Datasets:

pszemraj
/

govreport-summarization-8192

Tasks:

Summarization

Languages: English

Size Categories: 1K<n<10K

Source Datasets: ccdv/govreport-summarization

Tags: govreport long document

License:

Dataset card Viewer Files Files and versions Community

Dataset Viewer

Auto-converted to Parquet

View in Dataset Viewer

Viewer

Split (3)

train · 8.2k rows

report stringlengths 319 46.5k	summary stringlengths 127 5.75k	input_token_len int64 78 8.19k	summary_token_len int64 29 1.02k
There are some similarities in how Medicare pays ASCs and hospital outpatient departments for the procedures they perform. However, the methods used by CMS to calculate the payment rates in each system, as well as the mechanisms used to revise the Medicare payment rates, differ. In 1980, legislation was enacted that enabled ASCs to bill Medicare for certain surgical procedures provided to Medicare beneficiaries. Under the ASC payment system, Medicare pays a predetermined, and generally all- inclusive, amount per procedure to the facility. The approximately 2,500 surgical procedures that ASCs may bill for under Medicare are assigned to one of nine payment groups that contain procedures with similar costs, but not necessarily clinical similarities. All procedures assigned to one payment group are paid at the same rate. Under the Medicare payment system, when more than one procedure is performed at the same time, the ASC receives a payment for each of the procedures. However, the procedure that has the highest payment rate receives 100 percent of the applicable payment, and each additional procedure receives 50 percent of the applicable payment. The Medicare payment for a procedure performed at an ASC is intended to cover the direct costs for a procedure, such as nursing and technician services, drugs, medical and surgical supplies and equipment, anesthesia materials, and diagnostic services (including imaging services), and the indirect costs associated with the procedure, including use of the facility and related administrative services. The ASC payment for a procedure does not include payment for implantable devices or prosthetics related to the procedure; ASCs may bill separately for those items. In addition, the payment to the ASC does not include payment for professional services associated with the procedure; the physician who performs the procedure and the anesthesiologist or anesthetist bill Medicare directly for their services. Finally, the ASC payment does not include payment for certain other services that are not directly related to performing the procedure and do not occur during the time that the procedure takes place, such as some laboratory, X-ray, and other diagnostic tests. Because these additional services are not ASC procedures, they may be performed by another provider. In those cases, Medicare makes payments to those providers for the additional services. For example, a laboratory service needed to evaluate a tissue sample removed during an ASC procedure is not included in the ASC payment. The provider that evaluated the tissue sample would bill and receive payment from Medicare for that service. Because ASCs receive one inclusive payment for the procedure performed and its associated services, such as drugs, they generally include on their Medicare claim only the procedure performed. In 1997, legislation was enacted that required the implementation of a prospective payment system for hospital outpatient departments; the OPPS was implemented in August 2000. Although ASCs perform only procedures, hospital outpatient departments provide a much broader array of services, including diagnostic services, such as X-rays and laboratory tests, and emergency room and clinic visits. Each of the approximately 5,500 services, including procedures, that hospital outpatient departments perform is assigned to one of over 800 APC groups with other services with clinical and cost similarities for payment under the OPPS. All services assigned to one APC group are paid the same rate. Similar to ASCs, when hospitals perform multiple procedures at the same time, they receive 100 percent of the applicable payment for the procedure that has the highest payment rate, and 50 percent of the applicable payment for each additional procedure, subject to certain exceptions. Like payments to ASCs, payment for a procedure under the OPPS is intended to cover the costs of the use of the facility, nursing and technician services, most drugs, medical and surgical supplies and equipment, anesthesia materials, and administrative costs. Medicare payment to a hospital for a procedure does not include professional services for physicians or other nonphysician practitioners. These services are paid for separately by Medicare. However, there are some differences between ASC and OPPS payments for procedures. Under the OPPS, hospital outpatient departments generally may not bill separately for implantable devices related to the procedure, but they may bill separately for additional services that are directly related to the procedure, such as certain drugs and diagnostic services, including X-rays. Hospital outpatient departments also may bill separately for additional services that are not directly related to the procedure and do not occur during the procedure, such as laboratory services to evaluate a tissue sample. Because they provide a broader array of services, and because CMS has encouraged hospitals to report all services provided during a procedure on their Medicare claims for rate-setting purposes, hospital claims may provide more detail about the services delivered during a procedure than ASC claims do. CMS set the initial 1982 ASC payment rates based on cost and charge data from 40 ASCs. At that time, there were about 125 ASCs in operation. Procedures were placed into four payment groups, and all procedures in a group were paid the same rate. When the ASC payment system was first established, federal law required CMS to review the payment rates periodically. In 1986, CMS conducted an ASC survey to gather cost and charge data. In 1990, using these data, CMS revised the payment rates and increased the number of payment groups to eight. A ninth payment group was established in 1991. These groups are still in use, although some procedures have been added to or deleted from the ASC-approved list. Although payments have not been revised using ASC cost data since 1990, the payment rates have been periodically updated for inflation. In 1994, Congress required that CMS conduct a survey of ASC costs no later than January 1, 1995, and thereafter every 5 years, to revise ASC payment rates. CMS conducted a survey in 1994 to collect ASC cost data. In 1998, CMS proposed revising ASC payment rates based on the 1994 survey data and assigned procedures performed at ASCs into payment groups that were comparable to the payment groups it was developing for the same procedures under the OPPS. However, CMS did not implement the proposal, and, as a result, the ASC payment system was not revised using the 1994 data. In 2003, MMA eliminated the requirement to conduct ASC surveys every 5 years and required CMS to implement a revised ASC payment system no later than January 1, 2008. During the course of our work, in August 2006, CMS published a proposed rule that would revise the ASC payment system effective January 1, 2008. In this proposed rule, CMS bases the revised ASC payment rates on the OPPS APC groups. However, the payment rates would be lower for ASCs. The initial OPPS payment rates, implemented in August 2000, were based on hospitals' 1996 costs. To determine the OPPS payment rates, CMS first calculates each hospital's cost for each service by multiplying the charge for that service by a cost-to-charge ratio computed from the hospital's most recently reported data. After calculating the cost of each service for each hospital, the services are grouped by their APC assignment, and a median cost for each APC group is calculated from the median costs of all services assigned to it. Using the median cost, CMS assigns each APC group a weight based on its median cost relative to the median cost of all other APCs. To obtain a payment rate for each APC group, CMS multiplies the relative weight by a factor that converts it to a dollar amount. Beginning in 2002, as required by law, the APC group payment rates have been revised annually based on the latest charge and cost data. In addition, the payment rates for services paid under the OPPS receive an annual inflation update. We found many similarities in the additional services provided by ASCs and hospital outpatient departments with the top 20 procedures. Of the additional services billed with a procedure, few resulted in an additional payment in one setting but not the other. Hospitals were paid for some of the related additional services they billed with the procedures. In the ASC setting, other providers billed Medicare for these services and received payment for them. In our analysis of Medicare claims, we found many similarities in the additional services billed in the ASC or hospital outpatient department setting with the top 20 procedures. The similar additional services are illustrated in the following four categories of services: additional procedures, laboratory services, radiology services, and anesthesia services. First, one or more additional procedures was billed with a procedure performed in either the ASC or hospital outpatient department setting for 14 of the top 20 procedures. The proportion of time each additional procedure was billed in each setting was similar. For example, when a hammertoe repair procedure was performed, our analysis indicated that another procedure to correct a bunion was billed 11 percent of the time in the ASC setting, and in the hospital outpatient setting, the procedure to correct a bunion was billed 13 percent of the time. Similarly, when a diagnostic colonoscopy was performed, an upper gastrointestinal (GI) endoscopy was billed 11 percent of the time in the ASC setting, and in the hospital setting, the upper GI endoscopy was billed 12 percent of the time. For 11 of these 14 procedures, the proportion of time each additional procedure was billed differed by less than 10 percentage points between the two settings. For the 3 remaining procedures, the percentage of time that an additional procedure was billed did not vary by more than 25 percentage points between the two settings. See appendix III for a complete list of the additional procedures billed and the proportion of time they were billed in each setting. Second, laboratory services were billed with 10 of the top 20 procedures in the hospital outpatient department setting and 7 of the top 20 procedures in the ASC setting. While these services were almost always billed by the hospital in the outpatient setting, they were typically not billed by the ASCs. These laboratory services were present in our analysis in the ASC setting because they were performed and billed by another Medicare part B provider. Third, four different radiology services were billed with 8 of the top 20 procedures. Radiology services were billed with 5 procedures in the ASC setting and with 8 procedures in the hospital outpatient department setting. The radiology services generally were included on the hospital outpatient department bills but rarely were included on the ASC bills. Similar to laboratory services, hospital outpatient departments billed for radiology services that they performed in addition to the procedures. When radiology services were billed with procedures in the ASC setting, these services generally were performed and billed by another part B provider. Fourth, anesthesia services were billed with 17 of the top 20 procedures in either the ASC or hospital outpatient settings and with 14 procedures in both settings. In virtually every case in the ASC setting, and most cases in the hospital outpatient department setting, these services were billed by another part B provider. According to our analysis, ASCs did not generally include any services other than the procedures they performed on their bills. However, in the hospital outpatient setting, some additional services were included on the hospitals' bills. We believe this is a result of the structure of the two payment systems. As ASCs generally receive payment from Medicare only for procedures, they typically include only those procedures on their bills. In contrast, hospital outpatient departments' bills often include many of the individual items or services they provide as a part of a procedure because CMS has encouraged them to do so, whether the items or services are included in the OPPS payment or paid separately. With the exception of additional procedures, there were few separate payments that could be made for additional services provided with the top 20 procedures because most of the services in our analysis were included in the Medicare payment to the ASC or hospital. Under both the Medicare ASC and OPPS payment systems, when more than one procedure is performed at the same time, the facility receives 100 percent of the applicable payment for the procedure that has the highest payment rate and 50 percent of the applicable payment for each additional procedure. As this policy is applicable to both settings, for those instances in our analysis when an additional procedure was performed with one of the top 20 procedures in either setting, the ASC or hospital outpatient department received 100 percent of the payment for the procedure with the highest payment rate and 50 percent of the payment for each lesser paid procedure. Individual drugs were billed by hospital outpatient departments for most of the top 20 procedures, although they were not present on the claims from ASCs, likely because ASCs generally cannot receive separate Medicare payments for individual drugs. However, none of the individual drugs billed by the hospital outpatient departments in our analysis resulted in an additional payment to the hospitals. In each case, the cost of the particular drug was included in the Medicare payment for the procedure. In the case of the laboratory services billed with procedures in the ASC and hospital outpatient department settings, those services were not costs included in the payment for the procedure in either setting and were paid separately in each case. For both settings, the payment was made to the provider that performed the service. In the case of the hospital outpatient department setting, the payment was generally made to the hospital, while, for procedures performed at ASCs, payment was made to another provider who performed the service. Of the four radiology services in our analysis, three were similar to the laboratory services in that they are not included in the cost of the procedure and are separately paid services under Medicare. Therefore, when hospitals provided these services, they received payment for them. In the ASC setting, these services were typically billed by a provider other than the ASC, and the provider received payment for them. The fourth radiology service is included in the payment for the procedure with which it was associated. Therefore, no separate payment was made to either ASCs or hospital outpatient departments. With regard to anesthesia services, most services were billed by and paid to a provider other than an ASC or hospital. As a group, the costs of procedures performed in ASCs have a relatively consistent relationship with the costs of the APC groups to which they would be assigned under the OPPS. That is, the APC groups accurately reflect the relative costs of procedures performed in ASCs. We found that the ASC-to-APC cost ratios were more tightly distributed around their median cost ratio than the OPPS-to-APC cost ratios were around their median cost ratio. Specifically, 45 percent of all procedures in our analysis fell within 0.10 points of the ASC-to-APC median cost ratio, and 33 percent of procedures fell within 0.10 points of the OPPS-to-APC median cost ratio. However, the costs of procedures in ASCs are substantially lower than costs for the same procedures in the hospital outpatient setting. The APC groups reflect the relative costs of procedures provided by ASCs as well as they reflect the relative costs of procedures provided in the hospital outpatient department setting. In our analysis, we listed the procedures performed at ASCs and calculated the ratio of the cost of each procedure to the cost of the APC group to which it would have been assigned, referred to as the ASC-to-APC cost ratio. We then calculated similar cost ratios for the same procedures exclusively within the OPPS. To determine an OPPS-to-APC cost ratio, we divided individual procedures' median costs, as calculated by CMS for the OPPS, by the median cost of their APC group. Our analysis of the cost ratios showed that the ASC-to-APC cost ratios were more tightly distributed around their median than were the OPPS-to-APC cost ratios; that is, there were more of them closer to the median. Specifically, 45 percent of procedures performed in ASCs fell within a 0.10 point range of the ASC-to-APC median cost ratio, and 33 percent of those procedures fell within a 0.10 point range of the OPPS-to-APC median cost ratio in the hospital outpatient department setting (see figs. 1 and 2). Therefore, there is less variation in the ASC setting between individual procedures' costs and the costs of their assigned APC groups than there is in the hospital outpatient department setting. From this outcome, we determined that the OPPS APC groups could be used to pay for procedures in ASCs. The median costs of procedures performed in ASCs were generally lower than the median costs of their corresponding APC group under the OPPS. Among all procedures in our analysis, the median ASC-to-APC cost ratio was 0.39. The ASC-to-APC cost ratios ranged from 0.02 to 3.34. When weighted by Medicare volume based on 2004 claims data, the median ASC- to-APC cost ratio was 0.84. We determined that the median OPPS-to-APC cost ratio was 1.04. This analysis shows that when compared to the median cost of the same APC group, procedures performed in ASCs had substantially lower costs than when those same procedures were performed in hospital outpatient departments. Generally, there are many similarities between the additional services provided in ASCs and hospital outpatient departments with one of the top 20 procedures, and few resulted in an additional Medicare payment to ASCs or hospital outpatient departments. Although costs for individual procedures vary, in general, the median costs for procedures are lower in ASCs, relative to the median costs of their APC groups, than the median costs for the same procedures in the hospital outpatient department setting. The APC groups in the OPPS reflect the relative costs of procedures performed in ASCs in the same way that they reflect the relative costs of the same procedures when they are performed in hospital outpatient departments. Therefore, the APC groups could be applied to procedures performed in ASCs, and the OPPS could be used as the basis for an ASC payment system, eliminating the need for ASC surveys and providing for an annual revision of the ASC payment groups. We recommend that the Administrator of CMS implement a payment system for procedures performed in ASCs based on the OPPS. The Administrator should take into account the lower relative costs of procedures performed in ASCs compared to hospital outpatient departments in determining ASC payment rates. We received written comments on a draft of this report from CMS (see app. IV). We also received oral comments from external reviewers representing two ASC industry organizations, AAASC and FASA. In commenting on a draft of this report, CMS stated that our recommendation is consistent with its August 2006 proposed revisions to the ASC payment system. Industry representatives who reviewed a draft of this report did not agree or disagree with our recommendation for executive action. They did, however, provide several comments on the draft report. The industry representatives noted that we did not analyze the survey results to examine differences in per-procedure costs among single-specialty and multi-specialty ASCs. Regarding this comment, we initially considered developing our survey sample stratified by ASC specialty type. However, because accurate data identifying ASCs' specialties do not exist, we were unable to stratify our survey sample by specialty type. The industry representatives asked us to provide more explanation in our scope and methodology regarding our development of a relative weight scale for Medicare ASC-approved procedures to capture the general variation in resources associated with performing different procedures. We expanded the discussion of how we developed the relative weight scale in our methodology section. Reviewers also made technical comments, which we incorporated where appropriate. We are sending a copy of this report to the Administrator of CMS and appropriate congressional committees. The report is available at no charge on GAO's Web site at http://www.gao.gov. We will also make copies available to others on request. If you or your staff members have any questions about this report, please contact me at (202) 512-7119 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 members who made significant contributions to this report are listed in appendix V. The Medicare payment rates for ambulatory surgical centers (ASC), along with those of other facilities, are adjusted to account for the variation in labor costs across the country. To calculate payment rates for individual ASCs, the Centers for Medicare & Medicaid Services (CMS) calculates the share of total costs that are labor-related and then adjusts ASCs' labor- related share of costs based on a wage index calculated for specific geographic areas across the