousands of potentially dangerous people and significant amounts of illegal goods at ports of entry. According to CBP, in fiscal year 2006, CBP officers arrested more than 23,000 suspected criminals, denied entry to over 200,000 inadmissible aliens, seized more than 644,000 pounds of illegal drugs, intercepted nearly 1.7 million prohibited agricultural items, and seized over $155 million in illegal commercial merchandise, such as counterfeit footwear and handbags. CBP officers also intercepted 40,362 fraudulent documents used in attempts to enter the country illegally in fiscal year 2006. Over half (21,292) of the fraudulent documents intercepted by CBP involved the alteration or improper use of travel documents issued by the U.S Department of State. About 80 percent of these documents involved impostors—that is, people using authentic, unaltered documents that had been validly issued to another person. The remaining 20 percent attempted to enter with fraudulent documents that were altered in some way, such as a fake or altered U.S. visa, or were entirely counterfeit. CBP’s success in identifying inadmissible aliens and other violators has been enhanced by several new initiatives and programs that aim to further improve security at ports of entry. They include the following: New cargo inspection technology. According to CBP, it has installed nonintrusive inspection technologies at ports of entry that enable officers to rapidly inspect vehicles and truck containers for inadmissible aliens and other violators, nuclear or radiological weapons, or other contraband (see fig. 4). Other nonintrusive technologies, such as radiation detectors, allow CBP to inspect containerized truck and sea cargo without having to perform a time- intensive manual search or other intrusive examinations of the contents. Additional requirements for screening passengers. To improve its ability to target high-risk individuals that are on international flights bound for the United States, CBP in fiscal year 2007, issued a ruling that requires airlines provide passenger manifest information prior to departure. These data are critical in screening passengers against watch lists and other databases and identifying potentially dangerous individuals attempting to enter the United States. CBP also expanded the entry capability of the U.S. Visitor and Immigrant Status Indicator Technology (US-VISIT) program to a total of 116 airports, 15 seaports, and 154 land ports of entry. Through this program, CBP is able to collect, maintain, and share data, including biometric identifiers like digital fingerprints, on selected foreign nationals entering the United States to verify their identities as they Inspection Equipment arrive at air, sea, and land ports of entry. CBP also uses these data to screen persons against watch lists and other law enforcement databases to determine their eligibility to enter the country. Prescreening programs for low-risk travelers. As part of CBP efforts to facilitate legitimate trade and travel, CBP has implemented several initiatives to increase enrollment in its trusted traveler programs, such as the Secure Electronic Network for Travelers’ Rapid Inspection (SENTRI) program on the southern border and the NEXUS program on the northern border. These programs allow registered border residents and frequent cross-border travelers identified as low- risk individuals access to dedicated lanes and expedited processing with minimal inspection (see fig. 5). Participants undergo a thorough background check, a fingerprint law enforcement check, and a personal interview with a CBP officer. Enrollment in these two programs totaled nearly 260,000 members in fiscal year 2007. In addition, as part of an initiative among the United States, Canada, and Mexico, CBP operates a trusted traveler program called the Free and Secure Trade (FAST) program, for truck companies involved in transporting cargo through land ports of entry. Participants in FAST have access to dedicated lanes as well as reduced number of examinations. In 2006, CBP certified 124 new commercial partners and approved over 8,000 new drivers to participate in the program, bringing total FAST enrollment to 84,000 participants. Automated license plate and document readers. CBP has also increased deployment of automated license plate and document readers at land ports of entry. License plate readers automatically read front and rear license plates of vehicles as they enter the primary inspection area, with the data simultaneously queried against CBP and law enforcement databases (see fig. 6). In addition, CBP has installed document readers that electronically read documents, such as passports or border crossing cards, that allow CBP officers to automatically query law enforcement databases. With these readers in place, CBP officers spend less time manually inputting information, thereby reducing inspection times, improving the accuracy of the collected information, and affording the officers the ability to interact more with vehicle occupants. While CBP has had some success in interdicting inadmissible aliens and other violators, CBP acknowledges that it did not apprehend all inadmissible aliens and other violators who sought to enter the country at air and land ports of entry. CBP’s estimates of how many inadmissible aliens and other violators evade detection are based on a sample of travelers who arrive at land and air ports of entry. This program, called Compliance Examination (COMPEX), randomly selects travelers entering the country for more detailed inspections. CBP carries out this program at air and land ports of entry. At land ports, CBP randomly selects vehicles and conducts more detailed inspections of the vehicles and possessions of the traveler. At airports, CBP supervisors randomly select travelers. In both cases, the program is designed to select travelers who would not normally be referred to a secondary inspection and would therefore be allowed to enter the country. On the basis of the extent to which violations are found in the in-depth inspections, CBP estimates the total number of inadmissible aliens and other violators who seek to enter the country at locations where COMPEX is carried out. CBP then calculates an apprehension rate by comparing the number of violators it actually apprehends with the estimated number of violators that attempted entry. Using COMPEX, CBP estimates that several thousand inadmissible aliens and other violators entered the country through air and land ports of entry in fiscal year 2006. Weaknesses in how well inspection procedures were followed increased the potential that inadmissible aliens and other violators successfully entered the country. In the summer of 2006, CBP reviewed hundreds of hours of video from 150 large and small land ports of entry and determined that while CBP officers carried out thorough traveler inspections in many instances, they also identified numerous examples where officers did not comply with inspection requirements, according to CBP officials. At our request, CBP officials showed us a 15-minute video that CBP had prepared that documented noncompliance with inspection requirements. The following were examples of weaknesses that were on the video: In one instance, officers waved vehicles into the United States without stopping the vehicle or interviewing the driver or its passengers as required. In another instance, motorcycles passed through inspection lanes without stopping and making any contact with an officer. In a third instance, during “lane switches” when CBP officers were relieved of their duty and replaced by other officers, officers waved traffic through the lane while the officer logged into the computer. The proper procedure is for traffic to be stopped until the officer is logged into the system and is available to perform proper inspections. In another instance, while the CBP officer was reviewing information on his computer screen, he waved pedestrians through the lane without looking at them, making verbal contact, or inspecting travel documents. In another instance, travelers would simply hold up their identification cards and officers would view them without stepping out of the booth before waving the vehicle through. In these cases, the officers did not appear to make verbal contact with the passengers and did not interview any passengers sitting in the back seat of the vehicle. As a final example, officers did not board recreational vehicles to determine whether additional traveler inspections should be carried out. Without checking the identity, citizenship, and admissibility of travelers, there is an increased potential that dangerous people and inadmissible goods may enter the country and cause harm to American citizens and the economy. According to CBP interviews with apprehended alien smugglers, alien smuggling organizations have been aware of weaknesses in CBP’s inspection procedures and they have trained operatives to take advantage of these weaknesses. This awareness heightens the potential that failed inspections will occur at ports of entry when such procedural weaknesses exist. According to CBP senior management, the factors that may have contributed to these weaknesses included the following: Failure to engage, lack of focus, and complacency. According to CBP senior management, emphasis is not being placed on all missions, and there is a failure by some of its officers to recognize the threat associated with dangerous people and goods entering the country. Insufficient staffing. According to CBP senior management, they are unable to staff ports of entry to sufficiently accommodate the workload. Lack of sufficient staff contributes to officers working double shifts, sometimes resulting in fatigue that can affect decisions. Lack of supervisory presence in primary inspections. CBP senior management noted that lack of supervisory presence at primary inspection booths can contribute to less than optimal inspections. Lack of training. CBP senior management acknowledged that, in some cases, periodic and on-the-job training is not being delivered. CBP has taken action to address weaknesses in its inspection procedures by renewing its emphasis on the need to improve inspections at ports of entry and by revising traveler inspection policies and procedures. In July 2006, CBP headquarters showed field office directors the 15-minute videotape that documented the type of noncompliant inspections that were taking place at land ports of entry. CBP management emphasized the importance of thorough inspection procedures at all ports of entry, including airports and seaports, by requesting field office directors to review current procedures and identify best practices for more thorough inspections. As requested by the Assistant Commissioner of Field Operations, the field office directors conducted a series of meetings with senior port management to review and evaluate their ports’ performance, make corrections where necessary, and identify best practices when inspecting travelers. Through efforts such as these, CBP managers identified best practices that included (1) increased supervisory presence in primary inspection areas; (2) detailing specific steps that should be conducted during primary inspections, such as interviewing travelers and conducting thorough document review (e.g., handling and inspecting documents); and (3) personal visits to ports of entry by directors and managers. CBP also revised its policies and procedures for traveler inspections at land ports of entry to deal with weaknesses that were identified. In July 2007, CBP issued new policies and procedures for inspecting travelers at land ports of entry, including pedestrians and those who enter by vehicle. Among other things, the policies and procedures call on officers to obtain photo identification for all travelers in a vehicle and match the traveler with the photograph. In doing so, the CBP officer is required to obtain a declaration of citizenship, either in the form of travel documents, such as passports, or in the case of a U.S. citizen or Canadian citizen, an oral statement. To the extent possible, officers are required to query law enforcement databases for all travelers in a vehicle. The new policies identify roles and responsibilities of CBP officials at ports of entry, including directors of field offices, port directors, supervisory CBP officers, as well as CBP officers. In the near future, CBP officials are also planning to issue new policies and procedures for processing cargo at land borders and for inspecting travelers who enter the country at airports and seaports. However, issuing new policies and procedures alone does not ensure they will be carried out. For example, after CBP headquarters issued directives, held musters, and issued memorandums to field office and port managers that emphasized the importance of carrying out improved traveler inspections in July 2006, many of the same weaknesses they attempted to deal with continued to exist at ports of entry we visited. In October 2006 and January 2007, or as much as 5 months after managers informed officers of the need to carry out traveler inspections in a more rigorous way by interviewing travelers and examining their travel documents, our investigators identified weaknesses in traveler inspections that were similar to those identified in CBP’s 15-minute video. At several ports of entry, our investigators found that a CBP officer was not staffing the booth when they arrived for inspection. At other locations, CBP officers did not ask for travel documents from our investigators. For example, at one port, when our investigators arrived at the port of entry, one of them called over to three officers who were seated at desks behind a counter about ten feet away. One of the officers asked our investigator if he was a U.S. citizen and the investigator said “yes.” The CBP officers did not get up from their desks to ask for any identification, asked no other question, and allowed our investigator to enter the country. At another port of entry, a CBP officer was not present at the primary inspection booth when our investigator arrived for inspection and he had to wait approximately 3 to 4 minutes before an officer arrived. While CBP’s new policies and procedures are a step in the right direction, ensuring their proper implementation will be key to overcoming weaknesses in traveler inspections. An effective internal control environment is a key method to help agency managers achieve program objectives and enhance their ability to address identified weaknesses. CBP is taking positive steps to implement some control requirements. For example, one of the standards for internal control in the federal government involves monitoring to assess the quality of performance over time. To monitor how traveler inspections are conducted at ports of entry, CBP headquarters has developed a program to covertly test the integrity of existing security measures at ports of entry, including the work carried out by CBP officers. In addition, CBP headquarters officials are called on to conduct compliance reviews. Last, CBP’s new policies and procedures on traveler inspections call on field office directors to ensure compliance with the new inspection procedures at all ports of entry by conducting audits and assessments. Internal control standards state that information should be communicated to management to enable it to carry out its program responsibilities. However, CBP does not require that field offices share the results of their audits and assessments with CBP headquarters management. Without obtaining and receiving the results of field office audits and assessments, CBP management may be hindered in its ability to efficiently use the information to overcome weaknesses in traveler inspections and identify best practices that may occur during implementation of its new policies and procedures. Querying all travelers arriving at land ports of entry against CBP law enforcement databases represents a major challenge for CBP. As discussed earlier in this report, CBP’s new policies and procedures require officers, to the extent feasible, to query travel documents of all travelers who arrive at primary inspection at land ports of entry. In contrast, CBP officers at airports generally handle and query documents of all travelers. Taking the time to enter information into CBP’s law enforcement database for the several hundred million travelers arriving at primary inspection could hinder CBP’s ability to facilitate the movement of legitimate travel and commerce. DHS’s planned Western Hemisphere Travel Initiative, when implemented at land ports of entry, may allow CBP to query more travelers against law enforcement databases and could improve CBP’s ability to identify inadmissible aliens and other violators without harming commerce and travel. The initiative generally requires travelers to have a passport or passport-like document to enter the United States from Canada, Mexico, and other countries in the western hemisphere that is machine-readable and therefore can be more quickly and accurately checked against CBP’s law enforcement database than currently acceptable documents. CBP has already implemented the initiative at air ports of entry, but has yet to do so at land ports of entry. When the initiative is implemented at land ports of entry, CBP officers may be able to query more documents because DHS estimates that processing a traveler at primary inspection will be reduced by 15 to 25 seconds because all travelers will have documents that will be machine readable. CBP’s effectiveness at securing the nation’s borders depends not only on the quality of traveler inspections, but also on the degree to which physical infrastructure is in place to reduce the risk that inadmissible aliens and other violators could bypass inspection points and enter the country. During our site visits, we identified weaknesses in physical infrastructure at some land ports of entry. CBP has developed a process to identify and prioritize capital infrastructure needs at land ports of entry. One component of this planning process is called the Strategic Resource Assessment—an assessment that identifies capital needs at ports of entry by evaluating existing facility conditions, predicting future workload trends, performing space capacity analyses, and estimating costs for the recommended options. CBP’s Office of Finance has compiled resource assessments for 163 land crossings and has prioritized the ports with the greatest need. On the basis of the assessments, CBP estimates that the cost of making capital improvements at land crossings totals about $4 billion. In addition, the assessments identify a planning process to ensure that funding is allocated in a systematic and objective manner. While CBP has made progress in identifying its capital needs, making infrastructure changes to address the problems is not always easy, according to CBP officials responsible for infrastructure improvements. For example, these senior CBP officials noted that they do not have the discretion to make infrastructure improvements on their own, such as installing barriers and bollards, when they do not own the property and therefore need to coordinate their efforts with other entities, such as private bridge commissions or state highway departments. For capital improvements at ports of entry, such as building new vehicle lanes or secondary inspection facilities, the CBP officials said the lead time for making such improvements was long. For example, according to these CBP officials, for the 96 ports of entry that are owned by the General Services Administration (GSA), GSA approves and prioritizes capital improvement projects. The process of submitting a request for an infrastructure improvement and completion of the project is approximately 7 years from start to finish, according to a GSA official. For the 23 ports of entry that are privately owned and leased by GSA, CBP officials noted that coordinating with privately owned companies on infrastructure improvements is a difficult process because the private owner’s interest in facilitating commerce must be balanced with CBP’s interest in national security. According to CBP officials, the degree to which improvements will be made at land ports of entry and how long it will take depend on available funding and the results of discussions with various stakeholders, such as GSA and private port owners. Each year, depending upon funding availability, CBP submits its proposed capital improvement projects based upon the prioritized list it has developed. As of September 2007, CBP had infrastructure projects related to 20 different ports of entry in various stages of development, according to a CBP official. CBP has taken action to improve staffing and training at ports of entry by assessing staffing needs, adding staff, and developing an extensive training program, but it faces challenges in hiring and retaining staff and providing required training. To address staffing, CBP developed a staffing model to identify the resources needed at the nation’s ports of entry. While CBP has had a net increase of about 1,000 more staff since 2005, the results of the staffing model indicate that CBP may need additional officers at ports. Not having sufficient staff contributes to morale problems, fatigue, and safety issues for officers. It also makes it difficult for ports of entry to fully carry out anti-terrorism and other traveler inspection programs. The problems are exacerbated by difficulties in retaining experienced staff. Regarding training, CBP has made progress in developing 37 training modules for CBP officers and a national on-the-job training program for new officers. While it has delivered training to thousands of CBP officers, CBP faces challenges in (1) delivering the required training modules to those who need it and (2) providing on-the-job training to new CBP officers consistent with national program guidance. When staff do not receive required training or are not trained consistently with program guidance, it limits knowledge building and increases the risk that needed expertise is not developed. Senior CBP headquarters officials also stated that the lack of training and training that is inconsistently delivered may increase the risk that terrorists, inadmissible travelers, and illicit goods could be admitted into the country. Congressional concern about CBP’s ability to link resources to its mission led Congress to call on CBP to develop resource allocation models. In responding to language in the conference report for the fiscal year 2007 DHS appropriations and the SAFE Port Act of 2006, CBP developed a staffing model for its land, air, and sea ports of entry. The conference report directed CBP to develop the staffing model in a way that would align officer resources with threats, vulnerabilities, and workload. This directive stemmed, in part, from concern about CBP’s ability to effectively manage its growing workload, minimize wait times, and ensure that CBP officers receive adequate training in all relevant inspection functions. The staffing model is designed to determine the optimum number of CBP officers that each port of entry needs in order to accomplish its mission responsibilities. According to CBP staff involved in developing the staffing model, it is primarily driven by traveler volume and inspection processing times. The staffing model also incorporates assumptions for training, anti- terrorism activities, and staffing for special equipment, such as radiation portal monitors. According to CBP officials, the model’s assumptions will be recalculated each fiscal year in order to account for changes caused by new requirements, procedures, or changes in workload. For example, when the new inspection requirements come into effect under the Western Hemisphere Travel Initiative, CBP can adjust the processing times in the staffing model, which may result in changes in the number of staff needed, according to CBP officials. CBP plans to use the staffing model to help management decide on the number of staff needed and where they should be deployed. In July 2007, CBP provided us with the results for the staffing model. The model’s results showed that CBP would need up to several thousand additional CBP officers and agricultural specialists at its ports of entry. In addition, the staffing model showed the relative need among different CBP locations. CBP has determined that data from the staffing model are law enforcement sensitive. Therefore, we are not providing more detailed data and information from the model in this report. The staffing model was not finalized in time to prepare CBP’s fiscal year 2008 budget request. CBP officials told us that they plan to use the results of the staffing model to determine which locations are to receive additional staffing in fiscal year 2008, should Congress approve their request for additional positions. Before the staffing model was finalized, CBP used other data to determine staffing needs and provide an indication of the degree to which insufficient staffing affects operations at ports of entry. CBP’s 20 field offices and its pre-clearance headquarters office requested additional officers through quarterly resource assessment reports that quantified perceived staffing needs and provided justifications for the request. CBP used the quarterly resource assessment reports to help determine the number of officers to allocate to each office, but the majority of the requests went unfilled due, in part, to budget constraints. In January 2007, 19 of CBP’s 21 offices identified a need for additional officers to accomplish their anti-terrorism responsibilities through special operations and anti-terrorism teams; operate new equipment, such as radiation portal monitors and non- intrusive inspection technologies, both of which are relatively new additions to CBP’s mission responsibilities; and to deal with increased workload from increased traveler volume and the expansion of primary inspection lanes and other facilities. Managers, supervisors, and officers at seven of the eight ports of entry that we visited provided examples of how insufficient staffing affects their ability to carry out primary and secondary inspections: Anti-terrorism and other traveler inspection programs are not fully carried out. CBP uses a “layered” enforcement approach when it inspects travelers. In implementing this approach, port officials told us that when possible, they perform enforcement operations that include anti-terrorism teams and canine inspections (see fig. 7). While considered discretionary, according to CBP officials, these inspections can result in significant numbers of seizures and adverse actions and, thus, are a key tool in traveler inspection efforts. For example, one port conducted a 30-day pilot project during which it focused its efforts on such operations. During this time, CBP officers said they apprehended 96 criminals, inadmissible aliens, and other violators who were in line for primary inspection. Double shifts can result in officer fatigue. Due to staffing shortages, ports of entry rely on overtime to accomplish their inspection responsibilities. Officers at six of the eight ports of entry we visited indicated that officer fatigue caused by excessive overtime negatively affected inspections at their ports of entry. On occasion, officers said they are called upon to work 16-hour shifts, spending long stints in the primary passenger processing lanes in order to keep lanes open, in part to minimize traveler wait times. Further evidence of fatigue came from officers who said that CBP officers call in sick due to exhaustion, in part to avoid mandatory overtime, which in turn exacerbates the staffing challenges faced by the ports. CBP’s onboard staffing level is below its budgeted level, partly due to attrition. According to CBP officials at headquarters and the ports of entry we visited, the gap between the budgeted staffing level and the number of officers onboard is attributable in part to high attrition, with ports of entry losing officers faster than they can hire replacements. Through March 2007, CBP data show that, on average, 52 CBP officers left the agency each 2-week pay period in fiscal year 2007, up from 34 officers in fiscal year 2005. Port managers at five locations indicated that the rising attrition consistently keeps their ports of entry below the budgeted staffing level because of the lengthy amount of time—up to a year—that it can take to hire and train a new officer. On a case-by-case basis, CBP has allowed five field offices to hire above their budgeted staffing levels in order to account for the expected attrition before the next hiring cycle. For example, one field office was allowed to hire over its budgeted staffing level by 100 staff in anticipation of expected officer attrition. However, the use of this option is limited and port managers stated that attrition still outpaces hiring at such locations. Numerous reasons exist for officer attrition. As with other federal agencies, officer retirements are taking a toll on the agency’s workforce. In the next 4 years, over 3,700 CBP officers, or about 20 percent of CBP’s authorized level of 18,530 officers, will become eligible for retirement. In addition, according to CBP officials, CBP officers are leaving the agency to take positions at other DHS components and other federal agencies to obtain law enforcement officer benefits not authorized to them at CBP. In fiscal year 2006, about 24 percent of the officers leaving CBP, or about 339 officers, left for a position in another DHS component. Further, extensive overtime, poor officer morale, and the high cost of living in certain areas were frequently cited by employees who left as reasons for attrition. Our analysis of responses by nonsupervisory CBP staff to the 2006 OPM Federal Human Capital Survey corroborated that they have concerns about efforts to develop staff and agency leadership that could contribute to low morale and attrition. See appendix II for a more complete analysis of responses by nonsupervisory employees to OPM’s Federal Human Capital Survey. CBP recognizes that attrition of officers is adversely affecting its operations and that it must reassess aspects of its human capital approach if it is to hire and retain a high-performing, motivated workforce. CBP officials told us that CBP is considering a number actions including establishing personnel incentive programs, such as a tuition reimbursement program. In addition, the Office of Field Operations plans to work with CBP’s Office of Human Resources Management to develop and distribute a personnel satisfaction survey to obtain employee feedback so that leadership can better address the needs of its workforce. CBP has also revised the exit survey it gives to employees prior to their leaving the agency to better assess their reasons for leaving and to help CBP identify where it is losing employees. CBP plans to analyze data from OPM’s Human Capital Survey, the employee satisfaction and exit surveys, and attrition data to help identify what specific actions CBP may need to take to curb attrition. CBP plans to develop some initial retention strategies by December 2008 and by September 2009 develop approaches to retain staff based on areas of concern identified in the employee exit survey. Starting in 2003, CBP began developing a series of 37 training modules aimed at improving the skills of and to cross-train CBP officers in carrying out inspections at ports of entry. CBP recognized the importance of training in transforming the role of its officers, and has made officer training a focus of the agency. CBP initiated a multiyear cross-training program effort to equip new and legacy officers with the tools necessary to perform primary immigration and customs inspections, and sufficient knowledge to identify agricultural threats in need of further examination by the agricultural specialists. For example, through a combination of computer-based “fundamentals” courses followed by classroom and on- the-job training, a former customs inspector would take training that prepared him or her to conduct secondary inspections related to possible immigration violations. At airports, former customs officers might receive instruction so that they could better conduct traveler inspections. Legacy immigration officers in air and land ports of entry would be trained so that they could work in inspecting baggage or vehicles, respectively. The program involved developing training modules on such topics as anti- terrorism and detecting fraudulent documents. Through its efforts, CBP has cross-trained thousands of officers since 2004. For example, CBP has trained about 12,000 staff in the anti-terrorism module. In August 2007, CBP officials involved in developing the training program at ports of entry told us that CBP is in the process of changing its cross- training program. The officials told us that they hope to update existing cross-training materials and align them with recent changes in policies and procedures. Further, the officials said that the new program will be geared toward delivering training that provides specific expertise in immigration or customs-related inspection activities to new officers or CBP officers transferring to a different job function. According to these officials, they will begin implementing the program in January 2008. While CBP has made progress in developing training modules and in training its officers, CBP managers at seven of the eight ports of entry we visited said they had experienced difficulty in providing their officers with required training in a timely manner because staffing challenges force the ports to choose between performing port operations and providing training. In these instances, port of entry managers told us that training is often sacrificed. One port of entry director stated, “the port is thinking out of the box just to do basic functions cannot even begin to focus on training.” Managers at this port of entry also indicated that training courses are scheduled and then canceled because of staffing concerns. At two other ports of entry we visited, managers indicated that staffing challenges cause the ports of entry to use overtime to fill positions temporarily vacated by officers who participate in training. For example, to provide its officers with four basic cross-training courses, including a course in processing immigration cases, management at one port estimated they would need nearly $4 million in overtime—a condition that would make the port go over its budget for overtime and add to the problems we discussed earlier caused by excessive overtime. We also identified examples where ports of entry we visited did not consistently provide cross-training courses in the manner expected by CBP headquarters. For example, headquarters informed field offices that course content may not be shortened. However, according to a CBP official at one location, his port of entry trained officers to work in the immigration secondary area by pushing officers through a compressed 5- day version of the course rather than the 9-day version developed by headquarters. At another port, new officers we spoke with had not taken the immigration course after working for 3 years, even though CBP guidance states that new officers should take the course during their second year at the port. Challenges in providing training are not new. We have previously reported that staffing shortages have affected training efforts at ports of entry even before CBP was created in March 2003. Managers and supervisors at six of eight ports of entry we visited told us that vulnerabilities in traveler inspections occurred when officers did not receive cross-training before rotating to new inspection areas. Although CBP’s training policy calls for no officer to be placed in an area without receiving the proper cross-training module, officers and supervisors at ports of entry we visited told us that officers are placed in situations for which they had not been trained. While we cannot determine the degree to which this is happening in other ports of entry cross the country, we identified several examples where this policy is not being followed at the ports of entry we visited. For example, legacy customs officers at one port of entry reported feeling ill prepared when called upon to inspect passengers because they had not received the requisite training. One supervisor at this port of entry stated that he had “no confidence” that the officers he supervised could process the casework for a marijuana seizure correctly in order to successfully prosecute the violator because they had not received training. Supervisors at another port of entry told us that they were rotated to areas in which they had not received training. With responsibility over admissibility decisions, these supervisors were concerned that they could not answer questions from their subordinates or make necessary determinations beyond their area of expertise. As a result of not being trained, officers at this port stated that they relied heavily on senior officers from legacy agencies. The officers also told us that these senior officers have been leaving the agency. CBP managers in headquarters recognize that insufficient training can lead to a higher risk of failed inspections. In a presentation that was given to all field office directors, CBP headquarters officials stated that untrained officers increase the risk that terrorists, inadmissible travelers, and illicit goods could enter the country. CBP is attempting to capture information that better reflects whether training requirements are being met. In November 2006, CBP’s field offices submitted their revised training plans indicating how many additional officers needed to be cross-trained over the next several years. However, CBP officials told us that they do not track specifically which officers need to take a particular training module, nor do they track whether those officers have received the needed training. Without data on which CBP officers need which particular cross-training modules and whether they have received the training, CBP does not know the extent that its officers have received the necessary cross-training and are not in a position to measure progress toward achieving its cross-training program goals. Standards for internal control in the federal government provide a framework for agencies to achieve effective and efficient operations and ultimately to improve accountability. One of the standards involves having good controls in place to ensure that management’s directives are carried out. To do so, the standards call on agencies to compare actual performance to planned or expected results throughout the organization and to analyze significant differences. Having reliable data to measure the degree to which training has been delivered to those who are required to receive it would help meet this standard and put CBP management in a position to better gauge the results of its cross-training program. In addition to developing cross-training modules for its officers, CBP also has an on-the-job training program for new officers once they arrive at a port of entry. In a July 2003 report on inspections at land border ports of entry, we recommended that CBP develop and implement a field training program for new officers before they independently conduct inspections. In response to this recommendation, CBP issued guidance for on-the-job training of new CBP officers. According to the guidance, new officers should receive up to 12 and 14 weeks of on-the-job training at land and air ports of entry, respectively. The guidance provides an outline of the type of experiences that a port of entry needs to provide to an officer as part of the on-the-job training program, such as reviewing emergency port of entry procedures and computer systems used in primary inspections. However, at seven of the eight ports of entry we visited officials told us that they had difficulty in providing on-the-job training in compliance with the guidance. For example: Management at one land port of entry stated that it could not provide 12 weeks of on-the-job training to its new officers because of workload, budget, and staffing challenges, but indicated that it tried to provide 6 weeks of on-the-job training. CBP officers at another port of entry told us that the length of their on-the-job training varied from 2 weeks to 6 weeks and they told us that they needed more on-the-job training before inspecting travelers on their own. CBP’s on-the-job training guidance recommends, but does not require, new officers receive 3 weeks of the training under close supervision of a coach or field training officer in order to receive direct guidance and feedback in their performance. However, officials at seven of the ports of entry we visited said that their port of entry had difficulty providing new officers with field training officers. For example, at two ports of entry, experienced officers were unwilling to take on the extra responsibility of training new officers, according to CBP officials at these locations. Vulnerabilities in traveler inspections are created when new officers do not receive required training. For example, new officers who received as little as 2 weeks of on-the-job training rather than the recommended 12 weeks told us that they needed more training before inspecting travelers. In our July 2003 report, we reported that discrepancies in on-the-job training decrease the effectiveness of traveler inspections at ports of entry when little or no on-the-job training is given to new officers. For example, we found that the ports that graded their officers as being the least prepared to carry out traveler inspections were among the ports that provided the least amount of on-the-job training. In addition to new CBP officers not receiving on-the-job training consistent with CBP’s national program guidance, the training program lacks certain elements that may be limiting CBP’s ability to effectively train new officers. Internal control standards related to management of human capital state that management should ensure that the organization has a workforce that has the required skills necessary to achieve organizational goals. While CBP’s on-the-job training guidance requires supervisors to document the tasks officers have performed while in the on-the-job training program, the guidance does not require that officers perform certain tasks to develop needed skills nor does it call on officers to demonstrate proficiency in specific job tasks. The U.S. Border Patrol, an office within CBP, developed a field training program that contains mechanisms to help ensure new Border Patrol agents obtain the needed skills to do their job and demonstrate proficiency in those skills. For example, the Border Patrol identified 32 different specific skills, knowledge, and behavior traits intrinsic to Border Patrol operations, such as processing an expedited removal case, that agents must perform over the 12-week training period. If the new agent cannot gain experience in a specific task, the training officer must arrange for the new agent to conduct a practical exercise. The program requires that agents be evaluated in all 32 areas and be provided weekly feedback on those areas covered in training during the week. Agents are required to demonstrate competency in performing the 32 skills. In addition, training officers are required to write specific comments on performance that is rated as significantly deficient or exceptional. We discussed the utility of the Border Patrol’s on-the-job training program with CBP officials. CBP officials told us that they are planning to revise CBP’s on-the-job field training program for new CBP officers to make it a more robust program. They stated that they would review the Border Patrol’s field training program to identify best practices that they might incorporate into CBP’s on-the-job training program for new CBP officers. Similar to the issues discussed above, our analysis of OPM’s 2006 Federal Human Capital Survey shows that CBP staff expressed concern about training. Our analysis shows that less than half of nonsupervisory CBP staff were satisfied with how CBP assesses their training needs (43 percent), the extent to which supervisors support employee development (43 percent), and the degree to which supervisors provide constructive feedback on how to improve (42 percent). In responding to these three questions, a significantly lower percentage of nonsupervisory staff at CBP was satisfied with their training experiences than nonsupervisory staff in other federal agencies. CBP has developed strategic goals for its traveler inspection program, but it faces challenges in formalizing a set of performance measures that track what progress it is making toward achieving these goals. In September 2006, CBP’s Office of Field Operations issued its 5-year strategic plan called Securing America’s Borders at Ports of Entry, which defines CBP’s national strategy for securing America’s borders, specifically at ports of entry for fiscal year 2007 through fiscal year 2011. Building on the key themes in DHS’s and other CBP strategic plans and applying them specifically to ports of entry, the plan outlines the Office of Field Operation’s vision on establishing secure ports of entry where potential threats are deterred; threats and inadmissible people, goods, and conveyances are intercepted; legitimate trade and travel are facilitated; and operations and outcomes are consistent across locations and modes of transportation. The plan outlines five strategic goals. They are (1) expanding advance knowledge—increasing and improving the information and analysis CBP has about people, goods, and conveyances before they arrive at the ports of entry; (2) modernizing the inspection process to ensure that all people and goods are inspected appropriately; (3) ensuring a flexible enforcement focus to improve CBP’s effectiveness in assessing, detecting, and predicting threats; (4) strengthening physical security at the ports of entry to maintain a secure environment for officers to perform inspections; and (5) building organizational partnerships, maintaining a skilled workforce, and utilizing emerging technologies to achieve CBP’s mission. Although one of CBP’s main goals is to intercept inadmissible aliens and other violators, CBP’s reported performance measure does not address this goal. In its fiscal year 2006 Performance and Accountability Report, CBP reported on the degree to which travelers who arrive at the port of entry are in compliance with immigration, agricultural, and other laws, rules, and regulations as a way to gauge the success of its traveler inspection efforts. Using data from its COMPEX program, CBP uses a measure—called the compliance rate—which showed that in fiscal year 2006 about 99 percent of travelers who seek to enter the United States through 19 major airports and by vehicle at 25 major land ports were in compliance with laws, rules, and regulations. We have reported that linking performance to strategic goals and objectives and publicly reporting this information are important so that Congress and agency management have better information about agency performance and help to ensure accountability. CBP’s current performance measure, the compliance rate, shows the extent to which travelers arriving at ports of entry meet the legal requirements for entering the country. CBP does not use data that measure the extent to which it is intercepting inadmissible aliens and other violators, one of CBP’s key strategic objectives. As discussed earlier in our report, CBP calculates a measure known as the apprehension rate as part of its COMPEX program, which provides an estimate of the agency’s effectiveness in apprehending travelers seeking to enter the country illegally or in violation of other laws. The COMPEX program was originally developed by the former U.S. Customs Service to comply with the Government Performance and Results Act, which requires federal agencies to develop outcome-based performance goals and measures, when possible, as a way to assess the effectiveness and efficiency of their programs. During the course of our review, we discussed with CBP officials the potential of using the apprehension rate as one way of measuring the effectiveness of CBP interdiction efforts. In June 2007, CBP officials told us that CBP was in the process of selecting performance measures for fiscal year 2008 and a decision had not yet been made on whether to include the apprehension rate or some other similar outcome-based measure. Effective inspection of the millions of travelers entering the country each year is critical to the security of the United States. As CBP matures as an organization, having effective inspection procedures, retaining its officer corps, and developing the necessary skills in its officer corps are essential given the critical role that CBP plays in national security. Although CBP developed new inspection procedures that require CBP field office directors to monitor and assess compliance with the new procedures, a key internal control requiring field office directors to communicate with CBP management the results of their monitoring and assessment efforts is not in place. As a result, CBP management may not get information that would identify weaknesses in the traveler inspections process that need to be addressed. The initial set of actions that CBP has taken for dealing with challenges in training at ports of entry is a positive start, but it has not established a mechanism to know whether officers who need specific cross-training have received it and whether new CBP officers have experience in the necessary job tasks and are proficient in them. This means that some officers may be called on to perform certain inspection tasks without having the knowledge and skills to do them. It is also important to have performance measures in place to permit agency management to gauge progress in achieving program goals and, if not, to take corrective action. In regard to traveler inspections, CBP is missing an important performance measure that shows what results are achieved in apprehending inadmissible aliens and other violators. CBP has apprehension rate data that could be used to develop such a performance measure. Having performance measures related to the effectiveness of CBP interdiction efforts would help inform Congress and agency management of improvements resulting from changes in CBP’s traveler inspection program and what gaps in coverage, if any, remain. To mitigate the risk of failed traveler inspections at ports of entry, we recommended in our October 5, 2007 report that the Secretary of Homeland Security direct the Commissioner of Customs and Border Protection to take the following four actions: implement internal controls to help ensure that field office directors communicate to agency management the results of their monitoring and assessment efforts so that agencywide results can be analyzed and necessary actions taken to ensure that new traveler inspection procedures are carried out in a consistent way across all ports of entry; develop data on cross-training programs that measure whether the individuals who require training are receiving it so that agency management is in a better position to measure progress toward achieving training goals; incorporate into CBP’s procedures for its on-the-job training program (1) specific tasks that CBP officers must experience during on-the-job training and (2) requirements for measuring officer proficiency in performing those tasks; and formalize a performance measure for the traveler inspection program that identifies CBP’s effectiveness in apprehending inadmissible aliens and other violators. We provided a draft of the For Official Use Only version of this report to DHS for comment. In commenting on our draft report, DHS, including CBP, agreed with our recommendations. Specifically, DHS stated that CBP is taking action or has taken action to address each recommendation. For example, DHS stated that CBP will develop a measurement validation tool to help confirm that officers have received the necessary cross- training courses before they are assigned to a different work environment. In addition, CBP’s Office of Field Operations (OFO) will evaluate how the Border Patrol is implementing its on-the-job training program and analyze its compatibility to OFO. If effectively implemented, these actions should help address the intent of our recommendations. CBP took issue with an example we used in our draft report describing a situation where two GAO investigators who tested the traveler inspection process at land port of entry were not asked for any identification. We stated that as our investigators attempted to enter at the port, the CBP officer—who was seated behind a desk about 10 feet away—only asked our investigators if they were U.S. citizens and the investigators said “yes.” DHS stated that under current statute and regulation, a person claiming to be a United States citizen arriving at a port of entry is not required to provide identity documents as long as the subject can establish, to the satisfaction of the inspecting officer, citizenship. DHS stated that because CBP officers were satisfied with the citizenship of the two investigators at the time of inspection, identity documents were not required. We agree that an identity document is not required for U.S. citizens at land ports of entry. However, this example is meant to convey that some inspections were not meeting the intent of CBP’s July 2006 management guidance calling for more thorough inspections through traveler interviews and document review. Asking a traveler one question about citizenship when seated at a desk about 10 feet away does not seem to be consistent with the more thorough inspections called for in CBP’s management guidance. We modified our report to include additional information on this episode. DHS also provided technical comments, which we incorporated into the For Official Use Only version of this report as appropriate. Appendix III contains written comments from DHS. We are sending copies of this report to the Secretary of Homeland Security, the Director of the Office of Management and Budget, and interested congressional committees. We will also make copies available to others on request. In addition, this report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-8777 or by e-mail at stanar@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix IV. This report addresses the progress the U.S. Customs and Border Protection (CBP) has made and the remaining challenges it faces in conducting traveler inspections, staffing, and training at ports of entry. Specifically, we answered the following questions: (1) What success and challenges has CBP had in interdicting inadmissible aliens and other violators at its ports of entry? (2) What progress has CBP made in improving staffing and training at its ports of entry and how successful has it been in carrying out these workforce programs? (3) What progress and problems CBP has encountered in setting goals and performance measures for its traveler inspection program? On October 5, 2007, we issued a report that answered the above questions, but it contained information that DHS considered law enforcement sensitive. This version of the report omits sensitive information about CBP’s traveler inspection efforts, including information on the techniques used to carry out inspections, data on the number of inadmissible aliens and other violators that enter the country each year, and data on staffing at ports of entry. In addition, at DHS’s request, we have redacted the specific locations that we visited. The overall methodology used for our initial report is relevant to this version of the report since the information in this product is derived from our first report. Specifically, we, performed our work at the Department of Homeland Security’s (DHS) CBP offices, based in Washington, D.C. We also conducted work at 8 ports of entry—three airports and five land crossings. While we cannot generalize our work from our visits to all ports of entry, we chose these ports of entry to provide examples of operations at ports of entry. At each location, we held group sessions with CBP officers and supervisors. We also interviewed port management and staff involved in training. In addition, our investigators conducted vulnerability assessments of inspection procedures at 8 additional ports of entry. Our investigators conducted covert operations to evaluate screening procedures at small ports of entry. Although we cannot generalize our investigators‘ work at these locations to all ports of entry, we selected these ports of entry to provide examples of traveler inspections at small ports of entry. Our investigators did their work in accordance with quality standards for investigations as set forth by the President’s Council on Integrity and Efficiency. In assessing the adequacy of internal controls, we used the criteria in GAO's Standards for Internal Control in the Federal Government, GAO/AIMD 00-21.3.1, dated November 1999. These standards, issued pursuant to the requirements of the Federal Managers' Financial Integrity Act of 1982 (FMFIA), provide the overall framework for establishing and maintaining internal control in the federal government. Also pursuant to FMFIA, the Office of Management and Budget issued Circular A-123, revised December 21, 2004, to provide the specific requirements for assessing the reporting on internal controls. Internal control standards and the definition of internal control in Circular A-123 are based on the GAO Standards for Internal Control in the Federal Government. To determine what success and challenges CBP has had in interdicting inadmissible aliens and other violators at its ports of entry, we interviewed CBP headquarters officials, such as officials from the Offices of Field Operations, Policy and Planning, Finance, and Training and Development. We obtained and analyzed available DHS documents on traveler inspections, more specifically on COMPEX data (a compliance measurement to determine an overall estimated rate of compliance for travelers), and port infrastructure assessments. For example, we examined COMPEX data that estimate the total number of inadmissible aliens and other violators that seek to enter the country, and compared their compliance and apprehension rates. We assessed the reliability of the COMPEX data by (1) talking with knowledgeable officials about how COMPEX inspections are conducted, documented, and how the apprehension rate estimates are generated; (2) reviewing relevant documentation; and (3) replicating the calculations for the apprehension rates that were provided in the COMPEX reports. We determined the COMPEX estimates were sufficiently reliable for illustrating apprehension rates for the ports of entry the COMPEX program covers. Additionally, we also analyzed CBP’s Strategic Resource Assessment, an evaluation and planning tool designed to identify a port’s infrastructure needs and operational impact on traveler inspections. We also evaluated the CBP Inspector’s Field Manual to determine inspections-related requirements. During our eight site visits, we met with and interviewed field office directors and senior port management staff. During our interviews, we (1) discussed CBP’s success in interdicting inadmissible aliens and other violators and the vulnerabilities in the inspections procedures and concerns related to physical infrastructure and (2) obtained available documentation regarding traveler-related inspections policies and procedures. At each port of entry we visited, we observed both primary and secondary screening procedures and conducted discussion group sessions with officers and supervisors. At each port of entry we visited, we obtained a list of CBP officers scheduled to work during our site visit and from that list we randomly selected officers and supervisors to participate in our sessions at six of the eight ports we visited. We organized the discussion groups by whether they were from legacy organizations or became CBP officers after the merger. At two ports of entry, local management selected officers who would attend the discussion groups and interviews. The group discussions covered a variety of discussion topics, particularly officers’ perceptions and experiences with the “One Face at the Border” initiative and associated challenges in conducting inspections at ports of entry. Over 200 CBP officers participated in our discussion group sessions. In addition to the discussion groups, we also conducted meetings (usually groups of two to four) with CBP chiefs, line supervisors, and specialists (e.g., officers assigned to the intelligence or canine units). These meetings were designed to collect perceptions from CBP middle management and specialists. Additionally, we reviewed a videotape prepared by CBP that documented noncompliance with inspection requirements. Finally, we reviewed CBP’s new policies and procedures for traveler inspections at land ports of entry. To examine what progress CBP has made in improving staffing and training at its ports of entry and how successful has it been in carrying out these workforce programs, we interviewed CBP headquarters officials, including those from the Offices of Field Operations, Policy and Planning, Human Resource Management, and Training and Development. We obtained and analyzed available CBP reports on staffing and training data. For example, we analyzed staffing data from CBP’s Quarterly Resource Assessment, an allocation tool used by field offices to identify the port’s need for additional resources (e.g., request for additional officers). We also collected and analyzed data from CBP’s National Training Plan, a comprehensive guide that documents recommended training guidelines for CBP officers. At each major port we visited, we met with field office directors and senior port management. During our meetings we discussed staffing and training challenges that affected port operations. Follow-up meetings with CBP headquarters officials resulted in receiving staffing numbers from the Quarterly Resource Assessment—an assessment tool used by CBP to identify field office needs and resources—that documented field offices’ request for additional officers. We reviewed headquarters guidance on the on-the-job training program, then met with field office directors and training coordinators. We assessed the reliability of the staffing data by (1) talking with knowledgeable officials about staffing resources, (2) reviewing relevant documentation, and (3) comparing budgeted staffing numbers to officers currently onboard. Although CBP provided us with the results of the staffing model and not the model itself, we reviewed the model with knowledgeable officials, including the assumptions that were used to produce the estimated staffing needs. We understand that the staffing requirements the model produces will vary depending on the assumptions used and we present the key assumptions in the text of our report. Although we discussed the staffing model and its results with CBP officials responsible for the model, validating the model and its results was outside the scope of our review. During the course of our review, we analyzed November 2006 training data from ports of entry that showed the number of officers that had taken cross-training modules as well as the number of officers that local port management had identified as still needing to take a certain module. However, when we compared July 2007 training data with the November 2006 data from ports of entry, we identified inconsistencies with the data. For example, the July 2007 data showed that 120 fewer officers had taken training in a module when compared with the November 2006 data. Because of inconsistencies such as these, we did not use these data in our report. We also reviewed the Border Patrol’s on-the-job training program to identify best practices. Finally, we assessed nonsupervisory CBP employees’ perceptions of the effectiveness of CBP’s workforce management in areas such as job satisfaction, performance evaluation, providing employees sufficient resources to do their jobs, and meeting training needs by analyzing results from the 2004 and 2006 Office of Personnel Management’s (OPM) Federal Human Capital Survey. In addition, we discussed CBP’s training program with officers during discussion groups at the eight ports of entry we visited. To get a perspective on how these results ranked against other federal agencies, we compared the results of our analysis for nonsupervisory CBP employees with responses from nonsupervisory staff in the other DHS component agencies as well as the responses from the other 36 federal agencies included in the survey. To examine what progress CBP has made in setting goals and performance measures for its traveler inspection program, we interviewed and corresponded with officials in CBP’s Offices of Field Operations, Policy and Planning, and Human Resources Management. In addition, to identify CBP’s strategic goals and performance measures for inspecting travelers, we reviewed agency documents such as CBP’s Strategic Plan for 2005 to 2010, CBP Performance and Accountability Reports for fiscal years 2005 and 2006, and OFO’s strategic plan, Securing America’s Borders at Ports of Entry (FY 2007- 2011). We conducted our work from August 2006 through September 2007 in accordance with generally accepted government auditing standards. To gain a broader view of CBP nonsupervisory staff perspectives on workforce issues, we analyzed results from the 2004 and 2006 OPM Federal Human Capital Survey of 36 federal departments or agencies. OPM’s survey represents responses from over 220,000 federal employees, including staff from DHS and CBP. The survey has 73 questions designed to gauge employees’ perceptions about how effectively agencies manage their workforce in the following categories: Personal Work Experiences; Recruitment, Development and Retention; Performance Culture; Leadership; Learning (Knowledge Management); Job Satisfaction; and Satisfaction with Benefits. The following presents our analysis of responses from nonsupervisory CBP staff to questions from OPM’s 2004 and 2006 surveys. Estimates based on responses by CBP nonsupervisory staff to OPM’s 2006 survey show that weaknesses in the work environment generally outweighed the strengths. Our analysis of the survey data showed that CBP nonsupervisory staff identified strengths in 12 of the 73 survey questions. For example, we estimate that a high percentage of CBP staff (1) view their work as important, (2) use information technology to perform work, (3) like the kind of work they do, and (4) understand how their work relates to the agency’s mission. (See table 1 for the top 10 items.) Our analysis also showed that CBP nonsupervisory staff identified weaknesses in 22 of 73 areas. (See table 2 for the bottom 10 items.) When compared with the 2004 survey results, the survey results for 2006 showed that the only area where CBP demonstrated significant progress for non-supervisory staff was increasing employees’ electronic access to learning materials at their desks (an estimated 24 percent improvement from 2004 to 2006). For 19 of 71 items, we estimate that scores for nonsupervisory CBP staff declined by a statistically significant degree. Some of the items where CBP faces greater challenges today than it did in 2004 include (1) having worthwhile discussions with supervisors about performance (an estimated 9.4 percent fewer positive responses in 2006 compared with 2004); (2) rating the overall quality of work done by their unit (6.9 percent fewer); and (3) people I work with cooperate to get the job done (6.2 percent fewer). The estimates for nonsupervisory staff within CBP generally mirror those for the rest of DHS employees. Estimates based on responses from nonsupervisory CBP staff were about the same as those based on the rest of DHS on 47 of the 73 survey items. CBP scored higher on four items, including having a reasonable workload and electronic access to training. CBP was below DHS on the remaining 22 items, including work environment issues such as the quality of work done by the workgroup, feedback from supervisors, and having enough information to do the job well. Placing the results of our analysis in context with how DHS compared with the other 36 departments or agencies involved in OPM’s survey provides a baseline along which to examine a department or agency’s results. For 2006, DHS ranked at or near the bottom of four main categories measured by the survey. DHS ranked 35th on leadership and knowledge management, 36th on having a results-oriented performance culture, 33rd on talent management, and 36th on job satisfaction. To put the situation at CBP in this context, CBP’s survey results rank the agency 10th out of the 13 DHS subcomponents, which would suggest that CBP similarly ranks at or near the bottom in these categories when compared to other federal agencies. For 2006, nonsupervisory CBP staff scored the work environment as lower than elsewhere in the federal government on 61 of the survey’s 73 questions. For example, when we compared CBP with other federal agencies, we estimated that a significantly smaller percentage of CBP nonsupervisory staff said (1) supervisors or team leaders in their work unit support employee development, (2) their work unit recruits people with the right skills, and (3) they are given an opportunity to improve their skills. In contrast, there were no items where CBP staff scored the work environment as significantly better. When viewed in more detail, our analysis of OPM’s survey data shows that CBP faces challenges in staffing and training its personnel, especially when CBP is compared to other federal agencies. For staffing, we estimate that CBP staff gave low marks to CBP for (1) the adequacy of sufficient resources to get the job done and (2) their work unit being able to recruit people with the right skills. With respect to training, less than half of CBP’s staff were reportedly satisfied with (1) the quality of the training received, (2) CBP’s assessment of their training needs, and (3) supervisory support for employee development (see table 3). CBP acknowledges that it needs to improve its workforce management, particularly focusing on raising employees’ perceptions of CBP leadership, enhancing training and career development, and attitudes toward the performance culture at CBP. CBP has formulated a business plan that outlines a variety of corrective actions and initiatives it will take to achieve results in each of these areas. From a strategic standpoint, CBP will establish a Human Capital Advisory Board, composed of senior field leadership from the major CBP offices, that will serve as the central contact point for all program offices, advise and assist with implementing the initiatives outlined in the business plan, and asses the potential for forming an Employee Action Team Advisory Board. To facilitate communication with CBP employees about management actions, the plan sets forth a variety of potential actions, such as creating a Web site on the CBP intranet where CBP supervisors and employees can review the current workforce issues being addressed or results from actions taken, adding a link to CBP’s Web site where the public can access information to learn how CBP is addressing the survey results, and holding town hall meetings at key locations with the Commissioner and other high-level management. To better define the scope of the workforce issues and problems identified through the Federal Human Capital Survey, CBP also plans to conduct employee focus groups as well as administer the survey internally to a larger, more representative sample of CBP employees. Following an in-depth analysis of the results of these actions, CBP will update the business plan in the first quarter of 2008. As part of its leadership initiative, CBP is exploring options to improve employee perceptions of managers’ job performance, establish better communication of management’s goals and priorities, and encourage managers to build more trust and confidence with their employees. To accomplish these goals, CBP plans to create a leadership development checklist to make sure supervisors are addressing critical areas identified through the employee focus groups, and intends to increase the marketing of its recently implemented training course for incumbent supervisors as well as continue the development of training for supervisors newly promoted into management positions. These courses cover integrity, communication, conflict management, and holding effective roundtable discussions. Within the performance culture initiative, CBP wants to find better ways of recognizing employees’ performance that will improve their perceptions about the fairness of CBP’s performance recognition while also supporting a balance between work and family life, which employees also rated poorly. CBP’s plan includes, among other things, a call for improving the channels of communication used to inform supervisors and managers about the type and scope of discretionary performance awards they have at their disposal to issue throughout the year. It also suggests encouraging management at all levels of CBP to have more frequent employee recognition events, to publish award recipients and best practices, and to make awards management a component of performance standards for supervisory personnel. Finally, within the talent management initiative, the plan calls for Human Resources to complete its competency, skills, and needs assessment by the third quarter of fiscal year 2007, and for the Office of Training and Development to implement an automated development and career path system that will guide employees in their career development by providing occupational “road maps” and recommending training based on the occupations they intend to pursue. Richard M. Stana (202) 512-8777 or StanaR@gao.gov. In addition to the contact listed above, Michael Dino, Assistant Director; Neil Asaba; Frances Cook; Josh Diosomito; Kasea Hamar; Michael Meleady; Christopher Leach; Ron La Due Lake; and Stan Stenersen made key contributions to this report. Border Security: Security of New Passports and Visas Enhanced, but More Needs to Be Done to Prevent Their Fraudulent Use. GAO-07-1006. Washington D.C.: July 31, 2007. Homeland Security: Prospects for Biometric US-VISIT Exit Capability Remain Unclear. GAO-07-1044T. Washington, D.C.: June 28, 2007. Border Patrol: Costs and Challenges Related to Training New Agents. GAO-07-997T. Washington, D.C.: June 19, 2007. Homeland Security: Information on Training New Border Patrol Agents. GAO-07-540R. Washington, D.C.: March 30, 2007. Homeland Security: US-VISIT Program Faces Operational, Technological, and Management Challenges. GAO-07-632T. Washington, D.C.: March 20, 2007. Secure Border Initiative: SBInet Planning and Management Improvements Needed to Control Risks. GAO-07-504T. Washington, D.C.: February 27, 2007. Homeland Security: US-VISIT Has Not Fully Met Expectations and Longstanding Program Management Challenges Need to Be Addressed. GAO-07-499T. Washington, D.C.: February 16, 2007. Secure Border Initiative: SBInet Expenditure Plan Needs to Better Support Oversight and Accountability. GAO-07-309. Washington, D.C.: February 15, 2007. Homeland Security: Planned Expenditures for U.S. Visitor and Immigrant Status Program Need to Be Adequately Defined and Justified. GAO-07-278. Washington, D.C.: February 14, 2007. Border Security: US-VISIT Program Faces Strategic, Operational, and Technological Challenges at Land Ports of Entry. GAO-07-378T. Washington, D.C.: January 31, 2007. Border Security: US-VISIT Program Faces Strategic, Operational, and Technological Challenges at Land Ports of Entry. GAO-07-248. Washington, D.C.: December 6, 2006. Department of Homeland Security and Department of State: Documents Required for Travelers Departing from or Arriving in the United States at Air Ports-of-Entry From within the Western Hemisphere. GAO-07-250R. Washington, DC: December 6, 2006. Border Security: Stronger Actions Needed to Assess and Mitigate Risks of the Visa Waiver Program. GAO-06-1090T. Washington, D.C.: September 7, 2006. Illegal Immigration: Border-Crossing Deaths Have Doubled Since 1995; Border Patrol’s Efforts to Prevent Deaths Have Not Been Fully Evaluated. GAO-06-770. Washington, D.C.: August 15, 2006. Border Security: Continued Weaknesses in Screening Entrants into the United States. GAO-06-976T. Washington, D.C.: August 2, 2006. Border Security: Stronger Actions Needed to Assess and Mitigate Risks of the Visa Waiver Program. GAO-06-854. Washington, D.C.: July 28, 2006. Process for Admitting Additional Countries into the Visa Waiver Program. GAO-06-835R. Washington, D.C.: July 28, 2006. Intellectual Property: Initial Observations on the STOP Initiative and U.S. Border Efforts to Reduce Piracy. GAO-06-1004T. Washington, D.C.: July 26, 2006. Border Security: Investigators Transported Radioactive Sources across Our Nation’s Borders at Two Locations. GAO-06-940T. Washington, D.C.: July 7, 2006. Border Security: Investigators Transported Radioactive Sources across Our Nation’s Borders at Two Locations. GAO-06-939T. Washington, D.C.: July 5, 2006. Information on Immigration Enforcement and Supervisory Promotions in the Department of Homeland Security’s Immigration and Customs Enforcement and Customs and Border Protection. GAO-06-751R. Washington, D.C.: June 13, 2006. Homeland Security: Contract Management and Oversight for Visitor and Immigrant Status Program Need to Be Strengthened. GAO-06-404. Washington, D.C.: June 9, 2006. Observations on Efforts to Implement the Western Hemisphere Travel Initiative on the U.S. Border with Canada. GAO-06-741R. Washington, D.C.: May 25, 2006. Homeland Security: Management and Coordination Problems Increase the Vulnerability of U.S. Agriculture to Foreign Pests and Disease. GAO-06-644. Washington, D.C.: May 19, 2006. Border Security: Reassessment of Consular Resource Requirements Could Help Address Visa Delays. GAO-06-542T. Washington, D.C.: April 4, 2006. Border Security: Investigators Transported Radioactive Sources across Our Nation’s Borders at Two Locations. GAO-06-583T. Washington, D.C.: March 28, 2006. Border Security: Investigators Successfully Transported Radioactive Sources across Our Nation’s Borders at Selected Locations. GAO-06-545R. Washington, D.C.: March 28, 2006. Homeland Security: Better Management Practices Could Enhance DHS’s Ability to Allocate Investigative Resources. GAO-06-462T. Washington, D.C.: March 28, 2006. Combating Nuclear Smuggling: DHS Has Made Progress Deploying Radiation Detection Equipment at U.S. Ports-of-Entry, but Concerns Remain. GAO-06-389. Washington, D.C.: March 22, 2006. Combating Nuclear Smuggling: Corruption, Maintenance, and Coordination Problems Challenge U.S. Efforts to Provide Radiation Detection Equipment to Other Countries. GAO-06-311. Washington, D.C.: March 14, 2006. Border Security: Key Unresolved Issues Justify Reevaluation of Border Surveillance Technology Program. GAO-06-295. Washington, D.C.: February 22, 2006. Homeland Security: Recommendations to Improve Management of Key Border Security Program Need to Be Implemented. GAO-06-296. Washington, D.C.: February 14, 2006. Homeland Security: Visitor and Immigrant Status Program Operating, but Management Improvements Are Still Needed. GAO-06-318T. Washington, D.C.: January 25, 2006. Department of Homeland Security: Strategic Management of Training Important for Successful Transformation. GAO-05-888. Washington, D.C.: September 23, 2005. Border Security: Strengthened Visa Process Would Benefit from Improvements in Staffing and Information Sharing. GAO-05-859. Washington, D.C.: September 13, 2005. Border Security: Opportunities to Increase Coordination of Air and Marine Assets. GAO-05-543. Washington, D.C.: August 12, 2005. Border Security: Actions Needed to Strengthen Management of Department of Homeland Security’s Visa Security Program. GAO-05-801. Washington, D.C.: July 29, 2005. Border Patrol: Available Data on Interior Checkpoints Suggest Differences in Sector Performance. GAO-05-435. Washington, D.C.: July 22, 2005. Combating Nuclear Smuggling: Efforts to Deploy Radiation Detection Equipment in the United States and in Other Countries. GAO-05-840T. Washington, D.C.: June 21, 2005. Homeland Security: Performance of Foreign Student and Exchange Visitor Information System Continues to Improve, but Issues Remain. GAO-05-440T. Washington, D.C.: March 17, 2005. Homeland Security: Some Progress Made, but Many Challenges Remain on U.S. Visitor and Immigrant Status Indicator Technology Program. GAO-05-202. Washington, D.C.: February 23, 2005. Border Security: Streamlined Visas Mantis Program Has Lowered Burden on Foreign Science Students and Scholars, but Further Refinements Needed. GAO-05-198. Washington, D.C.: February 18, 2005. Border Security: Joint, Coordinated Actions by State and DHS Needed to Guide Biometric Visas and Related Programs. GAO-04-1080T. Washington, D.C.: September 9, 2004. Border Security: State Department Rollout of Biometric Visas on Schedule, but Guidance Is Lagging. GAO-04-1001. Washington, D.C.: September 9, 2004. Border Security: Consular Identification Cards Accepted within United States, but Consistent Federal Guidance Needed. GAO-04-881. Washington, D.C.: August 24, 2004. Border Security: Additional Actions Needed to Eliminate Weaknesses in the Visa Revocation Process. GAO-04-795. Washington, D.C.: July 13, 2004. Border Security: Additional Actions Needed to Eliminate Weaknesses in the Visa Revocation Process. GAO-04-899T. Washington, D.C.: July 13, 2004. Border Security: Agencies Need to Better Coordinate Their Strategies and Operations on Federal Lands. GAO-04-590. Washington, D.C.: June 16, 2004. Overstay Tracking: A Key Component of Homeland Security and a Layered Defense. GAO-04-82. Washington, D.C.: May 21, 2004. Homeland Security: First Phase of Visitor and Immigration Status Program Operating, but Improvements Needed. GAO-04-586. Washington, D.C.: May 11, 2004. Homeland Security: Risks Facing Key Border and Transportation Security Program Need to Be Addressed. GAO-04-569T. Washington, D.C.: March 18, 2004. Border Security: Improvements Needed to Reduce Time Taken to Adjudicate Visas for Science Students and Scholars. GAO-04-443T. Washington, D.C.: February 25, 2004. Border Security: Improvements Needed to Reduce Time Taken to Adjudicate Visas for Science Students and Scholars. GAO-04-371. Washington, D.C.: February 25, 2004. Homeland Security: Overstay Tracking Is a Key Component of a Layered Defense. GAO-04-170T. Washington, D.C.: October 16, 2003. Security: Counterfeit Identification Raises Homeland Security Concerns. GAO-04-133T. Washington, D.C.: October 1, 2003. Homeland Security: Risks Facing Key Border and Transportation Security Program Need to Be Addressed. GAO-03-1083. Washington, D.C.: September 19, 2003. Security: Counterfeit Identification and Identification Fraud Raise Security Concerns. GAO-03-1147T. Washington, D.C.: September 9, 2003. Land Border Ports of Entry: Vulnerabilities and Inefficiencies in the Inspections Process. GAO-03-1084R. Washington, D.C.: August 18, 2003. Federal Law Enforcement Training Center: Capacity Planning and Management Oversight Need Improvement. GAO-03-736. Washington, D.C.: July 24, 2003. Border Security: New Policies and Increased Interagency Coordination Needed to Improve Visa Process. GAO-03-1013T. Washington, D.C.: July 15, 2003. Land Border Ports of Entry: Vulnerabilities and Inefficiencies in the Inspections Process, GAO-03-782. Washington, D.C.: July 2003. Border Security: New Policies and Procedures Are Needed to Fill Gaps in the Visa Revocation Process. GAO-03-908T. Washington, D.C.: June 18, 2003. Border Security: New Policies and Procedures Are Needed to Fill Gaps in the Visa Revocation Process. GAO-03-798. Washington, D.C.: June 18, 2003. Homeland Security: Challenges Facing the Department of Homeland Security in Balancing its Border Security and Trade Facilitation Missions. GAO-03-902T. Washington, D.C.: June 16, 2003. Counterfeit Documents Used to Enter the United States from Certain Western Hemisphere Countries Not Detected. GAO-03-713T. Washington, D.C.: May 13, 2003. Information Technology: Terrorist Watch Lists Should Be Consolidated to Promote Better Integration and Sharing. GAO-03-322. Washington, D.C.: April 15, 2003. Border Security: Challenges in Implementing Border Technology. GAO-03-546T. Washington, D.C.: March 12, 2003.

U.S. Customs and Border Protection (CBP) is responsible for keeping terrorists and other dangerous people from entering the country while also facilitating the cross-border movement of millions of travelers. CBP carries out this responsibility at 326 air, sea, and land ports of entry. In response to a congressional request, GAO examined CBP traveler inspection efforts, the progress made and the challenges that remain in staffing and training at ports of entry, and the progress CBP has made in developing strategic plans and performance measures for its traveler inspection program. This is a public version of a For Official Use Only report GAO issued on October 5, 2007. To conduct its work, GAO reviewed and analyzed CBP data and documents related to inspections, staffing, and training, interviewed managers and officers, observed inspections at eight major air and land ports of entry, and tested inspection controls at eight small land ports of entry. Information the Department of Homeland Security (DHS) deemed sensitive has been redacted. CBP has had some success in identifying inadmissible aliens and other violators, but weaknesses in its operations increase the potential that terrorists and inadmissible travelers could enter the country. In fiscal year 2006, CBP turned away over 200,000 inadmissible aliens and interdicted other violators. Although CBP's goal is to interdict all violators, CBP estimated that several thousand inadmissible aliens and other violators entered the country though ports of entry in fiscal year 2006. Weaknesses in 2006 inspection procedures, such as not verifying the nationality and admissibility of each traveler, contribute to failed inspections. Although CBP took actions to address these weaknesses, subsequent follow up work conducted by GAO months after CBP's actions found that weaknesses such as those described above still existed. In July 2007, CBP issued detailed procedures for conducting inspections including requiring field office managers to assess compliance with these procedures. However, CBP has not established an internal control to ensure field office managers share their assessments with CBP headquarters to help ensure that the new procedures are consistently implemented across all ports of entry and reduce the risk of failed traveler inspections. CBP developed a staffing model that estimates it needs up to several thousand more staff. Field office managers said that staffing shortages affected their ability to carry out anti-terrorism programs and created other vulnerabilities in the inspections process. CBP recognizes that officer attrition has impaired its ability to attain budgeted staffing levels and is in the process of developing a strategy to help curb attrition. CBP has made progress in developing training programs, yet it does not measure the extent to which it provides training to all who need it and whether new officers demonstrate proficiency in required skills. CPB issued a strategic plan for operations at its ports of entry and has collected performance data that can be used to measure its progress in achieving its strategic goals. However, current performance measures do not gauge CBP effectiveness in apprehending inadmissible aliens and other violators, a key strategic goal.

Congress established the RFS as part of the Energy Policy Act of 2005, in response to concerns about the nation’s dependence on imported oil. The RFS initially required that a minimum of 4 billion gallons of renewable fuels be blended into transportation fuels in 2006, ramping up to 7.5 billion gallons by 2012. Two years later, the Energy Independence and Security Act of 2007 (EISA) increased and expanded the statutory target volumes for renewable fuels and extended the ramp-up period through 2022. More specifically, the act established overall target volumes for renewable fuels that increase from 9 billion gallons in 2008 to 36 billion gallons in 2022. The EISA volumes can be thought of in terms of two broad categories: conventional and advanced biofuels: Conventional biofuel: Biofuels from new facilities must achieve at least a 20-percent reduction in greenhouse gas emissions, relative to 2005 baseline petroleum-based fuels. The dominant biofuel produced to date is conventional corn-starch ethanol, although recently some conventional biodiesel has entered the fuel supply. Advanced biofuel: Biofuels, other than ethanol derived from corn starch must achieve at least a 50-percent reduction in life-cycle greenhouse gas emissions, as compared with 2005 baseline petroleum-based fuels. Advanced biofuel is a catch-all category that may include a number of fuels, including those made from any qualified renewable feedstock that achieves at least a 50-percent reduction in lifecycle greenhouse gas emissions, such as ethanol derived from cellulose, sugar, or waste material. This category also includes the following. Biomass-based diesel: Advanced biomass-based diesel must have life-cycle greenhouse gas emissions at least 50 percent lower than traditional petroleum-based diesel fuels. Cellulosic biofuel: Advanced biofuel derived from any cellulose, hemicellulose, or lignin that is derived from renewable biomass must have life-cycle greenhouse gas emissions at least 60 percent lower than traditional petroleum-based fuels. This category of fuel may include cellulosic ethanol, renewable gasoline, cellulosic diesel, and renewable natural gas from landfills that can be used to generate electricity for electric vehicles or used in vehicles designed to run on liquefied or compressed natural gas. EPA administers the RFS in consultation with DOE and USDA. EPA’s responsibilities for implementing the RFS include setting annual volume requirements. Each year, by November 30, EPA is required to establish via rulemaking the volumes of biofuel that must be blended into transportation fuels during the following calendar year (volume requirement). The statute provides EPA with waiver authority to set volumes below the targets specified in the statute under certain circumstances, such as when there is inadequate domestic supply. The structure of the volume targets emphasized conventional biofuels in the early years covered by the statute, while providing lead time for the development and commercialization of advanced, and especially cellulosic, biofuels. However, these fuels have not been produced in sufficient quantities to meet statutory targets through 2016. As a result, since 2010, EPA has used its waiver authority to deviate from the statutory target volumes and has reduced the volume requirement for cellulosic biofuel every year, citing inadequate domestic supply, among other things (see fig.1). Further, in December 2015—when EPA finalized the volume requirements for 2014, 2015, and 2016—the agency reduced the total renewable fuel requirement for those years. Effectively, this meant that EPA reduced the amount of conventional biofuels required under the program relative to statutory targets for those years. In this case, EPA cited constraints in the fuel market’s ability to accommodate increasing volumes of ethanol. EPA’s use of this waiver authority has been controversial among some RFS stakeholders, and EPA’s 2015 requirement currently faces legal challenges from multiple parties. However, in the volume requirement it finalized in November 2016, EPA effectively set the amount of conventional biofuels required under the program at 15 billion gallons, equal to the statutory target for 2017 (see fig.1). In our November 2016 report, we found that the federal government has supported R&D related to advanced biofuels through direct research or grants, and the target of this R&D is shifting away from cellulosic ethanol and toward drop-in biofuels. Unlike corn-starch-based or cellulosic ethanol, drop-in fuels, such as renewable gasoline, are fully compatible with existing infrastructure, such as vehicle engines and distribution pipelines. In fiscal years 2013 through 2015, the federal government obligated more than $1.1 billion for advanced biofuels R&D. Of this amount, DOE obligated over $890 million. For example, DOE’s Office of Science funds three bioenergy research centers affiliated with universities and national laboratories that conduct basic research for all stages of biofuel production. In addition, USDA obligated over $168 million in fiscal years 2013 through 2015 to support advanced biofuels. For example, USDA scientists developed a novel process to increase production of butanol, a drop-in fuel that lowered production costs by over 20 percent. The remaining federal obligations during these years were through EPA, DOD, and NSF, which obligated less for such R&D. According to agency officials, agencies are shifting their focus to drop-in fuels in part because these fuels are compatible with existing infrastructure. Officials from one federal funding agency said this compatibility makes drop-in fuels more desirable than cellulosic ethanol. As we reported in November 2016, experts told us that the technology to produce several advanced biofuels is well understood but noted that among those currently being produced there is limited potential for increased production in the near term. Experts further cited multiple factors making it challenging to significantly increase the speed and volume of production. In addition, current advanced biofuel production is far below overall RFS target volumes, and those volumes are increasing every year. Consequently, it does not appear possible to meet statutory target volumes for advanced biofuels in the RFS under current market and regulatory conditions. Biofuels that the experts identified as being technologically well understood include biodiesel, renewable diesel, renewable natural gas, cellulosic ethanol, and some drop-in fuels. A few of these fuels are being produced in significant volumes, but the overall volume being produced falls short of the volume target in the RFS. For example, in 2015, about 3.1 billion ethanol equivalent gallons of advanced biofuels were produced, falling short of the statutory target of 5.5 billion gallons for that year. By 2022, the advanced biofuels target increases to 21 billion gallons, so production would have to rapidly increase to meet this target. Even though a few of these fuels, such as biodiesel and renewable diesel, are being produced in significant volumes, it is unlikely that production of these fuels can expand much in the next few years because of feedstock limitations. Current production of cellulosic biofuels is far below the statutory volume targets and, according to the experts, there is limited potential for expanded production to meet future higher targets, in part because production costs are currently too high. Experts told us that technologies to produce other fuels, such as some drop-in fuels, are well understood, but that those fuels are not being produced because production is too costly. Experts identified a number of factors that will affect the speed and volume of advanced biofuel production, including the following. The low price of fossil fuels relative to that of advanced biofuels. This disparity in price is a disincentive for consumers to adopt greater use of biofuels and also a deterrent for private investors entering the advanced biofuels market. Uncertainty about government policy, including whether the RFS and federal tax credits that support advanced biofuels will continue to be in effect. While such policies should encourage investment, investors do not see them as reliable and thus discount their potential benefits when considering whether to invest. High cost of converting cellulosic feedstocks. These costs include transporting and handling feedstocks, processing them into a fuel, and disposing of wastes, among other things. Time and cost to bring a new technology to commercial-scale production. The timeline to bring a new technology from laboratory scale to commercial scale is 12 years if everything works well, and it can be considerably longer. Time and cost to secure fuel certification and acceptance. Before a fuel is brought to market, it must go through regulatory registration, certification by ASTM International, and other testing. Underdeveloped feedstock supply chain. Lack of logistics for the entire feedstock supply chain—from securing a contract to delivering and storing a feedstock—is an economic barrier to the production of advanced biofuels. As we found in our November 2016 report, it is unlikely that the goals of the RFS—reduce greenhouse gas emissions and expand the nation’s renewable fuels sector—will be met as envisioned because there is limited production of advanced biofuels and limited potential for expanded production by 2022. Advanced biofuels achieve greater greenhouse gas reductions than conventional biofuels, although the latter account for most of the biofuel blended into domestic transportation fuels under the RFS. As a result, the RFS is unlikely to achieve greenhouse gas emissions reductions as envisioned. For example, the cellulosic biofuel blended into the domestic transportation fuel supply in 2015 was less than 5 percent of the statutory target of 3 billion gallons. Partly as a result of low production of advanced biofuels, EPA has reduced the RFS targets for such fuels through waivers in each of the last 4 years. According to experts we interviewed, the shortfall of advanced biofuels is the result of high production costs, and the investments in further R&D required to make these fuels more cost-competitive with petroleum-based fuels, even in the longer run, are unlikely in the current investment climate. Given the relative scarcity of advanced biofuels, most of the biofuel blended under the RFS to date has been conventional corn-starch ethanol, which achieves smaller greenhouse gas emission reductions than advanced biofuels. The use of corn-starch ethanol has been effectively capped at 15 billion gallons. As a result, expanded use of biofuels will require increasing use of advanced biofuels, and experts told us the most likely advanced biofuel to be commercially produced in the near- to mid-term will be cellulosic ethanol. However, the ability to add ethanol to the transportation fuel market to meet expanding RFS requirements is limited by the incompatibility of ethanol blends above E10 (up to 10 percent ethanol) with the existing vehicle fleet and fueling infrastructure. Many experts and stakeholders refer to this infrastructure limitation as the “blend wall.” If ethanol continues to be the primary biofuel produced to meet the RFS, these infrastructure limitations will have to be addressed. Several experts raised concerns about the extent to which the RFS is achieving its goal for reducing greenhouse gas emissions, given that most biofuel blended under the RFS is corn-starch ethanol. More specifically, some experts were critical of the life-cycle analysis EPA used to determine the greenhouse gas emissions reductions for corn-starch ethanol. Further, corn-starch ethanol plants that were in operation or under construction before December 19, 2007, are not subject to the requirement to reduce greenhouse gas emissions by at least 20 percent. According to an August 2016 EPA Inspector General report, grandfathered production that is not subject to any greenhouse gas reduction requirements was estimated to be at least 15 billion gallons, or over 80 percent of today’s RFS blending volume. Moreover, some experts told us that the RFS creates a perverse incentive to import Brazilian sugarcane ethanol. Specifically, because sugarcane ethanol qualifies as an advanced biofuel, it is more profitable to import this fuel than to domestically produce advanced biofuels. According to these experts, the import of sugarcane ethanol, which occurs to meet RFS requirements, causes significant greenhouse gas emissions as a result of fuel burned during shipping. As we reported in November 2016, while advanced biofuels are not likely to be produced in sufficient quantities to meet the statutory targets, experts identified actions that they suggested could improve the existing RFS framework by incrementally increasing investment in advanced biofuels, which may lead to greater volumes of these fuels being produced and used in the longer term. For example, some experts stated that the Second Generation Biofuel Producer Tax Credit—an incentive to accelerate commercialization of fuels in the advanced and cellulosic biofuels categories—has expired and been reinstated (sometimes retroactively) about every 2 years, contributing to uncertainty among cellulosic fuel producers and investors. One expert told us that investment in cellulosic biofuels could be encouraged, in part, by maintaining the Second Generation Biofuel Producer Tax Credit consistently, rather than allowing it to periodically lapse and be reinstated. In addition, experts identified actions to increase compatibility of infrastructure with higher ethanol blends. For example, several experts suggested that expanding grants to encourage infrastructure improvements, such as USDA’s Biofuel Infrastructure Partnership, could increase both the availability and competitiveness of higher blends at retail stations nationwide. Through this partnership, USDA is investing $100 million to install nearly 5,000 pumps offering high-ethanol blends in 21 states. However, some experts also said that blender pumps are not being installed with the density required to test demand. One expert suggested that, instead of installing blender pumps at all the transportation fuel stations of a certain brand in a region, blender pumps should be installed at all the stations at a specific road intersection. That way, these stations would be forced to compete with each other, which this expert told us would result in more competitive prices at the pump and increased incentives to improve fueling infrastructure. As we reported in November 2016, several experts stated that the RFS is not the most efficient way to achieve the environmental goal of reducing greenhouse gas emissions, and they suggested policy alternatives—in particular, a carbon tax and a low carbon fuel standard (LCFS). Several experts suggested that these alternatives would be more efficient at reducing greenhouse gas emissions. Specifically, some experts said that, whereas the RFS creates disincentives for the production of cellulosic fuels that achieve the greatest reductions in greenhouse gas emissions, a carbon tax or LCFS would incentivize the technologies that achieve the greatest such reductions at the lowest cost. Under a carbon tax, each fossil fuel would be taxed in proportion to the amount of greenhouse gas (carbon dioxide) released in its combustion. In addition, one expert stated that a carbon tax is preferable to the RFS because it allows market effects to increase the price of emission-causing activities, which decreases demand for those activities. As a result, a carbon tax could sustain consumers’ interest in fuel-saving vehicles and result in a wide range of fuel-saving responses from all consumers (rather than just those purchasing a new vehicle). However, some experts also noted that a carbon tax would force further electrification of the light-duty vehicle fleet because the electric power sector is the cheapest sector from which to obtain greenhouse gas reductions. According to one expert, this electrification of the light-duty fleet might further limit biofuels R&D, in effect undermining the RFS goal to expand that sector. In light of these concerns, several experts said that an LCFS would be more flexible and efficient than the RFS or a carbon tax at developing biofuels that achieve the greatest greenhouse gas reductions. Specifically, an LCFS accounts for carbon in a given fuel on a cost per unit of carbon intensity, thereby supporting incremental carbon reductions. An LCFS can be implemented in one of two ways. The first involves switching to direct fuel substitutes (e.g., drop-in fuels) or blending biofuels with lower greenhouse gas emissions directly into gasoline and diesel fuel. The second involves switching from petroleum-based fuels to other alternatives, such as natural gas, hydrogen, or electricity, because an LCFS would allow a wider array of fuel pathways than the RFS. Under the first scenario, an LCFS would promote biofuel usage, rather than incentivizing electrification of the light-duty vehicle fleet. As a result, according to some experts, an LCFS is preferable to a carbon tax because it more efficiently reduces greenhouse gas emissions and promotes the expansion of the biofuel sector. However, other experts we spoke with critiqued an LCFS as being uneconomical. Specifically, one expert stated that, while an LCFS such as the one in California could force technology and create greenhouse gas reductions in the fuel market, the costs of implementing an LCFS are much higher than its benefits. Chairman Lankford, Ranking Member Heitkamp, and Members of the Subcommittee, this concludes my prepared statement. I would be pleased to answer any questions that you may have at this time. If you or your staff members have any questions concerning this testimony, please contact Frank Rusco, Director, Natural Resources and Environment, at (202) 512-3841 or ruscof@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Other individuals who made key contributions to this testimony include Karla Springer, Assistant Director; Jesse Lamarre-Vincent; Marietta Revesz; and Jarrod West. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Since 2006 the RFS has required that transportation fuels—typically gasoline and diesel—sold in the United States be blended with increasing volumes of biofuels to meet environmental and energy goals. Annual targets for the volumes of biofuels to be blended are set by statute. EPA is responsible for adjusting the statutory targets through 2022 to reflect expected U.S. industry production levels, among other factors, and for setting volume targets after 2022. Biofuels included in the RFS are either conventional (primarily corn-starch ethanol) or advanced biofuels (e.g., cellulosic ethanol and biomass-based diesel). Advanced biofuels emit fewer greenhouse gases than petroleum-based fuels and corn-starch ethanol. In November 2016, GAO issued two reports on the RFS. This testimony is based on those two reports: GAO-17-94 and GAO-17-108 . It provides information on whether the RFS is expected to meet its production and other targets, as well as expert views on any federal actions that could improve the RFS framework, among other things. For the reports on which this testimony is based, GAO analyzed legal requirements and EPA data. In addition, GAO worked with the National Academy of Sciences to convene a meeting of experts from industry, academia, and research organizations in May 2016. GAO also contracted with the National Academy of Sciences for a list of experts on issues related to the RFS. Further information on how GAO conducted its work is contained in the reports. It is unlikely that the goals of the Renewable Fuel Standard (RFS)—to reduce greenhouse gas emissions and expand the nation's renewable fuels sector while reducing reliance on imported oil—will be met as envisioned because there is limited production of advanced biofuels and limited potential for expanded production by 2022. Advanced biofuels, such as cellulosic ethanol and biomass-based diesel, achieve greater greenhouse gas reductions than conventional biofuels (primarily corn-starch ethanol), but the latter account for most of the biofuel blended into domestic transportation fuels under the RFS. As a result, the RFS is unlikely to achieve the targeted level of greenhouse gas emissions reductions. For example, the cellulosic biofuel blended into the transportation fuel supply in 2015 was less than 5 percent of the statutory target of 3 billion gallons. Partly as a result of low production of advanced biofuels, the Environmental Protection Agency (EPA), which administers the RFS in consultation with other agencies, has reduced the RFS targets for such fuels through waivers in each of the last 4 years (see figure). According to experts GAO interviewed, the shortfall of advanced biofuels is due to high production costs. The investments required to make these fuels more cost-competitive with petroleum-based fuels, even in the longer run, are unlikely in the current investment climate, according to experts.

