Datasets:

urialon

/

gov_report_test

like 0

gao_GAO-19-306

gao_GAO-19-306_0

Background Private health insurance is the leading source of health coverage in the United States. Small and large employers may offer fully insured group plans (by purchasing coverage from an issuer) or self-funded group plans (by setting aside funds to pay for employee health care). Most small employers purchase fully insured plans, while most large employers self- fund at least some of their employee health benefits. While the majority of health insurance coverage is provided through the small or large group market, Americans without access to group health coverage, such as those with employers that do not offer health coverage, may choose to purchase it directly from an issuer through the individual market. (See fig. 1 for total covered life-years reported by issuers to CMS in the individual and fully insured small and large group markets.) We previously identified several factors that can affect concentration in health insurance markets. High concentration levels have often been the result of consolidation—mergers and acquisitions—among existing issuers. In addition, concentration can persist because of the difficulty for new issuers to enter the market. For example, new issuers that do not yet have large numbers of enrollees may have greater challenges negotiating discounts with health care providers. PPACA contains provisions that may affect market concentration and competition among health issuers, such as the establishment of health insurance exchanges within each state’s overall individual and small group markets. One goal of the exchanges is for issuers to have an incentive to compete with one another on price and value because consumers can visit a website to compare and select among health plans participating in the exchanges. Issuer participation in the exchanges is a key factor in assuring that consumers have a choice of health plans. While PPACA does not require issuers offering coverage in an overall market to participate in the exchanges, issuers have an incentive to do so in order to access additional consumers. For example, certain consumers earning from 100 to 400 percent of the federal poverty level are eligible to receive premium tax credits that can reduce premium costs, but only for plans purchased through an exchange. The federal government and some states have also instituted other provisions to encourage issuers to participate in the exchanges. For example, PPACA required the establishment of the Consumer Oriented and Operated Plan (CO-OP) program, which provided loans to new consumer-governed, nonprofit issuers that are required to offer health plans in the individual and small group exchanges. In addition, in Maryland, certain issuers that offered plans outside of the exchange are also required to offer plans through the exchange. PPACA also established other key market reforms that apply both within and outside of the exchanges, such as requiring that issuers offer coverage to all individuals regardless of health status and limiting the ability of issuers to deny coverage or charge higher premiums to individuals and small groups based on health risks or certain other factors. Since the enactment of PPACA in 2010, there have been additional federal policy changes that may influence an issuer’s decision about whether to participate in health insurance markets. For example, in 2014, HHS announced that a program that made payments to issuers whose losses exceeded a certain threshold—known as risk corridor payments—would be budget neutral, which resulted in reduced payments for some issuers. One issuer told us that this lower than expected funding was one of multiple factors that contributed to its decision to reduce the number of insurance markets in which it participated. Overall Individual Health Insurance Markets and Exchanges Generally Remained Concentrated in Recent Years, with Increasing Concentration in Many States’ Exchanges Overall Individual Health Insurance Markets Generally Remained Concentrated in Recent Years States’ overall individual health insurance markets were generally concentrated in 2015 and 2016, similar to what we reported for previous years. Individual market exchanges—representing 57 percent of the overall individual market nationally in 2016—were also concentrated in most states and in many cases became more concentrated in recent years. States’ overall individual health insurance markets were generally concentrated among a small number of issuers in 2015 and 2016. On average, there were 16 issuers participating in each state in 2016. However, that same year, the 3 largest issuers cumulatively held 80 percent or more of the market—an indicator of high concentration—in 37 of 51 states, generally consistent with what we previously reported for years 2011 through 2014 (see fig. 2). The remaining issuers in each state often had significantly smaller market shares—on average, 12 of the 16 issuers in each state held less than 5 percent market share. We also found that in over half of states in 2016, a single issuer held at least 50 percent of the market, consistent with prior years. Specifically, a single issuer held at least 50 percent market share in 28 states in 2016. Of these states, a single issuer held between 80 and 90 percent market share in 5 states, and more than 90 percent market share in 2 states. For example, although West Virginia had 15 issuers in 2016, a single issuer, Highmark, held 91 percent market share. This largest issuer position was held by the same company in both 2015 and 2016 in 45 states; in 35 of these states, the largest issuer had been the same since 2011. While states’ overall individual markets generally remained concentrated, they experienced fluctuations in the extent of concentration in recent years. Specifically, from 2014—the last year of data on which we previously reported—to 2016, the market share of the three largest issuers increased in 30 states (with a median increase of 4 percentage points) and decreased in 21 states (with a median decrease of 6 percentage points). (See fig. 3.) However, despite these changes, states that were highly concentrated in 2014—that is, where the market share of the three largest issuers was at least 80 percent—generally remained highly concentrated in 2016. States’ Individual Market Exchanges Were Generally Concentrated, and Many Became More Concentrated from 2015 to 2017 Our analyses found that states’ individual market exchanges—collectively representing 57 percent of enrollment in the overall individual market nationally in 2016—were generally concentrated among a small number of issuers from 2015 to 2017. Each year during this time period, for the 49 states for which we had complete data, on average, between 3 and 5 issuers participated in the individual market exchanges across the states’ rating areas. Further, each year, the three largest issuers held 80 percent or more of the exchange market, on average, across the states’ rating areas, in at least 46 states. For example, in Wisconsin in 2017, the market share of the three largest issuers ranged from 75 percent in 2 of the state’s 16 rating areas to 100 percent in 6 rating areas; on average, the three largest issuers held 92 percent market share across the 16 rating areas. While the number of states meeting this 80 percent average threshold for high concentration remained relatively constant from 2015 through 2017, market share was increasingly concentrated among a smaller number of issuers in many states, as fewer issuers participated in the exchanges by 2017. The number of states with three or fewer issuers, on average, in their rating areas—and where the issuers therefore held, on average, 100 percent or nearly 100 percent market share—increased from 16 states in 2015 to 32 states in 2017. (See fig. 4.) Further, we found that in at least 35 states each year from 2015 through 2017, the average market share of the largest individual market exchange issuer across the states’ rating areas was at least 50 percent. For example, although Kansas had up to three participating exchange issuers in each of its rating areas in 2017, the largest issuer in each rating area— generally Blue Cross and Blue Shield of Kansas—had at least 88 percent market share. We also found that many states’ individual market exchanges became more concentrated from 2015 to 2017. In 32 of the 49 states, the average market share of the largest exchange issuer across the states’ rating areas increased between 2015 and 2017, with a median increase of 13 percentage points. (See fig. 5.) For example: In Arizona, the average market share of the largest exchange issuer across the state’s rating areas increased by about 60 percentage points, from 39 percent in 2015 to 98 percent in 2017. This increase corresponded with a decrease in issuer participation in the exchange; the state’s seven rating areas had between 7 and 12 issuers in 2015, but by 2017 had only 1 or 2 issuers. In 2015, a CO-OP, Compass Cooperative Health Plan, Inc., had 29 percent of the total exchange market share statewide in Arizona and was among the largest issuers in three rating areas, but it exited the exchange after that year. Other, smaller issuers also exited the exchange after 2015 and 2016, and, in 2017, Blue Cross Blue Shield of Arizona was left as the only issuer in five of the state’s rating areas. In South Carolina, the average market share of the largest exchange issuer increased by 41 percentage points, from 59 percent in 2015 to 100 percent in 2017. As in Arizona, the increase corresponded with a decrease in issuer participation in the exchange, from 2 to 4 issuers in the state’s 46 rating areas in 2015 to only 1 issuer—BlueCross BlueShield of South Carolina—in each rating area in 2017. In addition, a CO-OP, Consumers’ Choice Health Insurance Company, had 43 percent of the total exchange market share statewide and was the largest issuer in nearly half of the state’s rating areas in 2015, but it exited the exchange after that year. In contrast, BlueCross BlueShield of South Carolina had 42 percent of the total exchange market share statewide and was the largest issuer in about half of the rating areas in 2015, and by 2017 was the only remaining issuer in the state. In the remaining 17 states, the average market share of the largest exchange issuer across the states’ rating areas decreased between 2015 and 2017, with a median decrease of 12 percentage points. For example: In Maine, the average market share of the largest exchange issuer decreased by 39 percentage points, from 81 percent in 2015 to 42 percent in 2017. Maine Community Health Options, a CO-OP, remained the largest issuer in each of the state’s four rating areas during this period. However, the other two issuers in these rating areas captured more market share. For instance, Harvard Pilgrim Health Care Group had 1 percent or less market share in each rating area in 2015, but as much as 32 percent market share in one of the state’s rating areas in 2017. In Delaware—which only had one rating area—the market share of the largest exchange issuer decreased by 37 percentage points, from 92 percent in 2015 to 55 percent in 2017. Although the state had the same two issuers, Aetna Group and Highmark Group, throughout this time period—and Highmark Group was the largest issuer each year— Aetna Group’s market share increased from 8 percent in 2015 to 45 percent in 2017. Overall Small Group Health Insurance Markets and Exchanges Generally Remained Concentrated in Recent Years The Overall Small Group Market Remained Concentrated in Recent Years Our analyses found that the overall small group health insurance market remained concentrated in recent years, similar to our prior report. Small group exchanges often had low enrollment—typically less than 1 percent of the overall small group market—and also remained concentrated in recent years. State small group health insurance markets were concentrated among a small number of issuers in 2015 and 2016. On average, there were 8 issuers participating in each state in 2016. However, in that same year the 3 largest issuers cumulatively held 80 percent or more of the market—an indicator of high concentration—in about three-quarters of states, generally consistent with what we previously reported for years 2011 through 2014 (see fig. 6). The remaining issuers in each state often had significantly smaller market shares—on average, 5 of the 8 issuers in each state held less than 5 percent market share. Further, we found that the largest issuers held 50 percent or more of the market in 30 states in 2016. For example, though Louisiana had 6 issuers in 2016, the largest issuer held 76 percent market share. Overall, a single issuer held between 80 and 90 percent market share in 10 states, and more than 90 percent market share in 1 state. This largest issuer position was held by the same company in both 2015 and 2016 in 46 states; in 40 of these states, the largest issuer had been the same since 2011. While states’ overall small group markets remained concentrated, they experienced fluctuations in concentration in recent years. From 2014 through 2016, the market share of the 3 largest issuers increased in 35 states (with a median increase of 3 percentage points), remained the same in 1 state, and decreased in 15 states (with a median decrease of 1 percentage point). (See fig. 7.) However, despite these changes, states that were highly concentrated in 2014—that is, where the market share of the three largest issuers was at least 80 percent—generally remained highly concentrated in 2016. Small Group Exchanges Were Concentrated in a Few Issuers between 2015 and 2017 Our analyses found that states’ SHOP exchanges remained concentrated from 2015 to 2017, with only slight overall changes in issuer participation or market share. Further, as a proportion of the overall small group market, SHOP exchanges in most states had little enrollment. (See sidebar.) Small Group Health Options Program (SHOP) Enrollment as a Proportion of the Overall Small Group Market As a proportion of the overall small group market, SHOP exchanges in most states had little enrollment—that is, typically less than 1 percent of the overall small group market. For example, in 2016, Alaska’s small group market had 17,257 covered life-years, while its SHOP exchange had 96 covered life-years (0.6 percent). The District of Columbia, Rhode Island, and Vermont were the only states where the SHOP exchange was more than 3 percent of the overall small group market. The District of Columbia and Vermont require all small group plans to be purchased through the state’s SHOP exchange. (See app. III.) In each year, more than 31 of the 46 states for which we had data had three or fewer issuers in the SHOP exchange; therefore between one and three issuers held 100 percent of the market share for the state. Among states with four or more issuers, the market share of the three largest issuers was typically at least 80 percent. For example, California had 6 participating issuers from 2015 through 2017 in the SHOP exchange, and the market share for the three largest issuers in that state ranged from 92 to 93 percent across the 3 years. (See fig. 8.) On average, the number of participating issuers in the SHOP exchange decreased slightly from 2015 through 2017. However, in a few states, there were larger changes in issuer participation and concentration. For example, Ohio’s SHOP exchange had 7 participating issuers in 2015, decreasing to 4 issuers in 2017. Across this time period, the market share of the three largest issuers in Ohio’s SHOP exchange increased from 59 percent to 98 percent. Conversely, New York’s SHOP exchange had 10 issuers in 2015, decreasing to 8 issuers in 2017; but the market share of the three largest issuers decreased by almost 7 percentage points within that time. Further, we found that in at least 38 of 46 states each year from 2015 through 2017, the largest issuer held at least 50 percent of the SHOP exchange market share. In 23 states, the market share of the largest issuer increased during this period, with a median increase of 11 percentage points, and in 6 additional states the largest issuer was the only issuer in the SHOP exchange and thus held 100 percent market share for all 3 years. (See fig. 9.) For example: In Kentucky, the market share of the largest issuer in the SHOP exchange increased by 56 percentage points, from 42 percent in 2015 to 98 percent in 2017. In 2015, the largest issuer was Kentucky Health Cooperative, a CO-OP that exited the SHOP exchange after 2016. The second largest issuer in 2015, Wellpoint Inc. Group, increased market share over this time period, from 33 percent in 2015 to 98 percent in 2017, becoming the largest issuer. The market share of the other remaining issuer in Kentucky’s SHOP exchange, Baptist Health Plan, Inc., decreased from 19 percent in 2015 to 1 percent in 2017. In Ohio, the market share of the largest issuer in the SHOP exchange increased by 54 percentage points, from 29 percent in 2015 to 83 percent in 2017. Across this time period, the largest issuer changed from Medical Mutual of Ohio (which had 29 percent market share in 2015 and 8 percent in 2017) to Wellpoint Inc. Group (which had 12 percent market share in 2015 and 83 percent in 2017). This increase in the largest issuer’s market share corresponded with a decrease in issuer participation. The state had 7 participating issuers in 2015, decreasing to 4 in 2017. The market share of the 3 issuers that left ranged from 9 to 16 percent. In the remaining 17 states, the market share of the largest issuer decreased between 2015 and 2017, with a median decrease of 7 percentage points. In some states, these decreases were significant. For example, in Maine, the market share of the largest issuer, Maine Community Health Options—a CO-OP—decreased by 40 percentage points, from 89 percent in 2015 to 49 percent in 2017. During this time period, while Maine Community Health Options remained the largest issuer, the other two participating issuers gained additional market share. For example, Harvard Pilgrim Health Care Group increased market share from 6 percent in 2015 to 38 percent in 2017. Overall Large Group Health Insurance Markets Remained Concentrated in Recent Years In 2015 and 2016, states’ overall large group health insurance markets remained concentrated, as in prior years. On average, there were 10 participating issuers in each state in 2016. However, in that same year the 3 largest issuers held at least 80 percent market share in 43 of 51 states, which is generally consistent with prior years. (See fig. 10.) In 2016, 3 issuers held 99 or 100 percent of the large group market in 7 states—Alabama, Alaska, Mississippi, Nebraska, North Dakota, South Carolina, and Vermont. The remaining issuers in each state often had significantly smaller market shares—on average, 6 of the 10 participating issuers in each state held less than 5 percent market share. In more than 30 states in 2015 and 2016, market share was not only concentrated among a small number of issuers, but a single issuer held at least 50 percent of the overall large group market, as in prior years. A single issuer held at least 50 percent market share in 33 states in 2016, with significantly higher levels of concentration by the largest issuer in some states. For example, in 2016, a single issuer held at least 90 percent of the market in Alabama and at least 80 percent of the market in 5 other states (Alaska, Mississippi, Montana, South Carolina, and Vermont). Further, this largest issuer position was held by the same company in 2015 and 2016 in 49 states; and, in 47 of those states, the largest issuer position has been held by the same company since 2011. The extent of concentration in the overall large group market remained relatively constant when comparing 2014—the last year of data on which we previously reported—to 2016. The market share of the 3 largest issuers increased in 24 states and decreased in 24. (See fig. 11.) The largest increase was in Wisconsin, where the market share of the 3 largest issuers increased from 38 percent in 2014 to 44 percent in 2016, and the largest decrease was in New York, where the market share of the 3 largest issuers decreased from 55 percent in 2014 to 47 percent in 2016. Agency Comments We provided a draft of this report to HHS for review and comment. The department provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate Congressional committees, the Secretary of Health and Human Services, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or dickenj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IX. Appendix I: Individual Market Health Insurance Exchange Enrollment as a Proportion of the Overall Market, 2016 This table presents covered life-years in each state’s individual market health insurance exchange as a proportion of total covered life-years in each state’s overall individual market in 2016. Appendix II: Number and Market Share of Issuers in Each State’s Individual Market Health Insurance Exchange, 2015-2017 The four tables below present information on the number of participating issuers and market share of the largest issuers in each state’s individual market exchange from 2015 through 2017. Specifically, Table 2 presents the total number of exchange issuers in each state. Table 3 presents the average number of exchange issuers across each state’s rating areas. Table 4 presents the names and market shares of the single largest exchange issuer, and market share of the largest three issuers, for each state. Table 5 presents the average market share of the largest issuer across each state’s rating areas. Appendix III: Small Group Health Insurance Exchange Enrollment as a Proportion of the Overall Market, 2016 This table presents covered life-years in each state’s Small Business Health Options Program (SHOP) exchange as a proportion of total covered life-years in each state’s overall small group market in 2016. Appendix IV: Number and Market Share of Issuers in Each State’s Small Group Health Insurance Exchange, 2015-2017 The two tables below present information on the participation of issuers in each state’s small group health insurance exchange from 2015 to 2017 and the market share of the largest and three largest issuers from 2015 to 2017. Appendix V: Number and Market Share of Largest Issuers Participating in Each State’s Overall Individual Market The two tables below present information on the participation of issuers in each state’s overall individual health insurance market from 2011 to 2016, and the market share of the largest and three largest issuers from 2014 to 2016. Appendix VI: Market Share for Consumer Operated and Oriented Plans That Participated in the Exchanges Table 11 provides market share for the 23 consumer operated and oriented plans (CO-OPs) participating in state individual market and Small Business Health Options Program exchanges for 2015 through 2017. CO- OPs are new consumer-governed, nonprofit issuers created under the Patient Protection and Affordable Care Act. Out of the 23 CO-OPs originally operating in 2014, all but four, operating in five states, had ceased operations by the end of 2017. Appendix VII: Number and Market Share of Largest Issuers Participating in Overall Small Group Health Insurance Market The two tables below present information on the participation of issuers in each state’s overall small group health insurance market from 2011 to 2016 and the market share of the largest and three largest issuers from 2014 to 2016. Appendix VIII: Number and Market Share of Largest Issuers Participating in Each State’s Overall Large Group Health Insurance Market The two tables below present information on the participation of issuers in each state’s overall large group health insurance market from 2011 to 2016 and the market share of the single largest and three largest issuers from 2014 to 2016. Appendix IX: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, William D. Hadley, Assistant Director; Katherine L. Amoroso, Analyst-in-Charge; Priyanka Sethi Bansal; Giselle Hicks; John Lalomio; and Sarah-Lynn McGrath made key contributions to this report. Also contributing were Yesook Merrill; Laurie Pachter; Vikki Porter; Oliver Richard; and Emily Wilson.

Why GAO Did This Study A highly concentrated health insurance market may indicate less competition and could affect consumers' choice of issuers and the premiums they pay. In 2014, PPACA required the establishment of health insurance exchanges—a new type of marketplace where individuals and small groups can compare and select among insurance plans sold by participating issuers—and the introduction of other reforms that could affect market concentration and competition among issuers. GAO previously reported that enrollment through these newly established exchanges was also generally concentrated. PPACA included a provision for GAO to study market concentration. This report describes changes in the concentration of enrollment among issuers in (1) overall individual, small group, and large group markets, and (2) individual and small group exchanges. GAO determined market share in the overall markets using enrollment data from 2015 and 2016 that issuers are required to report annually to the Centers for Medicare & Medicaid Services (CMS) and compared that data to 2011 through 2014 enrollment data GAO analyzed in previous reports. GAO determined market share in the exchanges from 2015 through 2017 using other sources of enrollment data from CMS and states. For all data sets, GAO used the most recent data available. What GAO Found Enrollment in private health insurance plans continued to be concentrated among a small number of issuers in 2015 and 2016. In the overall large group market (coverage offered by large employers), small group market (coverage offered by small employers), and individual market (coverage sold directly to individuals), the three largest issuers held 80 percent of the market or more in at least 37 of 51 states. This is similar to what GAO previously reported for 2011 through 2014. GAO also found that within the overall individual and small group markets in each state, the health insurance exchanges established by the Patient Protection and Affordable Care Act (PPACA) were also concentrated from 2015 to 2017. For the individual market exchanges, in each year, three or fewer issuers held 80 percent or more of the market, on average, in at least 46 of the 49 state exchanges for which GAO had data. Further, the largest issuers increased their market share in about two-thirds of exchanges. The number of issuers participating in a market and their market shares can affect concentration, and many individual exchanges generally had a decreasing number of participating issuers over time. For the small group market exchanges, in each year, three or fewer issuers held 80 percent or more of the market in at least 42 of the 46 state exchanges for which GAO had data. The small group exchanges also had slight changes in issuer participation and market share over this time period. GAO received technical comments on a draft of this report from the Department of Health and Human Services and incorporated them as appropriate.