country. The wage index reflects how the average wage for health care personnel in each geographic area compares to the national average health care personnel wage. The geographic areas are intended to represent the separate labor markets in which health care facilities compete for employees. In setting the initial ASC payment rates for 1982, CMS determined from the first survey of ASCs that one-third of their costs were labor-related. The labor-related costs included employee salaries and fringe benefits, contractual personnel, and owners' compensation for duties performed for the facility. To determine the payment rates for each individual ASC, CMS multiplied one-third of the payment rate for each procedure--the labor- related portion--by the local area wage index. Each ASC received the base payment rate for two-thirds of the payment rate--the nonlabor-related portion--for each procedure. The sum of the labor-related and nonlabor- related portions equaled each ASC's payment rate for each procedure. In 1990, when CMS revised the payment system based on a 1986 ASC survey, CMS found ASCs' average labor-related share of costs to be 34.45 percent and used this percentage as the labor-related portion of the payment rate. In a 1998 proposed rule, CMS noted that ASCs' share of labor-related costs as calculated from the 1994 ASC cost survey had increased to an average of 37.66 percent, slightly higher than the percentage calculated from the 1986 survey. However, CMS did not implement the 1998 proposal. Currently, the labor-related proportion of costs from CMS's 1986 survey, 34.45 percent, is used for calculating ASC payment rates. Using 2004 cost data we received from 290 ASCs that responded to our survey request for information, we determined that the mean labor-related proportion of costs was 50 percent, and the range of the labor-related costs for the middle 50 percent of our ASC facilities was 43 percent to 57 percent of total costs. To compare the delivery of procedures between ASCs and hospital outpatient departments, we analyzed Medicare claims data from 2003. To compare the relative costs of procedures performed in ASCs and hospital outpatient departments, we collected cost and procedure data from 2004 from a sample of Medicare-participating ASCs. We also interviewed officials at CMS and representatives from ASC industry organizations, specifically, the American Association of Ambulatory Surgery Centers (AAASC) and FASA, physician specialty societies, and nine ASCs. To compare the delivery of additional services provided with procedures performed in ASCs and hospital outpatient departments, we identified all additional services frequently billed in each setting when one of the top 20 procedures with the highest Medicare ASC claims volume is performed. These procedures represented approximately 75 percent of all Medicare ASC claims in 2003. Using Medicare claims data for 2003, we identified beneficiaries receiving one of the top 20 procedures in either an ASC or hospital outpatient department, then identified any other claims for those beneficiaries from ASCs, hospital outpatient departments, durable medical equipment suppliers, and other Medicare part B providers. We identified claims for the beneficiaries on the day the procedure was performed and the day after. We created a list that included all additional services that were billed at least 10 percent of the time with each of the top 20 procedures when they were performed in ASCs. We created a similar list of additional services for each of the top 20 procedures when they were performed in hospital outpatient departments. We then compared the lists for each of the top 20 procedures between the two settings to determine whether there were similarities in the additional services that were billed to Medicare. To compare the Medicare payments for procedures performed in ASCs and hospital outpatient departments, we identified whether any additional services included in our analysis resulted in an additional payment. We used Medicare claims data from the National Claims History (NCH) files. These data, which are used by the Medicare program to make payments to health care providers, are closely monitored by both CMS and the Medicare contractors that process, review, and pay claims for Medicare services. The data are subject to various internal controls, including checks and edits performed by the contractors before claims are submitted to CMS for payment approval. Although we did not review these internal controls, we did assess the reliability of the NCH data. First, we reviewed all existing information about the data, including the data dictionary and file layouts. We also interviewed experts at CMS who regularly use the data for evaluation and analysis. We found the data to be sufficiently reliable for the purposes of this report. To compare the relative costs of procedures performed in ASCs and hospital outpatient departments, we first compiled information on ASCs' costs and procedures performed. Because there were no recent existing data on ASC costs, we surveyed 600 ASCs, randomly selected from all ASCs, to obtain their 2004 cost and procedure data. We received response data from 397 ASC facilities. We assessed the reliability of these data through several means. We identified incomplete and inconsistent survey responses within individual surveys and placed follow-up calls to respondents to complete or verify their responses. To ensure that survey response data were accurately transferred to electronic files for our analytic purposes, two analysts independently entered all survey responses. Any discrepancies between the two sets of entered responses were resolved. We performed electronic testing for errors in accuracy and completeness, including an analysis of costs per procedure. As a result of our data reliability testing, we determined that data from 290 responding facilities were sufficiently reliable for our purposes. Our nonresponse analysis showed that there was no geographic bias among the facilities responding to our survey. The responding facilities performed more Medicare services than the average for all ASCs in our sample. To allocate ASCs' total costs among the individual procedures they perform, we developed a method to allocate the portion of an ASC's costs accounted for by each procedure. We constructed a relative weight scale for Medicare ASC-approved procedures that captures the general variation in resources associated with performing different procedures. The resources we used were the clinical staff time, surgical supplies, and surgical equipment used during the procedures. We used cost and quantity data on these resources from information CMS had collected for the purpose of setting the practice expense component of physician payment rates. For procedures for which CMS had no data on the resources used, we used information we collected from medical specialty societies and physicians who work for CMS. We summed the costs of the resources for each procedure and created a relative weight scale by dividing the total cost of each procedure by the average cost across all of the procedures. We assessed the reliability of these data through several means. We compared electronic CMS data with the original document sources for a large sample of records, performed electronic testing for errors in accuracy and completeness, and reviewed data for reasonableness. Based on these efforts, we determined that data were sufficiently reliable for our purposes. To calculate per-procedure costs with the data from the surveyed ASC facilities, we first deducted costs that Medicare considers unallowable, such as advertising and entertainment costs. (See fig. 3 for our per- procedure cost calculation methodology.) We also deducted costs for services that Medicare pays for separately, such as physician and nonphysician practitioner services. We then separated each facility's total costs into its direct and indirect costs. We defined direct costs as those associated with the clinical staff, equipment, and supplies used during the procedure. Indirect costs included all remaining costs, such as support and administrative staff, building expenses, and outside services purchased. To allocate each facility's direct costs across the procedures it performed, we applied our relative weight scale. We allocated indirect costs equally across all procedures performed by the facility. For each procedure performed by a responding ASC facility, we summed its allocated direct and indirect costs to determine a total cost for the procedure. To obtain a per-procedure cost across all ASCs, we arrayed the calculated costs for all ASCs performing that procedure and identified the median cost. To compare per-procedure costs for ASCs and hospital outpatient departments, we first obtained from CMS the list of ambulatory payment classification (APC) groups used for the outpatient prospective payment system (OPPS) and the procedures assigned to each APC group. We also obtained from CMS the OPPS median cost of each procedure and the median cost of each APC group. We then calculated a ratio between each procedure's ASC median cost, as determined by the survey, and the median cost of each procedure's corresponding APC group under the OPPS, referred to as the ASC-to-APC cost ratio. We also calculated a ratio between each ASC procedure's median cost under the OPPS and the median cost of the procedure's APC group, using the data obtained from CMS, referred to as the OPPS-to-APC cost ratio. To evaluate the difference in procedure costs between the two settings, we compared the ASC-to- APC and OPPS-to-APC cost ratios. To assess how well the relative costs of procedures in the OPPS, defined by their assignment to APC groups, reflect the relative costs of procedures in the ASC setting, we evaluated the distribution of the ASC-to-APC and OPPS-to-APC cost ratios. To calculate the percentage of labor-related costs among our sample ASCs, for each ASC, we divided total labor costs by total costs, after deducting costs not covered by Medicare's facility payment. We then determined the range of the percentage of labor-related costs among all of our ASCs and between the 25th percentile and the 75th percentile, as well as the mean and median percentage of labor-related costs. We performed our work from April 2004 through October 2006 in accordance with generally accepted government auditing standards. Appendix III: Additional Procedures Billed with the Top 20 ASC Procedures, 2003 (percentage) N/A (percentage) In addition to the contact named above, key contributors to this report were Nancy A. Edwards, Assistant Director; Kevin Dietz; Beth Cameron Feldpush; Marc Feuerberg; and Nora Hoban.	Medicare pays for surgical procedures performed at ambulatory surgical centers (ASC) and hospital outpatient departments through different payment systems. Although they perform a similar set of procedures, no comparison of ASC and hospital outpatient per-procedure costs has been conducted. The Medicare Prescription Drug, Improvement, and Modernization Act of 2003 directed GAO to compare the relative costs of procedures furnished in ASCs to the relative costs of those procedures furnished in hospital outpatient departments, in particular, how accurately the payment groups used in the hospital outpatient prospective payment system (OPPS) reflect the relative costs of procedures performed in ASCs. To do this, GAO collected data from ASCs through a survey. GAO also obtained hospital outpatient data from the Centers for Medicare & Medicaid Services (CMS). GAO determined that the payment groups in the OPPS, known as ambulatory payment classification (APC) groups, accurately reflect the relative cost of procedures performed in ASCs. GAO calculated the ratio between each procedure's ASC median cost, as determined by GAO's survey, and the median cost of each procedure's corresponding APC group under the OPPS, referred to as the ASC-to-APC cost ratio. GAO also compared the OPPS median costs of those same procedures with the median costs of their APC groups, referred to as the OPPS-to-APC cost ratio. GAO's analysis of the ASC-to-APC and OPPS-to-APC cost ratios showed that 45 percent of all procedures in the analysis fell within a 0.10 point range of the ASC-to-APC median cost ratio, and 33 percent of procedures fell within a 0.10 point range of the OPPS-to-APC median cost ratio. These similar patterns of distribution around the median show that the APC groups reflect the relative costs of procedures provided by ASCs as well as they reflect the relative costs of procedures provided in hospital outpatient departments and can be used as the basis for the ASC payment system. GAO's analysis also identified differences in the cost of procedures in the two settings. The median cost ratio among all ASC procedures was 0.39 and when weighted by Medicare claims volume was 0.84. The median cost ratio for OPPS procedures was 1.04. Thus, the cost of procedures in ASCs is substantially lower than the corresponding cost in hospital outpatient departments.	6,596	529
IRS's mission is to provide America's taxpayers top-quality service by helping them to understand and meet their tax responsibilities and to enforce the law with integrity and fairness to all. During fiscal year 2015, IRS collected more than $3.3 trillion; processed more than 243 million tax returns and other forms; and issued more than $403 billion in tax refunds. IRS employs about 90,000 people in its Washington, D.C., headquarters and at more than 550 offices in all 50 states, U.S. territories, and some U.S. embassies and consulates. Each filing season IRS provides assistance to tens of millions of taxpayers over the phone, through written correspondence, online, and face-to-face. The scale of these operations alone presents challenges. In carrying out its mission, IRS relies extensively on computerized information systems, which it must effectively secure to protect sensitive financial and taxpayer data for the collection of taxes, processing of tax returns, and enforcement of federal tax laws. Accordingly, it is critical for IRS to effectively implement information security controls and an agency- wide information security program in accordance with federal law and guidance. Cyber incidents can adversely affect national security, damage public health and safety, and compromise sensitive information. Regarding IRS specifically, two recent incidents illustrate the impact on taxpayer and other sensitive information: In June 2015, the Commissioner of the IRS testified that unauthorized third parties had gained access to taxpayer information from its Get Transcript application. According to officials, criminals used taxpayer- specific data acquired from non-department sources to gain unauthorized access to information on approximately 100,000 tax accounts. These data included Social Security information, dates of birth, and street addresses. In an August 2015 update, IRS reported this number to be about 114,000, and that an additional 220,000 accounts had been inappropriately accessed. In a February 2016 update, the agency reported that an additional 390,000 accounts had been accessed. Thus, about 724,000 accounts were reportedly affected. The online Get Transcript service has been unavailable since May 2015. In March 2016, IRS stated that as part of its ongoing security review, it had temporarily suspended the Identity Protection Personal Identification Number (IP PIN) service on IRS.gov. The IP PIN is a single-use identification number provided to taxpayers who are victims of identity theft (IDT) to help prevent future IDT refund fraud. The service on IRS's website allowed taxpayers to retrieve their IP PINs online by passing IRS's authentication checks. These checks confirm taxpayer identity by asking for personal, financial and tax-related information. The IRS stated that it was conducting further review of the IP PIN service and is looking at further strengthening the security features before resuming service. As of April 7, the online service was still suspended. The Commissioner of Internal Revenue has overall responsibility for ensuring the confidentiality, integrity, and availability of the information and systems that support the agency and its operations. Within IRS, the senior agency official responsible for information security is the Associate CIO, who heads the IRS Information Technology Cybersecurity organization. As we reported in March 2016, IRS has implemented numerous controls over key financial and tax processing systems; however, it had not always effectively implemented access and other controls, including elements of its information security program. Access controls are intended to prevent, limit, and detect unauthorized access to computing resources, programs, information, and facilities. These controls include identification and authentication, authorization, cryptography, audit and monitoring, and physical security controls, among others. In our most recent review we found that IRS had improved access controls, but some weaknesses remain. Identifying and authenticating users--such as through user account-password combinations--provides the basis for establishing accountability and controlling access to a system. IRS established policies for identification and authentication, including requiring multifactor authentication for local and network access accounts and establishing password complexity and expiration requirements. It also improved identification and authentication controls by, for example, expanding the use of an automated mechanism to centrally manage, apply, and verify password requirements. However, weaknesses in identification and authentication controls remained. For example, the agency used easily guessable passwords on servers supporting key systems. Authorization controls limit what actions users are able to perform after being allowed into a system and should be based on the concept of "least privilege," granting users the least amount of rights and privileges necessary to perform their duties. While IRS established policies for authorizing access to its systems, it continued to permit excessive access in some cases. For example, users were granted rights and permissions in excess of what they needed to perform their duties, including for an application used to process electronic tax payment information and a database on a human resources system. Cryptography controls protect sensitive data and computer programs by rendering data unintelligible to unauthorized users and protecting the integrity of transmitted or stored data. IRS policies require the use of encryption and it continued to expand its use of encryption to protect sensitive data. However, key systems we reviewed had not been configured to encrypt sensitive user authentication data. Audit and monitoring is the regular collection, review, and analysis of events on systems and networks in order to detect, respond to, and investigate unusual activity. IRS established policies and procedures for auditing and monitoring its systems and continued to enhance its capability by, for example, implementing an automated mechanism to log user activity on its access request and approval system. But it had not established logging for two key applications used to support the transfer of financial data and access and manage taxpayer accounts; nor was the agency consistently maintaining key system and application audit plans. Physical security controls, such as physical access cards, limit access to an organization's overall facility and areas housing sensitive IT components. IRS established policies for physically protecting its computer resources and physical security controls at its enterprise computer centers, such as a dedicated guard force at each of its computer centers. However, the agency had yet to address weaknesses in its review of access lists for both employees and visitors to sensitive areas. IRS also had weaknesses in configuration management controls, which are intended to prevent unauthorized changes to information system resources (e.g., software and hardware) and provide assurance that systems are configured and operating securely. Specifically, while IRS developed policies for managing the configuration of its information technology (IT) systems and improved some configuration management controls, it did not, for example, ensure security patch updates were applied in a timely manner to databases supporting 2 key systems we reviewed, including a patch that had been available since August 2012. To its credit, IRS had established contingency plans for the systems we reviewed, which help ensure that when unexpected events occur, critical operations can continue without interruption or can be promptly resumed, and that information resources are protected. Specifically, IRS had established policies for developing contingency plans for its information systems and for testing those plans, as well as for implementing and enforcing backup procedures. Moreover, the agency had documented and tested contingency plans for its systems and improved continuity of operations controls for several systems. Nevertheless, the control weaknesses can be attributed in part to IRS's inconsistent implementation of elements of its agency-wide information security program. The agency established a comprehensive framework for its program, including assessing risk for its systems, developing system security plans, and providing employees with security awareness and specialized training. However, IRS had not updated key mainframe policies and procedures to address issues such as comprehensively auditing and monitoring access. In addition, the agency had not fully addressed previously identified deficiencies or ensured that its corrective actions were effective. During our most recent review, IRS told us it had addressed 28 of our prior recommendations; however, we determined