The United States has historically sought to attract international students to its colleges and universities. In recent years international students have earned about one-third or more of all of the U.S. degrees at both the master’s and doctoral levels in several of the science, technology, engineering, and mathematics (STEM) fields. In academic year 2002-2003 alone, international students earned between 45 percent and 57 percent of all the STEM degrees in the United States. Several federal agencies coordinate efforts to attract and bring international students to the United States and implement related requirements. The Department of State (State) manages the student visa application process, administers some student exchange programs, offers grants to facilitate international exchanges, and provides information promoting educational opportunities in the United States. State’s Bureau of Educational and Cultural Affairs supports a global network of more than 450 advising centers around the world that provide comprehensive information about educational opportunities in the United States and guidance on how to access those opportunities. In addition, the Undersecretary for Public Diplomacy and Public Affairs has undertaken ongoing efforts at outreach. For example, the office has organized several delegations of American university presidents to travel overseas with the Undersecretary in order to emphasize the United States’ interest in welcoming international students. The Department of Homeland Security enforces immigration laws and oversees applications for changes in immigration status. It also administers the Student and Exchange Visitor Information System (SEVIS), an Internet-based system that maintains data on international students and exchange visitors before and during their stay in the United States. Finally, the Department of Education (Education) sponsors initiatives to encourage academic exchanges between the United States and other countries, and the Department of Commerce offers various activities to help U.S. educational institutions market their programs abroad. Students or exchange visitors interested in studying in the United States must first be admitted to a U.S. school or university before starting the visa process. Most full-time students enter the United States under temporary visas, which usually permit them to stay for the duration of their studies but may require renewals if they return home before their studies are complete. In order to apply for a visa at a U.S. embassy or consulate, students are required to submit a SEVIS -generated document issued by a U.S. college or university or State-designated sponsor organization when they apply for a visa. State advises student applicants to apply early for a student or exchange visitor visa to make sure that there is sufficient time to obtain an appointment for a visa interview and for visa processing. Among the long-standing requirements for students applying for a visa is that they demonstrate an “intent to return” to their country of origin after they complete their studies. Graduates who wish to stay and work in the United States beyond the time allowed by their student visas generally need to receive approval for a change in status, for example, through a temporary work visa or through permanent residency. Although the United States continues to enroll more international students than any other country, the number of international students enrolled in U.S. higher education institutions leveled off and even dropped slightly after 2001, as shown in figure 1. Figure 2 shows that the U.S. share of international students worldwide decreased between 2000 and 2004. According to the Institute of International Education, the decline in the number of international students attending U.S. higher education institutions between 2002 and 2003 was the first drop in over 30 years. While some preliminary data suggest that international student enrollment numbers may be rebounding, enrollments have yet to return to previous levels. Nevertheless, the United States continues to be a prime study destination for international students for numerous reasons: its high- quality higher education institutions, top-ranked graduate programs, strong research funding, English-language curriculum, and a diverse foreign-born faculty. As worldwide demand for higher education continues to rise, changes in the global higher education landscape have provided students with more options. For example, technological advancements have spurred online courses and even completely online programs that cater largely to nontraditional students having work and family commitments. Between 1995 and 2001, enrollment in distance education at the college level nearly quadrupled to over 3 million students, according to Education’s most recent data. In addition, international partnerships allow institutions to share faculty members and facilitate study abroad opportunities. International branch campuses now provide international students the opportunity to receive an American education without leaving their home country. Greater competition has prompted some countries to embrace instruction in English and encouraged other systems to expand their recruiting activities and incentives. Germany alone offers nearly 400 courses in English that are geared toward international students. In terms of recruiting, several of the participants during our global competitiveness and higher education forum suggested that some countries appear more committed to attracting international students than the United States or are now competing with the United States for the best and the brightest students. Japan offers the same subsidized tuition rates to international students as domestic students, while Singapore offers all students tuition grants covering up to 80 percent of tuition fees as long as they commit to working in Singapore for 3 years after graduation. France and Japan have also strengthened and expanded their scholarship programs for international students. Some countries’ recruiting efforts include providing scholarships to international students who may not be able to afford the costs of obtaining a higher education degree in the United States. In addition, some countries have also developed strategic plans or offices that address efforts to attract international students. The German Academic Exchange Service and EduFrance offer examples where government agencies have been tasked with international student recruitment. Participants at GAO’s forum on global competitiveness expressed concerns that the United States lacked such a national strategy for recruiting international students and emphasized a need to both explore new sources of international students as well as cultivate U.S. domestic capacity. As the cost of attending college in the United States rises, international students may be discouraged from coming here to study. Higher education in the United States ranks among the most expensive in the world. As shown from OECD data in table 1, in 2003-2004 annual average tuition at public U.S. colleges and universities ($4,587) was second only to Australia ($5,289) and more than 2.5 times higher than Europe’s system with the highest tuition fees, that of the United Kingdom. In terms of private higher education providers, U.S. institutions ranked the highest at more than $17,000 per year followed by Australia ($13,420), Italy ($3,992), and Portugal ($3,803). Moreover, student costs at U.S. colleges and universities continue to rise. Figure 3 depicts average undergraduate tuition and room and board costs between 1976 and 2004 for full-time students in degree-granting programs at both 4-year public and private higher education institutions as well as public 2-year institutions. Average costs for private colleges and universities have risen the most since 1990, from $13,237 to $26,489. However, in percentage terms the most growth took place at 4-year public institutions; the change between 1990 and 2004 was approximately 118 percent compared to a 100 percent increase at 4-year privates and an 83 percent increase at 2-year institutions. International students generally do not rely on U.S. federal funding to study in the United States. According to the Institute of International Education’s Open Doors 2004/2005 report, which provides data on international student mobility patterns from U.S. universities, an estimated 71 percent of all international students reported their primary source of funding coming from personal and family sources or other sources outside of the United States. The effects of high and rising tuition and other factors on international enrollment patterns are difficult to estimate, but some policymakers are concerned that costs may be discouraging some international students from coming to U.S. higher education institutions. After September 11, State and Homeland Security, as well as other agencies, took various steps to strengthen the visa process as an antiterrorism tool. This has made the visa process more robust, but may have contributed to real and perceived barriers for international students as well as fueled perceptions that international students were not welcome. Almost all visa applicants must now be interviewed by a consular adjudicating officer at a U.S. embassy or post; this requirement has both affected the number of visas issued and extended wait times for visas under certain circumstances. We have reviewed aspects of the visa process and have made many recommendations to strengthen the process in a way that reduces barriers for international students while balancing national security interests. In October 2002 we cited the need for a clear policy on how to balance national security concerns with the desire to facilitate legitimate travel when issuing visas and made several recommendations to help improve the visa process. In 2003, we reported that the Departments of State, Homeland Security, and Justice could more effectively manage the visa process if they had clear and comprehensive policies and procedures as well as increased agency coordination and information sharing. In 2005 we reported on State’s management of J-1 exchange programs. Separately in 2005, we reported on the department’s efforts to improve the time required to process visas for international science students and scholars as well as others. In 2004 we found that the time to adjudicate a visa depended largely on whether an applicant had to undergo a Visas Mantis security check. Visas Mantis security checks target foreigners who might be involved in violation or evasion of U.S. laws by exporting goods, software, technology, or sensitive information, aiming to prevent proliferation of weapons of mass destruction and conventional weapons. Between January 2004 and June 2006, almost 28 percent of all visa applications sent for Mantis security checks were for students or exchange participants. State has acknowledged that long wait times may discourage legitimate travel to the United States, potentially costing the country billions of dollars in economic benefits, including from foreign students, and adversely influencing foreign citizens’ impressions and opinions of our nation. Much progress has been made over the years with respect to the visa process. Since 2002, State and other agencies have implemented many of our recommendations aimed at strengthening the visa process as an antiterrorism tool while improving processes to facilitate legitimate travel. In particular, State has issued standard operating procedures, in consultation with Homeland Security, to inform consular officers on issues such as special security checks and student visa requirements. In 2005, we reported a significant decline in both Visas Mantis processing times and cases pending more than 60 days. Recent visa data show an increase in the number of student visas issued in the last few years. According to State Department data, the combined student visa issuance levels for fiscal year 2006 increased by about 20 percent from fiscal year 2002. See figure 4 for the issuance trends for individual student visa categories. Broader efforts to facilitate travel to the United States for international students have also been implemented. State has expedited interviews for students. In addition, the length of time that some visa clearances are valid has been extended. In February 2007, State issued guidance to posts that applicants should receive an appointment for a student visa interview within 15 days or less. We are continuing to study aspect of these issues, including visa delays and Visas Mantis security checks, which we will be reporting on in the coming months. The United States must maintain an appropriate balance between protecting national security interests and ensuring our long-term competitiveness. The United States has relied on undergraduate and graduate students from other countries to support both economic and foreign policy interests. Changes designed to protect national security in the wake of September 11 may have contributed to real and perceived barriers for international students, and the subsequent decline in international enrollments raises concerns about the long-term competitiveness of U.S. colleges and universities. Rising U.S. tuition costs and growing higher education options worldwide further demonstrate that the United States cannot take its position as the top destination for international students for granted. While federal efforts to reduce barriers for international students have helped, monitoring current trends and federal policies is essential to ensuring that the United States continues to obtain talented international students in the face of greater global competition. Mr. Chairman, this concludes my prepared statement. I would be happy to respond to any questions you or other members of the subcommittees may have at this time. For further information regarding this testimony, please contact me at (202) 512-7215. Individuals making key contributions to this testimony include Sherri Doughty, Carlo Salerno, Marissa Jones, John Brummet, Eugene Beye, Carmen Donohue, Eve Weisberg, Melissa Pickworth, and Susannah Compton. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

More international students obtain a higher education in the United States than in any other country, and they make valuable contributions while they are here. For those students returning home after their studies, such exchanges support federal public diplomacy efforts and can improve understanding among nations. International students have earned about one-third or more of all U.S. degrees at both the master's and doctoral levels in several of the science, technology, engineering, and mathematics fields. Yet recent trends, including a drop in international student enrollment in U.S. colleges and universities, and policy changes after September 11, 2001, have raised concerns about whether the United States will continue to attract talented international students to its universities. This testimony is based on ongoing and published GAO work. It includes themes from a September 2006 Comptroller General's forum on current trends in international student enrollment in the United States and abroad. Invitees to the forum included experts from the Congress, federal agencies, universities, research institutions, higher education organizations, and industry. GAO identified key issues that may affect the United States' ability to continue attracting the world's most talented international students to our universities and colleges. First, the global higher education landscape is changing and providing more alternatives for students, as other countries expand their educational capacity and technology-based distance learning opportunities increase. For example, enrollment in college-level distance education has nearly quadrupled since 1995. In addition, U.S. universities are establishing branch campuses in other countries and partnerships with international institutions, allowing international students to receive a U.S. education without leaving home. Greater competition has prompted some countries to offer courses in English and to expand their recruiting activities and incentives. Some countries also have developed strategic plans or offices focused on attracting international students. Second, the cost of obtaining a U.S. degree is among the highest in the world and rising, which may discourage international students. Average tuition in 2003 at public U.S. colleges and universities was second only to Australia. Moreover, tuition and associated costs continue to rise. While the effects of high and rising costs and related factors are difficult to estimate, some policymakers are concerned they may be discouraging international students from coming to the United States. Lastly, visa policies and procedures, tightened after September 11 to protect our national security, contributed to real and perceived barriers for international students. Post-September 11 changes included a requirement that almost all visa applicants be interviewed, affecting the number of visas issued and extending wait times for visas under certain circumstances. GAO has made several recommendations to strengthen the visa process in a way that reduces barriers for international students while balancing national security, and recent changes have improved the process. Processing times for certain security reviews have declined, and recent data show more student visas issued in the last few years. The Department of State also has taken steps to ease the burden on students, including expediting interviews and extending the length of time that some visa clearances are valid. We are continuing to study aspects of these issues. The United States must maintain an appropriate balance between protecting national security interests and ensuring our long-term competitiveness. Monitoring current trends and federal policies is essential to ensuring that the United States continues to obtain talented international students in the face of greater global competition.

ICPs provide services associated with the acquisition, distribution, maintenance, and disposal of consumable and reparable parts, and supplies needed to operate weapon systems and components. DLA manages 5 ICPs at 5 locations, and the services manage 11 ICPs at 13 locations. DLA’s ICPs manage consumable items such as repair parts, personnel support items, fuel, and other bulk items and material. The services’ ICPs manage reparable components, subsystems, and assemblies and selected consumable items. The 16 ICPs employ about 24,000 people and manage parts valued at approximately $69 billion. The number of ICPs is expected to be reduced to 11 ICPs at 13 locations by fiscal year 2003.(See app. I for a list of service and DLA ICPs by location and by those that are scheduled for downsizing.) In past reports, we criticized DOD’s logistics system as being cumbersome, inefficient, and costly. Likewise, since at least the 1970s, DOD has recognized and been concerned about overlap and duplication in its logistics system and other inefficiencies. In 1989, OSD proposed a review to consolidate ICPs under a single service or agency manager, but the services strongly opposed the idea because they believed their ability to support weapon systems effectively would be adversely affected. However, in the National Defense Authorization Act for Fiscal Year 1996, Congress required the Secretary of Defense to review the management of all DOD ICPs by DLA, including service-managed reparable items. Thus, in April 1996, the Deputy Under Secretary of Defense for Logistics tasked the Logistics Management Institute (LMI) to conduct such a review. On November 19, 1996, OSD reported the results of its review to Congress and provided a copy to the Comptroller General of the United States. The report concluded that cumulative savings during fiscal years 1998 to 2010, ranging from $2.2 billion to $3.8 billion, might accrue if the management of all ICPs were transferred to DLA. The report also noted the services’ concerns regarding the transfer, principally the risk of disrupting the intraservice integration of material and weapon system management. The report noted, however, that actions could be taken to lessen the risks. Given the services’ concerns, the report stated that DOD, through its QDR and other future planning and programming efforts, would examine alternatives that might provide similar savings at less overall risk. LMI developed a scenario for consolidating the service ICPs under a single manager within DLA and identified the associated potential costs, benefits, and risks. LMI recognized that if the proposed consolidation were to occur, the implementation might differ from its scenario, and the major personnel reductions and site consolidations envisioned in the review would likely have to undergo a process similar to that recently used for BRAC actions. Therefore, LMI considered its analysis conceptual in nature because it did not address specifics, such as which ICPs to close and which to retain. The analysis was intended to indicate only whether the consolidation has merit. Under LMI’s scenario, the consolidation would take place during fiscal years 1998-2010, reduce the number of ICPs to either six or three, and affect at least 12,000 people. Figure 1 is a chronology of LMI’s scenario, the actions projected to occur, and the associated range of savings. No actions would be scheduled during this period of steady-state savings; the movement of ICP personnel would be completed by fiscal year 2008. $0.7 billion to $1.2 billion saved DLA would reduce the number of ICPs and standardize systems and procedures. Remaining business process improvements would be implemented. $0.9 billion to $1.6 billion saved Under DLA management, service ICPs would continue with the same service people, policies, systems, and procedures (i.e., transfer in place). DLA could elect to consolidate support functions regionally or at a single site to reduce the number of personnel required. Some business process improvements would be implemented. $0.6 billion to $1.0 billion saved A 1-year period of decision-making and pre-implementation planning. To identify the cost savings of its scenario, LMI considered three areas through which savings were possible: (1) a transfer in place, (2) site consolidation, and (3) business process improvements. (See app. II for a list of the business improvements identified by LMI.) LMI developed the cost savings for the transfer in place and site consolidations using the services’ and DLA’s ICP and supporting headquarters cost data. For the business process improvements, however, LMI could not obtain complete data from the services for all 16 improvements, but was able to price 4 individual initiatives that would result from the transfer. To develop the potential cost savings in these areas, LMI used cost factors and made assumptions that were conservative in nature. According to an LMI official, the team’s conservative approach was designed to avoid overstating the anticipated cost savings. After examining the report on consolidation, we believe OSD’s approach was reasonable, given the sensitive nature of the issue, the limited amount of time to perform the review, and the data available. However, we concluded that the cost savings estimates would have been $1.3 billion to $2.3 billion greater if BRAC principles had been used. Also, indications are that the savings estimates would be even greater if the review included the savings associated with all 16 business process improvements and likely future improvements to the material management information systems. Full achievement of these additional savings is dependent on the consolidation of the ICPs under a single manager. Given the short time frame LMI had to review the ICP consolidation, it performed a conceptual analysis to show whether savings were possible. It did not use the cost of base realignment actions (COBRA) model, which was used during the four BRAC rounds since 1988 to evaluate the cost of stationing alternatives. Although LMI was not required to use the model, COBRA was the proven, standard means for analyzing proposed consolidations. We recognize the difficulty in using the COBRA model because it requires the collection of a large amount of data and numerous assumptions, such as which sites to retain and which to close. Had LMI used some of the BRAC principles that were used in the COBRA model, such as a longer period of steady-state savings and a present value analysis in arriving at its cost savings estimates, the combined effect would have resulted in larger estimated savings. More importantly, using these BRAC principles provides a way of showing cost savings estimates that are consistent with how DOD projected costs and savings in previous BRAC rounds. To illustrate, BRAC legislation required that consolidations be completed in no more than 6 years and that DOD project savings over a 20-year period, thus ensuring at least 14 years of steady-state savings. BRAC also required the use of a present value analysis to reflect the value of money over time. LMI projected cost savings over a 13-year period (i.e., fiscal years 1998-2010), which included an 11-year implementation period and 2 years of steady-state savings. Its analysis also did not consider the time value of money. An LMI official told us that, given more time, it would have considered using a present value analysis and a longer time period. We adjusted LMI’s cost savings estimates by applying these two BRAC principles without changing LMI’s scenario or assumptions. Specifically, we extended LMI’s ending time frame from fiscal year 2010 to 2022 to allow 14 years of steady-state savings and performed a present value analysis on LMI’s cost savings estimates, using a rate of 4 percent. Table 1 shows the results of our adjustments. LMI’s analysis could be adjusted in many ways if the scenario assumptions were changed. We could have used a 20-year period (fiscal years 1998-2017), which would include 14 years of steady-state savings. Although we believe this alternative calculation would generate savings similar to or greater than those from our analysis, we would have had to make numerous assumptions about LMI’s consolidation scenario. For example, by achieving consolidation within the first 6 years (i.e., between fiscal year 1998 and 2003, or sooner), DOD could increase the potential cost savings even more. We have previously reported on the effect of implementing BRAC actions sooner and the resulting increase in savings. The savings identified in LMI’s analysis do not include potential savings from all 16 business process improvements and a DOD-wide material management information system. We were unable to quantify these associated costs and savings, but we believe their inclusion into LMI’s analysis would increase LMI’s cost savings estimates. Although LMI identified 16 business process improvements from which savings could be anticipated, it estimated costs and savings for only 4. These four, however, account for a significant portion of the overall estimated savings—ranging between $1.5 billion and $2.7 billion. Nevertheless, the additional 12 could also result in savings. According to LMI officials, these business process improvements are a sample of improvements that DLA could make as a single manager for all DOD ICPs, to include improving the contracting methodology and process, deleting inactive parts, and improving material acquisitions and inventory storage. According to an LMI official, LMI estimated savings for only four improvements because of the lack of data, time constraints, and limited resources. Service officials stated that the savings associated with these four process improvements duplicate ongoing service efforts and should not be considered in this analysis. However, they did not provide data to support their statements. We believe that even greater savings could be achieved if the business process improvements were implemented by a single manager across service lines for all of DOD’s ICPs. At the time of LMI’s analysis, DOD was planning to implement the Material Management Standard System to be used at its ICPs. In July 1995, DOD estimated it would spend about $5.3 billion to develop, deploy, and maintain the system at its ICPs, and it expected the effort to produce as much as $15 billion in savings over a 15-year period. According to an LMI official, Material Management Steering Group officials told the LMI team not to consider using these numbers because of the questionable costs and savings estimates. We later reported that DOD had underestimated the costs and overestimated the savings. Because of difficulty in developing the system, the strategy to develop and implement a standard material management system was abandoned. According to a former senior official involved in the development of the system, progress was marred by incompatible service goals that could be overcome if the ICPs were consolidated under a single organization such as DLA. DOD officials told us that they did not believe a standard system would work, considering the differences in how each service does business. However, LMI and several military officials said that a standard database that could be shared was needed. Although the costs and savings associated with a standard system are not easily quantifiable, we believe that successful implementation of a standard system or database would be more likely and savings would be achievable under a single organization. The National Defense Authorization Act for Fiscal Year 1997 established the QDR to examine defense requirements and strategy and develop a revised defense program through 2005. The act also established an NDP to review the QDR’s work and provide you with recommendations for improvements to the QDR’s review, which it did on May 15, 1997. In addition, the NDP will report to you on additional matters by December 1, 1997. DOD established a QDR Infrastructure Panel Logistics Task Force to examine DOD’s infrastructure issues, including ICP consolidation alternatives. The Logistics Task Force considered six alternatives (see app. III for a list of all six alternatives) and decided against consolidating service ICPs and reparable inventory under DLA, even though the savings estimates were much greater than any other alternative. Instead, the task force recommended establishing one ICP per service with multiple locations. Only the recommended alternative was forwarded to the NDP for its consideration. In the NDP’s May 15, 1997, report, the NDP reported on the QDR’s changes and reductions to DOD’s infrastructure but did not specifically address ICP consolidation. According to an NDP staff member, DOD infrastructure issues are still being considered by the Panel, but it is uncertain whether ICP infrastructure will be addressed in the NDP’s December 1, 1997, report. Although substantial savings are possible by consolidating the services’ ICPs under DLA, the services have resisted such proposals, citing potential risks that could affect operational effectiveness. Given this situation, we recommend that you ask the NDP to examine the savings and risks associated with ICP consolidation under DLA. As you know, 31 U.S.C. 720 requires the head of a federal agency to submit a written statement on actions taken on this recommendation to the Senate Committee on Governmental Affairs and the House Committee on Government Reform and Oversight not later than 60 days after the date of the report and to the Senate and House Committees on Appropriations with the agency’s first request for appropriations made more than 60 days after the date of the report. DOD generally concurred with our findings, but stated that without addressing the risks associated with the consolidation, our cost savings projections would not be very meaningful. (See app. IV for a reproduction of DOD’s comments.) We agree with DOD that the risks cannot be ignored. However, as indicated in the OSD report, these potential risks can be mitigated. Given these circumstances, we believe that the NDP should examine both the savings and risks associated with the consolidation of ICPs under DLA. Although this recommendation was not in the draft report DOD reviewed, our subsequent review of the QDR and NDP reports prompted us to add this recommendation. During our review, we evaluated matters related to the cost of the proposed transfer of service-managed ICPs to DLA. We did not address the risks associated with the proposed transfer, nor did we examine any of DOD’s ongoing initiatives in the logistics infrastructure area. However, we did obtain some information on pertinent matters considered by the Logistics Infrastructure Panel of the QDR. To obtain an overall service perspective on the cost aspects of the report, we held discussions with cognizant officials from OSD; the Joint Chiefs of Staff; and headquarters and installations of the Army, Navy, Marine Corps, Air Force, and Defense Logistics Agency, and reviewed documents provided by the services. Locations visited included the Communications and Electronics Command, Fort Monmouth, New Jersey; the Naval Inventory Control Point and Naval Supply Systems Command, Mechanicsburg, Pennsylvania; Naval Sea Systems Command, Washington D.C.; Air Force Materiel Command, Dayton, Ohio; and Oklahoma City Air Logistics Center, Oklahoma City, Oklahoma. To understand the report’s methodology for estimating costs, we talked with OSD and LMI officials, reviewed LMI-prepared data and spreadsheets, and randomly checked LMI’s calculations. To estimate additional potential cost savings, we adjusted LMI’s data to include a longer time period of steady-state savings and a present value analysis. We conducted our review between December 1996 and June 1997 in accordance with generally accepted government auditing standards. We are sending copies of this report to the Chairmen and Ranking Minority Members of the Senate Committee on Armed Services and the House Committee on National Security. We will make copies available to others on request. Please contact me on (202) 512-8412 if you or your staff have any questions about this report. Major contributors to this report were George Jahnigen, Kevin Perkins, and David Epstein. Improved contracting methodology and process: Improves contracting efficiency by emphasizing corporate contracting and reduced acquisition lead times. Deletion of inactive items: Deletes from DOD’s catalog items for which no current applications have been identified, thereby reducing item management costs. Catalog total quality management: Corrects catalog data, which will facilitate correct requirements computations and decisions to repair or procure. Improved demilitarization: Corrects coding errors dealing with demilitarization responsibilities and facilitates timely disposal of excess material. Improved stock positioning: Uses better data on requisitioner locations to reposition stock and decreases shipping and storage costs and response time. Item reduction and entry control: Reviews items during weapon system design phase to identify all equivalent items, leading to reductions in items to be managed and inventory investments. Secondary-item provisioning on end-item contracts: Establishes a DOD program to deal with provisioning line items with end items, thereby reducing procurements and potentially reducing prices as administrative costs are reduced. Source breakout: Strengthens DLA’s program to identify subcontractors and other less costly sources of supply. Workloading of depot maintenance: Provides maintenance depots with better reparable parts induction scheduling, resulting in reduced inventories. Integration of initial and replenishment requirements: Integrates requirements procedures used by program managers to combine computation of initial inventory and replenishment levels. Single set of ICP policies and procedures: Eliminates current duplication of policies and procedures among the services and DLA for secondary items, thereby generating personnel savings. Integration of wholesale and retail requirements: Reduces wholesale and retail inventory investment by using procedures that integrate wholesale and retail responsiveness and inventory costs. Reduction of service-unique catalog data: Eliminates unique service management codes, thus reducing costs associated with data management. Single design activity for materiel management system: Combines into one DLA activity the activities of service design agencies that develop and maintain service-unique software for managing secondary items. Single ICP managing items on a weapon system: Realigns item management along weapon system lines, eliminating file duplication and facilitating computations using weapon system readiness goals. Uniform credit policy for returns: Establishes a single policy for giving credit to organizations returning materiel, thereby simplifying budgeting and accounting at customer levels and industrial fund accounting. Consolidation of selective ICP functions at a single site within a region. One wholesale manager for a common-use reparable item (or for similar common use reparable items). Electronic networking and tasking to link ICPs and provide for a mechanism for executing partnership (intra- or inter-component). Reduction of each DOD component’s ICPs (e.g., 1 ICP per service and 1 or 2 ICPs for DLA). Single management element Assignment of ICP management to all services, except the Marine Corps, along weapon system lines (e.g., Air Force - aircraft, Navy - ships, and Army - ground equipment). Management of all DOD ICPs under DLA. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

GAO reviewed the Office of the Secretary of Defense's (OSD) report on its review of the Defense Logistics Agency's (DLA) management of all Department of Defense (DOD) inventory control points (ICP). GAO noted that: (1) OSD used conservative assumptions and cost factors in estimating cost savings from consolidating service ICPs under DLA; (2) its projected cost savings of $2.2 billion to $3.8 billion cover a 13-year period, fiscal years 1998 to 2010; (3) GAO believes this approach to be reasonable, given the sensitive nature of the issue, the limited amount of time to perform the review, and the data available; (4) however, the projected cost savings estimates would be at least $1.3 billion to $2.3 billion greater if OSD used base realignment and closure principles, such as estimating steady-state savings over a longer time period and a present value analysis instead of a constant dollar analysis; and (5) the potential savings would likely be greater yet if the analysis included: (a) savings from all business process improvements related to the consolidation; and (b) planned future improvements to DOD's existing material management information systems.

DOD is a massive and complex organization. Overhauling its business operations will take years to accomplish and represents a huge management challenge. In fiscal year 2005, the department reported that its operations involved $1.3 trillion in assets and $1.9 trillion in liabilities, more than 2.9 million military and civilian personnel, and $635 billion in net cost of operations. For fiscal year 2005, the department was appropriated approximately $525 billion. Large differences between the net cost of operations and amounts appropriated for any given fiscal year are not unusual in DOD. For the most part, they are attributed to timing differences. For example, net cost is calculated using an accrual basis of accounting (revenues and expenses are recorded when earned and owed, respectively) whereas appropriations are recorded on a cash basis (revenues and expenses are recorded when cash is received or paid.) Using the accrual basis versus the cash basis can result in DOD’s reporting of revenues and expenses in different periods. For instance, DOD may have received in 2005 an appropriation for the acquisition of a weapon system but may not incur expenses or make payments from the appropriation until several years later. Also, DOD’s net cost of operations includes non-cash expenses, such as depreciation related to buildings and equipment that will not require cash outlays until several years after the funds were appropriated. In addition, the department’s recording of expenses related to environmental cleanups and pension and retiree health cost liabilities can occur many years before the appropriations to fund payment of those liabilities are received. Execution of DOD’s operations spans a wide range of defense organizations, including the military services and their respective major commands and functional activities, numerous large defense agencies and field activities, and various combatant and joint operational commands that are responsible for military operations for specific geographic regions or theaters of operation. To support DOD’s operations, the department performs an assortment of interrelated and interdependent business functions—using more than 3,700 business systems—related to major business areas such as weapon systems management, supply chain management, procurement, health care management, and financial management. The ability of these systems to operate as intended affects the lives of our warfighters both on and off the battlefield. For fiscal year 2006, Congress appropriated approximately $16 billion to DOD to operate, maintain, and modernize these business systems, and for fiscal year 2007, DOD has requested another $16 billion for this purpose. To assist DOD in addressing its modernization management challenges, Congress included provisions in the Ronald W. Reagan National Defense Authorization Act for Fiscal Year 2005 that were consistent with our recommendations for establishing and implementing effective business system investment management structures and processes. During the past year, DOD has embarked on a series of efforts to transform its business operations and further comply with the act. In February 2005, DOD chartered the DBSMC to oversee transformation. As the senior most governing body overseeing business transformation, the DBSMC consists of senior leaders who meet monthly under the personal direction of the Deputy Secretary of Defense to set business transformation priorities and recommend policies and procedures required to attain DOD-wide interoperability of business systems and processes. In October 2005, DOD also established the BTA that is intended to advance DOD-wide business transformation efforts in general, but particularly with regard to business systems modernization. DOD believes it can better address agencywide business transformation—which includes planning, management, organizational structures, and processes related to all key business areas—by first transforming business operations that support the warfighter while also enabling financial accountability across DOD. The BTA reports directly to the vice chair of the DBSMC—the Under Secretary of Defense for Acquisition, Technology and Logistics—and includes an acquisition executive who is responsible for 28 DOD-wide business projects, programs, systems, and initiatives. The BTA is responsible for integrating and supporting the work of the Office of the Secretary of Defense principal staff assistants, some of whom function as the approval authorities and who chair the business system investment review boards (IRB). The IRBs serve as the oversight and investment decision-making bodies for those business capabilities that support activities in their designated areas of responsibility. Since the first GAO report on the financial statement audit of a major DOD component over 16 years ago, we have repeatedly reported that weaknesses in business management systems, processes, and internal controls not only adversely affect the reliability of reported financial data, but also the management of DOD operations. In March 2006, I testified that DOD’s financial management deficiencies, taken together, continue to represent the single largest obstacle to achieving an unqualified opinion on the U.S. government’s consolidated financial statements. These issues were also discussed in the latest consolidated financial audit report. To date, none of the military services or major DOD components has passed the test of an independent financial audit because of pervasive weaknesses in internal control and processes and fundamentally flawed business systems. DOD’s financial management problems are pervasive, complex, long- standing, deeply rooted in virtually all of its business operations, and challenging to resolve. The nature and severity of DOD’s financial management, business operations, and system deficiencies not only affect financial reporting, but also impede the ability of DOD managers to receive the full range of information needed to effectively manage day-to-day operations. Such weaknesses have adversely affected the ability of DOD to control costs, ensure basic accountability, anticipate future costs and claims on the budget, measure performance, maintain funds control, and prevent fraud, as the following examples illustrate. We found that hundreds of separated battle-injured soldiers were pursued for collection of military debts incurred through no fault of their own, including 74 soldiers whose debts had been reported to credit bureaus, private collection agencies, and the Treasury Offset Program. Overpayment of pay and allowances (entitlements), pay calculation errors, and erroneous leave payments caused 73 percent of the reported debts. We identified numerous problems with DOD’s processes for recording and reporting costs for the Global War on Terrorism raising significant concerns about the overall reliability of DOD’s reported cost data. As discussed in our September 2005 report, neither DOD nor Congress know how much the war was costing and how appropriated funds were spent, or have historical data useful in considering future funding needs. In at least one case, the reported costs may have been materially overstated. Specifically, DOD’s reported obligations for mobilized Army reservists in fiscal year 2004 were based primarily on estimates rather than actual information and differed from related payroll information by as much as $2.1 billion, or 30 percent of the amount DOD reported in its cost report. In March 2006, we reported that DOD’s policies and procedures for determining, reporting, and documenting cost estimates associated with environmental cleanup or containment activities were not consistently followed. Further, none of the military services had adequate controls in place to help ensure that all identified contaminated sites were included in their environmental liability cost estimates. DOD’s reported liability of $64 billion is primarily for the cleanup of hazardous wastes at training ranges, military bases, and former defense sites; disposal of nuclear ships and submarines; and disposal of chemical weapons. These weaknesses not only affected the reliability of DOD’s environmental liability estimate, but also that of the federal government as a whole. Uncertainties in environmental liabilities could materially affect the ultimate cost and timing of cleanup activities. In December 2005, we reported that the Army had not maintained accurate accountability over inventory shipped to repair contractors, thereby placing these assets at risk of loss or theft. Although DOD policy requires the military services to confirm receipt of all assets shipped to contractors, we found that the Army did not consistently record shipment receipts in its inventory management systems. In an analysis of fiscal year 2004 shipment data obtained from two Army inventory control points, we could not reconcile shipment records with receipt records for 42 percent of the unclassified secondary repair item shipments, with a value of $481.7 million, or for 37 percent of the classified secondary repair item shipments, with a value of $8.1 million. These weaknesses in the Army’s ability to account for inventory shipped to repair contractors increase the risk of undetected loss or theft because the Army cannot ensure control over assets after they have been shipped from its supply system. Moreover, inaccurate and incomplete receipt records diminish asset visibility and can distort on- hand inventory balances, leading to unnecessary procurement of items. Over the years, DOD recorded billions of dollars of disbursements and collections in suspense accounts because the proper appropriation accounts could not be identified and charged. Because documentation needed to resolve these payment recording problems could not be found after so many years, DOD requested and received authority to write off certain aged suspense transactions. While DOD reported that it wrote off an absolute value of $35 billion or a net value of $629 million using the legislative authority, neither of these amounts accurately represents the true value of all the individual transactions that DOD had not correctly recorded in its financial records. Many of DOD’s accounting systems and processes routinely offset individual disbursements, collections, adjustments, and correction entries against each other and, over time, amounts might even have been netted more than once. This netting and summarizing misstated the total value of the write-offs and made it impossible for DOD to identify what appropriations may have been under- or overcharged or to determine whether individual transactions were valid. In May 2006, we reported that some DOD inventory management centers had not followed DOD-wide and individual policies and procedures to ensure they were retaining the right amount of contingency retention inventory. While policies require the centers to (1) use category codes to describe why they are retaining items in contingency inventory, (2) hold only those items needed to meet current and future needs, and (3) perform annual reviews of their contingency inventory decisions, one or more centers had not followed these policies. For example, the Army’s Aviation and Missile Command was not properly assigning category codes that described the reasons they were holding items in contingency inventory because the inventory system was not programmed to use the codes. We found that items valued at $193 million did not have codes to identify the reasons why they were being held, and therefore we were unable to determine the items’ contingency retention category. We also found that some inventory centers have held items such as gears, motors, and electronic switches, even though there have been no requests for some of them by the services in over 10 years. By not following policies for managing contingency inventory, DOD’s centers may be retaining items that are needlessly consuming warehouse space, and they are unable to know if their inventories most appropriately support current and future operational needs. In June 2006, we reported that the military services had not consistently implemented DOD’s revised policy in calculating carryover. Instead, the military services used different methodologies for calculating the reported actual amount of carryover and the allowable amount of carryover since DOD changed its carryover policy in December 2002. Specifically, (1) the military services did not consistently calculate the allowable amount of carryover that was reported in their fiscal year 2004, 2005, and 2006 budgets because they used different tables (both provided by DOD) that contained different outlay rates for the same appropriation; (2) the Air Force did not follow DOD’s regulation on calculating carryover for its depot maintenance activity group, which affected the amount of allowable carryover and actual carryover by tens of millions of dollars as well as whether the actual amount of carryover exceeded the allowable amount as reported in the fiscal year 2004, 2005, and 2006 budgets; and (3) the Army depot maintenance and ordnance activity groups’ actual carryover was understated in fiscal years 2002 and 2003 because carryover associated with prior year orders was not included in the carryover calculation as required. As a result, year-end carryover data provided to decision makers who review and use the data for budgeting were erroneous and not comparable across the three military services. The department is provided billions of dollars annually to operate, maintain, and modernize its stovepiped, duplicative, legacy business systems. Despite this significant investment, the department is severely challenged in implementing business systems on time, within budget, and with the promised capability. The Clinger-Cohen Act of 1996 and Office of Management and Budget guidance provide an effective framework for information technology (IT) investment management. They emphasize the need to have investment management processes and information to help ensure that IT projects are being implemented at acceptable costs and within reasonable and expected time frames and that they are contributing to tangible, observable improvements in mission performance. Effective project management and oversight will be critical to the department’s success in transforming its business management systems and operations. Many of the problems related to DOD’s inability to effectively implement its business systems on time, within budget, and with the promised capability can be attributed to its failure to implement the disciplined processes necessary to reduce the risks associated with these projects to acceptable levels. Disciplined processes have been shown to reduce the risks associated with software development and acquisition efforts and are fundamental to successful systems acquisition. While the department invests billions of dollars annually in its business systems, the following examples highlight the continuing problem faced by the department in successfully implementing business systems. Logistics Modernization Program (LMP). In May 2004, we first reported our concerns with the requirements management and testing processes used by the Army in the implementation of LMP and the problems being encountered after it became operational in July 2003. At the time of our initial report, the Army decided that future deployments would not occur until it had reasonable assurance that the system would operate as expected for a given deployment. However, as we reported in June 2005, the Army’s inability to effectively address the requirements management and testing problems hampered its ability to field LMP to other locations. Our analysis disclosed that LMP could not properly recognize revenue or bill customers. Furthermore, data conversion problems resulted in general ledger account balances not being properly converted when LMP became operational in July 2003. These differences remained unresolved almost 18 months later. These weaknesses adversely affected the Army’s ability to set the prices for the work performed at the Tobyhanna Army Depot. In addition, data conversion problems resulted in excess items being ordered and shipped to Tobyhanna. As noted in our June 2005 report, three truckloads of locking washers (for bolts) were mistakenly ordered and received and subsequently returned because of data conversion problems. At the request of the Chairman and Ranking Minority Member of the Subcommittee on Readiness and Management Support, Senate Committee on Armed Services, we have initiated an audit of the Army’s efforts to achieve financial management visibility over its assets. One aspect of this audit will be to ascertain the Army’s progress in resolving the previously identified problems with LMP. Navy Enterprise Resource Planning (ERP). We reported in September 2005 that the Navy had invested approximately $1 billion in four pilot ERP efforts, without marked improvement in its day-to-day operations. The four pilots were limited in scope and were not intended to be a corporate solution for resolving any of the Navy’s long- standing financial and business management problems. The lack of a coordinated effort among the pilots led to a duplication of efforts in implementing many business functions and resulted in ERP solutions that carry out similar functions in different ways from one another. In essence, the pilots resulted in four more DOD stovepiped systems that did not enhance DOD’s overall efficiency and resulted in $1 billion being largely wasted. While the current Navy ERP effort has the potential to address some of the Navy’s financial management weaknesses, its planned functionality will not provide an all-inclusive, end-to-end corporate solution for the Navy. For example, the scope of the ERP project does not provide for real-time asset visibility of shipboard inventory. Asset visibility has been and continues to be a long-standing problem within the department. Furthermore, the project has a long way to go, with a current estimated completion date of 2011, at an estimated cost of $800 million. Defense Travel System (DTS). As we reported in January 2006, DTS continues to face implementation challenges, particularly with respect to testing key functionality to ensure that the system will perform as intended. Our analysis of selected requirements for one key area disclosed that system testing was not effective in ensuring that the promised capability was delivered as intended. For example, we found that DOD did not have reasonable assurance that flight information was properly displayed. This problem was not detected prior to deployment of DTS because DOD did not properly test the system interfaces through which the data are accessed for display. As a result, those travelers using the system may not have received accurate information on available flights, which could have resulted in higher travel costs. Our report also identified key challenges facing DTS in becoming DOD’s standard travel system, including the development of needed interfaces and underutilization of DTS at sites where it has been deployed. While DTS has developed 36 interfaces with various DOD business systems, it will have to develop interfaces with at least 18 additional business systems—not a trivial task. Additionally, the continued use of the existing legacy travel systems at locations where DTS is already deployed results in underutilization of DTS and affects the savings that DTS was planned to achieve. Naval Tactical Command Support System (NTCSS). The Navy initiated the NTCSS program in 1995 to enhance the combat readiness of ships, submarines, and aircraft. To accomplish this, NTCSS was to provide unit commanding officers and crews with information about maintenance activities, parts inventories, finances, technical manuals and drawings, and personnel. According to the Navy, it spent approximately $1.1 billion for NTCSS from its inception through fiscal year 2005 and expects to spend another $348 million from fiscal years 2006 through 2009, for a total of approximately $1.45 billion. As discussed in our December 2005 report, the Navy has not economically justified its ongoing and planned investment in NTCSS on the basis of reliable estimates of future costs and benefits. The most recent economic justification’s cost estimates were not reliably derived, and return on investment was not properly calculated. In addition, independent reviews of the economic justification to determine its reliability did not occur, and the Navy has not measured whether already deployed and operating components of the system are producing expected value. TC-AIMS II. In December 2005, we reported that the Army had not economically justified its investment in TC-AIMS II on the basis of reliable estimates of costs and benefits. TC-AIMS II was intended to be the single integrated system to automate transportation management function areas for the military services. As noted in our report, the most recent economic justification included cost and benefit estimates predicated on all four military services using the system. However, the Air Force and the Marine Corps have stated that they do not intend to use TC-AMIS II. Even with costs and benefits for all four services included, the analysis showed a marginal return on investment; that is, for each dollar spent on the system, slightly less than one dollar of benefit would be returned. The Army estimates the total life cycle cost of TC-AIMS II to be $1.7 billion over 25 years, including $569 million for acquisition and $1.2 billion for operation and maintenance. The Army reports that it has spent approximately $751 million on TC-AIMS II since its inception in 1995. To effectively and efficiently modernize its nonintegrated and duplicative business operations and systems, it is essential for DOD to develop and use a well-defined business enterprise architecture. In July 2001, the department initiated a business management modernization program to, among other things, develop the architecture. We have previously reported on DOD’s long-standing architecture management weaknesses. Despite spending almost 4 years and about $318 million, the architecture did not provide sufficient content and utility to effectively guide and constrain ongoing and planned business systems investments. DOD recognized the weaknesses that needed to be addressed and assigned a new business transformation leadership team in 2005. More specifically, as previously noted, in October 2005, DOD established BTA to advance DOD-wide business transformation efforts in general, but particularly with regard to business systems modernization. DOD’s complex and pervasive weaknesses cannot be fixed with short-term solutions, but require ongoing and sustained top management attention and resources. DOD’s top management has demonstrated a commitment to transforming the department and has launched key initiatives to improve its financial management processes and related business systems, as well as made important progress in complying with legislation pertaining to its business systems modernization and financial management improvement efforts. For example, we reported in May 2006 that DOD released an update to its business enterprise architecture on March 15, 2006, developed an updated enterprise transition plan, and issued its annual report to Congress describing steps taken and planned with regard to business transformation, among other things. These steps address several of the missing elements we previously identified relative to the legislative provisions concerning the architecture, transition plan, budgetary reporting of business system investments, and investment review. Further, we testified that in December 2005 DOD had issued its FIAR Plan, a major component of its business transformation strategy, to guide financial management improvement and audit efforts within the department. In addition, DOD developed SFIS that will be its enterprisewide data standard for categorizing financial information to support financial management and reporting functions. While this progress better positions the department to address the business systems modernization and financial management high-risk areas, significant challenges remain, particularly in implementing its tiered accountability investment approach. A major component of DOD’s business transformation strategy is its FIAR Plan, issued in December 2005. The FIAR Plan was issued pursuant to section 376 of the National Defense Authorization Act for Fiscal Year 2006, which for fiscal year 2006 limited DOD’s ability to obligate or expend funds for financial improvement activities until the department submitted a comprehensive and integrated financial management improvement plan to congressional defense committees that (1) described specific actions to be taken to correct deficiencies that impair the department’s ability to prepare timely, reliable, and complete financial management information; and (2) systematically tied such actions to process and control improvements and business systems modernization efforts described in the business enterprise architecture and transition plan. Further, section 376 required a written determination that each financial management improvement activity undertaken be (1) consistent with the financial management improvement plan and (2) likely to improve internal controls or otherwise result in sustained improvement in DOD’s ability to produce timely, reliable, and complete financial management information. The act also required that each written determination be submitted to the congressional defense committees. The FIAR Plan is intended to provide DOD components with a road map for achieving the following objectives: (1) resolving problems affecting the accuracy, reliability, and timeliness of financial information, and (2) obtaining clean financial statement audit opinions. Similar to the Financial Improvement Initiative, an earlier DOD improvement effort, the FIAR Plan uses an incremental approach to structure its process for examining operations, diagnosing problems, planning corrective actions, and preparing for audit. However, unlike the previous initiative, the FIAR Plan does not establish a specific target date for achieving a clean audit opinion on the departmentwide financial statements. Target dates under the prior plan were not credible. Rather, the FIAR Plan recognizes that it will take several years before DOD is able to implement the systems, processes, and other changes necessary to fully address its financial management weaknesses. This plan is an important and positive step that will help key department personnel to better understand and address its financial management deficiencies. As outlined in its FIAR Plan, DOD has established business rules and an oversight structure to guide improvement activities and audit preparation efforts. In December 2005, the U.S. Army Corps of Engineers, Civil Works, became the first major DOD component to assert, under DOD’s new process and business rules, that its fiscal year 2006 financial statement information was reliable. An independent public accounting firm has been hired to perform this component’s financial statement audit, under the oversight and direction of the DOD Inspector General. However, the effectiveness of DOD’s FIAR Plan, as well as the department’s leadership and business rules, in addressing DOD’s financial management deficiencies will be ultimately measured by the department’s ability to provide timely, reliable, accurate, and useful information for day-to-day management and decision making. Another key initiative is SFIS, which is DOD’s enterprisewide data standard for categorizing financial information to support financial management and reporting functions. DOD has recently completed phase I of the SFIS initiative, which focused on standardizing general ledger and external financial reporting requirements. SFIS includes a standard accounting classification structure that can allow DOD to standardize financial data elements necessary to support budgeting, accounting, cost management, and external reporting; it also incorporates many of the Department of the Treasury’s U. S. Standard General Ledger attributes. Additional SFIS efforts remain under way, and the department plans to further define key data elements, such as those relating to the planning, programming, and budgeting business process area. DOD intends to implement SFIS using three approaches. One approach requires legacy accounting systems to submit detail-level accounting transactions that are to be converted to SFIS-equivalent data elements. The second approach applies to business feeder systems and will require incorporation of SFIS data elements within systems that create the business transactions. Lastly, accounting systems under development, including new enterprise resource planning systems, are required to have the ability to receive SFIS data as part of source transactions and generate appropriate general ledger entries in accordance with the U.S. Standard General Ledger. To help improve the department’s control and accountability over its business systems investments, provisions in the fiscal year 2005 national defense authorization act directed DOD to put in place a specifically defined structure that is responsible and accountable for controlling business systems investments to ensure compliance and consistency with the business enterprise architecture. More specifically, the act directs the Secretary of Defense to delegate responsibility for review, approval, and oversight of the planning, design, acquisition, deployment, operation, maintenance, and modernization of defense business systems to designated approval authorities or “owners” of certain business missions. DOD has satisfied this requirement under the act. On March 19, 2005, the Deputy Secretary of Defense issued a memorandum that delegated the authority in accordance with the criteria specified in the act, as described above. Our research and evaluation of agencies’ investment management practices have shown that clear assignment of senior executive investment management responsibilities and accountabilities is crucial to having an effective institutional approach to IT investment management. The fiscal year 2005 national defense authorization act also required DOD to establish investment review structures and processes, including a hierarchy of IRBs, each with representation from across the department, and a standard set of investment review and decision-making criteria for these boards to use to ensure compliance and consistency with DOD’s business enterprise architecture. In this regard, the act required the establishment of the DBSMC—which serves as the highest ranking governance body for business system modernization activities within the department. As of April 2006, DOD identified 3,717 business systems and assigned responsibility for these systems to IRBs. Table 1 shows the systems by the responsible IRB and component. A key element of the department’s approach to reviewing and approving business systems investments is the use of what it refers to as tiered accountability. DOD’s tiered accountability approach involves an investment control process that begins at the component level and works its way through a hierarchy of review and approval authorities, depending on the size and significance of the investment. Military service officials emphasized that the success of the process depends on them performing a thorough analysis of each business system before it is submitted for higher-level review and approval. Through this process, the department reported in March 2006 that 226 business systems, representing about $3.6 billion in modernization investment funding, had been approved by the DBSMC—the department’s highest-ranking approval body for business systems. According to the department’s March 2006 report, this process also identified more than 290 systems for phaseout or elimination and approximately 40 business systems for which the requested funding was reduced and the funding availability periods were shortened to fewer than the number of years requested. For example, one business system investment that has been eliminated is the Forward Compatible Payroll (FCP) system. In reviewing the program status, the IRB determined that FCP would duplicate the functionality contained in the Defense Integrated Military Human Resources System, and it was unnecessary to continue investing in both systems. According to the department’s fiscal year 2007 IT budget request, approximately $33 million was sought for fiscal year 2007 and about $31 million was estimated for fiscal year 2008 for FCP. Eliminating this duplicative system will enable DOD to use this funding for other priorities. The funding of multiple systems that perform the same function is one reason the department has thousands of business systems. Identifying and eliminating duplicative systems helps optimize mission performance and accountability and supports the department’s transformation goals. Furthermore, based on information provided by BTA program officials, there was a reduction of funding and the number of years that funding will be available for 14 Army business systems, 8 Air Force business systems, and 8 Navy business systems. For example, the Army’s Future Combat Systems Advanced Collaborative Environment program requested funding of $100 million for fiscal years 2006 to 2011, but the amount approved was reduced to approximately $51 million for fiscal years 2006 to 2008. Similarly, Navy’s Military Sealift Command Human Resources Management System requested funding of about $19 million for fiscal years 2006 to 2011, but the amount approved was approximately $2 million for the first 6 months of fiscal year 2006. According to Navy officials, this system initiative will be reviewed to ascertain whether it has some of the same functionality as the Defense Civilian Personnel Data System. Funding system initiatives for shorter time periods can help reduce the financial risk by providing additional opportunities for monitoring a project’s progress against established milestones and help ensure that the investment is properly aligned with the architecture and the department’s overall goals and objectives. Besides limiting funding as part of the investment review and approval process, this process is also resulting in conditions being placed on system investments. These conditions identify specific actions to be taken and when the actions must be completed. For example, in the case of the Army’s LMP initiative, one of the noted conditions was that the Army had to address the issues discussed in our previous reports. In our May 2004 report, we recommended that the department establish a mechanism that provides for tracking all business systems modernization conditional approvals to provide reasonable assurance that all specific actions are completed on time. The department’s action is consistent with the intent of our recommendations. Notwithstanding the department’s efforts to control its business system investments, formidable challenges remain. In particular, the reviews of those business systems that have modernization funding of less than $1 million, which represent the majority of the department’s reported 3,717 business systems, are only now being started on an annual basis. The extent to which the review structures and processes will be applied to the department’s 3,717 business systems is still evolving. Given the large number of systems involved, it is important that an efficient system review and approval process be effectively implemented for all systems. As indicated in table 1, there are numerous systems across the department in the same functional area. Such large numbers of systems indicate a real possibility for eliminating unnecessary duplication and avoiding unnecessary spending on the department’s multiple business systems. While DOD’s recent efforts represent positive steps toward improving financial management and changing DOD’s business systems environment, the department still lacks key elements that are needed to ensure a successful and sustainable business transformation effort. We reiterate two major elements necessary for successful business transformation: (1) a comprehensive, integrated, and enterprisewide business transformation plan and (2) a CMO with the right skills and at the right level of the department for providing the sustained leadership needed to achieve a successful and sustainable transformation effort. Although some progress has been made in business transformation planning, DOD still has not developed a comprehensive, integrated, and enterprisewide strategy or action plan for managing its overall business transformation effort. The lack of a comprehensive, integrated, enterprisewide action plan linked with performance goals, objectives, and rewards has been a continuing weakness in DOD’s business management transformation. Since 1999, GAO has recommended a comprehensive, integrated strategy and action plan for reforming DOD’s major business operations and support activities. DOD’s efforts to plan and organize itself to achieve business transformation are continuing to evolve. Critical to the success of these efforts will be top management attention and structures that focus on transformation from a broad perspective and a clear, comprehensive, integrated, and enterprisewide plan that at a summary level, addresses all of the department’s major business areas. This strategic plan should cover all of DOD’s key business functions; contain results-oriented goals, measures, and expectations that link institutional, unit, and individual performance goals and expectations to promote accountability; identify people with needed skills, knowledge, experience, responsibility, and authority to implement the plan; and establish an effective process and related tools for implementation. Such an integrated business transformation plan would be instrumental in establishing investment priorities and guiding the department’s key resource decisions. DOD’s leadership has recognized the need to transform the department’s business operations. DOD released a major update to its business enterprise architecture in September 2005 and developed an updated transition plan in March 2006 for modernizing its business processes and supporting IT assets. The business enterprise architecture provides a foundational blueprint for modernizing business operations, information, and systems, while the enterprise transition plan provides a road map and management tool that sequences business systems investments in the areas of personnel, logistics, real property, acquisition, purchasing, and financial requirements. However, while the enterprise transition plan is an important step toward developing a strategic plan for the department’s overall business transformation efforts, it is still focused primarily on business systems. Business transformation is much broader; it encompasses areas such as support infrastructure, human capital, financial management, planning and budgeting, and supply chain management. DOD officials acknowledge that the enterprise transition plan may not have all of the elements of an overarching business transformation plan as we envision it. However, they consider the plan to be evolving. DOD continues to lack the sustained leadership at the right level to achieve successful and lasting transformation. We have testified on the need for a CMO on numerous occasions. Because of the complexity and long-term nature of DOD’s business transformation efforts, we reiterate the need for a CMO to provide sustained leadership and maintain momentum. Without formally designating responsibility and accountability for results, choosing among competing demands for scarce resources and resolving differences in priorities between various DOD organizations will be difficult and could impede DOD’s ability to transform in an efficient, effective, and reasonably timely manner. In addition, it may be particularly difficult for DOD to sustain transformation progress when key personnel changes occur. The National Defense Authorization Act for Fiscal Year 2006 directs the department to study the feasibility of a CMO position in DOD. In this regard, the Institute for Defense Analysis has initiated a study and the results are due by December 2006. Further, in May 2006, the Defense Business Board recommended the creation of a Principal Under Secretary of Defense, with a 5 year term appointment, to serve as CMO. Additionally, in July 2006, a major global consulting firm recommended the concept of a chief operating officer be instituted in many federal agencies as the means to help achieve the transformation that many agencies have undertaken. To provide for senior-level leadership, the CMO would serve as the strategic, enterprisewide integrator of DOD’s overall efforts to transform its business operations. The CMO would be an executive level II appointment, with a tenure of 5 to7 years and serve as the Deputy Secretary or Principal Under Secretary of Defense for Management. This position would elevate integrate, and institutionalize the attention essential for addressing key stewardship responsibilities, such as strategic planning, enterprise architecture development and implementation, IT management, financial management reform, and human capital reform while facilitating the overall business management transformation effort within DOD. It is important to note that theCMO would not assume the responsibilities of the undersecretaries of defense, the service secretaries, or other DOD officials for the day-to-day management of the department. Rather, the CMO would be responsible and accountable for planning, integrating, and executing the overall business transformation effort. The CMO also would develop and implement a strategic plan for the overall business transformational efforts. The Secretary of Defense, Deputy Secretary of Defense, and other senior leaders have clearly shown a commitment to business transformation and addressing deficiencies in the department’s business operations. During the past year, DOD has taken additional steps to address certain provisions and requirements of the fiscal year 2005 national defense authorization act, including establishing the DBSMC as DOD’s primary transformation leadership and oversight mechanism, and creating the BTA to support the DBSMC, a decision-making body. However, these organizations do not provide the sustained leadership needed to successfully achieve business transformation. The DBSMC’s representatives consist of political appointees whose terms expire when administrations change. Furthermore, it is important to remember that committees do not lead, people do. Thus, DOD still needs to designate a person to provide sustained leadership and have overall responsibility and accountability for this effort. DOD continues to face two formidable challenges. Externally, it must combat the global war on terrorism, and internally, it must address the long-standing problems of fraud, waste, and abuse. Pervasive, decades-old management problems related to its business operations affect all of DOD’s major business areas. While DOD has taken several positive steps to address these problems, our previous work has uncovered a persistent pattern among DOD’s reform initiatives that limits their overall impact on the department. These initiatives have not been fully implemented in a timely fashion because of the absence of comprehensive, integrated strategic planning; inadequate transparency and accountability; and the lack of sustained leadership. In this time of growing fiscal constraints, every dollar that DOD can save through improved economy and efficiency of its operations is important to the well-being of our nation and the legitimate needs of our warfighters. Until DOD resolves the numerous problems and inefficiencies in its business operations, billions of dollars will continue to be wasted every year. Furthermore, without strong and sustained leadership, both within and across administrations, DOD will likely continue to have difficulties in maintaining the oversight, focus, and momentum needed to implement and sustain the needed reforms to its business operations. In this regard, I would like to reiterate the need for a CMO to serve as the strategic and enterprisewide integrator to oversee the overall transformation of the department’s business operations. Mr. Chairman and Members of the Subcommittee, this concludes my prepared statement. I would be happy to answer any questions you may have at this time. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

The Department of Defense (DOD) bears sole responsibility for eight DOD-specific high-risk areas and shares responsibility for six governmentwide high-risk areas. These high-risk areas reflect the pervasive weaknesses that cut across all of DOD's major business operations. Several of the high-risk areas are inter-related, including, but not limited to, financial management, business systems modernization, and DOD's overall approach to business transformation. Billions of dollars provided to DOD are wasted each year because of ineffective performance and inadequate accountability. DOD has taken some positive steps to successfully transform its business operations and address these high-risk areas, but huge challenges remain. This testimony discusses (1) pervasive, long-standing financial and business management weaknesses that affect DOD's efficiency; (2) some examples that highlight a need for improved business systems development and implementation oversight; (3) DOD's key initiatives to improve financial management, related business processes, and systems; and (4) actions needed to enhance the success of DOD's financial and business transformation efforts. DOD's pervasive financial and business management problems adversely affect the economy, efficiency, and effectiveness of its operations, and have resulted in a lack of adequate accountability across all major business areas. These problems have left the department vulnerable to billions of dollars of fraud, waste, and abuse annually, at a time of increasing fiscal constraint. Further evidence of DOD's problems is the long-standing inability of any military service or major defense component to pass the test of an independent financial audit because of pervasive weaknesses in financial management systems, operations, and controls. To support its business operations, DOD invests billions of dollars each year to operate, maintain, and modernize its business systems. But despite this significant annual investment, GAO has continued to identify business system projects that have failed to be implemented on time, within budget, and with the promised capability. For example, in January 2006, GAO reported on problems with the implementation of the Defense Travel System--a project that was initiated in September 1998. DOD's many high-risk challenges are years in the making and will take time to effectively address. Top management has demonstrated a commitment to transforming the department's business processes. In December 2005, DOD issued its Financial Improvement and Audit Readiness Plan to guide its financial management improvement efforts. Also, DOD has developed an initial Standard Financial Information Structure, which is DOD's enterprisewide data standard for categorizing financial information. Because of the complexity and long-term nature of DOD transformation efforts, GAO would like to reiterate two missing critical elements that need to be in place if DOD's transformation efforts are to be successful. First, DOD should develop and implement a comprehensive, integrated, and enterprisewide business transformation plan. Second, GAO continues to support the creation of a chief management officer, with the right skills and at the right level within the department, to provide the needed sustained leadership to oversee the department's overall business transformation process.

The Office of National Drug Control Policy has reported that international drug trafficking organizations have become sophisticated, multibillion-dollar industries that quickly adapt to new U.S. drug control efforts. According to Customs’ Strategic Plan: Fiscal Years 1997 - 2002, drug smugglers have moved from (1) using small planes and fast boats to smuggle drugs into the Southeastern United States in the early 1980s, to (2) using commercial cargo and international carriers in the mid- to late-1980s, and (3) exploiting the Southwest border in the 1990s. In addition to collecting revenue from international trade, the mission of the Customs Service is to enforce customs and related laws. It also processes persons, carriers, cargo, and mail into and out of the United States. One of Customs’ major goals is to prevent the smuggling of drugs into the country by creating an effective drug interdiction, intelligence, and investigation capability that disrupts and dismantles smuggling organizations. Customs performs its mission with a workforce of about 19,000 personnel at its headquarters in Washington, D.C., and at 20 Customs Management Centers, 20 investigative offices, and 301 ports of entry around the country. Of the 301 ports, 24 are located along the Southwest border and—through 39 crossing points (such as bridges)—handle both passengers and commercial cargo entering the United States. At the end of fiscal year 1997, Customs had deployed about 28 percent of its inspectors and about 62 percent of its canine enforcement officers at ports along the Southwest border. The Commissioner of Customs has designated drug enforcement to be Customs’ highest priority. As 1 of more than 50 federal agencies involved in the national drug control effort, Customs is responsible for stopping the flow of illegal drugs through the nation’s ports of entry. Customs’ inspectional, investigative, intelligence, canine, marine, and air interdiction assets combine with the efforts of other agencies to reduce the supply of narcotics coming into the country. In addition to routine (primary) inspections to search passengers, cargo, and conveyances (including cars, buses, trucks, aircraft, and vessels), Customs’ drug interdiction efforts include (1) preprimary and postprimary inspections; (2) a more thorough, intensive inspection (secondary) of suspicious shipments or those automatically selected by Customs’ computer system; (3) canine enforcement inspections; (4) inspections using X-ray machines for cargo and trucks; and (5) inspections of randomly selected groups of vehicles using canines and other inspection tools. Line Release is one of two programs the Customs Service is using at its land-border cargo ports to segregate low-risk shipments from other shipments. The Line Release Program was established on the assumption that port officials would know enough about the companies—the brokers, importers, and manufacturers—that participated in the program to assume that they would be unlikely to smuggle drugs. Our review indicated that the internal controls over the Line Release Program at three ports were lax and that port officials could not be reasonably assured that companies approved as low risk under this program should have been designated as such and afforded the benefits that go with it. The theory behind the Line Release Program is that companies that routinely import goods through a port and are generally in compliance with trade laws and Customs regulations pose significantly less risk for drug smuggling than other companies. Customs believes that if ports could identify and designate certain companies as low risk for drug smuggling, inspectors would have more assurance that these companies’ shipments may pose a lower risk than those of other companies. Truck drivers transporting shipments for companies approved under the Line Release Program are not required to stop inside the port at the dock to process paperwork; this procedure expedites their entry processing, frees dock space for trucks that are required to stop, and allows inspectors to focus their attention on higher-risk shipments. The Line Release Program was first implemented in 1986 on the Northern border and was designed to expedite the release and tracking of high-volume, low-risk shipments by prescreening manufacturers, brokers, and importers to ensure that they did not present a threat of drug smuggling. In 1987, Customs began implementing the Line Release Program at cargo ports along the Southwest border; by the end of 1989, most of the major Southwest border cargo ports had fully implemented the program. Customs’ regulations for the Line Release Program, which became effective in 1992, are published in the Code of Federal Regulations.Customs port directors are responsible for screening, reviewing, and approving Line Release applicants. Program applicants are required to complete and submit an application to the port director for review and approval. In 1993, Customs issued guidance on the administration and use of the Line Release Program. According to Customs, this guidance combined all the Line Release policies and procedures issued since the program’s inception. The guidance instructed port directors to establish their own procedures for screening, reviewing, and approving applications; and it suggested that specific port personnel, such as import specialists, review the applications. The guidance did not specify what criteria ports should consider in approving applicants or what the reviews should entail, although it did state that the purpose of the reviews would be to conduct risk analyses of applicants to determine if they qualified for Line Release. The guidance did not require port officials to maintain any specific documentation on the review and approval process. In August 1997, Customs developed national Draft Line Release Quality Standards that, among other things, established volume and compliance eligibility criteria for program applicants and recertification standards for program participants. The volume criterion proposes that applicants should have had “at least 50 shipments . . . within the previous 12 months prior to the filing of the application.” The compliance criterion proposes that the applicants should have had “at least five Customs intensive examinations with no discrepant findings, or more than five Customs examinations with no more than a 10 percent discrepancy rate.” The recertification standards propose that Line Release participants be reviewed at least annually to ensure they have had 50 shipments within the preceding 12 months. Biennially, participants are to be reviewed to ensure they have met a minimum compliance rate of 90 percent. In May 1998, Customs convened a Line Release Conference in San Diego, CA, during which representatives from northern and southern land-border cargo ports discussed, among other things, the above eligibility criteria and recertification standards, and agreed to finalize and issue the Line Release Quality Standards at the end of fiscal year 1998. Each of the three ports we reviewed had developed a two-part process for screening, reviewing, and approving Line Release applicants, although the eligibility criteria and review procedures differed somewhat among the ports. Part one of each port’s process involved screening applicants to determine whether they met the port’s eligibility criteria for participating in Line Release—a high-volume of shipments each year and a history of compliance with trade laws and Customs commercial importing regulations. The second part of each port’s process involved several components, including (1) verification of the data submitted by the applicants (i.e., verification of company name, address, identification number, etc.); (2) review of the application by an import specialist to ensure that, among other things, the commodity (merchandise) was properly classified; and (3) a background check on the applicants to ensure they had no past history of drug smuggling. Each port had developed a Line Release checklist that was used to evaluate the applications and track them through this process, as well as a checklist that recorded the results of the background checks conducted on applicants. Although the ports’ Line Release checklists varied, they included some of the same elements, such as approval by an import specialist and the Line Release coordinator. In addition, the checklists used to record and track background checks also included many of the same elements, such as name and address verifications, Internal Revenue Service numbers, and smuggling history. In July 1996, Customs implemented the Land Border Carrier Initiative Program (Carrier Initiative Program). The program requires participating carriers to be prescreened by Customs—through background checks and site visits—and approved as low risk for drug smuggling. At the time this program was implemented, Customs established a new requirement that all Line Release participants (brokers, importers, and manufacturers) on the Southwest border use carriers (trucks and drivers) approved under the Carrier Initiative Program. In fiscal year 1996, cargo entries along the Southwest border totaled 1,408,790 of which 277,382 or about 20 percent, were Line Release entries. In fiscal year 1997, total entries increased by nearly 15 percent to 1,617,445, while Line Release entries dropped by almost 29 percent to 197,344, or about 12 percent of total entries. Customs officials attributed the drop in Line Release entries to the implementation of the Carrier Initiative Program in July 1996. Of the three ports we reviewed, Otay Mesa had the largest number of Line Release entries during fiscal years 1996 and 1997 (see table 1) and, in fiscal year 1997, the greatest number of participants. Officials at the Laredo and Nogales cargo ports told us that Line Release entries dropped significantly at their ports at the time the Carrier Initiative Program went into effect. Laredo and Nogales officials said companies did not want to participate in the program either because they already had contracts with nonprogram carriers or because they did not want to tie themselves to Carrier Initiative—approved carriers, many of whom were located near the border and not the Mexican interior, where many of the commodities were produced. Although each of the three ports we reviewed had developed a process for screening and approving applicants, we found internal control weaknesses in the procedures actually followed. These weaknesses included (1) the lack of specific criteria for determining applicant eligibility at two of the three ports, (2) incomplete documentation of the screening and review of applicants at two of the three ports, and (3) lack of documentation of supervisory review and approval of decisions. We also noted that the ports were not able to locate some of the application files and background checklists that served as support for approving applications, and that two ports had not recertified Line Release companies. Otay Mesa had specific criteria for determining program eligibility and had established a standard review process for assessing Line Release applicants. To be considered for the Line Release Program, applicants were expected to have a minimum of 50 shipments during the 12 months prior to filing an application and to have at least 5 negative examinations.However, until recently the other two ports—Nogales and Laredo—did not have specific criteria by which reviewers were to judge an applicant’s eligibility. Lack of specific eligibility criteria could allow individual reviewers at a port to reach different conclusions about an applicant’s eligibility. The former and current Nogales Line Release coordinators told us that, until recently, Nogales did not have specific eligibility criteria in place for screening Line Release applicants. Instead, each application was to be reviewed on a case-by-case basis. The current Line Release coordinator said that in fiscal year 1997, to screen applicants, the port adopted the volume and compliance criteria specified in Customs’ Draft Line Release Quality Standards—50 shipments within the prior 12 months and at least 5 Customs intensive examinations with no discrepant findings. However, the coordinator could not provide port guidance that addressed this change, nor was the port’s Line Release checklist revised to reflect the new eligibility criteria. The Laredo Line Release coordinator told us that until recently, Laredo had no specific eligibility criteria for volume and discrepancy rates, relying instead on the Line Release coordinator’s subjective evaluation of applicants. The coordinator said that applications are judged on a case-by-case basis and that theoretically all companies are eligible for the Line Release Program, except those that have a history of drug violations. To be approved, Laredo requires that applicants can only have had relatively minor compliance “discrepancies” or violations on their examination records, with no record of drug violations. The Line Release coordinator told us that in fiscal year 1997 Laredo also began using the volume standard cited in Customs’ 1997 Draft Line Release Quality Standards. However, the coordinator could not provide documentation to substantiate this change, nor did the port’s Line Release checklist reflect the new criteria. The Comptroller General’s Standards for Internal Controls in the Federal Government (June 1983) requires that “documentation of transactions or other significant events should be complete and accurate and should facilitate tracing the transaction or event and related information from before it occurs, while it is in process, to after it is completed” and that transactions and other significant events be promptly recorded and properly classified. Officials at all three of the ports we visited said they routinely reviewed applicants’ trade history—specifically, volume and compliance history—as part of their Line Release review process. However, Otay Mesa did not include volume and compliance history on the Line Release checklist, nor did reviewing officials document in the files we reviewed that this information had been verified. In addition, 20 of the 46 Line Release checklists we reviewed at Otay Mesa had not been fully completed. For example, in 12 cases, review officials had failed to check off all applicable review elements. For 8 of 46 Line Release checklists, reviewers had failed to either sign and/or date the checklist. Also, one of the application files did not have a Line Release checklist. None of the files we reviewed contained supporting documentation—the Line Release coordinator told us that the port did not require supporting documentation, such as computer printouts of applicants’ trade histories. The Comptroller General’s Standards for Internal Controls in the Federal Government specifies that “Internal controls systems and all transactions and other significant events are to be clearly documented, and the documentation is to be readily available for examination.” Both Laredo and Nogales had applicants’ trade history as an element to be checked off on either their Line Release or background checklist. At Laredo, 64 of the 65 background checklists we reviewed documented applicants’ trade history—volume of shipments and compliance with Customs regulations. In addition, 69 of the 72 Line Release checklists we reviewed had been completed. At Nogales, the port could locate only one of the seven Line Release checklists associated with the application files we reviewed. Although the applicants’ trade history was documented on the checklist as required, the entire checklist had not been completed. Further, the Nogales Line Release coordinator told us that there was no port requirement to retain supporting documentation for record checks conducted on Line Release applicants; however, one of the application files we reviewed included supporting documentation. Laredo had provided supporting documentation for 65 of the 66 files we reviewed. According to the Comptroller General’s Standards for Internal Controls in the Federal Government, qualified and continuous supervision is to be provided to ensure that internal control objectives are achieved. Assignment, review, and approval of a staff’s work should result in the proper processing of transactions and events including (1) following approved procedures and requirements; (2) detecting and eliminating errors, misunderstandings, and improper practices; and (3) discouraging wrongful acts from occurring or from recurring. We found that aspects of the ports’ Line Release review and approval processes lacked documentation of supervisory review. At Otay Mesa, the Line Release coordinator told us he is responsible for reviewing the Line Release checklists to ensure they have been completed, signed, and dated. The coordinator also said he is responsible for documenting the progress of the application through the approval process but is not required to review other officials’ research. None of the 46 checklists we reviewed documented a supervisory review, either by the coordinator or his supervisor. Further, the operations analyst told us there is no supervisory review required for the background checks he performs on importers and manufacturers. The Laredo Line Release coordinator also told us that he is responsible for ensuring that the port’s Line Release checklists are properly completed. The coordinator said he reviews the research performed on the applications, including the background checks and trade history recorded on the background checklists, but there was no documentation of supervisory review on either the 72 Line Release checklists or the 65 background checklists provided by the port. At Nogales, applications are researched by the Line Release coordinator and others, including import specialists. Although the Line Release checklist provides for the chief inspector to document whether the application was approved or disapproved, the one checklist located by the port did not indicate whether the chief inspector had reviewed the checklist. According to the coordinator, the checklist used to document background checks performed on applicants does not have to have supervisory review. Our work at the three ports raised other issues, which could compromise the integrity of the Line Release Program. First, Nogales officials were unable to locate two of the seven application files for the companies currently using Line Release; in addition, they could only locate one of the seven Line Release checklists identified with the application files. The current and former Line Release coordinators told us the port had not received any Line Release applications since July 1996, when the Carrier Initiative Program went into effect. At Otay Mesa, officials were unable to provide 15 of the background checklists for the 46 Line Release checklists we reviewed; at all three ports, background checks served as the basis for approving applicants. The operations analyst responsible—as of May 1998—for completing the background checklists at Otay Mesa told us that although he is not required to retain copies of the checklists or to provide documentation in support of his findings—e.g., database check printouts—he does both. Second, although neither the Code of Federal Regulations nor Customs’ implementing guidelines require ports to recertify companies already approved for the Line Release Program, Otay Mesa had recertified participants based on their volume criteria. The port does not recheck (recertify) participants for compliance or perform follow-up background checks. Without recertification, there is no assurance that the participants continue to meet the volume and compliance criteria or that they remain low risk for drug smuggling. We verified that Otay Mesa had performed the volume recertifications for the 42 application files we reviewed. These 42 files included 93 commodities; 52 were recertified as meeting Otay Mesa’s volume criteria. The remaining 41 were either inactive or had been on Line Release for less than 12 months. Officials at Laredo and Nogales told us that they are planning to recertify Line Release participants, as required in the Draft Line Release Quality Standards, as soon as the standards are finalized. Customs developed the Three Tier Targeting Program to help identify low- and high-risk shipments so that inspectors along the Southwest border could focus their attention on shipments determined to be high-risk for narcotics smuggling. Low-risk shipments were to receive expedited treatment for release, while high-risk shipments were to be subject to a higher rate of narcotics examinations. Customs headquarters defined how cargo shipments would be divided into three tier categories and allowed the ports to develop their own policies and procedures for assigning risk. Officials at the ports we visited said they did not think the Three Tier Targeting Program was a viable program because it did not appear to have Customs headquarters’ support. They also said they had little confidence in the system as a method for assessing risk because (1) there was little information available in any database for researching foreign manufacturers and (2) they doubted the reliability of the designations: two ports cited examples of narcotics seizures from shipments designated as “low risk” and the lack of a significant number of seizures from shipments designated as “high risk.” In addition, they said that the research necessary to assign and recertify tier designations has been very time consuming given the questionable reliability of the tier designations. In 1992, Customs implemented the Three Tier concept—a method of targeting shipments for narcotics examinations—at Southwest border ports. According to Customs’ draft Three Tier Targeting Directive, this concept was devised to assist ports in classifying shipments according to a narcotics risk assessment so that they could better identify or “target” shipments that were “high risk” for smuggled narcotics. The intent of the program was for ports to better focus inspectional resources. According to a 1994 report by Customs’ Office of Regulatory Audit, ports were to start using the program in April 1992. Under the Three Tier concept, ports were to conduct research on importers and foreign manufacturers who shipped through their ports. The draft directive called for port analysts to check Customs databases and other available sources for information on importers’ and manufacturers’ business histories and criminal activities. Commercial cargo shipments were to be divided into three categories, or tiers, according to perceived risk factors: Tier I: bearing little risk for narcotics smuggling, based on analytical assessment. Tier II: an unknown degree of risk for narcotics smuggling. (All shipments that are not clearly Tier I or Tier III were to fall into Tier II.) Tier III: a significant risk for narcotics smuggling. (Shipments designated as Tier III were to be identified as high risk in Customs’ Automated Commercial System so that inspectors would know they were to receive narcotics examinations.) Officials at the three ports we visited expressed reservations about the viability of the Three Tier Targeting Program. The officials remarked that the program did not appear to have the full support of Customs headquarters because formal program directives were issued in draft but were not finalized. Officials at Customs headquarters could not explain why the Three Tier directive was not finalized. Customs’ current Narcotics Interdiction Guide calls for continued use of the program. Officials at the three ports told us they had little confidence in the program as a method for assessing risk for two reasons. First, program officials said sufficient information is not available to assess the risk of foreign companies. For example, a Nogales official told us that it was impossible to get enough information on Mexican manufacturers on which to base a reliable narcotics risk assessment. He said that no matter how much research was conducted through Customs’ automated databases and other sources, there were no data available on Mexican companies, particularly data identifying those that had been involved in narcotics smuggling. Second, port officials told us that inspectors had become suspicious about the reliability of Three Tier designations. In Laredo, for example, a program official told us the port had made two marijuana seizures from shipments classified as Tier I, or low risk. Conversely, in Nogales, the analyst responsible for the program told us there had been no narcotics seizures found in Tier III, high-risk shipments. Laredo officials also told us that inspectors were more suspicious of shipments classified as low risk because they had doubts about the reliability of the tier designations. These doubts could lead them to order more examinations of low-risk shipments, in direct conflict with the original intent of the program—to process low-risk shipments quickly so that inspectors could focus their attention on high-risk shipments. Port officials also told us that the research necessary to assign and recertify tier designations has been very time consuming given the questionable reliability of the tier designations. In addition to conducting the initial research necessary to assign tier designations, ports are to annually recertify Tier I designations by updating the research. An official at Otay Mesa told us that because of time constraints, port analysts were unable to both recertify companies for Tier I and certify companies for the Line Release Program. He said that in fiscal year 1997, port analysts would have needed to do 50 recertifications per month to keep the database current; but they had only been able to recertify—update the research for—39 Tier I companies for the entire year from a total of 1,576 Tier I designations in their database. According to the port official, other operations, such as providing research support to the port’s investigative team, take priority over Tier I recertifications. At Laredo, a port official told us that for the past two years, the port has continued to maintain the Tier I database but has not added any new companies to the Tier I database. Officials at the three ports said that the Three Tier Targeting Program should be discontinued and that, although the program had worked well in facilitating cargo, it had not been effective in distinguishing between high- and low-risk shipments. In February 1994, Customs had also reported in its Management Review of the Three Tier Targeting Program that “. . . the Three Tier Targeting Program is a good cargo facilitation tool, however, because of the lack of reliable intelligence, it has not been effective in targeting narcotics in cargo shipments . . . .” Port officials told us their inspectors now rely on other cargo entry programs—such as Line Release—to identify shipments that are low risk for drug smuggling. Customs’ 1994 Management Review also stated that “. . . no narcotic seizures have resulted from Three Tier Targeting . . . .” Customs headquarters officials told us that they did not know if any seizures had been made from Tier III, high-risk shipments. Further, they said they did not know whether any of the 61 narcotics seizures in commercial cargo in fiscal year 1997 were made from Tier III shipments. The officials also told us that there is no headquarters oversight of the Three Tier Targeting Program, and consequently no evaluations of the program or measures of success. “Prefile” is a cargo entry process used at the Port of Laredo to expedite low-risk shipments. The Prefile Process, which began in 1989, requires participating brokers to file cargo entry paperwork at least 4 hours prior to a shipment’s arrival at the port. This advance filing is to enable port officials to review the paperwork and perform computerized background checks on the manufacturer, importer, and broker to assess the smuggling risk of each shipment before it arrives at the port. The Prefile Process is complemented by the Automated Targeting System, which evaluates and scores arriving shipments through the use of approximately 400 “rules” designed to identify or profile high-risk shipments. The higher the score, the more the shipment warrants attention. This process is being evaluated to establish its effectiveness. According to Laredo officials, the Prefile Process was designed to expedite processing cargo through the port. Customs officials said it facilitates processing by identifying, before the cargo reaches the port, low-risk shipments that can be released at the primary inspection gate and shipments that should be held at the dock for intensive examinations. Compared with other low-risk cargo entry programs (e.g., Line Release and the Three Tier Targeting Program), which rely on initial research of applicants before they are approved or designated as “low risk,” the Prefile Process involves reviewing the most current—“real-time”—information available on companies and their potential for drug smuggling before the shipments reach the port. Although the databases may not include information on foreign manufacturers, the data accessed is the most current information available. When a broker uses the Prefile Process, the port is to receive the hard-copy entry paperwork—the entry summary, for example—at least 4 hours before the shipment arrives at the port. Under Customs’ standard entry-filing process, drivers park their trucks at the dock and give a hard copy of the entry paperwork to Customs for processing. While the drivers wait, Customs compares the hard copy with entry information that was filed electronically in advance by the broker. Any comparison of the hard copy and the electronic filing for consistency must be conducted while the trucks are parked at the dock. Under the Prefile Process, Customs inspectors are to perform the same reviews of the electronic and hard-copy entry documents that they would do under the standard entry-filing process, including additional research;but receiving the hard-copy entry paperwork in advance allows the port to perform these reviews, and any necessary additional research, before the shipment arrives at the port. If the research does not provide a reason to inspect the shipment, it is to be cleared for release. The inspector in the primary inspection booth can then allow the cleared shipment to proceed directly to the exit gate when it arrives at the port. (For other reasons—such as a driver acting suspiciously or a random, computer-generated order for an inspection—a Customs inspector may order the shipment held at the dock for an intensive examination.) One official estimated that approximately 70 percent of cargo shipments at Laredo are Prefile shipments, although statistics were not maintained to confirm this figure. A disadvantage of the Prefile Process is that unlike the Line Release process, Prefile focuses on the importer, broker, and manufacturer and does not require the use of prescreened carriers. Companies participating in the Line Release Program are required to use preapproved carriers and drivers cleared under the Carrier Initiative Program. According to a September 1997 report from the Office of National Drug Control Policy,76 percent of the seizures made in the Southwest border commercial cargo environment during 1997 were found in the conveyance (truck and trailer), not in the actual cargo. In May 1997 Laredo began pilot testing, in conjunction with the Prefile Process, a computerized system called the Automated Targeting System. The Automated Targeting System assists the port in identifying shipments that could pose a high risk for drug smuggling. According to Customs, the system is designed to help the port prioritize shipments according to threat, in order to allow the port to more effectively use resources and to ensure that shipments that pose the highest risk for smuggling are researched first. The Automated Targeting System standardizes entry and entry-summary data received from the broker and creates integrated records called “shipments.” The shipments are to be evaluated and scored by the Automated Targeting System through the use of approximately 400 weighted “rules” designed to identify or profile high-risk shipments. According to the system’s program officer, the rules are based on targeting and evaluation methods successfully used by experienced Customs inspectors. The higher the score, the more the shipment warrants attention. Customs inspectors may use the score to determine whether the shipment should be detained for inspection after it reaches the port. For example, a shipment going to a “first-time importer” might be selected for an intensive inspection. One of the rules used for scoring a potentially high-risk shipment is a first-time importer because little information is available about first-time importers on which to assess the risk of drug smuggling. The Automated Targeting System also allows Customs inspectors to query several databases simultaneously to conduct background checks on importers, brokers, and manufacturers associated with a shipment. Because data from several systems are displayed on a computer screen at one time, inspectors are able to compare information for potential irregularities and inconsistencies. Customs officials told us that Laredo is the first land-border port of entry to test the Automated Targeting System. Depending on the outcome of Laredo’s pilot test, Customs may expand the system to all major seaports, airports, and land-border ports of entry. Customs is currently evaluating the pilot test at Laredo. According to the system’s program officer, Customs does not plan to expand the system to other land-border, cargo ports of entry until an evaluation has been completed. According to the program officer, the system will be assessed for use at other Southwest border ports on the basis of three factors: (1) drug threat, (2) volume of shipments and method of processing (i.e., Line Release, etc.), and (3) technological capability. Laredo port officials told us they are tracking drug seizures attributed to the Automated Targeting System; as of May 1998, three marijuana seizures had been made, totaling over 5,000 pounds. The Customs Service is faced with the challenge of facilitating the flow of legitimate cargo into the United States while, at the same time, detecting and intercepting illegal drug smuggling. Customs has developed several programs to try to identify shipments that are lower risk than others and give more inspectional attention to the higher-risk shipments. Theoretically, these programs would facilitate the processing of lower risk cargo and enable Customs to use its inspectional resources more efficiently and effectively. The key to the success of these programs is Customs’ ability to identify the risk that any given shipment poses. Our review of three programs at three Southwest border ports raises several concerns about the implementation of two of these programs. The weak internal controls over the Line Release Program at three ports may not assure Customs that program participants, at these three ports, are fully researched and properly designated as low risk. Further, port officials’ concerns about the Three Tier Targeting Program raise questions about the continued value or utility of the program at the three ports we visited. Officials at all three ports said that the program should be discontinued, and that they relied on other programs for distinguishing high- and low-risk shipments. These reasons cause us to conclude that the Three Tier Targeting Program may not be an effective tool for assessing narcotics risk. We recognize that under current operating conditions, Customs will not be able to subject all cargo entering the United States to intensive inspections to detect drug smuggling. We also recognize that inadequately controlled processes for identifying low-risk shipments can give Customs inspectors a false sense of confidence that those shipments are low risk for drug smuggling. While the Prefile Process, used in conjunction with the Automated Targeting System, seems to have the potential to offer the advantage of basing inspection decisions on more current information than the Line Release and the Three Tier Targeting programs, it does not cover the carriers, and has not been thoroughly evaluated. We recommend that the Commissioner of Customs strengthen internal control procedures for the Line Release application and review process to ensure fully researched and documented risk-assessment decisions on applicants; suspend the Three Tier Targeting Program until it can be determined if more complete and comprehensive data are available on which to base “low risk for narcotics smuggling” risk assessments; and evaluate the effectiveness and efficiency of the Automated Targeting System, as designed and implemented at Laredo, and use the evaluation results to determine whether other land-border cargo ports should implement the system or whether additional testing is needed. Treasury provided written comments on a draft of this report, and its comments are reprinted in appendix III. Overall, Treasury and Customs management generally agreed with our conclusions, and Customs is taking action, or is planning to take action, on all of our recommendations. Regarding our first recommendation, Treasury stated in its written comments that Customs’ Office of Field Operations plans to publish the Line Release Quality Standards in the form of a Headquarters Directive by the end of fiscal year 1998. According to Treasury, this directive will create consistent national criteria and guidance with regard to the application procedures. Included will be a requirement for ports to retain the original approved applications and supporting documentation on file for as long as the applicants are active participants in the program. Regarding our second recommendation, Treasury agreed that the Three Tier Program should be suspended until more reliable information is developed for classifying low-risk importations. Treasury stated in its written comments that Customs believes its other targeting methods, including the Line Release Program, the Automated Targeting System, the Prefile Program, and the Land Border Carrier Initiative Program, are better able to fulfill Customs’ narcotic interdiction goals and responsibilities. Regarding our third recommendation, Treasury said Customs is currently evaluating the Automated Targeting System as implemented at the port of Laredo. It also plans to assess data regarding cargo volume and cargo processes used—e.g., Line Release—on other Southwest border ports of entry to determine future deployment of the system. Customs management, in their written comments, acknowledged that the Prefile Process, used in conjunction with the Automated Targeting System, does not require the use of preapproved carriers and drivers cleared under the Carrier Initiative Program and that this could be seen as a disadvantage. They also stated that the Prefile approach narrows the scope of Customs’ interdiction efforts to focus on the driver and conveyance because the cargo has been determined to be low risk. Customs pointed out that Prefile shipments are also subject to other enforcement actions, including (1) random checks performed on all companies using the Prefile Process, (2) X-ray and detection dogs, and (3) the experience and knowledge of Customs inspectors. Nevertheless, we still feel that the fact that the Prefile Process does not require shipments to use carriers preapproved under the Carrier Initiative Program is a significant disadvantage. Line Release shipments are also subject to the same enforcement actions mentioned above. Yet, in July 1996, Customs strengthened the Line Release Program by requiring all participants on the Southwest border to use carriers approved under the Carrier Initiatives Program. We are sending copies of this report to the Secretary of the Treasury, the Acting Commissioner of Customs, and to the Chairmen and Ranking Minority Members of the congressional committees that have responsibilities related to these issues. Copies also will be made available to others upon request. The major contributors to this report are listed in appendix IV. If you or your staff have any questions about the information in this report, please contact me on (202) 512-8777 or Darryl Dutton, Assistant Director, on (213) 830-1000.