gao_GAO-19-441

gao_GAO-19-441_0

Background Roles and Responsibilities in Federal Counterdrug Missions and Activities As previously mentioned, multiple federal departments and components have responsibilities for combating the flow of illicit drugs into the United States. Figure 1 summarizes the missions and responsibilities of the federal departments and components primarily responsible for combating the trafficking of illicit drugs. DOD and DHS Joint Task Forces Involved in Counterdrug Missions and Activities DOD Joint Interagency Task Forces In 1989, DOD created several joint task forces, which aimed to bridge the military’s counterdrug efforts with those of civilian, federal law enforcement agencies. These task forces have evolved since then and eventually developed into the present-day iterations of JIATF-South under the U.S. Southern Command and JIATF-West under the U.S. Indo-Pacific Command. JIATF-South and JIATF-West both consist of representatives from DOD, DHS, and DOJ components, among others. Coast Guard admirals currently serve as the Directors of both of the JIATFs. Previously, DOD service components have led JIATF-South; however, while DOD is responsible for detection and monitoring of drug flow, it is precluded from taking law enforcement actions in counterdrug efforts. Task force officials stated that Coast Guard leadership encourages participation from both DOD and DHS because the Coast Guard is both a military and a law enforcement agency. The deputy and vice leadership positions at the JIATFs are held by officers and civilians from DOD, DHS, and DOJ components, which allow the task forces to leverage various experiences and authorities across these components, according to task force officials. DHS Joint Task Forces In 2014, DHS established three new joint task forces — (1) JTF–East, (2) JTF–West, and (3) JTF–Investigations—as pilot programs to, among other things, address the smuggling of illicit drugs over the southern border and approaches to the United States. Additionally, according to the DHS Southern Border and Approaches Campaign Plan, the JTFs were created to strengthen the unity of effort within DHS toward common goals. The 2017 National Defense Authorization Act subsequently codified these task forces and established new JTF requirements, such as establishing outcome-based and other appropriate performance measures to evaluate the effectiveness of each JTF. In 2017, DHS also created a JTF Coordination Cell to develop JTF performance measures and enhance awareness among DHS components about the role of the JTFs, among other things. The DHS JTFs primarily consist of representatives from CBP, ICE, and the Coast Guard, and a representative from each of these components serves as the Director for each of the three JTFs. The deputy leadership positions of each JTF are held by officers from the other two components. For example, DHS JTF-West’s director is an officer from CBP, and the deputy directors are officers from the Coast Guard and ICE. According to a DHS memorandum, in establishing the JTFs, DHS wanted each JTF to be led and supported by the different DHS components in order to integrate their varied capabilities. For more information on the task forces’ leadership and compositions, see table 1. Each of the five task forces are similarly organized by functional areas and all include areas such as administration and personnel, intelligence, and operations. For example, JIATF-South, JIATF-West, and each of the DHS JTFs have Planning sections, which help guide the task forces’ overarching strategic plans and operations, with input from other sections. One task force—JIATF-West—further tailored its organizational structure to its missions and activities. Specifically, JIATF-West reorganized in January 2016, at the direction of the former U.S. Indo-Pacific Combatant Commander, to operationalize and combine its Intelligence and Operations functional areas into a Counternarcotics Operations Center. According to JIATF-West leadership, the Counternarcotics Operations Center better reflects the nature of its intelligence gathering and sharing activities with other federal law enforcement agencies and foreign countries. JIATF-West officials also stated the task force merged its section that provided support and training of foreign law enforcement agencies into its Planning and Engagement section since that section directs activities related to JIATF-West’s engagement with partner nations. Areas of Responsibility As shown in figure 2, the two DOD task forces (JIATF-South and JIATF- West) and two of the three DHS task forces (JTF-East and JTF-West) have geographical areas of responsibility. In contrast, the third DHS task force (JTF-Investigations) is focused on coordinating investigations and information sharing to support DHS and the other two DHS JTFs. As a result, it does not have a geographical area of responsibility. Each of the Five Task Forces in Our Review Has a Counterdrug Mission, and Its Activities Vary Based on Threats in Its Area of Responsibility and Available Resources Example of Joint Interagency Task Force– South (JIATF-South) Activity to Combat Illicit Drug Trafficking When JIATF-South receives information about a potential illicit drug smuggling event, it will use available air and maritime assets allocated to it to detect and monitor the suspect smuggling vessel. Once JIATF-South locates the suspect vessel and has assets in place, JIATF-South turns over control of the assets to the relevant law enforcement agencies (e.g., the Coast Guard, CBP, etc.) to interdict the smuggling vessel and any illicit drugs that may be on board. JIATF-South: Focuses its activities on detecting, monitoring, and supporting the interdiction of bulk cocaine movements being smuggled on noncommercial maritime vessels. According to JIATF- South officials, this focus is partly because the key coca-producing countries are within its area of responsibility, and partly because cocaine is a key source of profit for transnational criminal organizations. JIATF-South is also allocated assets, such as ships and surveillance aircraft, from DOD and DHS components (such as the Coast Guard and CBP Air and Marine Operations), as well as from foreign partners. JIATF-South uses these maritime and air assets, in conjunction with available intelligence, to detect and monitor the trafficking of illicit drugs, such as cocaine, being smuggled north across its area of responsibility. Once JIATF-South detects a smuggling event occurring, it passes this information and control of the assets to law enforcement authorities to interdict the smuggling event. For an example of how this occurs, see the sidebar. Example of Joint Interagency Task Force– West’s (JIATF-West) Capacity Building Efforts to Combat Illicit Drug Trafficking JIATF-West has helped countries in its area of responsibility—such as Vanuatu—build their financial investigative capacity by providing law enforcement training on topics such as bank records analysis, money laundering theory, and accounting. This training is intended to help foreign law enforcement agencies better detect transnational criminal organizations’ transactions, thus making it more difficult for such organizations to operate in their area of responsibility. JIATF-West: Focuses its missions and activities on four priorities: (1) detecting precursor chemicals that can be used to manufacture illicit drugs, such as synthetic opioids; (2) supporting allies and foreign partners in combating illicit drug trafficking in its area of responsibility; (3) monitoring drug flows moving to, from, and through Asia and other countries in the Indo-Pacific region; and (4) detecting the flow of fentanyl and other synthetic opioids, according to JIATF-West documents. According to JIATF-West officials, JIATF-West’s activities primarily consist of intelligence gathering and collaboration with law enforcement partners within foreign countries where precursor chemicals are manufactured or combined to manufacture illicit drugs. JIATF-West also engages in capacity building with law enforcement authorities in foreign countries in the Pacific region, such as the Philippines and Thailand. For an example of JIATF-West’s capacity building efforts, see the sidebar. Unlike JIATF-South, JIATF-West does not have assets, such as ships or aircraft. However, JIATF-West officials stated that even if JIATF- West had assets, it would not alter the focus of its missions and activities because of the threat transnational criminal organizations pose and the nature of the flow of illicit drugs and precursor chemicals in its expansive area of responsibility. For example, JIATF-West officials told us that precursor chemicals are typically shipped in commercial cargo containers. Notably, all precursor chemicals are legal to manufacture and sell for legitimate uses, such as the production of pharmaceutical drugs and pesticides, and it is difficult to determine when such chemicals have been diverted for illicit use. Officials stated that JIATF-West would face legal and logistical challenges if they were to directly disrupt precursor chemicals being diverted, such as if the vessel was state-owned or was in a foreign country’s territorial waters. Thus, even if JIATF-West had assets, JIATF-West officials noted that the legal and logistical challenges would not change how the task force approaches its missions and activities. DHS JTFs: Focus on coordinating with DHS components (e.g., CBP, ICE HSI, Coast Guard) to facilitate awareness about cross- component, cross-geographic homeland security issues. The JTFs have broader missions than countering the flow of illicit drugs. For example, the JTFs also have responsibilities for coordinating migrant interdiction and counter-terrorism activities. Further, given their areas of responsibility, JTF-East primarily focuses on threats along the southern maritime border of the United States and JTF-West primarily focuses on threats along the southwest land border. In contrast, JTF-Investigations focuses on supporting DHS-wide investigations and sharing information to support the other two task forces. Similar to JIATF-West, the JTFs do not have physical assets to support these activities. According to JTF officials, this is partly because the 2017 National Defense Authorization Act requires the JTFs to be cost neutral. Additionally, JTF officials stated the JTFs were not meant to serve a similar function as the DOD combatant commands and, instead, are meant to help with planning and coordinating missions and activities across joint operating areas. The Five Task Forces in Our Review Generally Coordinated Effectively to Help Minimize Duplication of Counterdrug Missions and Activities, Using Various Mechanisms Task Force Officials Task force officials reported that the task forces effectively coordinated counterdrug missions and activities to minimize duplication of efforts. The extent to which the task forces coordinate varied based on whether they have (1) shared purposes and (2) areas of responsibility with overlapping or shared geographical boundaries. In particular, those task forces that have shared purposes and those task forces that have overlapping areas of responsibility or shared boundaries tended to coordinate with one another more than with the other task forces. We also found that the task forces use a variety of mechanisms to coordinate counterdrug missions and activities, such as the use of working groups and liaison officers, that our prior work has identified as best practices. Reported Effective Coordination with the Other Task Forces Officials we met with from each of the task forces stated that they are satisfied with the level of coordination that takes place with other task forces and that the coordination efforts have been effective. Our analysis of their responses found that of the five task forces, JTF-Investigations’ coordination activities were rated as the most effective by the other four task forces. JIATF-South was rated the second highest task force in terms of both the effectiveness of its coordination activities and the number of other task forces with which it coordinated. Figure 3 provides a visual representation of the task force officials’ views on the extent to which the task forces coordinate with one another and the effectiveness of the coordination efforts. of coordination with the other task forces varies based on the extent to which the task forces have shared purposes. Of the five task forces, JTF-Investigations was the one task force that coordinated with all the other task forces, which is consistent with its purpose to enhance DHS investigations, coordinate priorities, and share information with the other joint task forces. As a part of its process in designating cases as a Homeland Criminal Organization Target (HOMECORT), JTF-Investigations conducts a Comprehensive Criminal Network Analysis that identifies links between multiple cases and criminal organizations that can cross geographical and task force boundaries (see sidebar for more information on the HOMECORT process). According to JTF-Investigations officials, this analysis helps identify cases that may be related and helps to coordinate cases across task force jurisdictions to prevent duplication of missions and activities. forces told us that they coordinated more with those task forces with which they had a shared border or joint operating area. For example, JIATF-South shares a joint operating area or a geographical boundary with both JTF-East and JIATF-West and, as a result, officials from these three task forces provided more robust examples of coordination. The Task Forces Use a Variety of Mechanisms to Coordinate Missions and Activities According to task force documentation, such as operational guidance, and our discussions with task force officials, JIATF-South, JIATF-West, and the DHS JTFs coordinated with each other on missions and activities where they have a shared interest, such as a common illicit drug threat. These coordination activities include information sharing and joint operations, as well as mechanisms, such as the use of working groups and liaison officers, which our prior work has identified as best practices for coordination. According to task force officials, this coordination is intended to enhance counterdrug efforts and avoid duplication of missions and activities. As described earlier, the task forces have different mission focuses that depend on their geographically defined areas of responsibility, which also help the task forces avoid duplication of missions and activities. However, as shown earlier in figure 2, there are some areas of land, sea, and air in which more than one task force may conduct missions and activities (e.g., between JIATF-South and JTF- East). These areas of overlap are called joint operating areas. According to our review of task force documents and discussions with task force officials, within these joint operating areas, the task forces share intelligence information, coordinate missions and activities with one another, and sometimes conduct joint operations. For example, in 2018, JTF-East led and coordinated with JIATF-South on an operation to increase intelligence and targeting capabilities to disrupt illicit drug trafficking organizations operating within their joint operating area in the Caribbean. JTF-East personnel deployed to JIATF-South’s headquarters to facilitate coordination and information sharing. As a result of this joint operation, the law enforcement agencies involved seized over 3,700 pounds of cocaine and apprehended 69 migrants, one smuggler, and the smuggling vessel, according to JTF-East documentation. Officials from the task forces we spoke with reported coordinating most frequently through meetings and working groups, and through liaison officers, as detailed in examples below. Meetings and Working Groups JIATFs: In 2018, JIATF-South and JIATF-West officials developed a collaborative process to track and target shipments with potential illicit drugs and precursor chemicals moving between their respective areas of responsibility. For example, JIATF-West analysts traveled to JIATF- South to initiate the process, and officials stated they continue to work with JIATF-South analysts remotely on an ongoing basis on such collaborative efforts. JTFs: The JTF Coordination Cell hosts quarterly “synchronization meetings” with the three DHS JTFs to discuss emerging drug and smuggling trends, ongoing coordination efforts, and investigations. Liaison Officers All five task forces utilize liaison officers to enhance coordination with the other task forces and components. For example, in 2018, JIATF-West sent an analyst to JTF-Investigations to coordinate on a HOMECORT case related to drug threats in the Indo-Pacific region. Further, the five task forces coordinate with each other and their participating components through liaison officers that reside at the task forces. For example, JIATF- South officials told us that they coordinate with JTF-East through a Coast Guard liaison at JIATF-South. Liaison officers also provide direct access to their components’ information systems, which task force officials said further aids them in sharing information and coordinating missions and activities. Other Coordination Mechanisms In addition to meetings, working groups, and liaison officers, the task forces utilize other coordination mechanisms, such as memoranda of understanding and agreement, shared databases, and conferences, as detailed below. Memoranda of understanding or agreement: The two JIATFs have nine separate formal memoranda of understanding or agreement with various DHS and DOJ components, such as the Drug Enforcement Administration and ICE, that detail how the task forces and agencies will coordinate with one another and share resources. Shared databases: Each of the five task forces, along with other federal agencies, can submit information, sometimes known as a “critical movement alert” to shared databases, to alert JIATF-South about a potential drug event in its area of responsibility. According to JIATF-South officials we spoke with and our observations, JIATF- South uses these critical movement alerts, along with other intelligence that may exist, to determine whether it will dedicate assets to target a smuggling event, in conjunction with other, relevant law enforcement agencies. Conferences: Each of the five task forces participates in periodic in- person, telephone, or video conferences to coordinate with one another and share information on key issues. For example, JIATF- South officials stated they have ongoing discussions once a quarter via video conference with JIATF-West officials and other federal agencies and task forces to coordinate on illicit drug threats. Four of the Five Task Forces’ Performance Measures Do Not Allow Them to Determine the Effectiveness of Their Counterdrug Activities JIATF-South Uses Various Output and Outcome- Based Measures to Assess the Effectiveness of Its Counterdrug Activities JIATF-South uses both output-based and outcome-based performance measures to gauge the effectiveness of its counterdrug missions and activities, and it reports the results to the DOD Deputy Assistant Secretary of Defense for Counternarcotics and Global Threats in JIATF- South’s annual Performance Summary Reports. JIATF-South consistently assesses four key performance measures, called interdiction continuum measures, using data from the Consolidated Counterdrug Database to determine the effectiveness of its missions and activities. Specifically, JIATF-South measures (1) total maritime smuggling events, (2) targeted smuggling events, (3) detected smuggling events, and (4) seized or disrupted smuggling events. According to JIATF-South officials, data on these events allow JIATF-South to develop its primary measure to determine the effectiveness of its counterdrug missions and activities: the percentage of smuggling events JIATF-South detected that it handed off to law enforcement agencies that resulted in disrupted or seized illicit drugs. These measures and the results for fiscal years 2014 through 2018 are shown in table 2. From fiscal years 2014 through 2018, the rate at which JIATF-South successfully detected and handed off smuggling events for interdiction was generally 70 percent or higher. While JIATF-South officials acknowledged they have not met the target set by DOD, they noted there are many factors that influence the effectiveness of JIATF-South’s counterdrug missions and activities in any given year that are outside of its span of control. For example, drug trafficking organizations may adapt their tactics in response to JIATF-South’s activities to make it more difficult for the task force to target and detect their movements. This could include changing their trafficking routes or altering the size or type of smuggling conveyances the drug trafficking organizations use to transport the illicit drugs. JIATF-West Has a Well- Documented Methodology for Assessing Its Activities, but Its Measures Do Not Allow It to Demonstrate the Overall Effectiveness of Its Counterdrug Activities In September 2014, JIATF-West set up an Assessments Branch to provide an annual assessment of the task force’s counterdrug efforts that was intended to inform leadership about whether the task force was undertaking the best activities to achieve its mission and implementing them effectively. According to JIATF-West officials, the nature of JIATF- West’s missions and activities make it inherently more difficult to assess and quantify the effectiveness of its efforts relative to other task forces. For example, unlike JIATF-South, which is annually allocated assets to support its missions and activities and can measure results—such as tons of cocaine seized—JIATF-West’s initiatives and activities are primarily focused on information sharing and helping partner nations improve their counterdrug capabilities, activities for which results may be more difficult to quantify. To develop its annual assessment report, JIATF-West’s Assessments Branch evaluates and assigns scores for each of the approximately 20 counterdrug initiatives and more than 100 corresponding activities it conducts each year. (For an example of a JIATF-West initiative and a corresponding activity and a description of how they were assessed, see the sidebar.) In particular, JIATF-West evaluates its initiatives to determine the progress made toward achieving objectives defined in JIATF-West’s strategic documents, such as its Theater Counternarcotics Campaign Plan. Further, JIATF-West evaluates its activities to determine whether they were executed as planned, including considerations of whether the activities were done with the intended organizations, at the specified locations and times, and whether they met stated objectives. Nevertheless, we identified ways JIATF-West measures its performance that inhibit its ability to demonstrate its overall effectiveness of countering the flow of illicit drugs. Specifically, we found that JIATF-West (1) lacks a vital few performance measures that summarize its overall effectiveness that can be consistently assessed over time and (2) that it does not have established targets for assessing the effectiveness of its numerous missions and activities. JIATF-West Lacks a Vital Few, Comprehensive Performance Measures That Summarize Its Effectiveness and Are Consistent Over Time JIATF-West has focused its performance measures on assessing its numerous initiatives and activities; however, it has not developed a vital few, comprehensive performance measures that summarize the overall effectiveness of its numerous initiatives and activities in a manner that would convey essential information on its counterdrug activities to decision makers at the DOD command level and above. Such information could help these decision makers better understand the overall effectiveness of JIATF-West’s counterdrug missions and activities in relation to broader U.S. counterdrug efforts. For example, JIATF-West could develop a performance measure that calculates the percentage of leads it provides to foreign partners that result in seizures or apprehensions. Such a measure could demonstrate JIATF-West’s overall effectiveness in supporting allies and foreign partners in combating illicit drug trafficking in its area of responsibility, in keeping with one of its operational priorities. Guidance on performance measures from the DOD Office of the Deputy Assistant Secretary of Defense for Counternarcotics and Global Threats states that agencies should develop a vital few measures, no more than two or three, which convey essential information on counterdrug activities to decision makers. JIATF-West used to assess and report such measures as recently as fiscal year 2015. For example, it reported scores on the effectiveness of its mission and broader lines of effort, such as countering illicit drug and precursor chemical flows in its area of responsibility. However, JIATF-West officials told us they stopped reporting on these performance measures in fiscal year 2015 because, from the task force’s perspective, the measures did not provide meaningful insights into the effectiveness of the task force’s operations. However, such performance measures could provide meaningful information on the effectiveness of JIATF-West’s counterdrug activities to decision makers at the DOD command level and above, even if such summary information might not affect the effectiveness of operations at the task force level. We have previously reported on the importance of developing performance measures that demonstrate broader effectiveness and have also reported it is worthwhile for agencies to develop them to better determine and understand the overall effectiveness of their missions and activities. Further, JIATF-West is also unable to assess the effectiveness of its initiatives and activities over time because it has annually changed the way it measures the effectiveness of more than 100 counterdrug initiatives and activities. Specifically, JIATF-West has adjusted how it assesses the initiatives and activities each year since 2014—for example, by changing the weight scale for scoring its initiatives and activities to give more emphasis to some over others. These changes to the methodology make it difficult to compare results and assess the effectiveness of its activities over time. DOD guidance states that agencies should have measures that are consistent over time to capture trend results. In discussing these issues with JIATF-West officials, they stated that assessing the initiatives and activities provides valuable information on the effectiveness of the task force’s efforts for internal, task force management. They acknowledged that JIATF-West’s performance measures could be improved to allow for assessments of the effectiveness of the task force’s activities over time, but they added that they plan to use the same methodology to calculate the task force’s initiatives and activities scores in the future because this information is still needed internally. Given that JIATF-West’s individual initiatives and activities change year to year, however, it will be difficult for JIATF-West to assess trends in the effectiveness of its initiatives and activities over time. By also establishing a vital few, comprehensive performance measures that can be assessed consistently over time, as appropriate, JIATF-West will be able to better convey trends in the overall effectiveness of its counterdrug missions and activities over time. JIATF-West Lacks Specific Performance Targets for Its Initiatives and Activities JIATF-West has not developed specific performance targets (i.e., established acceptable levels of performance or outcomes) for its initiatives and activities as part of its documented assessment methodology, and no such targets appear in any of the task force’s annual assessment reports. When assessing its initiatives and activities, JIATF-West officials told us it aims to achieve the best possible outcome—or the highest possible score—for each of the initiatives and activities it undertakes and assesses. DOD guidance states that targets should be set for each performance measure to establish a minimum level of performance to be accomplished within a given time frame. Additionally, establishing specific performance measure targets that set a minimum level of performance to achieve could better encourage the task force to meet the targets and identify ways to improve, as needed. The DHS JTF Performance Measures Changed from Fiscal Year 2017 to 2018 and Do Not Reflect Outcomes The DHS JTFs were fully operational in fiscal year 2016 and began assessing their performance and producing performance reports in fiscal year 2017. Since they began reporting on their performance, the measures the JTFs reported changed in fiscal year 2018 and, according to JTF officials, will change again in fiscal year 2019. Specifically, in the fiscal year 2017 performance report, the JTFs reported on activities, such as the amounts of drugs seized, arrests made, and currency seized. However, according to task force officials, the 2017 report’s performance measures did not accurately reflect the strategic-level coordination the JTFs performed. For example, the measures the JTFs reported in fiscal year 2017 focused on drug seizures and arrests made by the DHS components. While the drug seizures and arrests made by the DHS components may have been made possible because of coordination activities of the DHS JTFs, using data on drug seizures and arrests as JTF performance measures resulted in double-counting because the components reported on the same seizures and arrests for their respective counterdrug programs. To address these issues for fiscal year 2018, the JTFs and the DHS Coordination Cell developed a new set of performance measures that were intended to better reflect the JTFs’ coordination activities and contributions. For example, a new JTF performance measure developed for fiscal year 2018 included the number of leads that the JTFs provided to a partner law enforcement agency, DHS component, or foreign government partner for interdiction or investigative action. Table 3 shows the evolution of the JTF performance measures from fiscal year 2017 to fiscal year 2018. The 2017 National Defense Authorization Act requires the Secretary of DHS to establish outcome-based and other appropriate performance measures to evaluate the effectiveness of each joint task force. Although the DHS JTF Coordination Cell and the JTFs developed performance measures in fiscal year 2018 that better reflect the specific missions and activities of the three task forces, these measures are focused on outputs—such as the number of operations conducted in combating transnational criminal organizations—and not outcomes, such as the number or percentage of leads that resulted in seizures of illicit drugs. According to JTF Coordination Cell officials, the fiscal year 2018 JTF performance measures are not outcome-based because it is difficult to quantify and capture the contributions of the JTFs through their roles as coordinators and facilitators of missions and activities that are conducted by DHS components. Table 4 illustrates each of the DHS JTF performance results for fiscal year 2018 under the revised measures. In addition to the changes to the performance measures made from fiscal years 2017 to 2018, JTF Coordination Cell officials told us in October 2018 they plan to further revise their performance measures for fiscal year 2019, as they believe their measures could continue to improve to better reflect the value added by the JTFs and their coordination and information-sharing activities. JTF Coordination Cell officials further stated that they had considered linking the fiscal year 2018 performance measures to relevant strategic-level outcomes in DHS plans. However, they noted that such outcomes—including the number of drug seizures and apprehensions—are already reported by the individual DHS components and they are trying to avoid the double-counting that occurred in the fiscal year 2017 performance report. We acknowledge that the types of coordination activities that the JTFs perform are inherently more difficult to measure, but developing and implementing outcome-based performance measures that reflect the value the JTFs add would better position the JTFs to demonstrate the effectiveness of their coordination efforts. For example, a performance measure that calculates the percentage of leads provided to components, partner law enforcement agencies, or foreign government partners that result in a successful seizure or arrest could help demonstrate the JTFs’ contributions to DHS counterdrug efforts. Further, in designing its outcome-based performance measures that are reflective of their coordination and information sharing activities, establishing a consistent set of performance measures across years, as appropriate, will allow the JTFs to better assess and convey their progress over time. Conclusions In 2017, 70,237 Americans died from an overdose involving synthetic opioids, heroin, cocaine, and other drugs. The number of annual overdose deaths has nearly doubled over the past decade. Combating the trafficking and availability of illicit drugs in the United States is a government-wide priority that requires a coordinated effort by federal departments and agencies with counterdrug responsibilities. JIATF- South, JIATF-West, and the three DHS JTFs are five task forces that are focused on strengthening interagency counterdrug efforts. While these task forces have worked together to coordinate and avoid duplicative activities, improvements to the performance measures used by four of the five task forces could enable them to better determine the effectiveness of their counterdrug missions and activities. In particular, by developing a vital few, comprehensive measures that are consistent from one year to the next, and establishing specific targets against which it can measure its missions and activities, JIATF-West will be better able to determine the effectiveness of its missions and activities and assess performance trends across years. In addition, by developing outcome-based performance measures that are consistent, the JTFs would be better positioned to demonstrate the effectiveness of their counterdrug efforts over time. Recommendations for Executive Action We are making a total of three recommendations: two for JIATF-West and one for DHS. The Director of JIATF-West should establish a vital few performance measures that are consistently measured over time. (Recommendation 1) The Director of JIATF-West should establish specific targets that set a minimal level of performance. (Recommendation 2) The Secretary of Homeland Security should develop outcome-based performance measures for the DHS JTFs that are consistent. (Recommendation 3) Agency Comments In May 2019, we provided a copy of this report to DOD, DHS, DOJ, and the Office of National Drug Control Policy (ONDCP) for review and comment. In written comments, which are included in appendix II, DOD stated that it concurred with the two recommendations directed to JIATF- West and noted that JIATF-West plans to conduct an internal evaluation to establish a vital few performance measures to allow it to measure performance over time. Additionally, JIATF-West has identified several areas where it can establish specific targets that set a minimal level of performance to support DOD priorities. In its written comments, which are included in appendix III, DHS stated that it concurred with its recommendation and plans to implement new performance measures in a phased approach. DHS also provided technical comments, which we have incorporated into the report, as appropriate. Additionally, ONDCP provided technical comments, which we have incorporated into the report, as appropriate. DOJ did not have any comments on the draft report. We are sending copies of this report to the appropriate congressional committees, the Acting Secretary of Homeland Security, the Acting Secretary of Defense, and other interested parties. In addition, the report is available at no charge on the GAO website at www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-3841or AndersonN@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix IV. Appendix I: Fiscal Year 2018 Performance Measures and Targets for the Department of Homeland Security Joint Task Forces This appendix provides further details regarding the performance measures and performance targets for the Department of Homeland Security Joint Task Forces. Appendix II: Comments from the Department of Defense Appendix III: Comments from the Department of Homeland Security Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact: Acknowledgements: In addition to the contact named above, Christopher Conrad (Assistant Director), Kelsey Hawley (Analyst-in-Charge), and Julia Vieweg made key contributions to this report. Also contributing to the report were Billy Commons, Pamela Davidson, David Dornisch, Eric Hauswirth, and Susan Hsu. Related GAO Products Drug Control: DOD Should Improve Its Oversight of the National Guard Counterdrug Program. GAO-19-27. Washington, D.C.: January 17, 2019. Colombia: U.S. Counternarcotics Assistance Achieved Some Positive Results, but State Needs to Review the Overall U.S. Approach. GAO-19-106. Washington, D.C.: December 12, 2018. Illicit Opioids: Office of National Drug Control Policy and Other Agencies Need to Better Assess Strategic Efforts. GAO-18-569T. Washington, D.C.: May 17, 2018. Illicit Opioids: While Greater Attention Given to Combating Synthetic Opioids, Agencies Need to Better Assess their Efforts. GAO-18-205. Washington, D.C.: March 29, 2018. Counternarcotics: Overview of U.S. Efforts in the Western Hemisphere. GAO-18-10. Washington, D.C.: October 13, 2017. Border Security: Additional Actions Could Strengthen DHS Efforts to Address Subterranean, Aerial, and Maritime Smuggling. GAO-17-474. Washington, D.C.: May 1, 2017. Coast Guard: Resources Provided for Drug Interdiction Operations in the Transit Zone, Puerto Rico, and the U.S. Virgin Islands. GAO-14-527. Washington, D.C.: June 16, 2014. Combatting Illicit Drugs: DEA and ICE Interagency Agreement Has Helped to Ensure Better Coordination of Drug Investigations. GAO-11-763. Washington, D.C.: July 28, 2011.