that 9 of these had not been effectively implemented. The collective effect of the deficiencies in information security from prior years that continued to exist in fiscal year 2015, along with the new deficiencies we identified, are serious enough to merit the attention of those charged with governance of IRS and therefore represented a significant deficiency in IRS's internal control over financial reporting systems as of September 30, 2015. To assist IRS in fully implementing its agency-wide information security program, we made two new recommendations to more effectively implement security-related policies and plans. In addition, to assist IRS in strengthening security controls over the financial and tax processing systems we reviewed, we made 43 technical recommendations in a separate report with limited distribution to address 26 new weaknesses in access controls and configuration management. Implementing these recommendations--in addition to the 49 outstanding recommendations from previous audits--will help IRS improve its controls for identifying and authenticating users, limiting users' access to the minimum necessary to perform their job-related functions, protecting sensitive data when they are stored or in transit, auditing and monitoring system activities, and physically securing its IT facilities and resources. Table 1 below provides the number of our prior recommendations to IRS that were not implemented at the beginning of our fiscal year 2015 audit, how many were resolved by the end of the audit, new recommendations, and the total number of outstanding recommendations at the conclusion of the audit. In commenting on drafts of our reports presenting the results of our fiscal year 2015 audit, the IRS Commissioner stated that while the agency agreed with our new recommendations, it will review them to ensure that its actions include sustainable fixes that implement appropriate security controls balanced against IT and human capital resource limitations. In addition, IRS can take steps to improve its response to data breaches. Specifically, in December 2013 we reported on the extent to which data breach policies at eight agencies, including IRS, adhered to requirements and guidance set forth by the Office of Management and Budget and the National Institute of Standards and Technology. While the agencies in our review generally had policies and procedures in place that reflected the major elements of an effective data breach response program, implementation of these policies and procedures was not consistent. With respect to IRS, we determined that its policies and procedures generally reflected key practices, although the agency did not require considering the number of affected individuals as a factor when determining if affected individuals should be notified of a suspected breach. In addition, IRS did not document lessons learned from periodic analyses of its breach response efforts. We recommended that IRS correct these weaknesses, but the agency has yet to fully address them. The importance of protecting taxpayer information is further highlighted by the billions of dollars that have been lost to IDT refund fraud, which continues to be an evolving threat. IRS develops estimates of the extent of IDT refund fraud to help direct its efforts to identify and prevent the crime. While its estimates have inherent uncertainty, IRS estimated that it prevented or recovered $22.5 billion in fraudulent IDT refunds in filing season 2014 (see figure 1). However, IRS also estimated, where data were available, that it paid $3.1 billion in fraudulent IDT refunds. Because of the difficulties in knowing the amount of undetectable fraud, the actual amount could differ from these estimates. IRS has taken steps to address IDT refund fraud; however, it remains a persistent and continually changing threat. IRS recognized the challenge of IDT refund fraud in its fiscal year 2014-2017 strategic plan and increased resources dedicated to combating IDT and other types of refund fraud. In fiscal year 2015, IRS reported that it staffed more than 4,000 full-time equivalents and spent about $470 million on all refund fraud and IDT activities. As described above, IRS received an additional $290 million for fiscal year 2016 to improve customer service, IDT identification and prevention, and cybersecurity efforts and the agency plans to use $16.1 million of this funding to help prevent IDT refund fraud, among other things. The administration requested an additional $90 million and an additional 491 full-time equivalents for fiscal year 2017 to help prevent IDT refund fraud and reduce other improper payments. IRS estimates that this $90 million investment in IDT refund fraud and other improper payment prevention would help it protect $612 million in revenue in fiscal year 2017, as well as protect revenue in future years. IRS has taken action to improve customer service related to IDT refund fraud. For example, between the 2011 and 2015 filing seasons, IRS experienced a 430 percent increase in the number of telephone calls to its Identity Theft Toll Free Line--as of March 19, 2016, IRS had received over 1.1 million calls to this line. Moreover, 77 percent of callers seeking assistance on this telephone line received it compared to 54 percent during the same period last year. Average wait times during the same period have also decreased--taxpayers are waiting an average of 14 minutes to talk to an assistor, a decrease from 27 minutes last year. IRS also works with third parties, such as tax preparation industry participants, states, and financial institutions to try to detect and prevent IDT refund fraud. In March 2015, the IRS Commissioner convened a Security Summit with industry and states to improve information sharing and authentication. IRS officials said that 40 state departments of revenue and 20 tax industry participants have officially signed a partnership agreement to enact recommendations developed and agreed to by summit participants. IRS plans to invest a portion of the $16.1 million it received in fiscal year 2016 into identity theft prevention and refund fraud mitigation actions from the Security Summit. These efforts include developing an Information Sharing and Analysis Center where IRS, states, and industry can share information to combat IDT refund fraud. Even though IRS has prioritized combating IDT refund fraud, fraudsters adapt their schemes to identify weaknesses in IDT defenses, such as gaining access to taxpayers' tax return transcripts through IRS's online Get Transcript service. According to IRS officials, with access to tax transcripts, fraudsters can create historically consistent returns that are hard to distinguish from a return filed by a legitimate taxpayer, potentially making it more difficult for IRS to identify and detect IDT refund fraud. Without additional action by IRS and Congress, the risk of issuing fraudulent IDT refunds could grow. We previously made recommendations to IRS to help it better combat IDT refund fraud: Authentication. In January 2015, we reported that IRS's authentication tools have limitations and recommended that IRS assess the costs, benefits and risks of its authentication tools. For example, individuals can obtain an e-file PIN by providing their name, Social Security number, date of birth, address, and filing status for IRS's e-file PIN application. Identity thieves can easily find this information, allowing them to bypass some, if not all, of IRS's automatic checks, according to our analysis and interviews with tax software and return preparer associations and companies. After filing an IDT return using an e-file PIN, the fraudulent return would proceed through IRS's normal return processing. In November 2015, IRS officials told us that the agency had developed guidance for its Identity Assurance Office to assess costs, benefits, and risk, and that its analysis will inform decision-making on authentication-related issues. IRS also noted that the methods of analysis for the authentication tools will vary depending on the different costs and other factors for authenticating taxpayers in different channels, such as online, phone, or in-person. In February 2016, IRS officials told us that the Identity Assurance Office plans to complete a strategic plan for taxpayer authentication across the agency in September 2016. While IRS is taking steps, it will still be vulnerable until it completes and uses the results of its analysis of costs, benefits, and risk to inform decision-making. Form W-2, Wage and Tax Statement (W-2) Pre-refund Matching. In August 2014 we reported that the wage information that employers report on Form W-2 is not available to IRS until after it issues most refunds, and that if IRS had access to W-2 data earlier, it could match such information to taxpayers' returns and identify discrepancies before issuing billions of dollars of fraudulent IDT refunds. We recommended that IRS assess the costs and benefits of accelerating W-2 deadlines. In response to our recommendation, IRS provided us with a report in September 2015 discussing (1) adjustments to IRS systems and work processes needed to use accelerated W-2 information, (2) the potential impacts on internal and external stakeholders, and (3) other changes needed to match W-2 data to tax returns prior to issuing refunds, such as delaying refunds until W-2 data are available. In December 2015, the Consolidated Appropriations Act of 2016 amended the tax code to accelerate W-2 filing deadlines to January 31. IRS's report will help IRS determine how to best implement pre- refund W-2 matching, given the new January 31st deadline for filing W-2s. Additionally, we suggested that Congress should consider providing the Secretary of the Treasury with the regulatory authority to lower the threshold for electronic filing of W-2s, which could make more W-2 information available to IRS earlier. External Leads. IRS partners with financial institutions and other external parties to obtain information about emerging IDT refund trends and fraudulent returns that have passed through IRS detection systems. In August 2014, we reported that IRS provides limited feedback to external parties on IDT external leads they submit and offers external parties limited general information on IDT refund fraud trends and recommended that IRS provide actionable feedback to all lead generating third parties. In November 2015, IRS reported that it had developed a database to track leads submitted by financial institutions and the results of those leads. IRS also stated that it had held two sessions with financial institutions to provide feedback on external leads provided to IRS. In December 2015, IRS officials stated that the agency sent a customer satisfaction survey asking financial institutions for feedback on the external leads process and was considering other ways to provide feedback to financial institutions. In April 2016, IRS officials stated they plan to analyze preliminary survey results by mid-April 2016. Additionally, IRS officials reported that the agency shared information with financial institutions in March 2016 and plans to do so on a quarterly basis, with the next information sharing session scheduled in June 2016. IRS and industry partners have characterized that returns processing and refund issuance during this filing season has been generally smooth. Through April 1, 2016, IRS had processed about 95 million returns and issued 76 million refunds totaling about $215 billion. While IRS experienced a major system failure in February that halted returns processing for about a day, the agency reported that it had minimal effect on overall processing of returns and refunds. In addition to filing returns, many taxpayers often call IRS for assistance. IRS's telephone service has generally improved in 2016 over last year. From January 1 through March 19, 2016 IRS received about 35.4 million calls to its automated and live assistor telephone lines, about a 2 percent decrease compared to the same period last year. Of the 13.4 million calls seeking live assistance, IRS had answered 9.1 million calls--a 75 percent increase over the 5.2 million calls answered during the same period last year. IRS anticipated that 65 percent of callers seeking live assistance would receive it this filing season, which runs through April 18, and 47 percent of callers would receive live assistance through the entire 2016 fiscal year. As of March 19, 2016, 75 percent of callers had received live assistance, an increase from 38 percent during the same period last year. Further, the average wait time to speak to an assistor also decreased from 24 to 9 minutes. As we reported in March 2016, however, IRS's telephone level of service for the full fiscal year has yet to reach the levels it had achieved in earlier years. IRS attributed this year's service improvement to a number of factors. Of the additional $290 million IRS received in December 2015, it allocated $178.4 million (61.5 percent) for taxpayer services to make measurable improvements in its telephone level of service. With the funds, IRS hired 1,000 assistors who began answering taxpayer calls in March, in addition to the approximately 2,000 seasonal assistors it had hired in fall 2015. To help answer taxpayer calls before March, IRS officials told us that they detailed 275 staff from one of its compliance functions to answer telephone calls. IRS officials said they believe this step was necessary because the additional funding came too late in the year to hire and train assistors to fully cover the filing season. IRS also plans to use about 600 full-time equivalents of overtime for assistors to answer telephone calls and respond to correspondence in fiscal year 2016, compared to fewer than 60 full-time equivalents of overtime used in fiscal year 2015. In December 2014, we recommended that IRS systematically and periodically compare its telephone service to the best in business to identify gaps between actual and desired performance. IRS disagreed with this recommendation, noting that it is difficult to identify comparable organizations. We do not agree with IRS's position; many organizations run call centers that would provide ample opportunities to benchmark IRS's performance. In fall 2015, Department of the Treasury (Treasury) and IRS officials said they had no plans to develop a comprehensive customer service strategy or specific goals for telephone service tied to the best in the business and customer expectations. Without such a strategy, Treasury and IRS can neither measure nor effectively communicate to Congress the types and levels of customer service taxpayers should expect and the resources needed to reach those levels. Therefore, in December 2015 we suggested that Congress consider requiring that Treasury work with IRS to develop a comprehensive customer service strategy. In April 2016, IRS officials told us that the agency established a team to consider our prior work in developing this strategy or benchmarking its telephone service. In summary, while IRS has made progress in implementing information security controls, it needs to continue to address weaknesses in access controls and configuration management and consistently implement all elements of its information security program. The risks IRS and the public are exposed to have been illustrated by recent incidents involving public- facing applications, highlighting the importance of securing systems that contain sensitive taxpayer and financial data. In addition, fully implementing key elements of a breach response program will help ensure that when breaches of sensitive data do occur, their impact on affected individuals will be minimized. Weaknesses in information security can also increase the risk posed by identity theft refund fraud. IRS needs to establish an approach for addressing identity theft refund fraud that is informed by assessing the cost, benefits, and risks of IRS's various authentication options and improving the reliability of fraud estimates. While this year's tax filing season has generally gone smoothly and IRS has improved customer service, it still needs to develop a comprehensive approach to customer service that will meet the needs of taxpayers while ensuring that their sensitive information is adequately protected. Chairman Hatch, Ranking Member Wyden, and Members of the Committee, this concludes my statement. I look forward to answering any questions that you may have at this time. If you have any questions regarding this statement, please contact Gregory C. Wilshusen at (202) 512-6244 or wilshuseng@gao.gov, Nancy Kingsbury at (202) 512-2928 or kingsburyn@gao.gov, or James R. McTigue, Jr. at (202) 512-9110 or mctiguej@gao.gov or Jessica K. Lucas-Judy at (202) 512-9110 or LucasJudyJ@gao.gov. Other key contributors to this statement include Jeffrey Knott, Neil A. Pinney, and Joanna M. Stamatiades (assistant directors); Dawn E. Bidne; Mark Canter; James Cook; Shannon J. Finnegan; Lee McCracken; Justin Palk; J. Daniel Paulk; Erin Saunders Rath; and Daniel Swartz. 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.	In collecting taxes, processing returns, and providing taxpayer service, IRS relies extensively on computerized systems. Thus it is critical that sensitive taxpayer and other data are protected. Recent data breaches at IRS highlight the vulnerability of taxpayer information. In addition, identity theft refund fraud is an evolving threat to honest taxpayers and tax administration. This crime occurs when a thief files a fraudulent return using a legitimate taxpayer's identity and claims a refund. In 2015, GAO added identity theft refund fraud to its high-risk area on the enforcement of tax laws and expanded its government-wide high-risk area on federal information security to include the protection of personally identifiable information. This statement discusses (1) IRS information security controls over financial and tax processing systems, (2) IRS actions to address identity theft refund fraud, and (3) the status of selected IRS filing season operations. This statement is based on previously published GAO work as well as an update of selected data. In March 2016, GAO reported that the Internal Revenue Service (IRS) had instituted numerous controls over key financial and tax processing systems; however, it had not always effectively implemented other controls intended to properly restrict access to systems and information, among other security measures. In particular, while IRS had improved some of its access controls, weaknesses remained in key controls for identifying and authenticating users, authorizing users' level of rights and privileges, encrypting sensitive data, auditing and monitoring network activity, and physically securing facilities housing its information technology resources. These weaknesses were due in part to IRS's inconsistent implementation of its agency-wide security program, including not fully implementing prior GAO recommendations. GAO concluded that these weaknesses collectively constituted a significant deficiency for the purposes of financial reporting for fiscal year 2015. As a result, taxpayer and financial data continue to be exposed to unnecessary risk. Identity theft refund fraud also poses a significant challenge. IRS estimates it paid $3.1 billion in these fraudulent refunds in filing season 2014, while preventing $22.5 billion (see figure). The full extent is unknown because of the challenges inherent in detecting this form of fraud. IRS has taken steps to combat identity theft refund fraud such as improving phone service for taxpayers to report suspected identity theft and working with industry, states, and financial institutions to detect and prevent it. However, as GAO reported in August 2014 and January 2015, additional actions can further assist the agency in addressing this crime, including pre-refund matching of taxpayer returns with information returns from employers, and assessing the costs, benefits, and risks of improving methods for authenticating taxpayers. In addition, the Consolidated Appropriations Act 2016 includes a provision that would help IRS with pre-refund matching and also includes an additional $290 million to enhance cybersecurity, combat identity theft refund fraud, and improve customer service. According to IRS and industry partners, the 2016 filing season has generally gone smoothly, with about 95 million returns and $215 billion in refunds processed through April 1, 2016. In addition, IRS increased its level of phone service to taxpayers, although it has not developed a comprehensive strategy for customer service as GAO recommended in December 2015. In addition to 49 prior recommendations that had not been implemented, GAO made 45 new recommendations to IRS to further improve its information security controls and the implementation of its agency-wide information security program. GAO has also made recommendations to help IRS combat identity theft refund fraud, such as assessing costs, benefits, and risks of taxpayer authentication options.	5,021	728