Pursuant to a congressional request, GAO reviewed the Custom Service's drug-enforcement operations along the Southwest border of the United States, focusing on: (1) Customs' low-risk, cargo entry programs in use at three ports on the Southwest border--Otay Mesa, California; Laredo, Texas; and Nogales, Arizona; (2) the results of GAO's evaluation of internal controls over the Line Release Program; and (3) the processes used to assess the risk of narcotics smuggling in other cargo entry programs. GAO noted that: (1) to balance the objectives of facilitating trade through ports and interdicting illegal drugs being smuggled into the United States, Customs has initiated and encouraged its ports to use several programs to identify and separate low-risk shipments from those with apparently higher smuggling risk; (2) the Line Release Program was designed to expedite cargo shipments that Customs determined to be repetitive, high volume, and low risk for narcotics smuggling; (3) in 1996, Customs implemented the Carrier Initiative Program, which required that the Line Release shipments across the Southwest border be transported by Customs-approved carriers and driven by Customs-approved drivers; (4) after the Carrier Initiative Program was implemented, the number of Southwest border Line Release shipments dropped significantly; (5) GAO identified internal control weaknesses in one or more of the processes used at each of the three ports it visited to screen Line Release applicants for entry into the program; (6) these weaknesses included: (a) lack of specific criteria for determining applicant eligibility at two of the three ports; (b) incomplete documentation of the screening and review of applicants at two of the three ports; and (c) lack of documentation of supervisory review and approval of decisions; (7) in May 1998, Customs representatives from northern and southern land-border cargo ports approved draft Line Release volume and compliance eligibility criteria for program applicants and draft recertification standards for program participants; (8) the Three Tier Targeting Program--a method of targeting high-risk shipments for narcotics inspection--was being used at the three Southwest border ports that GAO visited; (9) according to officials at the three ports GAO reviewed, the Three Tier program had two operational problems that contributed to their loss of confidence in the program's ability to distinguish high- from low-risk shipments; (10) one new targeting method--the Automated Targeting System--is being pilot tested at Laredo; (11) used in conjunction with the Prefile Program, this system is designed to enable port officials to identify and direct inspectional attention to high-risk shipments; (12) the Automated Targeting System, which automatically assesses shipment entry information for known smuggling indicators, is designed to enable inspectors to target high-risk shipments more efficiently; and (13) Customs is evaluating the Automated Targeting System for expansion to other land-border cargo ports.

To meet property management requirements and provide data for personal property reporting needs, VA field facilities use an inventory accounting system, the Automated Engineering Management System/Medical Equipment Reporting System. The system was originally designed to schedule preventive maintenance. In 1996, the system was expanded to incorporate the agency’s previously separate property management function, becoming the agency’s official record of inventory for capitalized and noncapitalized equipment. VHA’s Acquisition and Materiel Management (A&MM) service maintains the property management portion of the system while Engineering Services operates the property maintenance portion. The property management system includes a detailed listing of the agency’s personal property, providing information that among other things, is (1) a control for the accuracy of property cost information presented in the agency’s financial report, (2) the basis for physical inventories of agency personal property, and (3) the primary control record for accountability over the agency’s personal property. The system is used to prepare bar code labels that are affixed to nonexpendable property acquired by VHA to identify items as VHA property and to provide for efficient physical inventories using portable bar code readers. The property management software has been updated occasionally to incorporate, for example, the addition of a disposal date capability and changes in the agency’s cost thresholds for property accountability and capitalization. A VHA official told us that once a property item is entered in the system’s database, a system application control retains the record, even after disposal of the item. Pharmaceutical manufacturers allow VA a credit for certain drugs that are returned. Each manufacturer establishes its own criteria for issuing credit, which can change at any time and differ among a single manufacturer’s products. The differing criteria can include such attributes as units of packaging and length of time between the return date and the expiration date. Some drugs are not returnable and must be destroyed if not used before the expiration date. To obtain available credits with minimal agency resources, VA has arranged contracts with pharmaceutical returns vendors that individual VHA medical facilities may utilize. One hundred forty of VHA’s 160 medical centers use the services of Devos, Ltd., doing business as Guaranteed Returns, to assist them in returning drugs for credit or disposing of nonreturnable drugs in accordance with environmental standards and preparing required paperwork to monitor the movement of narcotic drugs. Guaranteed Returns receives a percentage of credits issued for returnable drugs and a fee based on weight for destruction of nonreturnable drugs. VHA also uses the services of part-time physicians where necessary to alleviate recruitment difficulties or when practicality would not indicate full-time employees. While VA policy states a part-time appointment requires a tour of duty scheduled in advance that normally does not significantly change from one pay period to another, it also provides that a part-time physician whose other responsibilities make adherence to the same schedule every pay period impractical may have an adjustable work schedule. Part-time physicians with adjustable tours of duty have a biweekly work requirement consisting of non-core hours that may be adjusted at the request of the employee and core hours that are the days and times when the employee must be present unless granted an appropriate form of leave or excused absence. VA policy requires core hours to be at least 25 percent of total scheduled hours. In April 2003, the OIG reported that part-time physicians were not working the hours established in their VA appointments. A February 2004 follow-up report by the OIG stated that while most part-time physicians were on duty as required, 8 percent of the part-time physicians tested were not on duty or on approved leave or authorized absence as scheduled. To gain an understanding of VHA’s policies and procedures and the related internal controls for the three areas of operation we assessed, to identify key control activities, and to assess the design effectiveness of those controls, we obtained and reviewed VA and VHA directives, handbooks, and other policy guidance and reports issued by VA’s OIG. We also conducted interviews and system walk-throughs with VHA personnel and reviewed our previous reports. To assess the implementation effectiveness of the key control activities for the three areas of operation, we used a case study approach, reviewing transaction documentation at six VA medical centers selected based on size and medical specialization diversity of the location’s part-time physicians and other factors. For personal property management, we discussed requirements and procedures with VHA headquarters and medical center personnel. We performed tests of each medical center’s property records to assess their accuracy. Because our initial review disclosed incomplete and inaccurate information in property database records from each location we visited, we could not design our work to make a statistically based projection on the results of our work. Instead, we tested a nonstatistical selection of 100 items from each location’s property records to verify property existence by locating the item and comparing bar code, serial number, and item description information in the records to the item that we observed. For drugs returned for credit, we discussed requirements and procedures for managing turned-in drugs with personnel at VHA headquarters, the selected medical centers, and a pharmaceutical return contractor. Also, for each of the six medical centers we visited, we obtained and reviewed inventory lists of returned drugs for one contractor pickup of drugs held for return and vendor credit documents. For part-time physician time and attendance, we discussed policy requirements with VHA headquarters personnel and asked medical center staff about the processes for collecting, approving, and recording time and attendance data for part-time physicians. We reviewed time and attendance and corresponding payroll documentation for a judgmental selection of 10 part-time physicians for two biweekly pay periods ending in September 2003 at each of the six medical centers that we visited. We also reviewed medical center procedures for monitoring part-time physician attendance. We reviewed and used as guides, our Standards for Internal Control in the Federal Government and the Internal Control Management and Evaluation Tool. The Comptroller General issued these standards to provide the overall framework for establishing and maintaining internal control. According to these standards, internal control, also referred to as “management control,” comprises the plans, methods, and procedures used to meet the missions, goals, and objectives of an organization. Internal control also serves as the first line of defense in safeguarding assets and preventing and detecting errors and fraud. Our Management and Evaluation Tool provides a systematic, organized, and structured approach to assessing internal control. We performed our work at VA medical centers in Atlanta, Houston, Los Angeles, San Francisco, Tampa, and Washington, D.C.; at VA headquarters; and for drugs returned for credit, at a return contractor’s facility in East Setauket, New York. Our work was performed using a case study approach, and therefore, results of our study cannot be projected beyond the locations and transactions we reviewed. We conducted our review from February 2003 through March 2004 in accordance with U.S. generally accepted government auditing standards. We requested comments on a draft report from the Secretary of Veterans Affairs or his designee. Written comments were received from the Secretary of Veterans Affairs and are reprinted in appendix II. We found that VHA’s property control databases did not provide a complete and accurate record of personal property on hand, compromising effective management and security of agency assets at the six locations we visited. Our tests to determine whether the six medical centers had adequate control over items that were recorded in the property control databases showed that property officials could locate only about one-third of the 600 items we selected. We found that in addition to noncompliance with VA property management requirements, current VA physical inventory and property accountability policies were a major cause of unreliable property records and reduced the opportunity to adequately control personal property at five of the six medical centers we visited. Standards for Internal Control in the Federal Government requires that agencies establish physical control to secure and safeguard vulnerable assets such as equipment, periodically count those assets, and compare the counts to control records. However, through our initial reviews, we found that the property control records for the six locations we visited contained incomplete or incorrect information, such as missing property location or acquisition cost information. The property control records were such that we could not select a statistical sample of test items that would allow our results to be projected to the location’s entire property universe. We proceeded instead with a case study approach, reviewing 100 property items selected from each of the six medical centers’ databases using a nonstatistical selection method. Medical center property officials told us that some of the incorrect or incomplete information in the databases resulted from the incorrect transfer of some information from the previous property control system to the current system in 1996. The lack of accurate property control records hampered medical center property managers’ efforts to effectively safeguard and manage VHA personal property. Property officials located only 201 of 600 items (or about one-third) that we selected from the six medical centers’ property control records to observe and verify. At five locations, VHA officials found from 13 to 39 of the 100 items we tested at each location to determine if they were on hand, while at the sixth medical center, Atlanta, 62 of 100 items were found. The 600 assets we selected to observe were recorded at a total value of $104,220,868 in the property control system. However, because 125 of the 600 test items selected had no acquisition cost entered in the databases, the total cost of our selection could not be determined. Table 1 summarizes the results of our property observation tests at all locations we visited. Each category of items not observed is discussed below. Database errors or omissions represented 156 items that were not observed ostensibly because of insufficient database information. These errors included assets for which the property database did not (1) indicate a disposal date, though property officials told us the item had been disposed of, or (2) did not indicate a location for the property item. One property official said the predecessor property control system had not included a field for a disposal date, and when the program was modified to add that capability, then-existing records were not updated. Other database errors included records that did not accurately identify the asset as building service equipment, which represent items that are essentially part of facility buildings rather than personal property, or entries with one equipment identification number that represented several component items constituting one system. At the Washington medical center, we could not locate three property items, each valued at over $1 million, because of other data entry errors. The explanation for two of the three items was that the assets were on order but had not yet been received. Center officials attributed these errors to property personnel entering these assets into the property control system prior to receipt. For the third item, center officials informed us that one bar code number had been issued for a system of several pill dispensing machines, the components of which were at various locations throughout the medical center and had a combined value over $1 million, rather than bar coding each component and entering it in the property record to provide a means of controlling each item. Inadequate labeling of property items prevented us from verifying the identity of 17 items with a total acquisition cost of $29,463,952 selected from the property records at four medical centers. According to VA officials, nonexpendable property costing $5,000 or more must be bar coded and recorded in the property system. However, none of these 17 items had bar code labels attached, and either serial number labeling was also not attached to the asset or the serial number was not entered in the property records. For example, at the Tampa medical center, five property items totaling $9,996,491, including a telephone operating system, two X- ray systems, and two components of an X-ray system, were inadequately labeled. Although items that we observed matched the general description and location indicated by the property records for these items, we were unable to specifically verify their identity because bar codes had not been placed on the items and serial number information could not be compared between the property records and the physical items. When both bar code and serial number information cannot be compared between the property records and the property item, the physical inventory process is impaired and property accountability is compromised. Under these circumstances, even the most effective physical inventory procedure cannot provide the requisite assurance that assets are controlled adequately. Forty-two mobile or portable items, such as a wheelchair, adjustable bed, and Intensive Care Unit module, also could not be located. Officials stated that these items are moved from one location to another within the medical centers to meet patients’ needs. At the Houston medical center, from our selection of 100 items, property officials were unable to locate 13 portable assets, totaling $19,997, before the end of our visit. Some of these items included patient beds, patient feeding pumps, and a portable defibrillator. Regarding the remaining 184 items, property officials at the six medical centers could neither locate them at the time of our visits nor provide documentation supporting the disposal, loan, or loss of the items, or otherwise explain why they were not found. A 1997 addition to VA’s Handbook 7127, “Materiel Management Procedures,” established a $5,000 threshold for property that must be inventoried. The handbook stated that it is a local decision to maintain inventory on "other" nonexpendable equipment not capitalized or accounted for and also required accountability for sensitive property regardless of cost. Referring to the inventory provisions, A&MM staff at four of the six locations we visited told us they were only accountable for property items costing $5,000 or more and items in the four categories of sensitive assets specifically identified by VA policy: handguns, ammunition, canines, and automobiles. By ignoring VA’s general requirement to account for sensitive property regardless of cost, property managers at those locations did not keep the property control database current for most items costing less than $5,000 and lost control of the items not tracked. Medical center property officials at two centers said we should not expect to locate items with a cost lower than $5,000 because they do not inventory these assets. This practice means that some items, such as computers, monitors, and other sensitive equipment, which by their nature are subject to theft, loss, or conversion to personal use, are not inventoried or tracked. Of the 184 items that were neither found nor had plausible explanations for not being found, over half (95) were sensitive assets. Table 2 shows the nature of these 95 sensitive items categorized as personal computers, laptop computers, scanners, printers, monitors, facsimile or copier machines, and videocassette recorders. The Information Resources Management department (IRM) at the six medical centers we visited had developed alternative procedures to maintain accountability for computer equipment that cost less than $5,000. However, we found many instances in which these procedures were not used effectively. For example, a separate listing prepared by the Los Angeles center’s IRM was not used to update the property location information in the property control record, which showed the initial IRM storage room instead of the final location to which computer equipment was assigned. Further, the IRM record was not kept up to date, a factor in IRM personnel being unable to locate 22 of the 30 IRM items we selected for observation. Our standards for internal control require that key duties be divided or segregated among different people to reduce the risk of error or fraud. However, one of the methods for taking physical inventory of property established in VA’s handbook provides that each party responsible for property items will (1) receive a listing of accountable property items charged to him or her according to the property management system; (2) conduct a physical count; and (3) sign and date the listing, certifying the existence of and continuing need for the property for which he or she was responsible. Allowing the party responsible for the custody of property assets to attest to the existence of those same assets is contrary to the segregation of duties standard and compromises the control provided by taking an independent physical inventory. To illustrate the minimal value of such procedures, property officials at two medical centers told us that some service line managers just sign the inventory list without verifying the existence of the equipment. These practices would result in creating or perpetuating property control record errors if listed items had been lost, stolen, loaned, transferred, or otherwise disposed of. VA’s handbook also requires the involvement of A&MM officials in quarterly spot checks to verify inventory accuracy, but A&MM officials at only two of the locations we visited indicated they perform regular spot checks. A&MM staff at the Atlanta medical center told us they conducted periodic inventories of personal property rather than delegating that control function to parties responsible for the property. They also told us that the Atlanta facility considers computer equipment to be sensitive and, therefore, accountable. At this location, we observed 62 of 100 test items compared to from 13 to 39 of the 100 items at each of the other five locations we visited, all of which performed physical inventories primarily by using equipment lists certified by property custodians. Subsequent to our visit, property officials from the San Francisco medical center told us that they had located all equipment items with an acquisition cost of $5,000 or more that we had selected for testing, and officials from the Washington medical center told us they had located 10 additional items that we selected for observation, one of which was over $5,000. However, because we were no longer on site and could not verify the existence of these items, the additional found items are not incorporated in the statistics we present. Agency officials provided us with a copy of proposed revisions to VA’s property policy guidance that address some of the weaknesses we identified. While the draft policy adds 27 specific categories of equipment that would require accountability regardless of cost, including computer equipment, it reduces the frequency of spot checks from quarterly to semiannually and addresses the physical inventory segregation of duties issue only minimally by requiring that 5 percent of inventory be verified by disinterested parties. Internal control over drugs held for return credit, which according to VHA officials is left to the discretion of medical center management, provided no assurance that the six pharmacies we visited were receiving the proper amount of credits for returned drugs. All six of the pharmacies used contractors to return the drugs, and agency officials said that using contractors had increased the amount of credits VA received for returned drugs. However, all six locations lacked information about which drugs qualified for credit, and only one pharmacy inventoried non-narcotic drugs before they were turned over to the contractor. Accordingly, none of the pharmacies had the basic information needed to verify that credits received were correct and complete. We also found that no analytical review of credits for returned drugs, focused on maximizing the amount of credits received, was performed at the location, network, or agency level. In addition, we identified security weaknesses. Non-narcotic drugs held for return without a control listing were stored in unsecured open bins readily accessible to anyone within the pharmacy at each facility except the San Francisco and Tampa medical centers. Standards for Internal Control in the Federal Government states that internal control should provide reasonable assurance that effective and efficient use of the entity’s resources is achieved. VHA officials told us that controls over returned drugs and related credits were left up to pharmacy managers at individual medical centers. However, we found that each of the six medical centers we visited essentially used an honor system for returning drugs to manufacturers for credit, relying on contractors that collected and processed recalled, expired, or deteriorated drugs. The contractors packaged the drugs at the medical centers and shipped them either to the contractors’ processing facilities or, if required by the manufacturers, to the manufacturers’ processing facilities. For drugs shipped to the contractors’ facilities, the contractors (1) determined which drugs were returnable; (2) returned drugs qualified for credit to the manufacturers and destroyed the nonreturnable drugs; and (3) provided the pharmacy an itemized list of drugs collected, and their disposition, and an itemized estimate of credits to be received. The drug manufacturers determined the final amount of credits issued. While reviewing documentation for drugs that were returned by the six medical centers in September 2002, we found none of the pharmacies had determined if they received appropriate credit for the drugs they turned over to the contractor. Further, none of the pharmacies could determine if the credits received were complete or correct because all lacked detailed information about which drugs the manufacturers accepted for credit. In addition, none of the medical centers except Tampa maintained lists of non-narcotic drugs turned over to the return drug contractor. Medical center pharmacy staff told us there are over 1,000 drug manufacturers, each with its own policies for returning drugs for credit. For example, one drug manufacturer might require that a drug be returned 30 days prior to its expiration to qualify for credit, another drug manufacturer might allow a credit for a drug 30 days past its expiration, and another might not allow credits at all. Furthermore, a VA pharmaceutical return contractor informed us that the manufacturers frequently change their policies. Consequently, medical center pharmacy managers lacked information that would enable them to determine whether the credits they received for returned drugs were correct. As a result, the pharmacies relied on the contractors’ determination of the type and quantity of drugs that were returnable and relied solely on the drug manufacturers’ determination of the final amount of credits issued for returned drugs. In addition to establishing a return policy for drugs, each drug manufacturer set its own requirements for the process of returning the drugs and issuing credits. Some drug manufacturers allowed the pharmacy’s contractor to process returned drugs and issued a credit through the pharmacy’s prime vendor. Other manufacturers would only accept returned drugs directly from VHA. Our review of the estimated credits for non-narcotic drugs returned in September 2002 showed most were processed through each pharmacy’s prime vendor. Table 3 shows the contractors’ estimated value of credits to be received by the six medical center pharmacies we visited for non-narcotic drugs returned during September 2002. The return contractor informed us that if pharmacies requested, it could provide a report on the actual credits issued through the prime vendor for specific returned drugs. None of the pharmacies we visited indicated they were aware of this capability. Using these reports might facilitate the pharmacies’ reconciliation of credits received with drugs returned. As shown in table 3, analyzing the credits processed through the prime vendor could account for 80 percent or more of estimated credits. Standards for Internal Control in the Federal Government calls for establishing performance measures that facilitate analysis so appropriate actions are taken. None of the six medical centers had established performance measures or any kind of mechanism to oversee credits received for returned drugs. For example, medical center pharmacy managers we interviewed did not review the lists of drugs processed for credit provided by the pharmaceutical return contractor to determine if unusual trends occurred that might indicate an opportunity to increase credits received. Periodic analysis of drugs turned in throughout the year could reveal whether specific drugs were not accepted for credit on a recurring basis. For instance, drugs being consistently turned in too late to receive credit would indicate a need to process the drugs differently. If pharmacy managers reviewed actual returned drugs and credit data and took necessary corrective action to optimize returns, the net cost of pharmaceutical operations might be reduced. For example, at the Los Angeles pharmacy we found that 23 percent, or more than $60,000, of the $274,000 estimated value of drugs returned in September 2002 did not receive a credit because the drug expiration dates exceeded the manufacturers’ requirements. Medical center pharmacy officials stated that it was not cost effective to perform any of these control activities for returned drugs. However, the pharmacies had done no studies or analyses to document this conclusion. In fact, at one pharmacy, we noted that the chief pharmacist was not aware of the value of his pharmacy’s yearly credits from returned drugs. Having initially told us that his pharmacy’s yearly credits from returned drugs were only about $10,000, he subsequently reviewed the return documentation and told us his pharmacy received over $124,000 in returned drug credits for fiscal year 2003. We inquired of VA’s Pharmacy Benefit Management staff whether any agencywide analysis or study had been done to determine the reasons why more returned drugs had not qualified for credit. They stated that they had not undertaken such an analysis but believed credits had greatly increased through the use of a contractor to return drugs to manufacturers. They also told us that under their previous system, the material management staff of each medical center returned the drugs to the manufacturers and credits received for returned drugs had been minimal. Guaranteed Returns, contractor for five of the six medical centers we visited, reported that of the $21.5 million estimated return value of drugs it processed for VHA in fiscal year 2003, VHA received $5.7 million in credits for returned drugs. Without review and analysis of return drug documentation, the pharmacies cannot determine what control procedures would be cost effective. Further, despite the improved results obtained from using pharmaceutical return contractors, without agency oversight of returned drugs and related credits based on established performance measures as called for in our standards for internal control, VHA cannot be reasonably assured that stewardship of agency resources is effective. Standards for Internal Control in the Federal Government states that access to resources should be limited to authorized individuals. We found that at four of the six pharmacies visited, physical control over non- narcotic drugs held for return was lacking. The San Francisco pharmacy stored such drugs in a locked bin and the Tampa facility limited access to a secured area, but the other four locations used open, unsecured bins. Anyone with access to the pharmacy also had access to the drugs. Thefts would be very difficult, if not impossible, to detect because the pharmacies did not maintain lists of the non-narcotic drugs held for return. The drugs were simply deposited in the bin. The lack of physical control over non- narcotic drugs held for return represents a potential lost opportunity to maximize return credits and to reduce the risk of theft or misuse of these drugs. During our review, lapses in security at two of the pharmacies we visited were reported. The VA OIG reported that three employees of the Houston medical center were convicted of conspiring to steal large amounts of non-narcotic pharmaceutical drugs from the pharmacy. These employees had stolen over $1.3 million of drugs over 3 years. At the Washington, D.C., medical center, as discussed in our recent report on VHA personnel screening procedures, we found that one employee of the pharmacy had been convicted for possession of illegal drugs prior to VHA employment. While these incidents may not relate directly to drugs held for return to manufacturers, they indicate the risks involved and underscore the clear need for effective control over these drugs. The combination of weaknesses in record keeping and physical controls over non-narcotic drugs held for return exposed them to potential loss, theft, or unauthorized use. Our review of part-time physician time and attendance documentation for the two pay periods ending in September 2003 showed that scheduled and actual hours worked were not always documented according to policy at the six medical centers we visited. Also, specific hours worked recorded by physicians on their time and attendance reports sometimes differed from information entered in the payroll system. We also found that latitude provided in VHA’s Directive 2003-001, issued in January 2003, on time and attendance of part-time physicians was a factor in the various ways the six locations carried out part-time physician attendance monitoring responsibilities. While newly emphasized policies stressed the importance of this matter, compliance in some cases had been slow to develop and oversight processes varied and were not fully effective. Our standards for internal control state that control activities, such as approvals and authorizations, are integral to an entity’s accountability for stewardship of resources. Consistent with that management control objective, VHA’s January 2003 directive called for specifying work schedules in writing in advance of the biweekly pay period, showing the specific days and hours that part-time physicians were to work, including core hours when employees working adjustable shifts must be present. Our review showed that schedules were not always established in advance of the pay period as required by VHA. For the two pay periods ending in September 2003, our review of records for 10 part-time physicians at each of the six locations we visited revealed that only the Houston medical center had documented preapproved schedules for all physicians whose records we tested. A contributing factor for this weakness was that an official at one location told us part-time physicians with fixed schedules did not require a documented preapproved schedule. Almost one-third of the part-time physician records that we reviewed did not include the required documented schedule. However, all those who had documented schedules also had core hours established as required. Table 4 summarizes the results of our work regarding part-time physician policies and their schedules at the centers we visited. Failure to document schedules can lead to confusion about when a physician should be at work. The VA OIG’s February 2004 report on VA medical center part-time physician time and attendance stated that 15 of 58 part-time physicians who were not present when scheduled during a 1-day test said they had changed their hours without getting written approval. Our internal control standard regarding accurate recording of transactions and events applies to the entire process or life cycle of a transaction or event from initiation and authorization through its final classification in summary records. Our comparison of manually prepared time and attendance records with computerized payroll system timecards indicated no differences between total hours worked and total hours entered in the payroll system for the cases we tested. However, we found that when part- time physicians temporarily modified their approved work schedules, the changes they noted on their forms 4-5631a, used to document, review, and approve actual hours worked, were sometimes not entered in the computerized payroll system. At five of the six medical centers, we compared information shown on the payroll system timecards to the forms 4-5631a that were signed by the part-time physicians, timekeepers, and the physicians’ supervisors. At four of those five medical centers, we noted at least one instance of a difference between the specific days and hours worked shown on a part-time physician’s form 4-5631a and that information shown on the corresponding payroll system timecard. Timekeepers and other medical center officials told us that recording temporary changes for actual time worked in VHA’s computerized payroll system is difficult because the system is inflexible. As a result, if total hours that a part-time physician actually worked during a pay period equaled the total hours scheduled, timekeepers often entered the physician’s scheduled hours into the computerized payroll system rather than the actual hours worked. However, accurate payroll system information about specific hours worked is important to satisfy VHA’s need to document whether part-time physicians fulfill their core hour requirements. Our standards for internal control state that an entity’s documentation of transactions and other significant events must be complete and accurate. At the six medical centers we visited, we found variation in the design and effectiveness of medical center procedures concerning the way supervisors and timekeepers checked and documented daily employee attendance and how facility management periodically monitored employee compliance with time and attendance requirements. VHA’s January 2003 directive on part-time physician time and attendance referred to VA’s underlying policy manual that established requirements for supervisors or timekeepers to have personal knowledge that part-time physicians worked the hours or days shown on their time and attendance forms. Timekeeping procedures that included keeping a record of each physician’s daily attendance throughout a pay period provided greater reliability than those that relied on the physicians’, their timekeepers’, or their supervisors’ memories. With working arrangements of part-time physicians, their supervisors, and timekeepers that vary among the service centers within a medical center, we found that the timekeepers at the medical centers we visited accounted for daily attendance of physicians using a wide variety of procedures. While the Houston medical center established a sign-in procedure for all part-time physicians, the other five medical centers relied primarily on the timekeepers’ observation of physicians’ daily attendance. At those facilities, the procedures often differed among service centers and included activities such as timekeepers making informal notes on their personal calendars or preparing calendar- like worksheets to check off the names of each part-time physician when he or she was observed at the center during a scheduled workday. While each process offered a level of control over time and attendance, they all had limitations and none provided assurance that part-time physicians were on duty during their core hours. For example, on the surface, a sign-in procedure would seem to offer more definitive assurance; however, effectiveness depends on how well the procedure is implemented. On the day we reviewed part-time physician sign-in sheets at the Houston medical center, we noted that only 2 of 15 physicians scheduled to work had signed in. Timekeepers told us they observed 5 other physicians in the facility and 1 had advance approval to attend a lecture. However, the timekeepers also told us 2 other part-time physicians scheduled to work had called in and stated that they were “accounted for,” and the remaining 5 had not reported in or otherwise confirmed their attendance. Houston’s failure to enforce its sign-in procedures for its part-time physicians is an example of compromised control effectiveness that impaired medical center management’s ability to know if part-time physicians worked when scheduled. While VHA’s January 2003 directive identified medical center management’s responsibility for monitoring compliance with part-time physician time and attendance policy, the methodology for implementing that responsibility was left to the discretion of facility management. Some of the methodologies adopted were less effective than others. For example, while the Atlanta medical center service areas checked attendance for 5 percent of part-time physicians one day each month, at the San Francisco center service areas checked attendance of all part-time physicians at least one day per quarter, and its Office of Human Resource Management made random spot checks. In addition to physical observation, other methods used for making these periodic surveys of attendance at the six locations included monitoring doctors logging into the facility’s computer network, monitoring doctors’ notes entered into VHA’s patient records system, and paging doctors to determine if physicians used medical center telephones to respond. The wide variety of part-time physician time and attendance procedures that have been developed by the medical centers we visited reduces VHA management’s level of assurance that controls are effective and agency objectives are being achieved. We believe an opportunity exists for the agency to study the various medical center and service area procedures so that VHA can provide more specific direction about the most effective ways to improve control over part-time physician time and attendance agencywide. The weaknesses in internal control that we identified at the six VA medical centers we visited leave the agency vulnerable to waste, fraud, and abuse. Improving the design and implementation of policies regarding personal property will help improve accountability for agency assets, especially sensitive property. VHA managers performed no analytical oversight of credits for returned drugs, and the six medical centers had no effective control over the amount of credits for drugs returned to manufacturers. Some analysis of drug return transactions would provide management with a basis to determine what control activities would provide an appropriate cost/benefit ratio. Current policies and procedures for monitoring part- time physician time and attendance, if implemented more effectively, may provide reasonable assurance that management’s objectives will be met. In addition, the wide range of physician attendance monitoring procedures developed by the various medical centers and service areas provides an opportunity to improve controls agencywide if their relative effectiveness is studied. While some medical centers have already taken positive steps to improve controls over these areas, appropriate direction from management will spur action agencywide and help reduce vulnerability to waste, fraud, and abuse. We are making the following 17 recommendations to improve the internal controls over the operating areas that were the subject of our work. Some of these recommendations require attention of VA management at the department level, others VHA, and still others VA medical center management. We recommend that the Secretary of Veterans Affairs direct the Assistant Secretary for Management to clarify existing guidance and establish consistent parameters for personal property that is required to be accounted for in the property control records and that is subject to physical inventory to include sensitive property, provide a more comprehensive list of the type of personal property assets that are considered sensitive for accountability purposes, direct that physical inventories of personal property be performed by the A&MM staff or other parties who are independent of those with property custodian responsibilities, and reinforce VA’s requirement to attach bar code labels to agency personal property. To improve accuracy of VA’s time and attendance records for part-time physicians, we recommend that the Secretary of Veterans Affairs direct the Assistant Secretary for Management to coordinate all time and attendance system changes with VHA, in order to ensure that the time and attendance system facilitates entry of actual hours and days worked by part-time physicians into VA’s permanent electronic time and attendance record. To improve oversight of medical center operations, we recommend that the Acting Under Secretary for Health designate a headquarters-level staff office to monitor medical facilities’ credits for returned drugs; review returned drug credits and related pertinent information for VA medical facilities and determine, especially for those with unusual performance patterns, whether there might be additional opportunities for credits; develop procedures to periodically test whether the amount of credits received for returned drugs is correct; implement procedures to periodically test whether the amount of credits that medical centers received for returned drugs is correct; conduct a best practices review of procedures implemented by VA medical centers and service areas to identify those most effective in documenting daily attendance of part-time physicians and periodically monitoring employee compliance with time and attendance requirements; and use the results of the best practices review to provide more definitive policy guidance to improve control effectiveness over part-time physician attendance monitoring. To address the weaknesses noted during our visits to six VA medical centers, we recommend that the Acting Under Secretary for Health require the directors of those medical centers to determine the location or disposition of personal property items not found during our site visits; review property records to identify and correct erroneous or incomplete prepare a running list of all non-narcotic drugs held for return in facility pharmacies as they are removed from current supplies to compare with contractor-prepared lists of returned drugs; improve physical security over non-narcotic drugs held for return in facility pharmacies as they are removed from current supplies; and analyze information regarding drugs returned to manufacturers to identify potential improvements that might increase the amount of credits received, such as improving the timeliness of returning drugs consistently turned in too late to qualify for credit. We also recommend that the Acting Under Secretary for Health determine whether the above recommendations pertaining to the facilities we visited are applicable to all VA medical facilities. VA provided written comments on a draft of this report. In its response, VA agreed with our conclusions and recommendations and reported that it is developing an action plan to implement them. Additionally, VA’s response stated that it is pursuing a number of strategies to improve the processing of expired medications held for credit, the monitoring of part-time physician time and attendance, and the inventory records of all equipment. VA also provided technical clarifications, which we incorporated where appropriate. VA’s written comments are reprinted in appendix II. We are sending copies of this report to the Ranking Minority Member, Subcommittee on Oversight and Investigations, House Committee on Veterans’ Affairs; the Chairman and Ranking Minority Member, House Committee on Veterans’ Affairs; the Chairman and Ranking Minority Member, Senate Committee on Veterans’ Affairs; the Secretary of Veterans Affairs; the Acting Under Secretary for Health, Veterans Health Administration; and other interested parties. We will also make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. Should you or your staff have any questions on matters discussed in this report, please contact me at (202) 512-6906 or by e-mail at williamsm1@gao.gov or Jack Warner, Assistant Director, at (202) 512-4679 or by e-mail at warnerj@gao.gov. Major contributors to this report are acknowledged in appendix III. In addition to those named above, the following individuals made important contributions to this report: Kwabena Ansong, Sharon Byrd, Cary Chappell, Lisa Crye, Fred Evans, Lou Fernheimer, Jeff Isaacs, Julia Matta, Bonnie McEwan, Christina Quattrociocchi, Donell Ries, Alana Stanfield, and Jason Strange. The Government Accountability Office, the audit, evaluation and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is through GAO’s Web site (www.gao.gov). Each weekday, GAO posts newly released reports, testimony, and correspondence on its Web site. To have GAO e-mail you a list of newly posted products every afternoon, go to www.gao.gov and select “Subscribe to Updates.”

The Department of Veterans Affairs (VA) provides health care to veterans through the $27 billion Veterans Health Administration (VHA) medical programs. VHA administers and operates VA's medical system, providing care to nearly 5 million patients in 2003. As of September 2003, VHA operated 160 hospitals, 847 outpatient clinics, 134 nursing homes, 42 domiciliaries, and 73 comprehensive home care programs, including facilities in every state, Puerto Rico, the Philippines, and Guam. VHA is responsible for effective stewardship of the resources provided to it by Congress, which asked GAO to review internal controls in three areas of operation at selected VHA medical centers. GAO conducted a review to assess the effectiveness of control activities over (1) personal property, (2) drugs returned for credit, and (3) part-time physician time and attendance. GAO's review found that six selected VA medical centers lacked a reliable property control database. The property databases for the six medical centers contained incomplete information. As a result, GAO could not select a statistical sample of test items so that results could be projected to each location's entire property universe. Key policies and procedures established by VA to control personal property provided facilities with substantial latitude in conducting physical inventories and maintaining their property management systems, which resulted in reduced property accountability. For example, VA's Materiel Management Procedures handbook allowed the person responsible for custody of VA property to attest to the existence of that property rather than requiring independent verification. Also, personnel at some locations interpreted a policy that established a $5,000 threshold for property that must be inventoried as a license to ignore VA requirements to account for lower cost items that are susceptible to theft or loss, such as personal computers and peripheral equipment. These weak practices, combined with lax implementation, resulted in low levels of accountability and heightened risk of loss. VHA personnel located fewer than half of the 100 items GAO selected at each of five medical centers and 62 of 100 items at the sixth medical center. The process for obtaining credit for recalled, expired, or deteriorated drugs was, in essence, an honor system. Each of the six pharmacies GAO visited used a contractor to return drugs to the manufacturer for credit, but only one of the pharmacies inventoried non-narcotic drugs before they were turned over to the contractor. None of the pharmacies had enough information about which drugs qualified for credit to be able to reconcile the credits they received with the drugs they had turned over to the contractor. There was no agency-level oversight of returned drug information to help identify improvements that might increase the credits that VA receives. At four of the six facilities, non-narcotic drugs held for return were stored in unsecured open bins accessible to anyone in the pharmacy. The combined lack of record keeping and physical controls over non-narcotic drugs held for return exposed them to potential loss, theft, or unauthorized use. Scheduled and actual hours worked by part-time physicians at the six locations GAO visited were not always documented in accordance with a January 2003 VHA directive. Five of the six locations had not prepared written work schedules for all part-time physicians as required. GAO found that latitude provided in the directive resulted in wide variation in procedures used by the six medical centers to verify physician compliance with work schedules. While some timekeepers used informal notes to record daily attendance, one facility required physicians to sign in. However, on the day of GAO's review, only two of 15 scheduled physicians had signed in. Attendance monitoring procedures at the six locations varied in frequency and included monitoring all part-time physicians once per quarter at one location and 5 percent of part-time physicians each month at another.

On January 6, 1993, the Institute of Medicine published a report that discussed secret U.S. chemical weapons programs during World War II. The report found that an estimated 60,000 military personnel participated as human experimental subjects in tests of exposure to mustard agents and lewisite and unknown numbers of additional servicemembers may have been exposed to these substances through their participation in the production, transportation, and/or storage of these chemical substances. On February 18, 1993, we issued a report that found VA lacked information about individuals who were exposed during secret DOD chemical tests. After Members of Congress, the President of the United States, and the Secretary of Defense exchanged a series of letters about this issue in 1993, the Deputy Secretary of Defense issued an agencywide memo that released all individuals from any nondisclosure restrictions that might have been placed on them, tasked the secretaries of the military departments to undertake efforts to declassify and provide to VA as soon as possible information about individuals who were potentially exposed, and directed OUSD (P&R) to establish a task force to monitor the status of DOD’s efforts. As a result, OUSD (P&R), the military services, and VA developed the Chemical Weapons Exposure Study Task Force to identify DOD personnel exposed to chemical substances during testing, training, transport, production, and storage. By conducting site visits and other research efforts, the task force identified approximately 6,400 servicemembers and civilians who were potentially exposed to mustard, lewisite, and other chemical substances. The office created a database with information about these individuals (hereafter referred to as OUSD (P&R) database) and, according to OUSD (P&R), sent certificates of commendation to more than 700 individuals for whom it could find contact information. In addition to its own research, OUSD (P&R), on behalf of the task force, issued a task order for a contractor to analyze, extract, and develop a database of information on all volunteers and/or other subjects potentially exposed to live chemical or biological substances. The contractor developed a database and issued a series of reports that identified the locations of human exposures to chemical substances, including those resulting from tests and a variety of other activities such as transportation, production, storage, and disposal. Congress continued to look into this issue during 1994 through a series of hearings and a staff report that was prepared for the U.S. Senate’s Committee on Veteran Affairs. The issue of servicemembers being used as human subjects during DOD’s chemical and biological tests received high-level attention again in 2000, when the acting Secretary of Veterans Affairs wrote a letter to the Secretary of Defense requesting assistance in obtaining information about a series of then-classified chemical and biological tests under DOD’s Project 112 program. OASD (HA) officials consequently initiated some actions to identify potentially exposed individuals. Subsequently, DOD, VA, and Congress exchanged a series of correspondence about the need to identify individuals who were potentially exposed during these tests. Eventually, the Defense Authorization Act for FY 2003 required DOD to submit to Congress and the Secretary of Veterans Affairs a comprehensive plan for the review, declassification, and submittal to VA of all DOD records and information on Project 112 that are relevant to the provision of benefits by the Secretary of Veterans Affairs to members of the armed forces who participated in that project. During this effort, DOD identified 5,842 servicemembers and estimated that 350 civilians had been potentially exposed during Project 112 tests, and this information was entered into a Project 112 database. The act further required the Comptroller General to evaluate the plan and its implementation. The Defense Authorization Act for FY 2003 also required DOD to work with veterans and veterans service organizations to identify DOD projects or tests outside of Project 112 that may have exposed members of the armed forces to chemical or biological substances. In June 2004, we reported that DOD had not yet begun its investigation to identify such projects or tests and recommended that the Secretary of Defense direct the appropriate office(s) to finalize and implement a plan for identifying DOD projects and tests conducted outside of Project 112 that might have exposed servicemembers to chemical or biological substances and ensure that the plan addresses the scope, reporting requirements, milestones, and responsibilities for those involved in completing this effort. According to an OASD (HA) official, OASD (HA) made an informal agreement with OUSD (AT&L) to undertake this effort since OASD (HA) did not have the resources to conduct an investigation itself or to fund a contractor to do the research. In September 2004, OUSD (AT&L)’s chemical and biological defense office issued a task order to fulfill this provision of the legislation. The research being done as a result of this task order is ongoing as of December 2007. In June 2003, after having identified several thousand servicemembers and hundreds of civilians as having been potentially exposed to chemical or biological substances during Project 112, DOD stopped actively searching for additional individuals. According to a knowledgeable DOD official, this decision was made without a sound and documented cost-benefit analysis. The Defense Authorization Act for FY 2003 required DOD to review records and information necessary to identify members of the armed forces who were or may have been exposed to chemical or biological substances as a result of Project 112. Subsequently, in June 2003, DOD issued a report to Congress that stated that 5,842 servicemembers and an estimated 350 civilians might have been exposed during Project 112 tests. The report also indicated that DOD had ceased its active search for individuals potentially exposed during Project 112 tests and that it would investigate any new information that may be presented as well as share any additional or changed information with VA and the public. In 2004, we reported that DOD performed a reasonable investigation of servicemembers who were potentially exposed to the substances used during Project 112 tests. However, we found that DOD had not exhausted all possibilities for identifying additional servicemembers and civilian personnel who had been potentially exposed. Therefore, we recommended that DOD determine the feasibility of addressing these unresolved issues. In response to our recommendation, DOD determined continuing an active search for individuals had reached the point of diminishing returns, and reaffirmed its decision to cease active searches. This decision was not supported by any objective analysis of the potential costs and benefits of continuing the effort. Instead, this decision was made by officials in OASD (HA) who had a working knowledge of Project 112 tests and the contents of chemical and biological test record repositories. These officials concluded that the record repositories that had been searched contained the majority of Project 112 documents; therefore, they believed that the bulk of exposures related to Project 112 tests had already been identified. Furthermore, the officials decided that the application of resources necessary to continue searching for Project 112 exposures would result in a diminishing return on their investment. The Office of Management and Budget has stated that a good cost-benefit analysis should include a statement of the assumptions, the rationale behind them, and a review of their strengths and weaknesses. This could include a full accounting of information known, related costs, benefits, and challenges of continuing to search for additional Project 112 participants. Moreover, our prior work has shown that there are elements integral to a sound cost-benefit analysis. For example, the analysis should include a thorough evaluation of the social benefits and costs of investments, identify objectives to ensure a clear understanding of the desired outcome, and include a list of the relevant impacts to ensure that all aspects are considered. DOD could not provide us with a quantitative analysis based on objective data or any documented criteria because OASD (HA) was not required to provide any support or basis for the decision. Since DOD’s June 2003 report to Congress and its decision to cease actively searching for additional exposures, additional individuals who may have been exposed as a result of Project 112 tests have been identified through various non-DOD sources, as shown in table 1. For example, the Institute of Medicine conducted a study on the long-term health effects of participation in the shipboard hazard and defense tests that were conducted as a subset of Project 112. This study identified 394 individuals who had been potentially exposed and who were previously unknown to DOD. According to DOD and Institute of Medicine officials, the additional names were discovered when the Institute of Medicine applied a more inclusive methodology in its research. In addition, our previous work in 2004 reported that DOD did not exhaust all possible sources of information during its investigation of Project 112 and our own research for that report resulted in the identification of 39 additional potentially exposed servicemembers. For example, DOD had limited success in identifying exposures during land-based tests because it was unable to find documentation, and it did not specifically search for individual civilian personnel in its investigation because it considered them to be outside of its scope. Furthermore, DOD officials have told us that veterans who participated in Project 112 tests have contacted DOD on their own initiative in search of information and documentation related to their exposures, which has resulted in 165 additional veterans being identified as having been potentially exposed during these tests. DOD’s current effort to identify individuals who may have been exposed to chemical or biological substances during activities outside of Project 112, discussed in the following section of this report, has also resulted in the discovery of information related to Project 112 tests. Specifically, the DOD contractor has found evidence that individuals who DOD already knew were potentially exposed to substances during at least one known Project 112 test were also potentially exposed during other Project 112 tests. In light of the increasing number of individuals who have been identified since DOD ceased actively searching, until DOD makes a sound and documented decision regarding the cost and benefits of actively searching for individuals potentially exposed during Project 112 tests, Congress and veterans may continue to question the completeness and accuracy of DOD’s effort. Although DOD has taken action to identify individuals who were potentially exposed during chemical or biological tests outside of Project 112, we identified several shortcomings in the current effort. Specifically, we found that DOD’s approach was hampered by (1) a lack of clear and consistent objectives, scope of work, and information needs; (2) management and oversight weaknesses; (3) a limited use of the work of other entities that previously identified exposed individuals; and (4) a lack of transparency in DOD’s efforts. In response to the Defense Authorization Act for FY 2003 and our May 2004 recommendation that DOD finalize and implement a plan to identify individuals who were potentially exposed during tests conducted outside of Project 112, DOD issued a task order in September 2004. The task order identified four sets of tasks that the contractor was to undertake to accomplish the task order’s objectives within 3 years—perform literature searches, conduct and review on-site data collections, data mine existing databases, and augment a database maintained by the contractor. The contractor has issued monthly reports on its work to OUSD (AT&L)’s chemical and biological defense office, which indicate that the contractor has taken action on each of these tasks. OUSD (AT&L)’s chemical and biological defense office and the contractor have agreed that the on-site reviews will be conducted at a total of 18 sites that were identified and prioritized based on established criteria, such as relevance and number of documents expected to be present. As of October 2007, the contractor has completed on-site data collection at 5 of these 18 sites, and as of December 2007 was collecting data at 3 additional sites. During its site visits, the contractor’s staff searches a variety of documents for information that pertains to human exposure to chemical or biological substances. The documents that are identified as having relevant information are then scanned into an electronic file and the information from those documents—such as the individual’s name, the substance to which the subject was exposed, and the activity that resulted in the exposure—is entered into a database. The contractor conducts a quality assurance review before this information is delivered to OASD (HA) officials. OASD (HA) officials told us that they perform a detailed review of this information, query the contractor to resolve errors or inconsistencies, and make modifications to the information provided by the contractor if they have received or read other information that they believe could add contextual sophistication. Once OASD (HA) officials complete their review of the information, it is added to the DOD chemical and biological test database that they maintain (hereafter referred to as the OUSD (AT&L) task order database). While the database information is not provided to OUSD (AT&L)’s chemical and biological defense office, the contractor’s monthly report to this office includes the number of identified individuals that the contractor has provided to OASD (HA). The task order identified specific locations for the contractor to review and was supposed to be completed in September 2007; however, the contractor was unable to complete its work within the 3-year schedule and has subsequently received a 3-year extension. This task order is valued at almost $4.5 million, and the estimated value of the extension is between $2.5 million and $3.7 million. Based on the project’s June 2007 concept of operations plan, which DOD developed as a result of this review, the contractor is expected to meet the project’s objectives and complete collection and analysis of information obtained from 18 data collection sites by September 2010. Since the remaining sites have been prioritized based on expected level of information and other criteria, DOD officials believe that the remaining data collection efforts could be completed more quickly. DOD’s current effort to identify individuals potentially exposed to chemical or biological substances lacks clear and consistent objectives, scope of work, and information needs, which affects DOD’s ability to know whether it has accomplished the project’s goals. First, the objectives of DOD’s current effort are inconsistent. The Defense Authorization Act for FY 2003, which was the genesis for DOD’s current effort, directed the Secretary of Defense to identify DOD projects or tests outside of Project 112 that may have exposed members of the armed forces to chemical or biological substances. However, the focus of the current effort has expanded to include other exposures, including those resulting from immunizations, transportation, storage, and occupational accidents. This occurred because the documents that are guiding this effort, including the project’s September 2004 statement of work and its June 2007 concept of operations plan, have been used interchangeably to define the scope of the work. We identified a difference of opinion between DOD and VA regarding the overall focus of the contractor’s research efforts. Officials in OUSD (AT&L)’s chemical and biological defense office stated that they believe the contractor should focus only on identifying participants in DOD tests since the Defense Authorization Act for FY 2003 was the genesis of this task order, and they believe that the primary interest is in individuals who were not aware of their exposures or are unable to report their exposures due to the classified nature of the tests. They also believe that individuals accidentally exposed at a work location might be protected under occupational health regulations and statutes. However, VA officials stated that they would prefer that DOD provide information on all exposures, including those not associated with DOD tests, since VA is responsible for adjudicating all claims by servicemembers, regardless of how they were exposed. The contractor conducting the search has included all types of exposures in its research, which according to DOD and contractor officials is based on VA’s stated preferences. Second, the scope of DOD’s current effort is unclear. Specifically, while the Defense Authorization Act for FY 2003 directed DOD to identify only members of the armed forces, the task order’s 2004 statement of work and the June 2007 concept of operations plan state that the objective of the project is to collect information on all servicemembers and civilian personnel who might have been exposed from 1946 to present. However, DOD’s current effort has not included an active search of civilian personnel. Instead, at the direction of DOD, the contractor is collecting information on civilians who may have been exposed to chemical or biological substances when it comes across those names while searching for servicemembers. DOD officials stated that they focused their efforts on servicemembers because VA has actively requested information about servicemembers from DOD for years and the department has not received any inquiries about the civilians. At the time of our review, the contractor had collected information on approximately 700 civilian personnel who were potentially exposed to chemical or biological substances. Third, the amount and type of information that the contractor needs to collect for this effort has been expanded from the original task order requirement. The task order specifies that the information to be collected should identify potential human exposure events, the names of test programs, chemical and biological substances involved, and the names of volunteers or participants. However, DOD has expanded the information that the contractor should collect, which may be lengthening the time for the contractor to complete its work. For example, in February 2007, officials from one of the repository sites provided the contractor a CD with names and exposure information for 2,300 individuals who were exposed to a series of biological tests at Fort Detrick, Maryland, known as Operation Whitecoat. However, as of October 2007, the contractor had not provided DOD with these names because it was adding information, such as the test objective and summary, and exposure and treatment information. Since most of these 2,300 individuals had been previously aware of their exposures due to Fort Detrick’s independent outreach efforts, a DOD official who has worked with these individuals has stated that it is unclear how much additional information the contractor needs to collect about this group. While OASD (HA) officials have said that the additional information has been helpful for their needs, they and VA officials have also acknowledged that the identity of the chemical or biological substance to which an individual was potentially exposed is the most pertinent information. Without consistent guidance about the objectives, scope of work, and information necessary to meet DOD’s goals and objectives, DOD’s current effort might not produce the desired results. After discussing this issue with DOD officials, in December 2007 officials in OUSD (AT&L)’s chemical and biological defense office stated that they plan to revise the task order’s statement of work, concept of operations plan, and a DOD implementation plan to clarify the scope of work and the focus of the research to servicemembers—the original focus as identified in the Defense Authorization Act for FY 2003. Until recently, DOD’s current effort has lacked adequate oversight of the contractor activities and results. We have previously reported that providing effective oversight is essential and, at times, DOD’s oversight was wanting, as it did not always task personnel with oversight duties or establish clear lines of accountability. While OUSD (AT&L)’s chemical and biological defense office established three different points of contact throughout the life of the task order who participated in meetings when the work started in 2004 and assisted the contractor undertaking the effort in accessing repository sites when requested, these points of contact were not performing active oversight activities nor were they designated as the project manager for this effort. During our review, officials in OUSD (AT&L)’s chemical and biological defense office realized that their predecessors had not selected a project manager and selected one of the office’s civilian employees to oversee the effort. We also found that DOD had not visited any of the repository sites where the contractor had proposed or completed its research to ensure that the work was effectively and efficiently meeting the task order’s objectives. We visited the three repository sites where the contractor was conducting its work during our review. At one location, a knowledgeable DOD official expressed concerns to us that the contractor’s presence and research in one of the site’s libraries might not be needed. However, since officials in OUSD (AT&L)’s chemical and biological defense office had not visited the site or met with site officials, they were unaware of these concerns and therefore were unable to decide whether the contractor should be conducting work at that particular site or whether the research funds and time should be spent at a site that they believe might provide more relevant information. In addition, until June 2007, OUSD (AT&L)’s chemical and biological defense office had not regularly evaluated the effectiveness or efficiency of the contractor’s work. For example, at the time of our review, officials in OUSD (AT&L)’s chemical and biological office told us that they did not know the extent to which each of the task order’s four tasks was meeting its objective to identify servicemembers and civilians who were potentially exposed to chemical or biological substances during testing and other activities. Therefore, DOD was not in a position to determine whether the task order needed to be modified to focus DOD’s resources and the contractor’s research efforts to those tasks that will best meet its objectives. Further, while the contractor had implemented its own quality assurance/quality control process that was approved by OUSD (AT&L)’s chemical and biological defense office, the office had not taken any action to independently assess the accuracy and characterization of the information that the contractor was providing to the OASD (HA), which maintains DOD’s databases of potentially exposed individuals. As a result, officials in OUSD (AT&L)’s chemical and biological defense office, who are responsible for overseeing the contractor’s efforts, have limited knowledge about the accuracy and characterization of the information that was being collected. Review and assessment of the contractor-provided data by the project manager are important because we identified potential problems with the accuracy of that information. For example, our work indicated that there are discrepancies between the number of individuals reported by the contractor in its monthly reports to OUSD (AT&L)’s chemical and biological office and the number of individuals that exist in OASD (HA)’s database that could not be adequately explained. In addition, at the time of our review, the characterization in the contractor’s monthly reports provided to OUSD (AT&L)’s chemical and biological defense office that all of these individuals were potentially exposed during chemical or biological tests gave the wrong impression to the project manager. For example, while the contractor has characterized the individuals it has identified as having been involved in DOD’s chemical and biological “tests”, an unknown number of these exposures resulted from immunizations, transportation, occupational, and storage accidents. This number also includes individuals who might have been associated with the tests but who were not exposed to any substances, such as those who participated in physical exercises to test the durability of chemical and biological suits or who could have been part of a test control group. OASD (HA) officials were able to identify at least 1,800 names in the database that were not exposed to any substances, which leaves about 7,100 names in the database that have been potentially exposed to chemical or biological substances, as shown in table 2. DOD and contractor officials stated that they have included these names in the database so that they could appropriately respond to these individuals’ concerns if they contact DOD or VA. Specifically, according to DOD, including these names in the database enables the department to refute any claims by individuals who participated in tests where they were not exposed to any chemical or biological substances. We identified a variety of factors affecting the ability of OUSD (AT&L)’s chemical and biological defense office to provide oversight, including a lack of consistent leadership, inadequate internal controls, a shortage of personnel, and a lack of defined roles and responsibilities. For example, the position that was identified as the office’s point of contact for the task order is a 1-year position. Consequently, the contractor has had to work with three different individuals during the first 3 years of the task order. The official holding this position during our review requested and was granted a 2-year extension in this position, and thus he has been able to implement a number of internal controls to improve the oversight and accountability of this project. In addition, until September 2007, the respective roles and responsibilities of OUSD (AT&L)’s chemical and biological defense office and OASD (HA) had not been clearly identified. In September 2007, in response to our review, OUSD (AT&L)’s chemical and biological defense office and OASD (HA) signed an implementation plan that identified their respective roles and responsibilities. In planning, executing, and evaluating DOD’s current effort, OUSD (AT&L)’s chemical and biological defense office did not fully leverage the work of other entities that had previously identified exposed individuals. Multiple DOD and non-DOD organizations have conducted a variety of independent efforts since the early 1990s, through which they have identified thousands of individuals who were potentially exposed during chemical or biological tests. These entities possess specific information about the tests—to include the location of test records—and the personnel conducting the work developed institutional knowledge. While OUSD (AT&L)’s chemical and biological defense office leveraged Project 112 information from the OASD (HA), it did not leverage information available from other DOD and non-DOD sources. For example, between 1993 and 1997, the joint DOD-VA task force identified approximately 6,400 individuals who were potentially exposed to sulfur mustard, lewisite, and other chemical substances. OUSD (P&R) led the effort by using some of its own personnel to conduct the research and visit several repository sites in addition to issuing a task order for a contractor—the same contractor DOD is currently using to research and identify tests and exposures—to develop a database containing information on the location, chemicals tested, and dates of the chemical weapons research program. During this period, OUSD (P&R) personnel involved with the research became very knowledgeable about the issues, collected boxes of information, and issued various reports. OUSD (P&R) officials transferred the names of the individuals who were identified to OASD (HA) officials in April 2005. According to OUSD (P&R) officials, however, officials in OUSD (AT&L)’s chemical and biological defense office had not met with any of the personnel with institutional knowledge or examined any of the documents that OUSD (P&R) still maintained. Since OUSD (P&R)’s reports identified locations of exposures, officials in OUSD (AT&L)’s chemical and biological defense office could have used this information as another source to help validate and prioritize the repository sites proposed by the contractor for its current effort, and to eliminate potential redundancy. Furthermore, as a result of independent research efforts by the Institute of Medicine about the health effects of DOD chemical tests using human subjects, the organization developed a database that contained the names and addresses of more than 4,000 servicemembers who were potentially exposed to chemical substances during a series of tests at Edgewood, Maryland. However, OUSD (AT&L)’s chemical and biological defense office was not aware of this database since the office had not coordinated with the organization. Institute of Medicine officials told us that they believe the names and contact information in this database could help DOD with its efforts since the names were collected from the same locations where the contractor for DOD’s current effort is doing its research. Subsequent to our September 2007 meeting with the Institute of Medicine, its officials contacted OASD (HA) to establish the protocols to transfer the names of identified individuals to DOD so that it can determine whether these individuals are already included in any of DOD’s databases. Without communicating and coordinating with DOD and non- DOD organizations that have previously conducted similar efforts, DOD’s current effort will not be able to take advantage of existing information so that it can focus its resources on the areas where information is missing. DOD’s current effort lacks transparency since it has not worked with veterans, and it has not kept Congress and veterans service organizations fully informed about the status of its efforts. Although DOD officials conducted outreach to veterans during its Project 112 research effort and the Defense Authorization Act for FY 2003 required DOD to work with veterans and veterans service organizations to identify projects and tests outside of Project 112 that may have exposed members of the armed forces to chemical and biological substances, DOD has not included veterans and veterans service organizations during its current effort. DOD also has not kept Congress, veterans, and the public informed on the status of its current effort as it did during its Project 112 investigation. Specifically, in 2002, DOD established a public internet site to provide interested persons with information on what happened during those tests that might have affected the health of those who served. The internet site included a status report on DOD’s efforts so that veterans and others could monitor the progress, and it also contained reports, documents, and links to related internet sites. The internet site, which was operated by OASD (HA), has not been updated with information about DOD’s current effort to identify individuals outside of Project 112. Representatives from a veterans service organization that has pursued information regarding DOD’s use of servicemembers as human subjects told us they were not aware of DOD’s current effort and they believe DOD has not been transparent and forthcoming with the information that it has obtained. These officials stated that the continuous lack of collaboration and transparency has negatively affected the level of trust veterans and the veterans service organization have in DOD regarding its commitment to fully identify and disclose information regarding these tests. The representatives stated that it is imperative for DOD to be as transparent as possible so that Congress, veterans, and the public have reason to believe the cloak of secrecy regarding these tests has been lifted and individuals who were potentially exposed could receive appropriate medical care and benefits. DOD officials acknowledged the importance of keeping veterans informed so that they know that these tests are no longer classified, they are entitled to a medical screening for long-term health effects, and they can assist in DOD’s efforts to identify other individuals who might have been exposed. Until DOD is more transparent about its efforts to identify individuals who were potentially exposed during these previously classified tests, Congress, veterans, and the public could have reason to believe that the cloak of secrecy has not been lifted and not realize the reasonableness, effectiveness, success, and challenges of DOD’s current effort. DOD and VA have had limited success in notifying individuals who were potentially exposed to chemical and biological substances during Project 112 tests or testing that occurred outside of Project 112 due to several factors. First, DOD has inconsistently transmitted information about identified servicemembers to VA. Second, VA has not used all available resources to obtain contact information for servicemembers who were identified as having been potentially exposed. Finally, DOD has not taken any actions to notify civilians who have been identified. While DOD and VA have a process in place to share the names of servicemembers who are identified as having been potentially exposed to chemical and biological substances, the transmission of information between the two agencies has been inconsistent. To date, DOD has provided information to VA as agreed upon through an informal arrangement. Under the arrangement, DOD generally provides VA with the servicemember’s name, as well as any information related to the potential exposure that DOD uncovered during its investigation, such as the chemical or biological substance that was used, the dosage of the chemical or biological substance, and the date of the exposure. As of October 2007, DOD had used this process to transmit to VA approximately 20,700 names of servicemembers who had been potentially exposed to chemical or biological substances. The informal arrangement between DOD and VA did not establish a schedule for the exchange of information, so DOD provides newly acquired exposure information to VA in batches of varying size and at inconsistent intervals. When we began our work we found that DOD had not provided VA with any updates after September 2006 even though, as of June 2007, DOD had added approximately 1,800 additional servicemember names to its chemical and biological exposure database. Subsequent to our inquiries, however, DOD provided VA with an update in September 2007. According to DOD officials, regular updates to VA have been delayed because of a number of factors, including competing priorities such as current military operations, lack of personnel, database management issues, and lack of an impetus to take a proactive approach. Although limited personnel and competing priorities might be valid issues, until DOD provides regular updates of identified servicemembers to VA in a timely manner, VA will be unable to notify identified veterans about their potential exposure to chemical or biological substances. VA has not used certain available resources to obtain contact information for and to notify veterans who were identified as having been potentially exposed to chemical or biological substances. To notify veterans who were potentially exposed to chemical or biological substances during DOD tests, VA matches the list of potentially exposed veterans it obtains from DOD against its own database of veterans to find either contact information or a Social Security number. If no Social Security number is located, VA matches the available veterans’ information to information contained in the National Personnel Records Center. Once a Social Security number is obtained, VA usually uses a private credit bureau and on occasion has used the Internal Revenue Service database to obtain contact information for the veteran. In responding to a draft of this report, VA notes that it uses the credit bureau for a variety of reasons, including its up-to-date data transmissions from the Social Security Administration, expedience in responding, and general accuracy of information. As shown in table 3, as of December 2007, VA had obtained contact information for and sent notification letters to 48 percent of the names that DOD provided to them and that they may be able to contact. VA officials noted that while the total number of notification letters sent is 48 percent of the number of names that DOD has provided to them and that they may be able to contact, it represents all of the individuals for whom they were able to obtain contact information. A number of factors beyond VA’s control have impeded its ability to notify veterans of their potential exposure to chemical or biological substances. For example, some records have been lost or destroyed, and existing documentation contains limited information and often does not identify names of participants, while others were not turned in by the scientists who were conducting the research. When the records can be found, they do not necessarily identify the participants, but may instead refer to control numbers that were issued to the participants, which cannot be cross-referenced to other documents for identification. For those records that do include identification of participants, the information may contain only the participants’ initials, nicknames, or only first or last names. Also, since a number of these records do not include the participant’s military service number or social security number, it is difficult to determine the exact identity of these individuals. Further, the contact information that VA is able to obtain may not be accurate. For example, more than 860 notification letters have been returned as undeliverable to VA. However, VA is not using other available resources to obtain contact information to notify veterans. For example, while VA told us that it was using a company that is able to provide current contact information as a source, it had not coordinated with the Social Security Administration to obtain contact information for veterans receiving social security benefits or to identify deceased veterans using the agency’s death index and had not regularly used the Internal Revenue Service’s information. VA officials acknowledged that they had not directly used the death index and that a memorandum of understanding with the Social Security Administration might facilitate a new way to accomplish this. However, they noted the credit bureau receives weekly updates from the Social Security Administration’s death index. VA officials also acknowledged that it planned to make more frequent use of IRS databases. Until VA implements a more effective process to obtain contact information for veterans, some veterans will remain unaware of their potential exposure or the availability of health exams and the potential for benefits directly related to an exposure. DOD has not taken any actions to notify civilians who have been identified as having been potentially exposed during Project 112 tests and other chemical and biological tests, due in part to a lack of specific guidance defining the requirements to notify civilians. The Defense Authorization Act for FY 2003 required DOD to identify its tests or projects that may have exposed members of the armed forces to chemical or biological substances, but did not specifically address civilian personnel who may have been affected by these tests. However, in our 2004 report we recommended that DOD address the appropriateness of and responsibility for reporting new information, such as the identification of additional potentially exposed servicemembers, civilian employees, contractors, and foreign nationals who participated in the tests. In its response to our report, DOD concurred with our recommendation and stated that it would determine the appropriate reporting channels for civilian employees, contractors, and foreign national participants who were identified as being potentially exposed. However, DOD has not taken any action with the approximately 1,900 civilian names that it maintains, as shown in table 4. Instead, DOD has focused its efforts on the identification and notification of servicemembers who were potentially exposed. DOD officials stated that they have focused on identifying and notifying servicemembers since the primary impetus for their efforts to identify and notify individuals who may have been exposed has been requests for information from veterans and VA. OASD (HA) has not acted in part because it is unclear whether it is required to notify civilians or transmit civilian exposure information to another agency for notification. During our review, DOD and Department of Labor officials stated that they were unaware of a requirement for them to notify civilians of their potential exposure. However, our April 2005 report about civilian and contractor exposures to chemical substances in Vietnam identified compensation programs that might be available for civilians who were exposed during these chemical and biological tests if they come forward and present evidence that they were potentially exposed. Specifically, federal employees can file claims for workers compensation with their employing agency, which refers the claims to the Department of Labor under the Federal Employees Compensation Act. Employees who work under contract to the U.S. government can file workers compensation claims through their employers with the employers’ insurance carrier. Without an effort to develop and provide guidance for notifying civilians, those civilians who have been identified may not be aware of their potential exposure. Since World War II, potentially tens of thousands of military personnel and civilians have been exposed to chemical or biological substances during previously classified DOD tests. As this population becomes older, it will become more imperative for DOD and VA to identify and notify these individuals in a timely manner because they might be eligible for health care or other benefits. While DOD has concluded that continuing an active search for individuals potentially exposed during Project 112 has reached a point of diminishing returns, it has not conducted an informed cost- benefit analysis, which could guide DOD in identifying the extent to which it might need to take additional actions. Without conducting a sound and documented cost-benefit analysis that includes a full accounting of information known and the challenges associated with continuing to search for Project 112 participants, DOD will not be in a position to make an informed and transparent decision about whether any of the remaining investigative leads could result in meaningful opportunities to identify additional potentially exposed individuals. Furthermore, until DOD conducts such an analysis, Congress, veterans, and the public may continue to question the completeness and accuracy of DOD’s efforts. Moreover, while DOD has undertaken efforts to identify and notify individuals who were potentially exposed during tests outside of Project 112, the department has not worked with veterans and veterans service organizations during its current effort as required by the Defense Authorization Act for FY 2003, and it has not coordinated its efforts with other DOD and non-DOD organizations. Until DOD and VA undertake more effective and efficient efforts to identify and notify potentially exposed individuals—including consistent guidance about the scope of work, such as clearly defined goals and objectives and agreement on the type and amount of information that is necessary to collect; effective internal controls and oversight practices; coordination with other entities to leverage existing information; regular updates to VA; and utilization of all available resources—Congress, veterans, and the public may continue to question DOD and VA’s commitment to this effort. Furthermore, in the absence of transparency about these previously classified tests and DOD’s efforts to identify individuals who were potentially exposed, Congress, veterans, and the public could have reason to believe that the cloak of secrecy has not been lifted and may not understand the success and challenges of DOD’s current effort. While DOD and VA have developed a process for notifying servicemembers who were potentially exposed, it is unclear whether DOD or any other agency, such as the Department of Labor, is required to notify potentially exposed civilians who are identified. Therefore, without specific guidance that defines the requirements, roles and responsibilities, and mechanisms to notify civilians who have been potentially exposed to chemical or biological substances, these individuals might continue to be unaware of their circumstances. We are suggesting the Congress consider the following two matters: To provide greater transparency and resolve outstanding questions related to DOD’s decision to cease actively searching for the identification of individuals associated with Project 112, Congress should consider requiring the Secretary of Defense to consult with and address the concerns of VA, veterans, and veterans service organizations; to conduct and document an analysis that includes a full accounting of information known, and the related costs, benefits, and challenges associated with continuing the search for additional Project 112 participants; and to provide Congress with the results of this analysis. Our draft report addressed this recommendation to the Secretary of Defense; however, because DOD disagreed, we elevated this to a matter for congressional consideration. To ensure that civilians who were potentially exposed to chemical or biological substances as a result of tests conducted or sponsored by DOD are aware of their circumstances, Congress should consider requiring the Secretary of Defense, in consultation with the Secretary of Labor, to develop specific guidance that defines the requirements, roles and responsibilities, and mechanisms to notify civilians who have been potentially exposed to chemical or biological substances. To ensure a sound and documented process for DOD’s decision regarding the identification of individuals associated with Project 112, we recommend that the Secretary of Defense direct the Office of the Under Secretary of Defense for Personnel and Readiness to conduct and document an analysis that includes a full accounting of information known, and the related costs, benefits, and challenges associated with continuing the search for additional Project 112 participants, and to provide Congress with the results of this analysis. In developing the analysis, DOD should consult with and address the identified concerns of VA, veterans, and veterans service organizations. To ensure that DOD’s current effort to identify individuals who were potentially exposed during chemical and biological tests outside of Project 112 are more efficient, effective, and transparent, and to ensure that its databases contain accurate information, we recommend that the Secretary of Defense direct the Office of Under Secretary of Defense for Acquisition, Technology, and Logistics to take the following four actions: in coordination with the Office of the Under Secretary of Defense for Personnel and Readiness and the Secretary of Veterans Affairs, modify the guidance about the scope of work for its current effort, such as the statement of work and concept of operations plan, to clearly define consistent, reasonable, and acceptable goals and objectives, and the type and amount of information that will need to be collected to meet these goals and objectives; implement effective internal controls and oversight practices, such as periodic site visits, regular assessments of the contactor’s efforts, and quality assurance reviews of the information provided by the contractor; coordinate and communicate with other entities that previously identified exposed individuals to leverage existing information, including institutional knowledge and documents; and make its efforts transparent with regular updates to Congress, the public, and veterans service organizations. To ensure that DOD has taken appropriate action in its efforts to notify servicemembers who were potentially exposed, we recommend that the Secretary of Defense direct the Office of the Under Secretary of Defense for Personnel and Readiness to take appropriate action to address the factors—such as competing priorities and database management weaknesses—affecting DOD’s ability to forward the names of potentially exposed individuals to VA in a timely and effective manner. To ensure that all veterans who have been identified as having been potentially exposed to chemical or biological substances have been notified, we recommend that the Secretary of Veterans Affairs take steps to increase its use of available resources, such as the Internal Revenue Service, to implement a more efficient and effective process for obtaining contact information for living veterans. We requested comments from DOD, VA, and the Department of Labor on a draft copy of this report. DOD generally agreed with five recommendations, but disagreed with the first recommendation to conduct and document a cost-benefit analysis associated with continuing the search for additional Project 112 participants, and to provide Congress with the results of this analysis. VA agreed with one recommendation and partially agreed with another recommendation that pertained to its activities. The Department of Labor did not provide us any comments. Because DOD disagreed with the recommendation to conduct and document a cost-benefit analysis associated with continuing the search for additional Project 112 participants and has not adequately addressed our May 2004 recommendation to determine the feasibility of addressing unresolved issues associated with Project 112, we added a Matter for Congress to consider directing the Secretary of Defense to conduct such an analysis. DOD and VA also provided technical comments, which we incorporated as appropriate. DOD’s and VA’s comments are reprinted in appendices II and III, respectively. DOD agreed to and has in some cases begun taking action to respond to five of the recommendations. Specifically, DOD stated that it has already coordinated on updating program goals and objectives for the identification of individuals who were potentially exposed during chemical and biological tests outside of Project 112 and is revising the statement of work, implementation plan, and concept of operations to ensure consistent guidance and deliverables. DOD also stated that it has taken steps to increase oversight of the project and has established an implementation plan with OASD (HA) delineating oversight responsibilities. In addition, DOD stated that it will take steps to determine if other organizations are conducting similar work to identify potentially exposed individuals and will coordinate and leverage all available information. The department also stated that it will expand its current efforts to update the public and make efforts more transparent. Finally, DOD and VA are in the process of discussing short-term and long-term improvements necessary for improving the transfer of information to VA in a timely and effective manner. We believe these are positive steps that, when completed, will address the intent of our recommendations. DOD did not agree with the first recommendation to conduct and document an analysis that includes a full accounting of information known, and the related costs, benefits, and challenges associated with continuing the search for additional Project 112 participants, and to provide Congress with the results of this analysis. DOD stated that it believes it made a full accounting of its efforts available to Congress in 2003, that it has not received any credible leads that would allow DOD to continue its research, and that it currently knows of no other investigative leads that would meaningfully supplement what it believes to be a total picture of Project 112. However, as discussed in our May 2004 report, we identified a number of credible leads that could possibly result in additional Project 112 information. In addition, as discussed in this report, almost 600 additional individuals who were potentially exposed during Project 112 (more than a 10 percent increase) have been identified by non- DOD sources since DOD’s 2003 report to Congress and its decision to cease actively searching for additional exposures. In light of the increasing number of individuals who have been identified since DOD provided its report to Congress in 2003 and ceased its active search for additional individuals, until the department provides a more substantive analysis that supports its decision to cease active searches for additional individuals potentially exposed during Project 112 tests, Congress and veterans may continue to question the completeness and level of commitment to this effort. Because DOD has disagreed with our recommendation and has not adequately addressed our May 2004 recommendation to determine the feasibility of addressing unresolved issues associated with Project 112, we have added a Matter for Congress to consider directing the Secretary of Defense to conduct such an analysis. In response to our recommendations, VA agreed to work with DOD to modify the guidance about the scope of work for its current effort to clearly define consistent, reasonable, and acceptable goals and objectives; and the types and amount of information that will need to be collected to meet these goals and objectives. VA also agreed to contact the Internal Revenue Service to determine if a more timely response can be obtained from them to assist VA in notifying individuals potentially exposed to chemical or biological substances. We believe these steps are consistent with the intent of our recommendations. However, VA disagreed with a part of our recommendation that it needs to pursue information from the Social Security Administration since the credit bureau that VA uses to obtain contact information already receives the same information from the Social Security Administration. Accordingly, we adjusted our recommendation to the Secretary of Veterans Affairs so that it did not refer to the Social Security Administration as another source of information. We are sending copies of this report to other interested congressional committees, the Secretary of Defense, the Secretary of Veterans Affairs, and the Secretary of Labor. We will also make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you have any questions about this report, please contact me at (202) 512-5431 or dagostinod@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made contributions to this report are listed in appendix IV. To assess the Department of Defense’s (DOD) efforts since 2003 to identify servicemembers and civilians who may have been exposed to chemical or biological substances used during tests conducted under Project 112, we reviewed and analyzed documents pertaining to Project 112, including DOD’s 2003 Report to Congress: Disclosure of Information on Project 112 to the Department of Veterans Affairs. We interviewed officials at the Office of the Secretary of Defense, Washington, D.C., including the Under Secretary of Defense for Acquisition, Technology, and Logistics, and the Under Secretary for Personnel and Readiness. We also interviewed officials at the Office of the Assistant Secretary of Defense for Health Affairs who were responsible for conducting DOD’s investigation of Project 112 tests and have been designated as the single point of contact for providing information related to tests and potential exposures during Project 112. We interviewed officials at the Institute of Medicine and reviewed their 2007 report on the long-term health effects of participation in the shipboard hazard and defense tests of Project 112. In addition, we reviewed and analyzed our prior reports as well as reports of other organizations to provide a historical and contextual framework for evaluating DOD’s efforts. To evaluate DOD’s current effort to identify servicemember and civilian exposures that occurred during activities outside of Project 112 tests, we reviewed and analyzed reports, briefings, and documents and interviewed officials at the Office of the Secretary of Defense, Washington, D.C., including the Under Secretary of Defense for Acquisition, Technology, and Logistics and the Under Secretary of Defense for Personnel and Readiness. We also interviewed officials at the Office of the Assistant Secretary of Defense for Health Affairs, who have been designated as the single point of contact for providing information related to tests and potential exposures outside of Project 112. In addition, we interviewed officials at the U.S. Army Medical Research Institute of Infectious Diseases and the U.S. Army Medical Research and Materiel Command, Fort Dietrich, Maryland; the Department of Veterans Affairs, Washington, D.C.; the Institute of Medicine, Washington, D.C.; the Vietnam Veterans of America, Silver Spring, Maryland; and DOD’s contractor currently conducting research to identify potential exposures that occurred outside of Project 112. We also evaluated DOD’s methodology for identifying servicemembers and civilians who may have been exposed to chemical or biological substances by observing the process the contractor uses to conduct research at repositories containing documents related to chemical and biological exposures from tests and other activities, such as the transportation and storage of chemical and biological substances. We interviewed officials and observed storage facilities at the three chemical or biological substance exposure record repositories where the contractor was currently conducting its work: Edgewood Chemical and Biological Center Technical Library, Aberdeen Proving Grounds, Maryland; U.S. Army Research, Development, and Engineering Command Historical Office, Aberdeen Proving Grounds, Maryland; and U.S. Army Medical Research Institute of Infectious Diseases Technical Library, Fort Detrick, Maryland. In addition, we interviewed officials and observed the records storage area at the U.S. Army Medical Research Institute of Infectious Diseases Medical Records Office, Fort Detrick, Maryland, where information about Operation Whitecoat is maintained. We also reviewed DOD’s outreach efforts and the extent to which DOD coordinated with other agencies that might have useful information, including the Department of Veterans Affairs (VA), the Department of Labor, the Institute of Medicine, and the Vietnam Veterans of America. To evaluate VA’s process to notify servicemembers whom DOD has determined may have been exposed to a chemical or biological substance, we interviewed VA officials with the Veteran’s Benefit Administration, Veteran’s Health Administration, and Office of Planning and Policy, and gathered data concerning their success in making notifications. In particular, we documented the number of servicemembers whose names had been provided to VA by DOD, the extent to which notification letters were sent, the extent to which veterans were deceased, and the number of cases where sufficient documentation was not available to obtain contact information to make notifications. We assessed the reliability of DOD’s and VA’s data by interviewing agency officials knowledgeable about the data and by reviewing existing information about the data and the systems used to maintain and produce them. Although we found that there were potential problems with the quality and reliability of the information, we determined that the data were sufficient for the purposes of this report. We conducted this performance audit from June 2007 to February 2008 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the contact named above, Robert L. Repasky (Assistant Director), Tommy Baril, Renee S. Brown, Brian D. Pegram, Steven Putansu, Terry L. Richardson, and Karen Thornton made key contributions to this report. Agent Orange: Limited Information Is Available on the Number of Civilians Exposed in Vietnam and Their Workers’ Compensation Claims. GAO-05-371. Washington, D.C.: April 22, 2005. Chemical And Biological Defense: DOD Needs to Continue to Collect and Provide Information on Tests and Potentially Exposed Personnel. GAO- 04-410. Washington, D.C.: May 14, 2004. Human Experimentation: An Overview on Cold War Era Programs. GAO/T-NSIAD-94-266. Washington, D.C.: September 28, 1994. Veterans Disability: Information From Military May Help VA Assess Claims Related to Secret Tests. GAO/NSIAD-93-89. Washington, D.C.: February 18, 1993.

Tens of thousands of military personnel and civilians were potentially exposed to chemical or biological substances through Department of Defense (DOD) tests since World War II. DOD conducted some of these tests as part of its Project 112 test program, while others were conducted as separate efforts. GAO was asked to (1) assess DOD's efforts to identify individuals who were potentially exposed during Project 112 tests, (2) evaluate DOD's current effort to identify individuals who were potentially exposed during tests conducted outside of Project 112, and (3) determine the extent to which DOD and the Department of Veterans Affairs (VA) have taken action to notify individuals who might have been exposed during chemical and biological tests. GAO analyzed documents and interviewed officials from DOD, VA, the Department of Labor, and a veterans service organization. Since 2003, DOD has stopped actively searching for individuals who were potentially exposed to chemical or biological substances during Project 112 tests, but did not provide a sound and documented basis for that decision. In 2003, DOD reported it had identified 5,842 servicemembers and estimated 350 civilians as having been potentially exposed during Project 112, and indicated that DOD would cease actively searching for additional individuals. However, in 2004, GAO reported that DOD did not exhaust all possible sources of information and recommended that DOD determine the feasibility of identifying additional individuals. In response to GAO's recommendation, DOD determined continuing an active search for individuals had reached the point of diminishing returns, and reaffirmed its decision to cease active searches. This decision was not supported by an objective analysis of the potential costs and benefits of continuing the effort, nor could DOD provide any documented criteria from which it made its determination. Since June 2003, however, non-DOD sources--including the Institute of Medicine--have identified approximately 600 additional names of individuals who were potentially exposed during Project 112. Until DOD provides a more objective analysis of the costs and benefits of actively searching for Project 112 participants, DOD's efforts may continue to be questioned. DOD has taken action to identify individuals who were potentially exposed during tests outside of Project 112, but GAO identified four shortcomings in DOD's current effort. First, DOD's effort lacks clear and consistent objectives, scope of work, and information needs that would set the parameters for its efforts. Second, DOD has not provided adequate oversight to guide this effort. Third, DOD has not fully leveraged information obtained from previous research efforts that identified exposed individuals. Fourth, DOD's effort lacks transparency since it has not kept Congress and veterans service organizations fully informed of the progress and results of its efforts. Until DOD addresses these limitations, Congress, veterans, and the American public can not be assured that DOD's current effort is reasonable and effective. DOD and VA have had limited success in notifying individuals potentially exposed during tests both within and outside Project 112. DOD has a process to share the names of identified servicemembers with VA; however, DOD has delayed regular updates to VA because of a number of factors, such as competing priorities. Furthermore, although VA has a process for notifying potentially exposed veterans, it was not using certain available resources to obtain contact information to notify veterans or to help determine whether they were deceased. Moreover, DOD had not taken any action with the civilian names, focusing instead on veterans since the primary impetus for the research has been requests from VA. DOD has refrained from taking action on civilians in part because it lacks specific guidance that defines the requirements to notify civilians. Until these issues are addressed, some identified veterans and civilians will remain unaware of their potential exposure.