Why GAO Did This Study The U.S. government has identified illicit drugs, as well as the criminal organizations that traffic them, as significant threats to the United States. In 2017, over 70,000 people died from drug overdoses, according to the Centers for Disease Control and Prevention. DOD and DHS created joint task forces to help facilitate and strengthen interagency efforts in combating the flow of illicit drugs, particularly in the maritime domain. GAO was asked to review the structure of these task forces and their ability to coordinate and conduct missions effectively. Among other objectives, this report (1) assesses the extent to which the task forces coordinate effectively to minimize duplication, and (2) examines how the task forces measure the effectiveness of their missions and activities. GAO reviewed and assessed documentation on the task forces' missions, coordination efforts, and performance assessments and compared them to best practices from prior work, departmental guidance, and federal internal control standards. GAO also met with task force officials to discuss and observe planning and coordination activities. What GAO Found Many federal agencies are involved in efforts to reduce the availability of illicit drugs by countering the flow of such drugs into the United States. Among them are the Department of Defense (DOD), which has lead responsibility for detecting and monitoring illicit drug trafficking into the country, and the Department of Homeland Security (DHS), which is responsible for securing U.S. borders to prevent illegal activity. DOD and DHS lead and operate task forces—Joint Interagency Task Force (JIATF)-South, JIATF-West, and three DHS Joint Task Forces (JTF)—to coordinate and conduct counterdrug missions and activities. Task force officials reported that the task forces coordinated effectively with each other when they had shared purposes and overlapping or shared geographical boundaries (see map). The task forces also used coordination mechanisms that align with best practices, such as working groups and liaison officers, to minimize duplication of their missions and activities. Note: DHS also has JTF-Investigations, which is a functional task force with no geographic area of responsibility. Each of the five task forces GAO reviewed has performance measures, but only JIATF-South uses output (e.g., number of detected smuggling events) and outcome-based measures to assess the effectiveness of its activities. Specifically, JIATF-South developed an outcome-based measure of its overall effectiveness: the percentage of smuggling events it detected and provided to law enforcement that resulted in disrupted or seized illicit drugs. JIATF-West evaluates its numerous initiatives and activities, for instance, by determining if they were executed as planned, but has not established a vital few performance measures that consistently convey the overall effectiveness of its activities. Lastly, the DHS JTFs' performance measures are not outcome-based and do not fully assess the effectiveness of the task forces' activities. Enhancing their measures would better position JIATF-West and the JTFs to demonstrate contributions and convey trends in the overall effectiveness of their activities. What GAO Recommends GAO is making three recommendations, including that JIATF-West establish a vital few, consistent performance measures for its overall performance; and that DHS develop outcome-based performance measures for the JTFs' activities. DOD and DHS concurred with the three recommendations.

gao_GAO-20-267

gao_GAO-20-267_0

Background The Administration of State and Federal Elections Involves Various Roles, Responsibilities, and Processes In the United States, authority to regulate elections is shared by federal, state, and local officials. Congressional authority to regulate elections derives from various constitutional sources, depending on the type of election. In addition, Congress has passed legislation in major functional areas of the voting process, such as voter registration and prohibitions against discriminatory voting practices. However, responsibility for the administration of state and federal elections resides at the state level. States regulate various aspects of elections including, for example, registration procedures, absentee and early voting requirements, and Election Day procedures. Within each state, responsibility for managing, planning, and conducting elections is largely a local process, residing with about 10,300 local election jurisdictions nationwide. Some states have mandated statewide election administration guidelines and procedures that foster uniformity in the way their local jurisdictions conduct elections, whereas other states have guidelines that generally permit local election jurisdictions considerable autonomy and discretion in the way they run elections. The result is that elections can be administered differently across states and local jurisdictions. Unless states require otherwise, local jurisdictions generally have discretion over activities such as election officials’ training and, in most states, the selection and purchase of voting technology. Among other things, local election officials register eligible voters; educate voters on how to use voting technology; provide information on the candidates and ballot measures; recruit, train, organize, and mobilize poll workers; prepare and test voting equipment for use; and count ballots. The election process is composed of pre-election, Election Day, and post-election activities: Pre-election activities include providing opportunities for eligible individuals to register to vote, maintaining and updating the voter registration database, recruiting and training poll workers, selecting polling locations, preparing voting materials, testing equipment, qualifying candidates for office, and administering absentee and vote-by-mail voting processes. Election Day activities include opening and closing polling places, setting up voting machines and voting booths, checking in voters and verifying registration status, and providing opportunities for voters to mark and cast ballots. Post-election activities include securing equipment and ballots, transferring physical ballots or records of vote counts to a central location for counting, determining the outcome of the election, publishing unofficial results, certifying official election results, and performing recounts, if required. Election Infrastructure Relies on Various Components and Assets and Is Susceptible to Threats The election process relies on various assets—such as information technology systems, networks, equipment, and facilities. These assets can be broadly categorized as physical, cyber, and human components of the Election Infrastructure Subsector, as described in table 1. Physical, cyber, and human assets comprising the election infrastructure are susceptible to unintentional and intentional threats. As we have previously reported, unintentional, or nonadversarial, threat sources include equipment failures, software coding errors, or the accidental actions of employees (human errors). Threat sources also include natural disasters and other events that can cause failure within sectors on which the election infrastructure is dependent, such as power grid failures in the energy sector. Intentional, or adversarial, threats can involve targeted and untargeted attacks from a variety of sources, including criminal groups, hackers, disgruntled employees, foreign nations engaged in espionage and information warfare, and terrorists. These adversaries vary in terms of the capabilities of the actors, their willingness to act, and their motives, which can include seeking monetary gain or pursuing an economic, political, or military advantage. Appendix I lists general cybersecurity threat sources that can impact information technology systems that support the election infrastructure. Cyber adversaries may make use of various techniques, tactics, and practices—or exploits—to adversely affect an organization’s computers, software, or networks, or to intercept or steal valuable or sensitive information. These exploits are carried out through various conduits, including websites, email, wireless and cellular communications, internet protocols, portable media, and social media. Further, adversaries can leverage common computer software programs, such as Adobe Acrobat and Microsoft Office, to deliver a threat by embedding malware or other exploits within software files that can be activated when a user opens a file within its corresponding program. DHS and others have identified general cyber and physical threats that are applicable throughout the election process. For example, voting equipment may be susceptible to a supply chain attack in which the malicious actor may use the voting equipment vendor as a pathway to plant malware to modify or compromise ballot definition files before they reach the hands of election officials. Also, the absence of or lack of consistent physical access controls, auditable chain of custody procedures, or vendor installed countermeasures may allow malicious actors or well-placed insiders to manipulate voting equipment and ballots at any stage of the process through unauthorized physical access. In addition, there are certain physical and cyber threats that are applicable to individual assets and stages in the election process. Figure 1 provides examples of threats to various assets and stages in the election process. Additionally, DHS has analyzed and identified common cybersecurity vulnerabilities associated with enterprise networks and voter registration systems supporting election infrastructure that could apply to multiple assets at all stages of the election process. Such vulnerabilities include user susceptibility to malicious email, outdated software patches, the use of default system configurations, passwords that are weak or presented in clear text, and the use of operating systems with known weaknesses that have not been properly addressed. DHS, through CISA, Is the Lead Federal Agency for the Election Infrastructure Subsector Presidential Policy Directive 21, issued in February 2013, shifted the nation’s focus from protecting critical infrastructure against terrorism to protecting and securing critical infrastructure and increasing its resilience against all hazards, including natural disasters, terrorism, and cyber incidents. The directive identified 16 critical infrastructure sectors and outlined roles and responsibilities for protecting these sectors. Further, the directive established sector specific agencies as the federal entities responsible for providing institutional knowledge and specialized expertise to facilitate or support federal, state, and local governments, as well as private sector entities, in protecting critical infrastructure. The National Infrastructure Protection Plan, updated by DHS in December 2013, further integrates critical infrastructure protection efforts between government and private sectors, among other things. It describes a voluntary partnership model as the primary means of coordinating government and private sector efforts to protect critical infrastructure. As part of the partnership structure, the designated sector-specific agencies serve as the lead coordinators for the security programs of their respective sectors. In accordance with the National Infrastructure Protection Plan, the National Protection and Programs Directorate within DHS was designated the sector-specific agency, or lead federal agency, for the Election Infrastructure Subsector. CISA subsequently assumed the role of sector-specific agency upon its establishment as the successor to the Directorate in November 2018. As the lead agency for the Election Infrastructure Subsector, CISA is responsible for coordinating partnership activities and information sharing and is the primary federal interface with the subsector’s stakeholders with respect to security. The Election Security Initiative, part of CISA’s National Risk Management Center, is responsible for managing the agency’s election subsector partnerships. To implement the voluntary partnership model, the subsector created two complementary coordinating councils—one for governments and one for private sector partners—to facilitate partnerships to support election infrastructure. Specifically, the Election Infrastructure Subsector Government Coordinating Council, created in October 2017, enables federal, state, and local governments to share information and collaborate on best practices to mitigate and counter threats to election infrastructure. The council is composed of 27 members, which include three voting members from the federal government—specifically one from DHS and two from the Election Assistance Commission (EAC)—and 24 from state and local governments. The Federal Bureau of Investigation (FBI), EAC, and National Institute of Standards and Technology (NIST) coordinate with each other, with CISA, and with state and local governments through the Election Infrastructure Subsector Government Coordinating Council. Additionally, the Subsector Coordinating Council was chartered in February 2018 and includes private sector entities whose services, systems, products, or technology are used by or on behalf of state or local governments in administrating the U.S. election process. The Election Infrastructure Subsector Specific Plan outlines actions that CISA, as the sector-specific agency, the Government Coordinating Council, and the Subsector Coordinating Council will take to support election infrastructure. Other Federal Agencies Also Have Key Roles in the Election Infrastructure Subsector Within the Department of Justice, the FBI supports the Election Infrastructure Subsector by countering foreign influence operations and collecting and processing threat information on election infrastructure. This effort is headed by the FBI’s Foreign Influence Task Force, which integrates the agency’s cyber, counterintelligence, counterterrorism, and criminal law enforcement resources to better understand threats posed by foreign influence operations. Among other things, the task force investigates cyber operations targeting election infrastructure or public officials, and covert influence operations designed to influence public opinion and sow division through disinformation and misinformation on social media. The FBI exchanges threat information with CISA and other federal partners to help states and local jurisdictions detect and prevent operations targeting the election infrastructure. Further, the EAC supports the Election Infrastructure Subsector by carrying out its responsibilities under the Help America Vote Act. Specifically, the EAC develops voluntary voting system guidelines and oversees the testing and certification of voting systems. Under the Help America Vote Act, the EAC works through the Technical Guidelines Development Committee to establish a set of principles, guidelines, and requirements specifying how voting systems are to meet standards of functionality, accessibility, and security. The EAC has also provided states with operational grants to replace voting systems. According to the Acting Executive Director of EAC, states also used the grants to increase the security of election systems, such as voter registration systems, and apply other cybersecurity enhancements. Additionally, the EAC and CISA have collaborated to develop select initiatives—such as web-based training for election officials—to expand outreach to states and local jurisdictions. NIST supports the Election Infrastructure Subsector by conducting research to develop and provide standards, tests, guidelines, best practices, and lab accreditation assistance that EAC and states and local jurisdictions may use at their discretion. The Director of NIST chairs the Technical Guidelines Development Committee. At the request of the Committee, the Director of NIST provides technical support for the Committee to carry out its duties, such as by participating in election and constituency working groups to provide technical leadership in support of the development of voluntary voting system guidelines. NIST also helps election officials identify and prioritize opportunities to improve their cybersecurity posture. For example, it established a joint working group with the Election Infrastructure Subsector Government Coordinating Council and Subsector Coordinating Council to develop a framework of cybersecurity practices tailored to elections. In doing so, NIST works with the election community to identify the resources and outcomes needed to ensure the security of the election infrastructure. As part of this effort, it receives feedback from states and local jurisdictions, as well as from CISA, through the Election Infrastructure Subsector Government Coordinating Council. DHS Provides Services to States and Local Election Jurisdictions, and Selected Election Officials Reported Being Satisfied with DHS’s Assistance DHS, through CISA, has taken steps to assist election officials in securing election infrastructure by providing services in three areas: regional support and assistance, education and awareness, and information sharing and analysis among federal, state, and local organizations. Appendix II provides a list of the services that CISA makes available to states and local election jurisdictions. Regional support and assistance. CISA employs personnel with cyber and physical security expertise in its 10 regional offices throughout the country. According to CISA, as of November 2019, these experts included 24 cybersecurity advisors and 100 protective security advisors who perform and coordinate security assessments for the 16 critical infrastructure sectors, including the Election Infrastructure Subsector. A single advisor may be responsible for performing and coordinating assessments for an entire state or region and across multiple critical infrastructure sectors. The cybersecurity advisors and protective security advisors consult with state and local election officials and identify services that CISA can provide on a voluntary, no cost basis. For example, according to CISA Election Security Initiative officials, cybersecurity advisors and protective security advisors have promoted CISA services and assessments, such as an assessment of network security vulnerabilities and an assessment of risks associated with information and communication technology suppliers and service providers. In addition, protective security advisors have conducted physical inspections of the protections over facilities that store election-related equipment such as voting machines or poll books. Protective security advisors told us that they also provide a web-based tool that states or local jurisdictions can use to identify security gaps and preparedness across facilities. CISA officials stated that, although regional personnel promote cybersecurity and physical security services to election officials, personnel based at CISA headquarters conduct the more advanced cybersecurity assessments. For example, the Vulnerability Management Branch provides vulnerability scanning and risk and vulnerability assessments, while the Threat Hunting Branch responds to cyber incidents. In September 2019, officials from the Election Security Initiative told us that, based on the CISA Director’s guidance, the agency gives requests from election infrastructure stakeholders a higher level of priority than requests from the other sectors. The precise length of the wait for service depends on the type of service. For some services, such as vulnerability scanning, there is no wait time, according to CISA officials, because CISA can activate the service within 24 hours. Education and awareness. CISA disseminates educational materials to raise awareness of election security-related issues and services available to state and local election officials. For example, CISA provides a web- based training course to help election officials understand the principles of information technology management and has developed guidance to help states and localities adopt recommended information technology practices to improve their security posture. According to CISA, as of November 2019, 1,201 individuals had completed the online course. Further, CISA conducted two election infrastructure tabletop exercises known as “Tabletop the Vote” in August 2018 and June 2019 to help the Election Infrastructure Subsector community collaborate and identify best practices and areas for improvement in election-related cyber incident planning, identification, response, and recovery. The 2018 tabletop exercise included 44 states, the District of Columbia, 16 federal entities, the National Association of Secretaries of State, and the National Association of State Election Directors. According to CISA officials, the June 2019 exercise included 47 states, the District of Columbia, 15 federal entities, the National Association of Secretaries of State, the National Association of State Election Directors, the National Governors Association, and the National Conference of State Legislatures. CISA officials also noted that CISA personnel, including regional personnel, have presented at numerous national and state meetings of election officials, such as the Election Center’s annual conference in August 2019. In addition, as part of CISA’s Last Mile initiative, the agency collaborates with state and local election officials to create customized posters that highlight efforts to strengthen election security. The purpose of the posters is to describe the state’s or local jurisdiction’s election infrastructure assets and systems, characterize risks, and offer specific measures it should implement to mitigate those risks. Election officials can present the posters to voters, lawmakers, and their own personnel to bolster confidence in the security of their election systems. As of November 2019, CISA reported that it had delivered Last Mile posters to 19 states (including six states since the 2018 election) and 1,202 local election jurisdictions. Information sharing and analysis. CISA collects and analyzes election security-related information—such as threat indicators, incident alerts, and vulnerability data—and shares this information with election officials to help them assess cybersecurity controls, detect threats, and mitigate risks. To further this goal, CISA partnered with the Center for Internet Security and the Election Infrastructure Subsector Government Coordinating Council to create the Election Infrastructure Information Sharing and Analysis Center (EI-ISAC) in February 2018. State and local election offices can join the EI-ISAC at no cost and receive election-focused cyber defense tools and products. According to the Director of the EI-ISAC, as of November 2019, its members included 50 states, the District of Columbia, and 2,267 local jurisdictions. CISA officials stated that the EI-ISAC is the primary mechanism that CISA uses to exchange information throughout the election community. For example, the EI-ISAC produces a quarterly threat report to assist the election community in the analysis of active information security threats. From its inception through September 2019, the EI-ISAC had sent out 263 alerts to its members, including weekly, spotlight, and other emails, according to EI-ISAC officials. EI-ISAC officials added that CISA funds the EI-ISAC to, among other things, deploy an intrusion detection sensor in each state specifically for voter registration systems and other supporting infrastructure to detect malicious activity and provide network security alerts. CISA officials stated that the agency, in coordination with the EI-ISAC, analyzes data from these sensors to identify trends in threats and vulnerabilities across states and local jurisdictions. CISA also manages the National Cybersecurity and Communications Integration Center (NCCIC), which receives reports of suspected malicious cyber activity from state and local officials, analyzes attempts to infiltrate election systems, and shares information about threats and vulnerabilities through the EI-ISAC. The NCCIC has also assisted election officials in responding to incidents, upon request. According to CISA officials, in fiscal years 2018 and 2019, NCCIC’s Hunt and Incident Response Teams provided services to 10 states and 16 local election jurisdictions, such as incident response activities and proactive reviews for malicious activity at the time of service. Table 2 identifies selected services that CISA provided to states and local jurisdictions in 2018 and 2019, as of November 6, 2019. Selected Election Officials Are Generally Satisfied with DHS’s Election Security Assistance and Identified Various Benefits and Challenges State election officials with whom we spoke were generally satisfied with CISA’s support to secure their election infrastructure. Specifically, officials from seven of the eight states we contacted said that they were very satisfied with CISA’s election-related work, while officials from the eighth state said that they were somewhat satisfied. Officials from five states told us that their relationship with CISA had improved markedly since early 2017, when the elections subsector was established. For example, state officials said that CISA has made progress in this area. The Secretary of Homeland Security’s designation of elections as critical infrastructure was initially controversial among state and local officials. For example, in February 2017, the National Association of Secretaries of State voted to oppose the designation of elections as critical infrastructure, citing the states’ constitutional authority to regulate elections. In addition, CISA officials told us that a lack of trust and communication between DHS and state and local election officials hindered initial efforts to establish the Election Infrastructure Subsector. However, officials from one state told us that, despite initial reservations about DHS’s role in election security, CISA has become a good partner over time. An official from another state expressed appreciation that CISA appears to be honestly and earnestly working to gain states’ trust. Officials representing the National Association of Secretaries of State and the National Association of State Election Directors also stated that CISA had worked to improve its relationships with state election officials. According to these officials, CISA has expanded outreach efforts by attending state association meetings and conferences to present information on CISA’s resources and the threat environment and has impressed election officials with the level of detail provided by CISA’s threat reporting. An official from the Election Center, which represents local election officials, stated that CISA officials attend every cybersecurity and critical infrastructure event hosted by the center. CISA officials stated that, while it is not possible to meet individually with all of the local election jurisdictions nationwide, it can engage with multiple local election jurisdictions at one time at these association conferences. Selected State and Local Election Officials Reported Benefits from CISA’s Efforts Election officials from selected states and local jurisdictions cited various benefits from CISA’s support to election security. According to officials from six states, CISA’s involvement in election security has increased the officials’ understanding of the threat environment that the election community faces. They also said that CISA’s involvement has helped them to plan for cybersecurity threats and to prioritize their election security efforts. For example, officials from one state said that CISA recommended that the state set priorities and focus on risk assessments and network segmentation. In addition, election officials from five states spoke highly of CISA’s expertise and availability. The officials said, for example, that CISA regional and headquarters personnel were easy to get in touch with and knowledgeable about the election community. As a result, the officials said that they had better access to training opportunities and informal advice. Further, officials from each of the eight states spoke positively about the information that the officials received from the EI-ISAC. For example, state officials said that the EI-ISAC updated them regularly on election security incidents and vulnerabilities nationwide, allowing them to prepare for potential incidents. Officials also stated that the EI-ISAC presented the information in a way that was understandable to election administrators who may not have backgrounds in information technology. For example, in a monthly “spotlight” email, the EI-ISAC defines a key cybersecurity term and explains to the election officials why it should matter to them. One official told us that through membership in the EI-ISAC, the state has learned about election security best practices from other states, and other officials said that EI-ISAC allows them to maintain visibility of nationwide threats and other election security issues. Officials from one state said that their contacts through the EI-ISAC helped them to identify a point of contact at social media companies so that they could inform the companies about election-related misinformation being spread online. In addition, election officials from two states said that they have encouraged or required local election jurisdictions to enroll in the EI-ISAC. According to EI-ISAC officials, the number of local election jurisdictions enrolled in the EI-ISAC increased from 1,384 at the end of 2018 to 2,267 in November 2019, and included all three election jurisdictions that we contacted. Officials from two of the three local jurisdictions said that the EI-ISAC emails were valuable. For example, election officials from one local jurisdiction said that communication from the EI-ISAC is meaningful and targeted to bolster their election security efforts. Election officials from the other jurisdiction said that they use the EI-ISAC information to improve their continuity of operations plans. On the other hand, officials from the third local jurisdiction said that the information provided in EI- ISAC emails was too general and not specific enough to their circumstances. Election officials from five states also spoke positively of the EI-ISAC situational awareness chat rooms, which DHS hosted on its Homeland Security Information Network. These officials stated that they participated in and monitored the chat rooms on Election Day to maintain awareness of any emergent election security issues nationwide. For example, officials from one state said that the chat rooms helped them receive real time notification of issues in other states and possible solutions to those issues. In addition, as previously mentioned, one of CISA’s major efforts was the 3-day tabletop exercises held in August 2018 and June 2019, which state and local officials were able to attend remotely by video teleconference from sites around the country. Elections officials from five states said that the exercises conveyed important information and prompted thoughtful discussions among state and local officials. Election officials from four states said that the exercises helped them build relationships within their states, and election officials from three states also said the exercises helped to build relationships with federal agencies as well. In addition, officials from four states said that they conducted or were planning to conduct tabletop exercises modeled on CISA’s exercises. Election officials from three states said that CISA’s cybersecurity assistance has helped them to assure voters that elections in their states are secure or to promote election security efforts. For example, officials from one state said that, when they get questions from the public about election security, they tell voters that CISA’s assessments have shown that the state’s election systems are free of malicious code. Election officials from another state said that CISA officials’ outspokenness has created opportunities for state officials to discuss the importance of election security issues with local officials. Additionally, officials from five states told us they encourage local election officials to request election security services from CISA to increase the security posture of the local jurisdictions. Officials Cited Challenges Linked to DHS’s Election Security Efforts Even though state and local election officials provided mostly positive feedback on DHS’s election security assistance, officials also identified two challenges linked to DHS’s assistance efforts. First, officials from three states stated that it is challenging to find time to schedule election security services. For example, officials from one state said that their biggest challenge is to find time in their state’s election schedule for receiving CISA services because the state has seven to nine elections in off years (that is, years without congressional or presidential elections). Officials from another state said that they might have requested additional election services from CISA if the state had more time in its election calendar. However, none of the state officials with whom we spoke attributed this difficulty to CISA, as election calendars are outside of CISA’s control. In commenting on this challenge, CISA officials said that they have tried to accommodate states’ and local election jurisdictions’ needs, when possible. For example, CISA started offering remote penetration testing as an alternative to the risk and vulnerability assessment. The officials said that the two services are similar, but the remote penetration testing can be completed in fewer days and does not require CISA personnel to be physically present in the election offices. CISA officials told us that smaller jurisdictions sometimes prefer this option. Election officials also identified an additional challenge related to the intelligence briefings that were provided by DHS’s Office of Intelligence and Analysis for state and local officials with security clearances leading up to the 2018 elections. According to Office of Intelligence and Analysis officials, the briefings allowed state and local officials to become more informed about the national threat picture, which in turn, allowed them to adjust to the threat more effectively. Election officials from two states said that the intelligence briefings had provided helpful contextual information about cyber threats. However, election officials in two other states said that the briefings were not as useful as the election officials had hoped because the briefings only provided information that was already available publicly, and election officials from another state said that they learned about a significant election security issue possibly related to their state through news reports. For example, an election official from a different state said that the state learned about threats from the Department of Justice’s July 2018 indictment against foreign intelligence officers. CISA officials stated that they are aware of this issue and have been trying to improve the communication of intelligence information to state and local election officials. For example, at a October 2019 hearing of the House of Representatives Committee on Homeland Security, a CISA senior cybersecurity advisor testified that DHS has begun working with the Intelligence Community to rapidly declassify relevant intelligence or provide as much intelligence as possible, at the lowest classification level possible, to state and local election officials. CISA officials also told us that the agency has started working with cybersecurity intelligence firms to provide election security information to state and local officials without the need for national security clearances or travel to secure facilities. According to CISA officials, two cybersecurity intelligence firms provided webinars to election officials in September and October 2019. CISA officials said that these firms have sophisticated capabilities that they use to analyze information that is not classified. As a result, the cybersecurity intelligence firms can more easily share information with states and local election jurisdictions. CISA officials said that state and local officials will benefit from these briefings because they provide actionable threat information to election officials without requiring them to have security clearances or travel to secure facilities. CISA Has Not Finalized Its Plans to Address Key Objectives and Challenges According to DHS planning guidance, strategic-level planning provides a framework for guiding homeland security activities and generates the objectives and priorities, which influence the roles, responsibilities, and actions that are detailed in the operational-level plans. Further, subsequent operational-level plans are to identify the tasks and resources needed to execute strategic plans. Prior GAO work has shown that strategic and operations plans can help further define capabilities, including opportunities to leverage resources. Such plans can also provide a roadmap for addressing identified gaps and better position an agency and its components to work collaboratively and strategically with external partners, such as states and local jurisdictions. CISA has begun developing strategic and operations plans for assisting states and local jurisdictions in securing election infrastructure in preparation for the 2020 elections. Specifically, CISA has developed a draft strategic plan for securing election infrastructure, known as the #Protect2020 Strategic Plan. According to the draft, CISA intends for its strategic plan to be used to achieve the high-level goals and outcomes called for in the agency’s August 2019 Strategic Intent. The draft strategic plan focuses on four areas, also referred to as lines of effort: (1) protecting election infrastructure, (2) supporting political campaigns, (3) raising public awareness on foreign influence threats and building resilience, and (4) sharing intelligence and identifying threats. In addition, the draft strategic plan identifies several objectives for each line of effort. For example, it includes three objectives for the protecting election infrastructure line of effort: building stakeholder capacity to manage risks and handle adversaries, through activities such as creating incident response and communication plans and encouraging states to adopt and practice them; providing technology services to stakeholders to monitor and secure their networks, by promoting the use of CISA’s voluntary services and assessments, among other things; and facilitating information sharing between the federal government, private sector, and state and local partners by, among other things, hosting situational awareness chat rooms prior to, during, and after state and federal elections. As another example, the draft strategic plan identifies three objectives for the sharing intelligence and identifying threats line of effort: partnering with private sector firms and vendors to improve cyber threat intelligence, through activities such as developing threat indicators and warnings; cooperating across federal partners—including federal law enforcement and the Intelligence Community—by, among other things, advocating for the creation of a joint memorandum to consolidate and highlight current knowledge on election threat intelligence; and monitoring threat activity through actions such as using network monitoring capabilities to spot malicious activity and reveal key trends. In addition, CISA officials stated that the agency has begun developing a draft operations plan, known as the 2020 Election Security Operations Plan. This plan is to—in conjunction with the strategic plan—describe key organizational functions, processes, and resources employed to carry out the agency’s efforts in support of elections in 2020. CISA officials stated, as of November 2019, that the agency intended to finalize the strategic and operations plans by January 2020. However, as of January 2020, CISA’s plans were not yet complete. According to a CISA official, the plans were not finalized due to an ongoing reorganization within CISA and limited staffing resources within the Election Security Initiative. While CISA has drafted the strategic plan, the agency has not yet completed a draft of its operations plan. CISA officials have noted the importance of the operations plan to help ensure the agency is adequately prepared to support election officials in securing election infrastructure in advance of elections, which begin with presidential primaries in February 2020, as well as subsequent primaries leading up to the November 2020 general election. Further, CISA’s operations plan may not fully address the four lines of effort outlined in its strategic plan when finalized. Specifically, according to CISA officials, the operations plan is expected to identify organizational functions, processes, and resources for certain elements of two of the strategic plan’s lines of effort—protecting election infrastructure and sharing intelligence and identifying threats. However, agency officials did not identify the extent to which the operations plan would address all of the objectives from these lines of effort in the strategic plan. CISA officials also stated that the agency is unlikely to develop additional operations plans for the other two lines of effort—providing security assistance to political campaigns, and raising public awareness on foreign influence threats and building resilience. The officials stated that, given the limited amount of time remaining before election preparation activities commence, the agency decided to prioritize developing a plan for the first line of effort that addresses the primary customers of the agency’s election services. In the absence of completed strategic and operations plans, a CISA official in one region stated in October 2019 that the region is moving forward with its own strategy for assisting states and local jurisdictions because the 2020 election cycle is scheduled to start with state primary elections in the region in March 2020. The lack of finalized plans can affect CISA’s achievement of higher-level objectives that take time to accomplish, such as building stakeholder capacity and public awareness. Until CISA finalizes its strategic and operations plans for supporting elections in 2020 and ensures that the operations plan fully addresses all of the aspects of its strategic plan, CISA will not be well-positioned to execute a nationwide strategy for securing election infrastructure prior to the start of 2020 election activities. CISA Identified Challenges Related to Its Efforts to Secure Election Infrastructure, but Has Not Documented How It Intends to Address Them DHS’s National Infrastructure Protection Plan, which provides strategic direction for national, critical infrastructure protection efforts, calls for sector-specific agencies to coordinate lessons learned and corrective actions and rapidly incorporate them to improve future efforts. Further, GAO’s Standards for Internal Control calls for management to document corrective action plans to remediate internal control deficiencies in a timely manner following the reporting and evaluation of issues. CISA has identified various challenges related to its election assistance efforts; however, the agency has not yet documented plans that address them. Following the 2018 midterm elections, CISA and the RAND Corporation conducted two reviews of CISA’s efforts supporting the elections, in order to inform strategic planning and strengthen future operations. The first review, conducted by the RAND Corporation under a contract with DHS, assessed election security operations that CISA undertook from January 2017 through the November 2018 midterm elections. The review relied upon input from DHS personnel, the EI- ISAC, associations representing state election officials, and election system vendors to identify lessons learned from CISA’s activities to assist in securing election infrastructure. In the second review, CISA conducted an after action review covering its efforts to assist in securing election infrastructure from September 2018 to December 2018, based on input from personnel within DHS and its federal partners who participated in the agency’s election security operations. Both reviews identified various challenges that CISA needed to address in its planning for 2020. For example, the RAND review cited challenges related to the services and threat briefings CISA provided to states and local jurisdictions. The review noted, among other things, that CISA: lacked an approach for prioritizing its activities based on election security risks, which could limit the agency’s ability to dedicate increased attention and resources to the jurisdictions with the highest risk; did not adequately tailor services, which could have made it more difficult to meet the resource and time constraints of customers such as local election jurisdictions; and did not always provide actionable recommendations in DHS classified threat briefings or make unclassified versions of the briefings available, which may have hindered election officials’ ability to effectively communicate with information technology and other personnel in their agencies who did not have clearances. Additionally, the CISA after action report identified a number of internal operational challenges associated with its election-related efforts in 2018. For example, the report cited: a lack of understanding by CISA headquarters staff of the roles and functions of regional field staff, which led to redundant requests for information from headquarters staff to regional staff; the lack of a single agency-wide platform to maintain an awareness of election threats, which resulted in confusion among CISA personnel about which threat information was accurate and current; and the inability of CISA personnel supporting election security operations to access social media websites from situational awareness rooms, which hindered their collection and analysis of threat information. Further, both reviews cited challenges regarding CISA’s ability to manage incident information and provide Election Day incident response capabilities in the event of a compromise. For example, with regards to the 2018 election, the reviews noted: few capabilities that CISA field staff could quickly provide on Election Day, which could limit the agency’s timeliness in mitigating or responding to an incident; a lack of clarity regarding CISA’s incident response capabilities in the event of a compromise that exhausts state and local resources, which may limit knowledge about agency capabilities that are available; and a lack of outreach and situational reporting on incidents, threats, and trends on Election Day from headquarters to regional staff following the closure of the polls on the East Coast, which hindered CISA’s coordination of such information with state and local officials. While CISA identified challenges related to its prior efforts, it has not developed plans to address them. According to a CISA official, the agency does not intend to develop a separate plan addressing how it will remediate the identified challenges in the RAND report. Rather, CISA officials noted that the agency plans to address the challenges from that report in the strategic plan and operations plan that it is developing. In addition, the officials noted that CISA may address challenges through other actions that the agency expects to take, such as hiring additional staff. However, CISA’s draft strategic plan, as of November 2019, had only addressed three challenges from the RAND report—countering the threat of disinformation, clarifying how CISA is to support political campaigns, and prioritizing outreach to local jurisdictions. The extent to which the strategic plan, when finalized, will address the other outstanding challenges remains unclear. In addition, the extent to which the operations plan will document how the agency is to address challenges in the RAND report remains uncertain as the operations plan has not yet been completed. Further, CISA has not documented how the agency is to address challenges in the RAND report through other actions that it expects to take before the 2020 elections. Similarly, CISA officials stated that the agency intends to address a subset of the challenges from CISA’s after action report in its anticipated operations plan. However, the extent to which the operations plan will document how the agency is to address challenges in the after action report remains unclear, given that the operations plan has not yet been completed. Without documented plans that address prior challenges, CISA will not be well-positioned to effectively address the challenges identified in prior reviews. This includes addressing how CISA will coordinate among its personnel and provide accurate threat information and other capabilities that address the needs of the election infrastructure community in the remaining months ahead of the 2020 elections. Conclusions With primary elections beginning in February 2020 and culminating in the general election in November 2020, CISA has limited time remaining to help states and local election jurisdictions protect their election infrastructure in advance of these elections. State and local election officials that we contacted have been generally satisfied with CISA’s election security efforts. However, CISA’s unfinished planning means the agency may be limited in its ability to execute a nationwide strategy for securing election infrastructure. In particular, the #Protect2020 Strategic Plan’s higher-level objectives—such as building stakeholder capacity and public awareness—necessarily take time to accomplish. In addition, CISA has not fully assessed and documented how it will address challenges identified in prior assessments, which limits the ability of CISA to address these challenges in its current efforts. Recommendations for Executive Action We are making three recommendations to the Director of the Cybersecurity and Infrastructure Security Agency: The CISA Director should urgently finalize the strategic plan and the supporting operations plan for securing election infrastructure for the upcoming elections. (Recommendation 1) The CISA Director should ensure that the operations plan fully addresses all lines of effort in the strategic plan for securing election infrastructure for the upcoming elections. (Recommendation 2) The CISA Director should document how the agency intends to address challenges identified in its prior election assistance efforts and incorporate appropriate remedial actions into the agency’s 2020 planning. (Recommendation 3) Agency Comments and Our Evaluation DHS provided written comments on a draft of this report, which are reprinted in appendix III. In its comments, the department concurred with all three of our recommendations and identified actions that it plans to take to implement each of the recommendations. For example, the department stated that CISA intends to finalize its strategic and operations plans by February 14, 2020. The department noted that these plans are to provide a strategic overview and operational framework in support of the primaries and the general election in 2020. Further, the department stated that the operations plan, when finalized, is to address all lines of effort in the strategic plan. In addition, the department noted that both the strategic and operations plans are to further document DHS’s plans to address challenges identified during the 2017-2018 election cycle. If implemented effectively, the actions that DHS plans to take in response to the recommendations should address the weaknesses that we identified during our review. DHS and CISA officials also provided technical comments, which we incorporated, as appropriate. We are sending copies of this report to the appropriate congressional committees, the Acting Secretary of Homeland Security, the Director of the Cybersecurity and Infrastructure Security Agency, and other interested parties. In addition, the report is available at no charge on the GAO website at https://www.gao.gov. If you or your staff have any questions about this report, please contact Vijay A. D’Souza, Director, Information Technology and Cybersecurity, at (202) 512-6240 or dsouzav@gao.gov or Rebecca Gambler, Director, Homeland Security and Justice, at (202) 512-8777 or gamblerr@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix IV. Appendix I: Sources of Cybersecurity Threats to the Election Infrastructure The election process relies on various assets—such as information technology systems, networks, equipment, and facilities—that can be broadly categorized as physical, cyber, and human components of the Election Infrastructure Subsector. The assets and components of the election infrastructure are susceptible to a variety of unintentional, or nonadversarial, and intentional, or adversarial, threats. The table below identifies sources of cybersecurity threats to election infrastructure. Appendix II: Voluntary Services for the Election Infrastructure Subsector Provided by CISA The Department of Homeland Security, through the Cybersecurity and Infrastructure Security Agency (CISA), has taken steps to assist state and local election officials in securing election infrastructure by providing a variety of services on a voluntary, no cost basis. These services include cybersecurity assessments, detection and prevention activities, and information sharing. The table below identifies voluntary services that CISA offers to states and local election jurisdictions. Appendix IV: GAO Contact and Staff Acknowledgments GAO Contacts Staff Acknowledgments In addition to the contacts named above, Josh Leiling (Assistant Director), Tom Jessor (Assistant Director), Torrey Hardee (Analyst-in-Charge), Roger Bracy, Rebecca Eyler, Richard Hung, Amanda Miller, Heidi Nielson, Monica Perez-Nelson, Jeff Tessin, Eric Warren, and Haley Weller made significant contributions to this report.