In 1991, we reported that, historically, INS leadership had allowed INS' organizational structure to become decentralized without adequate controls. Specifically, its regional structure had created geographical separation among INS programs and hampered resource allocation and consistent program implementation. The field structure designed to carry out INS' enforcement functions was bifurcated between districts and Border Patrol sectors, resulting in uncoordinated, overlapping programs. In addition, only a single senior INS headquarters manager supervised INS' 33 district directors and 21 Border Patrol chiefs. In 1994, with the appointment of a new Commissioner, INS implemented an organizational structure intended to remedy at least two problems. First, the headquarters operations office's unrealistically large span of control resulting in uneven and poorly coordinated field performance. Second, the lack of focus on program planning resulting from the operations office's preoccupation with matters that should have been handled by field managers. The Commissioner shifted some management authority to officials closer to field activities. While INS made some progress toward achieving its reorganization goals, its organizational structure is still in a state of flux and some problems persist. For example, in 1997 we reported that the responsibilities and authority of the Office of Field Operations and Office of Programs were unclear. We recommended, among other things, that the INS Commissioner provide written guidance on (1) the responsibilities and authorities of these two offices and (2) the appropriate coordination and communication methods between these two offices, and between the Office of Programs and field offices. Although INS has taken some steps to implement our 1997 recommendations, they have yet to be completed because, according to INS, these recommendations relate to INS restructuring that is currently under study. As previously mentioned, INS' mission involves carrying out two primary functions--enforcing immigration laws and providing services or benefits to eligible legal immigrants. These functions often translate into competing priorities at the program level that need to be balanced for effective program implementation. All too often, the emphasis placed on one over the other results in ineffective enforcement or poor benefit delivery. An example of this inability to balance these priorities can be found in our September 2000 report on the processing of visas for specialty occupations, called H-1B visas. The performance appraisal process for staff that evaluates the merits of applications filed with INS (called adjudicators) focused mainly on the number of applications reviewed, not the quality of the review. INS rewarded those adjudicators who processed the greatest number of applications over those who processed fewer applications. Some adjudicators told us that because of pressure to adjudicate cases quickly, they did not routinely use investigations staff to look into potentially fraudulent applications because doing so would take more time and reduce the number of applications they could complete. INS investigators following up on approved applications found instances of fraud; for example, they found employers who created shell corporations and false credentials and documents for aliens ineligible for H-1B employment. We found other examples where the goal of providing timely service delivery has negatively impacted INS' enforcement goal of providing benefits to only eligible aliens. In our May 2001 report on INS application processing, we stated that INS' policy is to grant work authorization to applicants who file for adjustment of status to that of a permanent resident before it adjudicates their application. This policy is intended to prevent aliens from having to wait for INS to adjudicate their application before they can work. However, in fiscal year 2000 INS denied about 80,000 applicants for adjustment of status (about 14 percent of all the adjustment of status applications completed) and had to revoke their work authorization. Because these aliens had work authorization while waiting for their application to be processed, they could have developed a work history that may have facilitated their obtaining employment even after INS' efforts to officially revoke their work authorization. A senior INS official stated that the policy to grant work authorization before the adjustment of status application is decided is intended to be fair to the majority of adjustment of status applicants who are approved. An investigation into INS' initiative to process naturalization applications more quickly found the initiative to be fraught with quality and integrity problems resulting in ineligible applicants receiving citizenship. According to a Department of Justice Office of Inspector General (OIG) report on INS' Citizenship USA initiative launched in 1995, INS made the timely completion of naturalization applications its guiding principle at the expense of accuracy and quality in determining eligibility. As a result of the problems found, INS instituted naturalization quality control procedures to enhance the integrity of the process. We are finding a similar situation in our ongoing review for this subcommittee of INS' efforts to deter immigration benefit fraud. We will discuss this and other issues related to immigration benefit fraud in a report to be released later this year. Other researchers have also found that INS had difficulty in balancing its enforcement and service delivery priorities. For example, the Visa Waiver Program allows nationals of certain counties to enter the United States with just a passport. No visa is required. According to a Department of Justice OIG report, abuse of the program poses a threat to national security and increases illegal immigration. The report found that aliens used stolen passports from Visa Waiver countries to illegally enter the United States. In one case, the OIG found that 27 stolen Icelandic passports had been used to smuggle children into the United States.Although the passport numbers of the stolen Icelandic passports had been entered into a lookout database, INS airport inspectors were not entering the passport numbers of passengers arriving with Icelandic passports into the lookout database. INS officials told the OIG investigators that manually keying in these passport numbers into the system would take too long and would hamper INS' ability to inspect all passengers from a flight within 45 minutes, as mandated by law. An INS contractor that evaluated INS' immigration benefits process in 1999 found that INS needed to strengthen the integrity of the process. The study found that INS had no standard quality control program for ensuring that applications were processed consistently. Although some adjudicators believed the number of fraudulent applications submitted was significantly higher than the number they were detecting, they received little training in fraud detection. According to the report, some management and operations personnel indicated that performance evaluations in large part are based on the quantity of applications processed. The report concluded that whether employees receive incentives and rewards depends more on the quantity of applications processed rather than on fraud detection. Therefore, adjudicators had no incentives to actively search out fraud. As we reported in our applications processing report, despite these pressures to complete applications more quickly, INS' backlog of applications increased to about 4 million applications by the end of fiscal year 2000, a four-fold increase since 1994. As of September 30, 2001 about 767,000 applicants out of almost 3 million with pending applications had been waiting at least 21 months for INS to process their application. In our 1997 management report, we found that poor communication was a problem, especially between headquarters and field units. For example, field and policy manuals were out of date and there was not one place that program staff could go for direction. Over one half of the employees we surveyed in preparing that report believed that INS had poor communications and that information was disseminated poorly. As noted earlier in our testimony, how INS' Office of Programs and Office of Field Operations were to coordinate was still unclear. Our recent work shows that coordination and communication is still a problem. For example, although both the Border Patrol and INS' Office of Investigations have anti-smuggling units that conduct alien smuggling investigations, these units operate through separate chains of command with different reporting structures. In May 2000, we reported that alien smuggling was a growing problem, and that the Border Patrol and Investigations anti-smuggling units operated autonomously, resulting in a lack of program coordination. Further, this lack of coordination sometimes led to different anti-smuggling units opening investigations on the same target. INS Investigations officials told us that the autonomy of the individual units and the lack of a single chain of command to manage INS' anti-smuggling investigations were major obstacles to building a more effective anti-smuggling program. Communicating the necessary information to the appropriate individuals has also been a problem. In our H-1B report, we stated that adjudicators told us that they did not have easy access to case-specific information that would have helped them correctly decide whether an application should be approved or denied. For example, evidence of a fraudulent employer or falsified worker credentials either was not available to the adjudicator or could only be accessed through a time-consuming and complicated process. Consequently, a previously denied application could be resubmitted and approved by a different adjudicator. At the time of our review, INS officials told us that INS was in the process of upgrading the computer system that tracks H-1B applications, which could make more accurate and up to date information available on-line for adjudicators. Our work and the work of an INS contractor both found that INS did not have a structure in place to manage the information that adjudicators needed to make correct decisions. Information systems were not easily accessible to all adjudicators, so these systems were generally not queried as part of the adjudication process. INS had no single repository of information where adjudicators could find the most up to date information on such things as adjudication processes and legal and regulatory policies. In one case, the lack of communication and unclear policies and procedures had tragic consequences. In January 1999, police in Texas obtained a warrant for the arrest of Rafael Resendez-Ramirez, the "railway killer" who traveled around the United States by freight train and committed murders near railroad lines. In early 1999 police contacted INS Investigations staff in Houston Texas several times about placing a "border lookout" for Resendez-Ramirez in case he was apprehended at the border. According to a Department of Justice OIG report, none of the Investigations staff contacted by the police thought to inform the police about the existence of IDENT, INS' automated fingerprint identification system. The Investigations staff also failed to enter a lookout in IDENT in case Resendez-Ramirez was apprehended trying to cross the border. On June 1, 1999, the Border Patrol apprehended Resendez-Ramirez trying to cross illegally and had him processed through the IDENT system. Because no border lookout had been placed, however, the Border Patrol voluntarily returned him to Mexico in accordance with standard Border Patrol practices. He subsequently returned illegally to the United States and committed four more murders before he was captured. INS' Houston investigations staff provided OIG investigators with various reasons as to why they did not mention IDENT or its lookout capability to police or enter a lookout in IDENT, including the following: They were unfamiliar with IDENT and how it worked. They never received any IDENT training. They were unaware IDENT had a lookout feature. They thought IDENT was a system primarily for the Border Patrol to use. The OIG concluded that the lack of knowledge about IDENT was largely the result of broader problems in the way INS implemented and monitored IDENT. INS failed to (1) (1) ensure that components outside of the Border Patrol, such as Investigations, understood IDENT policies, particularly the lookout policy and (2) provide adequate IDENT training for all INS staff. INS and the FBI are currently working on integrating IDENT with the FBI's automated fingerprint system to improve the quality and accuracy of criminal identification so that such mistakes can be averted in the future. Effective communication has also been a problem between INS and local communities. In August 2001, we reported that since 1994 as INS' Border Patrol has increased enforcement efforts in certain locations as part of its strategy to deter illegal entry along the southwest border, illegal alien traffic shifted to other locations. Officials from some border communities told us that they were caught by surprise by the increase in the number of illegal aliens apprehended in their communities. INS has recognized the need to improve communications with the public regarding its strategy and its potential implications and has increased its outreach efforts. INS has had long-standing difficulty developing and fielding information systems to support its program operations. In 1990, we reported that INS managers and field officials did not have adequate, reliable, and timely information to effectively carry out the Service's mission. We also reported that INS had not conducted a comprehensive agency-wide information needs assessment. As a result, program and management data were kept in a loose collection of automated systems as well as a number of ad-hoc labor-intensive manual systems. Effectively using information technology continues to remain a challenge for INS. In August 2000, we reported that INS did not have a "blueprint" to guide the development of its information systems. The absence of such a plan increases the risk that the information systems in which hundreds of millions of dollars are invested each year will not be well integrated or compatible and will not support mission needs. In December 2000, we reported that INS had limited capability to effectively manage its planned and ongoing information technology investments. While INS has some important information technology management capabilities in place, it has to do considerable work to fully implement mature and effective processes. The Department of Justice agreed with our recommendation that INS develop and submit a plan to Justice for implementing investment management process improvements. INS is in the process of developing this plan. The lack of adequate information technology systems has significantly impacted INS' ability to perform its core missions. As we reported in our applications processing report, INS headquarters and field staff cited automation problems as the number one factor affecting INS' ability to process applications in a timely manner to reduce backlogs. INS has no national case management system for applications filed at its 33 district offices. Most of these offices process applications manually. As a result, these offices cannot determine the number of pending cases, identify problem areas or bottlenecks, establish processing priorities, deploy staff based on workload, and ensure cases are processed in the order received. Due to the lack of any automated system, staff spend considerable time responding to applicants' inquires on the status of their case, which takes time away from application processing. Existing INS systems used to process applications do not provide accurate and reliable data. In our applications processing report we stated that the system INS Service Centers use to process some applications frequently fails to operate and does not always update data to INS' mainframe computer as it should. This lack of automation has resulted in INS expending considerable time and effort to obtain the data it needs. In our applications processing report we also stated that lack of reliable data was the primary reason INS undertook a time-consuming and costly hand-count of all pending applications in September 2000. INS undertook the hand-count to get an accurate count of pending applications hoping to obtain an unqualified opinion on its fiscal year 2000 financial statements. According to INS officials, the cost to complete this hand-count was high in terms of lost production and staff time. INS suspended nearly all case processing for 2-3 weeks. Due to the lack of accurate data in its computer systems, INS will have to do another hand-count of all pending applications at the end of fiscal year 2001 if it hopes to obtain an unqualified opinion on its financial statement. As a result of this lack of accurate data, INS has also approved more visas than the Congress has allowed. According to an INS contractor study, INS' system that tracks these visas was not designed to keep a running total of the number of visas issued and to compare it against the annual limit to ensure that only the allowable number is approved. Consequently, in fiscal year 1999, INS approved approximately 137,000 to 138,000 H-1B visas, well over the 115,000 limit. Program management issues at INS have caused continuing concern. Our work indicates that INS needs to improve its program management in several fundamental areas, including having efficient processes and clear policies and procedures, providing adequate staff training, and aligning its workforce with its workload. The INS contractor study on immigration benefits processing found that INS' processes were inefficient. For example, INS staff spends considerable time re-entering the same data into various INS computer systems. INS did not consistently adjudicate applications because the procedures used to process applications varied by office, most field offices allowed adjudicators to review cases using minimal guidelines, and standard quality controls were lacking. The study made numerous recommendations on how to make the processes more efficient and improve quality control. We stated in our applications processing report that INS was developing a strategic plan to reengineer applications processing. INS will make decisions regarding the contractor's recommendations after completing two related strategic plans - the plan to reengineer applications processing and the information technology strategic plan. Both are in the early planning stages. INS estimated that it will take 5 years or more to develop and implement the reengineered processes and implement a service-wide automated system to process applications. Adequate staff training is also a critical aspect of program management. As noted earlier in our testimony, an INS contractor study found that INS adjudicators received little training in fraud detection. According to a November 2000 INS report prepared as part of INS' Government Performance and Results Act reporting requirements, the INS workforce is not well supported in terms of training. Advanced training classes have been cut back or delayed. According to the report, because of the growing workforce and these training cutbacks, INS will have a larger portion of its workforce that is relatively inexperienced and inadequately trained for its work.	The Immigration and Naturalization's (INS) organizational structure has led to recurring management problems, including an inability to balance competing priorities, poor communications, and weaknesses in the development and fielding of critical information technology. Although restructuring may help, INS will still need to assemble the basic building blocks essential to any organization. These building blocks include clearly delineated roles and responsibilities, policies and procedures that effectively balance competing priorities, effective internal and external communication and coordination, and computer systems that provide accurate and timely information. Until these element are in place, it will be difficult to enforce the nation's immigration laws effectively.	3,844	130