The Army’s modular force transformation affected the Army’s combat units and the related command and support organizations in both the active and reserve components. The Army’s objective in redesigning its force structure was to create more units to meet operational needs and be more flexible in deploying independently while maintaining combat capabilities of division-based brigades. According to the Army, having more combat brigades with specialized equipment and specialist personnel would increase combat capability and add value for combatant commanders. To increase the flexibility of units, the Army standardized brigade combat teams in one of three designs—armored brigade, infantry brigade, or Stryker brigade (see fig. 1). The Army’s new modular units were designed, equipped, and staffed differently than the units they replaced, and thus the transformation required many changes, such as new equipment and facilities, a different mix of skills and occupational specialties among Army personnel, and significant changes to training and doctrine. A key change was the reduction in the number of maneuver battalions within the modular units from three battalions per brigade under the division-based organization to two battalions for most brigade combat teams. Critics of the decision to have only two battalions raised concerns about whether the new structure would maintain as much combat capability as the division-based battalion. However, the Army expected to increase the modular brigade combat teams’ capability through specialized equipment and personnel, called “key enablers.” Since 2004, when the Army introduced its modular restructuring initiative, the Army has made multiple adjustments to its original plans for restructuring its operational force. The Army’s initial restructuring plan called for 77 modular brigade combat teams—43 active-duty brigade combat teams and 34 National Guard brigades. As of fiscal year 2013, the Army had 71 brigade combat teams, consisting of 43 active-duty brigade combat teams and 28 National Guard brigades. In 2013, the Army announced plans for another change in the structure of the modular force related to the need to reduce the active-duty component from 570,000 to 490,000 soldiers by fiscal year 2015. As figure 2 shows, the Army plans to reduce the number of active-duty brigade combat teams from a high of 45 teams to 32 teams by fiscal year 2015. The number of National Guard brigades would remain at 28, bringing the total of brigade combat teams to 60. In addition to the reduction in the numbers of brigade combat teams, the Army plans to refine the designs of the remaining brigades to add engineering and artillery capabilities, as well as increase the number of maneuver battalions from two to three for most brigade combat teams. According to the Army, the modifications to the modular force would enable it to preserve operational capability and flexibility, while reducing the number of soldiers in the active-duty component. Our body of work on the Army’s modular force transformation includes seven reports and three testimonies. The related work is listed at the end of this report. Based on our work, we made several recommendations to the Secretary of Defense and to the Army intended to improve the information on the Army’s transition to a modular design that the Army provided to decision makers in Congress. Because of the magnitude of the Army’s transformation plans and concerns about their affordability, we initially conducted work under the Comptroller General’s statutory authority and examined both the force structure and cost implications of the Army’s transformation into a modular force. Subsequently, Congress enacted requirements that the Army submit an annual report on its progress on its modular force transformation and that we review the Army’s report. According to the Army, the transition to the modular force structure, which began in 2004, was completed by the end of fiscal year 2013. The National Defense Authorization Act for Fiscal Year 2014 repealed the requirement for the Army and GAO to prepare annual reports about the Army’s modular force restructuring. The Army’s annual report on its modular force generally met legislative requirements by providing information that either fully or partially addressed each of the requirements. Our analysis showed that of the 14 legislative requirements, the report fully addressed 9 and partially addressed 5. The fully addressed requirements included information related to the status of key enabler personnel and equipment, an assessment of the modular force capabilities, and the status of doctrine for the modular force, among others. The partially addressed requirements included risks associated with shortfalls; mitigation strategies for shortfalls; scheduling for repairing, recapitalizing, and replacing equipment; itemizing information by active-duty and reserve components; and comments by the National Guard and Army Reserve regarding key enabler personnel and equipment. By fully or partially addressing the requirements, the Army’s 2013 report provided more thorough information to congressional decision makers on the Army’s progress in its modular force transformation than previous reports. Army officials gave several reasons why the report did not fully address some of these requirements. For example, the Army’s report discussed mitigation strategies for personnel shortfalls but not for all equipment shortfalls. According to Army officials, the Army mitigated risk by providing equipment to the next deploying units. In other cases, the Army chose not to mitigate equipment shortfalls due to the costs involved or because the specific equipment item no longer met the needs of the modular force. Additionally, the Army did not fully report on a schedule for the repair of equipment because, according to Army officials, the number of battle losses and the related amount of wear and tear on equipment returning from overseas operations was unpredictable. However, the Army provided some general information about its repair schedule for fiscal years 2013 and 2014, such as when the Army expects to begin addressing postcombat equipment repairs. In addition, according to officials the report did not itemize information by component because the report included a separate section with comments from the reserve components. However, the comments by the reserve components did not include all required information, such as identifying risks and mitigation strategies associated with equipment shortfalls. Table 1 summarizes our assessment of the extent to which the Army’s annual modular force report included each of the legislative requirements. The Individual Ready Reserve is a subcategory of the Ready Reserve of the Army Reserve. Members of the Individual Ready Reserve include individuals who were previously trained during periods of active service, but have not completed their service obligations; individuals who have completed their service obligation and voluntarily retain their reserve status; and personnel who have not completed basic training. Most of these members are not assigned to organized units, do not attend weekend or annual training, and do not receive pay unless they are called to active duty. reported that Army officials responsible for providing information on the modular force progress were not given sufficient guidance to ensure the completeness of its report. We recommended that the Army provide guidance on the level and type of detail needed for each office within the Army responsible for providing information on the Army’s progress in meeting modularity requirements. In preparing the fiscal year 2013 report, the Army implemented our recommendation to provide guidance to Army officials to ensure the completeness of its report. The Director of the Army Staff sent a memorandum in March 2013 to Army staff that outlined the coordination process for preparing the fiscal year 2013 report, identifying each office responsible for providing information for the report. In addition, Army officials coordinating the report held meetings with each office early in the process to ensure they complied with the mandated language. Additionally, the Army provided the offices with a list of the key enabler items to report on rather than letting the offices interpret what to report. By implementing our recommendation, the Army’s fiscal year 2013 modularity report generally met legislative requirements and provided congressional decision makers with additional information on the Army’s progress in its modular force restructuring. The Army has completed its transition to modular brigade combat team designs, but it has not addressed the key challenges of creating a results- oriented plan, creating realistic cost estimates, and planning comprehensive assessments that we identified in our work since 2005. In our prior reports between 2005 and 2008 on the Army’s modular transformation, we made 20 recommendations intended to help the Army address these challenges. The Army generally agreed with 18 of those recommendations, but it has so far implemented only 3 of them. The Army has begun to create a results-oriented plan, but more work remains to create realistic cost estimates and plan comprehensive assessments. As the Army continues to make changes to the structure of its modular brigade combat team—including adding a maneuver battalion to the infantry and armored brigade combat teams—it has the opportunity to incorporate lessons learned and reduce the risk of repeating mistakes from its recent experience in changing its force structure. In order to improve the Army’s focus on the relationship between key enabler investments and results and the completeness of the information that the Army provides Congress, between 2005 and 2008 we made four recommendations regarding creating a results-oriented plan. Our recommendations were rooted in key practices that we have identified for assisting organizational transformations, suggesting agencies can be more results-oriented by focusing on a key set of principles and priorities at the outset of the transformation as well as setting implementation goals to show progress from the beginning of the transformation. Our recommendations to the Army included developing a plan to identify authorized and projected personnel and equipment levels as well as an assessment of the risks associated with any shortfalls. The Army generally agreed with three of the recommendations and ultimately implemented two of them. For example, the Army concurred with but did not implement our 2006 recommendation to provide the Secretary of Defense and Congress with details about the Army’s equipping strategy; when we reiterated a similar recommendation in 2008, the Army implemented it by providing more detailed information on its progress in providing the modular force with key personnel and equipment enablers. For a full list of our past recommendations and their implementation status regarding creating a results-oriented plan for the Army’s modular force transformation, see appendix II. Our work since 2005 found that the Army began its modular transition without creating a results-oriented plan with clear milestones to provide units with specially trained personnel and specialized equipment. In 2005 and 2006, we reported that the Army began its modular transformation without creating a staffing plan that considered the size and composition of the modular force. For example, in 2005 we testified that the Army had begun its modular transformation without deciding on the number of brigade combat teams or finalizing the design of supporting units. Without finalized designs or key decisions, the Army did not have a complete understanding of the personnel needed to achieve its goals. As a result, the Army could not assure decision makers when modular units would have the required key enabler staff in place to restore readiness, and it experienced cost growth and timeline slippage in its efforts to transform to a modular and more capable force. In 2006, we testified and reported that the Army did not plan to fill some key intelligence positions required by its new modular force structure. Without continued and significant progress in meeting personnel requirements, the Army had to accept increased risk in its ability to support its combat forces, and it ultimately sought support for an increase in overall personnel from the Department of Defense (DOD) and Congress. Additionally, in 2005 and 2006 we reported that the Army did not develop an equipping plan to provide modular units the required quantities of key enabler equipment considered critical for the transformation. For example, in 2005 we testified that although the Army had some of its key enabler equipment on hand at the start of its modular transformation, the amount of equipment provided to brigade combat teams was well below the levels tested by the Army Training and Doctrine Command. As a result, officials from two divisions that we visited expressed concern over their soldiers’ ability to train and become proficient with some of this high- tech equipment because the equipment was not available in sufficient numbers. In 2006, we similarly testified that although active modular combat brigades were receiving considerable quantities of equipment, they initially lacked required quantities of items such as communications systems that were key for providing the enhanced intelligence, situational awareness, and network capabilities needed to help match the combat power of the Army’s former brigade structure. At that time, the Army’s modular combat brigade conversion schedule outpaced the planned acquisition or funding for some equipment requirements, and the Army had not defined specific equipping plans for brigades. By not completing development of its equipping strategy, the Secretary of Defense and Congress were not in a good position to assess the Army’s equipment requirements and the level of risk associated with the Army’s plans. Moreover, in 2008 we reported that although the Army had established over 80 percent of its modular units, it did not have a results-oriented plan with clear milestones in place to guide efforts to staff and equip those new units. The Army extended its estimates of how long it would take to equip the modular force from 2011 to 2019, but it provided few details about interim steps. While the Army projected that it would have enough personnel and equipment in the aggregate, its projections relied on uncertain assumptions related to meeting recruiting and retention goals as well as restoring equipment used in current operations. For example, the Army centered its equipping strategy on the Future Combat System, a longer-term transformation effort that comprised 14 integrated weapon systems and an advanced information network. The Army expected brigade combat teams equipped with the Future Combat System to provide significant warfighting capabilities to DOD’s overall joint military operations. However, in 2009, after 6 years and an estimated $18 billion invested, DOD canceled the Future Combat System acquisition program and instead identified alternate plans to modernize equipment. The cancelation of the Future Combat System presented setbacks to the equipping of the modular force. Without a results-oriented plan for equipment and staffing with realistic milestones, the Army could not assure decision makers when modular units would have the required equipment and staff in place to restore readiness. In 2013, when the Army announced plans to change its modular force designs and add a third battalion to most brigade combat teams, it incorporated some lessons we identified in our prior work and took some initial steps to create a results-oriented plan to guide implementation of the changes. The Army based decisions on which units to inactivate on quantitative and qualitative analyses and developed a timeline for the changes, with associated tasks and milestones. For example, as we reported in December 2013, the Army established a planning team for the brigade combat team reorganization to assess factors such as strategic considerations, military construction costs, and proximity to embarkation points, among others, and to develop stationing options for decision makers. Furthermore, the Army has developed a plan to reduce risk to the readiness of the force during this reorganization by providing equipment, personnel, and training resources to units currently deployed or deploying for operations or contingencies and then to seven brigade combat teams that will maintain a high level of readiness for 18 to 24 months. According to Army officials, these seven teams will remain at the highest level of readiness in order to support any planned or unexpected operations while the remaining brigade combat teams undergo their reorganizations and accept a risk of low readiness to respond to potential contingencies. Moreover, the Army Structure memorandum for 2015 through 2019 documented interim steps in reorganizing the Army modular force structure. For example, the Army identified inactivation and reorganization dates as well as changes to the tables of equipment for the brigade combat teams. Additionally, once the Army identified which units would be inactivated or reorganized, officials developed an online tracking system that provides information such as when the reorganization and associated training will occur and what tasks each reorganized unit will have to complete. According to an Army official, senior leaders—including three-star generals and the Chief of Staff of the Army—reviewed the online system to track progress in implementing planned changes. However, the Army has not always been able to implement its plans to achieve its goals. For example, the Army was not able to fulfill plans for some key enabler equipment such as through its Future Combat System program that were deemed critical to achieving the combat effectiveness of the modular brigade combat teams. The Army could face risks in implementing current plans without sustaining attention and following through on its plans for changes to the modular force design. By creating a results-oriented plan for the inactivations and reorganizations, the Army has established a baseline against which to measure performance. If the Army follows through its initial steps to create a results-oriented plan for changes to the modular force design, it would help to provide senior officials and Congress the ability to identify and mitigate any potential problems that may arise. In order to improve information available to decision makers on the cost of the Army’s plan for its modular force transformation, between 2005 and 2008 we made 10 recommendations regarding creating a realistic cost estimate. In the John Warner National Defense Authorization Act for Fiscal Year 2007, Congress specifically required the Army to report on a complete itemization of the amount of funds expended to date on the modular brigades and itemization of the requirements for the funding priorities. Our recommendations to the Army included submitting an annual cost plan that incorporated a clear definition of the costs the Army considered to be related to the modular transformation, estimates for equipment and personnel, and divergences from the plan as stated in the prior year’s report, among others. The Army generally agreed with all the recommendations, but it did not implement any of them. For example, the Army did not develop a plan for overseeing the costs related to the Army’s transformation to a modular force as we had recommended, stating that the administrative costs of such an effort would outweigh any benefit. For a full list of our past recommendations and their implementation status regarding creating realistic cost estimates for the Army’s modular force transformation, see appendix II. From 2005 through 2013, the Army did not create realistic cost estimates for implementing its modular force transformation. We reported in 2005 that the Army might not have estimated all potential costs for its modular force transformation because it had not made decisions related to force design, equipment, facilities, and personnel. We reported that the Army likely understated its estimates for equipment costs because it did not entirely reflect the cost of purchasing all the equipment needed to bring the planned units to the modular design—and therefore to the level of capability—that the Army validated in testing. In addition, Army officials were uncertain whether the personnel authorization was enough to support the modular transformation, putting costs at risk of increasing if the Army determined that the transformation required additional personnel. Additionally, the Army was uncertain of the costs of constructing permanent facilities because it did not incorporate proposals for base realignment and closure and restationing of personnel from overseas. As costs grew due to these uncertainties, the Army required additional funding or needed to accept reduced capabilities among some or all of its units. By not developing a better understanding of costs associated with the modular force and a clearer picture of the effect of resource decisions on the modular force capability, DOD was not well positioned to weigh competing priorities or to provide congressional decision makers the information they needed to evaluate funding requests. Additionally, in 2007 we reported that Army officials did not identify how much additional funding they needed to fully equip modular units but they planned to request funds for additional equipment needs through DOD’s annual budget process. We noted that in the absence of a complete cost estimate, the Army would not be in a good position to identify detailed costs and provide transparency to Congress of its total funding needs. We also reported that the Army sought multiple sources of money without linking funding to its modular unit design requirements, thus complicating decision makers’ ability to assess the Army’s progress in fully equipping the modular force. In 2007, we reported that the Army estimated the modular restructuring could cost $52.5 billion—more than two-and-a-half times greater than its initial cost estimate of $20 billion in 2004. In 2009, the Army reported to Congress that it could no longer itemize modular costs because all Army personnel and equipment budgets support the modular force. Without linking funding to requirements, decision makers would have difficulty assessing the Army’s progress in meeting its goals, knowing what resources would be required to equip and staff modular units, and balancing funding requests for these initiatives with other competing priorities. As the Army continues planning for changes to its modular force design, Army officials compiling the fiscal year 2013 report were not aware of any cost estimates developed for inactivating 12 brigade combat teams and adding a third battalion to the infantry and armored brigade combat teams. According to Army officials, the Army expects that the costs will be low because 9 of the 12 inactivations would involve reorganizations within the same installation, minimizing military construction and personnel relocation costs. However, the Army did not provide us with any detailed cost analysis. Further, the Army did not provide cost estimates for military construction and personnel relocation costs for those reorganizations occurring across installations or for associated doctrine development or training for the reorganized units. Without realistic cost estimates, the Army may encounter many of the same risks that we reported previously. The Army plans to reduce its budget by $170 billion between fiscal years 2013 and 2022, and the reorganization of the brigade combat teams and the associated reduction of 80,000 personnel should contribute to the cost savings. However, given two decades of GAO reports delineating DOD’s overly optimistic planning assumptions in budget formulation, which often lead to costly program delays, we believe that not having a detailed cost analysis could lead to increases in the Army’s incremental costs for its reorganization. Specifically, if costs grow due to uncertainties regarding equipment and personnel movement costs, the Army may require additional funding or need to accept reduced capabilities among some or all of its units. In the absence of a complete cost estimate, the Army may be unable to assure Congress that the Army has identified the total funding needs for reorganizing modular forces. We continue to believe that realistic cost estimates would enhance DOD decision makers’ ability to weigh competing priorities in a fiscally constrained environment and provide Congress with the information needed to evaluate funding requests. In order to assess the implications of changes to the Army force structure in terms of the goals of modular restructuring, from 2005 through 2008 we made six recommendations regarding creating comprehensive assessment plans. Standards for Internal Control in the Federal Government state that agencies should provide reasonable assurance to decision makers that their objectives are being achieved and that decision makers should have reliable data to determine whether they are meeting goals and using resources effectively and efficiently. Our recommendations to the Army included developing a comprehensive plan for assessing the Army’s progress toward achieving the benefits of the modular transformation that incorporated quantifiable metrics and addressed a wide range of both traditional and irregular security challenges. The Army generally agreed with five of the recommendations but implemented only one. For example, the Army agreed with but did not implement our recommendation to develop a comprehensive assessment plan that includes steps to evaluate modular units in full-spectrum combat. However, the Army acted upon our recommendation by assessing aspects of the modular force and refining its modular designs based on lessons learned in the areas of equipment, doctrine, and training. For a full list of our past recommendations and their implementation status regarding completing a comprehensive assessment plan for the Army’s modular force transformation, see appendix II. Since 2004, the Army has made many refinements to its modular design based on lessons learned and limited assessments of specific capabilities, but it has not completed a comprehensive assessment plan to measure the extent that its modular force transformation is meeting performance goals. In 2006, we reported that the Army did not have a comprehensive and transparent approach to measure progress against stated modularity objectives, assess the need for further changes to modular designs, and monitor implementation plans. While DOD had identified the importance of establishing objectives that translate into measurable metrics that in turn provide accountability for results, the Army had not established outcome-related metrics linked to most of its modularity objectives. Further, we reported that although the Army analyzed lessons learned from Iraq and training events, the Army did not have a long-term comprehensive plan for further analysis and testing of its modular combat brigade designs and fielded capabilities. As a result, decision makers did not have sufficient information to assess the capabilities, cost, and risks of the Army’s modular force implementation plans. Moreover, in 2007 we reported that the Army was evaluating and applying lessons learned from its counterinsurgency operations. However, it did not have a comprehensive assessment plan to determine whether fielded modular unit designs met the Army’s original goals for modular units across the full spectrum of low- and high-intensity warfare, and it did not have outcome-oriented metrics that helped to measure progress in achieving the goals of the modular force. The Army evaluated the experiences of modular units deployed to Iraq and Afghanistan and had made some changes in unit designs based on these lessons; however, the Army did not develop a plan for assessing modular units in high- intensity combat operations. In seeking approval to establish modular units, the Army identified a number of planned benefits associated with them, such as providing the same or better combat effectiveness of the Army’s division-based brigades. However, the Army limited its evaluations to the performance of modular units during predeployment exercises and counterinsurgency operations and did not evaluate their performance across the full spectrum of combat operations that include large-scale, high-intensity combat operations. As a result, the Army did not have a clear way to measure the extent to which new modular brigades were as effective as its division-based brigades for a range of missions. Without a comprehensive assessment plan that included a wider range of potential missions, the Army may have missed opportunities to strengthen its designs. Additionally, we reported in our 2008 report on Army modularity that the Army tested its units with the full complement of required equipment and personnel, and not at the somewhat lower level of personnel and equipment the Army actually provided to units. As a result, the Army assessment of whether the capabilities that it was fielding could perform mission requirements did not capture realistic missions and outcomes. Without an analysis of the capabilities of the modular force at realistic personnel and equipment levels, the Army was not in a position to assess whether the capabilities that it was fielding could perform mission requirements. Faced with decreasing financial resources and increasing ambiguity regarding future missions, decision makers sought to determine how to organize combat formations to best position the Army for a range of possible missions. The Army Training and Doctrine Command Analysis Center prepared an analysis in May 2012 to consider whether the Army should add a third maneuver battalion to the armored and infantry brigade combat team designs. To begin this analysis, the Army assembled 23 commanders of brigade combat teams to gather insights into the effectiveness of the both the two- and three-battalion designs across a wide range of possible future demands, such as major combat operations, low-threat activities, and enhanced protective posture. In doing so, the Army considered several factors we reported on in our prior work, including reviewing the full spectrum of low- and high-intensity warfare and identifying assessment metrics such as security of vehicles. However, the analysis was not an assessment of the Army’s prior performance under a two-battalion construct. Rather, the analysis was a projection of how to organize the Army for future demands and thus did not meet the intent of our past recommendations. The Army has developed some plans to conduct assessments and capture lessons learned as it changes its modular force design, but it has not formalized these plans with a detailed methodology, data-collection procedures, or outcome-based metrics. According to Army officials, the Army plans to conduct assessments and capture lessons learned during the reorganization of the brigade combat teams. The Army issued an execution order for the Army Training and Doctrine Command to conduct assessments for this reorganization. According to a senior official from the Army Training and Doctrine Command, the Army plans to collect baseline metrics on the brigades both prior to and following their reorganization. Additionally, the Army has developed new mission- essential task lists for the infantry, Stryker, and armored brigade combat teams on which to base assessments. When the brigade combat teams participate in training exercises in their new organizational designs, assessors will evaluate how well the teams achieve their mission- essential tasks. The assessors can then adjust the training if the brigade combat teams are having difficulties understanding the new tasks. However, the Army officials stated that there is no checklist or detailed plan on how to conduct these assessments or what outputs to measure. Moreover, according to Army officials, the assessments are dependent upon receiving sufficient funding and potentially may not occur. If the Army created a comprehensive assessment plan, it could help enable the Army to clearly measure the extent to which it is achieving desired benefits in the design of its modular force. We are not making new recommendations in this report. However, this report’s analysis provides additional support for past recommendations to develop realistic cost estimates and to create a comprehensive assessment plan to measure achievement of desired benefits. We provided a draft of this report to DOD for comment. Army officials provided oral comments on the draft indicating that DOD concurred with our report. We are sending copies of this report to the appropriate congressional committees; the Secretary of Defense; and the Secretary of the Army. The report is also available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (404) 679-1816 or pendletonj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix III. To determine the extent that the Army addressed legislative requirements to report information regarding key equipment and personnel needs for its reorganized modular force, three analysts independently reviewed the Army’s fiscal year 2013 modularity report and compared the report to the legislative requirements. A fourth analyst adjudicated the differences in cases of dispute and determined a final categorization. The analysts used an evaluation tool that listed the legislative requirements and categorized the extent to which the Army’s report included information required for each reporting element from the mandate. The categories were “Addressed,” “Addressed in part,” and “Not addressed.” “Addressed” meant the report thoroughly addressed all components of the requirement. “Addressed in part” meant that one or more, but not all, components of the requirement were addressed, or that all components of the requirement were addressed, but the information provided was insufficient to answer the requirement fully due to limitations in the data or information provided. “Not addressed” meant that the report did not address any part of the requirement. To gain a full understanding of the method and data the Army used to prepare the report, clarify the significance of the information presented in the report, and obtain additional information that addressed the legislative reporting requirements, we met with Army officials knowledgeable about compiling information for the report, about key enabler personnel and equipment, and about equipment reset, doctrine, and force structure changes. Specifically we met with Army officials from the Offices of the Deputy Chiefs of Staff for Personnel (G-1), Logistics (G-4), Operations and Plans (G-3/5/7), and Programs (G-8); Training and Doctrine Command; Army National Guard; and Army Reserve who provided data for the Army’s fiscal year 2013 modularity report. To gain a full understanding of the data the Army used to prepare the report, analysts reviewed documents the Army used to compile the report, including a Director of the Army Staff memorandum, the list of key enabler personnel and equipment required to be included in the report, and the Army Equipment Reset Update. To gain a full understanding of the progress made in fulfilling modularity requirements in the fiscal year 2013 report, we reviewed the fiscal year 2012 report to determine the extent of progress made between fiscal years 2012 and 2013. After the initial interview, document review, and completion of the evaluation tool and adjudication of the differences, the analysts determined that of the 14 legislative requirements, 9 were addressed and 5 were addressed in part. Analysts conducted a follow-up interview regarding the five legislative requirements that were addressed in part to obtain additional information or documentation on why the elements were not fully addressed. To identify challenges in the Army’s modular force restructuring over the past 10 years, we reviewed prior GAO reports evaluating the Army force structure. We also reviewed prior GAO reports evaluating technology and equipment related to the Army’s modular restructuring to identify challenges. From our review of prior reports, we identified three main challenges the Army faced during its modular force restructuring. To determine how the Army is addressing these challenges, we reviewed whether the Army implemented the recommendations in our prior GAO reports that evaluated the Army’s modular force structure. To determine whether the Army plans in its modular force structure reorganization to address challenges previously identified in our reports, we reviewed Army documents and interviewed Army officials. We reviewed documents including the Brigade Combat Team inactivation execution order, Army Structure Memorandum for fiscal years 2014 to 2019, Army of 2020 Analysis Supporting the Brigade Combat Team Design Decision, and the Army Campaign Plan portal, which the Army uses to track the reorganization. We met with officials knowledgeable about the changes to the brigade combat team designs to gain a full understanding of the plans the Army is developing to execute the changes. We also spoke with an official from the Army Training and Doctrine Command to learn about the Army’s plans to conduct comprehensive assessments of its modular force reorganization. We conducted this performance audit from September 2013 to April 2014 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. This appendix presents a list of (1) the 20 recommendations that we had previously made regarding the Army’s modular force transformation, (2) the Department of Defense (DOD) response to those recommendations, and (3) our analysis of whether the Army has addressed the issues that gave rise to the recommendations. From 2005 to 2008, we made 20 recommendations regarding the Army’s modular force transformation in the following four reports: Force Structure: Actions Needed to Improve Estimates and Oversight of Costs for Transforming Army to a Modular Force. GAO-05-926. Washington, D.C.: September 29, 2005. (Referred to below as September 2005 recommendations). Force Structure: Army Needs to Provide DOD and Congress More Visibility Regarding Modular Force Capabilities and Implementation Plans. GAO-06-745. Washington, D.C.: September 6, 2006. (Referred to below as September 2006 recommendations). Force Structure: Better Management Controls Are Needed to Oversee the Army’s Modular Force and Expansion Initiatives and Improve Accountability for Results. GAO-08-145. Washington, D.C.: December 14, 2007. (Referred to below as December 2007 recommendations). Force Structure: The Army Needs a Results-Oriented Plan to Equip and Staff Modular Forces and a Thorough Assessment of Their Capabilities. GAO-09-131. Washington, D.C.: November 14, 2008. (Referred to below as November 2008 recommendations). We tracked the recommendations for 4 years following each report’s publication and closed each one as either implemented or not implemented. We grouped our recommendations by the types of issues and challenges on which the recommendations focused: (1) creating a results-oriented plan, (2) developing realistic cost estimates, and (3) planning comprehensive assessments. The appendix lists the recommendations by these three key challenges. In addition to the contact named above, Margaret Morgan (Assistant Director), Alice Paszel, Richard Powelson, Kelly Rubin, Jodie Sandel, Amie Steele, and Sabrina Streagle made significant contributions to this report. Force Structure: Army’s Annual Report on Modularity Progress Needs More Complete and Clear Information to Aid Decision Makers. GAO-13-183R. Washington, D.C.: January 16, 2013. Force Structure: Assessment of Army Report on Fiscal Year 2011 Progress in Modular Restructuring. GAO-12-527R. Washington, D.C.: March 26, 2012. Force Structure: Assessment of Army Progress in Modular Restructuring, Prepositioned Equipment, and Equipment Reset. GAO-10-507R. Washington, D.C.: April 26, 2010. Force Structure: The Army Needs a Results-Oriented Plan to Equip and Staff Modular Forces and a Thorough Assessment of Their Capabilities. GAO-09-131. Washington, D.C.: November 14, 2008. Force Structure: Restructuring and Rebuilding the Army Will Cost Billions of Dollars for Equipment but the Total Cost Is Uncertain. GAO-08-669T. Washington, D.C.: April 10, 2008. Force Structure: Better Management Controls Are Needed to Oversee the Army’s Modular Force and Expansion Initiatives and Improve Accountability for Results. GAO-08-145. Washington, D.C.: December 14, 2007. Force Structure: Army Needs to Provide DOD and Congress More Visibility Regarding Modular Force Capabilities and Implementation Plans. GAO-06-745. Washington, D.C.: September 6, 2006. Force Structure: Capabilities and Cost of Army Modular Force Remain Uncertain. GAO-06-548T. Washington, D.C.: April 4, 2006. Force Structure: Actions Needed to Improve Estimates and Oversight of Costs for Transforming Army to a Modular Force. GAO-05-926. Washington, D.C.: September 29, 2005. Force Structure: Preliminary Observations on Army Plans to Implement and Fund Modular Forces. GAO-05-443T. Washington, D.C.: March 16, 2005.

The Army considers its modular force transformation, which began in 2004, to be its most extensive restructuring since World War II. The Army expanded the number of deployable units and incorporated advanced equipment and specialized personnel, but removed a maneuver battalion from its brigades. Throughout the transformation, GAO reported, testified, and made recommendations on associated challenges the Army faced. In 2013, the Army stated it had completed its transformation and submitted its last required report to Congress on its modular progress. It also announced plans to restore a maneuver battalion to most brigades. Congress mandated that GAO report annually on the Army's modular force. For this report, GAO (1) evaluates whether the Army addressed the legislative requirements in its modular force report and (2) provides an overview of any challenges that the Army faced in its modular force transformation and describes how the Army is addressing these challenges as it implements further changes in its force structure. GAO analyzed the Army's report against the legislative requirements, reviewed key Army reports, and spoke to Army officials. The Army's annual report on its modular force either fully or partially addressed all of the requirements mandated by law. GAO's analysis showed that of the 14 legislative requirements, the report fully addressed 9 and partially addressed 5. The requirements that were fully addressed included an assessment of the modular force capabilities and the status of doctrine for the modular force, among others. Some of the requirements that were partially addressed included information related to risks and mitigation strategies associated with shortfalls; scheduling for repairing, recapitalizing, and replacing equipment; and itemizing information by active-duty and reserve components. The 2013 report provided more thorough information to congressional decision makers on the Army's progress in its modular force transformation than previous reports. GAO's body of work since 2005 on the Army's modular restructuring found that the Army faced challenges in creating a results-oriented plan, developing realistic cost estimates, and planning comprehensive assessments. GAO made 20 recommendations from 2005 through 2008 to help address these challenges; the Army generally agreed with 18 of the recommendations but so far has implemented only 3. As the Army plans to restructure its modular force it has made some progress in creating a results-oriented plan, but more work remains in developing realistic cost estimates and planning comprehensive assessments. Creating a results-oriented plan. As the Army plans further changes to its modular force design, it has taken initial steps to create a results-oriented plan by developing a timeline with associated tasks and milestones. When the Army began its modular force transformation it did not create a plan with clear milestones to guide its efforts to fully staff and equip the modular force. By incorporating lessons identified in GAO's prior work as it makes further changes, the Army has established a baseline against which to measure performance and may provide decision makers the ability to mitigate any potential problems that may arise. Developing realistic cost estimates . From 2005 through 2013, the Army did not create realistic cost estimates or provide a reliable accounting of past spending or future funding needs for implementing its modular force transformation. As the Army plans further changes to its modular force design, it has not developed cost estimates for military construction, personnel relocation, or training for the reorganized units. GAO continues to believe that realistic cost estimates would better position the Army to weigh competing priorities in a fiscally constrained environment and provide Congress with the information needed to evaluate funding requests. Planning comprehensive assessments. Since 2004, the Army has made many changes to its modular design based on limited assessments, but it has not completed a comprehensive assessment plan to measure the extent that its modular force transformation is meeting performance goals. As the Army continues to make changes to its modular design, the Army plans to conduct assessments but has not identified outcome-oriented metrics to measure progress. If the Army created a comprehensive assessment plan, it could help decision makers identify capability gaps and mitigate risks. GAO is not making new recommendations, but this analysis provides additional support for past recommendations to develop realistic cost estimates and to create a comprehensive assessment plan to measure achievement of desired benefits. In oral comments on a draft of this report, the Army concurred with the report.

The federal government encourages federal prime contractors’ use of small businesses as subcontractors by requiring prime contractors to develop plans with stated goals for subcontracting to various types of small businesses. Federal regulations require a subcontracting plan for each contract or contract modification that exceeds $500,000 ($1 million for construction contracts) and has subcontracting possibilities. The Subcontracting Assistance Program is SBA’s vehicle for increasing the percentage of subcontract awards to small businesses and ensuring that small businesses have the maximum practicable opportunity to participate in the performance of federal government contracts. SBA’s Office of Government Contracting (OGC) oversees this program. SBA/OGC has six Area Offices responsible for all prime contractors’ subcontracting performance. The program is implemented by CMRs, who promote small business subcontracting in two primary ways, as described in SBA regulations.First, CMRs review prime contractors’ compliance with the requirements of their subcontracting plans. They conduct on-site compliance reviews at prime contractors’ facilities and validate how well the prime contractors are implementing their subcontracting plans. They also conduct “desk reviews,” which are reviews of relevant subcontracting reports that are submitted by prime contractors and completed without on-site visits. Second, CMRs conduct various marketing activities, such as marketing small businesses to prime contractors or matching certain types of small business subcontractors with prime contractors. CMRs perform this “matchmaking” through both personal introductions and the use of Web- based tools that help to connect prime contractors and subcontractors. CMRs also provide various educational activities (e.g., seminars and workshops) for prime contractors, subcontractors, and agency officials as part of their marketing activities. In describing the duties required of CMRs, SBA regulations do not place a relative order of importance on any of these responsibilities. In addition, CMRs work on various SBA special initiatives as part of their marketing activities. For example, CMRs promote SBA’s special 8(a) subcontracting initiative, which focuses on increasing the number of subcontracts to small disadvantaged businesses. CMRs also promote several special initiatives involving SBA’s Procurement Marketing and Access Network (PRO-Net). PRO-Net is a Web-based system that allows prime contractors to advertise potential subcontracting opportunities and small businesses to advertise their capabilities as subcontractors. CMRs install PRO-Net access stations at large prime contractor sites and public libraries, and train interested businesses and agencies in the use of PRO- Net to match prime contractors and subcontractors. Finally, CMRs conduct special training for agencies and contractors on SBA’s historically underutilized business zones (HUB-Zone) Empowerment Contracting Program, which encourages economic development in HUB Zones. SBA/OGC also employs other contract specialists, who focus on small businesses as prime contractors rather than subcontractors. These include Procurement Center Representatives (PCRs), size determination specialists, and Certificate of Competency (COC) specialists. PCRs work with agencies to determine whether it is appropriate for acquisitions not set aside for small businesses to be set aside. Size determination specialists determine whether a small business meets existing size standards for all procurement programs for which status as a small business is required. COC specialists review a contracting officer’s determination that the small business in question is not competent to perform on a particular contract. The CMR role is conflicted and in decline. Over the past few years, the CMR role has become part-time, and CMRs now usually have additional roles that often take priority. CMRs appear to spend slightly more time on marketing activities than on compliance monitoring, and they now rely much more frequently on desk reviews than on-site visits for the latter. In addition, workloads and prime contractor coverage vary greatly between CMRs. SBA officials and CMRs have several concerns about the CMR role. The CMR position is usually part-time. At the end of fiscal year 2001, about 90 percent of the CMRs had other substantial responsibilities in addition to their CMR duties. At that time, only 4 of the 39 CMRs were full-time CMRs. As figure 1 shows, the number of part-time CMRs has grown over time, increasing twelve-fold since 1992. (SBA officials anticipate the continued assignment of additional roles to CMRs and other contracting specialists.) In addition, despite the fact that a larger number of staff had CMR duties at the end of fiscal year 2001 than in fiscal 1992, the number of CMR full- time equivalents (FTEs) declined 28 percent—from 25 to about 18—during this period. The SBA officials and CMRs we interviewed told us that part-time CMRs have a variety of additional responsibilities, including serving as PCRs, COC specialists, and/or size determination specialists. For example, in our survey, 87 percent of the CMRs responding reported that in fiscal year 2000, they served in one or even two additional roles, as figure 2 shows.PCR was the most frequent additional role; 59 percent of the part-time CMRs also served as PCRs. The CMRs and SBA officials we interviewed also told us that part-time CMRs often give lower priority to their CMR work than to their PCR, COC, and/or size determination work. This is consistent with our survey—57 percent of the part-time CMR respondents reported spending 60 to 89 percent of their time in fiscal year 2000 performing non-CMR duties. Only 31 percent reported spending less than 30 percent of their time on non- CMR duties. Apparently, there are several reasons for this, including not only the workload demands of these other roles, but also the time frames within which they must be performed. Generally, these roles are tied to specific procurements, and staff performing them must meet the procurement schedule. Regulations and guidance also mandate tight time frames for certain tasks. In contrast, CMR work is generally not as time sensitive, and several CMRs who perform these other roles told us they “fit in CMR work” when they have time. CMRs generally seem to spend slightly more time on marketing than on compliance monitoring. Figure 3 and table 1 show details of how CMRs spent their time in fiscal year 2000. However, the SBA officials and CMRs we interviewed told us that the amount of time that both full- and part- time CMRs spend on the two primary CMR duties—compliance monitoring and marketing—varies considerably between Area Offices and between individual CMRs. CMRs use desk reviews far more frequently than on-site reviews to monitor prime contractors’ compliance with subcontracting plans. In fiscal year 2001, 70 percent of all compliance reviews were desk reviews. Reliance on desk reviews has increased substantially since fiscal year 1992, when all compliance reviews were done on-site. Figure 4 and table 2 show more details. There is an uneven distribution of CMRs nationally and an uneven distribution of CMR workload. According to SBA, at the end of fiscal year 2001, there were about 18 CMR FTEs nationally to monitor the subcontracting activities of the 2,029 prime contractors under SBA’s cognizance. Figure 5 and table 3 illustrate the wide range in CMR workload. The SBA officials and CMRs we interviewed also acknowledged wide variations in workload between individual CMRs, whether full-time or part-time, in different Area Offices or within the same Area Office. For example, in one Area Office, one CMR had 13 prime contractors to monitor, while another had 65. Both spend about 10 percent of their time as CMRs. Similarly, in another Area Office, one CMR had 120 prime contractors to monitor, while a CMR in another had only 2. Both spend about 20 percent of their time as CMRs. Not surprisingly, the uneven distribution of CMRs and CMR FTEs has contributed to wide variations in the number of on-site reviews performed annually by Area Offices. For example, one Area Office conducted 15 on- site reviews in fiscal year 2001, while others conducted 40 or more. In addition, uneven workload distribution has contributed to the existence of large blocks of un-reviewed, that is, “uncovered,” prime contractors. For example, in one Area, there are no SBA reviews of any kind—on-site or desk—being done for 212 prime contractors, including all the prime contractors in one state and half of those in another. SBA officials told us that in fiscal year 2001, only 11.4 percent of the prime contractors nationwide received an on-site SBA review. CMRs and SBA officials have a range of concerns about the focus of the CMR role—particularly with regard to how CMRs should spend their time. Some believe that CMRs should concentrate more on monitoring prime contractors’ compliance with their subcontracting plans. Others believe they should focus more on helping small businesses connect with prime contractors for subcontracting opportunities. There are also differences of opinion and concerns within the agency about the relative merits of the methods that CMRs use to monitor compliance as well as with the uneven distribution of the CMR workload. There are disagreements within SBA about the most appropriate focus for the CMR role, that is, the balance between compliance monitoring and marketing. The Area Directors, Supervisors, and CMRs we interviewed had different views about CMR work priorities. For example, one Area Director stated that compliance monitoring should be the first priority for CMRs and matchmaking (i.e., marketing) should be secondary. Another Area Director said that the CMR role should be a combination of compliance monitoring and matchmaking but that matchmaking should get more emphasis. In contrast, the May 2001 report of the SBA Subcontracting Task Team clearly identifies compliance monitoring as the CMRs’ main function. The report concludes that over the past several years, CMRs have spent a disproportionate amount of their time on special initiatives. The report recommends that compliance reviews become the CMRs’ primary focus and that less emphasis be placed on ancillary duties, such as matchmaking or special initiatives. The report even proposes a new job title for the CMR, adding that “marketing” gives the wrong impression that matchmaking is the CMR’s main role. In contrast again, SBA’s Web site does not mention compliance monitoring as part of the CMR role. Rather, the Web site, which is an important means of communicating with both prime contractors and small businesses, describes the CMR role as “assisting small businesses in obtaining subcontracts by marketing small businesses and matching them with large prime contractors.” Concerns about the focus of the CMR role are not new. SBA’s Office of the Inspector General (OIG) issued a report in October 1995 citing the need to focus CMR efforts more effectively with respect to compliance monitoring and matchmaking. SBA introduced desk reviews in fiscal year 1996 as an economy measure. Because desk reviews take place in the CMRs’ offices rather than at the prime contractors’ locations, CMRs do not have to travel. In addition, the desk review takes substantially less time to accomplish than does the on- site review. Consequently, relying on desk reviews saves both travel money and CMR time. SBA staff opinions differ about the relative merits of on-site versus desk reviews. Some SBA officials and many CMRs believe that on-site reviews are more effective and question the relative value of desk reviews as a means of monitoring compliance. During on-site reviews, CMRs determine if prime contractors are complying with their subcontracting plans by going to the prime contractor’s location, reviewing files and documentation that support reported summaries of subcontracting activities, and interviewing officials. The CMRs we spoke with said that this type of in-depth review is one of their best tools to encourage the maximum use of small businesses as subcontractors. In contrast, other SBA officials believe that using desk reviews not only saves resources but also increases the total amount of monitoring that CMRs can do. SBA officials said they have never evaluated the effectiveness of the two types of reviews. SBA’s OIG observed the uneven workload distribution problem in October 1995. Noting the limited subcontracting resources and variable staffing levels and workloads, the OIG concluded that the unequal distribution of workload among CMRs was correlated with the uneven coverage of prime contractors. Accordingly, the OIG recommended that SBA more evenly distribute prime contractors among CMRs to improve coverage. However, SBA officials told us that where CMRs are located is driven by factors other than the location of prime contractors with subcontracting plans. For example, they said that although they can request CMRs to relocate to provide better coverage, they cannot require staff to relocate. It is likely that the full extent of the uneven workload and coverage problems has not yet been identified. SBA officials have estimated that as many as 1,500 additional prime contractors with subcontracting plans are not presently captured by SBA’s data systems. While SBA officials told us they are working to improve the methods and databases for identifying and tracking prime contractors, progress appears to be very slow. Again in October 1995, the SBA OIG noted that the number of federal prime contractors with subcontracting plans was unknown, complicating SBA’s efforts to focus subcontracting program activities and likely resulting in lost small business opportunities. Two primary factors have affected the CMR role: declines in resources and an ad hoc, piecemeal response to resource challenges. Both staffing and travel funds declined substantially over the past several years. With downsizing and retirements taking place and no staff assigned to replace lost personnel, the number of CMR FTEs declined significantly and resulted in workload imbalances across and even within SBA’s Area Offices. Travel fund reductions have meant fewer on-site visits and greater reliance on desk reviews. Staff at all levels agreed that insufficient resources—particularly staffing and travel funds—are the biggest obstacles to greater CMR coverage of prime contractors. SBA did not formulate a strategic plan for dealing with the impact of resource reductions on the CMR role. Without such a plan, SBA implemented ad hoc measures to deal piecemeal with resource declines. Nonetheless, these measures collectively have redefined the CMR role. SBA/OGC resources have declined substantially over the past few years. From fiscal year 1991 through fiscal 2000, staff declined 52 percent, from 333 to 159. During this same period, travel funds declined 54 percent, from $440,000 to $201,000. SBA officials agreed that these resource declines have significantly affected the CMR role. For example, SBA Area Directors told us that the need to assign multiple roles to field staff has increased since SBA experienced significant budget and staff cuts in the mid-1990s. Prior to these cuts, most field staff for both the prime and subcontracting programs were hired as specialists for one program. Now most, including CMRs, work on multiple programs. Similarly, the Subcontracting Task Team report pointed out that the merging of the CMR/PCR positions was workload driven and reflected the fact that positions were not being filled when incumbents retired. Finally, several SBA officials told us that to cope with the loss of staff and lack of travel funds, as well as the increased number of special initiatives, the assignment of multiple roles to staff became necessary and the desk review was created to replace the on-site review, when necessary. The decline in travel funds has especially affected on-site compliance monitoring. The lack of travel funds has become a critical factor in decisions about which prime contractors receive on-site reviews, particularly in those Area Offices that cover large geographic areas. For example, in one Area, one CMR said that one major factor determining which contractors receive on-site visits was whether they could be reached on a tank of gas, since travel funds are so limited. In another Area, the Director said that travel funds have been a problem for a long time, and that for many months of the fiscal year, CMRs do not have access to any travel money. This prevents them from being able to effectively select prime contractors for on-site review. Our survey indicated that the lack of travel funds also affected on-site reviews in fiscal year 2000. For example, 76 percent of the responding CMRs rated their Area Office’s travel budget as very important in determining which prime contractors to review on-site. In comparison, fewer—66 percent—rated problems with contractor reports as very important. Still fewer—58 percent and 50 percent, respectively—rated poor/marginal ratings of the contractor during the last on-site review and the fact that the contractor had never had an on-site review as very important. Similarly, 78 percent rated the lack of travel funds as a significant barrier to conducting on-site prime contractor reviews. Finally, the Task Team’s report expressed concern that without the requisite travel money and other resources, SBA will continue to monitor (on-site) about 12 percent of the total prime contractor portfolio, thus limiting the effectiveness of the Subcontracting Assistance Program. SBA’s response to these resource challenges has been largely ad hoc and reactive. As resources have declined, the agency has struggled to keep up with subcontracting program demands by making various piecemeal adjustments to address specific problems, such as assigning multiple roles to CMRs and instituting desk reviews to cover staffing shortages and travel fund declines. SBA has not stepped back—at either the agency or program level—and taken a broader, more strategic look at the CMR role, particularly in the context of today’s resource-constrained environment. Strategic planning, assessment, and evaluation are essential elements of good management. This has been recognized for federal agencies since 1993, when the Government Performance and Results Act became law. Effective management requires the establishment of goals and objectives as well as impact or outcome performance measures. SBA has not extended such planning to the CMR role to help address the difficult challenges that declining resources pose for the Subcontracting Assistance Program. SBA’s agency-level plan does not address the CMR role at all. Furthermore, SBA officials told us that SBA’s current strategic- planning process does not deal with the CMR role in subcontracting or provide for assessments, evaluations, or planning for the CMR role. In addition, they told us that neither OGC nor the Subcontracting Assistance Program has conducted strategic assessments or planning for the CMR role. In sum, while regulations and operating procedures describe a variety of duties and responsibilities that CMRs may perform, SBA has not developed goals and objectives for the CMR role. For example, the agency analyzed what might be desirable levels of prime contractor coverage or subcontracting plan compliance in today’s environment and how CMRs might contribute to achieving these goals. In addition, although SBA has productivity measures (e.g., the number of on-site and desk reviews conducted) to track CMRs’ performance, it does not have impact or outcome measures—or even such expectations—for CMRs. Finally, SBA has not strategically assessed, evaluated, or planned how best to address critical CMR role issues that have emerged during the past few years, such as the effect of multiple role assignments for CMRs, disagreements about the focus of the CMR role, the relative merits of on-site and desk reviews, and the impact of uneven distributions of CMRs and CMR workloads. Consequently, we do not know how effective the CMR role is. Effectiveness cannot be assessed without outcome and impact measures tied to program goals and objectives. These do not exist for the CMR role. Subcontracting on federal contracts is a large and growing marketplace for small businesses. CMRs have been long considered to be key to fostering small business participation in such subcontracting. However, the value of the CMR role and the effect of recent changes in it are unknown. Unless steps are taken to better assess, evaluate, and plan for the future of the CMR role, SBA will continue to lack an understanding of CMR contributions to small business subcontracting. Moreover, its approach to addressing challenges that CMRs face will continue to be ad hoc and piecemeal. We recommend that the Administrator of SBA assess, evaluate, and plan the CMR role, including addressing such issues as the impact of assigning multiple roles to CMRs, the appropriate CMR role focus, the effectiveness of compliance- monitoring methods, and the impact of uneven CMR workloads and prime contractor coverage; develop specific outcome and impact measures for CMRs’ clearly communicate the strategic plan and expectations for the CMR role to both SBA staff and small businesses. SBA provided us with written comments on a draft of this report. The comments, along with our responses, appear in appendix I. SBA neither concurred nor disagreed with our recommendations. However, SBA said that our report will be extremely helpful as it seeks ways to strengthen and improve its Subcontracting Assistance Program. In addition, SBA agreed that it needs to rethink the CMR role in today’s environment and noted that some aspects of the role may change in the future. SBA also agreed that it needs to develop outcome or impact measures to better assess the effectiveness of the CMR role. Furthermore, SBA noted that it has changed its Web site to include compliance monitoring as part of the CMR role. SBA objected to some of our conclusions, particularly (1) that SBA’s response to the challenges posed by declining resources has been ad hoc and piecemeal; (2) that SBA lacks a clear, strategic vision of the CMR role; and (3) that CMRs’ effectiveness is unknown. We continue to believe that our conclusions are correct. They are based on evidence that SBA has not strategically assessed, evaluated, or planned the CMR role in light of the current environment, including developing goals and objectives for the role and impact and outcome measures for CMRs. SBA’s comments do not provide any new evidence to the contrary. For example, the task force study that SBA cites as evidence of planning identified some specific problems with the CMR role and recommended some specific solutions. However, it did not strategically assess or plan the CMR role, nor was it tied to any SBA strategic-planning effort. In addition, the study did not use or establish measurable goals and objectives or outcome and impact measures for the CMR role. Similarly, the regulations and operating procedures that SBA references as evidence of vision are not tied to measurable goals and objectives and associated outcome and impact measures. Rather, they simply describe a variety of duties and responsibilities that CMRs are required to perform. In addition, the conditions that SBA cites as evidence of CMRs’ effectiveness may be the result of many factors other than CMRs’ efforts alone. SBA has not conducted the assessments and evaluations necessary to determine what effect CMRs actually had on these conditions. Finally, SBA’s agreement that it needs to rethink the CMR role and develop impact or outcome measures further supports our conclusions. To determine CMRs’ duties and responsibilities, we analyzed pertinent legislation, regulations, and operating procedures and reviewed other agency documentation, including staffing profiles and workload analyses. We extracted information from the General Services Administration’s Federal Procurement Data System database on prime and subcontractors. We also interviewed officials at SBA headquarters and Area Directors and CMRs located in all six SBA Area Offices. We interviewed all six Area Directors and several Area Supervisors. We also interviewed 15 of the 39 current CMR staff, accounting for about 51 percent of the total CMR FTEs. Finally, we analyzed data from a GAO survey of all 33 staff working as CMRs in fiscal year 2000, who constitute 85 percent of the current CMR population. Our overall response rate was 97 percent. On our survey, we asked a variety of questions about CMRs’ roles and responsibilities. To identify the factors affecting the CMR role, we reviewed pertinent legislation, SBA staffing information, workload analyses, and travel budget fund submissions. We also interviewed SBA headquarters and Area Office personnel and reviewed agency audit reports and task force studies. We also met with SBA officials responsible for strategic planning. We conducted our work from November 2001 through July 2002 in accordance with generally accepted government auditing standards. We are sending copies of this report to interested congressional committees; the Administrator, SBA; and the Director, Office of Management and Budget. We will also make copies available to others upon request. In addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you have any questions concerning this letter, please call me at (617) 565-7555. Key contributors to this review were Catherine Baltzell, Art Fine, David Bennett, Christina Chaplain, and Sylvia Schatz. The following are GAO’s comments on the Small Business Administration’s letter dated October 8, 2002. 1. We do not agree that 80-percent time is essentially equivalent to 100- percent time. Four staff that work 80 percent of their time as Commercial Marketing Representatives (CMRs) constitute about three full-time equivalents (FTEs), not four FTEs. This is a difference of about 20 percent. As we point out in our report, SBA’s use of part-time CMRs for whom the CMR role is, as SBA comments, “a collateral duty,” has not meant an increase in CMR staff resources. Rather, total CMR FTEs declined 28 percent from fiscal year 1992 through fiscal 2001. 2. SBA’s task force study identified some specific problems with the CMR role and recommended some specific solutions, and SBA’s efforts to implement some of these solutions can possibly lead to some incremental improvements. However, the task force study did not strategically assess or plan the CMR role, nor was it tied to any SBA strategic-planning effort. In addition, the study did not use or establish measurable goals and objectives or outcome and impact measures for the CMR role. In other words, the study did not address the larger issues of CMR strategic role assessment and planning. In the absence of such planning, SBA’s approach to addressing challenges that CMRs face will continue to be ad hoc and piecemeal. 3. Delegating decisions about the focus of the CMR role to Area Directors may be appropriate. However, SBA has not conducted any assessments or evaluations that address what Area Office factors and characteristics make such delegation essential or effective. Furthermore, SBA has not evaluated the effectiveness of local decisions regarding the balance between compliance monitoring and marketing. The resource pressures on the CMR role heighten our concern. It may be that local decisions about the focus of the CMR role are more influenced by the realities of local resource constraints than by local compliance monitoring and marketing needs. Without sound assessment and evaluation, however, this and many other issues of role focus will remain clouded. 4. The regulations and operating procedures that SBA references do not constitute a strategic vision tied to measurable goals and objectives. Rather, as we discuss in our report, they simply describe a variety of duties and responsibilities that CMRs are required to perform. Since the regulations and guidance do not place a relative order of importance on these duties, and the Small Business Act and the Federal Acquisition Regulations do not mention CMRs, there is no law, regulation, or guidance requiring or suggesting that CMRs prioritize their work according to a strategic vision. In contrast, a strategic vision is broader and more comprehensive. It is expressed in an overall strategic plan that articulates a mission and specific plans to fulfill that mission, including measurable goals and objectives and associated outcome and impact measures. SBA also appears to recognize at least to some degree the limitations of the regulations and operating procedures as a strategic vision because it agrees that it needs to rethink the CMR role and develop outcome or impact measures. 5. SBA recognizes that on-site reviews are both more effective than and preferable to desk reviews but asserts that they are “less cost- effective” and “not always necessary.” However, SBA does not offer persuasive evidence to support these assertions. During our review, SBA officials told us that they had not conducted assessments or evaluations of either of these compliance review methods or of their comparative effectiveness. Furthermore, SBA has not identified the criteria—that is, the strategic goals and objectives and associated impact or outcome measures—necessary to guide such assessments and evaluations. SBA also says that it uses the results of desk reviews to determine which prime contractors should receive on-site reviews. However, as we discuss in our report, the lack of travel funds is the primary driver of decisions about which prime contractors receive on-site reviews. 6. SBA observes that in fiscal year 2000, small businesses, in general; small disadvantaged businesses; and women-owned small businesses received “much higher” percentages of all subcontract awards than of prime contract awards. SBA also observes that the percentage of subcontract dollars awarded to small disadvantaged businesses has improved steadily over the past two decades. SBA then attributes both of these conditions solely to CMR efforts. SBA has not conducted the assessments and evaluations necessary to support this conclusion. Moreover, while CMRs’ efforts may well have contributed to these conditions, a number of other factors likely have had a significant impact. As we note in our report, federal regulations require a subcontracting plan for each contract or contract modification that exceeds $500,000 ($1 million for construction contracts) and has subcontracting possibilities. Federal agencies are independently responsible for complying with these regulations. Agencies may even conduct their own subcontract-monitoring efforts. For example, DOD reviews subcontracting plans for defense prime contractors. In fact, SBA commented that it has delegated the primary responsibility for monitoring the subcontracting plans of most DOD contractors to the responsible DOD agency so that SBA can focus its limited resources on monitoring civilian agencies’ plans. Since DOD accounted for 65 percent of all subcontracted dollars awarded in 2001, it is not likely that CMRs alone are responsible for all small business subcontracting achievements. There are also inherent business incentives for prime contractors to subcontract voluntarily with small businesses. For example, small business subcontractors can provide important specialized capabilities that the prime contractor does not have or wish to invest in developing. Small business subcontractors may also be able to provide some general services faster and more economically, thus saving prime contractor resources. Subcontracting also allows prime contractors to avoid permanent staffing increases that may not be sustainable in the face of market shifts. Finally, as we note in our report, the nature of federal contracting has changed. The number of prime contracts is shrinking, and many prime contracts have become so large that small businesses find it difficult to compete for them. This change alone may heavily influence the conditions that SBA cites as markers of CMRs’ effectiveness. 7. We do not agree that our report leaves a mistaken impression with regard to the goals and objectives assigned to CMRs. Rather, we state clearly that SBA has productivity measures (such as the number of on- site and desk reviews conducted) to track CMRs’ performance but that it does not have impact or outcome measures—or even such expectations—for CMRs. SBA agrees that it needs to develop outcome or impact measures to better measure CMR effectiveness. 8. We continue to believe that SBA’s development of its new database appears to be slow. SBA began developing this new database in early 2001, after experiencing repeated problems with its original database. In October 2001, the SBA officials we interviewed told us that the new database would be operational in January 2002. In February 2002, SBA told us that it would be operational by June. Now, in October 2002, SBA says that it is only in test mode and acknowledges that it will not have a complete list of prime contractors with subcontracting plans until the end of the second quarter of fiscal year 2003. 9. While SBA’s original database does provide some useful information, the SBA officials we interviewed told us that the data were not complete. The database did not contain a complete list of prime contractors with subcontracting plans. 10. We continue to have concerns about CMRs’ workload and prime contractor coverage. As we discuss in our report, about 18 CMR FTEs nationally monitor the subcontracting activities of the 2,029 prime contractors currently identified. In addition, CMRs have various marketing duties. (In fiscal year 2000, CMRs generally seemed to spend slightly more time on marketing than on compliance monitoring.) CMR compliance monitoring Area Office workloads currently range from 56 to 198 prime contractors per CMR FTE. There are wide variations in CMR workloads both between Area Offices and within more than one Area Office. This situation may be exacerbated when additional contractors with subcontracts are identified and added to CMR workloads.

Subcontracting on federal contracts is a large and growing marketplace for small businesses. The Small Business Administration's (SBA) Commercial Marketing Representatives (CMRs) have long been considered to be key to fostering small businesses' participation in subcontracts. GAO was asked to assess the role that CMRs are playing in administering SBA's subcontracting assistance program. CMRs are supposed to promote small business subcontracting in two primary ways. First, they review prime contractors' compliance with the requirements of their subcontracting plans--either through on-site visits to contractors or by simply reviewing their subcontractor activity reports. Second, they conduct various marketing activities, such as marketing small businesses to prime contractors. In recent years, however, additional duties placed on CMRs have often taken priority over these responsibilities. In fact, in fiscal year 2000, 87 percent of the CMRs had other substantial responsibilities. Moreover, workloads and prime contractor coverage now vary greatly between CMRs. Additionally, CMRs are relying more on "desk" reviews of subcontractors' activity to monitor compliance with subcontracting plans as opposed to on-site reviews. This is a concern to some SBA officials who believe that on-site reviews are more thorough, though others believe the desk review offers the potential for greater coverage. Declines in staffing and travel funds have contributed to the changing role of the CMR. With downsizing and retirements taking place and no staff assigned to replace lost personnel, the number of CMR full-time equivalents (FTEs) has decreased significantly and has resulted in workload imbalances. While there are concerns about the changing nature of the CMR role, SBA has not strategically planned for these changes or assessed their collective impact. Instead, it has implemented ad hoc measures to deal piecemeal with resource declines. Unless steps are taken to better evaluate and plan for the future of CMRs, SBA will continue to lack an understanding of their contributions to small business subcontracting.

Because of potential cost increases, the Air Force established a team—the Joint Estimating Team—to review the total estimated cost of the F-22 program. This team reported in 1997 that the cost of the F-22 production program could grow substantially from the amount planned, but that the contractors should design cost reduction plans to fully offset that cost growth. The Office of the Under Secretary of Defense for Acquisition, Technology and Logistics generally adopted the team's recommendations to change certain aspects of the program as well as a plan to define and implement cost reduction plans. The contractors have continued since 1997 to refine and increase the number and dollar amounts associated with the plans to reduce F-22 production costs. Ultimately, the savings to be achieved by production cost reduction plans must be reflected in lower production contract prices, and lower expenditures by the Air Force than would have been the case if the plans had not been implemented. The Air Force and contractors have entered into memoranda of understanding that relate the affordability of F-22 production to contract prices that will be negotiated for low-rate initial production. The memoranda established target price objectives against which the negotiated prices will be evaluated and financial incentives to achieve the target price objectives for the applicable production lots. To encourage a reduction in production costs, the Air Force and contractors agreed that production cost reduction plans would be proposed, approved and implemented as appropriate. The Air Force agreed to reimburse certain investment costs and to pay award fees to the contractors based on negotiating contracts for certain prices. Until contract prices are negotiated, cost estimates will continue to reflect judgements of estimators about the potential impact of cost reduction plans when implemented. In an effort to offset production cost increases, F-22 contractors have been developing production cost reduction plans to enhance production technology, improve manufacturing techniques, and improve acquisition strategies and subcontract agreements for buying materials. These cost reduction plans are categorized as: implemented, not yet implemented, or "challenge". The Air Force and contractors' criteria for determining if a cost reduction plan is implemented include whether the contractor has submitted a firm-fixed price proposal that recognizes the impact of the cost reduction, the impact of the reduction has been reflected in a current contract price—either with the prime contractor or a supplier to the prime contractor, the contractor has reduced the standard number of hours allocated to a specific task, the reduction has been negotiated in a forward pricing rate agreement, or the reduction has been negotiated with a subcontractor or vendor. Plans are categorized as not yet implemented if none of the criteria are met. Challenge plans represent additional potential savings in areas that have been identified, but that are not yet well defined. The contractors' estimated reductions in costs that are associated with F-22 cost reduction plans increased from $13.1 billion in January 1997, to $21 billion in mid-2000, to $26.5 billion in January 2001. According to the contractors, the $26.5 billion is distributed as follows by category: About $13.7 billion (52 percent) in cost reductions that have been About $8.5 billion (32 percent) in cost reductions not yet implemented; About $4.2 billion (16 percent) labeled as a challenge amount. Figure 1 shows examples of implemented, not yet implemented, and challenge cost reduction plans and the relative progression of these plans toward potentially achieving some cost reductions in the future. In late 2000, the Air Force cost estimators projected, in an estimate supporting the fiscal year 2002 budget request, that production costs of 333 F-22s were likely to exceed the $37.6 billion congressional cost limitation by $2 billion. The cost estimate produced by the Office of the Secretary indicates that costs will likely exceed the congressional cost limitation by $9 billion. Important reasons for the differences between Air Force and Office of the Secretary estimators are differing judgements about labor efficiencies, engine costs, and the viability of cost reduction plans and their potential impact on the cost of F-22 production. In late 1999, both the Air Force and Office of the Secretary cost estimators projected that production costs for 339 aircraft would exceed the congressional cost limitation of $39.8 billion in effect at that time. The Air Force cost estimators projected production costs at $40.8 billion, and the Office of the Secretary estimated $48.6 billion for the 339 production aircraft. Even though the cost estimates exceeded the $39.8 billion cost limitation in effect at that time, the Secretary of the Air Force maintained that the cost would not exceed the limitation, and established the Air Force's position on F-22 production cost at $39.8 billion. In estimates made in December 2000 to support the fiscal year 2002 budget request, both Air Force and Office of the Secretary cost estimators continue to project that F-22 production costs will exceed the congressional cost limitation. Table 1 shows the details of these estimates and the amounts by which the estimates exceed the congressional cost limitation. The current cost limitation of $37.6 billion has been adjusted to reflect planned acquisition of 333 production aircraft, 6 fewer than included in the cost limitation in effect in 1999. This change reflects congressional action on the fiscal year 2000 Air Force budget, in which the Congress approved funding for 6 aircraft using appropriations for Research, Development, Test and Evaluation. Accordingly, the 6 aircraft and associated costs of $1.575 billion (excluding about $200 million that had been appropriated in fiscal year 1999 for advanced procurement for those aircraft) were eliminated from the production cost limitation and added to the development cost limitation. If the Office of the Secretary's higher estimate is correct and additional cost reduction plans are not developed and implemented, we project that the Air Force would have to buy about 85 fewer F-22s (or about 25 percent) than the 333 aircraft now planned to stay within the cost limitation. In our August 2000 report, we had also calculated that the Air Force would not be able to procure about 85 F-22s if the Office of the Secretary's 1999 estimate was correct. The Air Force and the Office of the Secretary cost estimators included in their projections the effect of cost reduction plans that have been categorized as implemented. They also estimated the expected future impact of cost reduction plans that have not yet been implemented. Neither included challenge plans.. Air Force officials advised us that their cost estimates consider the same cost reduction plans as the Office of the Secretary estimators, but that differing judgements regarding the viability of the plans and potential amounts of cost reductions are applied. Table 2 compares the two cost estimates. Air Force and Office of the Secretary officials attributed the majority of the differences in the estimates to the Office of the Secretary having— Estimated higher labor costs than the Air Force relating to subcontractor Estimated higher costs for the F-22 engines, Excluded some cost reduction plans because of the limited viability, and Estimated more conservative savings from some cost reduction plans. Labor costs for subcontractors projected by estimators from the Office of the Secretary were $3.0 billion more than those projected by Air Force cost estimators. Projections of engine costs by the Office of the Secretary estimators were $1.2 billion higher. The Office of the Secretary also excluded some planned manufacturing cost reduction plans because they were not adequately detailed, and estimated $1 billion less would be saved by the planned manufacturing cost reductions than did the Air Force. Further, the Office of the Secretary estimated $800 million less in cost reductions than the Air Force for plans relating to productivity investments. Because F-22 production is in its early stages, few plans have resulted in actual cost reductions. However, analysis of plans categorized as implemented do show indications that lower costs can be achieved. The Air Force in mid-2000 asked DCAA to conduct a limited, independent review of some of these plans. In late 2000, the DCAA examined eleven cost reduction plans totaling $425 million of total estimated savings of $26.5 billion. These eleven plans were chosen so DCAA could examine cost reduction plans at different stages of development and at different locations including Lockheed Martin Aeronautical Systems, Marietta, Georgia; Lockheed Martin Tactical Aircraft Systems, Fort Worth, Texas; and Boeing Military Aircraft, Seattle, Washington. DCAA did not conduct detailed audits of these cost reduction plans. Their reviews focused primarily on methodologies used to calculate the reported savings or the verification of rates or material cost used in the calculations of contractor reported savings. DCAA did not take exception to the potential cost reductions for 8 of the 11 plans reviewed; found potential cost reductions on two others to be based on judgement, not discrete, measurable events; and found documentation on one to be lacking. Regarding the one plan where documentation was lacking, DCAA auditors were unable to validate contractor estimates totaling around $2 million that involved a new process developed to only require one step to drill a hole in the airframe, rather than two steps. Our August 2000 report recommended that the Secretary of Defense reconcile the number of F-22s that need to be procured with the cost limitation and report to the Congress on the implications of procuring fewer F-22s because of potentially higher costs. DOD partially agreed, stating that the affordability of the F-22 will be evaluated during an upcoming Quadrennial Defense Review. We also found that Air Force quarterly reports provided to the Under Secretary of Defense for Acquisition, Technology, and Logistics did not regularly highlight major changes associated with cost reduction plans. While the status of individual cost reduction plans are tracked by contractors and the F-22 Program Office, we believe regular reporting by the Air Force to the Under Secretary of Defense on the status of these plans is necessary to continuously assess their impact on the estimated cost of F-22 production. Achievement of the estimated cost reductions embodied in the plans is critical to completing F-22 production within the congressional cost limitation. Quarterly reporting of cost reduction plan information enhances its visibility. As a result, we recommended the Air Force report to the Under Secretary of Defense on the status of the cost reduction plans each quarter and that quarterly reports include summary information such as the total number of cost reduction plans identified, the number implemented, the total estimated cost reductions, cost reductions realized to date, and additions or deletions from the plans included in the prior report. DOD concurred with our recommendation in October 2000 and agreed to report cost reduction plan information in subsequent quarterly reviews to the Under Secretary of Defense for Acquisition, Technology and Logistics. DOD agreed that the achievement of cost reduction plans is essential to the execution of the F-22 program within the congressional cost limit. However, our recommendation has not been implemented. In the Air Force's March 2001 quarterly review to the Under Secretary, the information reported included only summary information on the total estimated cost reductions. In commenting on a draft of this report, DOD agreed there continues to be a notable difference between the Air Force and Office of the Secretary F-22 production cost estimates. They indicated that data would emerge toward verification of these estimates as the program begins to accumulate production cost data. DOD also agreed that the dollar amounts associated with the cost reduction plans have continued to increase since 1997. They indicated that as F-22 cost pressures have increased, so have the number of cost reduction plans and the cost reductions attributed to them. In commenting on the progress the Air Force has made toward complying with our recommendation from August 2000 for specific cost reduction plan information to be reported in quarterly reviews, DOD indicated the information reported in the last quarterly review (June 2001) contained more detailed information. We have examined the June 2001 quarterly review and agree it contains more information on cost reduction plans than previous quarterly reviews. Information on the total estimated cost reductions was reported. However, the information reported is still not consistent with what we recommended be reported in August 2000. Information was not reported as we recommended regarding the total number of cost reduction plans identified, the number implemented, the cost reductions realized to date, and any additions or deletions from the plans included in the prior report. To identify the amount of potential offsets attributable to production cost reduction plans by F-22 contractors we reviewed contractor cost reduction plans to determine the basis for the reductions expected to be achieved and whether the reduction was implemented or not yet implemented. We reviewed the documentation from the contractors and discussed the plans and the Air Force procedures for reporting on such plans with contractor and Air Force officials. To compare the latest F-22 production cost estimates of the Air Force and the Office of the Secretary with the congressional production cost limitation and to determine the extent to which cost reductions plans were considered in establishing these estimates, we reviewed the Joint Estimating Team's report and various Air Force briefings. We discussed the estimates with officials in the Office of the Secretary and the Air Force's F-22 Program Office to determine why they differed. We compared the two estimates, including the baseline estimate, the estimated reductions from cost reduction plans, and the net estimates. We obtained a description of the reasons for the variances between the two estimates. We also discussed the estimates and production cost limitation with Air Force and Office of the Secretary officials. The Office of the Secretary cost estimate shown in this report is recorded in briefing documents we obtained during the course of our review. The Office of the Secretary provided us neither its cost estimate nor documentation related to its cost estimate. Officials from the Office of the Secretary cited their policy of not allowing access to that information because they considered it predecisional. However, we corroborated the information contained in the briefing documents we analyzed. An Office of the Secretary official reviewed and agreed with the estimated and projected costs included in this report that are attributed to the Office of the Secretary. To calculate the number of F-22s that could not be procured within the cost limitation we allocated the dollars in the Office of the Secretary estimate to production lots 1 through 11 in the previous Air Force estimate. Starting with the adjusted costs for production lot 1, we determined how many aircraft could be purchased without exceeding the applicable cost limitation To evaluate whether the Office of the Secretary and the Air Force were complying with our prior recommendations, we determined whether a defense review, that potentially could reconcile the number of F-22s needed with the cost limitation, had been completed. We also reviewed recent quarterly briefings from the Air Force to the Under Secretary of Defense for Acquisition, Technology and Logistics to determine how the information included on production cost reduction plans compared to the information we recommended be included in the briefings. In performing our work, we obtained information or interviewed officials from the Office of the Secretary of Defense, Washington D.C.; the F-22 Program Office, Wright-Patterson Air Force Base, Ohio; the Defense Contract Management Agency, Marietta, Georgia; Lockheed Martin Aeronautical Systems, Marietta, Georgia; Lockheed Martin Tactical Aircraft Systems, Fort Worth, Texas; and Boeing Military Aircraft, Seattle, Washington. We performed our work from December 2000 through May 2001 in accordance with generally accepted government auditing standards. As arranged with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days after the date of this letter. At that time, we will send copies to appropriate congressional committees; the Secretary of Defense; the Secretary of the Air Force; and the Director, Office of Management and Budget. Copies will also be made available to others on request. Please contact me at (202) 512-4841 or Robert D. Murphy at (937) 258-7904 if you or your staff have any questions concerning this report. GAO staff acknowledgments to this report are listed in appendix II. Marvin E. Bonner, Christopher T. Brannon, Edward R. Browning, Arthur L. Cobb, Michael J. Hazard, Don M. Springman, and John Van Schaik made key contributions to this report. Tactical Aircraft: F-22 Development and Testing Delays Indicate Need for Limit on Low-Rate Production (GAO-01-310, Mar. 15, 2001). Supporting Congressional Oversight: Framework for Considering Budgetary Implications of Selected GAO Work (GAO-01-447, Mar. 9, 2001). Defense Acquisitions: Recent F-22 Production Cost Estimates Exceeded Congressional Limitation (GAO/NSIAD-00-178, Aug.15, 2000). Defense Acquisitions: Use of Cost Reduction Plans in Estimating F-22 Total Production Costs (GAO/T-NSIAD-00-200, June 15, 2000). Budget Issues: Budgetary Implications of Selected GAO Work for Fiscal Year 2001 (GAO/OCG-00-8, Mar. 31, 2000). F-22 Aircraft: Development Cost Goal Achievable If Major Problems Are Avoided (GAO/NSIAD-00-68, Mar. 14, 2000). Defense Acquisitions: Progress in Meeting F-22 Cost and Schedule Goals (GAO/T-NSIAD-00-58, Dec. 7, 1999). Fiscal Year 2000 Budget: DOD's Procurement and RDT&E Programs (GAO/NSIAD-99-233R, Sept. 23, 1999). Budget Issues: Budgetary Implications of Selected GAO Work for Fiscal Year 2000 (GAO/OCG-99-26, Apr. 16, 1999). Defense Acquisitions: Progress of the F-22 and F/A-18E/F Engineering and Manufacturing Development Programs (GAO/T-NSIAD-99-113, Mar. 17, 1999). F-22 Aircraft: Issues in Achieving Engineering and Manufacturing Development Goals (GAO/NSIAD-99-55, Mar. 15, 1999). F-22 Aircraft: Progress of the Engineering and Manufacturing Development Program (GAO/T-NSIAD-98-137, Mar. 25, 1998). F-22 Aircraft: Progress in Achieving Engineering and Manufacturing Development Goals (GAO/NSIAD-98-67, Mar. 10, 1998). Tactical Aircraft: Restructuring of the Air Force F-22 Fighter Program (GAO/NSIAD-97-156, June 4, 1997). Defense Aircraft Investments: Major Program Commitments Based on Optimistic Budget Projections (GAO/T-NSIAD-97-103, Mar. 5, 1997). F-22 Restructuring (GAO/NSIAD-97-100R, Feb. 28, 1997). Tactical Aircraft: Concurrency in Development and Production of F-22 Aircraft Should Be Reduced (GAO/NSIAD-95-59, Apr. 19, 1995). Tactical Aircraft: F-15 Replacement Issues (GAO/T-NSIAD-94-176, May 5, 1994). Tactical Aircraft: F-15 Replacement Is Premature as Currently Planned (GAO/NSIAD-94-118, Mar. 25, 1994).

The Air Force F-22 Raptor, an air superiority aircraft with an air-to-ground attack capability is set for completion in September 2003. However, contracts to begin 10 low-rate initial production aircraft for fiscal year 2001 have been delayed until after completion of the President's review of Department of Defense (DOD) programs. The Air Force plans to procure 333 production aircraft through 2013. The cost of F-22 production is limited by law, but the total number of aircraft to be procured is unspecified. This report (1) identifies the cost reduction plans by F-22 contractors, (2) compares the military's latest F-22 production cost estimates with the congressional cost limitation and determines the extent to which cost reduction plans were considered in establishing these estimates, and (3) provides the status of DOD's actions to implement GAO's earlier recommendations on production cost estimates and cost reduction plans for the F-22 program. GAO found that enhancing production technology, improving manufacturing techniques, and improving acquisition practices have contributed to cost reductions. Both the Air Force and the Office of the Secretary cost estimators projected that F-22 production costs would exceed the congressional cost limitation if the Air Force were to procure 333 F-22s. DOD and the Air Force have partially responded to the recommendations in GAO's August 2000 report on the F-22.

Data breaches involving PII can occur under many circumstances and for many reasons. They can be inadvertent, such as from the loss of an electronic device, or deliberate, such as from the theft of a device or a cyber-based attack by a malicious individual or group, foreign nation, terrorist, or other adversary. Incidents have been reported at a wide range of public- and private-sector institutions, including federal, state, and local government agencies; educational institutions; hospitals and other medical facilities; financial institutions; information resellers; retailers; and other types of businesses. The loss or unauthorized disclosure or alteration of the information residing on federal systems, which can include PII, can lead to serious consequences and substantial harm to individuals and the nation. Thus it is critical that federal agencies protect their systems and the information on them and respond to data breaches and cyber incidents when they occur. Over the last several years, federal agencies have reported an increasing number of information security incidents to the U.S. Computer Emergency Readiness Team (US-CERT). These include both cyber- and non-cyber- related incidents, and many of them involved PII. Figure 1 shows that the total number of security incidents reported annually more than doubled from fiscal year 2009 to fiscal year 2013. These incidents are categorized by type. Figure 2 shows the categories into which incidents reported in fiscal year 2013 fell. Moreover, a significant number of security incidents reported by agencies have involved PII. Figure 3 shows that the number of incidents involving PII for fiscal years 2009 through 2013 increased over 140 percent. Data breaches at federal agencies have received considerable publicity and raised concerns about the protection of PII at those agencies. Most notably, in May 2006, the Department of Veterans Affairs (VA) reported that computer equipment containing PII on about 26.5 million veterans and active duty members of the military was stolen from the home of a VA employee. More recent examples of incidents that compromised individuals’ personal information further highlight the impact that such incidents can have: In July 2013, hackers stole a variety of PII on more than 104,000 individuals from a Department of Energy system. Types of data stolen included Social Security numbers, birth dates and locations, bank account numbers and security questions and answers. According to the department’s Inspector General, the combined costs of assisting affected individuals and lost productivity—due to federal employees being granted administrative leave to correct issues stemming from the breach—could be more than $3.7 million. In May 2012, the Federal Retirement Thrift Investment Board (FRTIB) reported a sophisticated cyber attack on the computer of a contractor that provided services to the Thrift Savings Plan. As a result of the attack, PII associated with approximately 123,000 plan participants was accessed. According to FRTIB, the information included 43,587 individuals’ names, addresses, and Social Security numbers, and 79,614 individuals’ Social Security numbers and other PII-related information. In March 2012, a laptop computer containing sensitive PII was stolen from a National Aeronautics and Space Administration employee at the Kennedy Space Center. As a result, 2,300 employees’ names, Social Security numbers, dates of birth, and other personal information were exposed. In February 2009, the Federal Aviation Administration notified employees that an agency computer had been illegally accessed and that employee PII had been stolen electronically. Two of the 48 files on the breached computer server contained personal information about more than 45,000 agency employees and retirees. Title III of the E-Government Act of 2002, known as the Federal Information Security Management Act (FISMA), establishes a framework designed to ensure the effectiveness of security controls over information resources that support federal operations and assets. According to FISMA, each agency is responsible for, among other things, providing information security protections commensurate with the risk and magnitude resulting from unauthorized access, use, disclosure, disruption, modification, or destruction of information collected or maintained by or on behalf of the agency and information systems used or operated by an agency or by a contractor or other organization on behalf of an agency. These protections are to provide federal information and systems with integrity—preventing improper modification or destruction of information; confidentiality—preserving authorized restrictions on access and disclosure; and availability—ensuring timely and reliable access to and use of information. Under FISMA, agencies are required to develop procedures for detecting, reporting, and responding to security incidents, consistent with federal standards and guidelines, including mitigating risks associated with such incidents before substantial damage is done. The law also requires the operation of a central federal information security incident center that compiles and analyzes information about incidents that threaten information security. The Department of Homeland Security (DHS) was given the role of operating this center, which became US-CERT, by the Homeland Security Act. DHS’s role is further defined by Office of Management and Budget (OMB) guidance, which requires that incidents involving PII be reported to US-CERT within 1 hour of discovery. US- CERT is also responsible for providing timely technical assistance to operators of agency information systems regarding security incidents, including offering guidance on detecting and handling incidents. In addition to establishing responsibilities for agencies, FISMA assigns specific responsibilities to OMB, the National Institute of Standards and Technology (NIST) and inspectors general: OMB is to develop and oversee the implementation of policies, principles, standards, and guidelines on information security in federal agencies (except with regard to national security systems). It is also responsible for reviewing, at least annually, and approving or disapproving agency information security programs. NIST’s responsibilities include developing security standards and guidelines for agencies that include standards for categorizing information and information systems according to ranges of risk levels, minimum security requirements for information and information systems in risk categories, guidelines for detection and handling of information security incidents, and guidelines for identifying an information system as a national security system. Agency inspectors general are required to annually evaluate the information security program and practices of their agency. The results of these evaluations are to be submitted to OMB, and OMB is to summarize the results in its reporting to Congress. In July 2010, the Director of OMB and the White House Cybersecurity Coordinator issued a joint memorandum stating that DHS was to exercise primary responsibility within the executive branch for the operational aspects of cybersecurity for federal information systems that fall within the scope of FISMA. In September 2013 we issued the most recent of our periodic reports on federal agencies’ compliance with the requirements of FISMA. Specifically, we reported that, for fiscal year 2012, 24 major federal departments and agencies covered by the Chief Financial Officers Act had established many of the components of an agency-wide information security program, as required by FISMA, but had only partially established others. In particular, with regard to the eight components of an agency-wide security program, 18 agencies had fully implemented a program for managing information security risk, and 6 had partially implemented such a program; 10 agencies had fully documented security policies and procedures, while 12 had partially documented them; 18 agencies had selected security controls for their systems, but 6 had only partially implemented this practice; 22 agencies had established a security training program, and 2 had partially established such a program; 13 agencies were monitoring security controls on an ongoing basis, but 10 agencies had not fully implemented a continuous monitoring program; 19 agencies had established a program for remediating weaknesses in their security policies, practices, and procedures, while 5 had not fully implemented elements of a remediation program; 20 agencies had established an incident response and reporting program, but 3 agencies had not fully established such a program;and 18 agencies had fully established a program for ensuring continuity of operations in the event of a disruption or disaster, but 5 agencies partially implemented a continuity of operations program. The extent to which the agencies had implemented security program components showed mixed progress from fiscal year 2011 to fiscal year 2012. For example, according to inspectors general reports, the number of agencies that had analyzed, validated, and documented security incidents increased from 16 to 19, while the number able to track identified weaknesses had declined from 20 to 15. In addition, although most agencies had implemented elements of their security programs, we and inspectors general continued to identify weaknesses in elements of their programs, such as the implementation of specific security controls. Specifically, most major federal agencies had weaknesses in major categories of information security controls, as defined by our Federal Information System Controls Audit Manual. Table 1 shows, for fiscal year 2012, the number of the 24 major federal agencies that had weaknesses in the five major control categories. Illustrating the extent to which weaknesses continue to affect the 24 major federal agencies, in fiscal year 2013, inspectors general at 21 of the 24 agencies cited information security as a major management challenge for their agency, and 18 agencies reported that information security control deficiencies were either a material weakness or significant deficiency in internal controls over financial reporting in fiscal year 2013. These weaknesses show that information security continues to be a major challenge for federal agencies, putting federal systems and the information they contain, including PII, at increased risk. We and agency inspectors general have continued to make numerous recommendations to agencies aimed at improving their information security posture. Fully implementing these recommendations will strengthen agencies’ ability to ensure that their information, including PII, is adequately protected. Even when information security programs have been implemented effectively, data breaches can occur. Accordingly, OMB and NIST have specified key practices for responding to PII data breaches.include management practices such as establishing a data breach response team and training employees on roles and responsibilities for breach response, and operational practices, such as preparing reports on These suspected data breaches and submitting them to appropriate internal and external entities, assessing the likely risk of harm and level of impact of a suspected breach, offering assistance to affected individuals (if appropriate), and analyzing the agency’s breach response and identifying lessons learned. Table 2 provides more details on these key management and operational practices. In December 2013, we reported on our review of issues related to PII data breaches. The eight agencies in our review had generally developed, but inconsistently implemented, policies and procedures for responding to a data breach involving PII that addressed key practices. Specifically, with few exceptions, the agencies reviewed addressed the key management and operational practices in their policies and procedures. However, they did not consistently implement the operational practices, as summarized in figure 4. Of the seven agencies we reviewed, only the Internal Revenue Service (IRS) consistently documented both an assigned risk level and how that level was determined for PII-related data breach incidents; only the Army and IRS documented the number of affected individuals for each incident; and only the Army and the Securities and Exchange Commission notified affected individuals for all high- risk breaches. The seven agencies did not consistently offer credit monitoring to individuals affected by PII-related breaches. None of the seven agencies consistently documented lessons learned from PII breaches, including corrective actions to prevent similar incidents in the future or whether better security controls could help detect, analyze, and mitigate future incidents. Incomplete guidance from OMB contributed to this inconsistent implementation. For example, OMB’s guidance does not make clear how agencies should use risk levels in making a determination about notification to affected individuals. Further, OMB guidance states that the risk levels should help determine when and how notification should be provided, but it does not set specific requirements for notification based on agency risk determinations. In addition, OMB guidance for reporting on data breaches involving PII may be too stringent. Specifically, OMB guidance requires that DHS collect information about PII-related breaches within 1 hour, but officials at US-CERT and the agencies in our review generally agreed that this requirement was difficult to meet and may not provide US-CERT with the best information. For example, some agencies noted that it is difficult to provide a meaningful report on a breach within 1 hour since relevant information—such as how much PII was affected or the extent of the risk—may not be available within that time frame. Agency officials also questioned the value of reporting certain types of PII breaches, such as paper-based incidents or incidents involving the loss of hardware containing encrypted PII, individually to US-CERT, as currently required. Officials from US-CERT agreed that their office should not be receiving all PII-related incident reports individually as they occur. According to DHS officials, the PII-related incident data they collect are not generally used to help remediate incidents or provide technical assistance to agencies. Rather, the information is compiled in accordance with certain FISMA requirements and reported to OMB. We determined that the limited use of these data calls into question OMB’s requirement that such incidents be reported within 1 hour. US-CERT officials also noted that the vast majority of PII-related data breaches are not cybersecurity related—that is, they do not involve attacks on or threats to government systems or networks. Thus receiving information about such incidents on an individual basis may not be useful to the office in pursuing its mission. Finally, we reported that seven of the eight agencies in our review had not requested technical assistance from US-CERT when PII data breaches occurred. DHS officials said that US-CERT is not equipped to assist agencies in remediating paper-based incidents, and agencies agreed that issues they encounter in dealing with PII breaches are generally best addressed by agency general counsel staff or privacy officers. DHS’s Privacy Office has developed guidance that addresses agencies’ obligations to protect PII and procedures to follow when a suspected PII incident occurs, but this is geared more toward developing agency response capabilities in general rather than supporting decision-making related to specific incidents. In our report, we recommended that OMB revise its guidance on federal agencies’ response to PII-related data breaches to include (1) guidance on notifying affected individuals based on a determination of the level of risk; (2) criteria for determining whether to offer assistance, such as credit monitoring, to affected individuals; and (3) revised requirements for reporting PII-related breaches to US-CERT. In commenting on our draft report, officials from OMB’s Office of Information and Regulatory Affairs stated that our recommendation did not sufficiently specify what supplemental guidance was needed; we subsequently revised the draft recommendation to provide greater specificity. We also made a number of recommendations to the individual agencies in our review to improve their response to data breaches involving PII. Specifically, we recommended, among other things, that several of the agencies (1) consistently document risk levels and how those levels are determined for PII-related data breach incidents; (2) document the number of affected individuals for each incident; and (3) identify lessons learned from responses to PII breaches. Agencies varied in the extent to which they concurred with these recommendations, with some providing information pertaining to the recommendations. In response to agencies’ comments, we clarified or deleted three draft recommendations but retained the rest as still warranted. In a forthcoming report, to be issued later this spring, we plan to provide the results of our study of federal agencies’ ability to respond to cyber incidents. More specifically, we have determined the extent to which (1) federal agencies are effectively responding to cyber incidents, and (2) DHS is providing cybersecurity incident assistance to agencies. While these results are still subject to revision, we estimate, based on a statistical sample of cyber incidents reported in fiscal year 2012, that the 24 major federal agencies did not effectively or consistently demonstrate actions taken in response to a detected cyber incident in about 65 percent of reported incidents. For example, agencies identified the scope of incidents in the majority of cases, but did not always demonstrate that they had determined the impact of an incident. In addition, agencies did not consistently demonstrate how they had handled other key activities, such as whether actions to prevent the recurrence of an incident were taken. We also reviewed six selected agencies in greater depth and found that, while they had developed parts of policies, plans, and procedures to guide incident response activities, their efforts were not comprehensive or fully consistent with federal requirements. The inconsistencies in agencies’ incident response activities suggest that additional oversight, such as that provided by OMB and DHS during the CyberStat review process, not covered agencies’ incident response practices. may be warranted. However, these meetings generally have With regard to DHS’s role, we observed that DHS provides various services to agencies to assist them in preparing to handle incidents, maintain awareness of the current threat environment, and deal with ongoing incidents. However, opportunities exist to enhance the usefulness of these services, such as improving reporting requirements and evaluating the effectiveness of these services. To improve the effectiveness of government-wide cyber incident response activities, we are planning to make recommendations to OMB and DHS to address agency response practices. We also plan to make recommendations to the six selected agencies in our review to improve their cyber incident response programs. CyberStat reviews are in-depth sessions with National Security Staff, OMB, DHS, and an agency to discuss that agency’s cybersecurity posture and opportunities for collaboration. In summary, the increasing number of cyber incidents at federal agencies, many involving the compromise of PII, highlights the need for focused agency action to ensure the security of the large amount of sensitive personal information collected by the federal government. These actions include establishing comprehensive agency-wide information security programs and consistently and effectively responding to incidents when they occur. As we and inspectors general have long pointed out, federal agencies continue to face challenges in effectively implementing all elements of their information security programs. Likewise, agencies have not been consistent or fully effective in responding to data breaches and cyber incidents. Ongoing improvements in these areas are needed to help ensure that the personal information entrusted to the government by American citizens and other individuals will be protected. Chairman Carper, Ranking Member Dr. Coburn, and Members of the Committee, this concludes my statement. I would be happy to answer any questions you may have. If you have any questions regarding this statement, please contact Gregory C. Wilshusen at (202) 512-6244 or wilshuseng@gao.gov. Other key contributors to this statement include John A. de Ferrari and Jeffrey Knott (assistant directors), Larry E. Crosland, Marisol Cruz, and Lee McCracken. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

The federal government collects large amounts of PII from the public, including taxpayer data, Social Security information, and patient health information. It is critical that federal agencies ensure that this information is adequately protected from data breaches, and that they respond swiftly and appropriately when breaches occur. Since 1997, GAO has designated information security as a government-wide high-risk area. Further, data breaches at federal agencies have raised concerns about the protection of PII. Federal laws and other guidance specify the responsibilities of agencies in securing their information and information systems and in responding to data breaches. This testimony addresses federal agencies' efforts to secure their information and respond to data breaches. In preparing this statement, GAO relied primarily on previously published and ongoing work in this area. The number of reported information security incidents involving personally identifiable information (PII) has more than doubled over the last several years (see figure). As GAO has previously reported, major federal agencies continue to face challenges in fully implementing all components of an agency-wide information security program, which is essential for securing agency systems and the information they contain—including PII. Specifically, agencies have had mixed results in addressing the eight components of an information security program called for by law, and most agencies had weaknesses in implementing specific security controls. GAO and inspectors general have continued to make recommendations to strengthen agency policies and practices. In December 2013, GAO reported on agencies' responses to PII data breaches and found that they were inconsistent and needed improvement. Although selected agencies had generally developed breach-response policies and procedures, their implementation of key practices called for by Office of Management and Budget (OMB) and National Institute of Standards and Technology guidance was inconsistent. For example, only one of seven agencies reviewed had documented both an assigned risk level and how that level was determined for PII data breaches; two agencies documented the number of affected individuals for each incident; and two agencies notified affected individuals for all high-risk breaches. the seven agencies did not consistently offer credit monitoring to affected individuals; and none of the seven agencies consistently documented lessons learned from their breach responses. Incomplete guidance from OMB contributed to this inconsistent implementation. For example, OMB's guidance does not make clear how agencies should use risk levels to determine whether affected individuals should be notified. In addition, the nature and timing of reporting requirements may be too stringent. In its December 2013 report, GAO made 22 recommendations to the agencies included in its review aimed at improving their data breach response activities. GAO also recommended that OMB update its guidance on federal agencies' responses to PII-related data breaches. Agency responses to GAO's recommendations varied.