Why GAO Did This Study In January 2017, the Secretary of Homeland Security designated election infrastructure as a critical infrastructure subsector. The designation allowed DHS to prioritize assistance to state and local election officials to protect key election assets, including voter registration databases and voting equipment. The Conference Report (H. Rep. No. 116-9) accompanying the 2019 Consolidated Appropriations Act included a provision for GAO to examine how DHS is implementing key responsibilities to help protect the election infrastructure and the reported benefits and challenges of such efforts. This report addresses (1) DHS's election security efforts and selected election officials' perspectives on them, and (2) DHS's planning for the 2020 elections. GAO reviewed DHS's strategies, plans, and services provided to election officials. GAO also interviewed DHS officials, representatives of the EI-ISAC, a DHS-funded center responsible for sharing threat information nationwide, and election officials from eight states and three local jurisdictions. GAO selected the states and local jurisdictions to provide geographic diversity and variation in election administration, among other factors. The results from these states and localities are not generalizable, but provide insight into election officials' perspectives on DHS's efforts. What GAO Found Since the 2017 designation of election infrastructure as critical infrastructure, the Department of Homeland Security (DHS), through its Cybersecurity and Infrastructure Security Agency (CISA), has assisted state and local election officials in securing election infrastructure through regional support and assistance, education, and information sharing. Such efforts help state and local election officials protect various election assets from threats (see figure). In August 2019, the CISA Director identified election security as one of the agency's top five operational priorities. CISA security advisors, who are located throughout the country, consult with state and local election officials and identify voluntary, no cost services that CISA can provide. According to CISA, as of November 2019, 24 cybersecurity advisors and 100 protective security advisors perform and coordinate cyber and physical security assessments for the 16 critical infrastructure sectors, including the Election Infrastructure Subsector. Technical teams at CISA headquarters generally provide the services, once requested. To further assist state and local election officials, CISA conducted two exercises simulating real-world events and risks facing election infrastructure in August 2018 and June 2019. According to CISA, the 2019 exercise included 47 states and the District of Columbia. In addition, CISA has funded the Election Infrastructure Information Sharing and Analysis Center (EI-ISAC). According to CISA officials, the EI-ISAC is the primary mechanism for exchanging information about threats and vulnerabilities throughout the election community. The EI-ISAC director reported that, as of November 2019, its members included 50 states, the District of Columbia, and 2,267 local election jurisdictions, an increase from 1,384 local jurisdictions that were members in 2018. As a result of its efforts, CISA has provided a variety of services to states and local election jurisdictions in the past 2 years (see table). State election officials with whom GAO spoke were generally satisfied with CISA's support to secure their election infrastructure. Specifically, officials from seven of the eight states GAO contacted said that they were very satisfied with CISA's election-related work. Also, officials from each of the eight states spoke positively about the information that they received from the EI-ISAC. Further, officials from five states told GAO that their relationship with CISA had improved markedly since 2017 and spoke highly of CISA's expertise and availability. To guide its support to states and local election jurisdictions for the 2020 elections, CISA reported that it is developing strategic and operations plans. CISA intended to finalize them by January 2020, but has faced challenges in its planning efforts due to a reorganization within CISA, among other things. In the absence of completed plans, CISA is not well-positioned to execute a nationwide strategy for securing election infrastructure prior to the start of the 2020 election cycle. Further, CISA's operations plan may not fully address all aspects outlined in its strategic plan, when finalized. Specifically, according to CISA officials, the operations plan is expected to identify organizational functions, processes, and resources for certain elements of two of the four strategic plan's lines of effort—protecting election infrastructure, and sharing intelligence and identifying threats. CISA officials stated that CISA was unlikely to develop additional operations plans for the other two lines of effort—providing security assistance to political campaigns, and raising public awareness on foreign influence threats and building resilience. Moreover, CISA has not developed plans for how it will address challenges, such as concerns about incident response, identified in two reviews—one conducted by CISA and the other done by an external entity under contract—of the agency's 2018 election security assistance. Challenges that the reviews identified include: inadequate tailoring of services, which could have made it more difficult for CISA to meet the resource and time constraints of customers such as local election jurisdictions; not always providing actionable recommendations in DHS classified threat briefings or making unclassified versions of the briefings available, which may have hindered election officials' ability to effectively communicate with information technology and other personnel in their agencies who did not have clearances; the inability of CISA personnel supporting election security operations to access social media websites from situational awareness rooms, which hindered their collection and analysis of threat information; few capabilities that CISA field staff could quickly provide on Election Day, which could limit the agency's timeliness in responding to an incident; and a lack of clarity regarding CISA's incident response capabilities in the event of a compromise that exhausts state and local resources, which may limit knowledge about agency capabilities that are available. Although CISA officials said that the challenges identified in the reviews have informed their strategic and operational planning, without finalized plans it is unknown whether CISA will address these challenges. What GAO Recommends GAO is making three recommendations to the CISA Director to (1) urgently finalize the strategic plan and the supporting operations plan for securing election infrastructure for the upcoming elections, (2) ensure that the operations plan fully addresses all lines of effort in the strategic plan for securing election infrastructure for the upcoming elections, and (3) document how the agency intends to address challenges identified in its prior election assistance efforts and incorporate appropriate remedial actions into the agency's 2020 planning. DHS concurred with all three recommendations and provided estimated dates for implementing each of them.

gao_GAO-20-269SP

gao_GAO-20-269SP_0

State and Local Governments Will Need to Make Policy Changes to Achieve Fiscal Balance Our simulations suggest that the sector will likely continue to face a difference between revenues and expenditures during the next 50 years, as measured by its operating balance. We simulated the state and local government sector’s operating balance—a measure of the sector’s ability to cover its current expenditures out of current revenues—to understand the sector’s long-term fiscal outlook based on historical revenue patterns and other assumptions. Because a great majority of states and many local governments are required to balance or nearly balance their operating budgets, the operating balance illustrates the magnitude of fiscal pressures they face. Expenditures and revenues are both simulated to increase as a percentage of gross domestic product (GDP) during the simulation period. However, expenditures are generally expected to grow at a faster rate than revenues, resulting in a declining operating balance (see figure 1). One way of measuring the long-term fiscal challenges faced by the state and local government sector is through an indicator known as the “fiscal gap.” The fiscal gap is an estimate of annual changes in expenditures and in revenues our simulations suggest would be needed to maintain the operating balance equal to zero during the 50-year simulation period. The sector could close the fiscal gap through an increase in revenues, a reduction in expenditures, or a combination of the two of sufficient magnitude. Our simulations suggest that the fiscal gap is about 3.6 percent of GDP over the next 50 years. The sector will need to take actions in annually reducing its expenditures or raising revenues, to achieve fiscal balance. Assuming no change in simulated expenditures, the sector would need to take actions equivalent to increasing its total revenues by 4.2 percent each year to achieve fiscal balance. Alternatively, assuming no change in its simulated revenues, the sector would need to take actions equivalent to decreasing its noninterest expenditures by an amount equal to 3.2 percent of its total expenditures each year. Total expenditure reductions required by the sector are 20.7 percent each year, which includes interest payments on debt that are simulated to be 17.4 percent of annual spending. To eliminate the fiscal gap, the sector would most likely take actions that include a combination of expenditure reductions and revenue increases. Health Care Cost Growth and Other Factors Contribute to the State and Local Sector’s Fiscal Imbalance States’ Spending on Medicaid is a Key Driver of Long-Term Expenditures Our simulations suggest that growth in the sector’s overall expenditures is largely driven by health care expenditures. Medicaid will likely constitute a growing expenditure for state and local governments. In 2018, Medicaid spending was 2.9 percent of GDP compared to 0.85 percent of GDP for other kinds of health care spending such as non-Medicaid social benefit payments and employee health benefit contributions. At the end of our simulations, Medicaid is simulated to be 4.6 percent of GDP and the other kinds of health care spending are 1.3 percent of GDP. After 2029, Medicaid spending in our simulations is derived from Centers for Medicare & Medicaid Services’ (CMS) projections. On average, Medicaid expenditures are expected to rise by 1 percentage point more than GDP each year over the simulation period. Breaking this down, Medicaid expenditures per capita are expected to increase, on average, about 0.6 percent faster than GDP per capita—referred to as excess cost growth. Excess Cost Growth The extent to which health care costs per capita outpace gross domestic product (GDP) growth per person. As shown in figure 4, health care expenditures are simulated to increase from about 3.94 percent of GDP in 2019 to 5.9 percent of GDP in 2068. In comparison, nonhealth, noninterest expenditures, which include all other operational expenditures other than debt interest payments, will decrease as a share of GDP by 2.74 percentage points over the simulation period. Per capita, national health expenditures, which make up part of the health care expenditures in the figure below, are expected to grow on average 0.8 percent faster than GDP each year during the simulation period, according to CMS. Employee Compensation Decreases as a Share of the Sector’s Expenditures during the Simulation Period Employee compensation is the largest expenditure for the state and local government sector. It declines from 6.8 percent of GDP in 2018 to 6.1 percent of GDP in 2068. All spending components, including employee compensation, are simulated to increase in actual dollar amounts during our simulation period. Of the spending components included in employee compensation, only health benefits for employees and retirees increase as a share of employee compensation. In contrast, wages and salaries, pension contributions, and other forms of compensation decrease as a share of employee compensation (see figure 5). These percentages reflect a simulated decrease in state and local government employees’ compensation as a share of GDP. Our simulations suggest that spending on health benefits for state and local government employees and retirees is likely to rise, on average, by 0.9 percentage points more than GDP each year. Similar to the growth in Medicaid spending, growth in spending for these health benefits is due to an increase in the simulated number of employees and retirees enrolled as well as an increase in the simulated amount of health benefits for each employee and retiree. According to our simulations, if employee and retiree health benefits follow trends in overall national health spending, they will likely make up an increasingly larger share of total employee compensation going forward. Our simulations suggest that annual contributions to state and local government employee pension plans will need to remain at their historical 10-year average of 12.9 percent of wages and salaries for state and local governments to meet their long-term pension obligations. Prior to the last decade, from 1999 to 2008 the state and local government sector averaged about an 8 percent contribution rate, which was lower than what our current simulations show is necessary for meeting pension obligations. State and local government contributions to employee pension plans are simulated to decline as a share of GDP, as are wages and salaries of state and local government employees. Growth in Federal Medicaid Grants Drives Revenues Our simulations suggest that federal grants will increase slightly as a share of GDP. The largest grant receipts are for Medicaid which will likely grow more quickly than other types of federal grants making up an increasing share of revenues in the future (see figure 6). The increase in Medicaid expenditures simulated during this period will likely put increasing pressure on both federal and state governments. As a matching formula grant program, the simulated increase in federal Medicaid grants implies an expected increase in Medicaid expenditures that will be shared by state governments. Federal investment grants (i.e., grants intended to finance capital infrastructure investments) and other federal grants unrelated to Medicaid (i.e., grants intended to finance education, social services, housing, and community investment) are simulated to decline as a share of GDP. Further, our simulations suggest that if historical relationships between state and local governments’ tax revenues and tax bases persist, total tax revenues for the state and local government sector will increase from 8.7 percent of GDP in 2019 to 9.1 percent of GDP by the end of the simulation period. As shown in figure 7, the different components of total tax revenues are simulated to remain fairly consistent or slightly increase. The simulations suggest that personal income tax revenues will increase as a share of GDP by about 0.5 percentage points during the simulation period. Sales tax is expected to decrease by approximately 0.2 percentage points and property taxes are simulated to slightly increase as a share of GDP through 2068 from 2.73 percent to 2.86 percent. Economic Growth and Other Factors Could Affect the Sector’s Fiscal Outlook Sensitivity Analysis An analysis using alternative assumptions of one variable to determine the uncertainty, or sensitivity, of another variable. Several factors, or key model variables, could affect the state and local government sector’s long-term fiscal outlook, including economic growth, health care excess cost growth, and the rate of return on pension assets. To see how the outlook changes in response to them, we developed sensitivity analyses—simulations that use alternative assumptions about their growth. For each of these key variables we use a baseline assumption, a higher-than-baseline assumption, and a lower-than- baseline assumption. We determined that these alternative assumptions highlighted the operating balance’s sensitivity to changes, shifting the future fiscal outcomes for the sector. Economic Growth Future trends in GDP growth could affect the state and local government sector’s fiscal outlook. In our simulations, GDP growth is based on the most recent data from the Congressional Budget Office (CBO) and the Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds (OASDI Trustees) which project real GDP (adjusted for inflation) to grow by 1.9 percent per year on average from 2018 through 2029, and by 2 percent per year on average after 2029. Using these projections, our simulations suggest that maintaining current policies would cause the sector’s operating balance to become increasingly negative. Using the OASDI Trustees’ alternative assumptions of real GDP growth at a faster rate—2.7 percent—suggests that the operating balance, while remaining negative, would have an improved outlook compared to the baseline. While growth in revenue and health care spending is largely tied to GDP in our simulations, spending for other components is tied to inflation and population growth and grows more slowly than GDP. As such, increases in GDP growth improve the sector’s outlook. Our simulations, using the OASDI Trustees’ alternative assumptions also show that if GDP were to grow at a slower rate—1.4 percent—the difference between revenues and expenditures would expand, resulting in an increasingly negative operating balance (see figure 8). Excess cost growth in health care is another key determinant of the sector’s fiscal balance. In our simulations Medicaid spending per capita grows about 1.8 percent faster than GDP per capita on average for the period from 2020 through 2029. Medicaid spending per capita grows about 0.6 percent faster on average from the period from 2030 through 2068. Other health expenditures per capita grow about 0.8 percent faster than GDP per capita for the period from 2019 through 2068. Using these projections, our simulations suggest that maintaining current policies will cause the sector’s expenditures to exceed its revenues and this difference will become increasingly negative during the next several decades. The simulations developed assuming zero excess cost growth in Medicaid and national health expenditures suggest that spending would be lower as a share of GDP. The difference between revenues and expenditures would be significantly less negative than the baseline simulations around the middle of the simulation period before stabilizing, but remain negative over the simulation period. In the scenario using the alternative projections from CMS where excess cost growth rises faster— 0.6 percent on average for Medicaid for the period from 2030 through 2068 and 0.9 percent for national health expenditures for the period between 2019 through 2068—our simulations show that the difference between revenues and expenditures would persist for the remainder of the simulation period (see figure 9). The rate of return on pension assets could also affect the state and local government sector’s fiscal outlook. Based on an inflation-adjusted rate of return on pension assets of 5 percent, our simulations suggest that state and local governments would need to make pension contributions equivalent to about 13 percent of employees’ wages and salaries to meet their long-term pension obligations. The simulations we developed using a higher rate of return—7.5 percent—suggest that pension contributions would be about 3 percent of state and local government employees’ wages and salaries to meet pension obligations. Under this scenario, spending would need to be a lower share of GDP and the sector’s outlook would improve. The difference between revenues and expenditures would briefly narrow early on before becoming increasingly negative through the remainder of the simulation period. Alternatively, we estimated that if the rate of return on pension assets is relatively low—2.5 percent—required pension contributions would need to be about 24 percent of state and local government employees’ wages and salaries. Under this scenario our simulations show that spending would be a higher share of GDP and the sector’s outlook would worsen as the sector’s negative operating balance would continue to grow larger (see figure 10). This report was prepared under the direction of Michelle A. Sager, Director, Strategic Issues, who can be reached at 202-512-6806 or sagerm@gao.gov, and Oliver M. Richard, Director, Center for Economics, who can be reached at 202-512-8424 or richardo@gao.gov if there are any questions. GAO staff who made key contributions are listed in appendix II. Appendix I: Objectives, Scope, and Methodology Data This update of the state and local government fiscal model used aggregate data on the state and local government sector and national data on other variables from the following sources: the Agency for Healthcare Research and Quality, Bloomberg, the Board of Governors of the Federal Reserve System, the Board of Trustees of the Federal Old- Age, Survivors, and Disability Insurance Program (OASDI Trustees), BEA, the Bureau of Labor Statistics, the Census Bureau, the Centers for Medicare & Medicaid Services (CMS), the Congressional Budget Office (CBO), and the Social Security Administration. These data sources are generally the same data sources we used for our prior update. We used annual observations on historical data through 2018 where available. Objectives and Methodology This report updates GAO’s state and local fiscal model to simulate the fiscal outlook for the state and local government sector. This includes identifying the factors that are likely to contribute to the state and local sector’s fiscal imbalance. The level of receipts and expenditures for the state and local government sector as a whole in future years is based on current and historical spending and revenue patterns. We used Table 3.3 of the National Income and Product Accounts (NIPA)—State and Local Government Current Receipts and Expenditures—prepared by BEA as an organizing framework for developing our model, and we simulated state and local government receipts and expenditures using methods similar to those we have used in prior updates. Our simulations of real U.S. gross domestic product (GDP) were consistent with the growth path developed by CBO for the period from 2019 through 2029 and by the OASDI Trustees for the period thereafter. Our simulations of U.S. population was consistent with the growth path developed by the OASDI Trustees, and our simulations of excess cost growth for national health expenditures and for Medicaid were consistent with CMS projections, all for the entire simulation period. Our simulations of other variables, such as the GDP price index, personal income, and 3-month U.S. Department of the Treasury (Treasury) rates, were consistent with the growth paths for these variables developed by CBO for as much of the simulation period as possible. Otherwise, we developed our own assumptions about the likely future growth path of the variables in our model. In general, we assumed that current policies remain in place and that all levels of government continue to provide services at current per capita levels. A detailed description of the model is in appendix I of GAO, State and Local Governments’ Fiscal Outlook: 2018 Update, GAO-19-208SP (Washington, D.C.: December 2018). We describe below where we updated equations or added equations to the model. Otherwise our approach is the same as the approach we used in that update. State and Local Government Sector Receipts We simulated the future growth paths of the following types of state and local government revenues: current tax receipts, contributions to government social insurance, income on financial assets, current transfer receipts, the surplus from government enterprises, and capital transfer receipts. We also simulated the future growth path of state and local government long-term debt issuance. We updated some of the equations we used to simulate tax receipts(see table 1). We also added equations to simulate current transfers from the rest of the world to state and local governments, disaster-related insurance benefits to state and local governments, and other capital transfers to state and local governments, which we had not included in prior updates. The equations we used to simulate the other types of receipts are the same as the equations we used in GAO-19-208SP. We simulated the future growth paths of the following types of state and local government expenditures: consumption expenditures, current transfer payments, interest paid on outstanding state and local government debt, subsidies, capital outlays, and consumption of fixed assets (depreciation). We also simulated the future growth path of the state and local government sector’s net social insurance fund balance. We updated some of the equations we used to simulate the interest paid on outstanding state and local government debt (see table 1 above). We also added equations to simulate current transfer payments to the rest of the world, which we had not included in prior updates. Otherwise, the approach we used to simulate expenditures is the same as the approach we used in GAO-19-208SP. State and Local Government Sector Fiscal Balance Our main indicator of the sector’s fiscal balance is its operating balance net of funds for capital expenditures (henceforth, operating balance), which is a measure of the sector’s ability to cover its current expenditures out of current receipts. Operating balance is defined as total receipts minus (1) capital outlays not financed by long-term debt issuance, (2) current expenditures less depreciation, (3) current surplus of state and local government enterprises, and (4) net social insurance fund balance. We also estimated the annual changes in spending and in receipts that our simulations suggest would be needed to maintain the operating balance equal to zero during the 50-year simulation period, which we refer to as the “fiscal gap.” As discussed above, our baseline simulations assume that current policies remain in place and that all levels of government continue to provide services at current per capita levels. We then simulated the change in total expenditures needed to maintain the operating balance equal to zero. To estimate the annual change in spending needed to maintain balance we calculated the present value of that change as a percentage of the present value of baseline total expenditures and as a percentage of the present value of U.S. GDP, all for a 50-year period. We also calculated the interest and non-interest expenditure components of the change in total expenditures needed to maintain balance. We used a similar approach to estimate the annual change in total receipts needed to maintain balance. Sensitivity Analysis We assessed the sensitivity of our baseline results to alternative projections of real U.S. GDP growth, health care excess cost growth, and the real rate of return on state and local government pension fund assets. Following the same approach we used in GAO-19-208SP, for each of these variables, we selected an alternative projection associated with faster growth or rate of return and one associated with slower growth or rate of return. Real U.S. GDP. For our baseline simulations, we used CBO projections of real GDP for the period from 2019 through 2029 and the OASDI Trustees’ intermediate projections of real U.S. GDP growth for the years thereafter. For our sensitivity analysis, we used the OASDI Trustees’ high-cost and low-cost projections. Health care excess cost growth. For our baseline simulations, we used CMS’s baseline projection of national health expenditures excess cost growth and we estimated Medicaid excess cost growth based on CMS’s baseline projections. For our sensitivity analysis, we used CMS’s alternative projection of national health expenditures excess cost growth and we estimated Medicaid excess cost growth based on CMS’s alternative projections. As another alternative, we simulated the model assuming both zero excess cost growth for national health expenditures and Medicaid. Our simulations used CBO’s projection of federal spending on Medicaid, CHIP, and exchange subsidies as a fraction of GDP to simulate certain variables related to state and local government spending on Medicaid and other health spending. This projection incorporates excess cost growth for the period from 2019 through 2029 but assumes zero excess cost growth starting in 2030, so we could only vary Medicaid excess cost growth in the alternative simulations for 2030 and later. Real rate of return on state and local government pension assets. For our baseline simulations, we assumed a 5 percent real rate of return on state and local government pension assets. For our sensitivity analysis, we used 2.5 percent and 7.5 percent. Table 2 shows the growth rates or rates of return associated with the baseline and alternative projections of each variable for the simulation period. We simulated the model changing either real U.S. GDP growth, health care excess cost growth, or the real rate of return on pension assets, leaving the other variables fixed at their baseline values. Thus, our sensitivity analysis is in the spirit of a partial equilibrium comparative statics analysis that sheds light on how each of the individual variables may affect the state and local government sector’s fiscal outlook. However, these variables are likely to be correlated, so future changes in one would likely be associated with changes in others. Caveats and limitations Our approach has a number of limitations and the results should be interpreted with caution. First, the state and local fiscal model is not designed for certain types of analyses. The simulations are not intended to provide precise predictions. Even though we know that these governments regularly make changes in tax laws and expenditures, the model essentially holds current policies in place and analyzes the fiscal future for the sector as if those policies were maintained because it would be highly speculative to make any assumptions about future policy adjustments. In addition, fiscal outcomes related to the sector’s financial position and solvency may not reflect all aspects of the sector’s “health.” Other indicators include economic indicators that go beyond the sector’s financial position to include economic growth, income, or distributional equity, as well as indicators of the quality of services provided by the sector, including education, health care, infrastructure, and other public goods and services. Finally, our unit of analysis is the state and local government sector as a whole, so our results provide an assessment of the sector’s fiscal outlook. However, individual state and local governments likely exhibit significant heterogeneity in their expenditure and revenue patterns and their fiscal outlooks will likely differ from the sector as a whole. Nevertheless, it is informative to assess the overall fiscal outlook because doing so reveals the outlook for state and local governments as a sector. In addition, aggregate data on the sector is available on a more timely basis than data for individual state and local governments, allowing for a better assessment of the sector’s current fiscal outlook. Our results for the sector also provide a baseline from which to view the experiences of individual state and local governments. Finally, assessing the fiscal outlook of the sector as a whole can help mitigate the tendency to extrapolate from the most visible, but potentially not representative, experiences of individual states or localities. Appendix II: GAO Contacts and Staff Acknowledgments GAO Contacts Acknowledgments In addition to the contact named above, Peter Del Toro, Courtney LaFountain, Melissa Wolf (Assistant Directors), Silvia Symber (Analyst-in- Charge), Shelby Clark, Amalia Konstas, Dylan Stagner, Frank Todisco, Walter Vance, and Alicia White made significant contributions to this report. Related GAO Products State and Local Governments’ Fiscal Outlook: December 2018 Update, GAO-19-208SP. Washington, D.C.: Dec. 12, 2018. State and Local Governments’ Fiscal Outlook: December 2016 Update, GAO-17-213SP. Washington, D.C.: Dec. 8, 2016. State and Local Governments’ Fiscal Outlook: December 2015 Update, GAO-16-260SP. Washington, D.C.: Dec. 16, 2015. State and Local Governments’ Fiscal Outlook: December 2014 Update, GAO-15-224SP. Washington, D.C.: Dec. 17, 2014. State and Local Governments’ Fiscal Outlook: April 2013 Update, GAO-13-546SP. Washington, D.C.: Apr. 29, 2013. State and Local Governments’ Fiscal Outlook: April 2012 Update, GAO-12-523SP. Washington, D.C.: Apr. 5, 2012. State and Local Government Pension Plans: Economic Downturn Spurs Efforts to Address Costs and Sustainability, GAO-12-322. Washington, D.C.: Mar. 2, 2012. State and Local Governments’ Fiscal Outlook: April 2011 Update, GAO-11-495SP. Washington, D.C.: Apr. 6, 2011. State and Local Governments: Knowledge of Past Recessions Can Inform Future Federal Fiscal Assistance, GAO-11-401. Washington, D.C.: Mar. 31, 2011. State and Local Governments: Fiscal Pressures Could Have Implications for Future Delivery of Intergovernmental Programs, GAO-10-899. Washington, D.C.: July 30, 2010. State and Local Governments’ Fiscal Outlook: March 2010 Update, GAO-10-358. Washington, D.C.: Mar. 2, 2010. Update of State and Local Government Fiscal Pressures, GAO-09-320R. Washington, D.C.: Jan. 26, 2009. State and Local Fiscal Challenges: Rising Health Care Costs Drive Long- term and Immediate Pressures, GAO-09-210T. Washington, D.C.: Nov. 19, 2008. State and Local Governments: Growing Fiscal Challenges Will Emerge during the Next 10 Years, GAO-08-317. Washington, D.C.: Jan. 22, 2008. Our Nation’s Long-Term Fiscal Challenge: State and Local Governments Will Likely Face Persistent Fiscal Challenges in the Next Decade, GAO-07-1113CG. Washington, D.C.: July 18, 2007. State and Local Governments: Persistent Fiscal Challenges Will Likely Emerge within the Next Decade, GAO-07-1080SP. Washington, D.C.: July 18, 2007.