While TCE and perchlorate are both DOD-classified emerging contaminants, there are key distinctions between the contaminants that affect the extent to which they are regulated, and the information that may be needed before further steps are taken to protect human health and the environment. Since 1989, a maximum contaminant level (MCL) under the Safe Drinking Water Act has been in place for TCE. In contrast, EPA has not adopted an MCL for perchlorate, although recent government- sponsored studies have raised concerns that even low-levels of exposure to perchlorate may pose serious risks to infants and fetuses of pregnant women. We provided details about EPA's evolving standards for TCE and the evolving knowledge of its health effects in our May 2007 report and June 2007 testimony on issues related to drinking water contamination on Camp Lejeune. TCE is a colorless liquid with a sweet, chloroform-like odor that is used mainly as a degreaser for metal parts. The compound is also a component in adhesives, lubricants, paints, varnishes, paint strippers, and pesticides. At one time, TCE was used as an extraction solvent for cosmetics and drug products and as a dry-cleaning agent; however, its use for these purposes has been discontinued. DOD has used the chemical in a wide variety of industrial and maintenance processes. More recently, the department has used TCE to clean sensitive computer circuit boards in military equipment such as tanks and fixed wing aircraft. Because TCE is pervasive in the environment, most people are likely to be exposed to TCE by simply eating, drinking, and breathing, according to the Department of Health and Human Services' Agency for Toxic Substances and Disease Registry (ATSDR). Industrial wastewater is the primary source of release of TCE into water systems, but inhalation is the main route of potential environmental exposure to TCE. ATSDR has also reported that TCE has been found in a variety of foods, with the highest levels in meats, at 12 to 16 ppb, and U.S. margarine, at 440 to 3,600 ppb. In fact, HHS's National Health and Nutrition Examination Survey (NHANES) suggested that approximately 10 percent of the population had detectable levels of TCE in their blood. Inhaling small amounts of TCE may cause headaches, lung irritation, poor coordination, and difficulty concentrating, according ATSDR's Toxicological Profile. Inhaling or drinking liquids containing high levels of TCE may cause nervous system effects, liver and lung damage, abnormal heartbeat, coma, or possibly death. ATSDR also notes that some animal studies suggest that high levels of TCE may cause liver, kidney, or lung cancer, and some studies of people exposed over long periods to high levels of TCE in drinking water or workplace air have shown an increased risk of cancer. ATSDR's Toxicological Profile notes that the National Toxicology Program has determined that TCE is "reasonably anticipated to be a human carcinogen" and the International Agency for Research on Cancer has determined that TCE is probably carcinogenic to humans-- specifically, kidney, liver and cervical cancers, Hodgkin's disease, and non- Hodgkin's lymphoma--based on limited evidence of carcinogenicity in humans and additional evidence from studies in experimental animals. Effective in 1989, EPA adopted an MCL of 5 ppb of TCE in drinking water supplies pursuant to the Safe Drinking Water Act. Despite EPA's regulation of TCE as a drinking water contaminant, concerns over serious long-term effects associated with TCE exposures have prompted additional scrutiny by both governmental and nongovernmental scientific organizations. For example, ATSDR initiated a public health assessment in 1991 to evaluate the possible health risks from exposure to contaminated drinking water on Camp Lejeune. The health concerns over TCE have been further amplified in recent years after scientific studies have suggested additional risks posed by human exposure to TCE. ATSDR is continuing to develop information about the possible long-term health consequences of these potential exposures in a subregistry to the National Exposure Registry specifically for hazardous waste sites. As we previously reported with respect to Camp Lejeune, those who lived on base likely had a higher risk of inhalation exposure to volatile organic compounds such as TCE, which may be more potent than ingestion exposure. Thus, pregnant women who lived in areas of base housing with contaminated water and conducted activities during which they could inhale water vapor--such as bathing, showering, or washing dishes or clothing--likely faced greater exposure than those who did not live on base but worked on base in areas served by the contaminated drinking water. Concerns about possible adverse health effects and government actions related to the past drinking water contamination on Camp Lejeune have led to additional activities, including new health studies, claims against the federal government, and federal inquiries. As a consequence of these growing concerns--and of anxiety among affected communities about these health effects and related litigation--ATSDR has undertaken a study to examine whether individuals who were exposed in utero to the contaminated drinking water are more likely to have developed certain childhood cancers or birth defects. This research, once completed later in 2007, is expected to help regulators understand the effects of low levels of TCE in our environment. In addition, some former residents of Camp Lejeune have filed tort claims and lawsuits against the federal government related to the past drinking water contamination. As of June 2007, about 850 former residents and former employees had filed tort claims with the Department of the Navy related to the past drinking water contamination. According to an official with the U.S. Navy Judge Advocate General--which is handling the claims on behalf of the Department of the Navy--the agency is currently maintaining a database of all claims filed. The official said that the Judge Advocate General is awaiting completion of the latest ATSDR health study before deciding whether to settle or deny the pending claims in order to base its response on as much objective scientific and medical information as possible. According to DOD, any future reassessment of TCE toxicity may result in additional reviews of DOD sites that utilized the former TCE toxicity values, as the action levels for TCE cleanup in the environment may change. As we discussed in our May 2005 report and April 2007 testimony, EPA has not established a standard for limiting perchlorate concentrations in drinking water under the SDWA. Perchlorate has emerged as a matter of concern because recent studies have shown that it can affect the thyroid gland, which helps to regulate the body's metabolism and may cause developmental impairments in the fetuses of pregnant women. Perchlorate is a primary ingredient in propellant and has been used for decades by the Department of Defense, the National Aeronautics and Space Administration, and the defense industry in manufacturing, testing, and firing missiles and rockets. Other uses include fireworks, fertilizers, and explosives. It is readily dissolved and transported in water and has been found in groundwater, surface water, drinking water, and soil across the country. The sources of perchlorate vary, but the defense and aerospace industries are the greatest known source of contamination. Scientific information on perchlorate was limited until 1997, when a better detection method became available for perchlorate, and detections (and concern about perchlorate contamination) increased. In 1998, EPA first placed perchlorate on its Contaminant Candidate List, the list of contaminants that are candidates for regulation, but the agency concluded that information was insufficient to determine whether perchlorate should be regulated under the SDWA. EPA listed perchlorate as a priority for further research on health effects and treatment technologies and for collecting occurrence data. In 1999, EPA required water systems to monitor for perchlorate under the Unregulated Contaminant Monitoring Rule to determine the frequency and levels at which it is present in public water supplies nationwide. Interagency disagreements over the risks of perchlorate exposure led several federal agencies to ask the National Research Council (NRC) of the National Academy of Sciences to evaluate perchlorate's health effects. In 2005, NRC issued a comprehensive review of the health effects of perchlorate ingestion, and it reported that certain levels of exposure may not adversely affect healthy adults. However, the NRC-recommended more studies on the effects of perchlorate exposure in children and pregnant women and recommended a reference dose of 0.0007 milligrams per kilogram per day. In 2005, the EPA adopted the NRC recommended reference dose, which translates to a drinking water equivalent level (DWEL) of 24.5 ppb. If the EPA were to develop a drinking water standard for perchlorate, it would adjust the DWEL to account for other sources of exposure, such as food. Although EPA has taken some steps to consider a standard, in April 2007 EPA again decided not to regulate perchlorate--citing the need for additional research--and kept perchlorate on its Contaminant Candidate List. Several human studies have shown that thyroid changes occur in human adults at significantly higher concentrations than the amounts typically observed in water supplies. However, more recent studies have since provided new knowledge and raised concerns about potential health risks of low-level exposures, particularly for infants and fetuses. Specifically, in October 2006, researchers from the Centers for Disease Control and Prevention (CDC) published the results of the first large study to examine the relationship between low-level perchlorate exposure and thyroid function in women with lower iodine levels. About 36 percent of U.S. women have these lower iodine levels. The study found decreases in a thyroid hormone that helps regulate the body's metabolism and is needed for proper fetal neural development. Moreover, in May 2007, FDA released a preliminary exposure assessment because of significant public interest in the issue of perchlorate exposure from food. FDA sampled and tested foods such as tomatoes, carrots, spinach, and cantaloupe; and other high water content foods such as apple and orange juices; vegetables such as cucumbers, green beans, and greens; and seafood such as fish and shrimp for perchlorate and found widespread low-level perchlorate levels in these items. FDA is also planning to publish, in late 2007, an assessment of exposure to perchlorate from foods, based on results from its fiscal year 2005-2006 Total Diet Study--a market basket study that is representative of the U.S. diet. Some federal funding has been directed to perchlorate studies and cleanup activities. For example, committee reports related to the DOD and EPA appropriations acts of fiscal year 2006 directed some funding for perchlorate cleanup. In the Senate committee report for the Department of Health and Human Services fiscal year 2006 appropriations act, the committee encouraged support for studies on the long-term effects of perchlorate exposure. The Senate committee report for FDA's fiscal year 2006 appropriations act directed FDA to continue conducting surveys of perchlorate in food and bottled water and to report the findings to Congress. In the current Congress, legislation has been introduced that would require EPA to establish a health advisory for perchlorate, as well as requiring public water systems serving more than 10,000 people to test for perchlorate and disclose its presence in annual consumer confidence reports. Other pending legislation would require EPA to establish a national primary drinking water standard for perchlorate. DOD has certain responsibilities with regard to emerging contaminants such as TCE that are regulated by EPA or state governments, but its responsibilities and cleanup goals are less definite for emerging contaminants such as perchlorate that lack federal regulatory standards. As we have previously reported, DOD must comply with any cleanup standards and processes under all applicable environmental laws, regulations, and executive orders, including the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA), the Resource Conservation and Recovery Act (RCRA) and the Clean Water Act's National Pollutant Discharge Elimination System (NPDES), and the SDWA. DOD's designation of perchlorate as an emerging contaminant reflects the department's recognition that the chemical has a significant potential impact on people or the Department's mission. DOD's recognition of a substance as an emerging contaminant can lead DOD to decide to take to certain cleanup efforts even in the absence of a federal regulatory standard. In addition, federal laws enacted in fiscal years 2004 and 2005 required DOD to conduct health studies and evaluate perchlorate found at military sites. For example, the Ronald W. Reagan National Defense Authorization Act for fiscal year 2005 stated that the Secretary of Defense should develop a plan for cleaning up perchlorate resulting from DOD activities when the perchlorate poses a health hazard and continue evaluating identified sites. As we reported in our 2005 perchlorate report, DOD has sometimes responded at the request of EPA and state environmental authorities-- which have used a patchwork of statutes, regulations, and general oversight authorities--to act (or require others, including DOD, to act) when perchlorate was deemed to pose a threat to human health and the environment. For example, pursuant to its authority under the Clean Water Act's NPDES program, Texas required the Navy to reduce perchlorate levels in wastewater discharges at the McGregor Naval Weapons Industrial Reserve Plant to 4 parts per billion, the lowest level at which perchlorate could be detected. Similarly, after sampling required as part of a RCRA permit detected perchlorate, Utah officials required ATK Thiokol, an explosives and rocket fuel manufacturer, to install a monitoring well to determine the extent of perchlorate contamination at their facility and take steps to prevent additional releases of perchlorate. In addition, EPA and state officials also told us during our 2005 review that they have sometimes used their general oversight responsibilities to protect water quality and human health to investigate and sample groundwater and surface water areas for perchlorate. For example, EPA asked Patrick Air Force Base and the Cape Canaveral Air Force Station, Florida, to sample groundwater for perchlorate near rocket launch sites. Previously, both installations had inventoried areas where perchlorate was suspected and conducted limited sampling. DOD officials did not find perchlorate at Patrick Air Force Base and, according to an EPA official, the Department of the Air Force said it would not conduct additional sampling at either installation until there was a federal standard for perchlorate. Finally, according to EPA, in the absence of a federal perchlorate standard, at least eight states have established nonregulatory action levels or advisories for perchlorate ranging from 1 part per billion to 51 parts per billion. (See table 1.) Massachusetts is the only state to have established a drinking water standard--set at 2 ppb. The California Department of Health Services reports that California will complete the rulemaking for its proposed standard of 6 ppb later this year. States have used these thresholds to identify the level at which some specified action must be taken by DOD and other facilities in their state, in the absence of a federal standard. For example, Oregon initiated in-depth site studies to determine the cause and extent of perchlorate contamination when concentrations of 18 ppb or greater are found. Nevada required the Kerr-McGee Chemical site in Henderson to treat groundwater and reduce perchlorate concentration releases to 18 ppb, which is Nevada's action level for perchlorate. Utah officials told us that while the state did not have a written action level for perchlorate, it may require the responsible party to undertake cleanup activities if perchlorate concentrations exceed 18 ppb. DOD is undertaking a number of activities to address emerging contaminants in general, including the creation of the Materials of Evolving Regulatory Interest Team (MERIT) to systematically address the health, environmental, and safety concerns associated with emerging contaminants. As noted above, DOD is required to follow EPA regulations for monitoring and cleanup of TCE. In addition, DOD is working with ATSDR, which has projected a December 2007 completion date for its current study of TCE's health effects on pregnant women and their children. In the absence of a federal standard, DOD has adopted its own perchlorate policies for sampling and cleanup activities or is working under applicable state guidelines. DOD created MERIT to help address the health, environmental, and safety concerns associated with emerging contaminants. According to DOD, MERIT has focused on materials that have been or are used by DOD, or are under development for use, such as perchlorate, TCE, RDX, DNT and new explosives, naphthalene, perfluorooctanoic acid (PFOA), hexavalent chromium (i.e., chromium VI), beryllium, and nanomaterials. MERIT's initiatives include pollution prevention, detection/analytical methods, human health studies, treatment technologies, lifecycle cost analysis, risk assessment and risk management, and public outreach. Another of MERIT's activities was to create an Emerging Contaminant Action List of materials that DOD has assessed and judged to have a significant potential impact on people or DOD's mission. The current list includes five contaminants--perchlorate, TCE, RDX, naphthalene, and hexavalent chromium. To be placed on the action list, the contaminant will generally have been assessed by MERIT for its impacts on (1) environment, safety, and health (including occupational and public health), (2) cleanup efforts, (3) readiness and training, (4) acquisition, and (5) operation and maintenance activities. In 1979, EPA issued nonenforceable guidance establishing "suggested no adverse response levels" for TCE in drinking water. These levels provided EPA's estimate of the short- and long-term exposure to TCE in drinking water for which no adverse response would be observed and described the known information about possible health risks for these chemicals. However, the guidance for TCE did not suggest actions that public water systems should take if TCE concentrations exceeded those values. Subsequently, in 1989, EPA set an enforceable MCL for TCE of 5 micrograms per liter, equivalent to 5 ppb in drinking water. The new standard served as a regulatory basis for many facilities to take concrete action to measure and control TCE. According to EPA's Region 4 Superfund Director, for example, 46 sites on Camp Lejeune have since been identified for TCE cleanup. The Navy and EPA have selected remedies for 30 of those sites, and the remaining 16 are under active investigation. The first Record of Decision was signed in September 1992 and addressed contamination of groundwater in the Hadnot Point Area, one of Camp Lejeune's water systems. Remedies to address groundwater contamination include groundwater "pump and treat" systems, in-situ chemical oxidation, and monitored natural attenuation. DOD contends that it is aggressively treating TCE as part of its current cleanup program. It notes that the department uses much less TCE than in the past and requires strict handling procedures and pollution prevention measures to prevent exposure to TCE and the release of TCE into the environment. Specifically, DOD has replaced products containing TCE with other types of cleaning agents such as citrus-based agents, mineral oils and other non-toxic solutions. In the absence of a federal perchlorate standard, DOD has adopted its own policies with regard to sampling and cleanup. The 2003 Interim Policy on Perchlorate Sampling required the military services--Army, Navy, Air Force, and Marines--to sample on active installations (1) where a reasonable basis existed to suspect that a perchlorate release occurred as a result of DOD activities, and (2) a complete human exposure pathway likely existed or (3) where a particular installation must do so under state laws or applicable federal regulations such as the NPDES permit program. However, DOD's interim policy on perchlorate did not address cleanup responsibilities nor did it address contamination at closed installations. As we detailed in our previous work, DOD only sampled for perchlorate on closed installations when requested by EPA or a state agency, and only cleaned up active and closed installations when required by a specific environmental law, regulation, or program such as the environmental restoration program at formerly used defense sites. For example, at EPA's request, the U.S. Army Corps of Engineers (Corps) installed monitoring wells and sampled for perchlorate at Camp Bonneville, a closed installation near Vancouver, Washington. Utah state officials also reported to us that DOD removed soil containing perchlorate at the former Wendover Air Force Base in Utah, where the Corps found perchlorate in 2004. However, as we previously reported, DOD cited reluctance to sample on or near active installations because of the lack of a federal regulatory standard for perchlorate. In the absence of a federal standard, DOD has also worked with individual states on perchlorate sampling and cleanup. For example, in October 2004, DOD and California agreed to prioritize perchlorate sampling at DOD facilities in California, including identifying and prioritizing the investigation of areas on active installations and military sites (1) where the presence of perchlorate is likely based on previous and current defense-related activities and (2) near drinking water sources where perchlorate was found. In January 2006, DOD updated its policy with the issuance of its Policy on DOD Required Actions Related to Perchlorate. The new policy applies broadly to DOD's active and closed installations and formerly used defense sites within the United States, its territories and possessions. It directs DOD to test for perchlorate and take certain cleanup actions. The policy also acknowledges the importance of EPA direction in driving DOD's response to emerging contaminants. It stated, for example, that its adoption of 24 ppb as the current level of concern for managing perchlorate was in response to EPA's adoption of an oral reference dose that translates to a Drinking Water Equivalent Level of 24.5 ppb. The policy also states that when EPA or the states adopt standards for perchlorate, "DOD will comply with applicable state or federal promulgated standards whichever is more stringent." The 2006 policy directs DOD to test for perchlorate when it is reasonably expected that a release has occurred. If perchlorate levels exceed 24 ppb, a site-specific risk assessment must be conducted. When an assessment indicates that the perchlorate contamination could result in adverse health effects, the site must be prioritized for risk management. DOD uses a relative-risk site evaluation framework across DOD to evaluate the risks posed by one site relative to other sites and to help prioritize environmental restoration work and to allocate resources among sites. The policy also directs DOD's service components to program resources to address perchlorate contamination under four DOD programs-- environmental restoration, operational ranges, DOD-owned drinking water systems, and DOD wastewater effluent discharges. Under the 2006 perchlorate policy, DOD has sampled drinking water, groundwater, and soil where the release of perchlorate may result in human exposure and responded where it has deemed appropriate to protect public health. As we have reported, DOD is responsible for a large number of identified sites with perchlorate contamination, and the department has allotted significant resources to address the problem. According to DOD, sampling for perchlorate has occurred at 258 active DOD installations or