For almost a decade, the government of Mexico has sought to combat the growing power of criminal groups that initially emerged as DTOs in the 1980s and 1990s. This struggle became a national priority in 2006 when then-President Felipe Calderón mobilized the Mexican military and law enforcement agencies to disrupt DTO operations and target their leadership structures. As the Congressional Research Service reported, while these efforts have continued, under current President Enrique Peña Nieto, who was elected in 2012, there has been a shift in emphasis toward reducing criminal violence that threatens the security of civilians and the business sector. According to a RAND Corporation report, besides trafficking billions of dollars’ worth of narcotics into the United States annually, Mexican DTOs’ criminal activity now extends to other areas, including human trafficking, kidnapping, money laundering, extortion, bribery, racketeering, and weapons trafficking. According to the Strategy DTOs require a constant supply of firearms and ammunition to assert control over the territory where they operate, eliminate rival criminals, enforce illicit business dealings, and resist government operations. The Strategy indicates that firearms that criminal organizations acquire from the United States are primarily transported overland into Mexico using the same routes and methods employed when smuggling bulk cash south and drugs north across the U.S.-Mexico border. The Strategy also notes that within the United States, DTOs or their agents typically rely on “straw purchasers.” According to ATF, a “straw purchase” occurs when a person who is a convicted felon (or otherwise prohibited by federal law from purchasing a firearm) or who wishes to remain anonymous, uses a third party, the straw purchaser, to execute the paperwork necessary to purchase a firearm from a federally licensed firearms dealer. The straw purchaser is a person who, but for making false statements on the license application, would otherwise be eligible under federal law to purchase a firearm and is therefore able to pass the mandatory background check conducted by the federal firearms licensee. Although straw purchasers may legally purchase firearms for their own possession and use, when they purchase firearms on behalf of criminals or others, they violate federal law by making a false statement to a federal firearms licensee on the required forms. Firearm trafficking organizations also frequently obtain firearms from unlicensed private sellers in secondary markets, particularly at gun shows and flea markets or through classified ads or private-party Internet postings, according to ATF officials. The surge in criminal activity by DTOs along the U.S.-Mexico border has generated concern among policymakers that this violence is spilling over into the United States. Since 2009, according to the National Drug Threat Assessment—which is produced by the U.S. Department of Justice’s National Drug Intelligence Center, Mexican-based DTOs have been known to operate in more than a thousand cities in the United States. While the extent of violence seen in Mexico has not been reported in the United States, law enforcement officials in two border cities we visited told us that murders and other criminal activity on the U.S. side are often linked to Mexican DTO activities. The governments of the United States and Mexico have committed to work together to stem the activities of these criminal organizations, including illicit arms trafficking. From fiscal year 2008 to fiscal year 2015, Congress appropriated about $2.5 billion in assistance for Mexico that has been provided through the Mérida Initiative, including approximately $194 million provided in the Consolidated and Further Continuing Appropriations Act, 2015. For fiscal year 2016, the administration’s budget request for the Mérida Initiative is $119 million, from various accounts. The Mérida Initiative is a bilateral security partnership between the United States and Mexico to fight organized crime and build the capacity of Mexico’s justice sector and law enforcement institutions to uphold the rule of law. Among the many activities supported under the Mérida Initiative, some assistance is provided to help combat firearms trafficking, such as providing canines trained to detect weapons and ammunition, and non-intrusive inspection equipment to detect the flow of illicit goods, including firearms. DOJ’s ATF and DHS’s ICE are the two primary agencies combating illicit sales and trafficking of firearms across the Southwest border. ATF combats firearms trafficking within the United States and from the United States to other countries as part of its mission under the Gun Control Act (see table 1). ATF is responsible for investigating criminal and regulatory violations of federal firearms laws, among other responsibilities. In carrying out its responsibilities, ATF licenses and regulates federal firearms licensees to ensure that they comply with applicable laws and regulations. ATF also traces U.S. and foreign manufactured firearms for international, federal, state, and local law enforcement agencies to link a firearm recovered in a criminal investigation to its first retail purchaser. This information can be used to help link a suspect in the criminal investigation to a firearm or identify potential traffickers. ATF is the only entity within the U.S. government with the capacity to trace firearms seized in crimes in Mexico. The agency has conducted investigations to identify and prosecute individuals involved in firearms trafficking schemes and has provided training to Mexican law enforcement officials on firearms identification and tracing techniques, among other efforts. ICE enforces U.S. export laws, and ICE agents and other staff address a range of issues, including combating the illicit smuggling of money, people, drugs, and firearms (see table 2). As the primary federal law enforcement agency responsible for investigating international smuggling operations and enforcing U.S. export laws, ICE’s Homeland Security Investigations division targets the illegal movement of U.S.-origin firearms, ammunition, and explosive weapons with the goal of preventing the procurement of these items by DTOs and other transnational criminal organizations. ICE’s investigative strategy includes the identification and prosecution of criminal networks and individuals responsible for the acquisition and movement of firearms from the United States. Other U.S. agencies that contribute to the effort to stem firearms trafficking to Mexico include: CBP. DHS’s CBP is charged with managing, securing, and controlling the nation’s borders for both people and cargo entering and leaving the United States. CBP’s outbound mission is to facilitate the movement of legitimate cargo, while interdicting the illegal export of weapons and other contraband out of the United States. State. State’s Bureau of International Narcotics and Law Enforcement Affairs (INL) advises the President, Secretary of State, and government agencies on policies and programs to combat international narcotics and crime. INL programs support State’s strategic goals to reduce the entry of illegal drugs into the United States and to minimize the impact of international crime on the United States and its citizens. INL oversees funding provided to build the capacity of Mexico to fight organized crime under the Mérida Initiative, including funds to support efforts to combat firearms trafficking. ONDCP. ONDCP is a White House component whose principal purpose is to establish policies, priorities, and objectives for the nation’s drug control program. It produces a number of publications, including the Strategy—first issued in 2007. The Strategy is intended to serve as an overarching guide for combating criminal activity along the U.S.-Mexico border; since 2009 it has included a Weapons Chapter in recognition of the threat posed by the smuggling of firearms across the Southwest border. Given ATF’s and ICE’s roles in combating firearms trafficking, these agencies share responsibility for preparing the information presented in the Weapons Chapter of the Strategy. While ONDCP tracks progress by U.S. agencies in meeting these objectives, it is not directly involved in planning or implementing their activities. Data from ATF on firearms seized in Mexico and traced from calendar year 2009 to 2014 indicate that the majority originated in the United States. Because of the illicit nature of the trafficking, the exact number of firearms trafficked from the United States into Mexico is unknown. Similarly, ATF officials noted that since firearms seized in Mexico are not always submitted for tracing the same year they were seized, or are not submitted at all, it is not possible to develop data to track trends on firearms seized. However, ATF uses the number of firearms seized and traced as an indicator to estimate extent of illicit firearms trafficking. While the government of Mexico collects data on the number of firearms its law enforcement entities seize each year, our analysis and findings refer exclusively to the universe of firearms seized in Mexico that were submitted for tracing using eTrace. According to ATF data, of the 104,850 firearms seized by Mexican authorities and submitted for tracing from 2009 to 2014, there were 73,684, or 70 percent, found to have originated in the United States. About 17 percent of the total, 17,544 firearms, were traced to a country other than the United States. ATF could not determine the origin of 13,622 (13 percent) of these firearms because of incomplete information. See figure 1. From 2009 to 2011, numbers of firearms seized by Mexican authorities and submitted for tracing fluctuated significantly, followed by a steady decline after 2011. According to ATF officials, shifts in the number of guns seized and traced do not necessarily reflect fluctuations in the volume of firearms trafficked from the United States to Mexico from one year to the next. ATF staff explained that there are several factors that have influenced the year-to-year variance in the number of firearms traced since 2009. For example, they explained that the high number of firearms traced in 2009 reflects a single submission by the Mexican military to ATF for tracing of a backlog of thousands of firearms. Conversely, ATF officials noted there was a lower number of firearms submitted for tracing in 2010 because that is the year eTrace in Spanish was initially deployed in Mexico, and Mexican law enforcement officials at the local, state, and federal level had to be trained on using the system. In 2011, a much higher number of firearms were traced as Mexican officials became proficient in using the system. Finally, U.S. and Mexican officials suggest the decline since 2011 may reflect a period of adjustment in cooperation under the Peña Nieto administration. This included the centralization of access to eTrace in Mexico’s Attorney General’s Office and retraction of eTrace accounts from federal, state, and local law enforcement, which resulted in fewer Mexican law enforcement officials able to trace firearms using the system. According to Mexican law enforcement officials we interviewed, DTOs prefer high caliber weapons with greater firepower, including high caliber rifles or long guns, and military grade equipment. Officials explained that the firearms of choice for drug traffickers are high caliber assault rifles, such as AK type and AR 15 type, which are available for purchase in the United States and which can be converted to fully automatic fire (i.e., converted into machine guns). Officials also noted that in recent years they have seen DTOs acquire military equipment, such as .50 caliber machine guns, rocket launchers, and grenade launchers. However, they said that unlike firearms typically used by DTOs, which often can be traced back to the United States, this type of equipment is known to often be trafficked into Mexico from leftover Central American military stockpiles from past conflicts. See figure 2 for examples of long and short guns (also referred to as handguns). According to data provided by ICE, the agency seized 5,951 firearms that were destined for Mexico in the last 6 years. Of firearms seized by ICE from 2009 to 2014, 2,341, or 39 percent, were long guns—including rifles and shotguns. During the same period, ICE seized 3,610 short guns— including revolvers and pistols (see fig. 3). According to data provided by ATF, almost half of all firearms seized in Mexico and traced in the last 5 years were long guns. From 2009 through 2014, 49,566 long guns—rifles and shotguns—were seized and traced. During that same period, 53,156 short guns—including revolvers and pistols—were seized and traced. The data also show a substantial decline in the number of long guns traced since 2011 (see fig. 4). Mexican law enforcement officials said that in the last 2 years they often seized more handguns than rifles, but stated that the use of high caliber rifles by cartels is still widespread. According to ATF officials, steps the bureau has taken to combat firearms trafficking to Mexico have made it more difficult for firearms traffickers to acquire long guns. Specifically, they noted implementation of Demand Letter 3, which requires licensed dealers and pawnbrokers in Arizona, California, New Mexico, and Texas to report multiple sales of certain rifles. According to ATF, information from multiple sales reports on long guns allows the bureau to identify indicators of suspicious or high-volume purchasing by individuals, repetitive purchasing, and purchases by associates, as well as geographical trends for such sales. ATF officials reported that this information has helped them identify firearms traffickers and others involved in a timelier manner, which on several occasions has led to arrests and seizures of firearms intended for trafficking to Mexico. From 2011 to 2014, 490 long guns that had been recorded as part of multiple sales transactions under Demand Letter 3 were seized in crime scenes—259 in the United States, 209 in Mexico, and 22 in undetermined locations. Most of the firearms seized in Mexico that were traced and found to be of U.S. origin from 2009 to 2014 came from U.S. Southwest border states. While guns seized in Mexico of U.S. origin were traced to all of the 50 states, most came from Texas, California, and Arizona. As shown in figure 5, of all firearms seized in Mexico that were traced and identified to be of U.S. origin, about 41 percent came from Texas, 19 percent from California, and 15 percent from Arizona. According to ATF, in fiscal year 2014, there were about 10,134 licensed dealers and pawnbrokers in the four Southwest border states, many of them along the border. This represents about 16 percent of the approximately 63,311 licensed dealers and pawnbrokers nationwide. These licensed dealers and pawnbrokers can operate in locations such as gun shops, pawn shops, their own homes, or gun shows. According to ATF officials, most firearms seized in Mexico and traced back to the United States are purchased in the United States then transferred illegally to Mexico. ATF has been able to determine the original retail purchaser for about 45 percent of firearms seized in Mexico and traced to the United States from 2009 to 2014. However, ATF was unable to determine a purchaser for 53 percent, because of factors such as incomplete identifying data on trace request forms, altered serial numbers, no response from the federal firearm licensee to ATF’s request for trace information, or incomplete or never received out-of-business licensee records. ATF and Mexican government officials told us that a new complicating factor in their efforts to fight firearms trafficking is the use of weapons parts transported to Mexico to be later assembled into finished firearms. According to documents provided by ATF, firearm parts include unfinished receivers barrels, triggers and hammers, buttstocks, pistol grips, pins, bolts, springs, and other items. Figure 6 shows some of these firearms parts. None of these firearm parts are classified as firearms under the Gun Control Act. In general, U.S. federal laws and regulations requiring manufacturers and importers of firearms to identify firearms with a serial number do not apply to parts, unless otherwise specified by law. Federal firearms licensees and other retailers are not required to report on the acquisition and disposition of firearm parts as they must for firearms. Furthermore, any individual in the United States may legally acquire and possess certain firearm parts that are not otherwise proscribed by law, including persons prohibited from possessing firearms and ammunition, such as convicted felons. Firearms may be assembled by using parts kits that include all of the components of a fully operable firearm minus the firearm receiver or frame, which may be obtained separately. ATF officials explained that in order to circumvent marking requirements on transactions involving firearms and thus avoid tracing, criminals will sometimes use unfinished receivers, such as “castings” or “flats,” rather than fully functional receivers. A frame or receiver by itself is classified as a firearm by definition under the Gun Control Act. The receiver is the part of the firearm that houses the operating parts, typically the bolt or bolt carrier group, the magazine well, and the trigger group. A casting is essentially a piece of metal fabricated with the exterior features and contours of the firearm receiver for which it is intended to substitute, but that without further machining will not function as a firearm. Castings and flats are commonly referred to as 80 percent receivers in marketing materials and advertisements promoting their sale. The “80 percent” label is intended to convey that the product has been cast or fabricated with most of the features of a finished, functional firearm receiver, but it will require further machining to function as a firearm (see fig. 7). A receiver flat is a piece of metal that has the same dimensions as a receiver, but that has not been shaped into a firearm configuration. In this form, it cannot accept any component parts, but with the proper equipment it can be readily bent into shape and molded into a receiver (see fig. 8). According to documents provided by ATF, since kits, castings, and flats are not classified as firearms, transfers of those items are not regulated under the Gun Control Act or National Firearms Act. Although ICE officials noted they are subject to export control laws, they have no serial numbers and generally no markings; thus, firearms assembled with them are untraceable. In addition, receivers and firearms parts are small and when transported separately may not be easily identified as items intended for the production of firearms. They are also easy to conceal, making it more challenging for customs authorities to detect illicit shipments of such parts. According to ATF officials, there are no reliable data on the extent of firearm parts trafficking from the United States into Mexico. They noted, however, that recent seizures of firearms parts, firearms made with unmarked parts, and equipment used to assemble or manufacture firearms in Mexico suggest an emerging reliance by criminal organizations on this source of weapons. For example, law enforcement officials in Mexico described to us two high-profile cases in 2014 involving illicit firearm parts assembly of this type. One was in Guadalajara, where Jalisco state police seized hundreds of unfinished receivers and pieces of sophisticated equipment being used to complete high caliber rifles. The second was in Tijuana, where Baja California state police seized 25 rifles in the process of assembly with firearm parts from the United States. ATF and ICE have taken several steps to improve coordination on efforts to combat firearms trafficking that we previously identified. In 2009, we reported instances of dysfunctional operations, duplicative initiatives, and jurisdictional conflicts between ATF and ICE. In response to our recommendations on how to address these challenges, ATF and ICE updated and signed an interagency collaboration memorandum of understanding (MOU) in June 2009. In their revised MOU, the agencies committed to a shared goal of keeping the public safe by using the tools given to both agencies and which are vital to the effective control of domestic and international trafficking of firearms, ammunition, explosives, weapons, and munitions. Specifically, the MOU set forth roles and requirements for each agency with respect to (1) intelligence and information sharing, (2) general investigative guidelines, (3) specific investigative guidelines, (4) sources of information, and (5) conflict resolution. This effort to improve coordination and optimize use of the agencies’ expertise provided the basis to address the issues that had hampered interagency collaboration prior to the MOU’s implementation. ICE and ATF officials said that after the MOU was signed, they held joint training exercises and conferences to ensure that agents had knowledge of the MOU and its jurisdictional parameters and collaboration requirements. Officials from each agency in headquarters, Mexico, and border locations we visited indicated that personnel working on firearms trafficking to Mexico were generally aware of the MOU’s key provisions and collaborated on this basis. Agency officials also highlighted a more recent joint interagency conference in September 2014, which sought to provide participants with a common understanding of collaborative efforts and respective areas of jurisdiction. Additionally, senior agency headquarters officials asserted that there is extensive cooperation between ATF and ICE, at the headquarters and field office levels. ATF and ICE officials in border field offices we visited confirmed that they were familiar with the MOU and that it provides them guidance on interagency collaboration. Similarly, ATF and ICE officials in Mexico stated that since they are co-located physically, they have a greater opportunity to work together closely on firearms trafficking-related cases, and an ICE official said that they rely on the MOU to help define their respective roles. Nevertheless, we identified persistent challenges in information sharing and some disagreement on the agencies’ respective roles in investigations. For example, ATF and ICE disagree on the extent to which trace data on firearms seized in Mexico collected through eTrace should be shared to support ICE firearms trafficking investigations. According to an ICE assistant deputy director, these firearms trace data from Mexico are currently only shared on a limited basis with ICE. Several ICE officials expressed an interest in obtaining access to these data and indicated that this access would enhance their ability to identify methodologies used by firearms traffickers and trends in criminal activity along the Southwest border. ICE officials responsible for investigations said that trace data should be shared in accordance to the MOU, which states “ATF shall report to the appropriate ICE field office in a timely manner any intelligence received relating to the illegal exportation, attempted exportation, or planned exportation of any item on the United States Munitions List...” However, the MOU does not address how general trafficking information, such as that submitted through eTrace by a third law enforcement agency, may be shared. ATF officials asserted that their agency shares trace data on firearms seized in Mexico with ICE according to established agency polices, which currently only allow ATF to provide non-case-specific information to other agencies in aggregate form. With respect to the results of individual trace requests, ATF officials explained that they are provided only to the law enforcement agency that submits the trace information; generally, this information is not shared with third parties, including other law enforcement agencies. ATF would have to obtain authorization from the third-party law enforcement agency that submitted the trace information to share it with ICE. Thus, ATF cannot automatically share information with ICE on firearm traces submitted by Mexican law enforcement agencies without their authorization. ATF staff said these policies are set forth in the agreements ATF signs with each law enforcement agency for the use of eTrace. Officials from ATF and ICE said there are joint efforts under way to find a mechanism to share this information. Additionally, the 2009 MOU sets forth investigative guidelines to define the roles and responsibilities of ATF and ICE pursuant to their respective statutory authorities. For example, the MOU states that “the regulation and inspection of the firearms industry is within the sole purview of ATF” and that “all investigative activities at the port of entry, borders and their functional equivalents must be coordinated through ICE.” Notwithstanding these guidelines, we found some confusion among some agency officials about the appropriate roles of their counterparts in conducting investigations. For example, a senior ICE official responsible for investigations questioned the involvement of ATF in firearms trafficking investigations to Mexico, because, according to the official, ATF’s jurisdiction focuses on combating domestic firearms violations. ICE officials also expressed concerns that the involvement of ATF’s international desk with Mexican agencies may create confusion among Mexican government authorities over the roles that ICE and ATF play in addressing firearms trafficking cases. However, an ICE assistant deputy director explained that pursuant to the Arms Export Control Act, ICE has primary jurisdiction over violations related to the international trafficking of firearms, but many such trafficking investigations begin with domestic criminal activities for which ATF has jurisdiction. Therefore, he stressed that it is essential that the two agencies collaborate to leverage ICE’s international and ATF’s domestic legal authorities. He added that ATF’s international operations also provide much-needed capacity building regarding forensics and e-Trace activities in Mexico. However, ICE and ATF must work to ensure that confusion is not created among Mexican agencies regarding the responsibilities for the investigation of international firearms trafficking by U.S. authorities. ATF officials agree that their agency’s efforts to combat firearms trafficking are concentrated in the United States, and that they recognize the role of ICE in addressing transnational weapons trafficking. However, some ATF officials said that it is incorrect to suggest that ICE has exclusive jurisdiction with respect to illicit cross-border firearms trafficking to Mexico. According to these officials, most investigations involving the smuggling of firearms from the United States to Mexico implicate ATF jurisdiction, because they typically involve the illegal acquisition of firearms inside the United States. ATF’s jurisdiction extends to unlawful acquisition of firearms by prohibited persons, straw purchasing, and other unlawful transfers of firearms. ATF officials added that the bureau’s statutory responsibility for tracing firearms includes the deployment of eTrace to Mexican and other foreign law enforcement entities, and noted that eTrace entries from Mexico can result in the opening of firearm trafficking investigations focused on criminal activity in the United States. ATF officials also acknowledge that because of the nature of firearms trafficking to Mexico, many investigations involve overlapping jurisdiction with respect to cross-border offenses squarely within ICE’s jurisdiction. They also noted the critical role ATF plays in providing training and capacity building on firearms and explosives identification and tracing for Mexican law enforcement. During our fieldwork, Mexican law enforcement agencies confirmed the benefits they derived from ATF capacity-building efforts, and they said they regarded ATF as their lead U.S. counterpart in investigating firearms trafficking. Thus, although ATF has established productive cooperative relations with Mexican agencies, there may also be some confusion in Mexico over ATF’s and ICE’s roles in combating firearms trafficking, as expressed by some ICE officials. In prior work, we have identified several interagency mechanisms that can be used to improve collaboration among agencies working on a shared mission, such as information sharing, agency roles and responsibilities, and oversight and monitoring. We have also reported that written interagency agreements, such as MOUs, are most effective when they are regularly updated and monitored. We observed that when implementation of such agreements is not regularly monitored, there is sporadic and limited collaboration among agencies. We also have found that agencies that create a means to monitor, evaluate, and report the results of collaborative efforts can better identify areas for improvement. Immediately after the MOU was updated in 2009, the agencies committed to undertake efforts to ensure that its provisions would be implemented accordingly. For example, at that time, ICE informed GAO that headquarters had a process to obtain information from ICE field offices every 60 days to identify coordination issues with ATF that could not be resolved at the field level within the framework of the MOU. In such situations, ICE headquarters would then work with the appropriate ATF component to resolve the issue. ICE officials explained these initial monitoring efforts were designed to ensure that the updated MOU was being effectively followed as it introduced several provisions or guidelines on how ATF and ICE should collaborate on firearms trafficking. However, according to ICE officials, this process was only in place during the initial implementation period of the MOU, and the effort was not sustained. Currently, officials from both agencies acknowledged that there is no specific mechanism in place to monitor implementation of the MOU. However, each agency’s officials referred to different efforts that they said provide an opportunity to monitor interagency collaboration under the MOU. For example, a deputy assistant director for ICE stated that coordination between ICE and ATF on firearms trafficking cases occurs at the Export Enforcement Coordination Center as well as at the field level. ICE’s Export Enforcement Coordination Center is intended to serve as the primary forum within the federal government for executive departments and agencies to coordinate their export control enforcement efforts. The Center seeks to maximize information sharing, consistent with national security and applicable laws. Thus, it is likely that coordination challenges between ICE and ATF on firearms trafficking could potentially be detected at the Center. However, given the Center’s broader responsibility to enhance export control enforcement efforts with multiple agencies, it is not directly intended to monitor implementation of the MOU. Moreover, coordination challenges related to the MOU persist even though the Center has been in place for 5 years, indicating that this may not be an effective means to monitor the MOU’s implementation. Senior ATF officials said that although there is no formal arrangement to regularly monitor implementation of the MOU, they consider joint interagency training to be an effective approach to ensure that officials from both agencies are familiar with the provisions of the MOU and are working together effectively. However, only two such training exercises have taken place—one in 2014 and another in September 2015. The training is intended to acquaint officials from both agencies with how the agencies coordinate firearms trafficking efforts, and as part of the training, the MOU provisions are discussed, but these training exercises do not constitute a mechanism for consistent monitoring of implementation of the MOU. By not sustaining a monitoring process for the MOU, the agencies have no assurance that its provisions are being implemented effectively, and challenges we identified are continuing to persist without a process for resolution. Mexican and U.S. officials described how upon coming to power in December 2012, the current administration of Mexican President Enrique Peña Nieto undertook a reevaluation of U.S.-Mexico law enforcement collaboration, including efforts to combat firearms trafficking. According to some officials, the government of Mexico took steps to consolidate law enforcement cooperative activities under an approach termed Ventanilla Única—which translates to Single Window. Under Ventanilla Única, Mexico’s Interior Ministry has become the primary entity through which Mérida Initiative training and equipment are coordinated, including capacity-building activities related to firearms trafficking. The government of Mexico also established a single point of contact within Mexico’s Office of the Attorney General to approve joint investigations with U.S. counterparts. Additionally, Mexican officials explained that Mexican law categorizes firearms trafficking as a federal crime and permits only federal authorities to work on such cases. This has led to some notable changes in the way U.S. and Mexican authorities work together on firearms trafficking efforts. One of these changes stemmed from the decision to centralize access to ATF’s eTrace in the Mexican Attorney General’s Office. Consistent with our prior recommendations, in 2010, ATF reached an agreement for deployment of eTrace in Spanish in Mexico, with Mexican authorities. According to ATF, this was a significant investment for which ATF provided training to numerous officials from various Mexican federal, state, and local law enforcement agencies on the use of eTrace, while assigning accounts to allow them to access the system. However, by 2013 the Mexican government retracted access to many of these accounts, effectively limiting eTrace in Mexico to certain authorized officials in the office of Mexico’s Attorney General. Mexican officials explained that the decision to consolidate access at the Attorney General’s office was intended to provide the government of Mexico with more effective control over the information associated with eTrace, and to support a central repository of evidence related to federal crimes such as trafficking of firearms. However, U.S. officials and some Mexican authorities said that limiting access to eTrace to a single governmental entity has restricted opportunities for bilateral collaboration. Some U.S. officials based in Mexico similarly noted that limiting access to eTrace diminished tracing of total firearms seized by Mexican authorities. Another significant change following the reassessment of bilateral collaboration, which began in 2012, was the suspension of periodic meetings of a working group known as GC Armas, which brought together U.S. and Mexican officials from various agencies involved in combating firearms trafficking. According to ATF officials, prior to 2013, GC Armas held periodic meetings annually with the participation of approximately 70 to 100 officials from both governments. These officials shared useful information on firearms trafficking trends, trace data, investigations, collaboration questions, and many other issues. ATF officials said that oftentimes very productive cooperative efforts on firearms trafficking began informally at GC Armas meetings and were subsequently formalized. Mexican officials similarly characterized GC Armas meetings as contributing in a fundamental manner to reaching significant agreement between law enforcement in both countries on how to combat firearms trafficking. They noted one such bilateral effort that resulted in a comprehensive assessment of firearms and explosives trafficking with recommendations for each country on sharing information and cooperating on cross-border investigations. Officials from both countries explained that while bilateral coordination did not cease after the suspension of GC Armas meetings, overall collaboration slowed down with fewer opportunities to promote bilateral firearms trafficking initiatives. U.S. and Mexican authorities acknowledge the challenges to law enforcement efforts posed by continuing corruption among some Mexican officials. As we discussed in our 2009 report, concerns about corruption within Mexican government agencies often limit U.S. officials’ ability to develop a full partnership with their Mexican counterparts. Officials we met with from ATF, ICE, CBP, and State continued to express such concerns regarding corruption in Mexico. Some Mexico-based ICE officials, for example, stated that they are conscious that their U.S.-based colleagues will not always share with them all of the information they have on firearms trafficking investigations because of concerns about corruption. That is, ICE officials in the United States and along the U.S.- Mexican border are concerned about sharing information with ICE officials based in Mexico, fearing that the information may unintentionally reach corrupt Mexican authorities and compromise their investigations. According to ICE officials, concerns they had about corruption in Mexico were exacerbated early in the Peña Nieto administration when a vetted unit of Mexican law enforcement officials that they trusted and that ICE had trained and worked with for several years was disbanded. U.S. officials also highlighted the problems frequent turnover in Mexican law enforcement pose for bilateral efforts to combat criminal activities, including firearms trafficking. Some U.S. officials explained that recurring personnel changes aggravate the issue of corruption. In a country such as Mexico, where there is an underlying concern about government corruption, frequent turnover complicates efforts to develop trust with counterpart officials. Other U.S. officials noted that there are no civil service protections in Mexico, so there can be a virtually complete change in the staff of a government agency when a new administration comes into office, or even when the head of an agency is reassigned. As a result, all of the people who received specialized training, such as firearms recognition, can be removed suddenly leaving no institutional memory, which complicates planning future collaboration and program implementation. Similarly, ATF officials commented that oftentimes Mexican law enforcement personnel in key positions for whom they provided firearms training were subsequently replaced. While turnover has been a recurring challenge for U.S. agencies working in Mexico, various U.S. officials said that it appears to have been particularly frequent in the past few years. For example, the spokesperson for one U.S. agency in Mexico noted that in the past 5 years the division responsible for implementing professional development at a key Mexican law enforcement entity has been replaced seven times. While both U.S. and Mexican officials collaborating on firearms trafficking said that bilateral efforts had been scaled back after the Peña Nieto administration came into power, these officials noted that over the past year collaborative activities have been bolstered and are gaining momentum. For example, around the time of our fieldwork in Mexico, CBP was working with Mexican authorities to deploy specially trained canine units able to detect firearms and explosives around the country. Similarly, ATF was providing training on firearms identification for Mexican Customs. A Mexican Customs spokesman stressed the importance of such training in helping front-line customs officers recognize and safely secure not just firearms but also ammunition, firearms’ components, and explosives that criminals try to smuggle across the border. He explained that this training has been critical in allowing officers at the border to perform their mission. Mexican Attorney General officials also noted their increasing level of cooperation with U.S. authorities on firearms trafficking. They highlighted ATF training on the use of eTrace and the resumption of GC Armas meetings in 2015. ICE officials also told us that they have recently reestablished the vetted unit in Mexico, which improves trusted working relationships with Mexican counterparts. Finally, in addition to renewing existing collaborative efforts with Mexican law enforcement counterparts, ATF has also sought to reach out to other Mexican government entities. For example, this year ATF has been collaborating with the Mexican Navy on training for firearms and explosives detection, identification, and seizure. Mexican Navy officers expressed gratitude for this training, noting that they are increasingly confronting real-world situations that require this type of knowledge. The indicator used in the Strategy to track progress by U.S. agencies to stem firearms trafficking to Mexico does not adequately measure implementation of the strategic objective. The Strategy includes strategic objectives and indicators for each of its nine issue chapters to ensure effective implementation. The strategic objective for the Weapons Chapter is to “stem the flow of illegal weapons across the Southwest border into Mexico.” ONDCP’s indicator for this chapter is the “number of firearms trafficking/smuggling seizures with a nexus to Mexico.” The Strategy does not further define the indicator, but ONDCP staff explained that it refers to the number of firearms seized in Mexico that are traced by ATF. While ONDCP’s Strategy asserts that it is critical to have indicators that enable tracking the implementation of objectives, this indicator for the Weapons Chapter does not effectively track the status of efforts to stem the flow of illegal weapons across the Southwest border. As previously noted in this report, ATF officials readily acknowledge that shifts in the number of guns seized and traced do not necessarily reflect fluctuations in the volume of firearms trafficked from the United States to Mexico in any particular year. There are many factors that could account for the number of firearms traced in a given year beside the number of firearms smuggled from the Unites States. Moreover, as discussed above, for various reasons the number of firearms seized in Mexico and traced back to the United States shifted significantly year to year after 2009 and then declined steadily since 2011. Thus, while the number of firearms seized and traced by ATF is useful to provide an overall indication of firearms of U.S. origin found in Mexico, by itself it is not an adequate measure of progress agencies are making to stem the flow of firearms trafficked from the United States into Mexico. Additionally, ONDCP has not reported progress made on the strategic objective in the Weapons Chapter in 2011 or 2013. ONDCP staff said they anticipate that the 2015 Strategy will include a section to report on the outcomes of the last 2 years, and they plan to report on this indicator. Beside the strategic objective and indicator, the Weapons Chapter of the Strategy also includes five supporting actions, along with associated activities to achieve those actions; see table 3. According to an ONDCP spokesman, while the number of firearms seized in Mexico and traced by ATF may be an indicator of the flow of firearms across the border, these five supporting actions and their associated activities should also be considered to get a full picture of the agencies’ progress in combating arms trafficking. ONDCP officials said that they monitor progress in combating arms trafficking by obtaining periodic information from ATF and ICE on their implementation of these and other activities. Our review of the Weapons Chapter in the 2011 and 2013 Strategies determined that, generally, accomplishments under each supporting action were discussed. For example, in the 2011 Strategy, one supporting action called for ATF to increase staffing at the El Paso Intelligence Center Firearms and Explosives Trafficking Intelligence Unit through the incorporation of partner agencies. In 2013, the Strategy included an update that the unit had incorporated a CBP analyst dedicated to weapons-related intelligence. Similarly, in 2011, the agencies said they had plans under way to train over 200 Mexican law enforcement personnel in how to correctly use eTrace. The 2013 Strategy noted that 350 Mexican law enforcement personnel had received training on using eTrace. Nevertheless, the supporting actions described in the Strategy are not consistently linked to indicators or regularly measured. Currently, the narrative related to these supporting actions typically covers ongoing efforts by the agencies to address these actions, but it does not include a measure of overall progress. By including these supporting actions and activities in the Weapons Chapter as measures, ONDCP could better assess the agencies’ efforts in combating firearms trafficking because this would provide a more comprehensive assessment. Although ATF and ICE have pledged, through the 2009 MOU, to collaborate effectively to combat firearms trafficking, these agencies have not set up a mechanism to monitor implementation of the MOU that would allow them to identify and address information sharing and collaboration challenges. Consequently, gaps in information sharing and some disagreement about agency roles in the broader effort to combat firearms trafficking have emerged that weaken the effectiveness of the MOU. It is unclear to what extent ONDCP’s Strategy has advanced U.S. government efforts to combat firearms trafficking, since the indicator used to track progress, by itself, is not sufficient to measure progress made by U.S. agencies in stemming arms trafficking to Mexico. Other actions that agencies take to stem the flow of firearms from the United States into Mexico may be worth considering as additional measures of progress, such as the number of interdictions of firearms destined for Mexico, the number of investigations leading to indictments for firearms trafficking related to Mexico, and the number of convictions of firearms traffickers with a nexus to Mexico. By including these types of measures in a comprehensive indicator or set of indicators, ONDCP will be better positioned to monitor progress on stemming firearms trafficking across the Southwest border. We recommend that the Attorney General of the United States and the Secretary of Homeland Security convene cognizant officials from ATF and ICE to institute a mechanism to regularly monitor the implementation of the MOU and inform agency management of actions that may be needed to enhance collaboration and ensure effective information sharing. To ensure effective implementation of the strategic objective of the Weapons Chapter of the Strategy, we recommend that the ONDCP Director establish a more comprehensive indicator, or set of indicators, that more accurately reflects progress made by ATF and ICE in meeting the strategic objective. We provided a draft of this report for review and comment to the Departments of Homeland Security, Justice, and State; and the Office of National Drug Control Policy. DHS agreed with our recommendation regarding monitoring implementation of the MOU and provided written comments in response to the draft, reproduced in appendix II. In comments on the draft report provided via e-mail by a designated ATF Audit Liaison Officer, DOJ also agreed with this recommendation, noting that ATF officials will work with counterparts at DHS to create a mutually acceptable method to further enhance implementation of the MOU. State did not provide comments on the draft report. In e-mail comments provided by a designated General Counsel official, ONDCP concurred with our recommendation to establish a more comprehensive set of indicators for the Weapons Chapter of the Strategy. Accordingly, ONDCP indicated that it would work with ICE and ATF to develop additional indicators to evaluate their progress. The indicators developed through this collaborative process will be used in future iterations of the Strategy beginning with the next report in 2017. ICE and ATF also provided technical comments which we incorporated throughout the report where appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of Homeland Security; the Attorney General of the United States; the Director of the Office of National Drug Control Policy; the Secretary of State; and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6991 or farbj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. To identify data available on the origin of firearms trafficked to Mexico that were seized and traced, we relied primarily on the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) data compiled by its National Tracing Center (NTC).The data provided by NTC were obtained from ATF’s Firearms Tracing System, most of which is developed through eTrace submissions. We discussed with cognizant NTC officials the methodology used to collect these data and reviewed supporting agency documentation. Based on these discussions, we determined that NTC data were sufficiently reliable to permit an analysis of where the firearms seized in Mexico that were submitted for tracing had been manufactured and whether they had been imported into the United States before arriving in Mexico. For those firearms that were traced to a retail dealer in the United States before being trafficked to Mexico, NTC data also contained information on the states where they had originated. NTC trace data also contained information allowing identification of the types of firearms that were most commonly seized in Mexico and subsequently traced. We corroborated this information in discussions with U.S. and Mexican law enforcement officials. Since firearms seized in Mexico are not always submitted for tracing within the same year as they were seized, it was not possible for us to develop data to track trends on the types of firearms seized year to year. Similarly, we were unable to obtain quantitative data from U.S. or Mexican government sources on the users of illicit firearms in Mexico. However, there was consensus among U.S. and Mexican law enforcement officials that most illicit firearms seized in Mexico had been in the possession of organized criminal organizations linked to the drug trade. The involvement of criminal organizations with ties to drug trafficking in the trafficking of illicit firearms into Mexico was confirmed by law enforcement intelligence sources. We learned about the use of firearms parts for the assembly of firearms in Mexico through our interviews with cognizant U.S. and Mexican government and law enforcement officials and through review of ATF-provided documents. To learn more about U.S. government efforts to combat illicit sales of firearms in the United States and to stem the flow of these firearms across the Southwest border into Mexico, we interviewed cognizant officials from the Department of Justice’s (DOJ) ATF, the Department of Homeland Security’s (DHS) Immigration and Customs Enforcement (ICE) and Customs and Border Protection (CBP), and the Department of State (State) regarding their relevant efforts. We obtained data from ATF and ICE on funding for their respective efforts to address firearms trafficking to Mexico, and data from ICE on seizures of southbound firearms. To assess the reliability of the data, we discussed sources and the methodology use to develop the data with agency officials. We determined that the information provided to us was sufficiently reliable to describe agencies’ efforts to combat firearms trafficking. We also conducted fieldwork at U.S.-Mexico border crossings at El Paso, Texas, and San Diego, California. In these locations, we interviewed ATF, CBP, and ICE officials responsible for overseeing and implementing efforts to stem the flow of illicit firearms trafficking to Mexico and related law enforcement initiatives. We reviewed and analyzed DOJ and DHS documents relevant to U.S. government efforts and collaboration to address arms trafficking to Mexico, including funding data provided to us by ATF and ICE, the 2009 memorandum of understanding (MOU) between ICE and ATF, data from ICE on seizures of firearms destined for Mexico, data from ATF and ICE on efforts to investigate and prosecute cases involving arms trafficking to Mexico, and agency reports and assessments related to the issue. We also reviewed relevant prior GAO reports, Congressional Research Service reports and memorandums, and reports from DOJ’s Office of Inspector General related to ATF’s efforts to enforce federal firearms laws. We reviewed provisions of federal firearms laws relevant to U.S. government efforts to address firearms trafficking to Mexico, including the Gun Control Act of 1968, the National Firearms Act of 1934, and the Arms Export Control Act of 1976. We did not independently review any Mexican laws for this report. To determine how well agencies collaborated with Mexican authorities to combat illicit firearms trafficking, we conducted fieldwork in Mexico City, Guadalajara, and border locations in Ciudad Juarez and Tijuana, Mexico. In Mexico, we met with ATF, CBP, ICE, and State officials working on law enforcement issues at the U.S. embassy. We interviewed Mexican government officials engaged in efforts to combat firearms trafficking from the Attorney General’s Office (Procuraduría General de la República), the Federal Police (Policía Federal); the Ministry of Public Safety (Secretaría de Seguridad Pública); the Ministry of Defense (Secretaría de la Defensa Nacional); the Mexican National Intelligence Agency (Centro de Investigación y Seguridad Nacional, or CISEN); the Mexican Navy (Secretaría de Marina or Armada de Mexico); Customs (Servicio de Administración Tributaria); the Forensic Science Institute of Jalisco (Instituto Jalisciense de Ciencias Forenses); Attorney General Regional Offices, Federal Police, and State Police in Tijuana and Ciudad Juarez; and the State Attorney General in Guadalajara. Because our fieldwork was limited to selected locations along the Southwest border and in the interior of Mexico, our observations in these locations are illustrative but are not generalizable and may not be representative of all efforts to address the issue. To assess the extent to which the National Southwest Border Counternarcotics Strategy (Strategy) outlines U.S. goals and progress made in efforts to stem firearms trafficking to Mexico, we reviewed the 2011 and 2013 versions of the Strategy’s Weapons Chapter and the 2010 implementation guide. We also met with Office of National Drug Control Policy officials responsible for the implementation and monitoring the Strategy, as well as with ATF and ICE officials responsible for writing the Weapons Chapter and overseeing implementation and reporting on activities described within their respective agencies. In addition to the contact named above, Charles Johnson (Director), Juan Gobel (Assistant Director), Francisco Enriquez (Analyst-in-Charge), Danny Baez, and Julia Jebo-Grant made key contributions to this report. Ashley Alley, Karen Deans, Justin Fisher, and Oziel Trevino provided additional assistance.

Violent crimes committed by drug trafficking organizations in Mexico often involve firearms, and a 2009 GAO report found that many of these firearms originated in the United States. ATF and ICE have sought to stem firearms trafficking from the United States to Mexico. GAO was asked to undertake a follow-up review to its 2009 report ( GAO-09-709 ) addressing these issues. This report examines, among other things, (1) the origin of firearms seized in Mexico that have been traced by ATF, (2) the extent to which collaboration among U.S. agencies combating firearms trafficking has improved, and (3) the extent to which the National Southwest Border Counternarcotics Strategy measures progress by U.S. agencies to stem firearms trafficking to Mexico. To address these objectives, GAO analyzed program information and firearms tracing data from 2009 to 2014, and met with U.S. and Mexican officials on both sides of the border. According to data from the Department of Justice's Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF), 73,684 firearms (about 70 percent) seized in Mexico and traced from 2009 to 2014 originated in the United States. ATF data also show that these firearms were most often purchased in Southwest border states and that about half of them were long guns (rifles and shotguns). According to Mexican government officials, high caliber rifles are the preferred weapon used by drug trafficking organizations. According to ATF data, most were purchased legally in gun shops and at gun shows in the United States, and then trafficked illegally to Mexico. U.S. and Mexican law enforcement officials also noted a new complicating factor in efforts to fight firearms trafficking is that weapons parts are being transported to Mexico to be later assembled into finished firearms, an activity that is much harder to track. In 2009, GAO reported duplicative initiatives, and jurisdictional conflicts between ATF and the Department of Homeland Security's Immigration and Customs Enforcement (ICE). That year, in response to GAO's recommendations on these problems, ATF and ICE updated an interagency memorandum of understanding (MOU) to improve collaboration. ATF and ICE have taken several steps since then to improve coordination on efforts to combat firearms trafficking, such as joint training exercises and conferences to ensure that agents are aware of the MOU and its jurisdictional parameters and collaboration requirements. However, GAO found that ATF and ICE do not regularly monitor the implementation of the MOU. In the absence of a mechanism to monitor MOU implementation and ensure that appropriate coordination is taking place between the two agencies, GAO found that gaps in information sharing and misunderstandings related to their roles and responsibilities persist. The indicator used to track U.S. agencies' efforts to stem firearms trafficking to Mexico in the Office of National Drug Control Policy's (ONDCP) National Southwest Border Counternarcotics Strategy , by itself, does not adequately measure progress. ONDCP tracks progress based on the number of arms seized in Mexico and traced to the United States; however, this number does not reflect the total volume of firearms trafficked from the United States, and it does not take into account other key supporting agency actions and activities as measures. GAO recommends that the Secretary of Homeland Security and the Attorney General of the United States take steps to formally monitor implementation of the 2009 MOU between ATF and ICE. GAO also recommends that ONDCP establish comprehensive indicators that more accurately reflect progress made in efforts to stem arms trafficking to Mexico. The Departments of Homeland Security and Justice, and ONDCP agreed with GAO's recommendations.

States began adopting charter school laws in the early 1990s, beginning with Minnesota in 1991. Charter schools are permitted more flexibility in school operation for agreeing to accomplish specific academic goals contained in their charters. The specifics of these arrangements vary, as each state adopting a charter school law sets up its own charter school structure and guidelines, and states have continued to revise their charter school laws over time. However, not all states with a charter school law have operational charter schools, as shown in figure 2. The District of Columbia and Puerto Rico also have charter school laws. During the 2003- 04 school year, nearly 3,000 charter schools were operating nationwide, with nearly 700,000 students enrolled. Oversight authority for charter schools is established by the state’s charter school law and may rest with several entities, including state boards of education, which set educational policy, and state departments of education, which implement those policies. Some states have also created independent charter school boards that can authorize charter schools in the state. In addition, some states have created charter school offices, housed in the state department of education, that support and advocate for charter schools. State law also specifies which entities within the state can authorize the establishment of a charter school, such as state departments of education, state boards of education, local education agencies, institutions of higher education, and municipal governments. Depending on the state, a wide range of individuals or groups, including parents, educators, nonprofit organizations, and universities, may apply for permission to operate a charter school. The agreement that is reached between the applicants and the authorizer defines specific academic goals and outlines school finances and other operational considerations. In some states, including Texas and Arizona, a single charter may cover the establishment of multiple schools. Once charter schools are in operation, the authorizer is responsible for monitoring school performance and has authority to close the school or take other actions if academic goals or state financial requirements are not met. The schools are governed by a board of trustees, which is responsible for overseeing school operations. Requirements for charter school board membership vary across states, but the responsibilities are similar. Specifically, charter school boards oversee legal compliance, contracts with external parties, financial management and policies, and facilities and equipment acquisition and maintenance. NCLBA, signed into law in early 2002, increased federal funding for elementary and secondary education and created new requirements for all public schools, including charter schools. NCLBA requires states to test all children against reading and mathematics standards annually in grades 3-8 and once in grades 10-12 by the 2005-06 school year. States also had to develop performance goals for schools that identify what percentage of students must be proficient in reading and math each year for the school to achieve proficiency for all children by 2013-14. Standards for science proficiency are to be developed by the 2005-06 school year, with testing in science to begin in 2007-08. Schools are required to measure the performance of all students in meeting proficiency goals, as well as the performance of designated groups. These groups are students who (1) are economically disadvantaged, (2) represent major racial and ethnic groups, (3) have disabilities, and (4) are limited in English proficiency. NCLBA also requires that schools include at least 95 percent of students in each of these groups in statewide tests and meet at least one other academic indicator. States must use the graduation rate as the additional indicator for high schools. NCLBA also generally requires that teachers be highly qualified by the end of the 2005-06 school year, meaning that teachers must demonstrate subject matter expertise, have a bachelor’s degree and have full state certification as a teacher. NCLBA also specifies the actions that must be taken if schools receiving funds for children from low-income families under Title I of the act do not meet performance goals. Title I provides funds to states for local school districts to improve the education of low-income students in high-poverty schools. About half of all public schools nationwide receive a share of the federal funds—over $12 billion dollars in 2004—this program provides. NCLBA’s performance requirements specify that if a school receiving Title I funds does not meet its performance target for 2 consecutive years, it must provide professional development for the school’s staff and students must be offered the choice of attending another public school. If the school misses its performance goal for the third year, it must offer low- income students supplemental educational services, such as tutoring. If the school continues to miss its performance goal, additional actions are required, such as replacing the curriculum, hiring a new principal, turning the school into a charter school, letting a private company operate the school, or taking other action designed to improve student academic results. As with other public schools, funding for charter schools comes largely from state and local funds, augmented by federal aid. Eligible charter schools may receive funds under federal formula and discretionary grant programs. Formula grant programs, which provide funding to states on a noncompetitive basis, include Title I Grants to LEAs. This program is the largest federal program supporting elementary and secondary education. Title I provides funding for schools with disadvantaged children and accounts for about 2.5 percent of total education expenditures nationally. Impact Aid. This program provides funds to help educate children whose parents or residences are connected to the federal government through employment, the military, or federal housing programs. Special Education Grants to States. This program funds districts to assist them to provide special education and related services to children with disabilities. NCLBA continues to require Education and states to ensure that charter schools receive payment from 18 federal grant programs for which they are eligible, including Title I and Special Education Grants. New charter schools and charter schools with expanding enrollments are to receive these funds within 5 months of opening or expanding enrollment. This NCLBA requirement for timely payment of federal grant funds originated with the Charter School Expansion Act of 1998. Education has two data systems to support the department’s grant administration functions. One system, the Grant Administration and Payment System (GAPS), tracks the payment of federal grant funds to the grant recipient, frequently the state education agency (SEA) or LEA. Education is developing the second system, the PBDMI—a data management initiative for federal grant programs—to streamline the collection of performance and financial data across Education’s formula grant programs, including Title I. When fully implemented, this initiative would replace, in whole or in part, other data collections on the implementation of NCLBA’s academic performance requirements, including elements of the Consolidated State Performance Report. Education also administers two grant programs targeted to charter schools, the Charter School Program and the Credit Enhancement for Charter School Facilities Program. These programs provide funding to states, charter schools, and other entities on a competitive basis. The Charter School Program supports the planning, development, and initial implementation of charter schools. The Congress has appropriated over $1 billion to the Charter School Program since 1995. The Credit Enhancement for Charter School Facilities Program helps charter schools obtain school facilities, one of the greatest challenges faced by new charter schools. In total, the Congress has appropriated nearly $90 million for Credit Enhancement for Charter School Facilities since 2001. While Education relies on states for the most part to oversee the implementation of federal grant programs at the individual school level, Education does sponsor research on schools in areas that support the department’s strategic goals, such as improving student achievement. A considerable body of research related to charter school oversight by authorizers has been conducted since 1991 by a number of policy research organizations, as well as Education. In addition, to address the department’s interest in charter schools as an educational reform initiative, Education has sponsored three studies, focusing on the evolution of the charter school movement, the characteristics of charter schools and charter school students, and charter schools’ relationships with authorizers and their communities. The first evaluation, The State of Charter Schools, provides descriptive information about charter schools that were operating in the 1998-99 school year. This study addressed how charter schools have been implemented, under what conditions they have improved student achievement, and their impact on public education. The second evaluation, A Study of Charter School Accountability, by researchers at the University of Washington, examined charter schools’ relationships with authorizers and with their communities. The study included the perspectives of both charter schools and authorizers. The third study, A Decade of Public Charter Schools, evaluated the Public Charter School Program and documented the evolution of the charter school movement. None of these studies looked at the states’ role in ensuring that charter schools are held accountable for meeting their goals. In school year 2002-03, states reported that they provided flexibility through the authorizers they established and through releasing charter schools from traditional public school requirements. Some states provided charter schools flexibility in developing and operating their programs by allowing a degree of choice in selecting the authorizer that oversees operations. Twenty-six of 39 states in our survey reported that they allowed an appeal of an authorizer’s decision to deny an application to start a charter school. Nearly all states released charter schools from traditional public school requirements of some type. About half of the 39 states with operating charter schools in school year 2002-03 had established more than one type of authorizer to approve charters and oversee operations. In about a quarter of states, only a state agency—either the state education agency or the state board—could authorize; in another quarter, only LEAs could authorize. Figure 3, which is based on responses of charter school state agency officials to our survey, shows the types of authorizers states with a mix of authorizers have established as well as those in states that allow only one type of authorizer. A state with a mix of authorizers potentially provides charter school founders more opportunity to find support for a wider range of instructional approaches or educational philosophies than might be possible with a single authorizer, such as the local school district. In addition, providing more than one type of authorizer may expedite charter school authorization and insulate the decision to approve a charter school from the local political environment. The range of authorizers in states that offered a mix of authorizers included LEAs that are local school districts, other types of LEAs, state departments or boards of education, public or private colleges or universities, and nonprofit organizations. Two states—Ohio and Minnesota—had established more types of authorizers than other charter school states. In school year 2002-03, Ohio allowed the board of education, local school districts, an educational service center, and a private university to authorize. That same year in Minnesota, the department of education, local school districts and other types of LEAs, public and private universities, and nonprofit organizations served as authorizers. Examples of another type of authorizer can be found in Arizona and the District of Columbia: both created a charter school board independent of local school districts to authorize. In 11 of the 20 states with a mix of authorizers in school year 2002-03, more than one authorizer was available in at least some geographic areas that charter school developers could choose. (See fig. 3.) For example, in the District of Columbia, a charter school developer might petition either of the two established authorizers. Most of the 11 states offering choice emphasized that allowing choice was intended to provide charter school developers with options, and these states allowed choice among all of the state’s authorizers. In a few of the 11 states, charter school developers had choice in some areas of the state but not in other areas. For example, in Milwaukee, Wisconsin, charter school developers could choose an authorizer from several options: the LEA, the City of Milwaukee, the University of Wisconsin or a local technical college. In Racine, charter school developers could choose between the LEA and the University of Wisconsin, but the university was limited to overseeing just one charter school in Racine. However, outside of Milwaukee and Racine, charter school developers did not have a choice of authorizers. In 7 of the 20 states with a mix of authorizers, choice of authorizer was not available. Although LEAs and the state board of education served as authorizers in 6 of these 7 states, LEAs could authorize only within their jurisdiction and the board of education authorized only when a charter school developer appealed an LEA’s denial of an application. Table 7 in appendix II identifies the type and number of authorizers offered by each state with a mix of authorizers and those states that offer charter school developers a choice of authorizers. In 9 states—Alaska, Colorado, Nevada, New Mexico, Oklahoma, Pennsylvania, South Carolina, Virginia, and Wyoming—only LEAs served as authorizers. The number of LEAs that had operating charter schools in these states ranged from 1 in Wyoming and 2 in Oklahoma to 91 in Colorado and Pennsylvania. In 8 of these 9 states, the LEAs authorized only within their jurisdiction, and charter school developers were not allowed to choose an authorizer from a jurisdiction other than the one where they decided to locate their school. Ten states—Arkansas, Connecticut, Hawaii, Kansas, Massachusetts, Mississippi, North Carolina, New Jersey, Rhode Island, and Puerto Rico—allowed only a state agency to act as authorizer. Of the 39 charter school state agencies surveyed, 26 reported that they allowed appeals when authorizers denied applications. States reported taking different approaches to conducting the appeal process. In 17 states, the appeal is made to the state board of education, and in 3 states, the state department of education hears the appeal. In the other states, a charter school review panel or state board, the county office of education or city council, a state or district court, or an independent party hears the appeal. In 1 state, the reviewing body hears the appeal and, if the case has merit, asks the school developers to resubmit the application to the same authorizer. Nearly all of the 39 states in our survey reported that they released charter schools from some traditional public school requirements. All but 2 states offered release in one or more of 30 areas that states identified, from reporting requirements to staffing practices to student discipline. The greatest number of states released charter schools from teacher termination procedures and length of school day (21 states), teacher compensation and benefits (22 states), collective bargaining procedures (22 states), and requirements established for local school boards (23 states). Officials in 6 states reported that the state released charter schools from almost all traditional public school requirements, while a few requirements, such as the minimum number of teachers required, the use of district-approved text books, and graduation requirements, were released in only a few states. Michigan and Puerto Rico reported that they don’t release charter schools from traditional public school requirements but that their charter schools had some features that distinguished them from traditional public schools. Michigan’s state charter school agency representative explained that its charter schools operate as private, nonprofit corporations and that their contracts are individualized and set forth terms that are unique to each charter school. In addition, unlike public schools, Michigan’s charter school boards are appointed, not elected. Puerto Rico’s representative said that the state permits charter schools to introduce additional elements to the curriculum, as long as they meet state curriculum requirements and that the charter school computer equipment policy is different from the policy for traditional public schools. To promote charter school performance and financial integrity, states took various actions to oversee charter schools and provided oversight of and assistance to authorizers. Twenty-eight of 39 states reported that they collected information on the extent to which charter schools achieved the academic goals in their charters, goals that may not be related to NCLBA. A third of the 39 surveyed states reported having primary responsibility for monitoring the financial condition of charter schools. In overseeing authorizers, most states also reported having taken actions to determine whether authorizers were performing their oversight responsibilities or to address authorizer oversight problems. In addition, most of the states supported authorizers by providing them with assistance to help them oversee charter schools. Over half of the states either provided funding to authorizers or allowed authorizers to collect a fee from the charter schools they authorize. All but 5 states reported monitoring the enforcement of NCLBA school improvement requirements for charter schools. Title I schools are designated as in need of improvement if they miss state performance targets for 2 or more years in a row and certain school improvement actions are required. According to our survey, only 6 states reported being responsible for developing school improvement strategies. (See table 1.) Over half of the 39 states in our survey reported having responsibility for enforcing school improvement actions. Table 8 in Appendix II presents the detailed responses to these questions, organized by states’ authorizer structures. Twenty-eight of the 39 surveyed states reported that they also collected information on the extent to which charter schools achieved the academic goals in their charters, not including those goals related to NCLBA. Eight states reported that the state agency had primary responsibility for ensuring the charter goals are achieved, but most of these were states where the state agency is the only authorizer in the state. Table 9 in appendix II shows states’ detailed responses to these questions. States also reported varying roles in promoting financial integrity of charter schools. Thirteen states reported that a state agency was primarily responsible for monitoring the financial condition of charter schools. As shown in table 2, many of the 39 states reported multiple entities with varying degrees of responsibility for financial monitoring. Only 4 states— Arizona, Indiana, Wisconsin, and Puerto Rico—reported that a single entity had financial oversight responsibility. Table 10 and table 11 in appendix II show the complete range of answers given for each state reporting. Most of the 39 states relied on financial audits as one mechanism of financial oversight. Thirty states reported that the state required charter schools to obtain an audit of their financial statements on a regular cycle, usually each year, and most of these states said that state law required this audit. Generally, states reported that independent auditing firms conducted these audits; in 6 states, they were conducted by the state audit organization. Although 30 states reported that audits were required, many of these states did not provide information we requested about audit results for the 2002- 03 school year. Twenty-seven states reported that charter schools received audits in 2002-03, but only 14 states provided audit data. States gave several reasons for the incomplete information. Some state officials said that they collected the annual audit reports but did not compile the audit data for our survey or that their office did not receive copies of the audit reports. Other states reported that charter schools are included in school district audits, but results are not broken out for charter schools. In the 14 states that reported audit data, 360 of the 428 charter schools—84 percent—received an unqualified, or “clean,” opinion. An unqualified opinion means that financial statements present fairly the financial position, results of operations, and cash flows of the entity, in this case the charter school, in conformity with generally accepted accounting principles. Almost 70 charter schools in these states received something other than a clean opinion; most of these schools—59—received a qualified opinion as the result of a problem. For example, one state had several charter schools receive qualified opinions because of insufficient detail in their financial statements. Three schools received a disclaimer of opinion, indicating that the auditor did not express an opinion on the financial statements, and 6 schools received a going concern opinion, indicating that the school could not meet current operating costs without incurring debt or liquidating assets. Over one-third of the 39 surveyed states reported that in school year 2002- 03, at least one charter school was closed involuntarily, for reasons other than the charter holder’s request. One state—Connecticut—reported an involuntary charter school closure for academic reasons in 2002-03, and 7 states—Arizona, Colorado, Florida, Louisiana, New Jersey, Oregon, and Wisconsin—reported involuntary charter school closures for financial reasons. Other reasons given for involuntary closures included leadership and governance problems. Fourteen states, many of which were the same states reporting involuntary closures, also reported that at least one charter school closed voluntarily in 2002-03. Table 12 in appendix II provides more detailed information about charter school closures. Twenty-nine states reported that entities other than a state agency could authorize charter schools. These states reported that they had established a variety of statewide policies and procedures that authorizers and schools must use. Nearly all of the 29 states established procedures for administering standardized tests. Eight states established policies that required or allowed accreditation of charter schools. Over half of the 29 states reported that they prescribed accounting standards for authorizers, and 12 reported that they permitted authorizers to withhold state funds from charter schools. Most of the 29 states also reported having taken actions to determine whether authorizers were performing their oversight responsibilities or to address authorizer oversight problems. As table 3 shows, these actions sometimes involved audits or investigations of authorizers. Not shown in table 3, most of the 29 states provided notification of potential charter school noncompliance with educational or financial requirements. Table 13 in appendix II provides more detailed information about states’ actions to address authorizers not performing their oversight responsibilities. In addition to overseeing authorizers, states supported authorizers by providing them with assistance to help them oversee charter schools. As shown in table 4, nearly all of the 29 states provided assistance in at least one of four forms: state funding, fees for service, training, and technical assistance. Over half of the 29 states either provided funding to authorizers or allowed authorizers to collect a fee from the charter schools they authorize. Most of the 29 states also collected information from authorizers about the charter schools the authorizer oversaw. As shown in table 5, the majority required authorizers to submit schools’ charters and student attendance data to the state. Fewer than half required authorizers to submit performance reports about the schools they authorized, contracts approved by charter schools, and information about policy decisions made by charter holders. Under NCLBA, charter schools are required to meet the same performance requirements as other public schools, but the law permits certain flexibilities where allowed by state law. Charter schools, like other public schools, are subject to the law’s requirements for the assessment of school performance and the implementation of actions required when schools do not meet state performance goals. NCLBA requires that oversight responsibility be performed in accordance with state law. In addition, while NCLBA requires certification for all other teachers to meet the highly qualified teacher requirement, the law exempts charter school teachers from this requirement where state law contains such an exemption. Charter schools, like other public schools, are subject to the single statewide system for assessing school performance required by NCLBA and to the law’s parental notification requirements regarding the school’s performance on these assessments. If charter schools receiving Title I funds do not meet annual performance goals, they must also implement the school improvement actions NCLBA requires. Education’s guidance for charter schools specifies that NCLBA requirements are to be overseen in accordance with state law and that it is state law that determines the entity with responsibility for the performance of charter schools. According to the guidance, this generally means the authorizer. Our survey of states indicated that charter schools were included in statewide assessment systems. All 39 states indicated that charter schools administered the test used for states’ annual performance goals under NCLBA in 2002-03. Thirty-three states provided information on their charter schools’ performance in achieving the state performance goals in 2002-03. Of these 33 states, 21 reported that at least half of charter schools in the state achieved annual state performance goals in 2002-03, while 12 states indicated that fewer than half of their charter schools achieved annual performance goals. (See fig. 4.) For example, the percentage of charter schools achieving state performance goals ranged from 100 percent in Utah to 8 percent in Missouri. Table 14 in appendix II includes information for each state on charter schools’ achievement of state goals. In addition, the law requires schools receiving funds under Title I of NCLBA, including charter schools, to take certain improvement actions if they repeatedly do not achieve their states’ annual performance goals. As figure 5 shows, 31 states reported that some or all charter schools in their states received Title I funds in 2002-03; therefore, these Title I schools would potentially be subject to NCLBA school improvement actions. In 21 of those states, a majority of charter schools received Title I funds. Table 14 in appendix II provides detailed responses on the percentage of Title I charter schools by state. As mentioned previously in this report, our survey asked state officials which entity in their state had primary responsibility for NCLBA requirements, and consistent with Education’s guidance, states reported a variety of entities assuming these responsibilities. As shown in figure 6, officials most commonly reported that development of school improvement strategies is primarily the responsibility of the charter holder—the entity granted permission to establish the charter school. The enforcement role was most often seen as the responsibility of the state education agency, either the state department or board of education and, less frequently, authorizers. However, in 9 of the 19 states that identified a state agency as being primarily responsible, the state agency was the only authorizer in the state. In addition, while 7 states reported that enforcement is primarily an LEA responsibility, LEAs are the only authorizers in 4 of these states. Among the first of the improvement actions specified for Title I schools are the school choice transfer option and supplemental services. If a Title I school does not meet the state’s annual school performance goals for 2 consecutive years, it must be designated as in need of improvement. Students attending these schools must be given the option to transfer to another school in the district, and the transfer school offered must not be designated in need of improvement under NCLBA. If a school does not meet the target for a third year, students must be offered supplementary educational services, such as tutoring. Officials in 18 of the 31 states with Title I charter schools reported that at least one charter school in their state had been designated in need of improvement in school year 2002- 03—a total of 148 charter schools across the 18 states. Three of the 18 states reported a total of 15 charter schools implementing the school choice transfer option. An additional 5 of the 18 states reported implementing supplemental services rather than choice. Ten states reported that neither choice nor supplemental services was implemented or that they did not know if any school improvement action had been taken. Table 14 in appendix II provides this information for each state. One of the possible reasons that states reported relatively few charter schools implementing choice may be that many charter schools are single- school LEAs. In these cases, Education guidance says that to the extent practicable, arrangements should be made with a neighboring LEA to accept transferring students. If such arrangements cannot be made, supplemental services may be offered as an alternative. Officials in 8 of the 39 states we surveyed reported that charter schools were considered LEAs, and in 15 other states, some charter schools were LEAs. (See fig. 7.) However, when charter schools are parts of LEAs under state law, as reported by 16 states, and there are other eligible schools in the LEA to which students could transfer, LEAs are required to offer transfers. Charter school students who accept transfers under these conditions must be provided transportation to the offered school, even if a state’s charter law does not require that transportation funds be made available for charter schools. Those schools, including charter schools, that meet state performance goals may serve as schools of choice for students transferring under NCLBA or may provide supplemental services to students attending schools that did not meet state goals. Officials in 4 states reported on our survey that at least one charter school in their state received students transferring under the NCLBA school choice provision, and officials in 3 states reported charter schools serving as providers of supplemental services. A few states reported other reasons that school improvement actions, including school choice transfer, were not implemented in charter schools. In 1 state, timing was reported as a possible reason—that is, school improvement actions may have been planned but not yet taken by the end of school year 2002-03. In 2 states, officials said that charter schools are schools of choice and students may transfer at any time. In another state, officials said they believed that in most charter schools needing improvement, parents might have declined to transfer their children—or might not have been offered the transfer option because of a misunderstanding of the law. However, some states did not know what actions were taken in charter schools needing improvement. For example, officials in 3 states, with a total of 78 charter schools needing improvement in 2002-03, were unable to provide information about any school improvement actions that may have been taken in those schools. While most accountability provisions of NCLBA are applied in the same way to charter and traditional schools, the law makes a distinction in several areas. For one thing, NCLBA requirements for highly qualified teachers make an exception in the certification requirement for charter school teachers. In general, to be highly qualified under NCLBA, teachers in core academic subjects must have obtained state teacher certification, hold a bachelor’s degree, and have demonstrated subject matter knowledge. However, the law provides that teachers of core academic subjects in charter schools meet the certification requirement if they meet the requirements set forth in their state’s charter school law regarding certification or licensure. Officials in 13 of the 39 states in our survey reported that their state law exempted charter school teachers from certification requirements. In addition, Education’s NCLBA guidance for charter schools modifies the instructions regarding lotteries to give preference to students seeking to transfer to the charter school under the choice provision of NCLBA. Charter schools receiving funds under Education’s Charter School Program must use a lottery if they have more applicants than can be served by the school. The NCLBA guidance permits such charter schools to weight the lottery to increase the chances of admitting students seeking to change schools under the law’s choice provisions. Other areas where charter schools have had flexibility do not appear to be affected by NCLBA, such as having additional, unique academic goals and using additional assessments to measure progress for those goals. Officials in 30 of the 39 charter school states we surveyed reported that all charter schools included unique academic goals, not related to the state’s annual school performance goals, in their charters. Officials in 4 other states reported that at least some of the charter schools in their states included their own academic goals in their charters. Twenty-three states reported that charter schools in their states use a test, in addition to the standardized test required by the state, for their own assessment purposes. As it does for all public schools, Education plays a role in accountability for charter schools through the resources it provides: it administers grant programs that provide funds to charter schools, including a program designed specifically to encourage the development of charter schools, and sponsors research on charter school accountability. In addition, NCLBA reiterated Education’s additional responsibility for charter schools’ funds. The department must ensure that new and expanding charter schools receive timely payment of federal grant funds for which they are eligible. Education’s OIG has reported problems with the timeliness of receipt of Title I funds by charter schools and recommended that Education more closely monitor this situation. Although Education monitors states’ oversight systems and visits some school districts and schools, the data collected during these site visits can only be used to determine the timeliness of funds disbursed at the locations visited. Therefore, Education has little information to use in ensuring that charter schools receive their federal funds promptly or to know how well the schools perform. Education is in the process of developing new systems that are expected to provide both performance and financial reports for the department’s major grant programs, but the ability of the new systems to provide financial data for charter schools is questionable. Education also sponsors research that provides a better understanding of charter schools. Although charter schools receive funds from a variety of federal programs, Education’s monitoring of these programs provides little information that can help the department fulfill its responsibility under NCLBA to ensure timely payment to charter schools. Table 6 shows selected grant programs from which charter schools commonly receive federal funds. In the case of the larger grant programs, Title I and the Individuals with Disabilities Education Act (IDEA), Education makes grants to the states, which then distribute the money to local education agencies. In such cases, Education monitors state programs, including state systems for monitoring local programs, by reviewing annual performance reports and conducting site visits. As part of its monitoring process, Education visits a selected number of school districts and schools. However, the data collected during these site visits could be used to determine the timeliness only of funds disbursed to the districts and schools visited. These data cannot be used to check timeliness of funds dispersed to all of a state’s charter schools. NCLBA charges both Education and states with ensuring that new and expanding charter schools receive all federal formula grant funds for which they are eligible within 5 months of opening or expanding. However, Education’s OIG reported problems with the timeliness of Title I grant payments to charter schools. In 2003 and 2004, the OIG examined the timeliness of states’ Title I payments to charter schools in Arizona, California, and New York and found delays as long as 13 months in New York and 6 months in Arizona. In reporting these findings, the OIG included suggestions for improving Education’s monitoring of payment timeliness for charter schools. Education has generally accepted the OIG’s recommendations and proposes to take certain steps to improve its monitoring of these payments. Responses to our survey suggested that states varied in their ability to track federal funds flowing to charter schools in their states. For instance, according to our survey, although officials in 36 states reported that they monitored the federal funding that individual charter schools received, just 13 states were able to report the proportion. However, according to officials in at least two of these 13 states—Ohio and Texas—their states have developed the capability to track the flow of state and federal funds to charter schools through their automated financial information systems. The information on funding flows and the timeliness of payments these systems provide can assist state education agencies in Ohio and Texas in ensuring that charter schools receive federal grant funds. For most grant programs, Education’s financial data system does not identify individual schools, nor does it distinguish between charter schools and other schools. Instead, for most programs, funding is provided to, and data are collected on, the grant recipient, which is usually an SEA or LEA. The management information system that Education’s grant managers use to track funds and oversee their programs, GAPS, follows the payment and timing of grants according to the fiscal agent, frequently an SEA or LEA. Only when schools are single-school LEAs can they serve as fiscal agents in some programs. Even then, however, single-school charter school LEAs are not separately identified as charter schools in GAPS. For the programs shown in table 6, GAPS cannot be used to obtain complete information on grant receipt at the charter school level. Furthermore, although there are plans to develop a new management information system to improve grant monitoring, it is unlikely that the new system will have the capability to track Education’s grant funds to the school level. Without the capability in its financial information system to track federal funds to charter schools, Education must rely on states for information to perform its responsibility under NCLBA to ensure prompt payment of federal funds to individual charter schools. According to Education’s OIG, Education’s current monitoring of states does not systematically obtain information about timeliness. While monitoring team members sometimes asked about timely payment on their own initiative, Education’s monitoring procedures for Title I funds did not instruct team members to inquire about timely payment of funds to charter schools. At present, a similar lack of charter school-level information exists for school performance data. Education did not collect information on NCLBA-related annual school performance goal status for any public schools in school year 2002-03, and the performance data Education collected about school improvement status did not allow the department to distinguish charter schools from other public schools without additional analysis and reporting. Education required states to submit a Consolidated State Performance Report by December 2003 on specific aspects of NCLBA implementation for the 2002-03 school year. Of the Consolidated State Performance Report’s three main sections—student performance, schools needing improvement, and school choice and supplemental educational services—none requested separate data on charter schools. Education also administers the Charter School Program, a grant program designed specifically to encourage the development and expansion of charter schools. The Charter School Program obligated about $199 million in grants in fiscal year 2003. Typically, grants are awarded to state agencies, although other entities, including schools, can apply directly if their state does not have an approved application on file. The grant is competitive; that is, applications are ranked and awards are made on the basis of the applicant’s ability to meet program goals. To monitor this program, Education has collected information on how well the program is meeting its goal of developing and expanding charter schools, using the standard performance report used by most Education programs. This report does not require that states receiving Charter School Program grants provide this information in a standardized, uniform way. States provided information about how they are meeting the goal in formats of their own choosing, resulting in data that are not readily aggregated and making determination of overall program success difficult. For example, at least one state reported the number of grant applications received and awarded but did not report on the number of schools actually opened and operated that were funded by those grants, a piece of information critical to assessing the program’s goal of developing and expanding charter schools. In 2003, federal program officials developed a supplementary form that requested more specific indicators of performance; for instance, the form asks specifically for the number of charter schools opened each year. According to an Education official, the information provided on this form has been useful in monitoring the grant, but completing it is voluntary on the part of the states, and not all states choose to provide the information. Having the data from all states on the number of new charter schools started with Charter School Program funds would allow program officials to monitor the program’s goal of encouraging the development of charter schools more precisely than the standard performance report permits. To support Education’s grant management functions and to streamline collection of performance reports across grant programs, Education is in the process of developing the PBDMI. The PBDMI is designed to be a comprehensive system that will integrate information from numerous data sources. The design calls for the system to include both financial and academic performance information. For the academic performance information component, the PBDMI is expected to provide school-level academic performance information for the department’s major grant programs, including information on NCLBA implementation. Education officials also hope to use PBDMI’s financial information to examine the link between federal grant program resources and program results. Moreover, the system is designed to produce reports that break out charter schools, but this capability is contingent upon receiving data from states that identify charter schools. Education officials expect to implement the academic performance information component in spring 2005, and the PBDMI’s financial information component is in an early stage of implementation. For academic performance information, the system’s school-level data categories will include whether schools have achieved annual state school performance goals under NCLBA, whether or not students from the school have transferred to other schools under school choice, and the number of students receiving supplemental services. Reports covering these categories would allow Education to monitor and analyze charter schools’ NCLBA results. In fact, once the PBDMI is fully implemented, according to an Education official, elements of the Consolidated State Performance Report will be replaced, since the new system will allow Education to monitor implementation of NCLBA in all public schools. Although the academic performance information component is fully developed, the extent to which the PBDMI will provide school-level financial information, including information that could be used to track the timeliness of payments, still is unclear. The PBDMI is expected to draw on the new grant tracking and monitoring system that will replace GAPS, but that system is unlikely to have the capability to track Education’s grant funds to the school level, according to Education officials. Consequently, on the basis of our discussions with Education officials, it appears questionable that the PBDMI will be able to track federal funds to schools, either traditional public schools or charter schools, unless other school- level data sources are available, such as information states may be able to provide. However, according to Education officials, states and school districts in some cases may not maintain complete school-level records on federal grant fund disbursements. For instance, Education officials explained that because for some grant programs, federal funds bypass the states and go directly to school districts, states may not record the disbursements. States’ records also are incomplete in some cases because states may redistribute the funds for federal grant programs, such as Title I, and record those disbursements, but have no records of the disbursements made by school districts or schools. Thus, Education’s plan for collecting school-level financial information for the PBDMI is not yet complete. Through a separate process, the PBDMI also will make possible the collection of specialized data, which could provide useful information in understanding the performance of charter schools. Surveys will be used to obtain information on schools or subpopulations of schools that is not reported annually by state agencies. For charter schools, the specialized information could include data such as authorizer type. Researchers then would be able to use the PBDMI’s information on characteristics of charter school and the type of authorizer that oversees each charter school in the design of charter school studies, such as Education’s charter school impact evaluation of the effectiveness of charter schools. Collecting information on the entire charter school population, storing it in the PBDMI, and updating it periodically would avoid the need for researchers to collect basic descriptive information each time a charter school study is conducted. The first survey—which is not currently focused on charter schools—is planned for spring 2005. No surveys of charter schools are planned yet, according to an Education official, although they could be considered in the future. To contribute to understanding of the charter school movement and address the department’s interest in charter schools as an educational reform, Education has sponsored a range of research projects on charter schools. Some findings of these studies provide useful information related to accountability for school performance and financial integrity. Appendix III shows selected charter school research projects sponsored by Education. For example, the department has conducted a series of studies based on national surveys, published as The State of Charter Schools and A Decade of Public Charter Schools. These studies provided useful information on the methods authorizers use to hold charter schools responsible for academic performance and financial integrity, but they did not examine states’ oversight actions. These studies also have produced valuable information on the flexibilities available to and preferred by charter schools. Although they provided descriptive information on authorizers’ oversight methods, these studies were not designed to focus on states’ oversight actions and charter school performance. Thus, the studies did not attempt to associate states’ approaches to flexibility or oversight actions with charter school results. At present, Education is undertaking a major study of charter school performance, the charter school impact study shown in appendix III. This study, Education’s first evaluation designed to determine whether charter schools can make a difference in the academic achievement of their students, will track about 3,000 students in 50 schools in 10 states. It will compare the achievement on standardized tests over 3 years for a cohort of students accepted into the 50 charter schools and students who applied to those schools but were not accepted. This approach should result in a methodologically rigorous study design because acceptance into an oversubscribed charter school is supposed to be based on a lottery. Assignment by a lottery would yield a study group and a comparison group that are created from randomly selected students. In order to identify the conditions under which charter schools are most effective in improving student achievement, the study is likely to include a measure of the flexibility states offer charter schools, according to an Education official. As originally designed, the study did not include plans to examine states’ approaches to oversight or to associate these approaches with school performance. According to department officials, subsequent revisions were made to the study design to allow Education to examine how various aspects of state and authorizer policy may contribute to student achievement. However, the amended evaluation design does not make clear the extent to which states’ accountability practices will be taken into account. In recent years, charter schools have enjoyed widespread support as vehicles with potential to provide parental choice in education and promote innovation and creativity in the nation’s educational system. Flexibility and increased autonomy are thought to be important tools for successful charter schools. However, the many challenges of starting and operating a successful charter school are widely recognized, and these challenges could put charter schools at risk for academic and financial difficulties. Like the department’s role in administering other federal public school programs, Education’s oversight responsibility for charter schools seldom extends to individual schools. However, charter schools, as a group, are of particular interest to policy makers because they hold promise as an education reform. Education’s inability to disaggregate charter schools in its routine monitoring and analysis activities has limited its ability to provide policy makers information on the academic performance of charter schools, although the department’s planned comprehensive data system, PBDMI, is expected to improve this situation soon. However, the department’s plans for the financial component of PBDMI are much less developed. Although it appears that the system has the potential to assist the department in its responsibility to ensure timely grant payments to charter schools, critical questions remain about the capability of states to provide complete information on the timing and disbursements of federal funds. Some states, including Ohio and Texas, have developed financial information systems that include records of school-level disbursements of federal funds. Such improvements enable these states to track the timeliness of federal grant payments to charter schools. In an area where Education has a more direct monitoring role—oversight of the Charter School Program—program officials have recently taken steps to collect data that give better insight into program performance. For instance, program officials have begun to ask for standardized data on program progress, such as the number of charter schools opened. However, reporting the information is voluntary, and not all states choose to provide it. Without assurance that states will provide the requested information, Education is limited in its ability to gauge the Charter School Program’s accomplishments. Finally, as it does for other public schools, Education plays an important role in advancing knowledge about charter schools through research efforts it sponsors, and its planned evaluation of achievement in charter schools will further contribute to this knowledge base. The size of this study—about 3000 students in 50 charter schools—could afford an opportunity for researchers to further examine the relationship between states’ approaches to oversight of charter schools and their academic success. To help the department in carrying out its responsibilities related to monitoring federal funds for charter schools and to provide further information on charter schools as an educational reform, we recommend that the Secretary of Education 1. Support implementation of the PBDMI’s financial performance information component by assisting states in developing automated financial information systems to measure and track the disbursement of funds to the charter school level. 2. Require Charter School Program grantees to include in their annual performance reports standard indicators of program accomplishments, in particular, the number of schools started through use of grant funds. 3. Require that the planned charter school impact evaluation design include an analysis of the effects of accountability practices on charter schools’ performance. We provided a draft of this report to the Department of Education for review and comment. Education’s comments appear in appendix IV. Recommended technical changes have been incorporated in the report as appropriate. Education said that the report provides useful information about the application of NCLBA to charter schools, charter school authorizers, and states’ oversight of charter schools. Education strongly agreed with our recommendation about reporting requirements for the Charter School Program and indicated its intention to implement that recommendation. In the case of our recommendation that the department support the implementation of PBDMI’s financial component by assisting states in making improvements to their financial information systems, Education said that it would look more closely at the systems developed in Ohio and Texas and, if merited, will share this information with other states. Education also will use the review of Ohio and Texas’ information systems to assess the degree of burden states might incur in measuring and tracking financial information to the school level. Regarding our recommendation about examining the effects of accountability practices on school performance, Education agreed to examine expanding the impact evaluation to include a review of authorizers’ oversight and accountability practices. We have withdrawn our recommendation that Education collect information on authorizer type through a PBDMI survey. We think that Education’s plan to examine the strength of the relationship between authorizers and charter schools’ academic performance in the charter school impact evaluation is a good first step. We think using information from a small sample of schools to determine whether authorizer type should be collected for the entire charter school population through a PBDMI survey is a prudent use of resources. We will send copies of this report to the relevant congressional committees and other interested parties. We also will make copies available to others upon request. In addition, the report will be made available at no charge on GAO’s Web site at http://www.gao.gov. Please contact me at (202) 512-7215 if you or your staff have any questions about this report. Other contacts and major contributors are listed in appendix V. This appendix discusses in more detail our methodology for examining states’ oversight of their charter school systems and the role the Department of Education (Education) plays in charter school accountability. The study was framed around four questions: (1) how states allow charter schools flexibility in design and operation, (2) how states promote accountability for school performance and financial integrity in their charter school systems, (3) the implications of the No Child Left Behind Act (NCLBA) for charter schools, and (4) the role Education plays in charter school accountability for school performance and financial integrity. While individual charter schools often are locally initiated, the state legislature determines the basic structure of a state’s charter school system in legislation. Prior research on charter school accountability has focused on the relationship between charter school authorizers and their schools, but few studies examined the flexibility that state laws and agencies grant authorizers and schools or the accountability responsibilities the state exercises. To examine how states allow flexibility and promote accountability, we focused on the approaches all states and other political units with operating charter schools in school year 2002-03 took to managing the accountability relationship with charter school authorizers and schools. We also examined states’ NCLBA implementation practices and the role Education played in supporting charter school accountability through federal funds and research. The population for this survey included the state charter school agencies in the 37 states, the District of Columbia, and Puerto Rico, all of which had operating charter schools in school year 2002-03. We obtained usable data from all 39. We asked that the person most knowledgeable about the state’s charter schools coordinate completion of the data collection instrument and confer with representatives of other offices in the state department of education or other charter school agencies for questions that required more in-depth knowledge of particular areas. To develop survey questions, we conducted interviews with charter school researchers and reviewed existing studies on the relationship between authorizers and their schools. We also consulted with Education’s Charter School Program officials. In addition, we conducted an exploratory site visit to Ohio charter school organizations to develop an understanding of authorizers’ and charter schools’ perspectives on state oversight actions. We also discussed their charter school accountability research with representatives of the state legislative oversight agency and state audit agency. The survey included both a primary data collection instrument and a second data collection instrument used to verify state responses to the primary data collection instrument and to collect additional contextual data. In addition to an internal expert technical review by our survey coordination group, we conducted a two-stage pretest of the primary data collection instrument in 6 states to ensure that the data collection instrument was clear and could be answered accurately in a reasonable amount of time. These states were Arizona, Indiana, Louisiana, New Jersey, Oregon, and Texas. We modified the data collection instrument to incorporate findings from the pretest. On the basis of this work, we decided on a two-pronged approach: a primary self-administered survey and a follow-up telephone survey. The primary data collection instrument, used for the self-administered survey, was an electronic instrument that was sent to the state agencies by e-mail. As intended, state agency respondents completed the data collection instrument electronically and returned the completed instrument to GAO headquarters by e-mail. The second data collection instrument, used for the follow-up survey, was also an electronic document with open-ended questions that collected narrative data. This data collection instrument was administered by telephone by a GAO analyst. The second follow-up survey provided for clarification and verification of responses to the primary data collection instrument to ensure that any variation among charter school state agencies in approaches to flexibility, accountability practices, and organizational structures was in fact due to differences in their approaches. This follow-up interview was conducted with each of the 39 states that returned the primary data collection instrument. Another set of follow-up contacts for data that required inquiries with additional state agency representatives or searches of state agency records also was conducted by e-mail. Our approach also involved two additional data gathering and verification steps. During the follow-up telephone interview, additional information was obtained from all participating states on survey questions concerning actions state agencies took to oversee authorizers and set statewide policies and procedures for charter schools, and a short supplemental instrument was completed. In addition, the 6 states selected to pretest the data collection instrument were asked to complete a short modification instrument that included only those questions that had changed or been added from the pretest version to the final version of the primary data collection instrument. To expedite data preparation, responses on the electronic version of the primary data collection instrument were transferred electronically directly into the survey data file. Changes to responses on the electronic version of the instruments originally submitted by state respondents, arising from (1) the follow-up interviews and (2) the supplemental instruments on accountability practices, were recorded manually on a hard copy of the e- mail submission and entered into the survey data file through programming codes. For the pretest states, integrated electronic versions of the original survey responses on the primary data collection instrument, and responses to the second data collection instrument, the supplemental instrument, and changes collected by the modification instrument were transferred to a new instrument, and a 100 percent verification of this information was completed. As with other states’ responses, the pretest states’ responses on the integrated electronic data collection instrument were then read directly into the survey data file. Thirty-nine states responded to the survey, yielding a 100 percent response rate. In two instances, we adapted our survey data collection strategy to suit local circumstances. In Arizona, although we initially contacted the state education agency, that agency and the independent charter school board agreed that the independent board would take the lead in completing the survey. However, both entities worked together to provide a completed survey. In the District of Columbia, the Board of Education opted to answer the survey only for the charter schools it authorized. To obtain complete information, we asked the District of Columbia’s independent charter school board to complete a separate survey for its charter schools. Where appropriate, we have provided the additional information we obtained from the independent charter school board. The survey relied on state agency officials’ self-reporting of flexibility, accountability, and NCLBA implementation information. To ensure the reliability of the data collected, responses from the follow-up interviews were used to verify survey responses and to make corrections. Changes made to the original survey were verified independently. An independent analyst compared the programming done to incorporate the corrections arising from the follow-up interview with responses in the survey data file. The survey data were analyzed using descriptive statistics and cross- tabulations. In collecting and analyzing the financial statement audit data, we employed the following definitions of audit results: Unqualified opinion. The financial statements present fairly, in all material respects, the financial position, results of operations, and cash flows of the entity in conformity with generally accepted accounting principles. Qualified opinion. Except for the effects of the matter to which the qualification relates, the financial statements present fairly, in all material respects, the financial position, results of operations, and cash flows of the entity in conformity with generally accepted accounting principles. Disclaimer of opinion. A disclaimer of opinion states that the auditor does not express an opinion on the financial statements. Going concern. The entity cannot meet current operating costs without incurring debt or liquidating assets. However, in analyzing the data on states’ financial oversight actions, we did not assess how well the oversight was conducted, that is, whether the required financial audit reports were in fact reviewed by the oversight agencies or whether agency officials had the skills needed to make financial assessments. Because the school performance data reported were incomplete or not comparable with data reported by other states, we excluded data on charter schools’ performance in achieving state performance goals in school year 2002-03 reported by 5 states from our state performance summary: Arkansas. All seven of Arkansas’ charter schools were assessed, but all seven were not rated because the charter schools had not all been in operation for 3 years. Arkansas uses a 3-year model to determine student proficiency. Connecticut. The designated contact disclosed that the state tested charter schools in the fall of 2003 and attributed the results to spring 2003. Thus, the Connecticut charter school performance data did not reflect the same time period as data for other states. Indiana. Because test results for the 2002-03 school year were not yet available, Indiana reported results from the test administered in school year 2001-02. Indiana had no operating charter schools in 2001-02. New York. New York’s charter school performance data were incomplete. The designated contact reported results for 5 of the state’s 38 charter schools. Wisconsin. Wisconsin’s charter school performance data were incomplete. The designated contact reported results for the charter schools authorized by the state agency, a small proportion of the state’s 128 charter schools. To determine Education’s role in charter school flexibility and accountability, we conducted interviews with representatives of the Charter School Program, Title I, the Individuals with Disabilities Education Act (IDEA), selected other formula and discretionary grant programs, the Grant Administration and Payment System (GAPS) financial information system, the Office of Inspector General, and Education’s research offices. We also interviewed charter school experts, including Bryan Hassel, Public Impact; Katrina Bulkley, Rutgers University; Mark Cannon, National Association of Charter School Authorizers; Louann Bierlein-Palmer, Western Michigan University; and Anna Varghese, Center for Education Reform. We identified grant programs for inclusion in our assessment on the basis of program size and nomination by Education officials, and we asked Education officials to identify grant programs in which charter schools were likely to participate. We also reviewed documentation for the grant programs, the Performance-Based Data Management Initiative (PBDMI), GAPS, and the charter school impact study design. We also used findings from the interviews and document review to determine the implications of NCLBA for charter schools. To examine Education’s role in sponsoring charter school research, we reviewed existing studies of charter school accountability for school performance and financial integrity. This review was designed to identify the research sponsored by Education and other policy research organizations that assessed some aspect of charter school accountability. We framed our search for existing research broadly enough to identify studies that focused on the oversight relationship between authorizers and charter schools. We included research sponsored by Education and other organizations to identify the range and quality of research evidence on charter school accountability available to charter school state agencies and authorizers. To identify a potential population of studies, we consulted with knowledgeable Education officials and conducted a search of automated bibliographic data bases for studies of charter schools focused, in whole or in part, on any aspect of charter school accountability for school performance and financial integrity. We gave priority to studies that covered all or multiple states with operating charter schools or had highly salient findings. Appendix III lists the studies identified. Appendix II: Selected Data Tables from Survey education (DOE) education (BOE) agencies (LEAs) Choice is restricted to some locations or circumstances. Did not respond. Law changed April 2003 to further expand choices available. The respondent from Texas chose not to answer this question. However, the respondent acknowledged that when the legislation was passed, the intent was to have both state and LEA authorizers. Employed annual surveys of all charter schools, operating in school years 1995-96 to 1998-99. Employed random sample telephone surveys of charter schools and charter school authorizers Conducted site visits to seven charter schools in six states Analyzed data collected by the National Study of Charter Schools Conducted descriptive statistical analysis of survey data. Qualitative data are used to help explain some of the quantitative findings and provide examples. Students on charter school waiting lists who are accepted will be assigned to the treatment group. Students not selected will be assigned to control groups. No Child Left Behind Act: Education Needs to Provide Additional Technical Assistance and Conduct Implementation Studies for School Choice Provision, GAO-05-007, Washington, D.C.: December 2004. No Child Left Behind Act: Improvements Needed in Education’s Process for Tracking States’ Implementation of Key Provisions, GAO-04-734, Washington, D.C.: September 2004. Special Education: Additional Assistance and Better Coordination Needed among Education Offices to Help States Meet the NCLBA Teacher Requirements, GAO-04-659, Washington, D.C., July 2004. Charter Schools: New Charter Schools across the Country and in the District of Columbia Face Similar Start-Up Challenges, GAO-03-899, Washington, D.C.: September 2003. No Child Left Behind Act: More Information Would Help States Determine Which Teachers Are Highly Qualified, GAO-03-631, Washington, D.C.: July 17, 2003. Title I: Characteristics of Tests Will Influence Expenses; Information Sharing May Help States Realize Efficiencies, GAO-03-389, Washington, D.C.: May 8, 2003. Title I: Education Needs to Monitor States’ Scoring of Assessments, GAO-02-393, Washington, D.C.: April 2002. Title I Funding: Poor Children Benefit though Funding Per Poor Child Differs, GAO-02-242, Washington, D.C.: January 2002. School Vouchers: Publicly Funded Programs in Cleveland and Milwaukee, GAO-01-914, August 31, 2001. Charter Schools: Limited Access to Facility Financing, GAO/HEHS-00-163, Washington, D.C.: September 2000. Title I Program: Stronger Accountability Needed for Performance of Disadvantaged Students, GAO/HEHS-00-89, Washington, D.C.: June 2000. Charter Schools: Federal Funding Available but Barriers Exist, GAO/HEHS-98-84, Washington, D.C.: April 1998. Charter Schools: Issues Affecting Access to Federal Funds, GAO/T-HEHS-97-216, Washington, D.C.: September 16, 1997.