Why GAO Did This Study Fiscal sustainability presents a national challenge shared by all levels of government. Since 2007, GAO has published simulations of long-term fiscal trends in the state and local government sector, which have consistently shown that the sector faces long-term fiscal pressures. While a great majority of states have requirements related to balancing their budgets, deficits can arise for reasons including planned annual revenues are not generated at the expected rate, demand for services exceeds planned expenditures, or both, resulting in a near-term operating deficit. This report updates GAO's state and local fiscal model to simulate the fiscal outlook for the state and local government sector. This includes identifying the components of state and local expenditures likely to contribute to the sector's fiscal pressures as well as the effects of revenue changes on the sector's outlook. GAO's model uses the Bureau of Economic Analysis's National Income and Product Accounts as the primary data source and presents the results in the aggregate for the state and local government sector as a whole. The model shows the expected level of receipts and expenditures for the sector until 2068, based on historical spending and revenue patterns. In addition, the model assumes that the current set of policies in place across state and local governments remains constant to show a simulated long-term outlook. Because the model covers the sector in the aggregate, the fiscal outcomes for individual states and localities cannot be identified. What GAO Found GAO's simulations suggest that state and local governments will likely face an increasing difference between expenditures and revenues during the next 50 years as reflected by the operating balance—a measure of the sector's ability to cover its current expenditures out of its current revenues. While both expenditures and revenues are projected to increase as a percentage of United States' gross domestic product (GDP), a difference between the two is projected to persist because expenditures are expected to grow faster than revenues throughout the simulation period. The sector would need to make changes to avoid fiscal imbalance and assure that revenues are at least equal to expenditures. GAO's simulations suggest that growth in the sector's overall expenditures is largely driven by health care, with states' share of Medicaid spending as the primary driver. These expenditures are projected to grow more than GDP each year. Employee compensation, the largest share of operating expenditures, decreases as a share of GDP during the simulation period. Health benefits are the only component of employee compensation that increase as a percentage of GDP. Revenues from federal grants to states and localities are also expected to increase during the simulation period, in part because of Medicaid grants to states. GAO also conducts sensitivity analyses to see how the sector's outlook changes when using alternative assumptions of key model variables – economic growth, health care excess cost growth, and the real rate of return on pension assets. Using these alternative assumptions highlights the operating balance's sensitivity to changes and possible shifts in the future fiscal outcomes for the sector.