facilities. Through fiscal year 2006, DOD reported spending approximately $88 million on perchlorate-related research activities, including $60 million for perchlorate treatment technologies, $9.5 million on health and toxicity studies, and $11.6 million on pollution prevention. Additional funds have been spent on testing technology and cleanup. DOD also claims credit for other efforts, including strict handling procedures to prevent the release of perchlorate into the environment and providing information about perchlorate at DOD facilities and DOD's responses. For example, DOD posts the results of its perchlorate sampling, by state, on MERIT's Web site. As we have previously reported, DOD must comply with cleanup standards and processes under applicable laws, regulations and executive orders, including EPA drinking water standards and state-level standards. In the absence of a federal perchlorate standard, DOD has also initiated perchlorate response actions to clean up perchlorate contamination at several active and formerly used defense sites under its current perchlorate policy. For example, at Edwards Air Force Base in California, DOD has treated 32 million gallons of ground water under a pilot project for contaminants that include perchlorate. In addition, DOD has removed soil and treated groundwater at the Massachusetts Military Reservation and Camp Bonneville in Washington State. In conclusion, Mr. Chairman, DOD faces significant challenges, and potentially large costs, in addressing emerging contaminants, particularly in light of the scientific developments and regulatory uncertainties surrounding these chemicals and materials. To help address them, DOD recently identified five emerging contaminants for which it is developing risk management options. As in the case of TCE, DOD took action to address contamination after EPA established an MCL in 1989. DOD has stated that further efforts to address perchlorate would require a regulatory standard from EPA and/or the states. The fact that some states have moved to create such standards complicates the issue for DOD by presenting it with varying cleanup standards across the country. As the debate over a federal perchlorate standard continues, the recently- issued health studies from CDC and FDA may provide additional weight to the view that the time for such a standard may be approaching. Until one is adopted, DOD will continue to face the challenges of differing regulatory requirements in different states and continuing questions about whether its efforts to control perchlorate contamination are necessary or sufficient to protect human health. Mr. Chairman, this concludes my prepared statement. I would be happy to respond to any questions that you or Members of the Subcommittee may have at this time. For further information about this testimony, please contact John Stephenson 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 statement. Contributors to this testimony include Steven Elstein, Assistant Director and Terrance Horner, Senior Analyst. Marc Castellano, Richard Johnson, and Alison O'Neill also made key contributions. Defense Health Care: Issues Related To Past Drinking Water Contamination at Marine Corps Base Camp Lejeune, GAO-07-933T (June 12, 2007). Defense Health Care: Activities Related To Past Drinking Water Contamination at Marine Corps Base Camp Lejeune, GAO-07-276 (May 11, 2007). Perchlorate: EPA Does Not Systematically Track Incidents of Contamination, GAO-07-797T (April 25, 2007). Environmental Information: EPA Actions Could Reduce the Availability Of Environmental Information To The Public, GAO-07-464T (February 6, 2007). Military Base Closures: Opportunities Exist to Improve Environmental Cleanup Cost Reporting and to Expedite Transfer of Unneeded Property, GAO-07-166 (January 30, 2007). Perchlorate: A System to Track Sampling and Cleanup Results Is Needed, GAO-05-462 (May 20, 2005). Military Base Closures: Updated Status of Prior Base Realignments and Closures, GAO-05-138 (January 13, 2005). Environmental Contamination: DOD Has Taken Steps To Improve Cleanup Coordination At Former Defense Sites But Clearer Guidance Is Needed To Ensure Consistency, GAO-03-146 (March 28, 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.	DOD defines emerging contaminants as chemicals or materials with (1) perceived or real threat to health or the environment and (2) lack of published standards or a standard that is evolving or being reevaluated. Two emerging contaminants--trichloroethylene (TCE) and perchlorate--are of particular concern to DOD because they have significant potential to impact people or DOD's mission. TCE, a degreasing agent in metal cleaning which has been used widely in DOD industrial and maintenance processes, has been documented at low exposure levels to cause headaches and difficulty concentrating. High-level exposure may cause dizziness, headaches, nausea, unconsciousness, cancer, and possibly death. Similarly, perchlorate has been used by DOD, NASA, and others in making, testing, and firing missiles and rockets. It has been widely found in groundwater, surface water, and soil across the United States, Perchlorate health studies have documented particular risks to fetuses of pregnant women. GAO was asked for testimony to summarize its past work on perchlorate-, TCE-, and defense-activities related to (1) the state of knowledge about the emerging contaminants TCE and perchlorate, (2) DOD responsibilities for managing TCE and perchlorate contamination at its facilities, and (3) DOD activities to address TCE and perchlorate contamination. While TCE and perchlorate are both classified by DOD as emerging contaminants, there are important distinctions in how they are regulated and in what is known about their health and environmental effects. Since 1989, EPA has regulated TCE in drinking water. However, health concerns over TCE have been further amplified in recent years after scientific studies have suggested additional risks posed by human exposure to TCE. Unlike TCE, no drinking water standard exists for perchlorate--a fact that has caused much discussion in Congress and elsewhere. Recent Food and Drug Administration data documenting the extent of perchlorate contamination in the nation's food supply has further fueled this debate. While DOD has clear responsibilities to address TCE because it is subject to EPA's regulatory standard, DOD's responsibilities are less definite for perchlorate due to the lack of such a standard. Nonetheless, perchlorate's designation by DOD as an emerging contaminant has led to some significant control actions. These actions have included responding to requests by EPA and state environmental authorities, which have used a patchwork of statutes, regulations, and general oversight authorities to address perchlorate contamination. Pursuant to its Clean Water Act authorities, for example, Texas required the Navy to reduce perchlorate levels in wastewater discharges at the McGregor Naval Weapons Industrial Reserve Plant to 4 parts per billion (ppb), the lowest level at which perchlorate could be detected at the time. In addition, in the absence of a federal perchlorate standard, at least nine states have established nonregulatory action levels or advisories for perchlorate ranging from 1 ppb to 51 ppb. Nevada, for example, required the Kerr-McGee Chemical site in Henderson to treat groundwater and reduce perchlorate releases to 18 ppb, which is Nevada's action level for perchlorate. While nonenforceable guidance had existed previously, it was not until EPA adopted its 1989 TCE standard that many DOD facilities began to take concrete action to control the contaminant. According to EPA, for example, 46 sites at Camp Lejeune have since been identified for TCE cleanup. The Navy and EPA have selected remedies for 30 of those sites, and the remaining 16 are under active investigation. Regarding perchlorate, in the absence of a federal standard DOD has implemented its own policies on sampling and cleanup, most recently with its 2006 Policy on DOD Required Actions Related to Perchlorate. The policy applies broadly to DOD's active and closed installations and formerly used defense sites within the United States and its territories. It requires testing for perchlorate and certain cleanup actions and directs the department to comply with applicable federal or state promulgated standards, whichever is more stringent. The policy notes, that DOD has established 24 ppb as the current level of concern for managing perchlorate until the promulgation of a formal standard by the states and/or EPA.	6,287	966
As you know, the cost of the Decennial Census has steadily increased during the past 40 years, in part because the nation's population has steadily grown larger, more diverse, and increasingly difficult to enumerate. For example, at about $13 billion, the 2010 Census was the costliest U.S. census in history and was 56 percent more costly than the $8.1 billion 2000 Census (in constant 2010 dollars). To help save costs, in preparing for the 2020 Census, the Bureau has been researching and testing new methods and technologies to redesign the Census to more cost-effectively count the population while maintaining high-quality results. The Bureau's research and testing has focused on four redesign areas: Reengineering address canvassing: This involves reengineering processes for updating the Bureau's address list and maps of the nation to reduce the need for employing field staff to walk every street in the nation to verify addresses. Optimizing self-response: Includes efforts to maximize the self- response of households by, among other things, offering an Internet response option. As we have previously reported, to deliver the Internet response option, the Bureau would need to, among other things, design and develop an Internet response application, develop and acquire the IT infrastructure to support a large volume of data processing and storage, and plan communication and outreach strategies to motivate households to respond via the Internet. Using administrative records: This includes expanding the use of data previously obtained by other federal and state government agencies and commercial sources to reduce the need for costly and labor-intensive follow-up work. My colleague will address the Bureau's progress on using administrative records in his statement today. Reengineering field operations: This includes reducing the number of visits to households, automating the management of enumerator work to conduct non-response follow-up, and automating and optimizing case assignment and routing for enumerators to reduce the staffing, infrastructure, and field offices required for the 2020 Census. The Bureau has conducted several major field tests to examine the potential for each of these redesign areas: In mid-2014 the Bureau conducted the 2014 Census Test in the Maryland and Washington, D.C., areas to test new methods for conducting self-response and non-response follow-up. In early 2015 the Bureau completed the Address Validation Test, which was used to examine new methods for updating the Bureau's address list. In mid-2015 the Bureau conducted the 2015 Census Test in Arizona to test, among other things, the use of a field operations management system to automate data collection operations and provide real-time data and the ability to reduce the non-response follow-up workload using data previously provided to the government, as well as enabling enumerators to use their personally owned mobile devices to collect census data. Also in mid-2015, the Bureau conducted an optimizing self-response test in Savannah, Georgia, and the surrounding area, which was intended to further explore methods of encouraging households to respond using the Internet, such as using advertising and outreach to motivate respondents, and enabling households to respond without a Bureau-issued identification number. More recently, the Bureau began its National Content Test, which is currently ongoing and intended to, among other things, continue to test self-response modes and contact strategies and refine estimates of national self-response and Internet response rates. These tests were intended to inform the first version of the Bureau's 2020 Census Operational Plan, which is intended to outline design decisions that drive how the 2020 Census will be conducted. As part of these decisions, the Bureau has committed to aspects of the 2020 Census redesign. The operational plan articulated 326 total design decision points, which vary widely in their complexity, importance, and urgency. As of October 6, 2015, the Bureau had made decisions for about 47 percent of them related to each of the four redesign areas. For example, the Bureau has decided to conduct 100 percent of address canvassing (i.e., identifying all addresses where people could live) in the office, and target a subset of up to 25 percent for in-the-field address canvassing; offer an Internet self-response option, as well as alternative response options via telephone and paper for limited circumstances; allow people to respond without a unique census identification use mobile devices for enumerators to conduct field data collection; use administrative records to enumerate vacant units; use enterprise solutions to support the 2020 Census, when practicable; and reduce the field footprint by half in comparison to the 2010 Census (e.g., 6 regional census centers instead of 12 and up to 250 field offices instead of nearly 500). Figure 1 provides an overview of the Bureau's current plans and assumptions for the 2020 Census, resulting from the October 2015 operational plan. As a result of these decisions, the Bureau estimates saving $5.2 billion. Specifically, the Bureau estimated that repeating the design of the 2010 Census for 2020 would cost approximately $17.8 billion (in constant 2020 dollars), while successfully implementing the four redesign areas is expected to result in an overall 2020 Census cost of $12.5 billion (in constant 2020 dollars). Table 1 illustrates the estimated cost savings associated with each redesign area. Moving forward, the Bureau plans to conduct additional research and testing and further refine the design through 2018. By August 2017, the Bureau plans to begin preparations for end-to-end testing, which is intended to test all systems and operations to ensure readiness for the 2020 Census. Figure 2 shows the timeline for planned 2020 Census research and testing. Concurrent with redesigning the decennial census, the Bureau has also begun a significant effort to modernize and consolidate its survey data collection and processing functions. This is being undertaken through an enterprise-wide IT initiative called Census Enterprise Data Collection and Processing (CEDCAP). This initiative is a large and complex modernization program intended to deliver a system-of-systems for all the Bureau's survey data collection and processing functions--rather than continuing to rely on unique, survey-specific systems with redundant capabilities. For the 2020 Census, CEDCAP is expected to deliver the systems and IT infrastructure needed to implement the Bureau's redesign areas. For example: To reengineer field work, CEDCAP is expected to implement a new dynamic operational control system to track and manage field work. This system is to be able to make decisions about which visits enumerators should attempt on a daily basis using real-time data, as well as provide automated route planning to make enumerator travel more efficient. CEDCAP also includes testing the use of mobile devices, either government-furnished or employee-owned, to automate data collection in the field. To maximize self-response with the Internet response option, CEDCAP is responsible for developing and testing a web-based survey application and exploring options for establishing the IT infrastructure to support the increased volume of data processing and storage that will be needed. CEDCAP consists of 12 projects that are to deliver capabilities incrementally, over the course of at least 10 releases. The Bureau plans to roll out capabilities for the 2020 Census incrementally through 6 of these releases, while also deploying capabilities for other surveys such as the American Community Survey and Economic Census. The Bureau expects to reuse selected systems, make modifications to other systems, and develop or acquire additional systems and infrastructure. As of August 2015, the CEDCAP program was projected to cost about $548 million through 2020. However, the Bureau's past efforts to implement new approaches and systems have not always gone well. As one example, during the 2010 Census, the Bureau planned to use handheld mobile devices to support field data collection for the census, including following up with nonrespondents. However, due to significant problems identified during testing of the devices, cost overruns, and schedule slippages, the Bureau decided not to use the handheld devices for non-response follow-up and reverted to paper-based processing, which increased the cost of the 2010 Census by up to $3 billion and significantly increased its risk as it had to switch its operations to paper-based operations as a backup. Last month's issuance of the 2020 Census Operational Plan, which documents many key decisions about the redesign of the 2020 Census, represents progress; however, the Bureau faces critical challenges in delivering the IT systems needed to support the redesign areas. Specifically, with preparations for end-to-end testing less than 2 years away, the window to implement CEDCAP, which is intended to be the backbone of the 2020 Census, is narrow. Additionally, while the Bureau has demonstrated improvements in IT management, as we have previously reported, it faces critical gaps in its IT workforce planning and information security. Until it takes actions we have previously recommended to address these challenges, the Bureau is at risk of cost overruns, schedule delays, and performance shortfalls, which will likely diminish the potentially significant cost savings that it estimates will result from redesigning the census for 2020. The Bureau has not prioritized key IT-related decisions, which is a trend we have reported for the past few years. Specifically, in April 2014, we reported the Bureau had not prioritized key IT research and testing needed for the design decisions planned for the end of 2015. In particular, the Bureau had not completed the necessary plans and schedules for research and testing efforts and had not prioritized what needed to be done in time for the 2015 design decisions--a milestone that had already been pushed back by a year (see fig. 3). We concluded that, given the current trajectory and the lack of supporting schedules and plans, it was unlikely that all planned IT-related research and testing activities would be completed in time to support the 2015 design decisions--which ultimately came to fruition (as discussed later). In light of these ongoing challenges, we recommended in our April 2014 report that the Bureau prioritize its IT-related research and testing projects that need to be completed to support the design decisions and develop schedules and plans to reflect the new prioritized approach. The Bureau agreed with our recommendations and has taken steps to address them. For example, in September 2014, the Bureau released a plan that identified inputs, such as research questions, design components, and testing, that were needed to inform the operational design decisions expected in the fall of 2015. However, as we reported in February 2015, the Bureau had not yet determined how key IT research questions that had been identified as critical inputs into the design decisions--estimating the Internet self- response rate and determining the IT infrastructure for security and scalability needed to support Internet response--were to be answered. We therefore recommended that the Bureau, among other things, develop methodologies and plans for answering key IT-related research questions in time to inform key design decisions. While the recent 2020 Census Operational Plan documents many key IT- related decisions about the redesign of the census, other critical questions, including the ones identified in our February 2015 report, remain unanswered. Of greater concern, the Bureau does not intend to answer these and other questions until 2016 through 2018. Specifically, there are several significant IT decisions that are being deferred, which have implications on the CEDCAP program's ability to have production- ready systems in place in time to conduct end-to-end testing. For example, the Bureau does not plan to decide on the projected demand that the IT infrastructure and systems would need to accommodate or whether the Bureau will build or buy the needed systems until June 2016, at the earliest; the high-level design and description of the systems (referred to as the solutions architecture) until September 2016--leaving about a year to, among other things, build or acquire, integrate, and test the systems that are intended to serve as the backbone to the 2020 Census before preparations for end-to-end testing begins in August 2017; and the strategy for the use of mobile devices for field work until October 2017. Figure 4 illustrates several key IT-related decisions that have been deferred which will impact preparations for the end-to-end test and 2020 Census. Unless the Bureau makes these key decisions soon, it will likely run out of time to put CEDCAP systems in place. Institutionalizing key IT management controls, such as IT governance, system development methodology, and requirements management processes, helps establish a consistent and repeatable process for managing and overseeing IT investments and reduces the risk of experiencing cost overruns, schedule slippages, and performance shortfalls, like those that affected the previous