Charter schools are public schools that are granted increased autonomy by states in exchange for meeting specified academic goals. State law determines who approves the formation of a charter school, often the board of education. As public schools, charter schools are subject to the performance requirements of the No Child Left Behind Act (NCLBA) as well. In this environment, states' systems for allowing charter schools flexibility and ensuring school performance and financial integrity assume greater importance. GAO examined (1) how states allow charter schools flexibility, (2) how states promote accountability for school performance and financial integrity for charter schools, (3) the implications of NCLBA for charter schools, and (4) the role the Department of Education (Education) plays in charter school accountability. GAO surveyed the 39 states and jurisdictions with operating charter schools in 2002-03 and interviewed charter school experts and Education officials. In school year 2002-03, some states reported that they provided charter schools flexibility by allowing them to choose their authorizer. Authorizers--state education agencies, local education agencies, universities, and other nonprofit organizations--oversee the formation and operation of charter schools. Also, nearly all states provided flexibility by releasing charter schools from some traditional public school requirements, such as teacher hiring and termination practices, schedules, and collective bargaining agreements. To promote charter school performance and financial integrity, states reported that they took action to oversee charter schools and to oversee and provide assistance to authorizers. About half of the 39 states reported having primary responsibility for enforcing school improvement actions in charter schools not achieving annual school performance goals under NCLBA. Most states reported that they intervened when authorizers were not performing their responsibilities and conducted or required audits of authorizers' finances. About half of the states assisted authorizers with funding for their charter school oversight responsibilities or gave them fee collection authority. NCLBA requires charter schools to meet the same requirements as other public schools, but the law permits certain flexibilities where allowed by state law. Charter schools must be included in the statewide assessment system, and charter schools that receive NCLBA Title I funds must take school improvement actions if they do not meet state performance goals. However, NCLBA allows state law to determine the entity responsible for charter school oversight. In addition, while NCLBA requires certification for all other teachers to meet the highly qualified teacher requirement, the law exempts charter school teachers from this requirement where state law permits. As it does for all public schools, Education administers grant programs that provide funds to charter schools, monitors grant performance, and sponsors research on accountability for academic performance and financial integrity. Under NCLBA, the department and states must ensure that new and expanding charter schools receive timely payment of federal grant funds for which they are eligible and meet the act's academic achievement goals. However, in its monitoring and data collection, Education gathers little information on the timeliness of charter school grant payments or how well the schools perform. Moreover, Education's Office of the Inspector General (OIG) reported delays in states' Title I payments to charter schools. Education is in the process of developing new systems that are expected to provide both academic performance and financial reports for the department's major grant programs, but the ability of the new systems to provide financial reports for charter schools is uncertain.

The Medicare statute divides benefits into two parts: (1) “hospital insurance,” or part A, which covers inpatient hospital, skilled nursing facility, hospice, and certain home health care services, and (2) “supplementary medical insurance,” or part B, which covers physician and outpatient hospital services, diagnostic tests, and ambulance and other medical services and supplies. Part B can also cover home health services under certain conditions. In 1996, Medicare paid approximately $18 billion for both part A and part B home health services. By fiscal year 1998, Medicare’s home health spending is estimated to total nearly $22 billion, representing a 700-percent increase from 1989 when spending was $2.7 billion. During this period, coverage requirements changed so that more beneficiaries qualified for home health services. In addition, advances in medical technologies and changes in practice patterns resulted in more beneficiaries needing these services. The number of home health agencies certified to care for Medicare beneficiaries has also grown rapidly since 1989—from 5,700 to more than 10,000 in September 1997. part-time or intermittent basis. Required medical supplies are also covered. Services must be furnished under a plan of care prescribed and periodically reviewed by a physician. As long as the care is reasonable and necessary, there are no limits on the number of visits or length of coverage. Medicare does not require copayments or deductibles for home health care except for durable medical equipment. HCFA, the agency within HHS responsible for administering Medicare, uses six regional claims processing contractors (which are insurance companies) to process and pay home health claims. These contractors—called regional home health intermediaries (RHHI)—process the claims submitted by the 10,000-strong home health agencies, which are paid on the basis of the costs they incur up to predetermined cost limits. RHHIs are responsible for ensuring that Medicare does not pay home health claims when beneficiaries do not meet the Medicare home health criteria, when services claimed are not reasonable or necessary, or when the volume of services exceeds the level called for in an approved plan of treatment. They carry out these responsibilities through medical reviews of claims. HHS’ Office of the Inspector General has emphasized the importance of medical reviews. In the Office’s sampling of claims—which included not just home health but all Medicare services—it found that 99 percent of the improper payments the Office identified appeared to be correct on the surface and were detected only through medical record reviews. their reviews on providers that have unexplained utilization patterns. A similar kind of analysis led our Office of Special Investigations to identify the case being discussed today. Since Medicare’s inception, the home health benefit has undergone several changes in which coverage criteria and their enforcement have alternately tightened and relaxed. The net effect of the changes was that home health care became available to more beneficiaries, for less acute conditions, and for longer periods of time. The benefit was legislatively liberalized in 1980 when limits on the number of services and cost-sharing requirements were eliminated. When prospective payment for hospital services was initiated in 1983, the use of home health services was expected to increase significantly because of incentives for hospitals to discharge patients more quickly. However, HCFA’s relatively stringent interpretation of coverage criteria and emphasis on medical record review kept home health growth in check. Then in 1989, coverage rules relaxed following a court case brought in 1988 that challenged HCFA’s interpretation that individuals had to satisfy both the part-time and intermittent criteria to qualify for the home health benefit (Duggan v. Bowen). HCFA was obliged to revise its coverage guidelines to allow individuals to qualify by satisfying either criterion, which, as we reported in 1996, enabled home health agencies to increase the frequency of home visits. The requirements were also changed so that patients qualified for skilled observation by a nurse or therapist if a reasonable potential for complications or possible need to change treatment existed. The skilled observation, in turn, qualified the beneficiary for home health aide visits. The benefit also allowed maintenance therapy when therapy services were required to simply maintain function; previously, patients had to show improvement from such services to be covered. Medicare Home Health Benefit: Congressional and HCFA Actions Begin to Address Chronic Oversight Weaknesses An individual does not have to be bedridden. . . . the condition of these patients should be such that there exists a normal inability to leave home, and, consequently, leaving their homes would require a considerable and taxing effort. If the patient does in fact leave the home, the patient may nevertheless be considered homebound if the absences from the home are infrequent or for periods of relatively short duration or are attributable to the need to receive medical treatment. In our interviews for the 1996 study, HCFA and intermediary officials said that few denials were made on the basis that the beneficiary was not homebound. In particular, the “infrequent” and “short duration” language qualifying permissible absences from the home would likely result in the reversal of homebound-criterion-based denials at the reconsideration or appeals level. My colleague’s statement on improper activities by Mid-Delta Home Health describes patients whose eligibility on the basis of being homebound was highly questionable. The relationship between the funding levels for payment safeguard activities and the proportion of claims reviewed helps explain the weak oversight of Medicare’s home health benefit in the 1990s. In 1985, legislation more than doubled funding for contractors to conduct claims reviews, enabling intermediaries to review over 60 percent of the home health claims processed in 1986 and 1987. By 1995, however, when payment safeguard funding for medical review of all Medicare-covered part A services had substantially declined (from $61 million in 1989 to $33 million in 1995), RHHIs reviewed about 1 percent of home health claims. As a result of decreased review, agencies were less likely to be caught if they abused the home health benefit. During this period, however, the number of home health agencies participating in Medicare increased by more than a third, and the volume of home health claims processed more than tripled. In January 1998, HCFA announced an increase in the number of claims reviews to about 1.3 percent—still far short of the peak levels of the mid-1980s. efforts, and HCFA has not routinely given guidance on best practices. For example, HCFA has not issued any guidance suggesting that claims for unusually high dollar amounts per beneficiary trigger prepayment reviews. In a recent study of home health claims reviews, we conducted a test of 80 high-dollar claims at one RHHI. The RHHI had initially processed and approved the claims without review but denied them subsequent to our test. The following examples illustrate the importance of careful prepayment review: Of $18,132 in charges for the care of a beneficiary’s decubitus ulcer (open wound) for 30 days, more than a third ($6,483)—including the charges for almost half of the skilled nursing visits (four per day)—were for services not considered medically necessary. Of $4,100 in charges for supplies related to care provided over 4 weeks, 31 percent were denied because they were not adequately documented in the medical records or should have been included as part of the nurse’s visit and not billed separately. About half the amount denied was for supplies never received by the beneficiary. Of $17,953 in charges for medical supplies related to the treatment of a beneficiary’s salivary gland disease, the intermediary denied the entire amount because the medical documentation was not consistent with the itemized list of supplies provided, thus failing to support the claims for supplies the agency billed for. Nine of the 80 claims tested—representing nearly half ($61,250) of the total dollars disapproved—were denied because the home health agency did not submit any of the medical records the intermediary had requested for the review. would have increased before the recent infusion of new payment safeguard funds through HIPAA. Cost-report audits help identify providers’ attempts to shift inappropriate or unnecessary costs to the program. Providers paid under Medicare’s cost-based reimbursement systems—including home health agencies—are reimbursed not on the basis of a fee schedule or the charge for a service but on the basis of the actual cost to provide the service, subject to certain limits. RHHIs reimburse cost providers in several steps, including making periodic interim payments based on the provider’s historical costs and current cost estimates, determining an end of the year tentative settlement based on a report the provider submits that details operating costs and the share related to the provision of Medicare services, and—in relatively few cases—conducting a detailed review (audit) of the cost report to determine the appropriate final settlement amounts. Between 1991 and 1996, the chances, on average, that a provider’s cost report would be subject to an audit fell from about 1 in 6 to about 1 in 13. Much of our statement on Mid-Delta Home Health centers on improperly claimed and reimbursed costs included in cost reports that had not received an in-depth audit until our investigation prompted a closer look. In January 1998, HCFA announced its plans to double the number of comprehensive home health agency audits it performs each year—from about 900 to 1,800. care—required at least every 62 days—are not routinely reviewed by an independent party, such as Medicare’s RHHIs. In our December 1997 report on the home health survey and certification process, we noted that becoming a Medicare-certified home health agency has been too easy, particularly in light of the number of problem agencies identified in various studies in recent years. Until recently, there was little screening of those seeking Medicare certification. We found that the initial survey of an applicant occurred too soon after the agency began operating, offering little assurance that the agency was providing or capable of providing quality care. For example, Medicare certified an agency owned by an individual with no home health experience who turned out to be a convicted drug felon and who later pled guilty with an associate to having defrauded Medicare of over $2.5 million. Rarely were new home health agencies found to fail Medicare’s certification requirements, which call for agencies to (1) be financially solvent, (2) comply with antidiscrimination provisions in title VI of the Civil Rights Act of 1964, and (3) meet Medicare’s conditions of participation. Home health agencies self-certify their solvency, agree to comply with the act, and undergo a very limited survey that few fail. Until less than a year ago, HCFA had been certifying about 100 new home health agencies each month. Once certified, it was unlikely that home health agencies would be terminated from the program or otherwise penalized, even when they had been repeatedly cited for not meeting Medicare’s conditions of participation or for providing substandard care. From September 15, 1997, until January 13, 1998, the Administration placed a moratorium on admitting new agencies into the Medicare program. The moratorium was intended to stop the admission of untrustworthy providers while HCFA strengthened its requirements for entering the program. HCFA used this period of time to develop new surety bond regulations (as mandated by BBA), capital requirements to ensure adequate operating funds, and procedures to better scrutinize the integrity of home health agency applicants. HCFA plans to issue additional provider certification and renewal regulations in the coming months. With the passage of HIPAA and BBA, the Congress recently provided important new resources and tools to fight fraud and abuse in general and home health care offenses in particular. In addition to earmarking funds for anti-fraud-and-abuse activities, the legislation offers specific civil and criminal penalties against health care fraud as well as opportunities to improve detection capabilities. For example, HIPAA makes health care fraud a separate criminal offense and establishes fines and other penalties for federal health care offenses. BBA stiffens the exclusion penalties for individuals convicted of health care fraud. It also establishes civil monetary penalties for such offenses as contracting with an excluded provider, failing to report adverse actions under the new health care data collection program, and violating the antikickback statute. With respect to the home health benefit in particular, BBA targets historical abuses. For example, in an egregious case of home health fraud that our Office of Special Investigations reported on in 1995, the HHS Inspector General charged ABC Home Health Care with billing Medicare for items that were solely for the owner’s or his family’s personal use, including condominium utility expenses, maid services, and automobile lease payments. BBA mandates the elimination of cost-based reimbursement and its replacement by a prospective payment method. Under this method, home health providers will be expected to deliver care for a fixed payment, thus breaking the link in the future between the home health agency’s costs and Medicare’s payments. level of care in favor of those who would be less expensive to treat. The adjuster would not only protect access to care but would also help ensure that Medicare was paying agencies more appropriately. Base-rate development: Because HCFA intends to use historical data on cost of services to calculate a base rate of an episode of care, it must take care to avoid incorporating the inflated costs identified in the cost reports of problem home health agencies. For example, in 1995 we reported on a number of problems with payments by intermediaries for surgical dressing supplies, indicating that excessive costs are being included and not removed from home health agency cost reports. We have suggested at several hearings that HCFA audit thoroughly a projectable sample of home health agency cost reports so that the results could be used to adjust HCFA’s cost database to help ensure that unallowable costs are not included in the base for setting prospective rates. Until October 1999, when the law requires prospective payment for home health services to be implemented, Medicare will continue to reimburse for home health services on a cost basis. Addressing this situation, BBA prohibits Medicare payments for items that have historically been associated with inflated cost reports, such as entertainment, gifts, donations, educational expenses, and the personal use of automobiles. It also tightens per-visit limits and imposes new ones based on historical per-beneficiary costs. Other BBA provisions designed to improve home health oversight include clarifying the terms “part-time” and “intermittent” nursing care; requiring the HHS Secretary to recommend by October 1, 1998, criteria to clarify the term “homebound”; and requiring a $50,000-minimum surety bond from home health agencies. Medicare because of little scrutiny during the certification process. While HIPAA and BBA have given HCFA greater resources and tools to fight fraud and abuse, the home health benefit will continue to require concerted oversight. Mr. Chairman, this concludes my prepared statement. I would be pleased to answer any questions you or the Subcommittee Members may have. Medicare: Improper Activities by Mid-Delta Home Health (GAO/T-OSI-98-6, Mar. 19, 1998; GAO/OSI-98-5, Mar. 12, 1998). Long-Term Care: Baby Boom Generation Presents Financing Challenges (GAO/T-HEHS-98-107, Mar. 9, 1998). Medicare Home Health Agencies: Certification Process Ineffective in Excluding Problem Agencies (GAO/HEHS-98-29, Dec. 16, 1997). Medicare Home Health: Success of Balanced Budget Act Cost Controls Depends on Effective and Timely Implementation (GAO/T-HEHS-98-41, Oct. 29, 1997). Medicare Fraud and Abuse: Summary and Analysis of Reforms in the Health Insurance Portability and Accountability Act of 1996 and the Balanced Budget Act of 1997 (GAO/HEHS-98-18R, Oct. 9, 1997). Medicare: Need to Hold Home Health Agencies More Accountable for Inappropriate Billings (GAO/HEHS-97-108, Jun. 13, 1997). Medicare: Home Health Cost Growth and Administration’s Proposal for Prospective Payment (GAO/T-HEHS-97-92, Mar. 5, 1997). Medicare: Home Health Utilization Expands While Program Controls Deteriorate (GAO/HEHS-96-16, Mar. 27, 1996). Medicare: Excessive Payments for Medical Supplies Continue Despite Improvements (GAO/HEHS-95-171, Aug. 8, 1995). Medicare: Allegations Against ABC Home Health Care (GAO/OSI-95-17, July 19, 1995). The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

GAO discussed Medicare benefit fraud and abuse in the home health industry, focusing on: (1) the general nature of beneficiary eligibility criteria; (2) the diminished Medicare contractor review and audit effort; (3) weaknesses in Medicare's home health provider certification processes; and (4) new tools Congress provided to strengthen oversight of the home health benefit. GAO noted that: (1) several historical factors have produced an environment that, until recently, has enabled improper billing and cost-reporting practices to grow unchecked; (2) legislation and coverage policy changes in response to court decisions in the 1980's made it easier for beneficiaries to obtain home health coverage and harder for Medicare claims reviewers to deny questionable claims; (3) from 1989 until recently, the volume of claims reviews and cost-report audits plummeted, reducing the likelihood that improprieties would be detected; (4) because of the laxity of Medicare's survey and certification process, agencies with no experience or proof of capability were certified as providers; (5) moreover, home health agencies were unlikely to be terminated or penalized even when they were cited repeatedly for providing substandard care or otherwise failed to comply with conditions of participation; (6) recent legislation has enhanced the Health Care Financing Administration's ability to improve its oversight of the home health benefit; (7) in 1995, a multiagency government effort known as Operation Restore Trust launched a new anti-fraud-and-abuse campaign, targeting home health services, among others, for investigation; (8) the following year, the Health Insurance Portability and Accountability Act of 1996 provided dedicated funding to finance, in part, the investigative efforts of the Department of Health and Human Services' Office of the Inspector General and other federal agencies; and (9) a year later, the Balanced Budget Act of 1997 mandated reforming Medicare's method of paying for home health services and contained additional provisions designed to tighten the use and oversight of the home health benefit.

Highway public-private partnerships have the potential to provide numerous benefits to the public sector. There are also potential costs and trade-offs. Highway public-private partnerships created to date have resulted in advantages from the perspective of state and local governments, such as the construction of new infrastructure without using public funding and obtaining funds by extracting value from existing facilities for reinvestment in transportation and other public programs. For example, the state of Indiana received $3.8 billion from leasing the Indiana Toll Road and used those proceeds to fund a 10-year statewide transportation plan. As we reported in 2004, by relying on private-sector sponsorship and investment to build roads rather than financing the construction themselves, states (1) conserve funding from their highway capital improvement programs for other projects, (2) avoid the up-front costs of borrowing needed to bridge the gap until toll collections became sufficient to pay for the cost of building the roads and paying the interest on the borrowed funds, and (3) avoid the legislative or administrative limits that govern the amount of outstanding debt these states are allowed to have. All of these results are advantages for the states. Highway public-private partnerships potentially provide other benefits, including the transfer or sharing of project risks to the private sector. Such risks include those associated with construction costs and schedules and having sufficient levels of traffic and revenues to be financially viable. Various government officials told us that because the private sector more reliably analyzes its costs, revenues, and risks throughout the life cycle of a project and adheres to scheduled toll increases, it is able to accept large amounts of risk at the outset of a project, although the private sector prices all project risks and bases its final bid proposal, in part, on the level of risk involved. In addition, the public sector can potentially benefit from increased efficiencies in operations and life-cycle management, such as increased use of innovative technologies. Highway public-private partnerships can also potentially provide mobility and other benefits to the public sector, through the use of tolling. The highway public-private partnerships we reviewed all involved toll roads. These benefits include better pricing of infrastructure to reflect the true costs of operating and maintaining the facility and thus improved condition and performance of public infrastructure, as well as the potential for more cost effective investment decisions by private investors. In addition, through congestion pricing, tolls can be set to vary during congested periods to maintain a predetermined level of service, creating incentives for drivers to consider costs when making their driving decisions, and potentially reducing the demand for roads during peak hours. Although highway public-private partnerships can be used to obtain financing for highway infrastructure without the use of public sector funding, there is no “free money” in highway public-private partnerships. Rather, this funding is a form of privately issued debt that must be repaid. Private concessionaires primarily make a return on their investment by collecting toll revenues. Though concession agreements can limit the extent to which a concessionaire can raise tolls, it is likely that tolls will increase on a privately operated highway to a greater extent than they would on a publicly run toll road. Tolls are generally set in accordance with concession agreements and, in contrast to public-sector practices, allowable toll increases can be frequent and automatic. The public sector may lose control over its ability to influence toll rates, and there is also the risk of tolls being set that exceed the costs of the facility, including a reasonable rate of return if, for example, a private concessionaire gains market power because of the lack of viable travel alternatives. In addition, highway public-private partnerships also potentially require additional costs to the public sector compared with traditional public procurement, including the costs associated with (1) required financial and legal advisors, and (2) private-sector financing compared with public-sector financing. In addition to potentially higher tolls, the public sector may give up more than it receives in a concession payment in using a highway public-private partnership with a focus on extracting value from an existing facility. In exchange for an up-front concession payment, the public sector gives up control over a future stream of toll revenues over an extended period of time, such as 75 or 99 years. It is possible that the net present value of the future stream of toll revenues (less operating and capital costs) given up can be much larger than the concession payment received. Concession payments could potentially be less than they could or should be. Conversely, because the private sector takes on substantial risks, the opposite could also be true—that is, the public sector might gain more than it gives up. Using a highway public-private partnership to extract value from an existing facility also raises issues about the use of those proceeds and whether future users might potentially pay higher tolls to support current benefits. In some instances, up-front payments have been used for immediate needs, and it remains to be seen whether these uses provide long-term benefits to future generations who will potentially be paying progressively higher toll rates to the private sector throughout the length of a concession agreement. Both Chicago and Indiana used their lease fees, in part, to fund immediate financial needs. Both also established long-term reserves from the lease proceeds. Conversely, proceeds from the lease of Highway 407 ETR in Toronto, Canada, went into the province’s general revenue fund. Trade-offs from the public perspective can also be financial, as highway public-private partnerships have implications for federal tax policy. Private firms generally do not realize profits in the first 10 to 15 years of a concession agreement. However, the private sector receives benefits from highway public-private partnerships over the term of a concession in the form of a return on its investment. Private-sector investors generally finance large public-sector benefits early in a concession period, including up-front payments for leases of existing projects or capital outlays for the construction of new, large-scale transportation projects. In return, the private sector expects to recover any and all up-front costs, as well as ongoing maintenance and operation costs, and generate a return on investment. Furthermore, any cost savings or operational efficiencies the private sector can generate, such as introducing electronic tolling, improving maintenance practices, or increasing customer satisfaction in other ways, can further boost the return on investment through increased traffic flow and increased toll revenue. Unlike public toll authorities, private-sector firms pay federal income tax. Current tax law allows private sector firms to deduct depreciation on assets involved with highway public-private partnerships for which they have “effective ownership.” Effective ownership of assets requires, among other things, that the length of a concession agreement be equal to or greater than the useful economic life of the asset. According to financial and legal experts, including those who were involved in the lease of the Chicago Skyway in Chicago, Illinois, and the Indiana Toll Road, the useful economic life of those facilities was lengthy. The requirement to demonstrate effective asset ownership thus required lengthy partnership concession periods and contributed to the 99-year and 75-year concession terms for the Chicago Skyway and Indiana Toll Road, respectively. These financial and legal experts told us that as effective owners, the private investors can claim full tax deductions for asset depreciation within the first 15 years of the lease agreements. Determining the extent of depreciation deductions associated with highway public-private partnerships, and the extent of foregone revenue to the federal government, if any, from these deductions is difficult to determine because they depend on such factors as taxable income, total deductions, and marginal tax rates of private-sector entities involved with highway public-private partnerships. Financial experts told us that in the absence of the depreciation benefit, the concession payments to Chicago and Indiana would likely have been less than the $1.8 billion and $3.8 billion paid, respectively. However, foregone revenue to the federal government from tax benefits associated with transportation projects can potentially amount to millions of dollars. For example, as we reported in 2004, foregone tax revenue when the private-sector used tax-exempt bonds to finance three projects with private sector involvement—the Pocahontas Parkway, Southern Connector, and Las Vegas Monorail—were between $25 million and $35 million. The public interest in highway public-private partnerships can and has been considered and protected in many ways. State and local officials in the U.S. projects we reviewed heavily relied on concession terms. Most often, these terms were focused on, among other things, ensuring performance of the asset, dealing with financial issues, and maintaining the public sector’s accountability and flexibility. Included in the protections we found in agreements we reviewed were: Operating and maintenance standards: These standards are put in place to ensure that the performance of the asset is upheld to high safety, maintenance, and operational standards and can be expanded when necessary. For example, based on documents we reviewed, the standards on the Indiana Toll Road require the concessionaire to maintain the road’s condition, utility, and level of safety including a wide range of roadway issues, such as signage, use of safety features such as barrier walls, snow and ice removal, and the level of pavement smoothness that must be maintained. Expansion trigger requirements: These triggers require that a concessionaire expand a facility once congestion reaches a certain level. Some agreements can be based on forecasts. For example, on the Indiana Toll Road, when service is forecasted to fall below certain levels within 7 years, the concessionaire must act to improve service, such as by adding additional capacity at its own cost. Revenue-sharing mechanisms: These mechanisms require a concessionaire to share some level of revenues with the public sector. For example, on one Texas project, if the annual return on investment of the private concessionaire is at or below 11 percent, then the state could share in 5 percent of all revenues. If it is over 15 percent, the state could receive as much as 50 percent of the net revenues. While these protections are important, governments in other countries, including Australia and the United Kingdom, have developed systematic approaches to identifying and evaluating public interest before agreements are entered into, including the use of public interest criteria, as well as assessment tools, and require their use when considering private investments in public infrastructure. These tools include the use of qualitative public interest tests and criteria to consider when entering into public-private partnerships. For example, a state government in Australia uses a public interest test to determine how the public interest would be affected in eight specific areas, including whether the views and rights of affected communities have been heard and protected and whether the process is sufficiently transparent. These tools also include quantitative tests such as Value for Money and public sector comparators, which are used to evaluate if entering into a project as a public-private partnership is the best procurement option available. While similar tools have been used to some extent in the United States, their use has been more limited. For example, Oregon hired a consultant to develop public-sector comparators to compare the estimated costs of a proposed highway public-private partnership with a model of the public sector’s undertaking the project. According to the Innovative Partnerships Project Director in the Oregon DOT, the results of this model were used to determine that the added costs of undertaking the project as a public- private partnership (given the need for a return on investment by the private investors) were not justifiable given the limited value of risk transfer in the project. While this study was conducted before the project was put out for official concession, it was prepared after substantial early development work was done by private partners. Neither Chicago nor Indiana had developed public interest tests or other tools prior to the leasing of the Chicago Skyway or the Indiana Toll Road. Using up-front public interest analysis tools can assist public agencies in determining the expected benefits and costs of a project and an appropriate means to undertake the project. Not using such tools may lead to certain aspects of protecting public interest being overlooked. For example, concerns by local and regional governments in Texas helped drive statewide legislation requiring the state to involve local and regional governments to a greater extent in future highway public-private partnerships. Elsewhere, in Toronto, Canada, the lack of a transparency about the toll rate structure and misunderstanding about the toll structure of the Highway 407 ETR facility was a major factor in significant opposition to the project. Direct federal involvement in highway public-private partnerships has generally been limited to projects in which federal requirements must be followed because federal funds have or will be used. At the time of our February 2008 report, minimal federal funding has been used in highway public-private partnerships. While direct federal involvement has been limited, the administration and the DOT have actively promoted highway public-private partnerships through policies and practices, including the development of experimental programs that waive certain federal regulations and encourage private investment. For example, until August 2007, federal regulations did not allow private contractors to be involved in highway contracts with a state department of transportation until after the federally mandated environmental review process had been completed. Texas applied for a waiver to allow its private contractor to start drafting a comprehensive development plan to guide decisions about the future of the corridor before its federal environmental review was complete. These flexibilities were pivotal to allowing highway public- private partnership arrangements in both Texas and Oregon to go forward while remaining eligible for federal funds. The Federal Highway Administration (FHWA) and DOT also promoted highway public-private partnerships by developing publications to educate state transportation officials about highway public-private partnerships and to promote their use, drafting model legislation for states to consider to enable highway public-private partnerships in their states, creating a public-private partnership Internet Web site, and making tolling a key component of DOT’s congestion mitigation initiatives. Recent highway public-private partnerships have involved sizable investments of funds and significant facilities and could pose national public interest implications such as interstate commerce that may transcend whether there is direct federal investment in a project. For example, both the Chicago Skyway and the Indiana Toll Road are part of the Interstate Highway System; the Indiana Toll Road is part of the most direct highway route between Chicago and New York City and, according to one study, over 60 percent of its traffic is interstate in nature. However, federal officials had little involvement in reviewing the terms of either of these concession agreements before they were signed. In the case of Indiana, FHWA played no role in reviewing either the lease or national public interests associated with leasing the highway, nor did it require the state of Indiana to review these interests. Texas envisions constructing new international border crossings and freight corridors using highway public-private partnerships, which may greatly facilitate North American Free Trade Agreement-related truck traffic to other states. However, no federal funding had been expended in the development of the project. Given the minimal federal funding in highway public-private partnerships to date, few mechanisms exist to consider potential national public interests in them. For example, FHWA officials told us that no federal definition of public interest or federal guidance on identifying and evaluating public interest exists. The absence of a clear identification and furtherance of national public interests in the national transportation system is not unique to highway public-private partnerships. We have called for a fundamental reexamination of the nations surface transportation policies, including creating well-defined goals based on identified areas of national interest, incorporating performance and accountability into funding decisions, and more clearly defining the role of the federal government as well as the roles of state and local governments, regional entities, and the private sector. Such a reexamination provides an opportunity to identify emerging national public interests (including tax considerations), the role of the highway public-private partnerships in supporting and furthering those national interests, and how best to identify and protect national public interests in future public-private partnerships. Highway public-private partnerships show promise as a viable alternative, where appropriate, to help meet growing and costly transportation demands. The public sector can acquire new infrastructure or extract value from existing infrastructure while potentially sharing with the private sector the risks associated with designing, constructing, operating, and maintaining public infrastructure. However, highway public-private partnerships are not a panacea for meeting all transportation system demands, nor are they without potentially substantial costs and risks to the public—both financial and nonfinancial—and trade-offs must be made. Highway public-private partnerships are fairly new in the United States, and, although they are meant to serve the public interest, it is difficult to be confident that these interests are being protected when formal identification and consideration of public and national interests has been lacking, and where limited up-front analysis of public interest issues using established criteria has been conducted. Consideration of highway public- private partnerships could benefit from more consistent, rigorous, systematic, up-front analysis. Benefits are potential benefits—that is, they are not assured and can only be achieved by weighing them against potential costs and trade-offs through careful, comprehensive analysis to determine whether public-private partnerships are appropriate in specific circumstances and, if so, how best to implement them. Despite the need for careful analysis, the approach at the federal level has not been fully balanced, as DOT has done much to promote the benefits, but comparatively little to either assist states and localities weigh potential costs and trade-offs, nor to assess how potentially important national interests might be protected in highway public-private partnerships. We have suggested that Congress consider directing the Secretary of Transportation to develop and submit objective criteria for identifying national public interests in highway public-private partnerships, including any additional legal authority, guidance, or assessment tools that would be appropriately required. We are pleased to note that in a recent testimony before the House, the Secretary indicated a willingness to begin developing such criteria. This is no easy task, however. The recent report by the National Surface Transportation Policy and Revenue Study Commission illustrates the challenges of identifying national public interests as the Policy Commission’s recommendations for future restrictions—including limiting allowable toll increases and requiring concessionaires to share revenues with the public sector—stood in sharp contrast to the dissenting views of three commissioners. We believe any potential federal restrictions on highway public-private partnerships must be carefully crafted to avoid undermining the potential benefits that can be achieved. Reexamining the federal role in transportation provides an opportunity for DOT, we believe, to play a targeted role in ensuring that national interests are considered, as appropriate. Mr. Chairman, this concludes my prepared statement. I would be pleased to respond to any questions that you or other Members of the Subcommittee might have. For further information on this statement, please contact JayEtta Z. Hecker at (202) 512-2834 or heckerj@gao.gov. Individuals making key contributions to this testimony were Steve Cohen (Assistant Director), Bert Japikse, Richard Jorgenson, Carol Henn, Matthew Rosenberg, and James White. This is a work of the U.S. government and is not subject to copyright protection in the United States. This published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

The private sector is increasingly involved in financing and operating highway facilities under long-term concession agreements. In some cases, this involves new facilities; in other cases, firms operate and maintain an existing facility for a period of time in exchange for an up-front payment to the public sector and the right to collect tolls over the term of the agreement. In February 2008 GAO reported on (1) the benefits, costs, and trade-offs of highway public-private partnerships; (2) how public officials have identified and acted to protect the public interest in these arrangements; and (3) the federal role in highway public-private partnerships and potential changes in this role. The Senate Finance Committee asked GAO to testify on this report and to highlight its discussion of tax issues. GAO reviewed the experience of projects in the U.S. (including the Chicago Skyway and Indiana Toll Road agreements), Australia, Canada, and Spain. Highway public-private partnerships provide potential benefits, such as sharing risks with the private sector, more efficient operations and management of facilities and, through the use of tolling, increased mobility and more cost-effective investment decisions. There are also potential costs and trade-offs--there is no "free" money in public-private partnerships and it is likely that tolls on a privately operated highway will increase to a greater extent than they would on a publicly operated toll road. There are also financial trade-offs. Unlike public toll authorities, the private sector pays federal income taxes and can deduct depreciation on assets for which they have effective ownership. The extent of these deductions and the amount of foregone revenue, if any, to the federal government is difficult to determine. Demonstrating effective ownership may require lengthy concession periods and, according to experts involved in the lease of the Chicago Skyway and Indiana Toll Road, contributed to the 99-year and 75-year concession terms on these two facilities, respectively. Experts also told us that in the absence of the depreciation benefit, the concession payments to Chicago and Indiana would likely have been less than $1.8 billion and $3.8 billion, respectively. Highway public-private partnerships in the U.S. that GAO reviewed sought to protect the public interest largely through concession agreement terms prescribing performance and other standards. While these protections are important, governments in other countries, such as Australia, have developed systematic approaches to identifying and evaluating public interest and require their use when considering private investments in public infrastructure. Similar tools have been used to some extent in the United States, but their use has been more limited. Using up-front tools can also assist public agencies in determining the expected benefits and costs of a project and an appropriate means to deliver the project. Not using such tools may lead to certain aspects of protecting the public interest being overlooked. While direct federal involvement has been limited to where federal investment exists and while the DOT has actively promoted them, highway public-private partnerships may pose national public interest implications such as interstate commerce that transcend whether there is direct federal investment in a project. However, given the minimal federal funding in highway public-private partnerships to date, little consideration has been given to potential national public interests in them. GAO has called for a fundamental reexamination of our surface transportation policies, including creating well-defined goals based on identified areas of national interest. This reexamination provides an opportunity to identify emerging national public interests (including tax considerations), the role of the highway public-private partnerships in supporting and furthering those national interests, and how best to identify and protect national public interests in future highway public-private partnerships.

Spending for the elderly’s long-term care was $91 billion, or about $12,000 per disabled elderly person, in 1995, the last year for which data on expenditures from all sources are available. The elderly and their families represent the largest single group of purchasers of long-term care, spending almost $36 billion dollars out of pocket, or almost 40 percent of the total $91 billion expenditures for long-term care. (See table 1 for expenditures and fig. 1 for percentages by funding source.) This spending does not include the substantial unpaid support provided to the elderly by family and friends. Studies have found that about 65 percent of disabled elderly living in the community rely exclusively on unpaid sources for their care. Public funding for long-term care comes primarily from Medicaid, which finances almost one-third of long-term care—$28.5 billion in 1995—and Medicare, which funds one-fourth—$22.7 billion. Long-term care expenditures for the elderly are disproportionately used to purchase nursing home care; about 70 percent of total elderly long-term care expenditures are for nursing homes. The baby boom generation, about 76 million people born between 1946 and 1964, will contribute to rapid growth in the number of elderly individuals who need long-term care and the resources required to pay for it. Forecasts of the exact number who will need such care are uncertain because of differing conclusions about the effect of better health care and lifestyles on the subpopulation that may eventually need long-term care. Nevertheless, the number will be very large even if the most rosy scenario prevails. Today’s elderly make up about 13 percent of the total population. The number of individuals aged 65 and over will make up about 20 percent of the total population in 2030, when the first of the baby boomers will reach their 85th birthday. From 1997 to 2030, individuals 85 and older, the most rapidly growing age group and the group most likely to require long-term care, will more than double—from about 3.9 million to about 8.5 million individuals—and by 2050 will more than double again—to about 18 million individuals. (See fig. 2 for the distribution of the elderly in 1997, 2030, and 2050.) Nearly a quarter of the nation’s elderly population—an estimated 7.3 million in 1994 —require some assistance with either activities of daily living (ADL) or instrumental activities of daily living (IADL), or both. Almost 80 percent of these 7.3 million elderly live at home or in other community-based settings, and about 30 percent of them are severely disabled, requiring assistance with at least three ADLs or needing substantial supervision because of cognitive impairment or other behavioral problems. About 22 percent—or 1.6 million—live in nursing homes. An estimated 1 million individuals live in residential settings that have services available, such as assisted living facilities. Experts agree that population aging will increase the number of disabled elderly needing long-term care over the next several decades, but no consensus exists on the size of that increase. While the sheer number of baby boomers is expected to drive up demand for long-term care services, projections of the number of elderly needing long-term care in the next century vary because of different assumptions about the future prevalence of disability. Predicting the magnitude and composition of the growth in the elderly needing long-term care services is complicated by several factors. Some researchers argue that medical advances have increased life expectancy but have not changed the onset of illness. They predict that declining death rates may actually increase the need for long-term care if more people live to develop age-related disabling conditions or live longer with existing disabilities. Others argue that disability is becoming increasingly compressed into a shorter portion of the lifespan, decreasing the number of years long-term care is needed. Improved treatments or prevention of common disabling conditions among the elderly, such as strokes and arthritis, could lessen long-term care need, independent of death rates. Nonetheless, recent forecasts of the number of disabled baby boomers who will need long-term care have been developed but differ widely, ranging from 2 to 4 times the current number of disabled elderly. How this will translate into the need for long-term care services and actual spending will depend on the public and private resources devoted to purchasing long-term care. How the increased long-term care needs of the baby boom generation will be met or financed is uncertain. The past 2 decades have seen change in the types of long-term care services used by the elderly and in who paid for these services. The change has occurred in large part because of shifts in Medicare and Medicaid coverage as well as private purchases of long-term care. We still are experiencing considerable change, which makes it extremely difficult to project what type of services the baby boomers will need and who will pay for them. Historically, the vast majority of long-term care was supplied in nursing homes or at home by family members and friends. Nursing home care was financed almost equally by residents’ own resources and state Medicaid programs. Over the past 15 years, there has been a substantial increase in the number of people receiving paid services at home and relying less on nursing homes. A major contributor to this trend has been increased use of Medicaid-financed home care following passage of home and community-based waiver provisions in 1981. In addition, since 1989, Medicare expenditures for home care have grown rapidly. Medicaid is the largest public funder of long-term care. Most of Medicaid expenditures are for nursing home care, but in the past 15 years there has been a shift to home care. The result is a significant change in the proportion of people with the need for long-term care who are receiving Medicaid-financed services and in the average cost of those services. State Medicaid programs have, by default, become the major form of insurance for long-term care, but only after individuals have become impoverished by “spending down” their assets. Medicaid long-term care spending for many of the elderly results from Medicaid coverage of people who have become poor as the result of depleting assets to pay for nursing home care, the average costs of which exceed $40,000 per year. In most states, nursing home residents without a spouse cannot have more than $2,000 in countable assets before becoming eligible for Medicaid coverage of their care. About two-thirds of nursing home residents in 1994 relied on Medicaid to help pay for their care. Slightly more than 25 percent of Medicaid nursing home residents were admitted as private pay residents. Both multiple nursing home stays and lengths of stay affect whether a private pay resident spends down to Medicaid eligibility. For example, more than one-half of residents who entered as private pay residents and who have been in the nursing home 3 to 5 years are on Medicaid. Traditionally, states emphasized nursing home care. In attempts to control their long-term costs, states imposed controls on the number of nursing home beds. They required assessment and screening of prospective residents to ensure that Medicaid financed nursing home care for the people who were most disabled. Some states also implemented payment systems to provide these facilities incentives to admit and care for the more disabled and higher cost residents. States limited eligibility for home care out of concern about the potential cost of covering services for the large number of disabled who were cared for by their families at home. However, as part of the Omnibus Budget Reconciliation Act of 1981 (P.L. 97-35), the Congress established the home and community-based service waiver program: section 1915(c) of the Social Security Act gave states the option of applying for Medicaid waivers to fund home and community-based services for people who meet Medicaid eligibility requirements. These waivers gave states the ability to restrict the number and costs of eligible individuals. As states have become more experienced with the waivers and confident of their ability to manage these programs, they have expanded their financing of home and community-based care. All states now have home and community-based waivers, and over 200 waiver programs serve more than 250,000 individuals nationwide. Medicaid expenditures for home and community-based waivers have increased an average of 32.7 percent per year from 1987 to 1996, reaching a level of $5.8 billion in 1996. States have used home and community-based waiver services not just to serve additional people at home, but to reduce reliance on nursing homes. In an earlier report, we found that three states we reviewed had restricted construction of new nursing home beds as they financed more home care services. According to the National Academy for State Health Policy, 27 states provide waiver services in assisted living or board and care facilities. Such settings may provide an alternative to nursing homes for someone whose care needs or family resources make it difficult to stay at home. As they address the challenges identified with providing long-term care, states are expected to increasingly focus on Medicaid-funded care provided in the beneficiary’s home or a community-based setting rather than expanding long-term care in nursing homes. Spending on home care in 1996 increased about 24 percent in comparison to the 3-percent increase in the overall program. According to the National Academy for State Health Policy, seven more states plan to expand home care to community-based residential settings, such as assisted living or board and care facilities. In the last 5 years, a number of states also have created forums to consider the direction and financing of long-term care—the National Conference of State Legislatures reports that at least 23 states have formed task forces or study commissions on this issue. Since 1989, Medicare has become the largest funder of long-term home care, financing $14.3 billion in care—or more than half of the home care purchased for the elderly in 1995. A new home health payment system, mandated by the Balanced Budget Act of 1997, however, may reduce the amount of long-term home care financed by Medicare. Medicare traditionally had focused on acute care and consequently paid very little for long-term care. However, legislative and court decisions and consequent changes in guidelines have essentially transformed the home health benefit from one focused on patients needing short-term care after hospitalization to one that serves chronic, long-term care patients as well.As a result, Medicare, on a de facto basis, has financed an increasing amount of long-term care through its home health care benefit. The increase in Medicare home health care use has been dramatic. Emerging trends in home health use suggest that Medicare is covering long-term care for increasing numbers of beneficiaries, rather than just skilled home health care. Both the number of beneficiaries receiving home health care and the number of visits per user more than doubled from 1989 to 1996. A small but significant proportion of users receive extensive long-term support primarily from home health aides. The share of visits supplied by home health aides increased from about 25 percent of all home health visits in 1988 to almost 50 percent in 1995. At the same time, home health users without a prior hospitalization accounted for about one-third of all users in 1993. Figure 3 shows the growth of Medicare home health care expenditures and highlights major policy changes. Medicare’s role could shift significantly as a result of the Balanced Budget Act. The Balanced Budget Act will change the way that Medicare home health care is reimbursed from a cost-based per-visit payment system to a case-mix-adjusted prospective payment system in 1999. How this system will be designed to reflect differences in home health care needed by individuals with various disabilities and what incentives the system creates will have major implications for the amount of future Medicare funding for long-term care. The baby boomers, in general, are expected to be wealthier in retirement than their parents. Those who are single or less educated, or who do not own homes, however, may not do as well. At the same time that many baby boomers will have greater financial resources, they will have fewer social resources, since this generation has remained single longer and had fewer children. As a result, a smaller proportion of this generation will have a spouse or adult children to provide unpaid caregiving. Geographic dispersion of families and the large percentage of women who work outside the home also may reduce the number of unpaid caregivers available to elderly baby boomers, creating more need for purchased services. While many baby boomers will have more financial resources in retirement than their parents, what might be more important is whether they have insurance. Private long-term care insurance has been seen as a means of reducing the catastrophic financial risk for people needing long-term care, and relieving some of the financing burden currently falling on public programs. Some observers also believe private long-term care insurance could provide individuals greater choice in selecting services to satisfy their long-term care needs. Nevertheless, a very small proportion of the elderly or near-elderly have purchased long-term care insurance during the past 10 years. Concern exists that consumers are not knowledgeable about their risk for needing long-term care and about the limitations on Medicare and Medicaid long-term care coverage, and that this lack of awareness decreases demand for long-term care insurance. Questions also remain about the affordability of policies for the majority of elderly people and the value of the coverage relative to the premiums being charged. Private long-term care insurance is a relatively new product with a growing market. In 1986, approximately 30 insurers were selling long-term care insurance policies of some type, and an estimated 200,000 people had purchased these policies. The Health Insurance Association of America (HIAA) has found that by 1995 125 insurers were offering long-term care insurance policies, and more than 4 million policies had been sold. Many fewer individuals had coverage, since many policies sold did not remain in force as individuals stopped paying premiums or dropped one policy to purchase another. Long-term care insurance financed less than 1 percent of long-term care in 1995. Long-term care insurance is still struggling to gain a greater market share. A recent survey of the elderly and near-elderly found that only about 40 percent believe that they or their family will be responsible for paying for their long-term care. HIAA reports that the industry expects continued growth, however, and that the “tax deductibility” of qualified policies will help accelerate that growth. The affordability of long-term care insurance will have a large impact on its market share. Assessments of the ability of private long-term care insurance to provide coverage to a majority of people who will need long-term care are pessimistic. HIAA reports that in 1995 policies paying $100 a day for nursing home care and $50 a day for home health care averaged annual premiums of $1,881 when purchased at the age of 65 and $5,889 when purchased at the age of 79. Long-term care insurance, then, is most affordable for middle- and upper-income individuals. One recent study estimates that the proportion of elderly who can afford long-term care insurance ranges from 10 to 20 percent. Not only is the cost of long-term care insurance a problem for the elderly and near-elderly, but questions also remain about the value of the coverage relative to the premiums being charged. Individuals who consider and decide against purchasing long-term care insurance indicate skepticism about the policies’ providing adequate coverage. Also, as insurers have better understood their risks and competition has increased, premiums have decreased. Some potential purchasers may defer purchase of long-term care insurance because they expect a “better buy” in the future—that is, improved coverage at less cost. We have reported on a number of problems in the long-term care insurance market—including disclosure standards, inflation protection options, clear and uniform definitions of services, eligibility criteria, grievance procedures, nonforfeiture of benefits, options for upgrading coverage, and sales commission structures that reduce incentives for marketing abuses. By the end of 1996, all 50 states had adopted laws and regulations pertaining to long-term care insurance, and 38 states had adopted at least one-half of the provisions of the 1996 National Association of Insurance Commissioners (NAIC) Long-Term Care Insurance Model Act. The Health Insurance Portability and Accountability Act requires that long-term care insurance policies written after December 1996 meet requirements of NAIC Long-Term Care Insurance Model Act to qualify as tax-deductible. This requirement adds to consumers’ protection. In conclusion, even though we cannot know the exact numbers of the baby boom generation who will require long-term care services, we do know that the aging of the baby boomers will lead to a tremendous increase in the elderly population in the next 3 decades and an even larger increase in the 85-and-over population who are more likely to use long-term care services. Financing these services will be a challenge for the baby boomers, their families, and federal and state governments. Mr. Chairman, this concludes my statement. I would be happy to answer any questions you or Members of the Committee might have at this time. Long-Term Care: Consumer Protection and Quality-of-Care Issues in Assisted Living (GAO/HEHS-97-93, May 15, 1997). Medicare Post-Acute Care: Home Health and Skilled Nursing Facility Cost Growth and Proposals for Prospective Payment (GAO/T-HEHS-97-90, Mar. 4, 1997). Medicare: Home Health Utilization Expands While Program Controls Deteriorate (GAO/HEHS-96-16, Mar. 27, 1996). Long-Term Care: Current Issues and Future Directions (GAO/HEHS-95-109, Apr. 13, 1995). Long-Term Care: Diverse, Growing Population Includes Millions of Americans of All Ages (GAO/HEHS-95-26, Nov. 7, 1994). Medicaid Long-Term Care: Successful State Efforts to Expand Home Services While Limiting Costs (GAO/HEHS-94-167, Aug. 11, 1994). Health Care Reform: Supplemental and Long-Term Care Insurance (GAO/T-HRD-94-58, Nov. 9, 1993). Long-Term Care Insurance: High Percentage of Policyholders Drop Policies (GAO/T-HRD-93-129, Aug. 25, 1993). Long-Term Care Insurance: Risks to Consumers Should Be Reduced (GAO/T-HRD-91-14, Dec. 26, 1991). Long-Term Care Insurance: Consumers Lack Protection in a Developing Market (GAO/T-HRD-92-5, Oct. 24, 1991). The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

Pursuant to a congressional request, GAO discussed the challenges the country will face in financing long-term care for the baby boom generation, focusing on: (1) the current spending for long-term care for the elderly; (2) the increased demand that the baby boom generation will likely create for long-term care; (3) recent shifts in Medicaid and Medicare financing of long-term care; and (4) the potential role of private long-term care insurance in help finance this care. GAO noted that: (1) spending for long-term care for the elderly totalled almost $91 billion in 1995, the most recent year for which expenditures from all sources were available; (2) almost 40 percent of these dollars were paid for by the elderly and their families and almost 60 percent by Medicaid and Medicare; (3) these amounts, however, do not include many hidden costs of long-term care, since an estimated two-thirds of the disabled elderly living in the community rely exclusively on their families and other unpaid sources for their care; (4) according to current estimates by the Congressional Research Service, nearly a quarter of the nation's elderly population--over 7 million elderly people--have some form of disability for which they require assistance, such as help with bathing, dressing, eating, preparing meals, or taking medicine; (5) as the 76-million-strong baby boom generation ages, so too will its demand for long-term care increase; (6) long-range predictions of the magnitude of the baby boomers' long-term care needs, however, vary, with estimates of the disabled elderly ranging from 2 to 4 times the current disabled elderly; (7) estimates of cost are even more imprecise due to the uncertain impact of several important factors, including who will be needing care, the types of care they will need, and who will fund it; (8) Medicaid and Medicare, which currently finance almost two-thirds of long-term care, have undergone significant changes in recent years; (9) while historically the majority of Medicaid long-term care expenditures were for nursing home care, in recent years there has been a shift toward more financing of home and community-based care; (10) at the same time, Medicare, the largest public payer for home-based care, has been paying for care that more and more resembles long-term care; (11) private long-term care insurance, seen as a means of helping reduce the catastrophic financial risk for people needing long-term care and some of the financing burden that falls to public programs, has contributed little to date; (12) it is a relatively new form of insurance with a growing market; and (13) nevertheless, after 10 years, a very small proportion of the elderly or near-elderly have coverage.