gao_GAO-20-491

gao_GAO-20-491_0

Background Micronesia, the Marshall Islands, and Palau are among the smallest countries in the world. In fiscal year 2017, the three FASs had a combined resident population of approximately 175,000 (102,622 in Micronesia; 54,354 in the Marshall Islands; and 17,901 in Palau). Interior’s Office of Insular Affairs (OIA) has primary responsibility for monitoring and coordinating U.S. assistance to the FASs, and State is responsible for government-to-government relations. The U.S. relationship with the FASs began when American forces liberated the islands from Japanese control near the end of World War II. In 1947, the United States entered into a trusteeship with the United Nations and became the administering authority over Micronesia, the Marshall Islands, and Palau. Voters approved the Constitution of the Federated States of Micronesia in 1978 and approved the Constitution of the Marshall Islands in 1979. Both Micronesia and the Marshall Islands remained subject to the authority of the United States until 1986, when a compact of free association went into effect between the United States and the two nations. The Palau constitution took effect in 1981, and Palau entered into a compact of free association with the United States in 1994. Micronesia and Marshall Islands became members of the United Nations in 1991, while Palau joined the organization in 1994. Compacts of Free Association Economic Assistance Provisions Under its compacts with Micronesia, the Marshall Islands, and Palau, the United States provided economic assistance that includes access to certain federal services and programs, among other things, for defined time periods. Economic assistance to Micronesia and the Marshall Islands. The 1986 compact of free association between the United States and Micronesia and the Marshall Islands, respectively, provided about $2.6 billion in funding for fiscal years 1987 through 2003. In 2003, the United States approved amended compacts of free association with the two countries. According to Interior, economic assistance under the amended compacts is projected to total $3.6 billion, including payments for compact sector grants and trust fund contributions for both countries in fiscal years 2004 through 2023. Funding under the original compact and amended compacts has been provided to Micronesia and the Marshall Islands through Interior. Economic assistance to Palau. The compact of free association between the United States and Palau entered into force in 1994 and provided $574 million in funding through Interior for fiscal years 1995 through 2009 for assistance to the government, contributions to a trust fund, construction of a road, and federal services. In September 2010, the United States and Palau signed an agreement that would, among other things, provide for additional assistance to Palau, including contributions to its trust fund. The 2010 agreement and subsequent amendments entered into force in September 2018. According to Interior, direct assistance to Palau under the compact will total $229 million for fiscal years 2010 through 2024, including $105 million that Congress provided in annual appropriations in fiscal years 2010 through 2017. Defense-Related Provisions Under the compacts, the United States has responsibility for defense and security matters in, and relating to, each of the FASs, and subsidiary agreements pursuant to the compacts provide for U.S. military use and operating rights in these countries. According to the Department of Defense, the compacts have enabled it to maintain a critical strategic position in the Indo–Pacific region. The compact with the Marshall Islands also provided for a separate agreement that constituted a full and final settlement of all claims resulting from U.S. nuclear tests conducted in the Marshall Islands during the period 1946 through 1958. In addition, a subsidiary agreement with the Marshall Islands secured the United States’ access to the U.S. military facilities on Kwajalein Atoll, which are used for missile testing and space tracking activities. Migration-Related Provisions Under the compacts, eligible FAS citizens are exempt from certain visa and labor certification requirements of the Immigration and Nationality Act as amended. The migration provisions of the compacts allow eligible FAS citizens to enter the United States (including all states, territories, and possessions) and to lawfully work and reside in the United States indefinitely. The implementing legislation for the 1986 compact with Micronesia and the Marshall Islands stated that it was not Congress’s intent to cause any adverse consequences for U.S. territories and commonwealths and the state of Hawaii. The legislation further declared that Congress would act sympathetically and expeditiously to redress any adverse consequences. In addition, the legislation authorized compensation to be appropriated for these areas that might experience increased demands on their educational and social services from compact migrants from Micronesia, the Marshall Islands, and Palau. The legislation required the President to report and make recommendations annually to Congress regarding adverse consequences resulting from the compact and provide statistics on compact migration. In November 2000, Congress made the submission of annual reports about the impact of compact migration in affected jurisdictions—that is, compact impact reports—optional and shifted the responsibility for preparing these reports from the President to the governors of Hawaii and the territories. Legislative Actions to Address Compact Impact In December 2003, Congress took steps in the amended compacts’ implementing legislation to address compact impact in designated U.S. areas. The legislation restated Congress’s intent not to cause any adverse consequences for the areas defined as affected jurisdictions— Hawaii, Guam, the CNMI, and American Samoa. In addition, the legislation authorized and appropriated funding for compact impact grants to the affected jurisdictions, to be allocated on the basis of the proportion of compact migrants living in each jurisdiction. Further, the legislation required an enumeration of compact migrants to be undertaken at least every 5 years. The legislation also permitted affected jurisdictions to submit compact impact reports to the Secretary of the Interior. Compact Impact Grants to Affected Jurisdictions The implementing legislation for the amended compacts authorized and appropriated $30 million for each fiscal year from 2004 through 2023 for grants to the affected jurisdictions. According to the legislation, the grants are provided to aid in defraying costs incurred by these jurisdictions as a result of increased demand for services due to the residence of compact migrants. OIA reviews the affected jurisdictions’ annual proposals for the use of the funds and provides the funds to the jurisdictions as compact impact grants. The grants are to be used only for health, educational, social, or public safety services or for infrastructure related to such services. Figure 1 shows the locations of the FASs and the affected jurisdictions. Required Enumerations of Compact Migrants The implementing legislation for the amended compacts requires Interior to conduct an enumeration of compact migrants, which is to be supervised by the Census Bureau or another organization selected by Interior, at least every 5 years beginning in fiscal year 2003. On the basis of these enumerations, each affected jurisdiction is to receive a portion of the annual $30 million appropriation in proportion to the number of compact migrants living there. The legislation permits Interior to use up to $300,000, adjusted for inflation, of the annual appropriation for compact impact to conduct each enumeration. The amended compacts’ implementing legislation defines a compact migrant, for the purposes of the enumeration, as “a person, or their children under the age of 18, admitted or resident pursuant to [the compacts] who as of a date referenced in the most recently published enumeration is a resident of an affected jurisdiction.” Compact Migrant Eligibility for Selected Federal Programs Compact migrants have varying eligibility for certain U.S. federal government programs. Eligibility for some federal programs changed as a result of the 1996 Personal Responsibility and Work Opportunity Reconciliation Act. For example, when the compacts were signed, FAS citizens were eligible for Medicaid; however, the act removed this eligibility. Table 1 shows compact migrants’ eligibility status for selected federal benefit programs as of November 2019. Compact Migrant Population Has Grown, with About Half Residing on U.S. Mainland Total Compact Migrant Population in U.S. Areas Grew by 68 Percent over 9 Years From 2009 to 2018, the number of compact migrants living in U.S. states and territories rose by an estimated 68 percent, from about 56,000 to about 94,000. In 2011, we reported that combined data from the Census Bureau’s 2005-2009 American Community Survey and 2008 enumeration showed an estimated 56,345 compact migrants living in U.S. areas. During the period 2013 to 2018, an estimated 94,399 compact migrants lived in U.S. areas, according to combined data from the Census Bureau’s 2013-2017 American Community Survey and 2018 required enumeration in Guam, the CNMI, and American Samoa. This estimate includes Micronesian and Marshallese citizens who entered the United States after 1986, Palauan citizens who entered the United States after 1994, and certain U.S.-born children younger than 18 years. About Half of All Compact Migrants Resided on U.S. Mainland in 2013-2018 Data from the 2013-2017 American Community Survey and the 2018 enumeration indicate that an estimated 50 percent of compact migrants lived on the U.S. mainland and an estimated 49 percent lived in the affected jurisdictions during this period: 26 percent in Hawaii, 20 percent in Guam, and 3 percent in the CNMI. This estimate indicates growth in the number of compact migrants on the U.S. mainland since 2011, when we reported that the Census Bureau estimated 58 percent of compact migrants lived in the affected jurisdictions. The Census Bureau estimated that 11 states in the U.S. mainland, in addition to three of the four affected jurisdictions—Hawaii, Guam, and the CNMI—had compact migrant populations of more than 1,000, according to the 2013-2017 American Community Survey and the 2018 enumeration (see fig. 2). Stakeholders Expressed Concerns about Undercounting of Compact Migrants Stakeholders we interviewed—including FAS embassy and consular officials, FAS community members, state government officials, and representatives of private sector and nonprofit organizations—expressed concerns about the Census Bureau’s prior estimates of compact migrants. Some Arkansas stakeholders cited other, higher estimates of the FAS population in their state. Moreover, some stakeholders said that compact migrant populations are apprehensive or distrustful about being formally counted through surveys or the census. Stakeholders also noted that some compact migrant communities have felt frustrated at having been encouraged to respond to surveys and be counted but not experiencing any benefit from these efforts, according to a nonprofit official and FAS community members. Marshallese consular officials said that they believed the 2010 census undercounted their citizens, noting that the Census Bureau did not employ any Marshallese surveyors in the Arkansas counties with Marshallese populations. Stakeholders also expressed concern about the decennial census to be conducted in 2020, which, like the 2010 decennial census, will collect information on race. Nonprofit organization officials whom we interviewed expressed concern that the 2020 census could result in an undercounting of compact migrants because of language barriers and compact migrants’ difficulty accessing the census form online. Arkansas health care and private sector representatives and the Marshallese consulate described plans to address barriers to obtaining a more accurate count of the population in the 2020 census. Hawaii is making a statewide effort to ensure that compact migrants are counted in the 2020 census, according to Hawaii state officials. According to Guam officials, an outreach effort in Guam has leveraged “trusted voices,” or parties known to compact migrant communities there, to communicate the importance of responding to the 2020 census. Census Data Provide Additional Information about Compact Migrants in the States, the District of Columbia, and Puerto Rico Data from the American Community Survey showed an estimated 72,965 compact migrants living in the 50 states, the District of Columbia, and Puerto Rico in 2013 through 2017. An estimated 31,425 compact migrants living in these areas (43 percent) were U.S. citizens. The remaining estimated 41,540 (57 percent) were not U.S. citizens. The U.S. citizens who were counted included naturalized citizens and minor-age U.S. citizen children of compact migrants, who would no longer be counted as compact migrants after reaching 18 years of age. An estimated 25,555 compact migrants living in these areas were born in Micronesia; 20,545 were born in the Marshall Islands; and 3,435 were born in Palau. These totals do not include compact migrants born in the FAS and living in Guam, the CNMI, or American Samoa, because the American Community Survey does not cover these territories. An estimated 27,735 compact migrants living in these areas who were 18 years and older (69 percent) were in the civilian labor force. Of those, 24,540 (89 percent) were employed and 3,195 (12 percent) were unemployed. An estimated 1,660 compact migrants living in these areas—4 percent of compact migrants 17 years and older—were on active duty in the U.S. military or had served on active duty in the past. For additional American Community Survey data on compact migrant demographics, see appendix IV. Reasons for Migration to U.S. Areas Vary Compact migrants move to U.S. areas for a range of reasons, including greater economic and educational opportunities, better access to health care, a desire to join family members in the United States, and a wish for greater personal freedom. In some communities we visited, stakeholders noted that FAS citizens had come to the United States for school or work before the compact with Micronesia and the Marshall Islands and the compact with Palau went into effect but that the compacts had opened the option of migration to a broader range of individuals. Economic opportunities. Compact migrants described moving to U.S. areas for better, more reliable jobs and higher wages. Having a better-paying job in the United States sometimes allows individuals to send remittances or consumer goods to family members living in an FAS. Other compact migrants move to U.S. areas to join the military. Educational opportunities. Compact migrant families often move to U.S. areas so that their children will have access to improved primary and secondary education, according to compact migrants. Some compact migrants travel to U.S. areas to attend college and choose to stay to work, including to pay off their student loans, according to consular officials and compact migrants. Health care access. Compact migrants sometimes migrate to U.S. areas to obtain medical treatment for themselves or family members, according to FAS community members and consular officials. Some medical procedures or treatments, such as dialysis or access to specialists, are not available in the FASs, according to federal and nonprofit officials. Family. Many compact migrants relocate to the United States to join family members and communities already living there, according to consular and nonprofit officials. Personal freedom. Some compact migrants said that they have more personal, social, and cultural freedom in the United States than in their more traditional home country. Changes in the natural environment in the FASs have also prompted migration from those areas, according to FAS representatives. Depleted food resources and effects of climate change—including more-frequent typhoons, coral reef bleaching, and depletion of fishing stocks—have contributed to migration, according to an FAS official. In addition, members of Marshallese communities cited rising sea levels and frequent tidal flooding as reasons for migrating from the Marshall Islands to U.S. areas. Some Marshallese community members also noted that the legacy of U.S. nuclear testing had contributed to their decision or need to move. Compact migrants cited varied reasons for choosing to migrate to specific locations. For example, representatives of FAS communities in Guam and the CNMI noted the FASs’ closer proximity to those territories than to the U.S. mainland as well as the similarity of Guam’s and the CNMI’s island cultures to those of their home countries. Also, some compact migrants in Arkansas and Oregon cited the lower cost of living and a perception of less discrimination or greater safety there than in Hawaii. Marshallese community members often migrate to Arkansas for jobs in the poultry industry. Consular officials noted that, because of comparatively lower wages and fewer housing options in the FASs, returning to their countries after living in U.S. areas can be difficult for some compact migrants. Some compact migrants said that it is also difficult to find a good job in their home countries without family or political connections. According to an FAS official, some compact migrants retire to their home countries. However, several compact migrants we spoke with said they planned to stay in U.S. areas to be close to medical care or to children and grandchildren born there. Hawaii, Guam, and the CNMI Have Reported Compact Impact Costs and Received Annual Grants to Defray Them The affected jurisdictions of Hawaii, Guam, and the CNMI reported estimated compact impact costs (i.e., costs incurred as a result of increased demands on public services from compact migrants) that totaled $3.2 billion during the period fiscal years 2004 through 2018 and increased over time for Hawaii and Guam. Interior has provided compact impact grants totaling more than $30 million annually to the affected jurisdictions, each of which uses the funds differently. In October 2019, Census discovered an error in the 2013 and 2018 enumerations, which Interior had used to determine the distribution of compact impact grant funds and which resulted in misallocation of these funds for fiscal years 2015 through 2020. In February 2020, Interior officials told us that the department had developed a modified plan for compact impact grants in fiscal years 2021 through 2023 that, according to the officials, is intended to correct the misallocation. Hawaii’s and Guam’s Reported Compact Impact Costs Have Risen, while the CNMI’s Have Varied Hawaii, Guam, and the CNMI reported a total of $3.2 billion in estimated compact impact costs during the period fiscal years 2004 through 2018, with estimated annual costs increasing over time for Hawaii and Guam and fluctuating for the CNMI. Hawaii reported $1.8 billion in total estimated compact impact costs. Hawaii’s reported annual costs increased from $55 million in fiscal year 2004 to $198 million in fiscal year 2018. Guam reported $1.2 billion in total estimated compact impact costs. Guam’s reported annual costs increased from $33 million in fiscal year 2004 to $147 million in fiscal year 2017. The CNMI reported $116 million in total estimated compact impact costs. The CNMI’s reported annual costs amounted to $10 million in both fiscal year 2004 and fiscal year 2018 but fluctuated over time, ranging from a low of about $3 million in fiscal year 2011 to a high of $12 million in fiscal year 2014. For a summary of the estimated compact impact costs reported by the three affected jurisdictions, see figure 3. For more details of their compact impact reporting, see appendix V. The three affected jurisdictions reported compact impact costs for education, health, public safety, and social services (see table 2). As the table shows, the highest total costs in fiscal year 2017 were for education and health services. In November 2011, we found that Interior’s reporting to Congress on compact impact had been limited, and we identified weaknesses in existing compact impact reporting. We found that some jurisdictions did not accurately define compact migrants, account for federal funding that supplemented local expenditures, or include revenue received from compact migrants. Our November 2011 report recommended that the Secretary of the Interior disseminate guidelines to the affected jurisdictions that adequately addressed concepts essential to producing reliable impact estimates and that the Secretary call for the use of these guidelines in developing compact impact reports. Although Interior developed a draft of compact impact reporting guidelines in 2014, it had not disseminated such guidelines to the affected jurisdictions as of February 2020. In 2019, Interior awarded the Guam government a technical assistance grant for $280,000 to conduct a cost-benefit analysis to determine compact migrants’ economic contribution to the local economy. The effort will reportedly also seek to address weaknesses and methodological concerns related to compact impact costs calculated by Hawaii, Guam, and the CNMI. Guam officials said that the grant application was prepared in response to our prior critique of their compact impact estimation methodology. The grant was awarded to the Guam Bureau of Statistics and Plans, which contracted with University of Guam consultants to carry out the work beginning in October 2019. Guam officials expected this work to result in two reports—one identifying economic contributions by compact migrants (expected September 2021) and another proposing a methodology for determining compact impact costs (expected August 2022). Hawaii, Guam, and the CNMI Have Received Grants to Defray Compact Impact Costs Compact Impact Grant Funding During fiscal years 2004 through 2019, Hawaii, Guam, and the CNMI received a combined total of approximately $509 million in compact impact grant funding. This total includes (1) annual compact impact grant funding allocated from $30 million authorized and appropriated in the amended compacts’ implementing legislation and (2) additional compact impact grant funding allocated from annual appropriations. In fiscal years 2004 through 2019, Interior made annual allocations of the $30 million of compact impact grant funds authorized and appropriated in the amended compacts’ implementing legislation. Interior provided these allocations as compact impact grants to each affected jurisdiction to defray their costs due to the residence of compact migrants. Interior used the four most recent enumerations— conducted in 2003, 2008, 2013, and 2018—as the basis for these annual allocations. Since fiscal year 2012, Interior has provided additional compact impact grant funding to the affected jurisdictions from annual appropriations. This additional funding has ranged from approximately $3 million to $5 million per year since fiscal year 2012. Interior has allocated the additional funding on the basis of the 2013 and 2018 enumerations. Table 3 shows the total amounts that Hawaii, Guam, and the CNMI received as compact impact grant funding in fiscal years 2004 through 2019. Affected jurisdictions use their compact impact grant funding in varying ways and report on their use of the funds to Interior. Hawaii allocates the entirety of its compact impact grant—approximately $13 million annually since fiscal year 2015—to the state’s MedQuest division to defray costs of providing medical services to compact migrants. Guam has used some of its approximately $15 million of compact impact funding each year for new schools constructed through leasebacks (see fig. 4 for photos of several schools built by the Guam government with compact impact funds). The CNMI allocates its approximately $2 million of compact impact funding each year across the education, health care, public safety, and social service sectors. Hawaii, Guam, and CNMI officials have emphasized that compact impact funding does not fully compensate for the expenses associated with compact migration. For stakeholder suggestions related to compact impact funding and other issues, see appendix VII. Misallocation of Compact Impact Grant Funding in Fiscal Years 2015-2020 In October 2019, Census Bureau officials discovered an error in the 2013 and 2018 Census Bureau enumerations that caused inaccurate counts of compact migrants in Hawaii and, according to Interior officials, resulted in misallocation of compact impact funding for Hawaii, Guam, the CNMI, and American Samoa in fiscal years 2015 through 2020. Relative to the proportion of compact migrants in each jurisdiction, allocations to Hawaii were a total of $16.9 million lower than they would have been without the enumeration error while allocations to Guam, the CNMI, and American Samoa were higher than they would have been without the error. Table 4 summarizes the under- and overpayments of compact impact funding to each affected jurisdiction that, according to Interior officials, resulted from the enumeration error. The enumeration error was discovered in late October 2019, near the beginning of fiscal year 2020. As of February 2020, OIA officials had developed a modified planned allocation of compact impact funds for fiscal years 2021 to 2023. Beginning in fiscal year 2021, OIA plans to divide the $30 million of annual compact impact grant funding in fiscal years 2021 through 2023 using corrected base allocations from the updated 2018 enumeration from Census Bureau, according to an Interior preliminary assessment. The base allocations will be adjusted upward for Hawaii and downward for Guam, the CNMI, and American Samoa to correct for the erroneous payments in fiscal years 2015 through 2020. See table 5 for a comparison of the originally planned fiscal year 2020 allocation (based on the erroneous enumeration) and the revised allocation (based on the corrected enumeration) as well as the grant amounts that OIA proposed for fiscal years 2021 through 2023 to correct for the erroneous payments. Compact Migration Affects Government Programs, Workforces, and Societies The governments of some of the U.S. areas we visited identified effects of providing public education and health care services to compact migrants. Compact migration’s effects in U.S. areas we visited also include budgetary contributions from compact migrants’ payment of taxes and fees as well as budgetary costs of other government programs and services to compact migrants. Stakeholders in the U.S. areas additionally discussed the participation of compact migrants in those areas’ workforces and communities in terms of contributions and impacts of compact migration. U.S. Area Governments Identify Effects of Providing Programs and Services to Compact Migrants Education Children of compact migrants attending U.S. public primary and secondary schools sometimes receive additional or specialized services, such as support for English language learners, according to state and territorial officials. In the U.S. areas we visited, state and territorial departments and school districts have identified and counted compact migrant students by means of one or more criteria, including ethnicity, language, and place of birth. See table 6 for estimated numbers of compact migrant students in the states and territories we visited and the criteria that each state or territory used to count students as compact migrants. Compact migrants are eligible for in-state tuition at some U.S.-based colleges and universities, according to university, nonprofit, and state officials. For example, in Guam, compact migrants attending the University of Guam are eligible for in-state tuition. In Oregon, FAS citizens are eligible for in-state tuition after a 1-year residency period in the state, according to nonprofit officials. In Arkansas, Marshallese citizens are eligible for in-state tuition after a 3-year residency period in the state, according to state tuition policy and officials. Health Care States and territories have reported budget and program effects related to health care for compact migrants who are eligible for federal benefits as well as health care for individuals, including compact migrants, who are ineligible for federal benefits and lack private insurance or other means of payment. U.S. area governments sought to enable compact migrants’ access to health care in several ways, including extending access to the federal Children’s Health Insurance Program (CHIP) or Medicaid and leveraging federal health insurance tax credits and other federal funding. According to some U.S. area government officials, some of these programs are provided specifically because compact migrants are ineligible for certain programs at the federal level. Extended Access to Children’s Health Insurance Program or Medicaid The Children’s Health Insurance Program Reauthorization Act of 2009 included an option for states to cover children younger than 21 years and pregnant women in both CHIP and Medicaid who are lawfully residing in the United States—a definition that includes compact migrants—and who are otherwise eligible under the state plan. Therefore, in some U.S. areas, non-U.S. citizen compact migrants who are children or pregnant may access federal health insurance coverage through CHIP or Medicaid. As of February 2020, 38 states and territories and the District of Columbia had extended such coverage to lawfully residing non–U.S. citizen pregnant women or children, including compact migrants, who met all other eligibility requirements (see fig. 5). According to Arkansas officials, their state’s decision to extend this coverage was sought in part to address unmet needs of compact migrants living in Arkansas. Subsidized Coverage in Patient Protection and Affordable Care Act Exchanges Compact migrants are eligible to purchase individual market health insurance plans through health insurance exchanges established under the Patient Protection and Affordable Care Act (PPACA). Individuals purchasing coverage through the exchanges may be eligible, depending on their incomes, to receive financial assistance in the form of premium tax credits to offset the costs of their coverage. Premium tax credits, which are designed to reduce an eligible individual’s premium costs, may be either paid in advance on a monthly basis to an enrollee’s issuer (referred to as advance premium tax credits) or received after the individual files federal income taxes for the prior year. Some state governments have elected to cover the remaining balance of some individuals’ exchange plans, leveraging a combination of advance premium tax credits and state funds to fully cover health insurance premiums on certain exchange plans. For example, Hawaii created the Health Care Premium Assistance Program, a special state program that covers the cost of premiums on eligible plans for qualified residents who do not qualify for Medicaid. While Hawaii’s program was not created specifically in response to compact migration and is not limited to compact migrants, most of its enrollees are compact migrants, according to Hawaii government officials. Since its launch in 2015, the program pays the balance of health insurance premiums not covered by advance premium tax credits for those who would otherwise be qualified for federal Medicaid if not for their citizenship status, including compact migrants. According to state officials, the program covered 3,223 compact migrants residing in Hawaii as of June 2017. Oregon and Washington developed premium assistance programs specific to compact migrants that leverage advance premium tax credits to eliminate health care premium costs. In addition to covering premiums, these programs provide for out-of-pocket health care costs, according to the programs’ websites and state officials. Oregon COFA Premium Assistance Program. The Oregon COFA Premium Assistance Program was launched in 2017, expressly to help compact migrants gain access to health care. In Oregon, participants pay for out-of-pocket costs at the time of service and subsequently apply to the program for reimbursement. Oregon’s program covered 780 compact migrants as of October 2019, according to state officials. The officials estimated that this program leverages $9 of federal funds through advance premium tax credits for every $1 of Oregon state funds contributed. Washington COFA Islander Health Care. The Washington COFA Islander Health Care program was launched in 2019, expressly to help compact migrants gain access to health care, and was based in part on the Oregon program, according to state officials. Washington will also cover dental insurance costs for compact migrants beginning in 2021, according to the program’s website and state officials. Participants in Washington’s program receive a payment card with preloaded funds to use for out-of-pocket costs. The program covered approximately 1,100 compact migrants in 2019, according to state officials. (Fig. 6 shows an example of an advertisement for Washington’s program, presenting information in six languages spoken by compact migrants.) Additional Federal Health Care Funding in U.S. Territories All U.S. territories, including Guam and the CNMI, receive federal funding through Medicaid, which is subject to an annual cap. Section 2005 of the PPACA, as amended, increased the funding caps for the territories for the period beginning on July 1, 2011, and ending on September 30, 2019, and provided a total of $6.3 billion in additional federal funding for health care to the territories. Guam and the CNMI have used some of this funding, in addition to other federal funding for health care, to partially support compact migrants’ health care costs or to alleviate the burden on programs that cover compact migrants. Guam. PPACA Section 2005 funding partly alleviated the financial shortfall of Guam’s Medically Indigent Program, according to a territory official. The Medically Indigent Program pays for health care costs of primarily non-U.S. citizens living in Guam, including compact migrants, who do not have other health insurance. Most compact migrants in Guam qualify for this program after meeting the 6-month residency requirement, according to Guam officials. In fiscal year 2019, compact migrants participating in the program numbered 8,616, according to Guam officials, and made up 73 percent of the program’s total participation. The officials said that the program is also funded through Guam local appropriations and federal Medicaid Undocumented Emergency Services funding. CNMI. Territorial hospital officials said that PPACA Section 2005 funding available in fiscal years 2011 through 2019 partially covered patient care costs in excess of the territory’s annual Medicaid cap, including care for compact migrants. The CNMI Medicaid program uses federal Disaster Relief Assistance funding to reimburse the hospital for emergency services provided to compact migrants, according to CNMI officials. Other Health Care Services Available to Compact Migrants Non-U.S. citizens, including compact migrants, may access health care through the U.S. Department of Health and Human Services Health Resources and Services Administration’s Health Center Program and through state government–supported clinics. The Health Center Program was established in the mid-1960s to help low-income individuals gain access to health care services. Health centers are responsible for delivering affordable, accessible, high-quality, comprehensive primary health care regardless of recipients’ ability to pay, according to Department of Health and Human Services officials. Figure 7 shows the entrance to Kokua Kalihi Valley, a federally qualified health center in Honolulu that estimates one-third of its patient population to be compact migrants, mostly from Micronesia. State clinics provide health services such as screening and treatment of certain infectious diseases to compact migrants, among other state residents. For example, the Arkansas Department of Health established the Dr. Joseph Bates Outreach Clinic to provide public health services to Marshallese in the region. As of September 2019, approximately 95 percent of the clinic’s patients were Marshallese, according to clinic officials. In addition, the University of Arkansas for Medical Sciences Northwest Campus facilitates research and community health programs in the Marshallese community and has established a clinic focused on diabetes. Compact Migration Has Other Budgetary Effects The budgetary effects of compact migration in the U.S. areas we visited include contributions by compact migrants, such as payment of federal and state taxes and fees, and also include several types of government program costs related to compact migration. Budgetary contributions. Compact migrants pay payroll taxes, including income taxes, and contributions to Social Security and Medicare. They also pay fees associated with state or territorial documentation or licensing, including driver’s licenses. In general, reliable data on budgetary contributions of compact migrants are not available, because state and territorial tax filings and related databases do not provide data on citizenship or ethnicity, according to state and territorial officials. However, the Hawaiian government reported that in 2017, compact migrants generated an estimated $36.6 million in state revenue from fees and taxes, such as the individual income tax, general excise taxes, and taxes generated from state government spending. According to University of Guam officials and an FAS community member, the presence of FAS communities may have helped Guam institutions obtain funding, including funding for research. Budgetary costs. State and territorial officials identified budgetary costs related to compact migration. For example, officials cited costs of providing translators or interpreters for government programs and costs associated with compact migrant interactions with police and the justice system. Some states have elected to extend state-level programs for food or cash-based assistance to compact migrants who are ineligible for the federal equivalents. For example, Washington’s Cash Assistance and Food Assistance Programs provide financial support to FAS citizens who are ineligible for the federal Supplemental Nutrition Assistance Program and Temporary Assistance for Needy Families. In Guam, some compact migrants qualify for the federal earned income tax credit, according to officials of Guam’s Department of Revenue and Taxation. The officials noted that because Guam’s tax system mirrors the federal system, any earned income tax credit paid in Guam is an expense to the territorial government. Compact Migrants Contribute to Workforces and Face Reported Challenges Compact migrants are eligible to work in U.S. areas and have contributed to the workforces of receiving communities, holding jobs in a range of industries. According to stakeholders we interviewed, compact migrants have encountered challenges while participating in the workforce. Workforce Contributions by Compact Migrants In the U.S. areas where they reside, compact migrants participate in the local economies in part by serving in the workforce in a variety of fields, including manufacturing, service industries, and professional industries, according to stakeholders we interviewed. See table 7 for examples. The following describes compact migrants’ participation in the areas we visited. Arkansas. Arkansas private sector representatives described Marshallese workers as essential to poultry plant operations, comprising one-quarter to one-third of some plants’ workers. At one such plant, most Marshallese employed are line workers on the floor of the plant, while others work as trainers and translators. Other compact migrants in Arkansas work at an airport; in hotels; in retail; or as caregivers, including in adult day care, according to FAS consular officials and nonprofit representatives. CNMI. CNMI officials and a private sector representative described compact migrants as a valuable resource in supplementing the CNMI’s small labor pool. Officials also noted that without compact migrants, businesses would have to recruit more foreign labor and face more-severe hiring challenges than they do now. Officials and a private sector representative stated that several businesses and franchises were founded by, and employ, compact migrants. Guam. Guam Chamber of Commerce representatives indicated that compact migrant workers would not be easily replaced if they were no longer eligible to work in Guam and that hiring other foreign workers in Guam involves difficult visa processes. Compact migrants tend to hold entry-level and low-skill jobs in Guam and have high turnover rates, according to representatives from one company. Several businesses in Guam were founded by, or cater to, compact migrants, according to private sector representatives. Hawaii. Micronesian officials noted that established communities of compact migrants in Hawaii help other FAS citizens to migrate, network, and find job opportunities. FAS community members in Hawaii identified multiple local businesses that either are owned by compact migrants or employ a large number of compact migrants. Oregon. In Oregon, some compact migrants work as caregivers or in a plant manufacturing reusable plastic containers for food storage and transport, according to an FAS official and community members. Oregon state government officials noted that compact migrants play an important role in working with adults and children with intellectual and developmental disabilities and in other paid caregiver capacities. The Governor of Oregon noted that compact migrants bring a tremendous amount of value to Oregon communities as educators, social workers, caregivers, and as members of the U.S. military. Other jobs or industries in which compact migrants work include warehousing, fast-food restaurants, and airport jobs, according to FAS officials. Washington. Some compact migrants work in caregiving, including at senior care homes; in manufacturing, warehousing, fast-food restaurants, or nonprofits; as artisans; or at airports, according to state and FAS officials and FAS community members. Workforce Challenges Faced by Compact Migrants Stakeholders reported that compact migrants have encountered various challenges related to participation in the U.S. workforce. See appendix VII for additional challenges experienced by compact migrant communities. Form I-94. Compact migrants from Micronesia and the Marshall Islands may present an unexpired FAS passport and Form I-94 Arrival/Departure Record (known as Form I-94) to employers to demonstrate their identity and employment authorizations. Before 2013, compact migrants entering the United States received a paper copy of the form to document their legal entry and their ability to legally reside indefinitely in the United States. The DHS transition in 2013 from issuing Forms I-94 on paper to issuing them electronically created challenges for compact migrants, according to FAS community members. According to consulate officials, communities were not adequately notified that DHS would maintain these records in publicly accessible databases for only 5 years. As a result, some compact migrants who entered the United States after mid-2013 did not download their Forms I-94 before they became unavailable and thus did not have a Form I-94 to show to employers, according to stakeholders we interviewed. REAL ID–compliant driver’s licenses. Some employers require employees to have REAL ID–compliant driver’s licenses, according to FAS officials and community members. Before September 2019, DHS required compact migrants and other nonimmigrants applying for a REAL ID–compliant driver’s license to present an unexpired passport with an unexpired visa and Form I-94 or to present an employment authorization document. However, because compact migrants do not receive a visa and are not otherwise required to obtain an employment authorization document, they were unable to obtain the licenses. In September 2019, DHS changed its requirements specifically to allow compact migrants to receive REAL ID–compliant driver’s licenses by presenting an unexpired passport and Form I- 94. Some compact migrants in Guam said that challenges related to REAL ID before the DHS regulation change had negatively affected their employment because some military base jobs required these documents for employment or for base access. In addition, some compact migrants have lost jobs at airports because of difficulty in obtaining REAL ID–compliant identification, according to Marshallese embassy officials. See appendix X for information about legislative and DHS policy changes that affected compact migrants’ ability to access full-term REAL ID–compliant driver’s licenses and identification cards. Commercial driver’s licenses. Various stakeholders discussed difficulties that compact migrants had encountered in obtaining commercial driver’s licenses required by certain jobs and obtaining standard driver’s licenses that are compliant with REAL ID requirements in some states. Marshallese officials said that compact migrants’ inability to obtain or renew commercial driver’s licenses had prevented them from being able to work in related jobs, such as truck driving. Labor abuse and discrimination. In September 2019, the government of Micronesia requested that the Department of State provide assistance to investigate abuse and mistreatment of Micronesian citizens who were recruited to move to the United States to work for a U.S. company in Iowa. In addition, compact migrants in Hawaii, Guam, and Oregon told us that they had faced workplace discrimination or were seen as harming the local economy. For example, compact migrants in Guam said that they had experienced discrimination in hiring and pay and sometimes were made to feel like a burden on the community. Additionally, a March 2019 report by the Hawaii Advisory Committee to the U.S. Commission on Civil Rights concluded that some compact migrants find it difficult to report workplace discrimination because they are concerned about retaliation from employers. The report found, among other things, that compact migrants face discrimination in access to employment and housing and also face widespread negative public perception in Hawaii. Compact Migration Has Societal Effects Stakeholders expressed some concerns about compact migration with respect to public health and law enforcement interactions. In addition to participating in the workforce, compact migrants participate in social institutions and create diversity and cultural exchange in their receiving communities. Public Health State and territorial health department officials and health care providers in the U.S. areas we visited noted concerns about the prevalence of communicable diseases such as tuberculosis and Hansen’s disease in compact migrant communities. Tuberculosis. State and territorial health departments have worked to identify and treat cases of active and latent tuberculosis in compact migrant communities. About 15 to 20 percent of active, communicable tuberculosis cases in Hawaii have occurred in the FAS community, including several cases of antimicrobial drug–resistant variants of tuberculosis, according to Hawaii government officials. In 2019, 23 communicable tuberculosis cases were diagnosed in compact migrants in Hawaii. In Arkansas, public health officials estimated that they had screened 30 percent of the Marshallese population since 2000 and reported 202 active cases and 500 cases of latent tuberculosis infection between 1997 and 2019. Arkansas officials also said that they screened 1,728 Marshallese and reported five cases of active disease and 95 cases of tuberculosis infection in fiscal years 2018 and 2019. In 2017 and 2018, Arkansas officials traveled to the Marshall Islands to conduct screening for active and latent tuberculosis in addition to diabetes and Hansen’s disease. Hansen’s disease. Hansen’s disease affects some members of compact migrant communities, according to health care providers and state government officials. For example, the Hawaii Department of Health has a registry of 281 patients who are on active treatment or monitoring for recurrence of Hansen’s disease or complications from the disease. The department manages 10 to 20 new cases of Hansen’s disease each year. According to Hawaii public health officials, 95 percent of the individuals diagnosed with Hansen’s disease in the state were from the Micronesian or Marshallese communities. From 2003 to 2019, the Arkansas Department of Health reported that 54 individuals, including 42 compact migrants, had been diagnosed with Hansen’s disease. Public Order and Law Enforcement Interactions Some stakeholders reported concerns regarding public order and law enforcement interactions with compact migrants in Guam, Hawaii, and Washington. Guam. Guam law enforcement agencies report on crimes committed by, or attributed to, FAS groups in each location. Guam private sector representatives we interviewed expressed a belief that social tension with the FAS communities was driven in part by some compact migrants’ public drunkenness or violence. In addition, language barriers can hinder compact migrants’ social integration into receiving communities, according to Guam law enforcement officials. Hawaii. Common offenses for which compact migrants are cited or arrested in Hawaii include quality-of-life or social-order offenses, such as trespassing, disorderly conduct, drinking in public or driving under the influence of alcohol, assault, or harassment, according to state officials. These interactions with the public or with law enforcement officials may contribute to a strained relationship between compact migrants and receiving communities. Hawaii officials estimated that 20 to 25 percent of the population using the state’s homeless services self-identify as part of the FAS community. Compact migrants may sleep in public parks, which can lead to legal charges. A lack of affordable housing may be a cause for homelessness among FAS communities. Washington. Marshallese embassy officials cited sporadic problems with gang activity and drug use among some younger Marshallese community members, particularly those living in Washington. These officials suggested that some migrant children who feel bullied or pressured may band together, resulting in a negative or gang-like situation. Community and Volunteer Work In some U.S. areas we visited, stakeholders we interviewed said that compact migrants seek to contribute to, or engage with, their surrounding U.S. communities through volunteer work, including church activities, environmental work, and other efforts. For example, FAS communities described participating in environmental cleanup efforts, including efforts to control invasive species and leveraging their agricultural knowledge to help Hawaiian farmers grow a more resilient variety of taro. Several community representatives in multiple states noted that some compact migrants spend a significant amount of time supporting their fellow community members as translators or interpreters or volunteering to help others navigate complex systems in U.S. areas. FAS citizens also serve in the U.S. military. The FAS countries have a high rate of military service, according to FAS officials and State documentation. Increased Diversity Stakeholders in some U.S. areas we visited described compact migrant populations as contributing to the diversity of receiving communities and educational institutions. For example, University of Guam officials said that FAS student association groups sponsor cultural events and activities that help to define the character of the university. The officials also noted that FAS students contribute to research portfolios and bring FAS government and community perspectives to classroom discussions. The officials observed that the presence of compact migrants increases the university community’s diversity and its cultural awareness and competency. In Arkansas, Marshallese community members said that they had helped to teach local U.S. residents about Marshallese culture and history not otherwise taught in U.S. schools. Marshallese community members in Arkansas also expressed a belief that the community brought a greater emphasis on family and respect for elders to the region. Agency Comments, Third-Party Views, and Our Evaluation We provided a draft of this report for review and comment to the Departments of Agriculture, Commerce, Health and Human Services, Homeland Security, Housing and Urban Development, the Interior, State, and Transportation; the Social Security Administration; the Governors of Hawaii, Guam, the CNMI, Arkansas, Oregon, and Washington; and the Ambassadors of Micronesia, the Marshall Islands, and Palau to the United States. The Departments of Agriculture, Health and Human Services, Homeland Security, and State and the Social Security Administration provided technical comments, which we incorporated as appropriate. The Departments of Commerce, Housing and Urban Development, the Interior, and Transportation did not provide comments. U.S. areas and the FAS Ambassadors provided written comments that we have summarized below and reproduced in appendixes XI through XIX, and responded to their comments, where appropriate, at the end of those appendixes. Comments from U.S. Areas Hawaii. The government of Hawaii commented that the health and economic impacts of the coronavirus pandemic, in addition to Hawaii’s high cost of living and public charge concerns, affect the state’s compact migrant communities in particular. The government also observed that racial disparities and other determinants of health and well-being are exacerbated for compact migrants. Noting that compact migrants lack access to Medicaid and the Supplemental Nutrition Assistance Program, the government urged that compact migrants’ access to health care and food nutrition programs be treated as a federal priority. Guam. The government of Guam advocated, in the context of the coronavirus pandemic, for restoring debt relief provisions associated with compact migration to offset unreimbursed compact expenses. The government noted that from the time the compacts went into effect until 1996, FAS citizens maintained access to federal health coverage. The government also commented that in 2017, Interior’s Office of Insular Affairs reported to Congress that restoring this eligibility would be in line with Congress’ intent to never cause adverse consequences to the territories and Hawaii. In addition, the government observed that moving compact migrants from Guam’s locally funded Medically Indigent Program to Medicaid would help Guam provide government services to all residents who need them. The government of Guam noted that the ongoing absence of an agreed definition for compact migrant for the purposes of data collection creates confusion. CNMI. The government of the CNMI commented on the importance of compact migrants’ contributions to the territory’s workforce needs and noted that they enrich the cultural makeup of the CNMI. Separately, the government stated that the response to the enumeration error discovered by the U.S. Census Bureau that led to a misallocation of compact impact funds has penalized the territories. According to Interior’s modified plan, future allocations to the CNMI (in addition to Guam and American Samoa) would be adjusted downward to account for past overpayment. The CNMI commented that reducing the future amounts of compact impact funds because of an error of the federal government does not recognize the present needs of the CNMI. The CNMI government also noted that the territories receive less data collection support from the American Community Survey, the U.S. Bureau of Economic Analysis, and the U.S. Bureau of Labor Statistics than other U.S. areas receive. Arkansas. The government of Arkansas commented that it considered the Census Bureau data in our report to underestimate the compact migrant population in Arkansas, and it cited several higher estimates. The government noted that the state does not receive compact funding, despite its high population of Marshallese, because it is not an affected jurisdiction as defined in the Compacts of Free Association Amendments Act of 2003. The government projected that approximately 12,000 compact migrants reside in Arkansas and estimated its annual costs related to compact migration at about $72 million. We believe that the Census Bureau data are sufficiently reliable for our purposes of estimating the number of compact migrants in U.S. areas. However, our report includes a discussion of stakeholder concerns that the compact migrant population in Arkansas may have been undercounted. We reported that the Census Bureau had estimated the compact migrant population in Arkansas during the period 2013 to 2017 at 5,895 on the basis of the definition of “compact migrants” used for its enumerations—citizens of Micronesia, the Marshall Islands, and Palau who entered the United States after 1986 (from Micronesia and the Marshall Islands) or 1994 (from Palau) and their U.S.