census. However, in September 2012, we reported that the Bureau lacked a sufficiently mature IT governance process to ensure that investments are properly controlled and monitored, did not have a comprehensive system development methodology, and continued to have long-standing challenges in requirements management. We made several recommendations to address these issues, and the Bureau took actions to fully implement each of the recommendations. For example, the Bureau addressed gaps in policies and procedures related to IT governance, such as establishing guidelines on the frequency of investment review board meetings and thresholds for escalation of cost, risk, or impact issues; finalized its adoption of an enterprise system development life-cycle methodology, which included the short incremental development model, referred to as Agile, and a process for continuously improving the methodology based on lessons learned; and implemented a consistent requirements development tool that includes guidance for developing requirements at the strategic mission, business, and project levels and is integrated with its enterprise system development life-cycle methodology. As a result, the Bureau has established a consistent process for managing and overseeing its IT investments. Effective workforce planning is essential to ensure organizations have the proper skills, abilities, and capacity for effective management. While the Bureau has made progress in IT workforce planning efforts, many critical IT competency gaps remain to be filled. In September 2012 we reported, among other things, that the Bureau had not developed a Bureau-wide IT workforce plan; identified gaps in mission-critical IT occupations, skills, and competencies; or developed strategies to address gaps. Accordingly, we recommended that the Bureau establish a repeatable process for performing IT skills assessments and gap analyses and establish a process for directorates to coordinate on IT workforce planning. In response, in 2013 the Bureau completed an enterprise-wide competency assessment and identified several mission-critical gaps in technical competencies. In 2014, the Bureau established documents to institutionalize a strategic workforce planning process, identified actions and targets to close the competency gaps by December 2015, and established a process to monitor quarterly status reports on the implementation of these actions. However, as we reported in February 2015, while these are positive steps in establishing strategic workforce planning capabilities, the Bureau's workforce competency assessment identified several mission-critical gaps that would challenge its ability to deliver IT-related initiatives, such as the IT systems that are expected to be delivered by CEDCAP. For example, the Bureau found that competency gaps existed in cloud computing, security integration and engineering, enterprise/mission engineering life- cycle, requirements development, and Internet data collection. The Bureau also found that enterprise-level competency gaps existed in program and project management, budget and cost estimation, systems development, data analytics, and shared services. The Bureau has taken steps to regularly monitor and report on the status of its efforts to close competency gaps and has completed several notable actions. For example, in August 2015, the Bureau filled the position of Decennial IT Division Chief and in September 2015 awarded an enterprise-wide IT services contract for systems engineering and integration support. However, more work remains for the Bureau to close competency gaps critical to the implementation of its IT efforts. Most significantly, in July 2015, the Chief Information Officer resigned. As of October 2015, the Bureau was working to fill that gap and had an acting Chief Information Officer temporarily in the position. Additionally, there are other gaps in key positions, such as the Chief of the Office of Information Security and Deputy Chief Information Security Officer, Big Data Center Chief, Chief Cloud Architect, and the CEDCAP Assistant Chief of Business Integration, who is responsible for overseeing the integration of schedule, risks, and budget across the 12 projects. According to Bureau officials, they are working to address these gaps. Critical to the Bureau's ability to perform its data collection and analysis duties are its information systems and the protection of the information they contain. A data breach could result in the public's loss of confidence in the Bureau, thus affecting its ability to collect census data. To ensure the reliability of their computerized information, agencies should design and implement controls to prevent, limit, and detect unauthorized access to computing resources, programs, information, and facilities. Inadequate design or implementation of access controls increases the risk of unauthorized disclosure, modification, and destruction of sensitive information and disruption of service. In January 2013, we reported on the Bureau's implementation of information security controls to protect the confidentiality, integrity, and availability of the information and systems that support its mission. We concluded that the Bureau had a number of weaknesses in controls intended to limit access to its systems and information, as well as those related to managing system configurations and unplanned events. We attributed these weaknesses to the fact that the Bureau had not fully implemented a comprehensive information security program, and made 115 recommendations aimed at addressing these deficiencies. The Bureau expressed broad agreement with the report and said it would work to find the best ways to address our recommendations. As of October 29, 2015, the Bureau had addressed 66 of the 115 recommendations we made in January 2013. Of the remaining open recommendations, we have determined that 30 require additional actions by the Bureau, and for the other 19 we have work under way to evaluate if they have been fully addressed. The Bureau's progress toward addressing our security recommendations is encouraging. However, more work remains to address the recommendations. A cyber incident recently occurred at the Bureau, and while it appears to have had limited impact, it demonstrates vulnerabilities at the Bureau. Specifically, in July 2015, the Bureau reported that it had been targeted by a cyber attack aimed at gaining access to its Federal Audit Clearinghouse, which contains non-confidential information from state and local governments, nonprofit organizations, and Indian tribes to facilitate oversight of federal grant awards. According to Bureau officials, the breach was limited to this database on a segmented portion of the Bureau's network that does not touch administrative records or sensitive respondent data protected under Title 13 of the U.S. Code, and the hackers did not obtain the personally identifiable information of census and survey respondents. Given that the Bureau is planning to build or acquire IT systems to collect the public's personal information for the 2020 Census in ways that it has not for previous censuses (e.g., web-based surveys, cloud computing, and enabling mobile devices to collect census data), continuing to implement our recommendations and apply IT security best practices as it implements CEDCAP systems must be a high priority. As a result of the Bureau's challenges in key IT internal controls and its looming deadline, we identified CEDCAP as an IT investment in need of attention in our February 2015 High-Risk report. We recently initiated a review of the CEDCAP program for your subcommittees, and expect to issue a report in the spring of 2016. In conclusion, the Bureau is pursuing initiatives to significantly reform its outdated and inefficient methods of conducting decennial censuses. However, with less than 2 years remaining until the Bureau plans to have all systems and processes for the 2020 Census developed and ready for end-to-end testing, it faces challenges that pose significant risk to 2020 Census program. These include the magnitude of the planned changes to the design of the census, the Bureau's prior track record in executing large-scale IT projects, and the current lack of a permanent Chief Information Officer, among others. Moreover, the Bureau's preliminary decision deadline has come and gone, and many IT-related decisions have been deferred to 2016 through 2018. Consequently, it is running out of time to develop, acquire, and implement the production systems it will need to deliver the redesign and achieve its projected $5.2 billion in cost savings. The Bureau needs to take action to address the specific challenges we have highlighted in prior reports. If these actions are not taken, cost overruns, schedules delays, and performance shortfalls may diminish the potentially significant cost savings that the Bureau estimates will result from redesigning the census for 2020. Chairmen Meadows and Hurd, Ranking Members Connolly and Kelly, and Members of the Subcommittees, this completes my prepared statement. I would be pleased to respond to any questions that you may have. If you have any questions concerning this statement, please contact Carol Cha, Director, Information Technology Acquisition Management Issues, at (202) 512-4456 or chac@gao.gov. Other individuals who made key contributions include Shannin O'Neill, Assistant Director; Andrew Beggs; Lee McCracken; and Jeanne Sung. 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 cost of the nation's decennial census has steadily increased over the past 40 years; the 2010 Census was the most expensive to date, at about $13 billion. To achieve cost savings while still conducting an accurate count of the population, the U.S. Census Bureau is planning significant changes for the design of the 2020 Decennial Census, including major efforts to implement new technologies and IT systems supporting its surveys. For example, the Bureau is planning to offer an option for households to respond via the Internet, which requires developing new applications and IT infrastructure. This statement summarizes the critical challenges the Bureau faces in successfully delivering IT systems in time for testing redesigned 2020 Census operations. To develop this statement, GAO relied on previously published work, as well as information on steps the Bureau has taken to implement prior GAO recommendations. GAO has previously reported that the U.S. Census Bureau (Bureau) faces a number of critical challenges in developing and deploying the information technology (IT) systems and infrastructure it plans to rely on to conduct the significantly redesigned 2020 Census. Specifically, the Bureau has a major IT program under way to modernize and consolidate the multiple, duplicative systems it currently uses to carry out survey data collection and processing functions; however, with less than 2 years before preparations begin for end-to-end testing of all systems and operations to ensure readiness for the 2020 Census, there is limited time to implement it. While the Bureau documented many key decisions about the redesigned 2020 Census in the 2020 Census Operational Plan, released in October 2015, several key IT-related decisions have not been made. Specifically, the Bureau has not yet made decisions about the projected demand that the IT infrastructure would need to meet or whether it will build or buy the needed systems. This lack of prioritization of IT decisions has been a continuing trend, which GAO has previously identified. For example: In April 2014, GAO reported that the Bureau had not prioritized key IT research and testing needed for its design decisions. Accordingly, GAO recommended that the Bureau prioritize its IT-related research and testing projects. The Bureau had taken steps to address this recommendation, such as releasing a plan in September 2014 that identified research questions intended to inform the operational design decisions. In February 2015, however, GAO reported that the Bureau had not determined how key IT research questions that were identified in the September 2014 plan would be answered--such as the expected rate of respondents using its Internet response option or the IT infrastructure that would be needed to support this option. GAO recommended that the Bureau, among other things, develop methodologies and plans for answering key IT-related research questions in time to inform design decisions. However, this has not yet happened. In addition, while the Bureau has made improvements in some key IT management areas, it still faces challenges in the areas of workforce planning and information security. Specifically: It has taken steps to develop an enterprise-wide IT workforce planning process, as GAO recommended in 2012. However, the Bureau has yet to fill key positions. Most concerning, it is currently without a permanent chief information officer. The Bureau has taken steps to implement the majority of the 115 recommendations GAO made in 2013 to address information security weaknesses; however, completing this effort is necessary to ensure that sensitive information it will collect during the census is adequately protected. With the deferral of key IT-related decisions, the Bureau is running out of time to develop, acquire, and implement the systems it will need to deliver the redesign and achieve its projected $5.2 billion in cost savings. In its prior work, GAO made recommendations to the Census Bureau to prioritize IT testing and research and determine how key decisions for the 2020 Census were to be answered. GAO also made recommendations to improve IT management, workforce planning, and information security. The Bureau has taken steps to address selected recommendations, but more actions are still needed to fully address these recommendations.	4,549	823
DOE is responsible for a nationwide complex of facilities created during World War II and the Cold War to research, produce, and test nuclear weapons. Much of the complex is no longer in productive use, but it contains vast quantities of radioactive waste related to the production of nuclear material, such as plutonium-contaminated sludge, and hazardous waste, such as solvents and hazardous chemicals. Since the 1980s, DOE has been planning and carrying out activities around the complex to clean up, contain, safely store, and dispose of these materials. It is a daunting challenge, involving the development of complicated technologies and costing about $220 billion over 70 years or more. DOE has reported completing its cleanup work at 74 of the 114 sites in the complex, but those were small and the least difficult to deal with. The sites remaining to be cleaned up present enormous challenges to DOE. DOE's cleanup program is carried out primarily under two environmental laws. Under section 120 of CERCLA, EPA must, where appropriate, evaluate hazardous waste sites at DOE's facilities to determine whether the waste sites qualify for inclusion on the National Priorities List, EPA's list of the nation's most serious hazardous waste sites. For each facility listed on the National Priorities List, section 120(e) (2) of CERCLA requires DOE to enter into an interagency agreement with EPA for the completion of all necessary remedial actions at the facility. These agreements often include the affected states as parties to the agreements. These agreements may be known as Federal Facility Agreements or Tri- Party Agreements. Under amendments to RCRA contained in section 105 of the Federal Facility Compliance Act of 1992, DOE generally must develop site treatment plans for its mixed-waste sites. These plans are submitted for approval to states authorized by EPA to perform regulatory responsibilities for RCRA within their borders or to EPA if the state does not have the required authority. Upon approval of the treatment plans, the state or EPA must issue an order requiring compliance with the approved plan. The agreements are generally known as Federal Facility Compliance orders. DOE carries out its cleanup program through the Assistant Secretary for Environmental Management and in consultation with a variety of stakeholders. These include the federal EPA and state environmental agencies, county and local governmental agencies, citizen groups, advisory groups, Native American tribes, and other organizations. In most cases, DOE's regulators are parties to the compliance agreements. Other stakeholders advocate their views through various public involvement processes including site-specific advisory boards. Compliance agreements in effect at DOE sites can be grouped into three main types (see table 1). Agreements of the first type--those specifically required by CERCLA or by RCRA--are in effect at all of DOE's major sites. They tend to cover a relatively large number of cleanup activities and have the majority of schedule milestones that DOE must meet. By contrast, agreements that implement court-ordered settlements exist at only a few DOE sites, tend to be focused on a specific issue or concern, and have fewer associated schedule milestones. These agreements are typically between DOE and states. The remaining agreements are based on either federal or state environmental laws and address a variety of purposes, such as cleaning up spills of hazardous waste or remediating groundwater contamination, and have a wide-ranging number of milestones. Most of the milestones DOE must meet are contained in the compliance agreements at its six largest sites--Hanford, Savannah River, Idaho Falls, Rocky Flats, Oak Ridge, and Fernald. These six DOE sites are important because they receive about two-thirds of DOE's cleanup funding. In all, these sites account for 40 of the agreements and more than 4,200 milestones. DOE reported completing about two-thirds of the 7,186 milestones contained in its compliance agreements as of December 2001. Of the 4,558 milestones completed, about 80 percent were finished by the original due date for the milestone. The remainder of the completed milestones were finished either after the original due date had passed or on a renegotiated due date, but DOE reported that the regulators considered the milestones to be met. DOE's six largest sites reported completing a total of 2,901 of their 4,262 milestones and met the original completion date for the milestones an average of 79 percent of the time. As table 2 shows, this percentage varied from a high of 95 percent at Rocky Flats to a low of 47 percent at Savannah River. Besides the 1,334 milestones currently yet to be completed, additional milestones will be added in the future. Although DOE has completed many of the milestones on time, for several reasons DOE's success in completing milestones on time is not a good measure of progress in cleaning up the weapons complex. Specifically: Many of the milestones do not indicate what cleanup work has been accomplished. For example, many milestones require completing an administrative requirement that may not indicate what, if any, actual cleanup work was performed. At DOE's six largest sites, DOE officials reported that about 73 percent of the 2,901 schedule milestones completed were tied to administrative requirements, such as obtaining a permit or submitting a report. Some agreements do not have a fixed number of milestones, and additional milestones are added over time as the scope of work is more fully defined. For example, one of Idaho Falls' compliance agreements establishes milestones for remedial activities after a record of decisionhas been signed for a given work area. Four records of decision associated with the agreement have not yet been approved. Their approval will increase the number of enforceable milestones required under that agreement. Many of the remaining milestones are tied to DOE's most expensive and challenging cleanup work, much of which still lies ahead. Approximately two-thirds of the estimated $220 billion cost of cleaning up DOE sites will be incurred after 2006. DOE has reported that the remaining cleanup activities present enormous technical and management challenges, and considerable uncertainties exist over the final cost and time frame for completing the cleanup. Even though schedule milestones are of questionable value as a measure of cleanup progress, the milestones do help regulators track DOE's activities. Regulators at the four sites we visited said that the compliance agreements they oversee and the milestones associated with those agreements provide a way to bring DOE into compliance with existing environmental laws and regulations. They said the agreements also help to integrate the requirements under various federal laws and allow regulators to track annual progress against DOE's milestone commitments. Regulators have generally been flexible in agreeing with DOE to change milestone dates when the original milestone could not be met. DOE received approval to change milestone deadlines in over 93 percent of the 1,413 requests made to regulators. Only 3 percent of DOE's requests were denied. Regulators at the four sites we visited told us they prefer to be flexible with DOE on accomplishing an agreement's cleanup goals. For example, they generally expressed willingness to work with DOE to extend milestone deadlines when a problem arises due to technology limitations or engineering problems. Because regulators have been so willing to adjust milestones, DOE officials reported missing a total of only 48 milestones, or about 1 percent of milestones that have been completed. Even in those few instances where DOE missed milestone deadlines and regulators were unwilling to negotiate revised dates, regulators have infrequently applied penalties available under the compliance agreements. DOE reported that regulators have taken enforcement actions only 13 times since 1988 when DOE failed to meet milestone deadlines. These enforcement actions resulted in DOE paying about $1.8 million in