In the mid-2000s, the Navy was developing the DDG 1000 Zumwalt class destroyer—a new multimission land-attack ship—and laying the analytical framework to support a new air warfare cruiser acquisition program known as CG(X). The Navy planned to end DDG 51 production with the delivery of DDG 112 in 2011 (which would have completed the 62-ship program), and concentrate instead on DDG 1000—initially intended to be a class of up to 32 ships—and building up to 19 CG(X). However, at a July 31, 2008, hearing before the House Armed Services Committee, Seapower and Expeditionary Forces Subcommittee, the Navy stated that it faces a growing proliferation of ballistic missiles and antiship cruise missiles, requiring greater integrated air and missile defense capability and that the naval land attack capability provided by DDG 1000 had been obviated by improved precision munitions and targeting. Navy officials added that DDG 1000 had performance deficiencies compared to DDG 51, most notably in the areas of ballistic missile defense (BMD), area air defense, and some types of antisubmarine warfare. Most importantly, the Navy stated that at that time DDG 1000 could not carry the Standard Missile (SM) 2, SM-3, or SM-6 and was incapable of conducting BMD, though officials have since told us that DDG 1000 is now capable of carrying the SM-2 missile, and that the Mk 57 Vertical Launching System is expected to be capable of carrying any of the standard missiles. The Navy stated that DDG 51 was a proven ship with a proven combat system, and that the Navy intended on restarting production of DDG 51 to defend against substantial ballistic missile proliferation as a bridge to the deployment of CG(X). The Navy focused on building additional DDG ships, but did not discuss AMDR during this hearing. Following this hearing, the Navy began to initiate plans to truncate the DDG 1000 program and made preparations to restart the DDG 51 program. The DOD Joint Requirements Oversight Council had previously identified simultaneous defense against ballistic missiles and antiship cruise missiles as a capability gap and in 2006 validated that IAMD was an operational requirement not sufficiently addressed by other platforms. At the same time the Navy adopted BMD as a core Navy mission that it would perform in concert with MDA. In September 2009, the Joint Requirements Oversight Council also updated and revalidated IAMD requirements. In order to determine the appropriate type of ship and radar that would best address identified IAMD capability gaps, the Navy conducted an Analysis of Alternatives (AOA) known as the Maritime Air and Missile Defense of Joint Forces (MAMDJF). An AOA is an analytical comparison of the operational effectiveness, suitability, and life-cycle cost of alternative potential solutions to address valid capability needs. According to DOD acquisition guidance, an AOA examines potential material solutions with the goal of identifying the most promising option and is required to support a program’s initiation of the technology development phase at Milestone A. We have previously reported on the importance of a robust AOA as a key element in ensuring a program has a sound, executable business case prior to program initiation. Our work has found that programs that conduct a limited AOA tended to experience poorer outcomes—including cost growth. In 2007, as a result of conclusions identified in the MAMDJF AOA, the Navy determined that it needed a very large radar carried on a larger, newly designed surface combatant to counter the most stressing ballistic and cruise missile threats. Consequently, the MAMDJF AOA served as the AOA for both the CG(X) program and for a new, dual-band radar development effort called AMDR. The Navy initiated development of CG(X) and AMDR—a large radar designed to be scalable, meaning that it could be increased in physical size to allow it to provide increased capability to meet future threats. In January 2009, in response to the Navy’s planned changes to its surface combatant program, the Office of the Under Secretary of Defense for Acquisition, Technology and Logistics released a memorandum stating that the Navy’s plan to buy additional DDG 51 Flight IIA ships would be followed by a procurement of either DDG 1000- or DDG 51- based destroyers. The memorandum stated that this procurement would be referred to as the “Future Surface Combatant” until the appropriate hullform to carry AMDR was selected, and required that a study be conducted to identify this hullform. To meet this requirement, in 2009 the Navy conducted a limited study referred to as the Radar/Hull Study. In the Radar/Hull Study, the Navy examined only the two existing destroyer designs—DDG 51 and DDG 1000—with several different radar concepts to determine which pairing would best address the IAMD capability gap and would be more affordable than CG(X), which Navy officials told us was estimated to cost upwards of $6 billion per ship. A senior review panel—known as a “red team”—also independently assessed the study, its analyses, and alternatives considered and provided a separate report on its findings. Following the conclusion of the Radar/Hull Study, the Navy validated the MAMDJF AOA’s findings that a very large radar carried on a larger, newly designed surface combatant was necessary to counter the most stressing threats, but decided, based on the analysis of the Radar/Hull Study, that the preferred solution to meet the IAMD capability gap would be pairing a smaller AMDR with the familiar DDG 51 hullform and the Aegis combat system—which would be referred to as DDG 51 Flight III. The Navy at the same time also cancelled the CG(X) program, largely as a result of cost considerations. The timing of this analysis and key decision making was compressed, as reflected in figure 1. The Navy now plans to build 9 DDG 51s in an upgraded Flight IIA configuration. Construction of the first restart ship (DDG 113) began at Ingalls Shipbuilding in July 2011, approximately 4 years after construction started on the last DDG 51 at that yard. Though the restart program refers to all 9 restart ships, we focus on DDG 113-115 because these are the first restart ships built at both yards—Ingalls Shipbuilding and Bath Iron Works, the only two shipyards that currently build destroyers—and because contracts for these three ships were recently awarded (DDG 113 in June 2011; DDG 114, 115, and an option for DDG 116 in September 2011). After the first 9 ships, the Navy will then transition to building 22 DDG 51s in the new Flight III configuration including AMDR, starting with construction of the lead Flight III ship (DDG 123) in fiscal year 2016, with an initial operating capability planned for 2023. The Navy is currently reviewing technical considerations and options for Flight III as part of an ongoing flight upgrade study that was initiated in February 2010. The Navy also has a notional Flight IV DDG 51 in its long-range shipbuilding plans. The Radar/Hull Study may not provide a sufficient analytical basis given the magnitude of the Navy’s acquisition decision, including up to 43 destroyers (22 of which will be in the Flight III configuration and 21 in a later Flight IV configuration, and both may require significant ship redesign), a new radar, and major combat system upgrades. The cost of 22 Flight III ships is estimated to range from $58 to $64 billion (in constant 2012 dollars), including research and development and procurement. This study played a central role in determining future Navy surface combatant acquisitions by contributing to a selection of the Navy’s preferred radar, combat system and ship solutions, making it, in essence, an AOA. Namely, the Radar/Hull Study provided analysis of the capability of multiple ship and radar alternatives against a revised IAMD capabilities gap, informing the selection of DDG 51 with AMDR as its preferred ship and radar combination. However, it does not provide an adequate evaluation of combat system and ship characteristics, and does not include key elements that are expected in an AOA that would help support a sound, long-term acquisition program decision. Navy officials who were involved in the Radar/Hull Study told us that the capability of the technology concepts they evaluated was considered a major priority, and that the goal was identifying the most capable solution to meet the IAMD threat in the near-term that was also cost-effective. Within this context, the study team analyzed the capability of the radar variants considered. The Navy determined that a dual-band radar (S- and X-Band radars working together as an integrated unit) was required to effectively perform IAMD. As a result, the study team focused on assessing several different combinations of S- and X-Band radars, as show in table 1. The maximum radar size studied in the Radar/Hull Study was a 14-foot radar, since this was determined to be the largest size of radar that the DDG 51 hull could carry and the largest radar that DDG 1000 could carry without substantial deckhouse modifications. These radars were evaluated first against each other, and then combinations of radars were evaluated and compared with the capability of the current S-Band SPY- 1D(V) radar installed on recent DDG 51 ships. All provided enhanced power over and above that of SPY-1D(V); this difference was quantified as a “SPY+” (in decibels) equating to the increase in target tracking range for a fixed amount of resources over the SPY-1D(V) radar. SPY+15 has a 32 times better signal to noise factor—or intensity of the returning radar signal echoing off a target over the intensity of background noise—than a SPY-1D(V) radar. Radars with additional average power and larger antennas have enhanced sensitivity, and thus better performance in advanced threat environments. The Navy found that the SPY+15 S-Band radars performed better than the SPY+11 S-Band radars, and the Radar/Hull Study’s independent red team described the capability of SPY+15 as marginally adequate. The Navy also found that the AMDR-S performed IAMD better than the VSR+. For the X-Band, the Radar/Hull Study identified that SPY-3 performed better than SPQ-9B. Although the Navy considered capability as a driving factor in its decision making, the Radar/Hull Study did not include a thorough comparative analysis of the capabilities of the two combat system architectures— Aegis on DDG 51 and the Total Ship Computing Environment (TSCE) on DDG 1000—into which the radars would need to be integrated. Other than assessing the BMD capability that Aegis currently possesses and the absence of BMD capability in TSCE, the Navy evaluated Aegis and TSCE by focusing on the amount of new software code that it estimated would be required to integrate the radars and to effectively perform IAMD and the costs and risks involved in this development. Such analysis is important because selection of a combat system essentially determines the ship choice, and the combat system is the interface between the radar and the ship’s weapons. Since TSCE does not currently have an inherent BMD capability, the Navy identified several ways to add this capability using Aegis software and hardware. Similarly, changes were assessed to Aegis to provide it enhanced IAMD capability and the ability to leverage a dual-band radar. Table 2 depicts the combat system modifications that were considered. Though TSCE was intended to be the combat system architecture for CG(X) and thus would have been modified to perform BMD, the Radar/Hull Study states that developing a BMD capability “from scratch” for TSCE was not considered viable enough by the study team to warrant further analysis, particularly because of the investment already made in the Aegis program. The Navy concluded that developing IAMD software and hardware specifically for TSCE would be more expensive and present higher risk.preferred combat system option. Navy officials stated that Aegis had proven some BMD capability and was widely used across the fleet, and that the Navy wanted to leverage the investments it had made over the years in this combat system, especially in its current development of a version that provides a new, limited IAMD capability. Ultimately, the Navy determined that Aegis was its While the Navy’s stated goal for the Radar/Hull Study was to identify the most capable solutions with an additional goal of affordability, the Navy selected Aegis based largely on its assessment of existing BMD capability, development costs and risk, and not on an analysis of other elements of combat system capability. Specifically, beyond the fact that Aegis already has a level of proven BMD capability and TSCE does not, other characteristics of the two combat systems that can contribute to Table 3 summarizes some overall performance were not evaluated. examples of combat system characteristics that could have been evaluated; more characteristics may exist. Since this analysis was not conducted, any impact of these capabilities on IAMD or other missions or how each system compares with each other is unknown. For additional discussion on combat system capabilities, see Classified Annex A which will be made available upon request to those with the appropriate clearance and a validated need to know. Description Ability of the computer system to process data; metrics may include the throughput of data that the system can manage and the speed at which it can complete work (e.g.: time to solution). Offensive and defensive electronic and information operations may be a key component of future Navy missions. A combat system that enables the ship to defend against electronic attacks and possibly conduct electronic attacks of its own could contribute to enhanced capability and performance. A measure of how long the system can operate without incurring failures that may require corrective maintenance actions. Measures that protect and defend information and information systems by ensuring their availability, integrity, authentication, confidentiality, and nonrepudiation. This includes providing for restoration of information systems by incorporating protection, detection, and reaction capabilities. A combat system with robust information assurance capabilities would be less vulnerable to interference in the ship’s electronic network (e.g., viruses, hacking) than other systems. A human-system interface measure of the extent to which a system can be used to achieve specified goals with effectiveness, efficiency, and satisfaction. Level of proprietary software code, which dictates whether or not combat system development efforts can be openly competed. Competing combat system upgrades could lead to reduced costs. The ability of a system to handle an increased workload, either without adding or by adding additional resources. While considering the resident BMD capabilities of Aegis and comparing software development costs and risks are essential to making a decision, without a thorough combat system assessment, the Navy cannot be sure how other combat system characteristics can contribute to overall performance. Because Aegis is carried by DDG 51 and not DDG 1000 ships, selection of Aegis as the preferred combat system essentially determined the preferred hull form. The Radar/Hull Study did not include any significant analysis of the ships themselves beyond comparing the costs to modify the ships to carry the new radar configurations and to procure variants of both types. Several characteristics associated with the ships (such as displacement or available power and cooling) were identified in the study. The ships were evaluated on their ability to meet Navy needs and the impact of these ship characteristics on costs. However, there was no documented comparison or discussion of the benefits or drawbacks associated with any additional capabilities that either ship may bring. Navy officials told us that these characteristics were not weighted or evaluated against one another. Other ship variables that directly relate to ship capability and performance—such as damage tolerance and stealth features that were explicitly designed into DDG 1000—were not discussed in the Radar/Hull Study, even though they were discussed in the MAMDJF AOA. The MAMDJF AOA notes that a stealthy ship is harder for enemy forces to detect and target, thus making it more likely that a stealthy ship would be available to execute its BMD mission. However, senior Navy officials told us that the Radar/Hull Study did not consider the impact of stealth on performance because the study assumed that stealth would not have a significant impact on performance in IAMD scenarios. Navy officials added that any additional benefits provided by DDG 1000 stealth features were not worth the high costs, and that adding larger radars to DDG 1000 would reduce its stealth. However, no modeling or simulation results or analysis were presented to support this conclusion. Table 4 depicts ship characteristics that were evaluated in the MAMDJF AOA that could have been evaluated in the Radar/Hull Study. These characteristics influence performance, and each ship option has strengths and weaknesses that could have been compared to help provide a reasonable basis for selecting a ship. For example, DDG 1000 has enhanced damage survivability and reduced ship signatures, while DDG 51 is capable of longer time-on-station and endurance. The Radar/Hull Study did not include a robust trade-off analysis for the variants studied to support the Navy’s DDG 51 selection decision, which is currently planned to result in an acquisition of 22 modified Flight III DDG 51s and a further 21 modified DDG 51s known as Flight IV. DOD acquisition guidance indicates that a discussion of trade-offs between the attributes of each variant being considered is important in an AOA to support the rationale and cost-effectiveness of acquisition programs. A trade-off analysis usually entails evaluating the impact on cost of increasing the capability desired, essentially answering the question of how much more will it cost to get a greater degree of capability. A trade- off analysis allows decision makers to determine which combination of variables provides the optimal solution for a cost they are willing to pay. For the Radar/Hull Study, the Navy examined 16 different combinations of ship, radar, and combat system options based around DDG 51 and DDG 1000. These variants are depicted in figure 2. The Radar/Hull Study documents full cost data for only 4 of the 16 ship variants; 8 ship variants have no cost data, and 4 others do not have ship procurement and operations and support costs. Instead, the Radar/Hull Study provided full cost data for only the most expensive and least expensive DDG 51 and DDG 1000 variants (high and low), and operations and support costs for these four variants. Higher costs were largely driven by the combat system selected. For example, the high DDG 1000 variant included a 14-foot AMDR coupled with a SPY-3 radar, and the more expensive combat system solution, which comprised replacing the central core of DDG 1000's TSCE combat system with the core of the Aegis combat system. The high DDG 51 variant included a 14-foot AMDR coupled with a SPY-3 radar and the Aegis combat system. The low DDG 1000 variant coupled a 12-foot VSR+ with the SPY-3 radar and a less expensive combat system solution involving replacing only portions of TSCE with portions of Aegis. The low DDG 51 included VSR+ coupled with the SPQ-9B radar and the Aegis combat system. In both the DDG 1000 high and low cases, the combat system solutions would be equally capable; the difference was in the level of effort and costs required to implement the changes. Since only a high and low version of DDG 1000s were priced out, the study did not include a DDG 1000 variant with AMDR and the less complicated TSCE combat system upgrade that may be a less expensive—but equally capable—option. Because this variant was not included in the study, cost data were not provided. This study also presented a brief analysis of operations and support costs; the Navy concluded that it found only negligible differences between the operations and support costs for the DDG 51 and DDG 1000 variants. Previous DDG 1000 cost estimates had indicated 28 percent lower long-term costs than DDG 51. While both ships had increases in these costs, the Navy determined in the Radar/Hull Study that adding additional crew to DDG 1000 to perform BMD-related tasks and increased fuel costs were more significant for that ship, and made the costs essentially equal between the two ships. The costs of the 4 variants that the Radar/Hull Study priced are shown in table 5. Navy officials agreed that they could have developed cost estimates for all 16 of the variants, but stated that there was a time constraint for the study that prohibited further analysis, and that they believed that pricing the high and low options was enough to bound the overall costs for each ship class. Without complete cost data for all variants, the Navy could not conduct a thorough trade-off analysis of the variants that fell between the high and low extremes because the costs of these variants are unknown. DOD acquisition guidance highlights the importance of conducting a trade-off analysis. Conducting a trade-off analysis with costs for all the variants would have established the breakpoints between choices, and identified potential situations where a cheaper, slightly less capable ship or a more expensive but much more capable ship might be a reasonable choice. Figure 3 is a notional depiction of the limitations of missing cost data when conducting a trade-off analysis with only high and low data points. Further, the Navy also did not prioritize what aspects of the radar, combat system, and ship it valued more than others, which could also be used to inform a trade-off analysis. For example, if performance is valued more than cost, choosing a ship variant that has 10 percent more performance than another variant but with a 20 percent increase in cost might be in the Navy’s best interest. Alternatively, if cost was weighted more than performance, the Navy might choose the cheaper and slightly less capable ship as it would be able to get a 20 percent reduction in cost with only a 10 percent reduction in performance. Similarly, the study did not discuss the Navy’s preferences with regard to ship characteristics and the impact that differences in these characteristics might have on a trade-off analysis. For example, Navy officials told us that electrical power was a major concern for future destroyers, but the considerable difference in available power between DDG 51 and DDG 1000 (approximately 8,700 kilowatts for DDG 51 after the addition of a supplemental generator compared to 78,000 kilowatts for DDG 1000 with no additional generators required) was not compared in a trade-off analysis. Finally, the Navy did not assess potential impacts of ship selection on future fleet composition. The MAMDJF AOA found that more capability can be obtained by fewer, more capable ships (meaning those with larger radars) than a greater number of less capable ships (meaning those with smaller radars). This could change the acquisition approach and would result in different program costs as a result if it is found that fewer, more capable ships are more cost-effective than many, less capable ships. Navy officials told us that some of these trade offs were not done in the Radar/Hull Study because they were already studied in the MAMDJF AOA. However, that study, using a different threat environment and ship concepts, eliminated the DDG 51 variant from further consideration as a single ship solution; it also eliminated the DDG 1000 option without a radar larger than the 14-foot design that was considered in the Radar/Hull Study. Consequently, its analysis is not directly comparable or interchangeable with the Radar/Hull Study. When comparing the raw ship data from the Radar/Hull Study, we found that the two ships offer different features worth evaluating. For example, all DDG 1000 variants offer more excess cooling and service life allowance, meaning the ability of the ship to accommodate new technologies over the life of the ship without major, costly overhauls than DDG 51 variants, while DDG 51 variants offer greater endurance and lower procurement costs. Table 6 depicts a simplified presentation of this comparison. As this table shows, these two ships offer different characteristics. Both were deemed capable of carrying AMDR, but without conducting a trade- off analysis of these characteristics, the Navy did not consider their relative merit and the significance, if any, of any differences between the two. Senior Navy officials told us that it is now conducting these types of trade-off analyses; however, these analyses are focused only on assessing various DDG 51 configurations, and were not done to help inform the ship selection decision. A preliminary finding of these new analyses is that the cost of Flight III is estimated to range from $58 billion to $64 billion (in constant 2012 dollars), including research and development and procurement. The Radar/Hull Study assumed a significantly reduced threat environment compared to the earlier MAMDJF AOA and other Navy studies. How the threat is characterized is important because against a reduced threat environment, a less capable radar than what was identified as necessary in the MAMDJF AOA was described by the Radar/Hull Study as marginally adequate. Both the Radar/Hull Study and MAMDJF AOA analyzed the performance of radars in several different classified tactical situations that presented threats of varying levels of complexity. The most stressing situations involved a number of different air and missile threats and a complex timing of events. In the MAMDJF AOA, these tactical situations involved many different types of simultaneous threats and larger radars, and were developed in consultation with the Office of Naval Intelligence—the agency tasked to provide validated threat intelligence to support Navy and joint, Navy-led acquisition programs—as well as MDA. Conversely, the subsequent Radar/Hull Study assumed a significantly reduced threat environment and smaller radar solutions than did the MAMDJF AOA. This study modeled radar performance based on a very limited air and missile threat which are both quantitatively and qualitatively less stressing than the threat environment established in the MAMDJF AOA, in other Navy and DOD threat analyses, and in system guideline documents for AMDR. Also, the Office of Naval Intelligence was not actively engaged in the Radar/Hull Study. The system guideline documents for AMDR that were generated at approximately the same time as the Radar/Hull Study also included significantly more taxing tactical situations than the Radar/Hull Study, and in some cases they are even more stressing than those found in the MAMDJF AOA. The Office of Naval Intelligence also provided input to these AMDR system guidelines. The Navy believes that some of the differences in the threat environment result from the different timeframes for the Radar/Hull Study and the MAMDJF AOA; the MAMDJF AOA states that it is based on a 2024 through 2030 timeframe while the Radar/Hull Study states that it is based on a 2015 through 2020 timeframe. However, Navy officials also told us that the IAMD threats are actually emerging more rapidly than they had assumed in the MAMDJF AOA, which could mean that some of the MAMDJF AOA threats may be present earlier. The Navy does not document why the Radar/Hull Study based its analysis on a reduced threat environment compared to the MAMDJF AOA, since both studies are attempting to identify solutions to the same capabilities gap and set of requirements. Navy officials later told us that the assumption in the Radar/Hull Study was that no single Navy ship would likely have to deal with all the threats in the battlespace, compared to the threat environment in the MAMDJF AOA where more of a single-ship solution was considered. However, other Navy studies developed in a similar timeframe to the Radar/Hull Study describe a larger number of threats than the Radar/Hull Study. Further, while the Navy’s assumption may account for some of the quantitative differences between the Radar/Hull Study and all the other Navy studies we analyzed, it should have no bearing on the qualitative difference in the composition of the threat, since this is a variable that is independent of Navy concepts of operations and is a variable over which the Navy has no influence. According to the Navy and shipbuilders, the changes to the restarted DDG 51’s hull and mechanical systems appear less substantial than previous modifications to earlier DDG 51s. However, due in part to a break in production, an initially noncompetitive environment, and other factors, the restart ships are budgeted to cost more than previous DDG 51 Flight IIA ships. While the shipbuilders’ planned production schedules are generally in line with past shipyard performance, the delivery schedule for the first restart ship (DDG 113) may be challenging because of a significant upgrade in the Aegis combat system, where major software development efforts are under way and a critical component has faced delays. Although the Navy plans to install and test this upgrade on an older DDG 51 (DDG 53) prior to installation on DDG 113, delays in these efforts could pose risks to a timely delivery in support of DDG 113 and ability to mitigate risk. If this occurs, the Navy may need additional time to identify, analyze, and work to resolve problems with the combat system—adding pressure to the schedule for DDG 113. Even if current testing goes as planned, the Navy has not planned for realistic operational testing necessary to ensure that the Aegis upgrades are capable of performing IAMD against multiple ballistic and cruise missile targets. While the restart ships will have some changes to the ship’s design and physical structure, Navy officials told us that they are less substantial than prior modifications, despite changes to a large number of design drawings. The Navy has been building DDG 51s since the late 1980s, and over time the ship design has been modified, including additions such as helicopter hangars, additional missiles, and significant combat system upgrades. As shown in table 7, a large number of design drawing changes are required for the DDG 51 restart program, similar to those implemented as part of previous major upgrades, such as the upgrade from Flight II to Flight IIA (DDGs 79 and higher). While these design changes may not be complex, they affect numerous areas of the ship. According to shipyard officials, most design drawings for the restart ships will have applicability from previous hulls and will not require re-design, but the Navy told us that they currently expect 1175 drawings will be changed, and the design work is still underway. As figure 4 shows, some of the changes will affect the topside of the ship, and include removing some redundant or unneeded equipment from the ship (e.g. the forward kingpost and port anchor) while internal changes largely pertain to upgrading the Aegis combat system with new computer displays and computer cabinets. The Navy has budgeted approximately $17.5 billion for the 10 Flight IIA restart ships. The first three restart ships, beginning with DDG 113, cost 45 percent more than recently delivered DDG 51s. DDG 113 through DDG 115 are currently budgeted to cost a total of $5.8 billion, which is approximately $1.8 billion higher than the last three DDG 51s built. Unlike the previous 24 ships, the restart ships are not part of multiyear ship procurements, which can be more cost-efficient due to economies of scale. The Navy partially attributes the increase in procurement costs to a 4-year gap in production. Construction of the last DDG 51s began in late 2007 and production on DDG 113 began in July 2011. The shipbuilders and the Navy anticipate that additional labor hours will be required to build DDG 113-115 due in part to a loss of experienced workers who will have been laid off or otherwise left the shipyard during the production gap. This attrition—along with changes in equipment and processes associated with the shutdown of the production line—contributes to a loss of learning whereby a less experienced and less efficient workforce requires more time to complete tasks with additional hours spent on rework. While the Navy in part attributes the higher ship costs to the need for additional labor hours to build the ships, it does not associate increases with significant changes in the supplier base. In general, the Navy found the supplier base for ship equipment was primarily intact, with most of the DDG 51 suppliers still in production, which allowed the Navy to get the equipment it needed at prices it considered reasonable. In cases where the suppliers were no longer available, the Navy recompeted some key equipment contracts in order to maximize value and to compensate for some modest changes in its supplier base. The Navy’s initial noncompetitive acquisition strategy also contributed to a higher budgeted cost for the first three restart ships. In response to the truncation of the DDG 1000 program, the Navy and the two shipyards had agreed to allocate the construction of DDG 1000s and the first three DDG 51s (DDG 113-115) between Bath Iron Works and Ingalls Shipbuilding to ensure workload stability between the shipyards. The parties agreed, subject to negotiation of fair and reasonable prices and other conditions, that Bath Iron Works would be responsible for all of the remaining DDG 1000 design and construction work and construction of DDG 115, while Ingalls Shipbuilding would construct DDG 113 and DDG 114. After these first three ships, the Navy intended to competitively award contracts for future surface combatants. The Navy assumed that it would pay a premium for the first three ships because a lack of competition between the two shipyards would drive up costs. Indeed, Navy officials noted that a noncompetitive environment, along with disagreements on the impact of the production gap, were among the reasons that initial bids from the shipbuilders were unreasonably high and in excess of Navy budget estimates. In an effort to generate more competitive pricing, the Navy changed its acquisition strategy in May 2011 to “competitively allocate” DDG 114 and 115. This strategy change allowed the Navy to award contracts to each shipbuilder using a Profit Related to Offers strategy, whereby the shipbuilder that submitted the lowest cost bid for its allocated ship would receive a higher target profit percentage, and the shipbuilder that submitted the lower bid for DDG 116 would be awarded an option for construction of that ship. The Navy believed that through its new strategy it would be able to reduce the costs for DDG 114 and DDG 115, noting its successful use on 30 previous DDG 51 ships since 1996. Additionally, the strategy allowed the Navy to award both DDG 114 and DDG 115 to one shipbuilder in the event that it failed to arrive at a fair and reasonable price with each shipbuilder on its allocated ship. After prolonged negotiations with the shipyards and over a year delay from when the Navy planned to award the DDG 113 contract, the Navy awarded a contract to Ingalls Shipbuilding for DDG 113 in June 2011 and DDG 114 in September 2011, and awarded a contract to Bath Iron Works for DDG 115 in September 2011, with an option to build DDG 116. The Navy expects DDG 113 to be built in 47 months (from the start of construction to delivery), DDG 114 in 41 months, and DDG 115 in 58 months. As show in figure 5, Ingalls Shipbuilding—which is building the two first ships —averages 41 months to build a DDG 51, though in recent years has required more time due in part to after-effects of Hurricane Katrina. Bath Iron Works typically requires an average of closer to 54 months. Navy officials told us that this longer 58 month schedule planned for DDG 115 is due to the shipyard beginning construction earlier than planned in part to maintain stability in the shipyard labor force, while maintaining the delivery date. The schedules, while in line with past performance, are contingent on achieving an optimum build sequence, meaning the most efficient schedule for constructing a ship, including building the ship from the bottom up and installing ship systems before bulkheads have been built and when spaces are still easily accessible. Shipbuilders generate specific dates for when systems need to arrive at the shipyard in order to take advantage of these efficiencies. According to shipyard officials, approximately 10 percent to 12 percent of the suppliers for the restart ships will be new vendors. Some key pieces of equipment—like the main reduction gear, the machinery control system, and the engine controllers —will now be government-furnished equipment, meaning that the Navy will be responsible for ensuring an on-time delivery to the shipyard, not the shipbuilder. For the main reduction gear, the Navy is now contracting with a company that bought the gear production line from the past supplier, and while this supplier builds reduction gears for San Antonio class ships, it does not have experience building DDG 51 main reduction gears. An on-time delivery of this key component is particularly important to the schedule because it is installed early in the lower sections of the ship. A delay in a main reduction gear could result in a suboptimal build sequence as the shipbuilder has to restructure work to leave that space open until the gear arrives. The Defense Contract Management Agency reports production of the first gear ship set is progressing well, and that Navy officials are tracking the schedule closely. A major change for the restart ships is a significant upgrade to the Aegis combat system currently under way. This upgrade, known as Advanced Capability Build 12 (ACB 12), will be retrofitted on some of the current fleet of DDG 51s (starting with DDG 53); following DDG 53, the upgrade will also be installed on the restart ships (starting with DDG 113). The retrofit on DDG 53 will provide the Navy with a risk mitigation opportunity, since any challenges or problems can be identified and resolved prior to installation on DDG 113. The Navy believes this is the most complex Aegis upgrade ever undertaken and will enable the combat system to perform limited IAMD for the first time. This upgrade will also move the Navy towards a more open architecture combat system, meaning that there will be a reduction of proprietary software code and hardware so that more elements can be competitively acquired in the future. To date, Lockheed Martin maintains intellectual property rights over some Aegis components. ACB 12 requires both software and hardware changes, and consists of three related development efforts: (1) development of a multimission signal processor (MMSP), (2) changes to the ballistic missile suite (BMD 5.0), and (3) changes to the Aegis combat system core. While the Navy manages the development of MMSP and ACB 12, MDA manages the development of BMD 5.0. Table 8 describes each of the three efforts. While the Navy has made significant progress in developing the components of ACB 12, MMSP is proving more difficult than estimated and is currently 4 months behind schedule, with $10 million in cost growth realized and an additional $5 million projected. A substantial amount of software integration and testing remains before MMSP can demonstrate full capability and is ready for installation on DDG 53—and later DDG 113. While all of the software has been developed, only 28 percent of the eight software increments have been integrated and tested. The integration phase is typically the most challenging in software development, often requiring more time and specialized facilities and equipment to test software and fix defects. According to the Navy, the contractor underestimated the time and effort required to develop and integrate the MMSP software. In December 2010, MMSP was unable to demonstrate planned functionality for a radar test event due to integration difficulties, and MMSP more recently experienced software problems during radar integration which resulted in schedule delays. In response, the contractor implemented a recovery plan, which included scheduling additional tests and replanning the remaining work to improve system stability. However, the recovery plan compresses the time allocated for integrating MMSP with the rest of the combat system from 10 months to 6 months. In order to meet schedule goals and mitigate software development risk in the nearterm, the contractor also moved some development of MMSP capability to future builds. However, this adds pressure to future development efforts and increases the probability of defects and integration challenges being realized late in the program. The contractor already anticipates a 126 percent increase in the number of software defects that it will have to correct over the next year, indicating the significant level of effort and resources required for the remaining development. According to the program office, the high level of defects projected is due to the complexities of integrating and testing with Aegis. Each defect takes time to identify and correct, so a high level of defects could result in significant additional work and potentially further delays if the contractor cannot resolve the defects as planned. The Navy believes the schedule risk associated with this increase is understood and anticipates no further schedule impacts. However, the Defense Contract Management Agency, which is monitoring the combat system development for the Navy, has characterized the MMSP schedule as high risk. As shown in figure 6 below, the Navy will not test ACB 12’s IAMD capabilities with combined live ballistic and cruise missile tests until after it certifies the combat system. Certification is an assessment of the readiness and safety of ACB 12 for operational use including the ability to perform Aegis ship missions. The Navy and MDA plan to determine future opportunities for additional testing to prove the system. The Navy plans to leverage a first quarter fiscal year 2015 test that MDA does not actually characterize as an IAMD test to demonstrate IAMD capabilities. The Navy initially planned to test the combat system’s IAMD tracking capability during a BMD test event to occur by third quarter fiscal year 2013. The test—tracking and simulated engagement of BMD and air warfare targets—would have provided confidence prior to certification of ACB 12 that the software worked as intended. However, this event was removed from the test schedule The Navy now plans to test tracking and simulated IAMD engagement capability during a BMD test event in third quarter 2014. According to the Navy, this is the earliest opportunity for sea-based testing of the ACB 12 upgrade installed on DDG 53. This event will help demonstrate functionality and confidence in the system, but only allows five months between the test and certification of the system to resolve any problems that may be identified during testing. The Navy and MDA plan on conducting a live ballistic missile exercise in second quarter fiscal year 2014, this will only test the combat system’s BMD capability, not IAMD. Consequently, the Navy will certify that the combat system is mission ready without validating with live ballistic and cruise missile targets that it can perform the IAMD mission. The first IAMD test with live targets is not scheduled until first quarter fiscal year 2015. Delays in MMSP could also lead to concurrence between final software integration and the start of ACB 12 installation on DDG 53. Although the Navy has stated that the contractor is currently on schedule, if the contractor is unable to resolve defects according to plan, Aegis Light-Off (when the combat system is fully powered on for the first time) on DDG 53 could slip or the test period could move closer to the start of installation on DDG 113, which could limit risk mitigation opportunities. Contractor officials told us that they plan to deliver the combat system hardware to the shipyard for installation on DDG 113 in May 2013. While the Navy believes the current schedule allows time for the Navy and contractor to remedy any defects or problems found with ACB 12 before it is scheduled to be installed on DDG 113, we have previously reported that concurrent development contributes to schedule slips and strains resources required to develop, integrate, test, and rework defects, which could encroach into this buffer. Additionally, if DDG 53 is not available when currently planned to begin its upgrade, this process could also be delayed. DDG 53’s upgrade schedule already slipped from May 2012 to September 2012, and any significant shifts could mean further schedule compression, or if it slipped past the start of installation on DDG 113 this new-construction ship could become the ACB 12 test bed, which would increase risk. At present, DOD weapons testers and Navy and MDA officials are unsure to what extent the new IAMD capabilities of Aegis will be fully operationally tested and evaluated. Operational testing involves the employment of a new system in a realistic operational environment to determine the operational effectiveness and suitability of the system. This testing is required to: (1) determine if performance thresholds are met, (2) assess impacts to combat operations, and (3) provide additional information on the system’s operational capability. Since the ACB 12 upgrade of Aegis is central to the combat capabilities of the ship, Navy weapons testers believe that Aegis should have a rigorous operational testing program—similar in scope to what was done for the first DDG 51s—in order to validate that the combat system still functions in all areas. According to DOD officials, there should be a high level of coordination between the Navy and MDA with regard to testing the IAMD capability of ACB 12. However, creation of robust test plans for IAMD is complicated because of the division of responsibility between MDA and the Navy. While IAMD consists of both defense against cruise missile and aircraft threats and BMD, MDA is responsible for funding and testing BMD functionality while the Navy is responsible for funding and testing everything else. According to MDA officials, the Aegis combat system first demonstrated the potential to be used for IAMD during a flight test on April 26, 2007, when Aegis engaged a BMD target and a target simulating a high-performance aircraft, but this test did not use the ACB 12 version of Aegis. be done via modeling and simulation, the Navy still needs sufficient data from flight tests conducted in an operationally relevant environment in order to validate the simulation models with actual performance data. Similarly, MDA told us that model validation requires making comparisons between previous flight test results and the results of the models. Without actual operational tests, the Navy’s IAMD models will lack vital real-world data needed to validate how accurately they model the performance of Aegis. The Navy plans to procure the first of 22 Flight III DDG 51s in 2016 with the new AMDR and plans to achieve Flight III initial operational capability in 2023. Other than AMDR, the Navy has not identified any other technologies for inclusion on Flight III or decided on the size of AMDR. Although the analysis supporting Flight III discusses a 14-foot AMDR, senior Navy officials recently told us that a 12-foot AMDR may also be under consideration. While the Navy is pursuing a thoughtful approach to AMDR development, it faces several significant technical challenges that may be difficult to overcome within the Navy’s current schedule. The red team assessment of an ongoing Navy Flight III technical study found that the introduction of AMDR on DDG 51 leads to significant risks in the ship’s design and a reduced future capacity and could result in design and construction delays and cost growth on the lead ship. Further, the Navy’s choice of DDG 51 as the platform for AMDR limits the overall size of the radar to one that will be unable to meet the Navy’s desired (objective) IAMD capabilities. If the Navy selects a 12-foot AMDR—which may reduce the impacts on the ship and design—it may not be able to meet the requirements for AMDR as currently stated in the Navy’s draft capabilities document. Given the level of complexity and the preliminary Navy cost estimates, the Navy has likely underestimated the cost of Flight III. However, since the DDG 51 program is no longer in the DOD milestone review process, decision makers currently cannot take advantage of knowledge gained through a thorough review of the program typically provided at a milestone. Further, since the Navy is responsible for acquisition oversight of the program, there is no requirement for a DOD-level assessment before making further investments in the program. AMDR represents a new type of radar for the Navy, which the Navy believes will bring a significantly higher degree of capability than is currently available to the fleet. AMDR is to enable a higher degree of IAMD than is possible with the current legacy radars. Further, the Navy believes that through the use of active electronically scanned array radars, AMDR will be able to “look” more places at one time, thus allowing it to identify more targets with better detection sensitivity. It will also allow the radar to view these targets with better resolution. AMDR is conceived to consist of three separate parts: AMDR-S: a 4 faced S-Band radar providing volume search for air and ballistic missile defense; AMDR-X: a 3 faced, 4-foot by 6-foot X-Band radar providing horizon search (as well as other tasks such as periscope and floating mine detection); and Radar suite controller: interface to integrate the two radars and interface with the combat system. Figure 7 depicts a notional employment of AMDR’s two radar bands. Three contractors are under contract to mature and demonstrate the critical AMDR-S radar technology required; the acquisition of the AMDR- X portion is still in the preliminary stage, and the Navy plans to award a contract for it in fiscal year 2012. The Navy recognized the risks inherent in the AMDR-S program early on, and implemented a risk mitigation approach to help develop and mature specific radar technologies that it has identified as being particularly difficult. Additionally, the Navy used an initial AMDR-S concept development phase to gain early contractor involvement in developing different concepts and earlier awareness of potential problems. In September 2010, the Navy awarded three fixed-priced incentive contracts to three contractors for a 2-year technology development phase. All three contractors are developing competing concepts with a goal of maturing and demonstrating S-Band and radar suite controller technology prototypes. In particular, the contractors are required to demonstrate performance and functionality of radar algorithms in a prototype one-fifth the size of the final AMDR-S. The Navy has estimated that AMDR will cost $2.2 billion for research and development activities and $13.2 billion to procure at most 24 radars. At the end of the 2-year phase, the Navy will hold a competition leading to award of an engineering and manufacturing contract to one contractor. As shown in figure 8, AMDR is first scheduled to be delivered to a shipyard in fiscal year 2019 in support of DDG 123—the lead ship of Flight III. AMDR-S relies on several cutting-edge technologies. Three of the most significant of these pertain to digital beamforming, the transmit/receive modules, and the radar/combat system interface. Table 9 highlights these technologies and key challenges. Though the Navy has been pursuing risk mitigation efforts related to some key AMDR technologies, realizing AMDR will require overcoming several significant technical challenges. For example, though the Navy worked with the United Kingdom on a radar development program to demonstrate large radar digital beamforming, including limited live target testing, the technical challenges facing the development of AMDR have not been fully mitigated by these efforts. The joint radar development program used a digital beamforming architecture different than what is intended for AMDR, and the demonstrator was much smaller than what is envisioned for AMDR-S. Further, the Navy’s previous effort also did not demonstrate against BMD targets, which are the most stressful for the radar resources. The Navy told us that the contractors have been successful in their AMDR development efforts to date, and that power and cooling requirements may be less than initially estimated. However, substantial work remains, and failure to achieve any of these technologies may result in AMDR being less effective than envisioned. AMDR development is scheduled for 10 years, compared with 9 years for the DDG 1000’s VSR. Integration with the Aegis combat system may also prove challenging: Aegis currently receives data from only a single band SPY-1D(V) radar, and adding AMDR will require modifying Aegis to receive these data, to accommodate some new capabilities, and to integrate Aegis with the radar suite controller. The Navy has deferred this integration, as it recently decided to eliminate AMDR integration work from its upcoming Aegis upgrade (ACB 16) contract, although Navy officials pointed out that this work could be started later under a separate contract. If the Navy does not fund AMDR integration work in ACB 16, this work may not be under way until the following ACB upgrade, which could be completed in 2020 at the earliest if the Navy remains on the same 4 year upgrade schedule. With an initial operating capability for Flight III planned for 2023, this could leave little margin for addressing any problems in enabling AMDR to communicate with the combat system. DDG 51 is already the densest surface combatant class; density refers to the extent to which ships have equipment, piping, and other hardware tightly packed within the ship spaces. According to a 2005 DOD- sponsored shipbuilding study, the DDG 51 design is about 50 percent more dense and complex than modern international destroyers. High- density ships have spaces that are more difficult to access; this results in added work for the shipbuilder since there is less available space to work efficiently. As a legacy design, the ship’s physical dimensions are already fixed, and it will be challenging for the Navy to incorporate AMDRs’ arrays and supporting equipment into this already dense hullform. Some deckhouse redesign will be necessary to add the additional radar arrays: a current DDG 51 only carries four SPY radar arrays, while Flight III is envisioned to carry four AMDR-S arrays plus three additional AMDR-X band arrays. The deckhouse will need to be redesigned to ensure that these arrays remain flush with the deckhouse structure. Adding a 14-foot AMDR to DDG 51 will also require significant additional power generating and cooling equipment to power and cool the radar. Navy data show that as a result of adding AMDR the ships will require 66 percent more power and 81 percent more cooling capacity than current DDG 51s. If the Navy elects to use a smaller AMDR for Flight III these impacts may be reduced, but the ship would also have a significant reduction in radar performance. The addition of AMDR and the supporting power and cooling equipment will significantly impact the design of Flight III. For example, additional large cooling units—each approximately 8 feet by 6 feet—required to facilitate heat transfer between the radar coolant and the ship’s chilled water system will have to be fit into the design. Similarly, a new electrical architecture may be required to power AMDR, which would result in changes to many electrical and machinery control systems and the addition of a fourth large generator. The red team assessment of the Navy’s ongoing Flight III technical study found that modifying DDG 51 to accommodate these changes will be challenging with serious design complexity. Since Flight III design work is just in the concept phases, it is currently unknown how the additional cooling and power generating units added to support AMDR will be arranged, or any impact they will have on ship spaces and habitability. For example, the Navy is currently considering five possible cooling unit configurations. Of these, one cannot be arranged within the existing spaces, another will be very difficult to arrange, and three of these options will require significant changes to the arrangements of the chilled water systems. Similarly, all of the options the Navy is considering for possible power generation options will require rearrangement and some impact on other spaces, including encroachment on storage and equipment rooms. Navy officials told us that hybrid electric drive is being researched for Flight III, and the Navy has awarded a number of contracts to study concepts. The Navy told us that this technology has the ability to generate an additional 1 megawatt of electricity, and thus could potentially obviate the need for an additional generator to support AMDR. However, adding hybrid electric drive would require additional design changes to accommodate the new motors and supporting equipment. Not only can density complicate design of the ship as equipment needs to be rearranged to fit in new items, but Navy data also show that construction of dense vessels tends to be more costly than construction of vessels with more open space. For example, submarine designs are more complicated to arrange and the vessels are more complicated and costly to build than many surface ships. DDG 1000 was designed in part to have reduced density, which could help lower construction costs. According to a 2005 independent study of U.S. naval shipbuilding, any incremental increase in the complexity of an already complex vessel results in a disproportionate increase in work for the shipbuilder, and concluded that cost, technical and schedule risk, and the probability of cost and schedule overrun all increase with vessel density and complexity. incorporate AMDR is likely to result in higher construction costs and longer construction schedules than on Flight IIA ships. First Marine International Findings for the Global Shipbuilding Industrial Base Benchmarking Study, First Marine International (London: August 2005). thickness. Weight and center of gravity are closely monitored in ship design due to the impact they can have on ship safety and performance. The Navy has required service life allowances (SLA) for weight and center of gravity for ships to allow for future changes to the ships, such as adding equipment and reasonable growth during the ship's service life— based on historical data—without unacceptable compromises to hull strength, reserve buoyancy, and stability (e.g., tolerance against capsizing). Adding new systems or equipment may require mitigating action such as removing weight (e.g., equipment, combat systems) from the ship to provide enough available weight allowance to add desired new systems or equipment. A reduced center of gravity may require mitigation such as adding additional weight in the bottom of the ship to act as ballast, though this could also reduce the available weight allowance. These changes all require redesign which can increase costs, and this design work and related costs can potentially recur over the life of the ship. The Navy is considering a range of design options to deal with adding AMDR and its supporting power and cooling equipment. None of the DDG 51 variants under consideration as part of an ongoing Navy study meet Navy SLA requirements of 10 percent of weight and 1 foot of center of gravity for surface combatants. Figure 9 shows that several variants provide less than half of the required amounts. The Navy has determined that only by completely changing the material of the entire fore and aft deckhouses and the helicopter hangars to aluminum or composite as well as expanding the overall dimensions of the hull (especially the width, or beam) can the full SLA be recovered for a Flight III with a 14-foot AMDR. Though a decision has not yet been made, at this time Navy officials do not believe that a composite or aluminum deckhouse will be used. The Navy also told us that removing combat capability from DDG 51 may be required in an effort to manage weight after adding AMDR, effectively reducing the multimission functionality of the class. Navy officials stated that SLA has not always been required, and that this allowance is included in designs to eventually be consumed. They pointed to other classes of ships that were designed with less than the required SLA margins and that have performed adequately. However, as shown in Table 10, our analysis of the data indicates that these ships have faced SLA-related issues. According to Navy data, delivery weight of DDG 51s has gotten considerably heavier over the course of building the class, with current 51s weighing approximately 700-900 long tons (a measure of ship displacement) more than the first DDG 51s. Further, while the current DDG 51s all can accept both an increase in weight or rise in the center of gravity, the ships are already below the required center of gravity allowances, though Navy officials told us that this could be corrected with ballasting if the Navy opted to fund the change. In commenting on the ongoing Navy study, the independent red team identified reduced SLA as a significant concern for Flight III, and noted that if the Navy does not create a larger hullform for Flight III, any future ship changes will be significantly constrained. Flight III with a 14-foot AMDR will not be powerful enough to meet the Navy’s objective, or desired IAMD capabilities. The shipyards and the Navy have determined that 14-foot radar arrays are the largest that can be accommodated within the confines of the existing DDG 51 configuration. Adding a radar larger than 14 feet to DDG 51 is unlikely without major structural changes to the ship. AMDR is being specifically developed to be a scalable radar—meaning that it can be increased in size and power to provide enhanced capability against emerging threats. According to AMDR contractors, the Navy had originally contracted for an investigation of a Variant 2 AMDR with a sensitivity of SPY+40, but this effort was cancelled. They added that the maximum feasible size of AMDR would be dictated by the ship and radar power and cooling demands, but that they had investigated versions as large as 36 feet. Leveraging AMDR’s scalability will not be possible on DDG 51 without major changes, such as a new deckhouse or adding to the dimensions of the hullform itself by broadening the beam of the ship or adding a new section (called a plug) to the middle of the ship to add length. Navy officials have stated that adding a plug to DDG 51 is not currently a viable option due to the complexity, and that a new ship design is preferable to a plugged DDG 51. According to senior Navy officials, since the MAMDJF AOA was released the Navy has changed its concept on the numbers of Navy ships that will be operating in an IAMD environment. Rather than one or a small number of ships conducting IAMD alone and independently managing the most taxing threat environments without support, the Navy now envisions multiple ships that they can operate in concert with different ground and space-based sensor assets to provide cueing for AMDR when targets are in the battlespace. This cueing would mean that the shooter ship could be told by the off-board sensors where to look for a target, allowing for earlier detection and increased size of the area that can be covered. According to the Navy, this concept—referred to as sensor netting—can be used to augment the reduced radar capability afforded by a 12 or 14-foot AMDR as compared to the larger radars studied in the MAMDJF AOA. For example, the Navy cited the use of the Precision Tracking Space System program as an example of sensors that could be leveraged. However, this program (envisioned as a constellation of missile tracking satellites) is currently in the conceptual phase, and the independent Radar/Hull Study red team stated that the development timeline for this system is too long to consider being able to leverage this system for Flight III. Navy officials told us that another option would be to leverage the newly completed Cobra Judy Replacement radar ship and its very powerful dual-band radar to provide cueing for DDG 51s. This cueing could allow the DDG 51s to operate a smaller AMDR and still be effective. The Cobra Judy Replacement ship is comparatively cheaper than DDG 51s (approximately $1.7 billion for the lead ship), and was commercially designed and built. However, it is not a combatant ship, which would limit its employment in a combat environment and make it difficult to deploy to multiple engagement locations. Senior Navy officials told us that the concept of sensor netting is not yet well defined, and that additional analysis is required to determine what sensor capabilities currently exist or will be developed in the future, as well as how sensor netting might be conceptualized for Flight III. Sensor netting requires not only deployment of the appropriate sensors and for these sensors to work alone, but they also need to be able to share usable data in real-time with Aegis in the precise manner required to support BMD engagements. Though sharing data among multiple sensors can provide greater capabilities than just using individual stand- alone sensors, officials told us that every sensor system has varying limitations on its accuracy, and as more sensors are networked together and sharing data, these accuracy limitations can compound. Further, though there have been recent successes in sharing data during BMD testing, DOD weapons testers responsible for overseeing BMD testing told us that there have also been issues with sending data between sensors. Although sensor technology will undoubtedly evolve in the future, how sensor netting will be leveraged by Flight III and integrated with Navy tactics to augment Aegis and the radar capability of Flight III is unknown. The Navy has added a future DDG 51 flight (known as Flight IV) to its annual long-range shipbuilding plan submitted to Congress, with procurement of 21 ships to begin in 2032. According to the Navy, this Flight IV ship could be notionally based on the DDG 51 hullform, but it may be largely or entirely a clean sheet design. DOD officials stated that no decisions have been made with respect to the capabilities of this future platform, and that Executive Office of the President and DOD decisions may ultimately dictate further analysis on the capabilities needed for future surface combatants. If additional studies are completed and materiel solutions are recommended, DOD officials stated that an AOA may be warranted. Senior Navy officials told us that they do not know if Flight IV will carry a larger, more powerful radar or not or what the overall improvements in capabilities will be, even though AMDR is being built with the capability to be scaled up in size. In its recent annual long-range shipbuilding plan, the Navy currently estimates that its notional Flight IV ships will cost approximately $2.1 billion each—the same as the Flight III ships, which implies no expectations of changes to the hullform. Navy officials told us that this amount was a placeholder. Officials told us that a major consideration in the future will be electrical power. While Flight III will most likely not leverage technologies developed as part of the DDG 1000 program because of DDG 51’s design constraints, Navy officials stated that Flight IV may carry some form of the integrated power system developed for DDG 1000. The Navy examined the use of the integrated power system for Flight III in the Flight Upgrade Study, but found that it was not currently viable due to current component technology. The constrained nature of Flight III will likely limit the ability of the Navy to add future weapon technologies to these ships— such as an electromagnetic rail gun or directed energy weapons as these technologies mature—unless the Navy wants to remove current weapon systems. For example, the ongoing Navy Flight Upgrade Study examined an option to add a small rail gun by removing the ship’s main 5-inch gun and the forward 32-cell missile launcher system. It is unknown when these future technologies may be used. Costs of the lead Flight III ship will likely exceed current budget estimates. Although the Navy has not yet determined the final configuration for the Flight III ships, regardless of the variant it selects, it will likely need additional funding to procure the lead ship above the level in its current shipbuilding budget. The Navy has estimated $2.6 billion in its fiscal year 2012 budget submission for the lead Flight III ship. However, this estimate may not reflect the significant design and construction challenges that the Navy will face in constructing the Flight III DDG 51s— and the lead ship in particular. In fact, the Navy’s most current estimates for a range of notional Flight III options are between $400 million and $1 billion more than current budget estimates, depending on the configuration and equipment of the variant selected (see table 11 below). Further, across the entire flight of 22 ships, the Navy currently estimates Flight III research and development and procurement costs to range from $58.5 billion to $64.1 billion in constant 2012 dollars. However, the Navy estimated in its 2011 long-range shipbuilding plan to Congress that these same 22 ships would cost approximately $50.5 billion in constant 2012 dollars. As shown in figure 10 below, depending on the extent of changes to hullform, the Navy may need at least $4.2 billion to $11.4 billion more to procure DDG 51 Flight III ships. Based on past experience, the Navy’s estimates for future DDG 51s will likely increase further as it gains greater certainty over the composition of Flight III and beyond. At the beginning of a program, uncertainty about cost estimates is high. Our work has shown that over time, cost estimates become more certain as the program progresses—and generally increase as costs are better understood and program risks are realized. Recent Navy shipbuilding programs, such as the Littoral Combat Ship program, initially estimated each ship to cost less than $220 million. This estimate has more than doubled as major elements of the ships’ design and construction became better understood. In the case of Flight III, the Navy now estimates 3 to 4 additional crew members will be required per Flight III ship to support the IAMD mission and AMDR than it estimated in the earlier Radar/Hull Study. Increases in the cost of Flight III would add further pressure to the Navy’s long-range shipbuilding plan. Beginning in 2019, the Navy will face significant constraints on its shipbuilding account as it starts procuring new ballistic missile submarines to replace the current Ohio class. The Navy currently estimates that this program will cost approximately $80.6 billion in procurement alone, with production spanning over a decade. Despite uncertainty in the costs of the DDG 123, the Flight III lead ship, the Navy currently plans to buy the ship as part of a multiyear procurement, including 8 DDG Flight IIA ships, and award the contract in fiscal year 2013. Multiyear contracting is a special contracting method to acquire known requirements for up to 5 years if, among other things, a product’s design is stable and technical risk is not excessive. According to the Navy, from fiscal year 1998 through 2005, the Navy procured Flight IIA ships using multiyear contracts yielding significant savings estimated at over $1 billion. However, the Navy first demonstrated production confidence through building 10 Flight IIAs before using a multiyear procurement approach. While Flight III is not a new clean sheet design, the technical risks associated with AMDR and the challenging ship redesign as well as a new power and cooling architecture coupled with the challenges to construct such a dense ship, will make technical risk high. Further, technical studies about Flight III and the equipment it will carry are still underway, and key decisions about the ship have not yet been made. DDG 123 is not due to start construction until fiscal year 2016. If the Navy proceeds with this plan it would ultimately be awarding a multiyear contract including this ship next fiscal year, even though design work has not yet started and without sufficient knowledge about cost or any construction history on which to base its costs, while waiting until this work is done could result in a more realistic understanding of costs. Our prior work has shown that construction of lead ships is challenging, the risk of cost growth is high, and having sufficient construction knowledge is important before awarding shipbuilding contracts. Given the potential technology, design, and construction risks, and level of the investment, the current level of program oversight for DDG 51 Flight III may not be sufficient. The DDG 51 program has a long history and has already passed through all of the DOD acquisition milestone reviews (formerly Milestones 0 through IV, now Milestones A through C), and is now an Acquisition Category (ACAT) 1C program. A program’s acquisition category is based on its location in the acquisition process, dollar value, and Milestone Decision Authority special interest, and the acquisition category determines the program’s decision authority. For an ACAT 1C program, the Assistant Secretary of the Navy (Research, Development, and Acquisition) is ultimately the Milestone Decision Authority. As the Milestone Decision Authority, the Assistant Secretary is designated as having the authority to approve entry of an acquisition program into the next phase of the acquisition process, and is accountable for cost, schedule, and performance reporting to higher authority, including congressional reporting. This differs from the higher- level ACAT 1D designation, where the Undersecretary of Defense for Acquisition, Technology and Logistics is the Milestone Decision Authority. The ACAT 1D designation provides a higher level of oversight to the program, and also provides enterprisewide visibility over acquisition program decisions. Although it is a potentially $64 billion investment spanning decades, DDG 51 program office officials do not believe that the Flight III changes are significant enough to warrant a return to ACAT 1D oversight. According to officials, since the AMDR program—which they believe is the risky element of Flight III—is already an ACAT 1D on its own and is also progressing through the milestone process, the ship does not warrant ACAT 1D designation. Similarly, program officials have stated that they believe AMDR has sufficient oversight for Flight III and that it is unnecessary for the ship to repeat any milestones. However, significant re-design and changes to the hull and mechanical and electrical systems will be required for Flight III, which could bring potentially significant risks not being captured by AMDR oversight alone. For example, the addition of AMDR requires a challenging ship redesign and software modifications to Aegis to integrate the new radar. Further, the program has historically switched from ACAT 1C to ACAT 1D during the transition from Flight I to Flight II which introduced new capabilities. Our analysis shows that Flight III meets DOD criteria for ACAT ID (see table 12 below). Officials from the Office of the Secretary of Defense have indicated support for designating the Flight III program an ACAT 1D program, though a final decision is not expected until 2012 at the earliest. It has also not been decided if the program will be required to return to a prior milestone, a decision also not expected until 2012 at the earliest. Typically, a milestone review gives decision makers an opportunity to evaluate important program documentation to help demonstrate that the program has the appropriate knowledge to proceed with development or production. In preparation for a milestone, programs submit documents for well over 10 information requirements, including an independent cost estimate, and technology readiness and affordability assessments. Though the Navy is working on a draft capabilities document for Flight III, without a milestone decision there may be no requirement to compel the Navy to develop this document. Further, without a milestone there will be no requirement for the Navy to seek an independent cost estimate from the office of Cost Assessment and Program Evaluation, typically submitted at a milestone review. According to Navy officials, they may consider developing a life-cycle cost estimate prior to requesting approval for the multiyear procurement approach. The DDG 51 program last conducted an independent cost estimate in 1993. The Navy is in the early stages of a potential $80 billion investment in up to 43 DDG 51 destroyers to provide IAMD capability for potentially up to the next 60 years. Such investment decisions cannot be made without some degree of uncertainty; they will always involve risks—especially in the early stages of a program. Yet, a decision of this magnitude should proceed with a solid base of analysis—regarding the alternatives, cost, and technical risks—as well as a plan for oversight that provides sufficient leverage and flexibility to adapt to information as it emerges. These pieces are not sufficiently in place, at least with respect to Flight III and AMDR. To its credit, the Navy’s goal was to move towards a lower-cost solution that could be rapidly fielded; however, there are a number of key shortfalls in the Navy’s analysis in support of its decisions. As it stands, the Navy risks getting a solution that is not low cost, will not be fielded in the near-term, or meet its long-term goals. DDG 51 may ultimately be the right decision, but at this point, the Navy’s analysis has not shown this to be the case. Specific issues include: The Navy’s choices for Flight III will likely be unsuitable for the most stressful threat environments it expects to face. While the Navy potentially pursued a lower-cost ship solution, it did not assess the effect of this decision in terms of long-term fleet needs where more of these ships may be required to provide the same capability of a smaller number of more costly, but more capable, ships. Though the Navy hopes to leverage sensor netting to augment the capability of these ships, there is a shortage of analysis and testing with operational assets to demonstrate that this is a viable option. The Navy clearly states in recent AMDR documents that a new, as-of- yet undefined ship is required to meet its desired IAMD capability. However, it has not yet articulated its long-term plans for a new surface combatant that is sized to be able to carry a larger AMDR, and such a ship is not currently in the Navy’s long-range shipbuilding plan. Without a robust operational test program that will demonstrate both DDG 51 with the modified Aegis combat system and the new AMDR, the Navy cannot be sure that the ships can perform the IAMD mission as well as planned. In addition to these issues about the analysis underpinning the DDG 51 program, oversight of the program moving forward could be limited by two factors: If the milestone decision authority remains at its current level, needed scrutiny may not occur. While the proper milestone entry may be discretionary, it is clear that the cost and risk of Flight III and AMDR warrant additional oversight. If the Navy pursues a multiyear shipbuilding contract that includes the lead ship of Flight III, visibility over the risks inherent in lead ship construction could be obscured. We recommend that the Secretary of Defense direct the Secretary of the Navy to take the following three actions: 1. Conduct a thorough AOA in accordance with DOD acquisition guidance for its future surface combatant program to include: (a) a range of representative threat environments developed in concert with the intelligence community; (b) results of its ongoing Flight III studies and full cost estimates in advance of awarding DDG 51 Flight III production contracts; (c) implications of the ability of the preferred ship to accommodate new technologies on future capabilities to determine the most suitable ship to carry AMDR and meet near-term IAMD requirements and provide a path to far-term capabilities; (d) implications on future fleet composition; and (e) an assessment of sensor netting—conducted in consultation with MDA and other cognizant DOD components—to determine the risks inherent in the sensor netting concept, potential current or planned programs that could be leveraged, and how sensor netting could realistically be integrated with the selected future surface combatant to assist in conducting BMD. This AOA should be briefed to the Joint Requirements Oversight Council. 2. Report to Congress in its annual long-range shipbuilding plan on its plans for a future, larger surface combatant, carrying a more capable version of AMDR and the costs and quantities of this ship. 3. In consultation with MDA and DOD and Navy weapons testers, define an operational testing approach for the Aegis ACB-12 upgrades that includes sufficient simultaneous live-fire testing needed to fully validate IAMD capabilities. We also recommend that the Secretary of Defense take the following two actions: 1. Upgrade the oversight of the Navy’s future surface combatant program to ACAT 1D status, and ensure that the appropriate milestone entry point is selected to provide cost baselines and assessments of design and technical risks and maturity. 2. Ensure that the planned DDG 51 multiyear procurement request does not include a Flight III ship. We provided a draft of this report to DOD for review and comment. DOD provided a written response which is reprinted in appendix II. DOD also submitted technical comments that were incorporated into the report as appropriate. DOD concurred with our second recommendation that the Navy report to Congress in its annual long-range shipbuilding plan on its plans for a future larger surface combatant carrying a more capable version of AMDR. Given the assessments that the Navy is currently conducting on surface combatants, the Navy’s next submission should include more specific information about its planned future surface combatant acquisitions. DOD also agreed with our third recommendation on live-fire testing of Aegis ACB-12 upgrades, stating that the Navy and the MDA— working under Office of the Secretary of Defense oversight—are committed to conducting adequate operational testing of ACB-12, but did not offer concrete steps they would take to address our concerns. Moving forward, DOD should demonstrate its commitment to fully validating IAMD capabilities by including robust simultaneous operational live-fire testing of multiple cruise and ballistic missile targets in its Aegis Test and Evaluation Master Plan that is signed by Director, Operational Test and Evaluation. DOD did not agree with our first recommendation to conduct an AOA to support its future surface combatant selection decision, stating that its previous analyses—specifically the MAMDJF AOA and the Radar/Hull Study—comprise a body of work that satisfies the objectives of an AOA. However, DOD did not present any additional evidence to refute our findings. DOD did agree that an assessment of sensor netting needs to be performed. Our analysis shows that the Radar/Hull Study, which was the key determinant in the DDG 51 decision, was a departure from the MAMDJF AOA. These studies are neither complementary nor can they be aggregated. While both sought to determine the best solution to address identified integrated air and missile defense gaps, the Radar/Hull Study essentially answered a different question than the MAMDJF AOA. In essence, it was attempting to identify a cost-constrained, less robust solution, which makes analysis from one study not always appropriate to apply to the other. Specifically, the MAMDJF AOA considered a significantly more taxing threat environment than the Radar/Hull Study, requiring ships carrying very large radars to independently manage these threats. Alternatively, the Radar/Hull Study considered a much less taxing threat environment, allowing for ships carrying smaller radars but that would need to work together to be effective. Ultimately, the MAMDJF AOA eliminated DDG 51 from consideration as a single-ship solution. DOD also states that it is currently conducting additional studies on Flight III, but since these are solely focused on DDG 51, they do not provide any additional insight into the decision as to the appropriate ship that might be used to supplement the Navy’s existing analysis. As we note in this report, the proposed program calls for an investment of up to approximately $80 billion for 43 destroyers, and likely more if the Navy chooses to pursue a Flight IV concept. Given the scope of the Navy’s plans, a thorough AOA is essential to affirm that the decision made is the right one and a sound investment moving forward. This AOA should be briefed to Joint Requirements Oversight Council because of the magnitude of this potential acquisition and because of the joint service interest in IAMD that make it important to have an overarching body review the Navy’s analysis and decisions. We believe that this recommendation remains valid. DOD disagreed with our fourth recommendation to upgrade the acquisition category designation of the Navy’s future surface combatant program to ACAT ID at this time, stating that a determination on the ACAT designation of DDG 51 Flight III will be made by the fourth quarter of fiscal year 2012, once sufficient information is available. If the results of the Navy’s analysis continue to support a DDG 51 Flight III as the appropriate solution, our analysis shows that Flight III already meets criteria for ACAT ID status, and that this status provides an enhanced level of oversight appropriate for a program of this magnitude. This strategy is also in line with the past flight upgrades that were also conducted under ACAT ID status. We therefore believe this recommendation remains valid. Regarding our fifth recommendation that DOD not include a Flight III ship in its planned DDG 51 multiyear procurement request, DOD partially concurred, stating that it is following the statutory requirements for multiyear procurement authority. DOD commented that it will select an acquisition approach that provides flexibility and minimizes the cost and technical risk across all DDG 51 class ships. DOD expects to make a determination on including or excluding Flight III ships within the certification of the planned multiyear procurement that is due to Congress by March 1, 2012. While the Secretary can certify that due to exceptional circumstances, proceeding with a multiyear contract is in the "best interest" of DOD, notwithstanding the fact that one or more of the conditions of the required statutory certification are not met, requesting a multiyear procurement in March 2012 that includes the lead Flight III ship carries significant risk. DOD will be committing to a cost with no actual construction performance data on which to base its estimates and a ship concept and design that are not finalized. While DOD argued that it has in the past included DDG 51’s that were receiving major upgrades in multiyear procurements, as this report shows, planned changes for Flight III could far exceed those completed in past DDG 51 upgrades. We therefore believe that, in view of the current uncertainty and risk, our recommendation remains valid to exclude a Flight III ship from the upcoming multiyear procurement request. In view of the Navy’s disagreement with a number of our recommendations, we are elevating these issues to the attention of Congress. In the coming years, the Navy will ask Congress to approve funding requests for DDG 51 Flight III ships and beyond. Without a solid basis of analysis, we believe Congress will not have assurance that the Navy is pursuing an appropriate strategy with regard to its future surface combatants, including the appropriate level of oversight given its significant cost. To help ensure that the department makes a sound investment moving forward, Congress should consider directing the Secretary of Defense to: 1. require the Navy to submit a thorough, well-documented AOA for the its future surface combatant program that follows both DOD acquisition guidance and the elements outlined in our first recommendation prior to issuing solicitations for any detail design and construction contracts of DDG 51 Flight III ships; 2. elevate the ACAT status of the DDG 51 Flight III to an ACAT ID level if the decision is made to continue pursuing the program; and 3. include the lead DDG 51 Flight III ship in a multi-year procurement request only when the Navy has adequate knowledge about ship design, cost, and risk. We are sending copies of this report to the Secretary of Defense. We are also sending copies to the appropriate congressional committees. In addition, the report is available at no charge on GAO’s website at http://www.gao.gov. If you or your staff has any questions about this report, please contact Belva Martin at (202) 512-4841 or martinb@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix III. The overall objectives of this review were to assess (1) the Navy’s determination of the most appropriate platform to meet current and future surface combatant requirements; (2) the differences in cost, schedule, and design of the restart DDG 51 destroyers compared with previous ships, and the risks associated with the restart; and (3) the feasibility of the Navy’s plans for maturing and integrating new technologies into the future DDG 51 ships. To assess how the Navy determined the most appropriate platform to meet current and future surface combatant requirements, we analyzed the Navy’s Radar/Hull Study, which was the main tool the Navy used for assessing the radar and ship options and reviewed the accompanying “red team” assessment. We compared this study with other Navy studies related to ballistic missile defense (BMD) and integrated air and missile defense (IAMD), including the Navy’s Maritime Air and Missile Defense of Joint Forces (MAMDJF) analysis of alternatives, the Navy BMD “Knee in the Curve Study,” a Navy Cruiser and Destroyer analysis study, and Office of Naval Intelligence threat assessment studies. We also reviewed the Operational Requirements Document for the DDG 1000 and the draft Capability Definition Document for the Air and Missile Defense Radar (AMDR). We also obtained and reviewed internal Navy briefing slides used to present the findings of the Radar/Hull Study to Navy decision makers. To assess the steps taken by the Navy in making this decision, we reviewed relevant Department of Defense (DOD) policy and guidance documents addressing, among other things, acquisition program initiation including DOD Instruction 5000.02 and the Weapon Systems Acquisition Reform Act of 2009. We compared the Radar/Hull Study with DOD analysis of alternatives guidance found in the Defense Acquisition Guidebook, DOD Instruction 5000.02, and a July 2008 Air Force Analysis of Alternatives handbook. We also analyzed key contractor data submissions related to the ship variants and the radar concepts that were provided to the Navy to support its decision. We met with officials from the Radar/Hull Study team, the Applied Physics Laboratory at Johns Hopkins University who were technical consultants on the study, the DDG 51 and DDG 1000 program offices, representatives from the Office of the Chief of Naval Operations Surface Warfare Division, officials from the Program Executive Office for Ships (PEO Ships), the Program Executive Office for Integrated Warfare Systems (PEO IWS) program offices responsible for the Aegis combat system and for AMDR, and contractor officials from Raytheon, Lockheed Martin, and Northrop Grumman. We met with officials from the Office of Naval Intelligence to discuss the threat environment, and we met with officials from the Joint Integrated Air and Missile Defense Organization to discuss the recent Joint Capabilities Mix study which established required numbers of Navy BMD capable ships. We also met with an official from the Joint Staff to discuss the role of the Joint Requirements Oversight Council in the DDG 1000 truncation and DDG 51 restart decisions. To assess the differences in cost between the restart DDG 51 ships and previous DDG 51 ships, we examined the Navy budget for DDG 51 restart ships and compared it with the budget for prior ships. We also spoke with the DDG 51 program office and Navy cost estimating officials, and discussed their methodology for estimating the impact of the production gap on prices, and spoke to officials from Bath Iron Works in Bath, Maine and Ingalls Shipbuilding in Pascagoula, Mississippi—the shipyards responsible for building DDG 51 destroyers—and the officials from the Navy’s Supervisor of Shipbuilding at both sites about the impact of the gap on cost estimates. We also spoke to shipyard officials at both sites about their readiness to begin construction. We analyzed the Navy’s revised acquisition strategy for hulls DDG 114 through DDG 116. To assess differences in production schedules we compared the Navy’s projected schedules for the Flight IIA restarts with the actual schedule performance on previous Flight IIA ships. We also spoke with Navy and shipyard officials at both shipyards. To assess the design changes for the restart ships, we compared the estimated number of design drawing changes and engineering change proposals for Flight IIA restart ships with those for previous Flight IIA ships. We examined Navy and contractor-provided analyses pertaining to the Aegis upgrade (ACB 12) with specific focus on the source lines of code (SLOC), and compared SLOC estimates with SLOC actual numbers. We also reviewed software defect rates and development schedules related to the ACB 12 upgrade, and we analyzed the ACB-12 development and test schedules, risk matrices, and results from relevant test events that might impact ACB 12 availability for installation on DDG 113. We analyzed Defense Contract Management Agency (DCMA) reports on ACB 12 development, and spoke to relevant DCMA officials. We also reviewed Navy, Missile Defense Agency (MDA), and Director, Operational Test and Evaluation (DOT&E) proposed operational test schedules and plans to assess integration efforts to verify IAMD capability, and interviewed relevant Lockheed Martin, MDA, DOT&E, and DOD Development Test and Evaluation officials. To address the feasibility of the Navy’s plans for maturing new technologies intended for DDG 51 Flight III ships, we analyzed key Navy documentation including the DDG 51 Flight Upgrade Study (Phase I) and the accompanying “red team” assessment, contractor AMDR concept development documents, and AMDR Top Level Radar Performance documents. We compared the development of AMDR and its development schedule with previous Navy radar development programs (e.g. Cobra Judy Replacement radar, Dual Band Radar) to determine the feasibility of the technology and the development schedule. We also discussed development, testing, and in-yard date schedules with the Navy. We met with each of the three AMDR contractors: Raytheon, Lockheed Martin, and Northrop Grumman. To determine the feasibility of integrating AMDR and other technologies into Flight III, we compared the Navy’s Flight III concepts with Navy service life allowance guidelines, and spoke with officials from both shipyards and a former Navy ship designer. To assess the feasibility of the Navy’s acquisition strategy for Flight III we analyzed relevant DOD acquisition guidance including DOD Instruction 5000.02, and spoke with officials from the Office of the Secretary of Defense for Acquisition, Technology, and Logistics. We also used GAO’s Cost Estimating and Assessment Guide. We are providing you with a classified annex containing supplemental information. We conducted this performance audit from January 2011 through January 2012 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives, and that the data we obtained and analyzed are sufficiently reliable for the purposes of our assessment. Belva Martin, 202-512-4841 or martinb@gao.gov. In addition to the contact above, Diana Moldafsky, Assistant Director; Jennifer Echard; Laura Greifner; Kristine Hassinger; Jeremy Hawk; Ioan Ifrim; C. James Madar; G. Michael Mikota; Karen Richey; W. Kendall Roberts; Roxanna Sun; and Alyssa Weir made key contributions to this report.

After nearly a decade and almost $10 billion in development on Zumwalt class destroyers, the Navy changed its acquisition approach from procuring Zumwalts to restarting production of Arleigh Burke class destroyers (DDG 51) and building a new version, known as Flight III. As requested, GAO reviewed the Navy’s plans for DDG 51 and missile defense capabilities by (1) evaluating how the Navy determined the most appropriate platform to meet surface combatant requirements; (2) identifying and analyzing differences in design, cost, and schedule of the restart ships compared with previous ships; and (3) assessing the feasibility of Navy plans for maturing and integrating new technologies and capabilities. GAO analyzed Navy and contractor documentation and interviewed Navy, contractor, and other officials. The Navy relied on its 2009 Radar/Hull Study as the basis to select DDG 51 over DDG 1000 to carry the Air and Missile Defense Radar (AMDR) as its preferred future surface combatant—a decision that may result in a procurement of up to 43 destroyers and cost up to $80 billion over the next several decades. The Radar/Hull Study may not provide a sufficient analytical basis for a decision of this magnitude. Specifically, the Radar/Hull Study: focuses on the capability of the radars it evaluated, but does not fully evaluate the capabilities of different shipboard combat systems and ship options under consideration, does not include a thorough trade-off analysis that would compare the relative costs and benefits of different solutions under consideration or provide robust insight into all cost alternatives, and assumes a significantly reduced threat environment from other Navy analyses, which allowed radar performance to seem more effective than it may actually be against more sophisticated threats. The Navy’s planned production schedules of the restart DDG 51 ships are comparable with past performance and officials told us that hull and mechanical systems changes are modest, but these ships will cost more than previous DDG 51s. A major upgrade to the ship’s combat system software also brings several challenges that could affect the restart ships, due in part to a key component of this upgrade that has already faced delays. Further delays could postpone delivery to the shipyard for the first restart ship, and could also jeopardize the Navy’s plan to install and test the upgrade on an older DDG 51 prior to installation on the restart ships. This first installation would serve to mitigate risk, and if it does not occur on time the Navy will be identifying, analyzing, and resolving any combat system problems on the first restart ship. Further, the Navy does not plan to fully test new capabilities until after certifying the upgrade as combat-ready, and has not planned for realistic operational testing necessary to fully demonstrate its integrated cruise and ballistic missile defense performance. The Navy faces significant technical risks with its new Flight III DDG 51 ships, and the current level of oversight may not be sufficient given these risks. The Navy is pursuing a reasonable risk mitigation approach to AMDR development, but it will be technically challenging. According to Navy analysis, selecting the DDG 51 hullform to carry AMDR requires significant redesign and reduces the ability of these ships to accommodate future systems. This decision also limits the radar size to one that will be at best marginally effective and incapable of meeting the Navy’s desired capabilities. The Navy may have underestimated the cost of Flight III, and its plan to include the lead ship in a multiyear procurement contract given the limited knowledge about the configuration and the design of the ship creates potential cost risk. Finally, the current level of oversight may not be commensurate with a program of this size, cost, and risk and could result in less information being available to decision makers. GAO is making several recommendations to the Secretary of Defense, including requiring the Navy to conduct thorough analyses of alternatives for its future surface combatant program and conduct realistic operational testing of the integrated missile defense capability of the DDG 51’s upgrade, ensuring that the Navy does not include the lead Flight III ship in a multiyear procurement request, and raising the level of oversight for this program. DOD agreed with the recommendations to varying degrees, but generally did not offer specific actions to address them. GAO believes all recommendations remain valid and has included matters for congressional consideration to ensure the soundness of the Navy’s business case.

According to data from the NTD, there were about 2,200 transit agencies in the U.S. in 2013, most of which used buses. The bus fleets ranged in size from large agencies operating more than a thousand buses to small agencies with a single bus. For example, MTA New York City Transit reported that it had more than 3,300 buses and the city of Goodland, Kansas, reported that it had one. Although bus fleets of transit agencies in most major cities number in the hundreds, 85 percent of all transit agencies operated fewer than 50 buses. Bus transit services from these agencies include: fixed routes, in which buses operate according to a set route and schedule; demand response, including paratransit service required under the Americans with Disabilities Act 1990, as amended (ADA), in which buses are scheduled in response to calls from passengers; and, deviated-fixed routes, which are fixed routes that allow for minor route deviations in response to passenger calls. To provide these services, transit agencies reported that they used a total of nearly 92,000 buses, representing a wide range of types and sizes. For the purposes of this report, we categorized buses as follows: Heavy-duty buses—the largest buses used for public transportation, typically used for fixed-route service (see fig. 1). Almost all heavy-duty buses are made specifically for the transit market, although many major components—including the engine, transmission, axles, and brakes—are also used in heavy-duty trucks. Cutaway buses—predominately used for demand response (see fig. 1). On cutaways, a bus body is manufactured and mounted on a chassis built by another manufacturer. Virtually all are equipped to accommodate wheelchairs. Buses most commonly use diesel as a fuel source, but alternative fuel options are used, including compressed natural gas, biodiesel, and hybrid electric. Table 1 compares various characteristics of the types of buses used by transit agencies. Transit agencies procure new buses to replace aging existing fleets or when expanding service and FTA policies specify how agencies manage, use, and dispose of buses acquired with federal funds. For example, FTA has established a minimum useful life for buses, in terms of years in service or accumulation of miles, and requires that buses are maintained and remain in use for at least their minimum life. To avoid the purchase of unnecessary vehicles, FTA policy limits the size of each agency’s bus fleet to those the agency needs to provide its services. Thus, for each new replacement bus purchased, FTA requires agencies to provide plans for disposition of the vehicle to be replaced. While some transit agencies purchase buses frequently, most do not make purchases every year and some go years between purchases. When transit agencies purchase buses using federal funds, there are four procurement methods recognized by FTA that they typically use, depending on their needs: Stand-alone procurement—conducted by a single agency and resulting in a contract between it and a bus manufacturer or vendor. Joint procurement—conducted by two or more agencies where each is a party from the outset to a single contract with a bus manufacturer or vendor. Assignment of options (also referred to as “piggybacking”)— conducted by two or more agencies where one agency with an existing bus procurement contract determines that it inadvertently contracted to purchase more buses than it needs. The agency may assign the rights to purchase some or all of its unneeded buses to other transit agencies. Purchasing from a state schedule—many states create purchasing schedules by which the state and its subsidiaries may acquire goods, including buses. Some subsidiaries—including cities and counties— are also transit agencies and purchase buses from the state’s purchasing schedule. FTA approves of this procurement method as long as the state ensures that procurement contracts include any clauses required by federal law. Recipients of FTA assistance are required to conduct procurements in accordance with FTA procurement rules including, for example, that all procurement transactions, such as stand-alone, joint, and assignment of options, be conducted in a manner that provides full and open competition and that recipients agree to purchase a finite number of vehicles based on the agency’s needs. Purchases from a state schedule also must follow FTA procurement rules such as the inclusion of FTA- required clauses and certifications but are typically completed by the state and not a transit agency. Over the last decade, the number of heavy-duty transit bus manufacturers in the U.S. has declined due to business failures and consolidation. The final assembly of heavy-duty buses purchased by transit agencies using funds from FTA generally takes place in the U.S., in order to meet Buy America requirements. In 2013, four companies produced virtually all heavy-duty buses used in the U.S. Transit agencies that receive federal funds are required to report operational data at least annually to the NTD, including information on the agencies’ bus fleets. Using these data, we identified the year and manufacturer of buses that transit agencies reported they owned or used in 2004 and 2013 and compared the two time periods. Our analysis of this data shows: The number of heavy-duty manufacturers declined—in 2004, 10 manufacturers produced almost all of the reported buses; by 2013, four manufacturers produced almost all of the buses. Smaller manufacturers stopped bus production—most manufacturers that stopped producing heavy-duty buses between 2004 and 2013 were relatively small, with each making 4 percent or fewer of the 2004 reported buses, although one manufacturer made 11 percent. The market share for the largest manufacturers increased —in 2004, the largest three manufactures produced 68 percent of the reported heavy-duty buses; in 2013, the same three manufacturers produced 81 percent of the buses. In contrast with heavy-duty manufacturers, the number of manufacturers of cutaway transit buses has remained about the same. As with heavy- duty buses, final assembly of cutaway buses purchased by transit agencies using funds from FTA generally takes place in the U.S., to meet Buy America requirements. According to FTA, transit agencies generally procure cutaways through dealers representing the manufacturers; each bus is to be produced to an agency’s specifications. There is also a substantial private market for cutaway buses. Specifically, hotels, rental car agencies, and universities also purchase cutaway buses. In 2007, FTA estimated that sales to these entities surpassed transit-related sales. Our analysis of bus information reported by transit agencies in 2004 and 2013 shows that: Almost all cutaway bus manufacturers continued to produce buses— of the 13 largest manufacturers in 2004, all continued to produce transit buses in 2013. At least two companies started producing buses—two companies not identified in 2004 each produced about 4 percent of the cutaway buses in 2013. Market share is more evenly distributed—in 2004, the three largest manufacturers made 64 percent of the cutaway buses; in 2013, the three largest manufacturers made 47 percent. In general, because the market share of the largest manufacturers declined, other manufacturers had larger shares. The decline in the number of heavy-duty bus manufacturers has not affected the number of buses produced. Excluding buses purchased with funds from the Recovery Act, the number of buses procured by transit agencies annually varied slightly from 2009 through 2013 (see table 2). Approximately half of the bus purchases were for heavy-duty buses and half were for cutaway buses, although the percentages vary from year to year. The Recovery Act provided additional funds for capital projects, including bus procurement, in 2009 and 2010. Specifically, transit agencies and others purchased 7,544 buses using Recovery Act funds over those 2 years. The number of transit buses produced annually is relatively small compared to heavy-duty trucks. Specifically, there are approximately 2,500 heavy-duty buses produced each year and approximately 300,000 heavy-duty trucks. Similarly, of the 370,000 cutaway chassis sold annually, 2,500 are purchased and modified for transit use. According to a 2007 FTA report, this situation is both beneficial and problematic for the transit industry. Components shared by multiple industries likely make their costs lower than if they were used only for transit. Conversely, the transit industry as a whole exerts little influence on availability of components or their design. FTA provides support for public transportation by awarding federal funding—about $11 billion in 2013—to transit agencies in the form of grants. Grant recipients have specific responsibilities associated with the use of this funding. For example, transit agencies are required to conduct procurements involving federal funds in compliance with applicable federal requirements. FTA is responsible for ensuring compliance with these requirements and for ensuring that these agencies use the funding prudently. In addition, FTA describes its role as one where, in general, FTA does not substitute its judgement for that of its recipients by making contract decisions for recipient transit agencies entering into procurement contracts with third parties, such as bus manufacturers or vendors. There were at least 11 federal grant programs in fiscal year 2013 that allowed funds to be used for bus procurement. The rules associated with use of grant money vary by program, but generally transit agencies have flexibility in deciding how to use these funds. For example, an agency may use most of its grant funding one year on a commuter rail project and use most of its funding for bus procurement in the next year. Transit providers may, and often do, receive funding from more than one program. Further, in each year, FTA may obligate funds—that is, enter into a grant agreement with a transit agency to, for example, procure buses—that were appropriated in that year or in prior years. As a result, the amount obligated by FTA in any given year is not the same as is appropriated. Four federal grant programs provide 95 percent of the funding awarded to transit agencies for bus procurement: Urbanized Area Formula Program (49 U.S.C. §5307)—provides funds to urban areas, defined as those with a population of at least 50,000 based on census information. Funds provided under this program may be used for capital projects, such as purchasing buses, planning, job access and reverse commute projects, and operating and other expenses. Bus and Bus Facilities Formula Program (49 U.S.C. §5339)—provides funds to allow transit agencies to overhaul and retrofit their buses as well as purchase new buses and construct bus-related facilities, such as bus maintenance facilities. Enhanced Mobility of Seniors and Individuals with Disabilities Program (49 U.S.C. §5310)—provides funds to transit agencies and other organizations for transportation programs for seniors and individuals with disabilities that go beyond the scope of public transportation fixed-route service as well as providing funds for ADA complementary paratransit services. At least 55 percent of these funds must be used on capital expenses, such as bus procurement. Rural Area Formula Program (49 U.S.C. §5311)—provides funds to states and tribal areas for use in areas with populations of less than 50,000. States distribute these funds to non-urban and rural areas. Funds can be used to support public transportation in rural areas in order to support the needs of rural areas such as access to healthcare, education, employment, public services, and recreation. Funds can also be directed towards capital, operating, and other expenses. Transit agencies have recently used increasing amounts from their grant funds to purchase buses. Specifically, from fiscal years 2009 to 2013, the most recent year data are available, obligations made by FTA to transit agencies for bus procurement increased (see table 3). These figures represent only the money that transit agencies have decided to use towards bus purchases and do not represent all other forms of transit spending. In addition, transit agencies have used funds from other sources to procure buses. For example, transit agencies and others used approximately $1.7 billion provided by the Recovery Act for procuring 7,544 buses from 2009 to 2010. Some of these grant programs changed over this 5-year period. For example, the Bus and Bus Facilities program was a discretionary program through fiscal year 2012, after which it became a formula program pursuant to changes made by the Moving Ahead for Progress in the 21st Century Act (MAP-21). When using federal funds from FTA to procure buses, transit agencies must comply with a range of federal requirements that include government-wide requirements, such as complying with the ADA, as well as FTA-specific requirements, such as providing a local funding match for federal grants. FTA includes federal requirements in the FTA Master Agreement, which contains the terms and conditions applicable to each grant. Some of the relevant clauses that apply to bus procurements are shown in table 4. FTA is responsible for ensuring that grant recipients comply with federal bus procurement requirements. According to FTA, it relies on transit agencies to annually certify, for each grant, that its procurement system meets all federal requirements and that those requirements are also met by third party contractors. FTA conducts periodic reviews to assess transit agencies’ performance and adherence to FTA requirements and policies. For example, FTA conducts triennial reviews of transit agencies receiving Urbanized Area grant funds. According to FTA officials, about 600 transit agencies receive Urbanized Area funds and approximately one-third of these agencies (200) undergo a triennial review each year. These triennial reviews include a review of an agency’s procurement system but may also include an Enhanced Review Module (ERM) with a specific focus on procurement. ERMs can be triggered by factors including high funding levels, open or repeated findings from previous reviews, unusual or complex procurements, a large number of change orders, bid protests, or Buy America issues. Based on its own risk assessment findings, FTA may also conduct a Procurement System Review (PSR). According to FTA officials, these PSRs examine at least 56 different elements of a transit agency’s procurement system plus its Buy America certifications to determine if it meets FTA procurement requirements and to identify needed corrective actions. FTA has found errors in the way transit agencies procure buses. For example, in 2013, FTA found that several bus procurement contracts did not comply with federal procurement requirements that limit the number of buses an agency may purchase to those needed for its public transportation services. According to FTA, these contracts, in which the parties contracted to buy more buses than they currently or could reasonably be expected to need, were apparently specifically designed for piggybacking—or assigning contract rights to other transit agencies— at a later date. Under FTA rules, piggybacking is allowed only when an agency inadvertently contracts for more buses than it needs. FTA also found that agencies procured buses from a Minnesota state contract even though the agencies were located outside that state, in violation of FTA rules. FTA noted that under this contract, to purchase a bus, a transit agency needed only to submit a simple, one-third page application to the State of Minnesota, receive a membership permit number, and send a purchase order to the vendor of its choosing listed in the contract. While some states have established purchasing schedules for use by state agencies and other authorized entities, FTA prohibits transit agencies using federal funds from FTA to procure buses from out-of-state contracts. Despite that, at the time of the FTA review, at least 13 transit agencies located outside of Minnesota had pending orders to procure buses through the Minnesota state contract with a single bus manufacturer, according to FTA. In both the piggybacking and Minnesota state contract cases, FTA prohibited future purchases but allowed pending purchases to stand. FTA provides guidance and technical assistance to help transit agencies comply with federal bus procurement requirements. Transit agencies access this guidance through FTA’s website or by contacting officials in FTA regional offices. According to FTA, the primary source of procurement guidance from FTA is FTA Circular 4220.1F – Third Party Contracting Guidance, which describes how transit agencies can comply with federal requirements when using federal funds for procurements, such as for buses. FTA issued the current version of the circular in November 2008, and revised it most recently in March 2013. In addition, individual circulars for some of the grant programs used for bus procurement provide guidance related to that program. For example, the circular for the Urbanized Area grant program provides information on the eligible uses of funding, such as for bus replacement, overhaul, or expansion of service. Further, FTA periodically issues Administrator’s Policy Letters to clarify FTA procedures or indicate changes to an existing policy. For example, FTA issued a policy letter in March 2013 to clarify the conditions under which transit agencies could piggyback off another transit agency’s contract. FTA also provides supplemental guidance through its Best Practices Procurement Manual (BPPM), Third Party Procurement Frequently Asked Questions (FAQ), and an online tool that transit agencies can use to assess their procurement systems. This supplemental guidance is designed to assist transit agency officials throughout the procurement process and facilitate their compliance with FTA regulations. However, this guidance is outdated. The BPPM addresses each step in the procurement process using a standardized format consisting of requirements, discussion, and best practices. Requirements include relevant sections from FTA circulars, the Master Agreement, federal laws and regulations, and FTA Administrator’s Policy Letters. The discussion sections provide definitions and guidance concerning the meaning or purpose of the topic being presented, and the best practices section describes practices that have proved to be effective in the past, according to FTA. Although FTA states that it envisions the BPPM to be an ongoing and expanding document that will be updated periodically, FTA has not updated sections of the manual in at least 10 years. As a result, the manual includes outdated references to important procurement policy documents. For example, it includes a previous version of FTA Circular 4220.1E, which was replaced by the current version, FTA Circular 4220.1F, in 2008. Moreover, the fiscal year 2015 FTA Master Agreement notes that the BPPM may lack the necessary information for compliance with certain federal requirements. In June 2015, an FTA official told us that FTA made draft revisions to the BPPM but FTA is awaiting internal approval before releasing it and has not set a time frame for approval. The FAQs on FTA’s website is an online resource for transit agencies to review FTA’s answers to frequently asked procurement-related questions. Visitors to the site can also submit questions to FTA. FTA adds new answers to the website but also retains undated answers with outdated guidance. For example, the bus procurement category includes answers added in 2014 but undated answers refer to the outdated Circular 4220.1E. FTA’s online Procurement System Self-Assessment Guide can be used by transit agencies to assess whether their own procurement system complies with FTA procurement requirements in 10 common areas of deficiency. However, the guide also includes outdated information. For example, the guide references the BPPM for additional guidance; as discussed previously, the BPPM includes outdated information. Furthermore, the guide does not include instructions for completing the self-assessment or obtaining results, and a user is not able to answer the self-assessment questions online. Although there is an area that allows for electronic submission of the self-assessment form to FTA, any form submitted this way would be blank because the form does not allow a user to enter any text. It is important to have updated guidance because transit agencies and state departments of transportation rely on the guidance when procuring buses to ensure compliance with federal requirements. Standards for Internal Control in the Federal Government states that management should ensure that there are adequate means of communicating with external stakeholders that may have a significant impact on the agency achieving its goals. Without updated guidance, transit agencies may not be able to purchase buses as efficiently as possible; for example, they may need to spend additional time researching the guidance or they may have to repeat a required step in the procurement process. As discussed in the next section, transit agency officials in four of our six discussion groups identified outdated guidance as a challenge in bus procurement. FTA also provides technical assistance and training on bus procurement. According to FTA, the National Transit Institute (NTI) developed two-day training on bus procurements in 2011 and held 12 training sessions from 2011 through 2013. The training is now held annually and transit agencies can use FTA grant funds to send personnel to the training. In addition, an FTA headquarters official stated that he provides in-person bus procurement trainings to transit agencies approximately 8 times per year. FTA and APTA developed a standardized format for transit agencies to use for bus procurements. This format is designed to provide a common method of contracting and is intended to save time and effort for the parties to a particular transaction. According to APTA, industry stakeholders such as bus manufacturers and transit agencies contributed to the standard’s development. Use of the standardized format is not required but FTA and APTA encourage its use. Finally, the National Rural Transit Assistance Program (RTAP) offers technical assistance and training, including for procurement, to transit agencies in rural areas. For example, RTAP developed and made available a free web-based application that is designed to guide agencies through FTA procurement procedures as well as provide required federal clauses and certifications. Transit agency officials we spoke to told us that they face an array of challenges in procuring buses. A 2006 FTA report concluded that “the greatest challenges cited by both transit agencies and bus manufacturers were related to bus procurement and contracting.” As discussed previously, to identify and understand procurement challenges faced by transit agencies, we conducted six discussion groups with officials of 36 rural and urban transit agencies. We recruited discussion group participants through mass e-mail solicitation and allocated participants into homogeneous groups based on their location and size of the transit agency. Our six discussion groups included the following: two groups of officials from small, rural transit agencies; two groups of officials from small, urban transit agencies; one group of officials from medium, urban transit agencies; and one group of officials from large, urban transit agencies. Participants in all six discussion groups identified some aspect of FTA’s guidance, as well as some federal legal and regulatory requirements, as posing challenges to procuring new buses. For example: FTA guidance: Participants from four groups noted that some FTA guidance is outdated, as discussed in the previous section. Participants in three groups also cited that getting technical assistance from FTA was challenging. For example, a participant said that they contacted FTA officials about the applicability of a particular requirement but FTA officials would not provide a specific answer, leaving it to the transit agency to interpret. Participants in two rural groups said that they do not have the resources or experience to interpret these FTA requirements. While some participants appreciated the assistance from FTA regional staff, other participants in the large and small urban groups reported variation in the assistance received from FTA; some participants stated that FTA staff gave inconsistent advice or showed a lack of technical expertise. Participants in one group mentioned that FTA could be more proactive and provide more assistance to agencies during the procurement process. Requirements: Participants from two groups described instances where procurement requirements are seemingly unnecessary or not relevant. For example, participants in the discussion groups representing large and medium urban-transit agencies said that some FTA requirements do not reflect the reality of the transit bus market. According to these participants, FTA requires transit agencies to conduct a price analysis to ensure the price is reasonable, but this requirement seems unnecessary when there is only a single manufacturer that builds a specific type of bus. However, FTA does not require a price analysis when there is only a single manufacturer of a bus but does require a different type of analysis. This discrepancy illustrates that some transit agency officials are confused about FTA requirements. Participants in the medium urban group said they find a lack of potential bidders when they attempt to procure buses, and also have difficulty demonstrating that pricing is fair and reasonable, which is required by federal regulations. Buy America: Participants from four groups told us that the Buy America certification requirements can be burdensome, complex, and costly. FTA requires transit agencies procuring buses with federal funds from FTA to certify compliance with Buy America requirements through pre-award and post-delivery reviews of bus manufacturers.Participants in two discussion groups from rural and medium urban transit agencies explained that their agencies needed consultants’ help with required pre- and post-award audits for the Buy America certification. Contract Length: Participants in one small urban group told us that a federal requirement intended to encourage competition among manufacturers creates challenges. Specifically, the federal statutory provision prohibits transit agencies from entering into multiyear rolling- stock contracts with options to buy additional rolling stock or replacement parts for longer than 5 years if federal funds are used. Although FTA officials told us that the requirement is in place to ensure that the market is fair and open, participants told us that the restriction is burdensome; if they decide to procure new buses after the 5th year, they must initiate the procurement process all over again, which can be lengthy and costly. Joint Procurement and “Piggybacking:” While participants in some of the groups we spoke with said piggybacking in particular is a useful procurement tool, other participants cited challenges and limitations associated with piggybacking. For example, one small urban agency said that piggybacking was the only way they could procure a small quantity of buses—such as one or two—because they did not receive any bids in response to a request for proposal. Participants in three groups from rural, small, and medium urban-transit agencies said that the opportunities for joint procurement and piggybacking may be limited because there is no formal mechanism to identify mutually beneficial opportunities for joint procurement or piggybacking and the regulations regarding piggybacking are confusing. As a result, some participants who had previously used piggybacking began to handle their own procurements. Participants in two groups told us that transit agencies are not generally aware of other transit agencies’ procurement plans, and there is no entity to formally help facilitate joint purchases. Some participants said they have learned of these types of procurements through informal means, such as existing relationships with other agencies or at industry conferences, and through representatives of bus manufacturers. In addition, even if agencies identify piggybacking opportunities, some participants found the rules confusing. For example, officials in three discussion groups (rural, small, and medium urban) were confused about whether FTA allows agencies to participate in piggybacking options. FTA officials told us that piggybacking options are allowed as long as an agency unintentionally orders more buses than it requires. Transit agency officials in two groups stated that piggybacking rules limit agencies’ options related to the buses to be procured. These rules allow some changes to the buses originally ordered that are within the general scope of the contract—such as seat fabric or exterior paint color—but prohibit changes outside the scope of the contract. One small urban participant said these rules effectively prevent them from piggybacking because the features his agency prefers to include in their buses—such as an extra wheel-chair securement location—are not features most other agencies include in contracts. Instead, the participant said his agency prefers to conduct its own procurement so it can get the features it needs. In addition to the challenges above, some transit agency officials in our discussion groups cited challenges that specifically affect rural, urban small and medium-sized agencies: Participants in four groups noted that the procurement process is difficult and resource intensive. Participants in four groups (rural, small, and medium urban) noted that smaller agencies tend to have few or no dedicated procurement staff and lack technical, engineering, or specification-writing expertise needed for bus procurement. Several participants noted that the procurement process can take from 6 months to a year. Further, small agencies said that they procure buses infrequently, such as every few years, so it is difficult for them to become familiar with the process. In addition, these procurements are often for 1–3 buses and sometimes generate few bids from manufacturers. Participants in three groups stated that procurement training opportunities through the NTI are not always designed to meet the needs of smaller agencies. Some participants told us that the material in the training courses is targeted to larger agencies, for example by assuming a level of support staff similar to what a larger transit agency would have. When these participants asked questions specific to their situations, they said the course instructors were not able to answer them. Additionally, participants in the rural group explained that with their limited staff resources, it was difficult to spare anyone to attend off-site training for multiple days. Participants in all six discussion groups cited the following challenges stemming from changes in the transit bus industry: Participants in all six discussion groups said there is limited competition to produce heavy-duty buses. Specifically, participants told us that there are currently three heavy-duty bus manufacturers and one that makes 60 foot buses. According to participants in one small urban group, the limited number of heavy-duty bus manufacturers has led to a longer delivery time (18 months) and the price of new buses has increased faster than inflation. Furthermore, participants also noted a lack of vendors for vehicle components. Specifically, participants told us that there is currently only one engine manufacturer for heavy-duty buses that complies with Buy America requirements. Participants in four discussion groups stated that Buy America and DBE requirements and consolidation of the industry have contributed to the lack of available vehicle component vendors. Participants in four discussion groups stated that it was a challenge to secure funding for transit buses. Transit agencies may use federal, state, and local funds to purchase or replace buses but must weigh these purchases against other capital and operating needs. Transit agencies also told us that the federal surface transportation authorizations and appropriations process creates year-to-year uncertainty that can be a challenge for long-term planning. Obtaining state or local funding to meet federal local match requirements was identified as a challenge by participants in three discussion groups. Constrained state and local budgets can make securing these funds difficult. We have previously found that some states provide limited or no state funds for rural transit, a process that increases the pressure on rural transit agencies to secure local funds. Participants in all six discussion groups offered ideas about possible ways that the federal government could address some of the procurement challenges they identified (see table 5). Participants in four discussion groups from rural, urban-medium, and urban small-transit agencies suggested that FTA develop a national procurement schedule similar to that of the GSA’s. GSA establishes long-term government-wide contracts for millions of products with commercial firms and provides federal agencies and others with access to these sources of supply. According to GSA officials, GSA negotiates volume discounted prices and its policies and procedures are designed to ensure that vendors comply with federal procurement requirements. Transit agencies are not permitted to purchase through GSA because they are not federal agencies or other authorized users. In the past, entities other than federal agencies have been authorized to purchase items through GSA. For example, states and units of local government have been authorized pursuant to statute to purchase equipment, including vehicles, in support of counter-drug, homeland security, and emergency response activities. Purchasing through GSA could allow for more streamlined procurement for transit agencies and potentially cost savings because GSA would be responsible for ensuring that any bus procured through GSA complies with federal and FTA requirements. As a result, transit agency officials would not need to develop their own specifications or procurement requirements and would not be required to ensure the buses comply with the Buy America certification. In addition, GSA officials noted that they are usually able to secure very good pricing because the purchasing power of the federal government results in volume discounts. For example, they said the price of vehicles they procure averages 17 percent below the dealer invoice price. Officials said these prices compare favorably to prices paid by other large capacity vehicle purchasers, such as rental car companies. According to FTA and GSA officials, both agencies have explored the feasibility of establishing a process to allow transit agencies to procure buses through GSA and both agencies support the concept. GSA officials said they have the capability to provide a streamlined procurement process to transit agencies that complies with federal and FTA requirements. However, GSA officials told us that Congress would have to authorize transit agencies to purchase through GSA. Neither agency has submitted a legislative proposal to Congress. Buses are an important part of public transit. Transit agencies maintain a fleet of buses to provide transit service. In order to do so, transit agencies purchase buses to either replace their existing fleet or expand service. In 2013, transit agencies used over $1 billion of federal funding to purchase buses. While some large agencies purchase buses each year and have a dedicated procurement staff for doing so, other transit agencies purchase buses only occasionally. Moreover, participants in our discussion groups explained that smaller agencies tend to have few or no dedicated procurement staff and lack expertise needed for bus procurement. Thus, they are challenged to knowledgably and efficiently complete all of the steps required to procure buses with FTA funds. Therefore, FTA guidance and assistance in helping transit agencies procure buses is important. While this guidance is widely available on-line, some FTA guidance is out of date and refers to incorrect regulations and information. As a result, transit agencies may not be able to purchase buses as efficiently as possible, as they may need to spend additional time researching the guidance or they may have to repeat a required step in the procurement process. While transit agencies face several challenges in procuring buses, many of these challenges could be alleviated if agencies were allowed to purchase buses through GSA. By purchasing through GSA, agencies would know that they are getting a bus that complies with federal procurement requirements. In addition, transit agency staff—particularly those at smaller agencies who may lack procurement expertise—would be able purchase buses more easily and efficiently. Specified non-federal entities have been authorized to purchase through GSA in the past. Specifically, states and local governments have been authorized under federal statute to purchase through GSA for emergency equipment and homeland security-related items, among other things. Transit agencies could experience cost savings if they were able to purchase buses through GSA, as agencies will not need to spend as much time on the procurement process. Allowing transit agencies to make such purchases could also lead to cost savings because GSA staff have expertise in price negotiation and they have procured vehicles that, given the volume of purchases, are on average 17 percent below the invoice price. Ultimately, the efficient purchasing of buses may also save money for the federal government, as transit agencies may require fewer federal funds to operate their transit programs. To ensure that transit agencies have appropriate and current guidance to assist them when procuring transit buses, we recommend that the Administrator of FTA update its Best Practices Procurement Manual and assess its other related guidance identified in this report and update that guidance as needed. To provide a more efficient and cost-effective way for transit agencies to procure transit buses while complying with federal procurement requirements, we recommend that the Administrator of FTA, in conjunction with the Administrator of the General Services Administration, submit a legislative proposal to Congress that would authorize transit agencies that are recipients of FTA grants to access GSA sources of supply for the purchase of transit buses. We provided a draft of this report to the Department of Transportation and the General Services Administration for review and comment prior to finalizing the report. The Department of Transportation provided written comments, which are reprinted in appendix II, and agreed with our recommendations. The General Services Administration provided written comments, which are reprinted in appendix III, and agreed with our recommendations. GSA also provided technical comments on the draft that we incorporated as appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Department of Transportation and the General Services Administration. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2834 or wised@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. The objectives of this report were to examine (1) the characteristics of the U.S. transit bus market, including manufacturing capacity and the production of new buses; (2) the federal role in transit bus procurement, including funding, procurement requirements, and oversight and guidance; and, (3) the views of selected transit agencies on challenges, if any, that they face when procuring new buses and any federal actions that could address those challenges. To address all three of our objectives, we reviewed relevant literature from the Federal Transit Administration (FTA), industry, and academia and interviewed officials representing FTA; General Services Administration (GSA); American Public Transportation Association (APTA); Community Transportation Association of America (CTAA); transit agencies; and bus manufacturers. To determine the characteristics of the U.S. transit bus market, we reviewed FTA’s Annual Statistical Summaries from fiscal years 2009 to 2013, the latest year data are available, and American Recovery and Reinvestment Act (Recovery Act) Statistical Summaries. We reviewed data from the National Transit Database (NTD) for the years 2004 and 2013 to evaluate changes to transit agencies’ bus fleets over time. We chose those years because 2013 was the latest data available and 2004 represented data 10 years earlier, which we believe to be sufficient to show any changes in the industry. We included vehicles used for fixed- route, demand-response, and deviated-fixed route services and excluded vehicles used for other services—commuter buses, inter-city buses, trolley or specialty buses, vanpool vans, and automobiles—because these vehicles represent a small portion of all vehicles used. To assess the reliability of data from NTD, we interviewed officials about data quality control procedures and reviewed relevant documentation and determined that the data were sufficiently reliable for the purposes of this report. We categorized buses generally and for the purposes of this report as heavy-duty or cutaway. FTA also uses these categories as well as three others, which we did not include for ease of reporting and because the heavy-duty and cutaway categories encompassed over 95 percent of transit buses, according to FTA in 2006. To identify bus manufacturers, the number of buses produced, and any changes over time, we reviewed NTD urban vehicle inventory reports from 2004 and 2013, the most recent year data are available. Transit agencies receiving federal grant funding from FTA are required to report operating information to NTD, including detailed information on their bus fleet. For each year, we limited the data to reported vehicles that are 1) used to provide bus or demand response transportation services; 2) shown as bus and articulated bus vehicle types; and 3) within a 5-year model range of the year of the data report (2004 to 2000 model years for the 2004 data and 2013 to 2009 for the 2013 data). We further grouped the data set by vehicle length: generally, vehicles 30 feet in length or greater (heavy-duty buses) and vehicles less than 30 feet (cutaway buses). For each group, we identified the number of vehicles from each manufacturer and calculated an approximate market share by dividing that number by the total number of vehicles. Within the data set, some records included missing data, particularly for manufacturer’s name and for model year. We excluded these records from our analysis. The number of vehicles associated with missing data records represented 11 percent of the total number of vehicles in the 2004 data and 10 percent in the 2013 data. We also manually reviewed all records and identified some information that appeared to be incorrect based on other data in the record. Most commonly, vehicles were categorized as a vehicle type van when other evidence, such as the manufacturer name, indicated they should be categorized as a bus. We included these vehicles in our analysis. In more limited cases, vehicles were categorized as a bus when other evidence indicated they should be categorized as a van. We excluded these vehicles from our analysis. The net change in vehicles because of these changes represented less than 3 percent of the total number of vehicles in the 2004 and 2013 analyses. To determine the federal government’s role in procurement of transit buses—including funding, oversight and guidance, and procurement requirements—we reviewed applicable federal law, regulations, guidance, and FTA documentations on federal grant programs including FTA’s Master Agreement; Circular 4220.1E; Circular 4220.1F; Administrator’s Policy Letters; Best Practices Procurement Manual; Triennial, State Management, and Procurement Systems reviews; and APTA’s Standard Bus Procurement Guidelines. To identify any challenges transit agencies face when procuring transit buses and federal actions that could address those challenges, we conducted six discussion groups with 36 rural and urban transit agencies. We recruited discussion group participants through mass e-mail solicitation to approximately 1,900 transit agencies that report to the NTD. In our email, we requested participation from transit agencies that had received funding from FTA since 2009 and that were interested in speaking with us on challenges they faced regarding transit bus procurement. We also solicited the help of three national transit industry groups to send our solicitation e-mail to rural transit agencies— APTA, CTAA, and American Association of State Highway and Transportation Officials (AASHTO) —that work closely with transit and state transportation agencies. For each volunteer received, we identified whether it was an urban or rural recipient, the number of buses each reported, and their NTD identification number to identify the FTA region. From our pool of volunteers, we recruited participants based on a range of criteria, including: agencies that purchase and operate buses; a mix of urban and rural FTA grantees; a mix of the size of bus fleet and type of bus used; and geographic diversity of various FTA regions. In order to include a wide geographical distribution of participating agencies, we conducted all six discussion groups remotely via WebEX. We established three size categories to ensure we received input from a range of sizes of urban agencies: large (more than 100 buses); medium (50–100 buses), and small (less than 50 buses). We did not group rural agencies into size categories because 97 percent of these agencies have less than 50 buses. After grouping the full list of volunteers, we allocated transit agency participants into six homogeneous groups. Our six discussion groups included the following: two groups of rural transit agencies (total of 9 agencies); two groups of small, urban transit agencies (total of 13 agencies); one group of medium, urban transit agencies (total of 6 agencies); one group of large, urban transit agencies (total of 8 agencies). In order to ensure the appropriateness of the questions for the discussions, we pretested our moderator guide with three transit agencies and modified the guide based on those pretest results as needed. We conducted four of the six groups with rural and small urban transit agencies due to the proportion of rural and small agencies in the total population of transit agencies. As we were unable to accommodate all volunteers with our discussion groups, we gave the non-selected volunteers the opportunity to send us written responses to the discussion questions; nine transit agencies provided written responses about the challenges they faced procuring transit buses. Based on our findings from the discussion groups, written responses and a review of literature, we summarized the challenges related to bus procurement into general categories. While these challenges were consistently identified across agencies we spoke with, our findings are based on a self-selecting, non- representative sample of transit agencies, and thus the results are not generalizable to all transit agencies. We conducted this performance audit from October 2014 to September 2015 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. In addition to the contact named above, John W. Shumann (Assistant Director), Melissa Bodeau, Timothy Guinane, Geoffrey Hamilton, Rosa Leung, Erik Shive, Travis Thomson, Elizabeth Wood and William T. Woods made key contributions to this report.