-born children (biological, adopted, and step-) and grandchildren younger than 18 years. Oregon. The government of Oregon advocated for more reporting on the effects that U.S. military access to, and U.S. testing of 67 nuclear weapons in, the Marshall Islands has had on compact migration, citing the devastating impact of nuclear fallout on inhabitants’ health and the environment. The government of Oregon also cited a need to report on compact migrants’ positive contributions to receiving areas. Our report provides qualitative descriptions of compact migrants’ contributions, including budgetary, workforce, and social contributions, and also provides high-level data on estimated mean and median incomes among compact migrants. We have incorporated additional statements by the government of Oregon about compact migrant contributions in our report. The government noted that it has taken steps at the state level to provide health care access to compact migrant populations while also urging Congress to restore this populations’ access to federal programs such as Medicaid and Temporary Assistance for Needy Families. Further, the government called on Interior to expand the definition of “affected jurisdiction” and appropriate grant funds equitably. As we note in our report, this definition and the associated grant funding were established by Congress in the amended compacts’ implementing legislation. Washington. The government of Washington commented that our report did not provide a detailed history of U.S. military nuclear testing in the FASs and subsequent impacts on them and their citizens. The government noted that such information is necessary to explain FAS citizens’ current challenges and why additional resources are required to meet their needs. Further, the government commented that our report omits the personal narratives that are critical to a holistic account of the FAS experience in the United States, including the struggles many compact migrants face. Our report incorporates information that we obtained through our interviews with members of compact migrant communities, including those in Washington, such as reasons for migration, workforce and other challenges they faced, and stakeholder suggestions for improving experiences or outcomes of compact migration (see app. VII). The government of Washington stated that it hoped our report would prompt the federal government to make additional resources available to U.S. areas with sizeable compact migrant populations, and it called for inclusion of Washington among affected jurisdictions receiving compact impact grant funding. Comments from Freely Associated States Micronesia. The Embassy of the Federated States of Micronesia emphasized the importance of quantifying not only costs but also economic benefits of compact migration, including job creation, taxes paid, and community contributions. The embassy also called for guidelines and enumeration methods that better capture actual costs and revenue. The embassy noted the relationship between FAS citizens’ ineligibility for federal programs such as Medicaid and the costs borne by local governments and communities in the absence of these federal programs. According to the embassy, the continuing challenge of Micronesian citizens’ ineligibility for Medicaid since 1996, compounded by the effects of relevant social determinants of health, make their successful integration in U.S. areas more difficult. Noting that these circumstances have a direct effect on Micronesian migrants’ ability to contribute positively in receiving areas and become less reliant on public assistance programs, the embassy expressed support for the restoration of FAS citizens’ eligibility for Medicaid and for expanded veterans’ health care in Micronesia. The embassy commented that compact impact grant funding is a domestic issue and that discussions related to this issue should not diminish the priority of ongoing U.S. assistance to Micronesia under the compact. The embassy also raised concerns about challenges facing compact migrants, including the challenges described in our report. Marshall Islands. The Embassy of the Republic of the Marshall Islands described the migration rights provided in the compact as fundamental and essential to its country’s relationship with the United States. Additionally, the embassy observed that restoring Medicaid eligibility for its citizens living and working in the United States would greatly benefit its citizens and substantially reduce impact costs to certain areas. The embassy noted that, although Marshall Islands citizens living in the United States are eligible to purchase individual market health insurance plans through exchanges established under the Patient Protection and Affordable Care Act, many who are employed lack access to affordable health care because of the limited insurance benefits offered by most service industries or the high cost of covering family members. Furthermore, the embassy called for an objective accounting of revenue received from compact migrants and depiction of their contributions to, for example, the health and food security of the United States through employment in the food processing industry and other essential work. Last, the embassy commented that the addition of Marshallese workers to the 2020 census effort may remedy the potential undercounting of its citizens in the previous census. Palau. The Embassy of the Republic of Palau observed that it would be helpful to know the number of compact migrants from each FAS country who are able to access the federal programs for which they are eligible. This question was outside the scope of our review. Further, the embassy commented that it would like the U.S. federal government to inform and educate state departments of motor vehicles regarding the special status of FAS citizens in the United States, and it highlighted the difficulties that compact migrants historically have faced in obtaining REAL ID–compliant identification. We are sending copies of this report to the appropriate congressional committees and to the Departments of Agriculture, Commerce, Health and Human Services, Homeland Security, Housing and Urban Development, the Interior, State, and Transportation; the Social Security Administration; and the Governors of Arkansas, the CNMI, Guam, Hawaii, Oregon, and Washington; and the Ambassadors of Micronesia, the Marshall Islands, and Palau. In addition, the report is available at no charge on the GAO website at https://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-3149 or gootnickd@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix XX. Appendix I: Objectives, Scope, and Methodology We were asked to review topics related to migration to U.S. areas from the freely associated states (FAS)—the Federated States of Micronesia (Micronesia), the Republic of the Marshall Islands (Marshall Islands), and the Republic of Palau (Palau)—under those countries’ compacts of free association with the United States. This report (1) presents estimates of compact migrant populations and describes recent trends in compact migration; (2) summarizes the reported costs related to compact migration (compact impact costs) in three affected jurisdictions—Hawaii, Guam, and the Commonwealth of the Northern Mariana Islands (CNMI); and (3) describes effects of compact migration on governments, workforces, and societies in these and other U.S. areas. Migration Data To present estimates of compact migrant populations in U.S. areas and describe recent trends in compact migration, we obtained special tabulations of data from the U.S. Census Bureau’s 2013-2017 American Community Survey (ACS) for the 50 U.S. states, the District of Columbia, and Puerto Rico. For Guam and the CNMI—U.S. territories that are not covered by the ACS—we used the Census Bureau’s revised 2018 enumeration of compact migrants in these areas. The special tabulations of ACS data and the Census Bureau’s 2013 and 2018 enumeration reports defined compact migrants as individuals residing in U.S. areas who were born in the FASs and entered the United States after 1986 (for Micronesia and the Marshall Islands) or 1994 (for Palau) and their U.S.-born children (biological, adopted, step-) and grandchildren younger than 18 years. We calculated percentage changes in states and territories that had more than 1,000 estimated compact migrants (or were designated as affected jurisdictions by the 2003 compacts’ implementing legislation) by comparing 2005-2009 ACS data and the 2008 enumeration with 2013-2017 ACS data and the revised 2018 enumeration. To identify and describe changes in Census Bureau methods and definitions for enumerating compact migrants over time (see app. VI), we reviewed the definitions of “compact migrant” in the bureau’s enumeration reports for tabulations before and including 2018. We also interviewed Census Bureau and Department of the Interior (Interior) officials. For example, we asked when and where grandchildren were counted among compact migrants younger than 18 years. To estimate net arrivals to U.S. areas by travelers with FAS passports (see app. III), we analyzed data from the Department of Homeland Security’s (DHS) Customs and Border Protection’s (CBP) Arrival and Departure Information System (ADIS). According to CBP officials, ADIS consolidates data from several DHS systems to create unique, person- centric travel records for all travelers regardless of citizenship. We calculated monthly FAS net arrivals to U.S. areas from 2017 through 2019 by using ADIS data that DHS provided, showing numbers of individuals with FAS-issued passports entering and exiting U.S. ports of entry each month during the period. To assess the reliability of ADIS data, we spoke with DHS officials to identify potential data reliability concerns and other limitations of ADIS. Officials said that any compact migrant who enters on an FAS passport and holds U.S. citizenship will be masked or not appear in the ADIS system. Officials also said that compact migrants who become U.S. citizens after arrival or are later discovered to be U.S. citizens are removed from the data; CBP officials believed these numbers to be small. We also conducted statistical checks for consistency and completeness of the ADIS data, including validating the ADIS data against publicly available passenger data from the U.S. Department of Transportation Air Carrier Statistics (TranStats) T-100 database for 2015 to 2019 (data for 2019 were partial). We used flight segment data from the T-100 database containing total passenger counts reported by both U.S. and foreign air carriers for flights that compact migrants take to U.S. areas. We found that data from ADIS and the T-100 database were positively correlated for 2015, 2017, 2018, and 2019 but were not correlated for 2016. According to CBP officials, ADIS was significantly changed in 2016 and may contain duplicate entries for that year. As a result, we determined that ADIS data for 2017 and later were sufficiently reliable for our intended use. Reported Costs Related to Compact Migration To quantify costs related to compact migration that were reported by the affected jurisdictions included in our review—Hawaii, Guam, and the CNMI—we reviewed documents that they had published or provided to Interior, such as compact impact reports submitted by Hawaii and Guam and grant documents submitted by the CNMI. We used the most recent data available for 2004 through 2018. To identify the amount of funding distributed by Interior as compact impact grants to the affected jurisdictions, we interviewed Interior officials and reviewed relevant documentation. Effects of Compact Migration on Governments, Workforces, and Societies To identify and describe effects of compact migration on governments, workforces, and societies of receiving U.S. areas, we reviewed relevant documentation and conducted interviews with stakeholders in six U.S. areas that we visited. Documentation that we reviewed included program information and counts of compact migrants using state-level benefits programs, treated by state or local health clinics, enrolled in public schools or higher-education institutions, or using interpreters. Because various sources may define compact migrants by ethnicity, place of birth, language of origin, or other metrics, we noted the definition used for each count in this report. To identify the eligibility of compact migrants for selected federal programs, we reviewed relevant statutes and regulations and held discussions with officials from the U.S. agencies that oversee the programs. We selected the programs included in table 1 on the basis of those we included in a prior report, and we added other selected programs that we learned about in the course of interviews for our current report. We traveled to, and interviewed stakeholders in, six U.S. states and territories where compact migrants live, including three of the U.S. areas designated in the 2003 amended compacts’ implementing legislation as affected jurisdictions—Hawaii, Guam, and the CNMI— and three mainland states—Arkansas, Oregon, and Washington. We selected these areas on the basis of previously reported compact migrant population distributions in U.S. areas and of the locations of consulates or honorary consuls established by Micronesia, the Marshall Islands, and Palau. Stakeholders we interviewed included officials from nine federal agencies; state and territorial government officials in areas we visited; private sector and nonprofit organization representatives such as chambers of commerce, employers of compact migrants, and nonprofit service providers; officials from the FAS embassies and consulates or honorary consuls in areas we visited; and compact migrants living in areas we visited (see table 8). FAS embassy officials in Washington, D.C., connected us with local community members who helped us promote and organize the local community meetings in areas we visited. Participants whom we interviewed in the meetings do not represent a generalizable sample of compact migrants, and the challenges they discussed are not comprehensive (see app. VII for a discussion of challenges faced by compact migrants, according to stakeholders we interviewed). To describe academic studies of workforce and fiscal impacts of new migrants (see app. IX), we conducted a search, using keywords relevant to the economic impact of migration, in American and European economics academic journals published during the period 2015 to 2019. We reviewed a subset of these articles that we deemed most relevant to the context of compact migration, including articles that related to migration of lower-skilled workers and that included empirical analysis of the impact of this migration on various economic aspects. We also reviewed survey articles reviewing the conclusions of prior relevant publications. We conducted this performance audit from March 2019 through June 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Estimates of Compact Migrants in U.S. Areas Table 9 presents Census Bureau data for U.S. states and territories that had estimated compact migrant populations of more than 1,000 in 2013 through 2018 and shows percentage changes in these populations from 2005-2009 to 2013-2018. Data for U.S. areas not covered by the American Community Survey, including Guam, the Commonwealth of the Northern Mariana Islands (CNMI), and American Samoa, are from compact migration enumerations that the Census Bureau performed on behalf of the Department of the Interior. According to 5-year data from the Census Bureau’s 2013-2017 American Community Survey, 72,965 compact migrants resided in the 50 U.S. states, the District of Columbia, and Puerto Rico. (The American Community Survey does not cover American Samoa, the CNMI, Guam, or the U.S. Virgin Islands.) For estimates of the number of compact migrants in each of the 50 U.S. states, the District of Columbia, and Puerto Rico, see table 10. The American Community Survey captures, among other things, respondents’ place of birth (by country) and state of residence. Table 11 provides Census Bureau estimates, using 2013-2017 American Community Survey data, of the numbers of compact migrants born in the freely associated states—the Federated States of Micronesia, the Republic of the Marshall Islands, and the Republic of Palau—and residing in each of the 50 states, the District of Columbia, and Puerto Rico. Appendix III: Federal Travel Data Showing Compact Migration to U.S. Areas From 2017 through 2019, an average of about 366 more migrants from the Federated States of Micronesia (Micronesia), Republic of the Marshall Islands (Marshall Islands), and Republic of Palau (Palau) arrived in U.S. areas per month (4,390 per year) than departed, according to the Department of Homeland Security’s Customs and Border Protection’s Arrival and Departure Information System (ADIS). As figure 8 shows, this trend was driven by migrants from Micronesia and the Marshall Islands (3,343 and 1,487 per year on average, respectively). Each year during this period, an average of about 440 more Palauan citizens departed from the United States than arrived. Appendix IV: Demographics and Characteristics of Compact Migrants in the 50 U.S. States, the District of Columbia, and Puerto Rico The Census Bureau’s American Community Survey is an ongoing survey that provides information on a yearly basis, including employment status, educational attainment, veteran status, and age of survey respondents, among other topics. The survey covers the 50 U.S. states, the District of Columbia, and Puerto Rico. (The survey does not cover American Samoa, the Commonwealth of the Northern Mariana Islands, Guam, or the U.S. Virgin Islands.) Some data for compact migrant populations are available through the American Community Survey. See table 12 for demographic information about compact migrant populations in the 50 U.S. states, the District of Columbia, and Puerto Rico in 2013 to 2017. See table 13 for demographic information about the compact migrant population in Hawaii only. Appendix V: Compact Impact Costs Reported by Hawaii, Guam, and the Commonwealth of the Northern Mariana Islands Since 1986, Hawaii, Guam, and the Commonwealth of the Northern Mariana Islands (CNMI) have submitted to the Department of the Interior (Interior) intermittent compact impact reports and other documents that include descriptions of, and estimated costs for, education, health, public safety, and social services that local government agencies provided to compact migrants. Hawaii and Guam have submitted compact impact reports, which are available on Interior’s Office of Insular Affairs’ website. The CNMI has not submitted a compact impact report since fiscal year 2003 but reports compact impact costs to Interior in the CNMI’s annual plan for the use of compact impact grants. Table 14 shows the estimated costs that these affected jurisdictions reported to Interior or provided to us for 1986 through 2018. Appendix VI: Compact Migrant Enumeration Methods, Definitions, and Error The Census Bureau, working under an interagency agreement with the Department of the Interior (Interior), has conducted six sets of enumerations of compact migrants in affected jurisdictions for the purpose of allocating compact impact grant funding and has performed the enumerations every 5 years. Enumeration methods and definitions have changed over time. During the course of our work, an error was discovered that affected the accuracy of the 2013 and 2018 enumerations and also affected Interior’s allocations of compact impact grants for several fiscal years. Compact Migrant Enumeration Methods Census Bureau methods of gathering new data or analyzing existing data for compact migrant enumerations on behalf of Interior have changed over time. In 1993,1998, and 2003, the bureau used the “snowball” technique; in 2008, 2013, and 2018, the bureau employed a two-pronged approach. For enumerations in 1993, 1998, and 2003, the Census Bureau employed a survey method known as snowball sampling to count compact migrants in Hawaii, Guam, and the Commonwealth of the Northern Mariana Islands (CNMI). Because the surveys relied on referrals by respondents to identify new respondents, they were likely to undercount compact migrants who were not referred. For the enumerations in 2008, 2013, and 2018, the Census Bureau used a combination of existing and new survey data to count or estimate the numbers of compact migrants in Hawaii, Guam, and the CNMI. However, for the enumerations in Hawaii, the bureau used a different approach than it used for the enumerations in Guam and the CNMI. Also, for the enumerations in Guam and the CNMI, the bureau used a different approach in 2013 than it used in 2008 and 2018. To estimate the number of compact migrants in Hawaii, the Census Bureau used existing American Community Survey data. To estimate the numbers of compact migrants in Guam and the CNMI, the bureau used existing decennial census data in 2013 and gathered new survey data in 2008 and 2018. See table 15 for a summary of the enumeration methods that the Census Bureau has used over time. Definitions of “Compact Migrant” Used in Enumerations The definition of “compact migrant” that the Census Bureau used for the enumerations has changed over time. Each enumeration has counted as a compact migrant any individual, of any age, who was born in the Federated States of Micronesia (Micronesia), the Republic of the Marshall Islands (Marshall Islands), or the Republic of Palau (Palau) and who entered the United States after the effective date of their country’s compact. However, the enumerations in 2003, in 2008, and in 2013 and 2018 used various criteria for counting U.S.-born (U.S. citizen) individuals as children of compact migrants and therefore as compact migrants. 2003 enumeration. The definition of “compact migrant” in the 2003 amended compacts’ implementing legislation indicates that the children of compact migrants were to be considered compact migrants until 18 years of age. Interior interpreted the legislation’s definition of “compact migrant” as including all children younger than 18 years who were born to a compact migrant or migrants in the United States, thus including some U.S. citizens. 2008 enumeration. For the 2008 enumeration, the Census Bureau, on behalf of Interior, counted as compact migrants all children (biological, adopted, and step-) younger than 18 years who were born in the United States to a compact migrant head of household or to his or her spouse, were adopted by a compact migrant head of household or by his or her spouse, or were stepchildren of a compact migrant head of household or of his or her spouse. 2013 and 2018 enumerations. Starting with the 2013 enumeration, the Census Bureau also began counting as compact migrants all children (biological, adopted, and step-) younger than 18 years who were born in the United States to a compact migrant or to his or her spouse, regardless of whether they were the children of the head of household or of his or her spouse, and all grandchildren of a compact migrant who were younger than 18 years, regardless of whether they were the grandchildren of the head of household or of his or her spouse. Also starting with the 2013 enumeration, the Census Bureau introduced a requirement that to be counted as a compact migrant, a child or grandchild of a compact migrant must never have been married. For a summary of “compact migrant” definitions used for the enumerations over time, see table 16. Compact Migrant Enumeration Results, 1993-2018 The six sets of enumerations of compact migrants that the Census Bureau conducted on behalf of Interior in affected jurisdictions from 1993 through 2018 showed these populations growing in Hawaii and Guam and fluctuating in the CNMI. During our work with the Census Bureau to obtain American Community Survey data related to compact migrant populations, bureau officials discovered a programming error in the 2013 and 2018 enumerations of compact migrants that had resulted in an underestimate of certain compact migrants in Hawaii. The bureau revised these estimates in October 2019 to correct for the error in Hawaii. Figure 9 shows the Census Bureau’s revised estimates of compact migrants in the affected jurisdictions as of October 2019. In February 2020, Interior requested that the Census Bureau further revise its estimates for 2013 and 2018 to no longer count grandchildren. As of March 2020, the results of this revision were not yet available. 2013 and 2018 Enumeration Error Appendix VII: Stakeholder Suggestions to Address Challenges Related to Compact Migration In the U.S. areas we visited, stakeholders from state and territorial governments, private sector and nonprofit organizations, and freely associated state (FAS) consulates and communities made suggestions for improving experiences or outcomes of compact migration for both the receiving areas and the migrants themselves. Stakeholders recommended that some actions be taken in both the United States after compact migrants’ arrival and in the FASs before the migrants’ departure. Federal Policies, Operations, and Funding Provide more information and education about the compacts. Several stakeholders said that U.S. agencies should better understand the compacts and coordinate their related work. These stakeholders, including members of compact migrant communities, noted that U.S. government officials in some cases have seemed uncertain or unaware that compact migrants are able to live and work in U.S. areas without a visa or other documentation and have asked them to present immigration documents they do not possess or are not required to obtain. An FAS official and community members noted a need for more education of employers and state government officials regarding the migration terms of the compacts and the migration status of FAS citizens in the United States. Restore Medicaid eligibility and expand benefits access. State government officials and health care providers advocated restoring Medicaid access to FAS populations. An FAS Consul General advocated restoring Medicaid eligibility to its pre-1996 status for compact migrants. FAS community members suggested extending Supplemental Nutrition Assistance Program benefits and expanding federal student loan access to compact migrants. Provide more information and guidelines about federal programs and policies. State government officials suggested that changes to federal government policies should include specific information about the applicability of the changes to FAS citizens. Health care providers suggested that the federal government should share more data about compact migration and noted a need for federally established guidelines to support accurate, rather than exaggerated, cost reporting. The providers noted that compact impact estimating was chaotic and had a negative effect on the community. FAS community members expressed interest in federally provided educational sessions and clear eligibility criteria for federal benefits. Simplify Form I-94 access for compact migrants. FAS consular officials and community members said that compact migrants entering the United States should receive information about the importance of their Form I-94 Arrival/Departure Record (Form I-94) and how to retrieve it online. Because compact migrants have had difficulty in accessing these forms, and given the cost of replacing them, FAS community members requested that federal agencies be enabled to retrieve migrants’ Forms I-94 for them. FAS consular officials recommended that compact migrants’ Forms I-94 be made accessible on the Customs and Border Protection website indefinitely, not only for the current 5-year period, since compact migrants’ forms do not expire. Provide more and broader funding to U.S. states and territories. State government officials, nonprofit representatives, and FAS community members said that more federal funding and resources were needed to accommodate the compact migrant population or to support the receiving states and territories. State government officials also said that the federal government should increase compact impact funds to a “reasonable amount,” even if the full costs cannot be covered. They noted that the compacts represent a federal obligation and expressed a belief that the federal government should take care of compact migrants. According to some health care providers, the United States’ treatment of the compact migrant population in U.S. areas could affect the FASs’ compact negotiations with the U.S. government. State government officials also suggested that allowing compact migrants access to more federal benefits would help alleviate compact impact on states and territories. Clarify immigration provisions under the compacts of free association. FAS community members in some locations we visited expressed a need for clarification about the status of migration provisions of the compacts. Specifically, they expressed concern that they might have to leave the United States in 2023. For example, in one FAS community we visited, community members registered confusion about whether provisions of the compacts (including migration provisions) are scheduled to end in 2023 and whether FAS citizens in U.S. areas can become U.S. citizens. One community member expressed concern that compact migrants would be “chased” out of U.S. areas after 2023 and that “all of their rights” under the compacts would be revoked. FAS community members also sought clarification about the implementation of the DHS rule for considering public charge while determining admissibility to U.S. areas. According to community members and other stakeholders, the rule has caused uncertainty in compact migrant communities, which may result in some compact migrants’ not enrolling in, or unenrolling from, public benefits programs. FAS community members said that they are uncertain whether and how the rule change will apply to them and whether enrolling in public benefits or enrolling eligible children will make them ineligible to reenter the United States. FAS consular officials and community members also suggested revising certain immigration provisions—for example, changing compact migrants’ nonimmigrant status to allow them access to a wider range of jobs, including law enforcement and military officer positions. Health Care Expand health care access and clinics in U.S. areas. State government officials said they believed that more health education and outreach to FAS communities were needed. A nonprofit representative noted that FAS communities lack vision care and that the extension of postpartum care to FAS communities would improve maternal and child health. FAS community members suggested the creation of a Pacific Islander–specific health clinic in the Pacific Northwest, with translators on staff and on-site enrollment for health insurance. Representatives of a nongovernmental organization in Hawaii that is led and staffed by compact migrants noted that a series of changes in compact migrants’ eligibility for the Hawaii state health care program, Med-QUEST, had caused confusion about compact migrants’ eligibility for public health care benefits. Address preventative care, dialysis needs, and communicable diseases in the FASs. State and territorial government officials and health care providers said that greater access to in-country care, including more resources for primary care, was needed in the FASs. They recommended making more preventative treatment available in the FASs, including diabetes prevention, and establishing clinics in the FASs to potentially reduce the number of individuals moving to the United States for health care. Health care providers suggested that the Department of the Interior (Interior) should produce or fund a study on dialysis in the FASs, including an analysis of whether high-quality dialysis services in the FASs would decrease migration solely for access to dialysis. Territorial government officials suggested that compact migrants should receive health screenings before departing for the United States to identify any serious conditions or communicable diseases. Some health care providers and state government officials proposed that the U.S. federal government focus on reducing or eliminating the transmission of tuberculosis in the FASs. Compact Migrant Orientation and Services Offer predeparture education to compact migrants in the FASs. State government officials and nonprofit representatives suggested that videos be aired on television in the FASs to support predeparture education, to explain differences they would find in the United States, and to reduce culture shock after arrival. Some state government officials and health care providers suggested that FAS citizens be encouraged to gather documentation, such as immunization and medical records, school records, and anything necessary to obtain a U.S. driver’s license, before departing for the United States. State government officials also suggested that lists of community-based organizations, by U.S. state or territory and city, be provided to FAS citizens before their departure. Offer orientation and information to compact migrants arriving in the United States. State government officials said that U.S. areas should offer and fund location-specific orientations for FAS citizens after arrival. The officials suggested that these orientations should cover how health care eligibility works, what resources are available to compact migrants, and how they can contact interpreters. State government officials also said that proactive education about U.S. laws could help compact migrants avoid behavior or circumstances that might cause them to run afoul of the law, given cultural differences and misunderstandings. Health care providers noted that compact migrants could be given more information to encourage better nutritional choices and more exercise. Expand and professionalize translation and interpretation resources. Compact migrants who are not fluent in English may experience challenges accessing or navigating health care, the judicial system, and educational institutions, according to state government officials, FAS consulate officials, private sector and nonprofit organization representatives, and compact migrant communities. State government officials reported frequent difficulty in finding interpreters and translators for the multiple languages spoken by compact migrants. State government officials recommended that grants be made available to help pay for interpreters until more FAS community members graduate from college and become qualified. The officials also said that interpreters should be encouraged to develop greater proficiency in fields such as law and medicine so that they can serve in multiple capacities. In addition, the officials identified a need for more in-person interpreters in hospitals and medical facilities. State government officials noted that FAS communities speak many different languages, and they acknowledged the need for a culturally-specific approach for each group. They said that, in addition to translating content, interpreters should fully explain the context of programs to ensure compact migrants’ understanding. FAS community members proposed the creation of a group of paid, full-time interpreters and a language certification requirement to guarantee the availability and quality of language services. Create “one-stop shops” with information and resources for compact migrants. State government officials and health care providers identified a need for one-stop shops—centers that serve compact migrant populations—in areas that do not currently have them. According to stakeholders in a U.S. state without such a center, a one-stop shop could reduce duplication and increase coordination among the many groups that serve the FAS community. Other stakeholders suggested that each state government establish a single point of contact for compact migrants. FAS community members and nonprofit representatives identified a need for a cultural center or other physical space that could be used to hold events and provide centralized communication and resources for the FAS community in the Pacific Northwest, in particular. Emphasize community-based approaches to supporting compact migrants. State government officials noted the importance of community- based approaches to supporting compact migrants. For example, stakeholders recommended hiring community health workers from the FAS population to engage with their communities in U.S. areas. According to the officials, community health workers, as known and trusted entities, are better sources of information for FAS communities than any government agency. The officials also acknowledged the importance of engaging with FAS community leaders (including embassy or consular officials and church leaders) in U.S. areas to successfully connect with FAS community members. Provide compact migrant–dedicated housing. State government officials, FAS consulate officials, and nonprofit organization representatives discussed discrimination that compact migrants experienced in housing. For example, stakeholders in some areas we visited described landlords who failed to maintain or repair housing leased to compact migrants, who targeted compact migrants for evictions, or who avoided renting units to compact migrants. Officials in one state suggested that FAS communities need access to dedicated housing options that align with their community traditions and cultural norms, such as units that can accommodate large or multiple families. Appendix VIII: Nonprofit and Private Sector Organizations Supporting Compact Migrants Nonprofit organizations provide compact migrants with a range of assistance, such as assistance with housing or rent, food, documentation and legal matters, and enrollment in health insurance. Some organizations, such as “one-stop shops” (i.e., centers serving compact migrant populations), serve only compact migrants, while other organizations serve compact migrants among other members of the receiving community. Additionally, some companies that employ compact migrants offer programs intended to help them adjust to life in the United States. The information presented in this appendix is based mainly on documentation provided by the organizations and interviews with their representatives. Nonprofit Organizations That Serve Compact Migrants Only Several nonprofit organizations in U.S. areas that we visited target their services to compact migrants. Two of these organizations—one-stop shops in Hawaii and Guam—aim to support the compact migrant communities by connecting the migrants to existing resources and, in some cases, creating new programs and services to support freely associated state (FAS) communities, according to nonprofit and government officials and documentation. These one-stop shops receive funding from the U.S. Department of the Interior (Interior) as well as other governmental and nongovernmental sources. In Guam, the Micronesian Resource Center One-Stop Shop was developed with input from various communities in Guam and government agencies and launched in October 2015, according to one-stop shop officials and Interior documentation. The one-stop shop has received an Interior grant each year starting in fiscal year 2016. The amount of the grant has steadily increased, rising from $210,000 in fiscal year 2016 to $217,000 in fiscal year 2017, $250,000 in fiscal year 2018, and $267,000 in fiscal year 2019, according to Interior documentation and officials. The one-stop shop employs both case workers and cultural mediators and uses a mobile van to bring services directly to FAS communities, according to one-stop shop officials. These services include outreach to communities, including youths; workshops for parenting and driving; and assistance with lost or replacement documentation. For example, when conflict escalated among compact migrants living in a Guam apartment complex, Interior and Guam officials noted that the one-stop shop worked with police to facilitate meetings and participation in neighborhood watch programs. The one-stop shop has hosted “Welcome to Guam” orientations to educate compact migrants about finding housing, setting up utilities, and opening a bank account in Guam; employees’ rights; medical insurance; deportable offenses; and the danger of human trafficking, according to one-stop shop officials. In Hawaii, the one-stop shop We Are Oceania was established with Interior funding in 2015. The organization provides case management, helping compact migrants to find jobs, address housing or legal issues, and enroll in health insurance through Hawaii’s Premium Assistance Program, according to one-stop shop officials and documentation. We Are Oceania has also provided cultural consultations and trainings to Hawaii public school teachers and service providers to educate them about cultural differences and potential challenges that compact migrants may face, according to nonprofit representatives and documentation. The officials also said that the organization hosts a youth summit and helped open a newcomer welcome center at a middle school. Figure 10 shows photos of the We Are Oceania facility, including desks where compact migrants can apply for health insurance and other services. Other nonprofit organizations were also founded specifically to assist the compact migrant community in navigating various U.S. systems, such as education and health care, and obtaining documentation such as driver’s licenses or Forms I-94 Arrival/Departure Records. The Arkansas Coalition of Marshallese in Springdale, Arkansas, according to representatives of the organization, helps local compact migrants with tasks such as retrieving new Forms I-94 from Customs and Border Protection; translating state driver’s license applications into Marshallese; providing education about diabetes prevention and management; and enrolling compact migrant children in ARKids, the state’s public health insurance program that extends federal health insurance coverage for children younger than 19 years. In 2018, the Micronesian Islander Community Organization in Oregon announced a study among local compact migrants to identify barriers that they faced in the region, such as a lack of certified health care interpreters. Additionally, the Oregon-based COFA Alliance National Network conducts policy and advocacy work aimed at supporting compact migrant communities, according to representatives of the organization. Nonprofit Organizations That Serve Compact Migrants Charities, legal services, and other programs assist compact migrants and other eligible individuals in selected U.S. areas. For example: In Hawaii, the Salvation Army of the Hawaiian and Pacific Islands provides assistance with rent, utilities, and food; interpreters to assist non–English speakers with accessing health and legal services; and digital literacy training (e.g., how to use email), according to Salvation Army officials. In Guam, the Salvation Army Guam Corps provides assistance with rent, utilities, food, and clothing and also provides case management services, according to Guam officials. In the Commonwealth of the Northern Mariana Islands (CNMI), Karidat provides a food pantry, clothing assistance, rental assistance, and victim advocacy, among other services. (Fig. 11 shows a public bulletin board and donated clothing in Karidat’s offices.) In 2018, compact migrants made up 20.4 percent of individuals accessing Karidat’s food pantry and 39.5 percent of individuals receiving clothing assistance, according to Karidat estimates. The Hawaii and Arkansas chapters of the Legal Aid Society provide legal services to local residents, such as victims of crime, according to Hawaii and Arkansas officials. According to Hawaii chapter officials, they served 569 compact migrants (8.5 percent of their total clients) in fiscal year 2019. The Asian Family Center within Oregon’s Immigrant and Refugee Community Organization provides similar services, including defense for parties engaged in deportation removal proceedings, according to representatives of the organization. Private Sector Organizations Some employers with compact migrant workers provide employee services, programs, or accommodations specific to these workers’ needs. In Arkansas, Tyson, Inc., provides written materials in Marshallese and operates a program that appoints chaplains to help the company’s Marshallese workers, as well as other non–U.S. citizen employees, navigate life in the United States generally and in Arkansas specifically, according to private sector representatives. Additionally, the representatives told us that the company provides free classes in financial literacy and English as a second language to its employees, including compact migrants. Another company in the region, Cargill Protein, has partnered with local nongovernmental organizations to educate its compact migrant employees about U.S. driving laws and help prepare them for driver’s license tests. Appendix IX: Review of Academic Studies of the Workforce Effects of Migration Similar to Compact Migration We examined academic studies published from 2015 through 2019 to determine what is known about the likely effects of migration similar to compact migration on the workforces of receiving countries. Because we were unable to identify articles published during this period that focused specifically on compact migration, we focused our search on studies examining the effects of migration by other groups with relatively few skills. Studies that we reviewed sometimes reached differing conclusions about whether migration is associated with a negative, neutral, or positive effect on the employment and earnings of nonmigrant workers in the receiving countries. Some studies found that migration may result in worsened employment prospects or wages—particularly in the short term and if the influx of migrants is sudden—for nonmigrant workers who are most similar to the migrants in terms of demographics and skills. If the migrant workers are close substitutes for nonmigrants, they may intensify competition for jobs, increasing unemployment and lowering wages for such nonmigrant workers as well as for similar prior migrants. In the case of compact migration, this might include younger and less educated nonmigrants. However, according to other studies and survey papers that we reviewed, nonmigrants, both low and high skill, could benefit as a whole from migration. For example, one study of the effects of migration on 20 countries found that both low- and high-skill nonmigrants clearly benefited from an influx of migrant workers about two-thirds of the time. Nonmigrant workers may benefit from migration if the migrant workers specialize in different skills and vocations than the nonmigrant population, leading to complementary effects from scale and specialization. For example, larger numbers of construction workers may result in greater efficiency and quality in the building of more restaurants and bars, benefitting workers in nonconstruction trades as well as nonmigrant investors and business owners. Institutions may play an important role in determining the effects of migration on the receiving country’s workforce. For example, a study estimating the effect of migrant workers in European Union countries and controlling for institutional and noninstitutional factors showed that the effect of migrants varied between countries, driven in part by differences in their institutional environment, such as the extent of unemployment insurance, fiscal redistribution, and government spending on services and public goods. This study found that, while fiscal redistribution to migrant workers through taxation and unemployment benefits somewhat worsens outcomes for nonmigrants, this effect is often outweighed by the economic contribution of these migrants. Distinctions in statistical methodologies and assumptions may explain studies’ seemingly contradictory conclusions about the effects of migration on the workforce of receiving countries. According to a survey paper reviewing other previously published work, the statistical controls selected for studies of the impact of migration can result in subtle but economically important distinctions in what the studies attempt to measure, such as the total effects of migration on a given region or the effects of migration on a specific group (e.g., a particular education or skill group). According to this and a second survey paper we reviewed, contradictory conclusions may also result from differing assumptions about factors such as the extent to which migrants “downskill” (compete for jobs for which they may be overqualified) and, therefore, about the nonmigrants that should be used as a comparison group to examine the effect of migrants of a particular skill and education level. According to a third survey paper we reviewed, studies also vary in whether they measure the shorter- or longer-term effects of migration; the survey found that negative effects are more often reported when studies measure migration’s shorter-term effects. Academic journal articles that we examined also discuss the potential fiscal effect of migration. Several studies argue that evaluations of migration’s fiscal effect should consider the potential effects over multiple generations and should also consider the indirect fiscal effect of migrants’ influence on native workers. For example, a panel discussion report of the National Academies of Sciences, Engineering, and Medicine states that descendants of immigrants are often studied only as children, in cross- sectional data providing a point-in-time snapshot. As a result, according to the report, the average immigrant household is counted as a net fiscal burden in part because young children of immigrants, like the children of natives, receive public education. The report stated that studying the descendants of immigrants as they complete their education, become workers, and start paying taxes provides a more complete measure of migration’s fiscal effect, because such an analysis may include not only the cost of their education but also the delayed fiscal benefits of that education: larger tax payments made possible by the investment in human capital that education represents. Another paper we reviewed argues that because migrant workers can positively influence the upward mobility of native workers, the higher taxable income from these native workers should be considered, in addition to the low taxable income of the migrants, to avoid negatively biasing the estimated fiscal effect of migrants. Appendix X: Compact Migrant Eligibility for, and Access to, REAL ID–Compliant Identification The REAL ID Act, passed by Congress in 2005, set minimum document requirements and issuance standards for driver’s licenses and personal identification cards. The act also prohibits federal agencies from accepting for certain purposes driver’s licenses and identification cards from states that do not meet the act’s minimum standards. Citizens of the freely associated states (FAS)—the Federated States of Micronesia, the Republic of the Marshall Islands, and the Republic of Palau—have always been eligible for REAL ID–compliant driver’s licenses or identification. However, the term of the licenses’ or identification’s validity and the documents that the Department of Homeland Security (DHS) required to establish compact migrants’ identity have varied over time. Currently, compact migrants are eligible for full-term REAL ID–compliant identification. Since September 2019, they have been required to present an unexpired FAS passport and the most recent Form I-94 Arrival/Departure Record (Form I-94) as evidence of identity to obtain the identification. Compact Migrant Eligibility for, and Access to, REAL ID–Compliant Identification DHS regulations previously required compact migrants to provide documents they were not required to have. Before September 2019, DHS required compact migrants applying for REAL ID–compliant identification to present, in addition to their unexpired foreign passport and Form I-94, either an unexpired U.S. visa (affixed to the passport) or an employment authorization document (EAD). However, under the compacts of free association, compact migrants are not required to obtain a visa or an EAD. On September 4, 2019, DHS modified its policy, designating an unexpired passport from one of the FAS countries, in combination with an individual’s most recent Form I-94, as acceptable evidence of identity that compact migrants may present to obtain REAL ID–compliant identification. Federal law previously restricted the term of the REAL ID–compliant identification that compact migrants could receive. Before December 2018, compact migrants were eligible to receive temporary, limited-term REAL ID–compliant identification, valid until the expiration date on their EAD, which could be a maximum of 5 years, according to DHS officials. In December 2018, the REAL ID Act Modification for Freely Associated States Act made FAS citizens eligible for full-term REAL ID–compliant identification. Since then, compact migrants have been eligible for full- term REAL ID–compliant identification, valid for the maximum number of years for any license or identification as set by individual U.S. states and territories, according to DHS officials. Compact Migrant Challenges Related to REAL ID–Compliant Identification In several areas that we visited, compact migrant communities described challenges they had experienced in obtaining or renewing their REAL ID– compliant identification. Some compact migrants spoke of difficulty due to the requirement to present a visa or an EAD as evidence of identity. In one location, FAS community members said that other members of the community had lost employment on a military base because they were unable to obtain REAL ID–compliant identification. (We heard many of these observations before September 2019, when DHS modified the policy that required applicants for REAL ID–compliant identification to present a visa or EAD.) Some compact migrants reported being unable to obtain REAL ID– compliant identification for other reasons. number on their current foreign passports. When an FAS citizen’s passport expires and he or she renews it while in the United States, the new passport has a different number than the former passport number displayed on the FAS citizen’s Form I-94. Appendix XI: Comments from the Government of Hawaii Appendix XII: Comments from the Government of Guam GAO Comments 1. Section 104(e)(9)(A) of the amended compacts’ enabling legislation authorized the President of the United States, at the request of the Governor of Guam or the Governor of the Commonwealth of the Northern Mariana Islands (CNMI), to reduce, release, or waive all or part of any amounts owed by the Guam or CNMI government (or either government’s autonomous agencies or instrumentalities), respectively, to any department, agency, independent agency, office, or instrumentality of the United States. According to section 104(e)(9)(B)(iv), that authority expired on February 28, 2005. 2. The Census Bureau data that we report reflect a definition of “compact migrants” that includes citizens of the Federated States of Micronesia (Micronesia), Republic of the Marshall Islands (Marshall Islands), and Republic of Palau (Palau) who entered the United States after 1986 (from Micronesia and the Marshall Islands) or 1994 (from Palau) and their U.S.-born children (biological, adopted, and step-) and grandchildren younger than 18 years. Appendix XIII: Comments from the Government of the Commonwealth of the Northern Mariana Islands Appendix XIV: Comments from the Government of Arkansas GAO Comments 1. The Arkansas Department of Education’s data estimating the number of compact migrant students at 4,175 is based on students’ ethnicity (Hawaiian and Pacific Islander) in the 2018-2019 school year. As a result, Arkansas’s estimate may include students who are not Marshallese. In addition, Arkansas’s estimate may include second- generation U.S. citizens, including Marshallese children born in the United States to Marshallese parents who were also born in the United States. The American Community Survey data that we report reflect a definition of “compact migrants” that includes only citizens of the Federated States of Micronesia (Micronesia), Republic of the Marshall Islands (Marshall Islands), and Republic of Palau (Palau) who entered the United States after 1986 (from Micronesia and the Marshall Islands) or 1994 (from Palau) and their U.S.-born children (biological, adopted, and step-) and grandchildren younger than 18 years. The 5,895 compact migrants that the Census Bureau estimated resided in Arkansas during the period 2013 to 2017 (a different time period from that of the data cited by the government of Arkansas) includes only adults and children who met those criteria. We believe that the Census Bureau data are sufficiently reliable for our purposes of estimating the number of compact migrants in U.S. areas. However, our report includes a discussion of stakeholder concerns that the compact migrant population in Arkansas may be undercounted. 2. The population estimate cited in the published study from Arkansas is based in part on a 2013 statement by a Marshallese consulate official. The Arkansas Department of Education estimated there were 4,175 Hawaiian and Pacific Islander students in Arkansas schools in the 2018-2019 school year. 3. Costs related to compact migration in U.S. areas not considered affected jurisdictions are outside the scope of our review. 4. We updated our report to reflect the data that the government of Arkansas cites for the period 1997 to 2019. Appendix XV: Comments from the Government of Oregon GAO Comments 1. We have previously reported on defense issues in the Federated States of Micronesia and the Republic of the Marshall Islands (Marshall Islands). For more information about the United States’ right to use part of the Kwajalein Atoll in the Marshall Islands for missile tests and space tracking operations, see GAO, Foreign Relations: Kwajalein Atoll Is the Key U.S. Defense Interest in Two Micronesian Nations, GAO-02-119 (Washington, D.C.: Jan. 22, 2002). For more information about the Marshall Islands’ Nuclear Claims Trust Fund, see GAO, Marshall Islands: Status of the Nuclear Claims Trust Fund, GAO/NSIAD-92-229 (Washington, D.C.: Sept. 25, 1992). 2. Our report provides some information about contributions by compact migrants, including qualitative statements about their budgetary, workforce, and community contributions as well as high-level data on their average per-capita income (see app. IV). We have added the government of Oregon’s statements about the contributions of compact migrants to our report. 3. As our report notes, the affected jurisdictions are defined in the amended compacts’ implementing legislation, which also establishes funding for the associated compact impact grants for those jurisdictions. 4. We made revisions in our report to help direct readers to stakeholders’ suggestions for improving experiences or outcomes of compact migration, presented in appendix VII. Appendix XVI: Comments from the Government of Washington GAO Comments 1. Our report incorporates the results of our interviews with members of compact migrant communities, including their reasons for migrating to U.S. areas, workforce challenges and other challenges they face, and their contributions to U.S. communities. Our report also includes these and other stakeholders’ suggestions for improving experiences or outcomes of compact migration (see app. VII). 2. We have previously reported on defense issues in the Federated States of Micronesia and the Republic of the Marshall Islands (Marshall Islands). For more information about the United States’ right to use part of the Kwajalein Atoll in the Marshall Islands for missile tests and space tracking operations, see GAO, Foreign Relations: Kwajalein Atoll Is the Key U.S. Defense Interest in Two Micronesian Nations, GAO-02-119 (Washington, D.C.: Jan. 22, 2002). For more information about the Marshall Islands’ Nuclear Claims Trust Fund, see GAO, Marshall Islands: Status of the Nuclear Claims Trust Fund, GAO/NSIAD-92-229 (Washington, D.C.: Sept. 25, 1992). Appendix XVII: Comments from the Government of the Federated States of Micronesia GAO Comments (from Micronesia and the Marshall Islands) or 1994 (from Palau) and their U.S.-born children (biological, adopted, and step-) and grandchildren younger than 18 years. Given this definition, any individual older than 18 years who was not born in an FAS would not be counted as a compact migrant in the Census Bureau enumerations or the American Community Survey data in this report. 8. Table 10 in appendix II of our report includes estimates of the number of compact migrants in states with fewer than 1,000 estimated compact migrants, except when the data were suppressed by the Census Bureau or the number was unreportable because the margin of error exceeded the estimate. 9. Our report notes that some FAS citizens move to U.S. areas to join the military and that the FASs have a high rate of U.S. military service, according to FAS officials and Department of State documentation. 10. The amended compacts’ implementing legislation permitted the affected jurisdictions (Hawaii, Guam, the Commonwealth of the Northern Mariana Islands, and American Samoa) to submit compact impact reports to the Secretary of the Interior. The definition of “affected jurisdictions” in the legislation did not include any mainland states. 11. Our report notes that compact migrants work in professional industries, including jobs in government and education. 12. We updated our report to include information about the locations of COFA Alliance National Network chapters in states other than Oregon. Appendix XVIII: Comments from the Government of the Republic of the Marshall Islands GAO Comments 1. Our report describes policies allowing compact migrants to access in- state tuition at colleges and universities in some U.S. areas but does not include a comprehensive description of such policies in all U.S. areas. 2. Our report describes this and other challenges related to Form I-94 and includes freely associated state consular officials’ recommendations to their citizens experiencing this challenge (see app. VII). Appendix XIX: Comments from the Government of the Republic of Palau GAO Comments Pacific in the 2005 legislation was an error, it had no impact on FAS citizens’ eligibility for limited-term REAL ID–compliant identification. Appendix XX: GAO Contact and Staff Acknowledgements GAO Contact Staff Acknowledgments David Gootnick, (202) 512-3149 or gootnickd@gao.gov In addition to the contact named above, Emil Friberg (Assistant Director), Caitlin Mitchell (Analyst-in-Charge), Topher Hoffmann, Andrew Kurtzman, Reid Lowe, Moon Parks, and Nicole Willems made key contributions to this report. Kathryn Bernet, Justin Fisher, Rebecca Gambler, Christopher Keblitis, Ty Mitchell, Mary Moutsos, and Michael Simon provided technical assistance.