monetary penalties, as shown in table 3. In addition to or instead of regulators assessing monetary penalties, several DOE sites agreed to other arrangements valued at about $4 million. For example, for missing a milestone to open a transuranic waste storage facility at the Rocky Flats site, the site agreed to provide a $40,000 grant to a local emergency planning committee to support a chemical-safety-in- schools program. At the Oak Ridge site, because of delays in operating a mixed waste incinerator, site officials agreed to move up the completion date for $1.4 million worth of cleanup work already scheduled. Also, at three sites--Paducah, Kentucky; Lawrence Livermore Main Site, California; and Nevada Test Site, Nevada--the regulators either did not impose penalties for missed milestones or the issue was still under discussion with DOE at the time of our review. The President's budget submitted to the Congress does not provide information on the amount of funding requested for DOE's compliance requirements. DOE sites prepare budget estimates that include compliance cost estimates and submit them for consideration by DOE headquarters. However, DOE headquarters officials evaluate individual site estimates and combine them into an overall DOE-wide budget, taking into account broader considerations and other priorities that it must address as part of the give-and-take of the budget process. As a result, the final budget sent to the Congress has summary information on DOE's programs and activities, but it provides no information on the portion of the budget needed to fund compliance requirements. DOE is not required to develop or present this information to the Congress. The President's budget typically states that the DOE funding requested is sufficient to substantially comply with compliance agreements, but it does not develop or disclose the total amount of funding needed for compliance. Officials at DOE headquarters told us that budget guidance from the Office of Management and Budget does not require DOE to develop or present information on the cost of meeting compliance requirements, and they said doing so for the thousands of milestones DOE must meet would be unnecessarily burdensome. They said their approach has been to allocate funds appropriated by the Congress and make it the sites' responsibility to use the funds in a way that meets the compliance agreement milestones established at the site level. Individual DOE sites develop information on the estimated cost of meeting compliance agreements, but the annual estimates are a flexible number. Sites develop these estimates because many of the compliance agreements require DOE to request sufficient funding each year to meet all of the requirements in the agreements. Also, DOE must respond to Executive Order 12088, which directs executive agencies to ensure that they request sufficient funds to comply with pollution control standards. Accordingly, each year DOE's sites develop budget estimates that also identify the amount needed to meet compliance requirements. The sites' process in developing these compliance estimates shows that a compliance estimate is a flexible number. For example, two budget estimates typically completed by the sites each year are the "full requirements" estimate and the "target" estimate. The full requirements estimate identifies how much money a site would need to accomplish its work in what site officials consider to be the most desirable fashion. The target estimate reflects a budget strategy based primarily on the amount of funding the site received the previous year and is considered a more realistic estimate of the funding a site can expect to receive. For each of these budget estimates, DOE sites also include an estimate of their compliance costs. As a result of this process, DOE sites usually have at least two different estimates of their compliance costs for the same budget year. Table 4 shows how the compliance cost estimates related to compliance agreements changed under different budget scenarios at four DOE sites. The multiple estimates of compliance costs developed by individual DOE sites indicate that DOE sites have alternative ways of achieving compliance in any given year. DOE site officials said that how much DOE plans to spend on compliance activities each year varies depending on the total amount of money available. Because many of the compliance milestones are due in the future, sites estimate how much compliance activity is needed each year to meet the future milestones. If sites anticipate that less money will be available, they must decide what compliance activities are critical for that year and defer work on some longer-term milestones to future years. On the other hand, if more money is available, sites have an opportunity to increase spending on compliance activities earlier than absolutely necessary. DOE's compliance agreements focus on environmental issues at specific sites and do not include information on the risks being addressed. As a result, they do not provide a means of setting priorities for risks among sites or a basis for decision-making across all DOE sites. Risk is only one of several factors considered in setting the milestones in compliance agreements. Other factors include the preferences and concerns of local stakeholders, business and technical risk, the cost associated with maintaining old facilities, and the desire to achieve demonstrable progress on cleanup. The schedules for when and in what sequence to perform the cleanup work reflect local DOE and stakeholder views on these and other factors and may not reflect the level of risk. For example, regulators at DOE's Savannah River site told us that they were primarily concerned that DOE maintain a certain level of effort and they expected DOE to schedule cleanup activities to most efficiently clean up the site. DOE developed a decision model to determine how to allocate its cleanup dollars at Savannah River to achieve this efficiency. A group of outside reviewers assessing the system at the request of site management concluded that the model was so strongly weighted to efficiency that it was unlikely that serious risks to human health or the environment could alter the sequencing of work. DOE officials said they revised the model so that serious risks receive greater emphasis. In response to concerns expressed by the Congress and others about the effectiveness of the cleanup program, DOE has made several attempts to develop a national, risk-based approach to cleanup, but has not succeeded. For example, in 1999, DOE pilot-tested the use of site risk profiles at 10 DOE offices. The profiles were intended to provide risk information about the sites, make effective use of existing data at the sites, and incorporate stakeholder input. However, reviewers found that the site profiles failed to adequately address environmental or worker risks because the risks were not consistently or adequately documented. In 2001, DOE eliminated a support group responsible for assisting the sites with this effort, and the risk profiles are generally no longer being developed or used. A 1999 DOE-funded study to evaluate its efforts to establish greater use of risk-based decision-making concluded that none of the attempts had been successful. Common problems identified by the study included poor documentation of risks and inconsistent scoring of risks between sites. The study reported that factors contributing to the failure of these efforts included a lack of consistent vision about how to use risk to establish work priorities, the lack of confidence in the results by DOE personnel, the unacceptability of the approaches to stakeholders at the sites, and DOE's overall failure to integrate any of the approaches into the decision- making process. However, the study concluded that the use of risk as a criterion for cleanup decision-making across DOE's sites not only was essential, it was also feasible and practical, given an appropriate level of commitment and effort by DOE. DOE plans to shift its cleanup program to place greater focus on rapid reduction of environmental risk, signaling yet again the need for a national risk-based approach to cleanup. Without a national, risk-based approach to cleanup in place, DOE's budget strategy had been to provide stable funding for individual sites and to allow the sites to determine what they needed most to accomplish. However, in a February 2002 report, DOE described numerous problems with the environmental management program and recommended a number of corrective actions. The report concluded that, among other things, the cleanup program was not based on a comprehensive, coherent, technically supported risk prioritization; it was not focused on accelerating risk reduction; and it was not addressing the challenges of uncontrolled cost and schedule growth. The report recommended that DOE, in consultation with its regulators, move to a national strategy for cleanup. In addition, the report noted that the compliance agreements have failed to achieve the expected risk reduction and have sometimes not focused on the highest risk. The report recommended that DOE develop specific proposals and present them to the states and EPA with accelerated risk reduction as the goal. DOE's new initiative provides additional funds for cleanup reform and is designed to serve as an incentive to sites and regulators to identify accelerated risk reduction and cleanup approaches. DOE's fiscal year 2003 budget request includes a request for $800 million for this purpose. Moreover, the Administration has agreed to support up to an additional $300 million if needed for cleanup reforms. The set-aside would come from a reduction in individual site funding levels and an increase in the overall funding level for the cleanup program. The money would be made available to sites that reach agreements with federal and state regulators on accelerated cleanup approaches. Sites that do not develop accelerated programs would not be eligible for the additional funds. As a result, sites that do not participate could receive less funding than in past years. To date, at least five major DOE sites with compliance agreements have signed letters of intent with their regulators outlining an agreement in principle to accelerate cleanup--Hanford, Idaho, Los Alamos, Oak Ridge, and Nevada Test Site. However, the letters of intent generally also include a provision that the letters do not modify the obligations DOE agreed to in the underlying compliance agreements. At Hanford, DOE and the regulators signed a letter of intent in March 2002 to accelerate cleanup at the site by 35 years or more. DOE and the regulators agreed to consider the greatest risks first as a principle in setting cleanup priorities. They also agreed to consider, as targets of opportunity for accelerated risk reduction, 42 potential areas identified in a recent study at the site. While accelerating the cleanup may hold promise, Hanford officials acknowledged that many technical, regulatory, and operational decisions need to be made to actually implement the proposals in the new approach. DOE is proceeding with the selection and approval of accelerated programs at the sites, as well as identifying the funding for those accelerated programs. At the same time, DOE is considering how best to develop a risk-based cleanup strategy. DOE's Assistant Secretary for Environmental Management said that in developing the risk-based approach, DOE should use available technical information, existing reports, DOE's own knowledge, and common sense to make risk-based decisions. Because DOE's approach to risk assessment is under development, it is unclear whether DOE will be able to overcome the barriers encountered during past efforts to formalize a risk-assessment process. In the interim, DOE headquarters review teams were evaluating the activities at each site and were qualitatively incorporating risk into those evaluations. Compliance agreements have not been a barrier to previous DOE management improvements, but it is not clear if the agreements will be used to oppose proposed changes stemming from the February 2002 initiative. DOE has implemented or tried to implement a number of management initiatives in recent years to improve its performance and address uncontrolled cost and schedule growth. For example, in 1994, it launched its contract reform initiative; in 1995, it established its privatization initiative; and in 1998, it implemented its accelerated path- to-closure initiative. These initiatives affected how DOE approached the cleanup work, the relationship DOE had with its contractors, and, in some cases, the schedule for completing the work. Based on our review of past evaluations of these initiatives and discussions with DOE officials and regulators at DOE sites, it appears that DOE proceeded with these initiatives without significant resistance or constraints as a result of the compliance agreements. Because DOE's cleanup reform initiative is in its early stages, and site- specific strategies are only beginning to emerge, it is unclear how the site compliance agreements will affect implementation of DOE's latest cleanup reforms. For example, it is not yet known how many sites will participate in DOE's initiative and how many other sites will encounter cleanup delays because of reduced funding. However, early indications suggest caution. Parties to the agreements at the sites we visited were supportive of DOE's overall efforts to improve management of the cleanup program, but expressed some concerns about proposals stemming from the February 2002 review of the program. They said that they welcome DOE's efforts to accelerate cleanup and focus attention on the more serious environmental risks because such initiatives are consistent with the regulators' overall goals of reducing risks to human health and the environment. Most regulators added, however, that DOE generally had not consulted with them in developing its reform initiative and they were concerned about being excluded from the process. Furthermore, they said DOE's initiative lacked specific details and they had numerous questions about the criteria DOE will use to select sites and the process it will follow at those sites to develop an implementation plan to accelerate cleanup and modify cleanup approaches. Most regulators said they would not view as favorable any attempt by DOE to avoid appropriate waste treatment activities or significantly delay treatment by reducing funding available to sites. In such a case, these regulators are likely to oppose DOE's initiative. They told us that they most likely would not be willing to renegotiate milestones in the compliance agreements if doing so would lead to delays in the cleanup program at their sites. In addition, these regulators said that if DOE misses the milestones after reducing the funding at individual sites, they would enforce the penalty provisions in the compliance agreements. The effect of compliance agreements on other aspects of DOE's initiative, especially its proposal to reclassify waste into different risk categories to increase disposal options, is also unclear. Some of the proposed changes in waste treatment would signal major changes in DOE assumptions about acceptable waste treatment and disposal options. For example, one change would eliminate the need to vitrify at least 75 percent of the high- level waste, which could result in disposing of more of the waste at DOE sites. In addition, DOE is considering the possibility of reclassifying much of its high-level waste as low-level mixed waste or transuranic waste based on the risk attributable to its actual composition. However, at all four sites we visited, regulators said that it is unclear how DOE's proposed initiatives will be implemented, what technologies will be considered, and whether the changes will result in reduced cost and accelerated cleanup while adequately protecting human health and the environment. DOE generally did not seek input from site regulators or other stakeholders when developing its latest initiative. DOE's review team leader said that when the review team visited individual sites, the team had not formulated its conclusions or recommendations and so did not seek regulators' views. Furthermore, the team leader said that, during the review, DOE was holding internal discussions about improving ineffective cleanup processes, such as contracting procedures. To include regulators on the review team during these discussions, according to the team leader, could have created the impression that the criticism of DOE processes came from the regulators rather than from DOE and contractor staff. According to the Associate Deputy Assistant Secretary for Planning and Budget, since the review team's proposals were made public in February, DOE has held discussions with regulators at all sites and headquarters about implementing the proposals. In summary, Mr. Chairman, DOE faces two main challenges in going forward with its initiative. The first is following through on its plan to develop and implement a risk-based method to prioritize its various cleanup activities. Given past failed attempts to implement a risk-based approach to cleanup, management leadership and resolve will be needed to overcome the barriers encountered in past attempts. The second challenge for DOE is following through on its plan to involve regulators in site implementation plans. DOE generally did not involve states and regulatory agencies in the development of its management initiative. Regulators have expressed concerns about the lack of specifics in the initiative, how implementation plans will be developed at individual sites, and about proposals that may delay or significantly alter cleanup strategies. Addressing both of these challenges will be important to better ensure that DOE's latest management initiative will achieve the desired results of accelerating risk reduction and reducing cleanup costs. Thank you, Mr. Chairman and Members of the Subcommittee. This concludes my testimony. I will be happy to respond to any questions that you may have. For future contacts regarding this testimony, please contact (Ms.) Gary Jones at (202) 512-3841. Chris Abraham, Doreen Feldman, Rich Johnson, Nancy Kintner-Meyer, Tom Perry, Ilene Pollack, Stan Stenersen, and Bill Swick made key contributions to this report.	Compliance agreements between the Department of Energy (DOE) and its regulators specify cleanup activities and milestones that DOE has agreed to achieve. The 70 compliance agreements at DOE sites vary, but can be divided into three main types. These are: (1) agreements specifically required by the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 to address cleanup of federal sites on EPA's national priorities list or by the Resource Conservation and Recovery Act of 1976 to address the management of mixed radioactive and hazardous waste at DOE facilities; (2) court-ordered agreements resulting from lawsuits initiated primarily by states; and (3) other agreements, including state administrative orders enforcing state hazardous waste management laws. DOE reported completing about 80 percent of its milestones by the time originally scheduled in the agreements. The cost of complying with these agreements is not specifically identified in the DOE budget submitted to Congress. Individual DOE sites develop annual compliance cost estimates as part of their budget requests. However, DOE headquarters officials adjust those individual site estimates to reflect national priorities and to reconcile various competing demands. Compliance agreements are site-specific and are not intended to provide a mechanism for DOE to use in prioritizing risks among the various sites. Compliance agreements have not been a barrier to previous DOE management initiatives, but it is not clear if the compliance agreements will be used to oppose DOE's latest initiative to focus on accelerating risk reduction at sites.	5,126	296
"Under TRICARE, beneficiaries may obtain health care through either the direct care system of milita(...TRUNCATED)	"DOD offered health care to about 9.6 million eligible beneficiaries through TRICARE, which provides(...TRUNCATED)	4,517	1,005
"Although VA has been authorized to collect third-party health insurance payments since 1986, it was(...TRUNCATED)	"The Department of Veterans Affairs (VA) collects health insurance payments, known as third-party co(...TRUNCATED)	3,720	564
"The federal government buys a myriad of goods and services from contractors. Federal agency acquisi(...TRUNCATED)	"This report focuses on the trends in federal procurement for small businesses during the 1990s. Som(...TRUNCATED)	5,297	246
"In June 1997, we reported on the results of our interviews with state WIC officials in 8 states tha(...TRUNCATED)	"Pursuant to a congressional request, GAO discussed its completed reviews of the Special Supplementa(...TRUNCATED)	4,807	353

End of preview. Expand in Dataset Viewer.

GovReport Summarization - 8192 tokens

ccdv/govreport-summarization with the changes of:
- data cleaned with the clean-text python package
- total tokens for each column computed and added in new columns according to the long-t5 tokenizer (done after cleaning)

train info

RangeIndex: 8200 entries, 0 to 8199
Data columns (total 4 columns):
 #   Column             Non-Null Count  Dtype 
---  ------             --------------  ----- 
 0   report             8200 non-null   string
 1   summary            8200 non-null   string
 2   input_token_len    8200 non-null   Int64 
 3   summary_token_len  8200 non-null   Int64 
dtypes: Int64(2), string(2)
memory usage: 272.4 KB

token length distribution (long-t5)

Downloads last month: 2

Edit dataset card

Size of downloaded dataset files:

315 MB

Size of the auto-converted Parquet files:

152 MB

Number of rows:

10,019

Models trained or fine-tuned on pszemraj/govreport-summarization-8192