Buses are critical to the nation's public transportation services. According to data from FTA's National Transit Database, buses carry more passengers than all other modes combined. FTA provides grants to transit agencies to buy buses. When making these purchases, agencies must comply with a range of federal requirements. GAO was asked to review the transit bus market and federal role in bus procurement. This report examines (1) the characteristics of the U.S. transit bus market, (2) the federal role in transit bus procurement, and (3) views of selected transit agencies on challenges, if any, agencies face when procuring new buses and federal actions that could address those challenges. GAO reviewed FTA's National Transit Database data from 2004 and 2013 and applicable federal law, regulations, and grant documents; conducted six discussion groups with representatives of 36 rural and urban transit agencies; and interviewed FTA and GSA officials, national transit industry organizations, and transit bus manufacturers. Overall, the number of manufacturers of transit buses has declined in recent years, but bus production has remained constant. Transit agencies purchase over 5,000 buses per year—about half are heavy-duty buses and half are smaller buses called “cutaways” because they consist of a bus body on top of a chassis built by another manufacturer. The number of firms that produce most heavy-duty transit buses declined from 10 in 2004 to 4 in 2013, the latest year data were available, due to business failures and consolidation. In contrast, the number of firms that produce most cutaways increased over the same time frame from 13 to 15. The number of buses procured annually by transit agencies from 2009–2013, using grant funds other than American Recovery and Reinvestment Act (Recovery Act) funds, ranged from 4,670 to 5,652. Transit agencies and others used Recovery Act funds to purchase 7,544 more buses in 2009 and 2010. The amount of Federal Transit Administration's (FTA) grant funding that transit agencies used for bus procurement increased from $794 million in 2009 to $1.3 billion in 2013. Also, agencies spent $1.7 billion in funds from the Recovery Act on buses from 2009 to 2010. FTA oversees transit agencies by requiring them to certify their compliance with a range of federal requirements and by periodic reviews. FTA also provides bus procurement guidance and technical assistance. However, GAO found that some resources provided by FTA, such as its Best Practices Procurement Manual , reference obsolete FTA documents. Without updated guidance, transit agency officials may not be able to purchase buses as efficiently as possible because they may need to spend additional time researching the guidance or to repeat a required step in the procurement process. Transit agency officials GAO spoke with identified a range of challenges they face when procuring buses. Those challenges include difficulties in complying with federal procurement requirements. For example, officials reported that it is time consuming to comply with a requirement to certify that at least 60 percent of the bus's components are made in the U.S. because the agency must conduct pre- and post-award reviews of bus manufacturers. Transit agency officials in four of six discussion groups also identified the procurement process as difficult and resource-intensive, particularly for those transit agencies that do not purchase buses each year and may lack procurement capacity. Transit agency officials in four of six discussion groups stated these challenges could be addressed by allowing agencies access to General Services Administration's (GSA) sources of supply. Purchasing through GSA could allow agencies to decrease the time they spend on bus procurements since GSA would be responsible for ensuring that vehicles comply with federal procurement requirements. Further, purchasing through GSA could result in lower prices for buses, given GSA's ability to purchase vehicles well below the dealer invoice. According to FTA and GSA officials, both agencies have explored the feasibility of establishing a process to allow transit agencies to procure buses through GSA, but neither agency has developed a legislative proposal requesting that Congress grant authority to allow transit agencies to do so. In the past, nonfederal entities, such as state and local governments, have been authorized to purchase items through GSA. GAO recommends FTA update its Best Practices Procurement Manual , assess its other related guidance, and update that guidance as needed. GAO also recommends FTA work with GSA to develop a legislative proposal to authorize transit agencies that receive relevant FTA grants to access GSA sources of supply for the purchase of transit buses. FTA and GSA agreed with our recommendations

Two “secret shopper” surveys of bank and thrift sales of mutual funds have been issued since we released our report. One was done by a private research organization called Prophet Market Research & Consulting and was completed in April 1996. The other was done for FDIC by another research organization, Market Trends, Inc., and was completed May 5, 1996. Both surveys indicated that many banks and thrifts still were not fully disclosing to their customers the risks associated with mutual fund investing. The results of the FDIC-sponsored survey, which was the most comprehensive, indicated that, in about 28 percent of the 3,886 in-person visits, bank and thrift representatives did not disclose to the shoppers that nondeposit investment products, including mutual funds, are not insured by FDIC. The results were worse for the 3,915 telephone contacts—with no disclosure in about 55 percent of the contacts. Similarly, in about 30 percent of the in -person visits, bank and thrift representatives did not inform shoppers that nondeposit investment products were not deposits or other obligations of, or guaranteed by, the institution (about 60 percent nondisclosure for telephone contacts). Finally, in about 9 percent of the in-person visits, bank and thrift representatives did not tell shoppers that their investment was subject to loss, including loss of principal (about 39 percent nondisclosure for telephone contacts). The survey’s findings on the physical location of the mutual fund sales area were nearly the same as ours, with about 37 percent of the institutions not clearly having separated the mutual fund sales area from the deposit-taking area. The survey’s findings reaffirm our earlier findings and indicate that a significant number of banks and thrifts continue to inadequately disclose four basic risks associated with mutual fund investing. Neither the FDIC-sponsored survey nor ours followed the sales process through to the point at which a mutual fund was purchased and an account was opened. However, the interagency guidelines emphasize that bank customers should clearly and fully understand the risks of investing in mutual funds, and that these risks should be orally disclosed to the customer during any sales presentation. Written disclosures or other documentation are to be available to customers during the sales process that may eventually fully inform them of the risks involved. Nevertheless, making these disclosures orally during initial sales presentations is particularly important because written disclosures may not always be read or understood until after the investors’ funds are committed, if at all. In responding to our report, the Federal Reserve and OCC indicated that bank practices generally complied with the interagency guidelines by mid-1995. However, FDIC’s survey results indicated that many banks and thrifts still need to improve their compliance with the guidelines so that their customers are adequately informed of the risks associated with mutual fund investing. According to banking and securities regulators, additional actions are being planned or taken to improve disclosures to bank customers. Some of these actions affect only those banks or thrifts under one regulator’s jurisdiction—such as FDIC’s efforts to improve its data systems to provide its examiners up-to-date information for more targeted examinations, or each regulator’s efforts to improve its examination guidelines. Other efforts are also being undertaken by all four bank and thrift regulators. These interagency efforts include efforts to adopt requirements that bank personnel engaged in the sale of nondeposit investment products take the securities industry’s standard qualifying examinations, better training for bank personnel selling uninsured investment products, reexamination of the interagency policy statement on mutual fund sales. business, and pass relevant qualifications examinations administered for the industry by NASD. The Securities Exchange Act of 1934 excludes banks from its broker-dealer registration requirements. As a result, banks have been able to choose whether to have their own employees sell mutual funds without the need to be associated with a Securities and Exchange Commission (SEC)-registered broker-dealer or subject to NASD oversight. If bank employees are to take NASD’s qualifying examination as the banking regulators propose, they are not to be registered with NASD because they would not be associated with a broker-dealer. However, under the proposal, they will have met the same initial qualifications as NASD-registered representatives. In addition, to maintain their qualifications, they would be subject to the same continuing education requirements imposed on NASD-registered representatives. FDIC officials told us that, in addition to the NASD testing and education requirements, the banking regulators plan to do further training to improve bank and thrift employees’ awareness of the importance of complying with the interagency guidelines. They said that although they found better compliance by NASD-registered representatives, the difference between these representatives and other employees was small, indicating that additional training might help further improve compliance. Banking regulators told us that efforts to reexamine the interagency policy statement are focused on clarifying (1) what situations do or do not constitute a sales presentation and (2) what the institution’s obligation is in assuring that an investment recommendation meets the customer’s needs. An FDIC official told us that the banking regulators want to make the interagency statement less vague so that banks and thrifts can better understand what is expected of them and their employees. restrictions on brokers’ use of confidential financial information from bank or thrift customer files were stricter than the interagency guidance and NASD’s proposed prohibition on the payment of referral fees by broker-dealers to employees of the bank differed from the interagency guidance, which allows payment of these fees. After analyzing nearly 300 comment letters, NASD made changes to its proposed rules. The revised proposal defines confidential financial information and allows its use, but only with the prior written approval of the customer; the prohibition on referral fees remains. NASD forwarded its revised proposal to SEC for approval. SEC published the proposal for public comment and received 86 comment letters by the end of the comment period in May 1996. Most of the letters were from banking organizations or bank-affiliated broker-dealers. SEC is currently analyzing the comment letters before deciding whether to approve the proposed rules. Ensuring that salespersons provide bank customers with appropriate risk disclosures during all mutual fund sales presentations presents a difficult challenge to regulators and to banks and thrifts. Over time, this task may become easier as distinctions among financial service providers continue to fade and customers become more aware of the differences between insured and uninsured products. The bank and securities regulators’ proposed actions for additional training of investment representatives, requiring testing of employees, and reexamining the interagency guidelines should help improve bank and thrift compliance with disclosures required by these guidelines. However, additional steps, which may have the potential to help improve compliance with the risk disclosure guidelines, could also be taken. Such actions, for example, could include regulators (1) continuing to monitor bank and thrift disclosure practices through periodic secret shopper surveys, (2) encouraging banks and thrifts to adopt this kind of testing procedure as part of their own internal compliance audits, if legal concerns can be overcome and it is cost effective; and (3) segmenting and publicizing the results of regulatory reviews of compliance with the interagency guidelines, including the results of secret shopper surveys, when appropriate. sales presentations between customers and bank employees, and they would have difficulty doing so without affecting the customer’s privacy or the performance of the employee. FDIC reported that it plans to evaluate the need for another secret shoppers survey on the basis of the results of bank examinations over the next 2 years. Because of the difficulty in monitoring oral sales presentations through examinations, it seems to us that decisions concerning the need for secret shopper surveys should not be based solely on examination results. Instead, using such surveys to supplement examination results could give banks and thrifts an additional incentive to better ensure that their personnel are providing proper disclosures. Bank regulators told us that some banks are using secret shopper surveys to monitor their own employees. A Federal Reserve official said that banks could make them part of their internal compliance audits. The need for federal regulators to do such surveys may decrease if more banks and thrifts do their own and if disclosure of mutual fund risks improves. Federal regulators could encourage banks and thrifts to adopt these surveys as part of their internal compliance audits if legal concerns can be overcome and it is cost effective. For example, some self-assessment activities, like self-testing, pose a dilemma for lending institutions in that under current law the results of self-testing programs may not be privileged or protected from disclosure to federal regulatory agencies or private litigants. Hence, despite the obvious preventative benefits to be gained from having lenders adopt continuous self-testing programs, many institutions are reluctant to undertake such programs out of fear that the findings could be used as evidence against them, especially by third-party litigants. One way to help resolve this issue would be to remove or diminish the disincentives associated with self-testing by alleviating the legal risks of self-testing when conducted by banks who, in good faith, are seeking to improve their mutual fund risk disclosures. Banking regulators suggested to us that they might also encourage depository institutions to consider methods other than secret shopper surveys to test compliance with disclosure requirements, such as calling their customers to determine if the sales person made the proper disclosures. investors than they are the safety and soundness of a depository institution. Therefore, bank and thrift regulators may want to consider the feasibility of segmenting the results of their reviews of compliance with disclosures required by the interagency guidelines, including the results of any secret shopper surveys, from other examination results and of making those results available to the public. Such segmentation and disclosure is already required in connection with regulators’ assessments of bank and thrift compliance with the Community Reinvestment Act. In summary, the results of our survey and the more recent surveys, indicate that there may be a persistent problem with many banks and thrifts failing to make the basic risk disclosures required under the interagency guidelines. These disclosures are important because they can help investors fully understand the risks of investing in bank mutual funds. Banking regulators and some banks and thrifts are taking steps to better ensure that the required disclosures are made. While these actions are positive, other steps, which may have the potential to help increase compliance with these guidelines and better ensure that investors are adequately informed of the risks of their investment decisions, could also be taken. The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary. VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. U.S. General Accounting Office P.O. Box 6015 Gaithersburg, MD 20884-6015 Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (301) 258-4066, or TDD (301) 413-0006. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists.

GAO discussed the Federal Deposit Insurance Corporation's (FDIC) survey concerning the risks associated with mutual fund investing. GAO noted that: (1) sales of mutual funds through banks and thrifts have increased dramatically; (2) the value of assets managed by these institutions doubled from $219 billion in December 1993, to $420 billion in March 1996; (3) 2,800 banks sold over $40 billion in both proprietary and nonproprietary mutual fund shares during 1995; (4) in February 1994, FDIC, the Office of the Comptroller of the Currency, the Federal Reserve, and the Office of Thrift Supervision jointly issued guidelines on the policies and procedures for selling nondeposit investment products; (5) these interagency guidelines require that bank and thrift customers be fully informed of the risks of investing in mutual funds; (6) the guidelines also require that banks' mutual fund sales activities be physically separated from bank deposit activities; (7) the results of the FDIC survey indicate that many banks and thrifts are not disclosing the risks associated with mutual fund investing; and (8) all four bank and thrift regulators are making an effort to ensure that bank personnel pass qualifying examinations and receive better training in selling uninsured investment products, and reexamine the current interagency policy on mutual fund sales.

The Employee Retirement Income Security Act of 1974 (ERISA) created PBGC as a government agency to help protect the retirement income of U.S. workers with private-sector defined benefit plans by guaranteeing their benefits up to certain legal limits. PBGC administers two separate insurance programs for these plans: a single-employer program and a multiemployer program. The single-employer program covers about 34 million participants in approximately 26,000 plans. The multiemployer program covers 10 million participants in another 1,500 collectively bargained plans that are maintained by two or more unrelated employers. If a multiemployer pension plan is unable to pay guaranteed benefits when due, PBGC will provide financial assistance to the plan, in the form of a loan, so that benefits may continue to be made up to the guaranteed benefit limits. However, if the sponsor of a single-employer plan is in financial distress and does not have sufficient assets to pay guaranteed promised benefits, the plan will be terminated and PBGC will likely become the plan’s trustee, assuming responsibility for paying benefits to participants as they become due, up to the guaranteed benefit limits. Most of PBGC’s $102.5 billion in liabilities are due to future benefit payments owed to participants of underfunded plans terminated under the single- employer insurance program. To finance these liabilities, PBGC currently has approximately $80 billion in assets. PBGC’s funds primarily come from three sources: insurance premiums paid by sponsors of defined benefit plans, assets acquired from terminated plans, and investment income earned on these assets. For example, in 2010 all plans insured by PBGC paid a total of approximately $2.3 billion in premiums. In addition, PBGC took over the assets from 147 defined benefit plans in fiscal year 2010, which totaled approximately $1.8 billion. Finally, over the course of the same year, PBGC recorded $7.8 billion in earnings from its investment portfolio, including interest, dividends, and capital gains. PBGC holds its assets in essentially two separate funds: the PBGC trust fund and the PBGC revolving fund (see fig. 1). The PBGC Trust Fund holds assets received from terminated plans and the return on investing the assets held in the trust fund, while the PBGC Revolving Fund consists of premium receipts and the return on investing the premium receipts. Benefit payments and financial assistance are paid from the revolving fund, and then the trust fund reimburses the revolving fund through a proportional payment at least annually. PBGC has grown significantly since the end of its first year of operation. As of June 30, 1975, the agency had $34 million in assets and $1.2 million in liabilities. At the end of fiscal year 2010, however, the trust and revolving funds combined contained about $80 billion in assets (see fig. 2), to cover total liabilities of $102.5 billion, leaving a deficit of approximately $23 billion. PBGC is governed by a three-member board of directors, which consists of the Secretary of Labor (Chair), the Secretary of the Treasury, and the Secretary of Commerce. The board is responsible for policy direction and oversight of PBGC’s finances and operations. The board of directors is responsible for establishing and overseeing the policies of the corporation, including the approval of the corporation’s investment policy statement. Under its bylaws, the board is required to review the corporation’s investment policy statement at least every two years and approve the investment policy statement at least every four years. Each board member must designate an official (not below the level of an assistant secretary) to support the board’s oversight. Board representatives are given the authority to act for all purposes under the bylaws, subject to some actions—such as approving the corporation’s investment policy statement—being ratified by the board members. The board members often rely on these department representatives to conduct much of their PBGC related work on their behalf. PBGC uses institutional investment management firms to invest its assets, subject to the agency’s oversight and in accordance with the investment policy statement as approved by its board and applicable legal requirements. For example, ERISA provides different requirements concerning how the assets held in the revolving fund and the trust fund can be invested. ERISA requires the PBGC Revolving Fund to be, in part, invested in U.S. obligations. PBGC has more flexibility to invest trust fund assets in other investments, and, along with revolving fund investments, the corporation’s investment policy statement provides direction on how these assets are to be invested. With respect to the trust fund, PBGC does not determine the specific investments to be made, but instead relies on its investment managers’ discretion to invest a portion of the funds consistent with the benchmarks and risk criteria provided to each investment manager. When PBGC receives assets from terminated plans, PBGC determines whether the assets fit into the agency’s current investment policy objectives. For incoming assets that do not fit with their current policy, PBGC uses investment managers to liquidate them, as soon as practicable, and then reinvests the proceeds into assets that do align with PBGC policy. In its role as an insurer, PBGC’s responsibilities are similar to those of other institutions that conduct such functions. However, the corporation also faces structural challenges that are not shared by other insurers, which gives the corporation less control over the terms by which it insures pension plans and constrains its ability to manage its risk of loss (see table 1). For example, in contrast with information provided by pension insurers in Canada and the Netherlands, PBGC tends to have less control over the terms by which it insures pension plans. Only Congress, through legislation, can change premiums or plan funding requirements for defined benefit plans in the United States. Beginning in 2003, recognizing PBGC’s long-term financial challenges, we included PBGC’s single-employer insurance program on our list of “high- risk” programs needing attention and congressional action; in 2009, we added PBGC’s multiemployer program as a program of concern. Both programs remain on our high-risk list today. PBGC’s investment policy has changed frequently since 1990, alternating between more conservative and more aggressive approaches to investment. The frequent changes in policy have had a moderate impact on PBGC’s actual allocation of assets since 1976 because there were no allocation targets in place prior to 1990 and the policy targets after that time were rarely ever met. Meanwhile, the transaction costs for the reinvestment of assets during each policy period have fluctuated with shifts in the market. Since 1990, PBGC has shifted its investment policy five times. The shifts in investment policy that occurred in 1990 and 2004 were aimed at strategies that immunized against interest rate exposure by increasing the allocation of fixed-income securities. PBGC’s investment policy was shifted in 2009 to a more conservative strategy of taking on a higher allocation of fixed- income securities. In contrast, shifts in investment policy that occurred in 1994 and 2008 were aimed at strategies that maximized returns by increasing the allocation of equities. Shifts in policy of this frequency are thought to reflect an undisciplined approach to investing. Experts we interviewed stated that a more disciplined approach would require that PBGC change its investment policy no more than once every 5 to 7 years, except to review the policy during unusual circumstances, such as the recent market crash or when taking over the assets of a large terminated plan. They noted that it can take up to 5 years for a policy to be fully implemented and to have an impact that can be evaluated. Moreover, these experts stated that a long-term and disciplined investment policy is needed in order to minimize the costs associated with shifts in policy. Since 1990, PBGC’s investment policy was in place for more than 5 years only once— during the period from 1994 to 2004. All other policies were in place for shorter periods, generally about 2 to 4 years (see fig. 3). PBGC’s actual allocation of its total assets (both revolving fund and trust fund combined) reflects these changes in policy to some extent, but the impact has been tempered by a number of factors. First, PBGC must comply with certain statutory investment restrictions. Therefore, because PBGC only invest the assets of its revolving fund in U.S. obligations which are fixed-income securities, accomplishing its investment policy goal is, in effect, limited to reallocating the assets of its trust fund —that is assets acquired from terminated plans under PBGC trusteeship. Second, during the period between 2004 and 2008, PBGC adopted the practice of using only assets of newly terminated plans to move toward new allocation targets, rather than reinvesting assets already in its trust fund. When in place, this practice further limited the amount of assets PBGC could use to meet its target allocations. Third, market conditions, at times, hindered PBGC in reaching its allocation targets by reducing the overall value—and as a result, the proportion—of assets invested in a particular sector. Finally, the frequency with which allocation targets changed also affected PBGC’s ability to make significant changes in its allocation. During each period a policy was in place, PBGC made progress toward reaching new allocation targets with varying success before a new policy was adopted. In May 1990, PBGC adopted a new investment policy calling for a decrease in the proportion of equities to no more than 35 percent of its portfolio. This policy was initiated by PBGC’s then newly appointed executive director in response to both an increase in unfunded liabilities and to the results of a commissioned study that examined PBGC’s liabilities and investment options. The study of PBGC’s trust and revolving funds together recommended that PBGC reduce its equity exposure and increase its allocation in long-duration fixed-income assets. Accordingly, PBGC adopted a new investment policy that focused on matching its assets with its liabilities and targeted an asset allocation of no more than 35 percent in equities and no less than 65 percent in fixed income. In 4 months, PBGC decreased its equity allocation from 43 percent to 33 percent and was able to maintain this allocation range throughout the period for which the 1990 policy was in effect. In October 1994, PBGC adopted a new investment policy that focused on maximizing the return on its investments by investing more heavily in equities in order to reduce the agency’s deficit by achieving higher rates of return. Although no explicit asset allocation was specified in the investment policy statement, PBGC’s 1994 annual report stated that, along with the adoption of the new investment policy, the agency had raised its ceiling for its equity allocation to 50 percent. The assets in PBGC’s revolving fund is, pursuant partially to statute and partially to PBGC policy, invested only in U.S obligations which are fixed-income assets, hindering PBGC’s efforts to increase its overall equity allocation. However, in the years that followed, the agency attempted to raise equity levels by investing all its trust fund assets into equities. In this way, over the course of fiscal year 1994, PBGC increased its actual equity allocation from 17 percent to 30 percent. During the subsequent 7 years this policy was in place, PBGC’s equity allocation peaked in 1999 at 44 percent, and over the period, averaged about 35 percent according to data provided by the agency. In 2004, PBGC adopted a new investment policy that would, similar to the 1990 policy, match PBGC’s assets to its liabilities by emphasizing fixed- income investments and limiting exposure to market risk. The 2004 policy reduced the allocation target for equities down to the 15 to 25 percent range and raised the allocation target for fixed-income securities to 75 to 85 percent. To implement this policy, however, PBGC’s board directed staff to use only assets acquired from newly terminated plans, rather than to transition core trust fund assets already under management. As a result, according to PBGC officials, the volume of assets available to transition toward the target allocations was limited and the agency was not able to lower its allocation of equities down to the target range during the time this policy was in effect. In 2006, PBGC adopted a new policy as a result of its biennial review process. It allowed PBGC to invest in international securities, a departure from the past. The agency’s overall investment policy, however, remained the same, with equity allocation targets set at 15 to 25 percent and fixed- income allocation targets set at 75 to 85 percent. Despite this new policy, once again, PBGC officials said that the agency did not receive enough in newly trusteed assets to be able to shift its equity allocation down to this target range. Also, during most of this period, the returns on PBGC’s equity investments outpaced those of its fixed-income investments, further hindering the agency’s attempt to reach this allocation target. Equities were achieving returns of 11 to 17 percent in fiscal years 2006 and 2007, while the returns of its fixed-income investments were around 1 to 3 percent annually. Hence, according to PBGC, the actual allocation hovered between 27 to 32 percent in equities and 67 to 72 percent in fixed income throughout this period. PBGC changed its investment policy again in 2008 with the goal of seeking to maximize returns on its investment. To this end, PBGC adopted an investment policy with target asset allocations of 45 percent in equities; 45 percent in fixed income; and 10 percent in alternative investments, such as real estate and private equity. In addition, the policy called for expansion into two new subclasses of fixed-income securities: high yield and emerging market debt. In February 2008, when the policy was adopted, 28 percent of PBGC’s assets were invested in equities. To move quickly toward its newly adopted allocation targets, PBGC decided to abandon its practice of relying only on newly acquired assets from terminated plans to transitioning a portion of core trust fund assets as well. PBGC transitioned nearly $5.7 billion from its existing trust fund investments in fixed-income securities to equities. Despite these efforts, the financial crisis and a 35 percent decline in the New York Stock Exchange Composite Index between early February 2008 and May 2009 caused PBGC’s actual equity allocation to drop to as low as 23 percent during this period. In May and June 2009, PBGC’s three board members issued a resolution instructing staff to cease implementing the 2008 investment policy. This resolution was in response to an investigation, conducted by PBGC’s Inspector General, concerning potential conflicts of interest involving PBGC’s then Director with securing asset managers for the agency’s portfolio. Transactions already initiated were allowed to proceed, but no new transactions were permitted until the board representatives issued investment policy guidance in July 2009, since the board had not also issued a new investment policy statement after it ceased the 2008 policy. Instead, this new interim policy called for a return to the actual portfolio composition as it was on March 31, 2009, which was 26.5 percent in equities and 73.5 percent in fixed income. This interim guidance served as the official policy. Since then, PBGC has transitioned its newly acquired assets to fixed-income investments. Nevertheless, the performance of the equities market improved enough that as of September 2010, equities made up 31 percent of PBGC’s portfolio. While the actual distribution of PBGC assets has remained within a fairly narrow range since 1990, the transaction costs incurred for the reinvestment of assets during each period a policy was in place have fluctuated with shifts in the market. Some transaction costs are always incurred with the assumption of assets from newly terminated plans and with the management of existing investments, but the magnitude of these costs can vary dramatically depending on the volume and type of assets being transitioned, the investment policy or goal in place, and the market conditions during the transition period. PBGC does not have a routine process for tracking the transaction costs associated with different investment policies, and does not consider these costs when developing new investment strategies. Transaction costs for reinvestment of assets generally consist of commissions, fees and certain taxes (referred to as explicit costs), and opportunity costs, due to market changes during the transaction (referred to as implicit costs). PBGC typically uses specialized transition investment managers when transitioning large pools of assets to keep explicit costs down through economies of scale and by taking advantage of other services offered by these managers. However, opportunity costs can vary widely based on market conditions, and can result in either a net loss or a net gain. Taking both explicit and implicit costs together, when transactions net an amount lower than the original value of the assets, a loss occurs; when transactions net an amount greater than the original value of the assets, a gain occurs. Although PBGC does not routinely track and conduct analytics on the transaction costs associated with implementing different investment policies, we were able to compile the costs incurred during each period a policy was in place from 2004 forward by obtaining records from PBGC officials as well as PBGC’s external transition managers, as summarized in table 2. From 2004 to 2008, PBGC’s investment policy remained primarily the same: to transition assets from newly terminated plans to increase the level of fixed-income investments. When the 2004 policy was being implemented, assets valued at $8.8 billion were transitioned, and positive market conditions helped PBGC realize a net gain of $40.5 million (or 46 basis points). When the 2006 policy was being implemented, assets of about $2.6 billion were transitioned, but declining market conditions towards the end of this period contributed to a loss of $7.6 million (or 30 basis points). In 2008, PBGC’s investment policy shifted to increasing the level of equity investments and certain subclasses of fixed-income securities and the agency opted to use assets already in the trust fund, as well as newly terminated plan assets, to accelerate implementation of the policy. In total, assets of about $13 billion were transitioned while this investment policy was in place, with $5.4 billion moving from fixed-income securities to equities and $7.6 billion moving from one type of fixed-income securities to others (specifically, from long-duration securities to high- yield and emerging market debt). These transactions were completed in three phases. According to PBGC’s own records, phase one was performed in an “ad hoc” manner and transaction costs were not tracked. Assets transitioned during this phase totaled approximately $3.7 billion. Phase two was more structured (referred to as “coordinated sales”), with PBGC assigning each fixed-income investment manager an amount of trust fund assets to sell over a 5-month period, allowing trades to be made on favorable trading days at the discretion of the investment manager. About $7.9 billion in assets were transitioned during this phase. During phase three, termed the “runoff” phase, the 2008 policy had been suspended, but PBGC officials told us they decided not to cancel the trades for about $1.4 billion in assets that their investment managers already had initiated. Due in part to the market downturn during the period the 2008 policy was in place, the transaction costs associated with asset trades of about $9.3 billion that were tracked during the last two phases of the transition totaled nearly $74.6 million (or 80 basis points). According to one PBGC investment manager, some trades related to the 2008 transition incurred opportunity costs of 400 to 500 basis points. In July 2009, a new interim directive was issued to decrease the level of equity investments back to the asset distribution held as of March 31, 2009. PBGC staff estimated that implementing this new policy could incur transaction costs of as much as $52 million. In January 2011, PBGC provided data indicating that between June 2009 and September 2010, $7.4 million in transaction costs had accrued since implementation of this 2009 directive. Our analysis of PBGC’s investment performance found that PBGC’s investments performed better than most on an asset-only basis compared with the seven benchmark portfolios (see table 3). However, PBGC’s investment portfolio tended to underperform these benchmarks when returns were assessed together with the liability return (or growth in liabilities). Specifically, in the asset-only comparison, PBGC’s portfolio achieved better risk-adjusted performance on its investments than that achieved by six of the seven benchmark portfolios. When assessed with liabilities, however, all seven benchmark portfolios performed better than PBGC’s investment portfolio. This occurred for either one of two reasons: either the benchmark had a mix of assets that were better correlated (that is, moved more in tandem) with PBGC’s liability return (growth in liability), or, when this was not the case, the benchmarks had returns sufficient to compensate for the lower correlations for the period examined. The best performing benchmark (the Pension Protection Act benchmark) incorporated elements of both features, with a mix of relatively high returns on assets and relatively high correlation of their assets with PBGC’s liabilities. Our analysis looks at the single historical period from 1976 to 2009, since the purpose of the analysis is performance assessment, not asset allocation recommendations. Typically, analyses for the purpose of asset allocation would project forward over multiple potential future economic scenarios in order to more fully assess potential risk and reward. The various alternative static portfolios used in this report were analyzed for the purpose of a “what-if” analysis—a historical comparison of alternative investment strategies versus the fluctuating asset allocation that PBGC actually employed—they were not for the purpose of recommending a particular asset allocation going forward. Further, the fact that a particular portfolio performed well over the 1976 to 2009 period in this particular analysis does not necessarily mean that such a portfolio would be appropriate for PBGC going forward. We assessed performance by calculating risk-adjusted returns for PBGC’s portfolio and for each benchmark, where higher returns improve performance while higher volatility reduces performance. The comparative benchmarks used for this analysis represent a range of equity and fixed-income allocations. Six of the benchmarks are largely static (fixed) allocations among asset classes; however, we also included one Dynamic Benchmark that had allocations that varied among asset classes over time. Our analysis of PBGC’s investment returns for the period 1976 to December 2009 found that, on an asset-only basis, PBGC’s portfolio achieved better risk-adjusted performance on its investments than that achieved by six of the seven benchmark portfolios. Specifically, our analysis found that on an asset-only basis, PBGC’s portfolio outperformed five of six fixed-allocation benchmarks, as well as the Dynamic Benchmark. In each instance, the results were maintained regardless of whether or not PBGC investment returns were net of investment expenses. Within this framework, the PBGC and benchmark portfolios were evaluated solely on how well the assets performed relative to the risks taken. For details see appendix III. When consideration of changes in liabilities was included in our analysis, we found that PBGC’s investments did not perform as well as the seven benchmark portfolios. PBGC must cover the liabilities from the underfunded plans it trustees in order to pay benefits to participants and beneficiaries. Other than the premiums assessed on plan sponsors that are statutorily set, the only revenue that PBGC has to cover its liabilities is the return on the assets it manages. Given this context, analyzing PBGC’s investment performance in a framework that explicitly incorporates liabilities provides useful information. We found that PBGC’s investments underperformed all seven of the benchmark portfolios on a risk-adjusted basis when the returns were analyzed net of the liability return. In simple terms, this means that all seven of the constructed benchmarks had a mix of assets with some combination of risk, return, and correlation levels that made their investment strategies achieve a higher level of risk-adjusted performance than PBGC’s investment policy for the 1976 to 2009 period. This occurred because either the benchmark portfolio had a mix of assets that had a higher correlation with the liability return, or, in cases where the correlations were lower, the benchmark portfolio had sufficient returns to compensate for the lower correlations for the period we examined. However, the dynamic portfolio, which maintains the same asset allocation as the PBGC total fund, performed as well as the S&P 500 benchmark and out performed the Barclays Capital and Post Fiscal Year 2002 benchmarks as well as the PBGC portfolio—three portfolios that have significant allocations to bonds. (For additional information, see app. III). According to our analysis, the best performing portfolio for the 1976 to 2009 period was the PPA Benchmark Portfolio, with a mix of 40 percent bonds and 60 percent equities. Because, this analysis is strictly based on past performance, this result does not guarantee or imply that a PPA-like portfolio will perform better than the current PBGC allocation going forward. Moreover, the PPA benchmark and other portfolios with a significant weighting toward equity would likely not perform as well if incoming cash-flows from new plan terminations were included in the analysis. Rather than determining a particular asset allocation, this analysis highlights that an approach that was not only mindful of returns, but also accounted for the correlation between asset returns and the liability return, was more likely to result in an investment policy for PBGC that achieved higher risk-adjusted performance for the 1976 to 2009 period. Our analysis found no link between the frequent changes in PBGC’s investment policy since 1990 and the actual allocation between equity and fixed-asset investments. This is because while the stated policy shifts were significant, changes to the actual allocation were moderate. Hence, changes remained within a narrow range of a portfolio mix between fixed- income and equity allocations. As a result, although some shifts in actual allocations did occur, we found no conclusive evidence that fluctuations in the proportional allocation between equity and fixed-income investments had a notable adverse impact on PBGC returns. This was the case for both types of analysis—asset-only and assets net the liability return. Finally, in the assets net of liability context, our finding that PBGC’s portfolio underperformed relative to the Dynamic Benchmark suggests that factors other than asset allocation are causing the underperformance—such factors could include the inflows of new assets, timing of shifts to meet allocation goals and their associated costs, or could reflect that there are no costs or fees in the Dynamic Benchmark. However, detailed information would be required to determine the reasons for the underperformance of the PBGC total fund relative to the Dynamic Benchmark. In our review of PBGC’s internal documents, we found that the agency has largely functioned without complete investment policy statements and operating procedures. Compared to industry-recommended standards for pension funds and insurance companies, PBGC’s investment policy statements are missing important provisions that provide implementation guidance. Further, PBGC staff have largely functioned without the benefit of fully developed and documented operating procedures. The investment policies issued by PBGC’s board for strategic guidance in the planning and execution of investments have generally lacked a number of provisions recommended for sound investment management or have been insufficiently detailed to provide adequate guidance for staff concerning certain investment objectives. One expert we interviewed stated that while PBGC is unique and may not be obligated to articulate the same policy provisions as other institutions with similar responsibilities—such as foreign pension insurers, domestic pension funds, and private insurance companies—the agency faces similar investment problems, opportunities, and solutions as many investment programs do. Hence, it is equally important for PBGC to have a well- developed investment policy statement as it is for these other institutions. According to one expert, “an investment policy statement (IPS) is a foundational document for a pension fund’s investment program. The essential purposes of the IPS are to articulate the consensus view of the board regarding the overall investment program and to document policies and procedures regarding major issues.” However, we found items included in the PBGC’s policy statements often are insufficiently detailed to provide adequate guidance for staff concerning certain investment objectives. For example, we found that prior to 1990, PBGC operated without a formal investment policy statement, and that the six different policy statements the PBGC board has issued since then have been silent in many areas cited as important by professional organizations such as the Chartered Financial Analyst Institute, the Association of Public Pension Fund Auditors, and the Foundation for Fiduciary Studies. We compiled a list of 25 items these organizations recommended be included in an investment policy statement in order to provide sound strategic guidance across the key areas of governance, investment objectives, and risk management. We then examined PBGC’s policy statements against these items and found certain items were often missing (see table 4). The agency’s 2008 policy statement has been the most thorough to date (including 15 of the items) while PBGC’s most recent investment guidance, adopted by board representatives in 2009, included the fewest to date (only 6 of the items). Further, some of the provisions that were covered were, according to some staff, insufficiently detailed to offer adequate guidance. In the governance area, PBGC’s investment policy statements have not assigned responsibility for managing, monitoring, and reporting on portfolio risk. According to PBGC officials, those responsibilities were either informally communicated to staff or staff assumed responsibility for these activities on their own. Further, while most of PBGC’s statements include a discussion of hiring and monitoring asset managers, they do not assign responsibility for these tasks to a specific group. By contrast, the investment policy statement of the United Kingdom’s pension insurer, Pension Protection Fund, and most of the public pension plans that we reviewed do assign responsibility for these tasks to specific groups, such as the public plan’s investment advisory committee. Also, while PBGC’s investment policy statements assign responsibility for the execution of the investment program, they generally do not assign responsibility for developing or monitoring the implementation of the policy. According to statute, the PBGC board is responsible for establishing policy. In addition, the board has an oversight responsibility to ensure that PBGC is executing the board’s policy in appropriate ways. According to PBGC staff, because of the lack of specific guidance in the policy statements, there have been instances when staff have had to request further policy guidance from PBGC’s board and the board had not always been responsive. For example, in 2004, the board had instructed staff to limit costs by using only incoming assets to transition to the new allocation target. When adherence to this directive, together with a low level of liquid, incoming assets caused the agency to miss its new allocation targets, staff told us they asked for guidance but did not receive it. More recently, in May and June 2009, the board members issued a resolution directing staff to cease implementation of the 2008 investment policy, but did not approve a new investment policy statement and did not provide further investment guidance. In response, PBGC’s Corporate Investments Department’s (CID) staff wrote a memo to PBGC’s acting director indicating that they were concerned about the lack of a defined policy to provide direction to CID staff with respect to asset allocation. Principal areas of concern outlined were: (1) oversight and management, (2) investment of newly trusteed assets, and (3) asset allocation risk. Subsequently, policy guidance was provided by the board representatives until a new investment policy statement was approved by the board. In addition, while we have found that the board and board representatives are meeting more frequently than in the past, we could find no formal oversight or formal feedback mechanism in place for the board and board representatives—a mechanism that is a necessary element for ensuring that PBGC is executing the policy in appropriate ways. According to one expert we interviewed, the inventory of critical subjects regarding an investment program is extensive, and the board is ultimately responsible for assessing and overseeing all of them. Some of the key elements the expert noted that should require the board’s focus include clearly articulated governance policies; a comprehensive, written investment policy statement; a well thought out asset allocation process; clearly defined and appropriate measures; monitoring processes; and monitoring of investment costs. Although PBGC staff told us that these things were accomplished below the level of the board members, we could find no documentation that indicated that such a formal oversight mechanism was in place. We reviewed decades of board meeting notes—up through the most recent meetings—in search of such evidence, but could find none. In the area of investment objectives, PBGC’s statements have remained silent with respect to several items, such as return targets and statements of risk tolerance. By comparison, the United Kingdom’s Pension Protection Fund board, in its policy statement, has specifically set a long- term target investment return of 1.8 percent above liabilities and a risk level equivalent to a tracking error of 4 percent against liabilities. The Pension Protection Fund also identified nine risks that might affect its investments and identified approaches to mitigate those risks. Six of the eight public pension plans we reviewed also included a return target and a risk tolerance in their investment policies. One expert stressed, in particular, the importance of documenting tolerance for risk in the investment policy cautioning that without such documentation, a firm risks making changes at a bad time (selling at a deep discount) or in response to political pressure. In order to keep the investment policy out of the political realm, a well-documented, long-term, and disciplined view with an effective governing board is necessary, while following a well established allocation model that keeps long term perspective in mind. In the area of risk management, although most items were covered in PBGC’s policy statements, almost all lacked sufficient detail to provide adequate guidance. For example, the cost management provision of PBGC’s statements generally identified the types of investment expenses involved and offered a low-cost policy for investing, but did not provide guidance on how to monitor these costs. As noted by some experts, ultimately, investing is not about seeking returns but about managing risks, with well-grounded policies to ensure adequate monitoring of risks over time. Typically missing from PBGC’s investment policy statements has been the practice of portfolio rebalancing. A provision for rebalancing was provided for the first time in 2008. All of the public pension plans that we studied included such tolerance ranges. Most also specified a time frame for rebalancing or assigned responsibility for determining a course of action. The PBGC’s CID staff has largely operated without fully developed and documented operating procedures, although it has recently begun to create them. According to a PBGC staff member, the mission of the CID is twofold: (1) to transition newly trusteed assets into PBGC’s investment portfolio and (2) to manage PBGC assets. Further, to transition newly trusteed assets into PBGC’s investment portfolio, CID staff are responsible for transferring assets so that they are commingled in compliance with PBGC policies, and are consistent with PBGC’s asset allocation. However, the staff member also said that PBGC historically has not had formal procedures for executing the investment policy and transitioning assets. As a result, according to PBGC’s Inspector General, when the former board established the 2008 investment policy, certain tasks were not performed in the proper order by CID staff. For example, according to PBGC’s Inspector General, PBGC had actually undergone several transition related activities—such as the selection of three investment management firms for strategic partnership contracts for managing $2.5 billion in PBGC assets—before risks and mitigating methods related to the transition were even documented. In addition, CID staff provided a group of documents covering a number of transition related activities that had several notable weaknesses. For example, these documents indicated timelines for implementation, but provided no risk analysis, accountability measures to monitor progress, or a delineation of roles and corresponding responsibilities related to the transition. According to a PBGC staff member, to manage PBGC’s assets, at a high level, CID staff are responsible for five operational tasks: (1) select, hire, and terminate investment managers; (2) oversee managers: (3) oversee the aggregate investment program: (4) implement board asset allocation and any other board investment policy; and (5) oversee all aspects of the PBGC investment programs including cash management and securities lending. In 2010, CID staff began to draft more complete working procedures for their investment operations, however, PBGC’s CID staff and the Inspector General recently told us that this effort has been a slow undertaking. PBGC’s CID staff stated that creating procedures takes away from their ability to do their mission work and, thus far, they have only been able to provide preliminary and incomplete drafts of some of the needed procedures. However, while complete operational procedures are lacking for most of the operational tasks under the purview of the CID, PBGC’s CID staff have recently completed a draft compendium of formal procedures that detail processes and procedures for managing their securities lending program—the smallest program operated within the CID. According to one expert, well functioning operational policies and procedures are an essential mechanism for ensuring linkages between a fund’s governance structure, which includes policy making, and its management systems. This expert wrote that with regard to operational policies, directors should (1) identify and address aspects of the fund’s investment operations, organization, and portfolio necessary to control undue risk and expenses, minimize inefficiency, and achieve the desired long-term return; (2) evaluate the fund’s organization and procedures relative to those of its peers and industry best practices; and (3) find ways to enhance public trust and confidence in the pension insurance system. The board must oversee and approve such policies and procedures. PBGC has grown from a relatively small agency with about $34 million in assets in its first year after its establishment in 1974, to one with almost $80 billion in assets in fiscal year 2010. As the agency has grown, so too has the frequency of changes to its investment policies. The agency’s policies and procedures for asset management still reflect its small agency past. Indeed, there are few formally documented procedures and the investment policy statements are insufficiently detailed for the agency to manage its investments and apply the investment policy consistently during a transition period and during times of political change. Without a detailed investment policy and formal investment procedures, the agency operates in an environment that is ripe for costly transactions and sub-par returns. When factoring in the frequent changes to the investment policy with the incomplete policies and procedures, a picture emerges that suggests PBGC lacks a disciplined approach to investing—an unsettling picture of an agency with responsibility for a large asset portfolio and a challenging financial future. As the guarantor of basic pension benefits for 44 million Americans, PBGC must take a more disciplined and long-term approach to investment by developing and adhering to a long-term comprehensive investment policy and developing a complete compendium of operational policies and procedures. Well-functioning operational policies and procedures are an essential mechanism for ensuring linkages between pension funds’ governance structure and management systems. Current work under way by PBGC’s CID staff to develop such policies and procedure is an important first step, but greater commitment is needed from both the PBGC board and its management to assure that PBGC can effectively and consistently meet its obligation to conduct the many investment related functions it performs. We are making the following two recommendations: 1. To ensure a disciplined and long-term approach to investment, we recommend PBGC and its board of directors develop and maintain a comprehensive investment policy statement that provides clear organizational accountability, well-defined goals, and risk management parameters. 2. To ensure proper stewardship of PBGC’s assets and effective implementation of its investment policy, we recommend that PBGC develop a complete set of operating procedures and guidelines consistent with recognized best practices in industry and government. We obtained written comments on a draft of this report from PBGC and from the Department of Labor (Labor), which are reproduced in appendixes IV and V, respectively. PBGC and Labor also provided technical comments, which we incorporated into the report as appropriate. PBGC and Labor generally concurred with our recommendations and outlined actions the agency has taken to address many of the concerns we raised. For example, PBGC and its board recently issued a more comprehensive investment policy statement that has incorporated many of the policy items that we identified as missing from previously issued policy statements. In addition, PBGC is in the process of developing a complete set of operating procedures and guidelines. We are pleased to learn of the steps already taken and those underway to address our recommendations. In our view, these initial actions and continued efforts to implement our recommendations fully can only strengthen the stewardship of PBGC’s investments to better assure that PBGC can effectively and consistently meet its obligation to conduct the many investment-related functions it performs. Underscoring its concern with the importance of PBGC’s mission, Labor highlighted the increased oversight activity by the current board, its representatives and their staffs. The Secretary noted that the board also exercises its oversight responsibilities through monthly transition and investment reports written documentation and other activities. We acknowledge this increased oversight and appreciate the efforts by the current board to play a greater role in monitoring the PBGC. The increased oversight by the current board members and their representatives indeed represents an improvement in the way policies and processes are adopted and overseen at the PBGC, but we believe such improvements must be documented and institutionalized to assure that such levels of effort are sustained through subsequent boards. Our prior recommendations to Congress to improve governance at the PBGC through an expanded and restructured board continue to be needed to assure that such appropriate and continuous oversight is carried out, not only today but in the future. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of the report to the Secretary of Labor, the Director of the PBGC, and other interested parties. We will also make copies available to others on request. This report is also available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions regarding this report, please contact Barbara Bovbjerg at (202) 525-7215 or bovbergb@gao.gov. Contact points for our Congressional Relations and Office of Public Affairs can be found on the last page of this report. Key contributors are listed in appendix VI. To determine how Pension Benefit Guaranty Corporation’s (PBGC) investment objectives have changed over time and whether policy goals have been met, we collected and reviewed investment policies used by PBGC from 1990 through the policy dated October 2009. We started our review with PBGC’s 1990 investment policy because it was the first investment policy that specified asset allocation targets, such as the proportion of assets to be invested in fixed-income assets versus equities. For each of these policies, we identified the overall objective, such as whether the policy attempted to maximize earnings using a higher proportion of equities or reduce risk by increasing the proportion of fixed- income securities matched to the duration of their liabilities. We also identified the percentages of each type of asset required by the policy, such as the percentage allocated to equities versus fixed-income investments, and compared these target allocations to actual allocations as stated in PBGC’s annual reports, internal trust and revolving fund data, and other financial information received from PBGC officials. We then looked at the conditions leading up to each change in policy, such as changes in investment philosophy, incoming assets from terminated plans, and changes in leadership at the executive director level. We obtained this information through interviews with PBGC officials, and from other information provided by the agency, including internal memos, e-mails, inspector general audit reports, summary information prepared for board and advisory committee members, asset and liability studies, and other reports and memos prepared by various PBGC investment managers and consultants. We also performed a detailed review of PBGC board and advisory committee meeting minutes using NVivo content analysis software. To identify and summarize discussions related to investment policy development, review, and implementation, we reviewed investment policy statements, related investment advisory committee meeting notes and documentation, and board meeting notes when available. We obtained and reviewed this information from PBGC’s inception in 1974 through the current policy, but focused our analysis on the period between 1990 and the 2009 policy because the policies in place during that period were the focus of our review. We also interviewed past PBGC directors, board representatives, advisory committee members, and the PBGC Inspector General. We also reviewed relevant federal laws and regulations. To determine how PBGC’s changes in investment policy compared to other entities, we interviewed officials from several pension consultants, investment and transition managers, life and property and casualty insurers, and large state pension plans. We also interviewed and reviewed investment policy-related information provided by foreign pension insurers in Canada, the Netherlands, Switzerland, and the United Kingdom. However, we did not conduct an independent legal analysis to verify the information provided by state pension plans or foreign pension insurers. Finally, we reviewed past our work on PBGC’s investment policies and oversight structure. To determine how PBGC transitions assets between investment policies and the resulting costs, we interviewed PBGC officials responsible for transitioning assets and reviewed transition related documentation provided by PBGC officials, as available. For transitions where data was not available from PBGC, we obtained this information directly from PBGC’s transition managers and interviewed officials from those firms to determine both implicit and explicit transaction costs associated with changing investment policies. We also looked at the procedures and costs associated with transitioning assets from terminated plans taken over by PBGC to determine whether or not it was possible to separate these costs from costs associated with changing policies. We interviewed the PBGC Inspector General and past PBGC directors to obtain additional information about PBGC’s transition related policies and other adopted practices. In order to understand asset transitions more generally, we interviewed transition and investment managers, financial industry consultants, and officials at several large state pension plans. We also looked at market conditions and returns on equity and fixed-income investments during the periods in which PBGC was transitioning assets. We limited our analysis of transaction costs to the policies in place from 2004 through 2009 because of the lack of detailed cost data available from PBGC and their transition managers for transactions made prior to the 2004 policy. To assess the performance of PBGC’s investments, we conducted a portfolio performance evaluation of the agency’s Single-Employer Total Fund monthly returns from the period October 1976 to December 2009. This analysis focused on the single-employer program, which accounted for 96 percent—or $21.08 billion—of the $21.95 billion total deficit from the single-employer and multiemployer programs, as of September 30, 2009. For those portions of this analysis involving PBGC liabilities, we used data on the liabilities associated with (terminated) trusteed plans within the single-employer program. The Single-Employer Total Fund represents the pool of trusteed assets that supports the liabilities associated with terminated defined benefit plans that have been trusteed by PBGC. As part of the portfolio performance evaluation, we compared PBGC’s Total Fund portfolio return performance to the returns on several well- diversified benchmark portfolios via a number of portfolio performance statistics. We selected well-diversified benchmark portfolios for the portfolio performance evaluation to ensure that the variability of the benchmark portfolio returns almost exclusively represented systematic risk and not the idiosyncratic risk associated with individual securities. Also, the portfolios were selected such that they represented exposure to the systematic risks that are reflected in the returns on several specific, broad asset classes. The asset classes are the domestic equity asset class (United States), the foreign equity asset class, the short maturity, risk-free asset class, and the long maturity fixed-income asset class. These particular asset classes were chosen because they are the ones emphasized in asset allocation data provided by PBGC. The benchmark portfolios used in this analysis are also distinguished by whether their asset class composition varies dynamically over time (“dynamic” composition portfolios) or is constant over time (“static” compositions portfolios). The benchmark portfolios and their characteristics are as follows: S&P 500. Asset class composition: 100 percent equities. This is a static composition portfolio that represents the equity asset class. Wilshire 5000. Asset class composition: 100 percent equities. This is a static composition portfolio. It represents the equity asset class with a greater allocation to smaller capitalization stocks than the S&P 500. Barclays Capital Long-Term Government Credit Index. Asset class composition: 100 percent fixed income. This is a static composition portfolio representing the fixed-income asset class, including both corporate and U.S. government fixed-income asset classes. Pension Protection Act Benchmark Portfolio. Asset class composition: 60 percent equities and 40 percent fixed income. This is a static composition portfolio. Life Insurance Benchmark. Asset class composition: 85 percent fixed income and 15 percent equities. This is a static composition portfolio, and is intended as a stylized representation of the asset portfolio typically held by life insurance firms in their general accounts (with grouping mortgage assets into the fixed-income category). Post Fiscal Year 2002 Benchmark. Asset class composition: 30 percent equities, 60 percent fixed income, and 10 percent cash. This is a static composition portfolio, and is intended as a stylized representation of the average asset allocation of the PBGC total fund during what is later termed “asset allocation period 4.” This roughly corresponds to the period from beginning of fiscal year 2002 to the present. Dynamic Benchmark. Asset class composition: equivalent to the asset class composition for the PBGC Total Fund. This is a dynamic composition portfolio, where the asset allocation varies over time in such a fashion so as to match that of the PBGC Total Fund for the broad asset classes domestic equity, foreign equity, fixed income, and riskless short maturity fixed-income assets (e.g., cash). The purpose of the Dynamic Benchmark in the PBGC Total Fund portfolio performance evaluation is to reflect the systematic risk exposure of the PBGC Total Fund as closely as possible while at the same time abstracting from any active tactical asset allocation undertaken by the PBGC Total Fund management, such as tactical allocations in specific subsectors within an asset class or investments in specific individual securities. The comparisons allowed us to analyze various aspects of the PBGC Total Fund’s risk-adjusted performance. Given that the primary function of PBGC is to support its liabilities—the pension benefits associated with terminated, trusteed plans—the portfolio performance evaluation was conducted using asset-only returns and asset returns net of the liability return. To determine how well PBGC’s investment policies and operations comport with best practices in the industry, we interviewed PBGC’s Inspector General, current PBGC board member’s representatives, and PBGC staff. We also reviewed relevant federal laws and regulations. To evaluate PBGC’s operational guidelines and procedures, we obtained procedures manuals and documents that PBGC’s staff uses to manage and oversee their operations. To evaluate PBGC’s investment policy statements against industry best practices, we obtained information and documentation of actual practices used by industry experts, foreign pension insurers in Canada, the Netherlands, Switzerland, and the United Kingdom, investment committee documents from large state pension plan providers, and a property and casualty insurance provider. To identify a list of items that could be included in an investment policy we first conducted a literature search for documents with guidance on investment policy statements. We found documents from expert organizations which provide standards that financial industry professionals follow to ensure they are meeting the fiduciary requirements under relevant state and federal laws. These organizations include the Chartered Financial Analyst Institute, the Foundation for Fiduciary Studies, the Association of Pension Plan Fund Auditors, the Government Finance Officers Association, and Independent Fiduciary Services. We started with the Chartered Financial Analyst Institute’s documents and listed elements of an investment policy identified in a document created by the institute and then compared that list to elements identified in the documents we reviewed created by other organizations. We also considered the investment policy statements of other entities and the elements that were frequently found in those statements. In our list, we kept items that were mentioned in more than one of the documents from the five expert organizations. We also added one item, transition policy, which was not found in the documents we used but we believe that it is specific and unique to the mission of PBGC since the agency transitions assets and liabilities from the defined benefit plans that are terminated. This list contains elements that multiple industry organizations have identified as desirable elements of investment policy statements, but, should not be considered an exhaustive, customized checklist. While we believe that PBGC should have some of the items contained in these lists, because every investor is unique, the actual items that PBGC should include in its investment policy needs to be tailored to their particular needs and situation. Receipt of newly-terminated plan assets is a multi-step process. Assets are evaluated by an analyst with PBGC’s Corporate Investment Department (CID).CID policy calls for various documents to be compiled into a file (including, for example, a plan asset listing, investment statements, trusteeship agreement, contact information), records receipt of the plan in CID’s plan tracking worksheet, and assigns the plan to a CID analyst. The analyst then reviews the file and makes contact with the party/parties that have custody of the assets (typically more than one) to initiate the transfer, and a plan asset transfer methodology is determined. PBGC officials noted that it is CID priority to transfer all assets in-kind, but that is not always permitted (per contractual agreements between the former plan sponsor and the asset custodian and/or proprietary investment products) or optimal (for example, with small dollar mutual funds). To transfer the assets, the analyst prepares a direction letter that will include a copy of the trusteeship agreement and transfer instructions at a minimum. This letter is signed by authorized PBGC personnel and sent to the asset custodian. The assets are then transferred to PBGC’s asset custodian and placed in a holding account until liquidity is determined and a certain dollar threshold is met. Illiquid assets, such as real property, are generally transferred to PBGC's Special Situation manager, where the manager seeks liquidation of the asset in a timely manner. Private equity (generally in the form of limited partnerships) is transferred to one of PBGC's private market overseers. We conducted a portfolio performance evaluation of the PBGC Single- Employer Total Fund monthly returns from the period October 1976 to December 2009. This analysis focused on the single-employer program, which accounted for 96 percent—or $21.08 billion—of the $21.95 billion total deficit from the single-employer and multiemployer programs, as of September 30, 2009. For those portions of this analysis involving PBGC liabilities, we used data on the liabilities associated with (terminated) trusteed plans within the single-employer program. The Single-Employer Total Fund represents the pool of trusteed assets that supports the liabilities associated with terminated defined benefit plans that have been trusteed by PBGC. As part of the portfolio performance evaluation, we compared PBGC’s Total Fund portfolio return performance to the returns on several well- diversified benchmark portfolios via a number of portfolio performance statistics. The comparisons allowed us to analyze various aspects of the PBGC Total Fund’s risk-adjusted performance. Given that the primary function of PBGC is to support its liabilities—the pension benefits associated with terminated, trusteed plans—the portfolio performance evaluation was conducted using asset-only returns and asset returns net of the liability return. The liability return refers to the rate of growth in the total value of the then-existing liabilities or terminated benefits, (i.e., exclusive of newly terminated plans). In computing the asset returns net of the liability return, we use what we term the “scaled” liability return—the product of the liability return and the inverse of the funding ratio (PBGC Total Fund aggregate assets to PBGC total fund aggregate liabilities). Our analysis also entailed examining patterns in the PBGC Total Fund’s asset allocations (PBGC Total Fund portfolio “weights” across asset classes) over time in order to assess the effect of fluctuations in the PBGC Total Fund asset allocations on the performance of the PBGC Total Fund. This analysis included characterizing the behavior of the PBGC Total Fund portfolio weights and identifying asset allocation periods in the PBGC Total Fund. The result of this analysis was used in selecting some of the benchmark portfolios. The results immediately below provide a two-way comparison, on an asset-only basis, of the PBGC Total Returns to the Dynamic Benchmark— the two dynamic portfolios among those included in our portfolio performance evaluation analyses. Due to the design of the Dynamic Benchmark, these results reflect PBGC Total Fund under- or over- performance linked to influences other than the asset class allocation, such as asset allocations to specific subsectors within an asset class or investments in specific securities. Then, in the following subsection, we assess the effect of variation in the PBGC Total Fund’s asset allocation in an asset-only context by comparing the performance of the Dynamic Benchmark and the PBGC Total Fund against the static benchmark portfolios. Special emphasis was placed on analyzing the differences in performance between the portfolios that have particularly strong performance and the two portfolios with dynamic asset allocations. This special emphasis allows us to then assess whether the time variation in the asset allocations associated with the PBGC Total Fund and the Dynamic Benchmark appeared to hurt or help their risk-adjusted performance. Also, we examine whether the data suggests other aspects of asset allocation aside from variation in portfolio weights that might bolster or harm risk- adjusted return performance. The performance statistics for this section are shown in table 6 unless otherwise noted. The phrases “decade subperiods” and “decade” will be used to denote the four subperiods—October 1976 through December 1979, 1980–1989, 1990–1999, and 2000–2009 for which statistical estimates are shown in table 6. The results summarized in table 6 indicate that, over the particular historical period studied, the PBGC Total Fund outperformed the Dynamic Benchmark on a risk-adjusted basis, where risk is measured in terms of the volatility of month returns. In particular, all risk adjusted measures (Sharpe, Adjusted Sharpe, Sortino, and Omega ratios) are slightly higher for the PBGC Total Fund than for the Dynamic Benchmark for the overall period. For those three decade subperiods where the Sharpe ratio is positive for the PBGC Total Fund (1980–1989, 1990–1999, and 2000–2009) the PBGC Total Fund outperformed the Dynamic Benchmark for two out of the three subperiods (1990–1999, and 2000–2009). In the sub-period where the Sharpe and Sortino ratios are negative, the Adjusted Sharpe measure for the PBGC Total Fund is greater than that of the Dynamic Benchmark, again indicating that the PBGC Total Fund outperformed the Dynamic Benchmark in that period. Disaggregation of the PBGC Total Fund’s return performance statistics reveals that PBGC Total Fund returns tended to underperform the Dynamic Benchmark returns on a risk-adjusted basis when the Total Fund’s allocation to equities is higher, not lower. For instance, during allocation period 1—when the PBGC Total Fund equity allocation had an upward trend and the fixed-income allocation had a downward trend—the PBGC Total Fund Sharpe ratio was below that of the Dynamic Benchmark. In contrast, the PBGC Total Fund outperformed the Dynamic Benchmark in terms of the Sharpe ratio in allocation period 2, when the total fund equity allocation was falling. Also, the average allocation to equities was lower in allocation regime 4 than it was in allocation period 3, and the PBGC Total Fund outperformed the Dynamic Benchmark in allocation period 4 but slightly underperformed the Dynamic Benchmark in allocation period 3. The Sortino and Omega ratios show similar (and more pronounced in the case of the Omega ratio) under- or over-performance patterns across the weight regimes, thus corroborating and affirming the Sharpe ratio results. Looking more closely at potential sources of the under- or over- performance of the PBGC Total Fund returns versus the Dynamic Benchmark returns, a driver of the PBGC Total Fund’s under- or over- performance appears to be the mean of the returns, inasmuch as the pattern of under- or over-performance in the risk-adjusted return metrics across decades matches that of the pattern of under- or over-performance in the mean. This holds whether one views the disaggregations by “decade” or by allocation period. According to our results, the Dynamic Benchmark outperformed every static benchmark except the PPA and the Post Fiscal Year 2002 Benchmark—the benchmark portfolio with composition that roughly reflects the PBGC Total Fund’s portfolio allocation during allocation period 4—while the PBGC Total Fund outperformed all the same benchmarks the Dynamic Benchmark did as well as the PPA. In addition, another finding is that all of the static benchmarks that involve mixtures of fixed income and equity securities outperform those benchmarks that allocate all funds to either bonds or equities. With regard to question of whether fluctuations in asset allocations had an adverse impact on PBGC Total Fund returns, the variable nature of the results precludes concluding that the time variation in asset allocations necessarily adversely impacted the PBGC Total Fund return performance. Both the PBGC Total Fund and the Dynamic Benchmark have fluctuating asset allocations, yet both have higher values for the risk-adjusted performance metrics—Sharpe, Adjusted Sharpe, Omega, and Sortino ratios—than the majority of the fixed allocation portfolios. On the other hand, there is one fixed-allocation benchmark—the Post Fiscal Year 2002 Benchmark portfolio—that, for the overall period, had risk-adjusted performance metrics that were superior to both the PBGC Total Fund and the Dynamic Benchmark. However, even this fixed allocation portfolio is really based upon the PBGC Total Fund, for the portfolio weights for the Post Fiscal Year 2002 Benchmark portfolio are a stylized representation of the PBGC Total Fund weights over the course of allocation period 4. In addition, note that, despite having fluctuating asset allocations, the PBGC Total Fund has outperformed the Post Fiscal Year 2002 Benchmark—in the sense of having higher risk-adjusted performance measure values— over significant subperiods of time in the past. For instance, the PBGC Total Fund has performed better than the Post Fiscal Year 2002 Benchmark portfolio on a risk adjusted basis for two out of the four “decade” subperiods—that is, the subperiods 1990–1999 and 2000–2009 for a total of 20 years out of the 33 1/4 years from October 1976 to December 2009. Thus, when returns on assets are considered in isolation from the growth in the liabilities, we did not find strong support in the data to indicate that the variations in the PBGC asset allocations adversely impacted the PBGC Total Funds’ performance. Lack of diversification across asset classes appeared to have a more adverse impact on risk-adjusted performance. Additionally, the portfolios with 100 percent allocations to equities had some undesirable characteristics. Notably, out of the eight portfolios considered in the portfolio performance analysis, all the portfolios that represented a single asset class were among the bottom half of the portfolios in terms of the Sharpe ratio for the entire data sample, including the 100 percent fixed- income benchmark. The dominant contributing factor to their relatively weak risk-adjusted return performance is risk—all three had among the highest standard deviation and kurtosis scores for the entire historical sample period. The two portfolios that were 100 percent equities—the S&P 500 and the Wilshire 5000—had an additional undesirable feature: negative skewness. These two had the “most negative” skewness scores for the total sample period out of the eight portfolios. The combination of magnified negative skewness and kurtosis evident in the returns of the two equity benchmark portfolios is undesirable because it implies that investment in such portfolios has the potential to contribute of extreme negative returns. Although both equity portfolios have the highest average returns for the overall sample, the relatively low Sharpe ratio scores associated with them imply that they do not offer enough reward per unit of risk—that is, robust enough reward to risk trade-offs—to outperform those portfolios, both static and dynamic, that contain a diversified mix of both bonds and equities. A two-way comparison of the PBGC Total Fund and the Dynamic Benchmark enabled us to evaluate PBGC Total Fund under- or over- performance associated with factors other than the PBGC Total Fund asset allocation. Next we examine the impact of fluctuations in the PBGC Total Fund’s asset allocation in the asset-liability context. A comparison of the PBGC Total Fund net-of-liability return performance to the net-of-liability return performance of the Dynamic Benchmark indicates that the PBGC Total Fund portfolio underperforms the Dynamic Benchmark in risk-adjusted terms on an asset-liability basis. In contrast to the results for the asset-only analysis, the PBGC Total Fund had weaker performance than the Dynamic Benchmark in that its Adjusted Sharpe ratio was lower for the overall time period (October 1976 to December 2009). Despite the switch in the performance rankings of the PBGC Total Fund and the Dynamic Benchmark, there are many similarities between the asset-liability performance analysis results and the asset-only performance analysis returns. The areas of similarity are as follows: 1. Under- or over-performance pattern across “decade” and asset allocation periods. The PBGC Total Fund underperformed the Dynamic Benchmark for two out of the four decade subperiods and two out of the four asset allocation regimes on a risk-adjusted, net-of- liability return basis, according to the Adjusted Sharpe ratio statistic values. 2. Lack of materiality of investment expenses. Deducting investment expenses from the PBGC Total Fund returns in the asset-liability context did not affect the performance ranking of the PBGC Total Fund relative to the Dynamic Benchmark on an Adjusted Sharpe ratio basis (in those periods for which investment expenses data were available). 3. Tendency to perform relatively worse in regimes where equity asset allocation is greater or rising. The PBGC Total Fund’s relative performance has tended to be worse in asset allocations periods where there is an elevated or rising allocation to the equity asset class. For example, as in the asset-only analysis, the PBGC Total Fund returns, net of the liability returns, had an Adjusted Sharpe ratio below that of the Dynamic Benchmark in allocation period 1 (which was characterized by a rising allocations to the equity sector). Also, as in the asset-only case, the PBGC Total Fund underperformed the Dynamic Benchmark on a net of liability return basis in allocation period 3, according to the Adjusted Sharpe ratio scores. In allocation period 4, when the average allocation to equities in the PBGC Total Fund portfolio was lower than in allocation regime 3, the PBGC Total Fund had a higher Adjusted Sharpe ratio than the Dynamic Benchmark did. However, unlike the asset-only case, the PBGC Total Fund Adjusted Sharpe ratio was less than that of the Dynamic Benchmark in allocation period 2, when the PBGC Total Fund allocation to equities was falling and to bonds was rising. The similarity of the seemingly inverse relation between the PBGC Total Fund Adjusted Sharpe ratio value and the magnitude of the allocation to the equities asset class reinforces the impression that elevated allocations of the PBGC Total Fund to the equity asset class had adverse impact on PBGC Total Fund returns net of the liability returns in an asset-liability context as well as when the portfolio returns are considered in isolation from the liability returns in an asset-only context. However the patterns in the equity asset allocation and its relationship to performance should not be viewed as implying causality. 4. Mean excess return prominence as a driver of risk-adjusted performance metric values across subperiods. In every sub-period and asset allocation regime where the excess mean return (net of the liability return) for the PBGC Total Fund exceeded the excess mean return (net of the liability return) for the Dynamic Benchmark portfolio, the Adjusted Sharpe ratio for the PBGC Total Fund exceeded the Adjusted Sharpe ratio for the Dynamic Benchmark. Given all of the similarities between the results of the performance comparisons in the asset-liability and asset-only contexts, the reason for the contrast between the PBGC Total Fund’s underperformance of the Dynamic Benchmark in the asset-liability context and its outperformance in the asset-only context appears to be risk. In the asset-only performance comparison analysis, the PBGC Total Fund’s riskiness—as measured by the standard deviation and semi-standard deviation of the returns—-was lower than that of the Dynamic Benchmark portfolio. However, in the asset-liability context, the results indicate that the PBGC Total Fund returns have greater standard deviation and semi-standard deviation values than the returns for the Dynamic Benchmark, suggesting that the PBGC Total Fund returns (net of the liability returns) are riskier and more volatile than the Dynamic Benchmark returns (net of the liability returns). One factor that likely played a role in elevating the PBGC Total Fund’s riskiness above that of the Dynamic Benchmark is the correlation of the actual asset returns with the liability returns (not the correlation between the liability returns and the asset returns net of the liability returns). For the overall sample period, the PBGC Total Fund actual returns had lower correlation with the liability returns —both scaled by the funding ratio and unscaled—than the Dynamic Benchmark. Higher correlation makes it more likely that movements in the liability return are accompanied by movements in the asset portfolio return in the same direction and of similar magnitude. Such co-movement of the actual asset returns and the liability returns helps reduce the volatility of the asset returns net of the liability returns. Lower correlation has the opposite effect of higher correlation—lower correlation reduces co-movement between asset returns and liability returns and thus elevates the riskiness of asset returns net of the liability returns. Thus, the extent to which the riskiness of the PBGC Total Fund exceeds the riskiness of the Dynamic Benchmark on a net-of-liability return basis probably reflects, at least in part, the extent to which the liability returns are less correlated with the PBGC Total Fund’s actual returns than with the Dynamic Benchmark actual returns. However, the question of why the PBGC Total Fund would be less correlated with liability returns than the Dynamic Benchmark would require a more detailed investigation. The results of comparing the performance measures of the PBGC Total Fund and the Dynamic Benchmark returns, net of the liability return, to the performance measures of the static portfolios are that the Dynamic Benchmark outperforms two out of the six static portfolios—the Post Fiscal Year 2002 Benchmark and the Barclays Capital Long-Term Government Credit Index—and is roughly tied in performance with the S&P 500. However, the PBGC Total Fund did not outperform any of the benchmarks. Moreover, two out of the three best-performing portfolios have significant allocations to bonds and equities versus representing only a single asset class. In order to detect potential sources of underperformance, as measured by the Adjusted Sharpe measure, in the PBGC Total Fund and the Dynamic Benchmark, we conducted a detailed comparison of various performance metrics for these two portfolios to performance metrics for the PPA benchmark portfolio—the portfolio that had the highest Adjusted Sharpe measure for the overall October 1976 through December 2009 sample period and, by that measure, the best risk-adjusted performance. This detailed comparison suggests that a major source of the underperformance of the PBGC Total Fund and the Dynamic Benchmark relative to the PPA benchmark portfolio was weakness in the mean excess return, for both portfolios had lower mean excess returns for the overall sample period the mean excess return of the PPA benchmark portfolio. However, both the PBGC Total Fund and the Dynamic Benchmark portfolios appeared to be less risky than the PPA portfolio inasmuch as the returns on both portfolios had lower standard deviation and semi- standard deviation than the returns on the PPA portfolio. Thus, the primary immediate reason both portfolios have lower Adjusted Sharpe ratios than the PPA benchmark is that their returns (net of the liability return) had lower mean excess returns than the PPA benchmark return (net of the liability return) not because they were more risky than the PPA benchmark. One other feature of the results that underscores the extent to which both portfolios were less risky than the PPA benchmark on an asset-liability basis is that the returns (net of the liability return) for both portfolios had lower standard deviations than the return (net of the liability return) for the PPA benchmark for every decade sub-period in the case of the Dynamic Benchmark and for three out of the four decade subperiods in the case of the PBGC Total Fund. Although the statistics clearly suggest that weakness in the mean excess return played a role in lowering the risk-adjusted performance of both the PBGC Total Fund and the Dynamic Benchmark portfolios, the evidence provided by the performance measures about whether the variation over time in asset allocations associated with both portfolios necessarily lowered the risk-adjusted performance of their returns on a net-of-liability basis is less clear. For example, on the one hand, the Dynamic Benchmark has a lower Adjusted Sharpe ratio for the overall sample period—and by that metric, weaker risk-adjusted performance—than several static portfolios; however, on the other hand, it also outperforms other fixed allocation portfolios on an Adjusted Sharpe ratio basis, which suggests that fluctuations in asset allocations alone do not immediately imply underperformance on a risk-adjusted basis. In general, the evidence from the performance metrics is too ambiguous to support the conclusion that the variation in the asset allocations inherent in the PBGC Total Fund and the Dynamic Benchmark portfolio necessarily lowered the risk-adjusted performance of the returns of both portfolios within the asset-liability analysis. Furthermore, there are subperiods where the returns (net of the liability returns) for both dynamic portfolios have higher Adjusted Sharpe ratios than the PPA benchmark, even though this portfolio had the highest Adjusted Sharpe ratio for the overall sample period. In particular, both the PBGC Total Fund and the Dynamic Benchmark have higher Adjusted Sharpe ratios than the PPA benchmark for two out of the four decade subperiods; also, the Adjusted Sharpe ratios for both portfolios exceed that of the PPA benchmark for one of the four asset allocation regimes, a period of 8 years. The lengths of time over which the PBGC Total Fund and the Dynamic Benchmark outperform, on a risk-adjusted basis, the best static portfolio over significant subperiods of the overall sample period does not indicate that the fluctuations in the asset allocations for the PBGC Total Fund and the Dynamic Benchmark alone are a preeminent source of weakness in the risk-adjusted performance of the returns for both portfolios in the asset-liability context. This analysis has primarily (although not exclusively) focused on the comparison of the risk-adjusted performance of the two dynamic portfolios to the performance of the PPA benchmark, the static portfolio which had the strongest risk-adjusted performance. However, if the focus is expanded to include comparisons across all of the static and dynamic portfolios, another feature of the results emerges. That is, the influence of the correlation between the portfolio returns and the liability return on the riskiness and the risk-adjusted performance of the portfolio returns net of the liability return. This influence is emphasized in the results for the best three performing portfolios over the period studied—the PPA benchmark, the Wilshire 5000, and the Life Insurance Benchmark portfolio performance results. The PPA benchmark and the Life Insurance Benchmark both have adjusted Sharpe ratios that equal or exceed that of the Wilshire 5000 for the overall sample period even though the mean excess return of the Wilshire 5000 for the overall sample period is 49 percent greater than that of the Life Insurance Benchmark and 31 percent greater than that of the PPA benchmark. Because both the PPA and the Life Insurance Benchmark have lower mean excess returns than the Wilshire 5000, the source of their strong adjusted Sharpe ratio performance in comparison to the Wilshire 5000 rests in the riskiness of the returns for those two portfolios (in comparison to the Wilshire 5000). As shown in table 7, both portfolios have distinctly lower standard deviations and semi-standard deviations—two measures of riskiness— than the Wilshire 5000. Specifically, the annualized standard deviation of the Life Insurance Benchmark returns is 49 percent less than that of the Wilshire 5000 returns, and the semi-standard deviation of the Life Insurance Benchmark portfolio returns is 46 percent less than the semi- standard deviation of the Wilshire 5000 returns. Similarly, the annualized standard deviation of the PPA benchmark returns is 32 percent less than the annualized standard deviation of the Wilshire 5000 returns, and the annualized semi-standard deviation of the PPA benchmark returns is 30 percent less than the semi-standard deviation of the Wilshire 5000 returns. By examining the correlation of the PPA benchmark, the Life Insurance Benchmark, and the Wilshire 5000 returns with the liability return in conjunction with the standard deviation values for the returns on those three portfolios, one can observe the potential role of the correlation in reducing the riskiness of the PPA and Life Insurance Benchmark return (net of the liability returns). In particular, the correlation of the returns on both the PPA and Life Insurance Benchmark portfolios with the liability return are distinctly higher than the correlation of the return on the Wilshire 5000 with the liability return. As shown in table 7, the correlation of the Wilshire 5000 return with the scaled liability return is 0.29, but the analogous correlation statistic for the PPA benchmark return is 0.51 (76 percent higher than the Wilshire 5000 correlation statistic) and for the Life Insurance Benchmark portfolio is 0.74 (155 percent higher than the Wilshire 5000 correlation statistic). Furthermore, it appears that, the higher the correlation, the lower the risk, since the benchmark portfolio that has returns with the highest correlation with the liability return—the Life Insurance Benchmark—has the least risk. When considering all eight portfolios being studied (instead of only the three portfolios with the strongest risk-adjusted performance), the four portfolios with the highest correlations with the liability return have the four lowest standard deviations, and the four portfolios with the lowest correlations have the four highest standard deviation estimates. Also, with the exception of the two portfolios with the highest correlation scores and lowest standard deviation values, the rankings of the standard deviation values matches the rankings of the correlation estimates (in ascending order by standard deviation and descending order by correlation). The strong relation between extent of correlation with the liability return and risk highlights how the relatively strong correlation of the PPA and the Life Insurance Benchmark returns with the liability return seems to contribute to lowering the riskiness of the returns (net of the liability return) of these two portfolios, enhancing their Adjusted Sharpe ratio values and risk- adjusted performance (according to the Adjusted Sharpe ratio statistic). The apparent linkage between the risk-adjusted performance metric and the correlation between the actual portfolio return and the liability return is most likely a reflection of the effect of the correlation between the actual portfolio returns and the liability returns on the volatility of the portfolio returns net of the liability return. As was discussed in the comparison between the performance of the PBGC Total Fund and the Dynamic Benchmark, higher correlation between the (actual) portfolio returns and the liability returns implies a greater degree of co-movement between the asset returns and the liability returns, and greater co- movement between the asset returns and the liability returns weakens or lowers the volatility of the portfolio returns net of the liability return. One reflection of the lowered volatility for the portfolio returns net of the liability return is a lowered standard deviation value, and a lower standard deviation value helps elevate the Adjusted Sharpe ratio value, implying stronger risk-adjusted performance. The strong relation between the correlation statistic and the Adjusted Sharpe measure values provide at least a partial explanation of why two out of the three portfolios that have the best risk-adjusted performance (as indicated by their Adjusted Sharpe ratio scores) all have allocations to the bond asset class sector of 40 percent or more. The portfolio of the liabilities consists mostly of annuities and annuity-like instruments, all of which are obligations of the PBGC. To the extent that annuities are fixed- income contracts, the portfolio of liabilities is essentially bond-like in nature. Hence, the fact that the asset portfolios that have a significant allocation to bonds have return behavior that is more similar to, and thus would have higher correlation with, the liability returns is not surprising. Among the three portfolios that have the returns with the strongest risk- adjusted performance, the returns for the two portfolios that have the highest allocation to bonds (the PPA benchmark and the Life Insurance Benchmark) also have other desirable characteristics. For instance, these two portfolios (as opposed to the Wilshire 5000) have returns which, net of the liability return, have higher skewness for the overall sample than the Wilshire 5000. The higher skewness of the returns for the PPA benchmark and the Life Insurance Benchmark portfolios suggests that those portfolios are less likely to have months where the return on the asset portfolio falls extremely short of the growth in the PBGC liabilities than the Wilshire 5000. Minimization of the instances where the asset returns fall extremely short of the liability returns would help prevent further growth of the already sizeable PBGC funding deficit. The desirable implications of the higher skewness can be seen through other statistics. Note that the minimum values for the PPA Benchmark and the Life Insurance Benchmark portfolios are less extreme. That is, they are less negative than for the Wilshire 5000. To get a sense of how much “less extreme” they are, note that the minimum monthly net-of-liability return for the PPA Benchmark is negative 12.39 percent and for the Life Insurance Benchmark is negative 15.09 percent in contrast to the minimum negative return value for the Wilshire 5000 of negative 24.42 percent. Because returns on the two highest-performing benchmark portfolios with significant allocations to bonds—the PPA Benchmark and the Insurance Benchmark portfolios—have less extreme negative values, lower semi-standard deviations, and lower standard deviations than the returns on Wilshire 5000, there is a strong possibility that the probability distributions associated with these returns have assign less probability to on extreme negative values than the probability distribution associated with the Wilshire 5000 returns, characteristics that are consistent with and are associated with the PPA and Life Insurance Benchmark portfolio returns having skewness statistic estimates that exceed the (negative) skewness statistic for the Wilshire 5000 returns. The fact that the three best performing portfolios over the 1976 through 2009 period in this particular analysis were the PPA benchmark, the Wilshire 5000, and the Life Insurance Benchmark does not necessarily mean that any of these portfolios would be appropriate for the PBGC going forward. The results of any particular analysis will depend on the performance statistics used and how these performance statistics balance risk and reward, and criteria would also need to be established for meaningful differences in these measurements. Also, as noted earlier, an asset allocation exercise would look at multiple possible future scenarios, not one particular historical period. High allocation to equities would be a particularly controversial matter for the PBGC because of the lower, and potentially, negative correlation between equity returns and new claims. The various alternative static portfolios used in this report were analyzed for the purpose of a “what-if” analysis—a historical comparison of alternative investment strategies versus the fluctuating asset allocation that the PBGC actually employed; they were not for the purpose of recommending a particular static asset allocation going forward. In addition to the above, Orice Williams-Brown, Charles A. Jeszeck, Thomas McCool, Frank Todisco, Serena Agoro- Menyang, James Bennett, Susan Bernstein, Lawrance Evans Jr., Charles Ford, Kimberley M. Granger-Heath, Michael Hoffman, Gene Kuehneman, Sheila McCoy, Michael Morris, Christopher Ross, Margie Shields, and Craig Winslow made important contributions to this report. Amenc, Noel and Veronique Le Sourd. Portfolio Theory And Performance Analysis. New York: John Wiley & Sons, 2003. Bacon, Carl R. Practical Portfolio Performance Measurement And Attribution. New York: John Wiley & Sons, 2008. Brealey, Richard A. and Stewart Myers. Principles Of Corporate Finance. New York: McGraw-Hill, Inc. 1981. Dowd, Kevin. Measuring Market Risk. New York: John Wiley & Sons, 2005. Israelsen, Craig L. “Refining The Sharpe Ratio.” The Journal Of Performance Measurement. (Spring 2009): 23–27. Maginn, John L., Donald Tuttle, Dennis McLeavey, and Jerald Pinto (eds.). Managing Investment Portfolios. New York: John Wiley & Sons, 2007. Matlab R2010a Documentation. Natick, MA: Mathworks, Inc., 2010. Pearson, Neil D. Risk Budgeting. New York: John Wiley & Sons, 2002. Sharpe, William F. “Budgeting And Monitoring Pension Fund Risk.” Financial Analysts Journal. Vol. 58 (2002): 74–86. Sharpe, William F. “Mutual Fund Performance.” Journal of Business. Vol. 39 (1966): 119–138. Sharpe, William F. and Lawrence Tint. “Liabilities—A New Approach.” Journal of Portfolio Management. Vol. 16 (1990): 5–10. Sortino, Frank A. and Price, Lee N. Price, “Performance Measurement in a Downside Risk Framework.” Journal of Investing. Vol. 3 (1994): 59–64. Waring, Barton M. “Liability-Relative Investing II.” Journal of Portfolio Management. Vol. 31 (2004): 40–53.

The Pension Benefit Guaranty Corporation's (PBGC) insures the pension benefits of more than 44 million people. Since its inception in 1974, PBGC's assets have grown from about $34 million to almost $80 billion in 2010, largely through assets received in plan terminations. Despite significant swings in PBGC's investment history, there has been little focus on the extent to which it has met its investment goals, the nature of its investment policies or how they compare with best practices in the industry. GAO examined (1) how PBGC's investment objectives have changed over time and the outcomes associated with those changes, (2) the performance of PBGC's investments, and (3) how well PBGC's investment policies and operations comport with best practices in the industry. To address these questions, GAO reviewed PBGC's investment policy statements and operational procedures; analyzed data on investments; and interviewed PBGC officials, officials from several state pension plans and foreign pension insurers, and other experts. PBGC's investment objectives and stated asset allocation targets have changed frequently in the last 8 years, alternating between more conservative and more aggressive approaches to investing. Yet these changes in stated objectives had only a moderate effect on PBGC's actual asset allocation because, for a variety of reasons, PBGC did not meet its targets. In our review of their investment history, we found that PBGC did not routinely monitor transaction costs related to its policy shifts and, at certain times, significant transaction costs were incurred. For example, we determined based on data obtained from PBGC's investment managers that nearly $75 million in transaction costs were incurred during the economic downturn which coincided with the period when the 2008 policy was being implemented and subsequently suspended. Using seven benchmarks, one of which was a Pension Protection Act benchmark that GAO constructed, GAO's analysis shows that PBGC's investments performed better than most benchmarks on an asset-only basis, but tended to underperform all seven of the benchmarks when returns were assessed together with the growth in liabilities. GAO notes that both analyses have limitations and can be seen by some experts as incomplete. However, GAO's method of analysis is consistent with how financial economics literature suggests investment performance analysis should be conducted. Finally, GAO's analysis found no apparent adverse effect on PBGC's investment performance as a result of changes in policy. PBGC's policy statements and operating procedures are incomplete and do not provide sufficient guidance to ensure sound implementation of its investment policies. The investment policies issued by PBGC's board for strategic guidance in the planning and execution of investments have generally lacked a number of provisions recommended by the Chartered Financial Analyst Institute; Independent Fiduciary Services; and other experts of sound investment management, such as the Government Finance Officers Association. Moreover, according to our review and based on interviews with PBGC staff, the policy statements have been insufficiently detailed to provide adequate guidance for staff. In addition, PBGC's Corporate Investments Department's staff have largely functioned without the benefit of fully-developed and documented operating procedures. Although PBGC has grown from a relatively small agency to one holding almost $80 billion in assets, its policies and procedures still reflect in many ways its small agency past. To ensure that PBGC can effectively and consistently meet its obligation to manage a fund of this size and its liabilities, PBGC's board and its management must enact better stewardship, standards, and procedures to ensure that PBGC can effectively and consistently meet its obligation to conduct the many investment related functions it performs. GAO recommends that the PBGC and its board of directors (1) develop and maintain comprehensive investment policy statements and (2) develop a complete set of operating procedures and guidelines for its investment activities. GAO received comments from the Department of Labor and the PBGC. They generally agreed with our recommendations.