Why GAO Did This Study The U.S. compacts of free association permit eligible citizens from the freely associated states (FAS), including Micronesia, the Marshall Islands, and Palau, to migrate to the United States and its territories without visa and labor certification requirements. In fiscal year 2004, Congress authorized and appropriated $30 million annually for 20 years to help defray costs associated with compact migration in affected jurisdictions, particularly Hawaii, Guam, and the CNMI. This funding ends in 2023, though migration to U.S. areas is permitted to continue and is expected to grow. GAO was asked to review topics related to compact migration. This report describes (1) estimated compact migrant populations and recent trends in compact migration; (2) reported costs related to compact migration in Hawaii, Guam, and the CNMI; and (3) effects of compact migration on governments, workforces, and societies in these and other U.S. areas. GAO reviewed Census Bureau data to determine the numbers of compact migrants in U.S. areas. In addition, GAO interviewed federal, state, and territory government officials; representatives of private sector and nonprofit groups employing or serving compact migrants; FAS embassy and consular officials; and members of compact migrant communities. In commenting on a draft of this report, U.S. area governments and FAS Ambassadors to the United States identified areas for additional study related to compact migration and impact. Some also discussed policy considerations, including restoration of Medicaid benefits to compact migrants. What GAO Found More than 94,000 compact migrants—that is, citizens of the Federated States of Micronesia (Micronesia), the Republic of the Marshall Islands (Marshall Islands), and the Republic of Palau (Palau) as well as their U.S.-born children and grandchildren younger than 18 years—live and work in the United States and its territories, according to Census Bureau data. Data from Census Bureau surveys covering the periods 2005-2009 and 2013-2017 and an enumeration in 2018 show that the combined compact migrant populations in U.S. areas grew by an estimated 68 percent, from about 56,000 to about 94,000. Historically, many compact migrants have lived in Hawaii, Guam, and the Commonwealth of the Northern Mariana Islands (CNMI). From 2013 to 2018, an estimated 50 percent of compact migrants lived on the U.S. mainland. Hawaii, Guam, and the CNMI track and report the financial costs related to compact migration, or compact impact, for their state or territory. These areas reported estimated costs totaling $3.2 billion during the period fiscal years 2004 through 2018. In fiscal years 2004 through 2019, Hawaii, Guam, and the CNMI received a combined total of approximately $509 million in federal grants to help defray the costs of providing services to compact migrants. In the U.S. areas GAO visited—Arkansas, the CNMI, Guam, Hawaii, Oregon, and Washington—state and territorial officials identified effects of providing public education and health care services to compact migrants. Some area governments use a combination of federal and state or territorial funds to extend health care coverage to compact migrants. For example, some states help compact migrants pay for coverage through health insurance exchanges, created under the 2010 Patient Protection and Affordable Care Act, by covering the cost of premiums not covered by advanced premium tax credits available to eligible compact migrants. Effects of compact migration in these U.S. areas also include compact migrants' budgetary contributions through payment of taxes and fees as well as their workforce contributions—for example, through jobs in hotels, manufacturing, the U.S. military, poultry processing, caregiving, and government.

gao_GAO-19-573

gao_GAO-19-573_0

Background EEOC provides leadership and guidance to federal agencies on all aspects of the federal government’s EEO program. EEOC assures federal agency and department compliance with EEOC regulations, provides technical assistance to federal agencies concerning EEO complaint adjudication, monitors and evaluates federal agencies’ affirmative employment programs, develops and distributes federal sector educational materials, conducts training for stakeholders, provides guidance and assistance to Administrative Judges who conduct hearings on EEO complaints, and adjudicates appeals from administrative decisions made by federal agencies on EEO complaints. EEOC’s MD-715 Reporting Requirements EEOC’s MD-715 requires agencies to take appropriate steps to establish a model EEO program and to ensure that all employment decisions are free from discrimination. It also sets forth the standards by which EEOC will review the sufficiency of agency Title VII and Rehabilitation Act programs, which include periodic agency self-assessments and the removal of barriers to free and open workplace competition. Under MD-715, federal agencies, and any subordinate component that enjoys autonomy from its parent agency, are required to submit annual MD-715 EEO program status reports to EEOC. Completed MD-715 reports include: Part F: Requires designated agency officials to certify that the agency has completed an annual self-assessment (Part G) and established plans to correct any program deficiencies (Part H), as well as conducted comprehensive barrier analyses and established plans to eliminate identified barriers (Part I). Part G: Contains a self-assessment checklist for an agency to assess its compliance with essential EEO program elements to operate a model EEO program and identify any deficiencies. Part H: Describes the agency’s plans to address identified deficiencies. Part I: Shows identified EEO triggers and barriers for race, gender, and national origin; how the agency plans to address them; and who the responsible officials are. Part J: Contains the agency’s affirmative action plan for individuals with disabilities and individuals with targeted disabilities. Additionally, federal agencies are required to identify and eliminate barriers that impede free and open competition in their respective workplaces. EEOC reporting requirements state that a barrier is an agency policy, procedure, practice, or condition that limits or tends to limit employment opportunities for members of a particular gender, race, or ethnic background, or for individuals based on disability status. According to EEOC’s instructions, many employment barriers are built into the organizational and operational structures of an agency, and are embedded in the day-to-day procedures and practices of the agency. Agencies are also required to identify EEO program deficiencies and develop plans to address them. According to EEOC’s instructions, deficiencies are weaknesses in an agency’s EEO program where agency officials need to provide more attention. For example, a deficiency might be that the EEO director is not under the direct supervision of the agency head, or that an EEO Director or Officer lacks a regular, effective means of informing the agency head and other top management of the effectiveness, efficiency, and legal compliance of the agency’s EEO program. EEOC’s Office of Federal Operations instructs agencies on how to complete their MD-715 reports, provides training and technical assistance, and offers additional informal assistance, such as sharing best workplace practices. Based on agency MD-715 reports, EEOC includes assessments of agency progress in its Annual Report on the Federal Workforce, and in notice and feedback letters addressed to individual agencies. In addition, according to EEOC officials, EEOC meets with each agency every 3 years to review the status of its compliance with federal EEO laws, regulations, and management directives. If EEOC determines that areas of noncompliance exist in an agency’s program, it may take compliance actions. Compliance actions include requiring the agency to provide an update on the status of its plans to correct deficiencies in its MD-715 submission, or to submit a Compliance Report to EEOC explaining the agency’s progress in correcting deficiencies within 6 months of the date of a feedback letter. If agencies do not comply, EEOC may choose to initiate its noncompliance process, which could include conducting a program evaluation, issuing a notice to the Secretary of Homeland Security, or publicly identifying DHS as a noncompliant agency. As part of its noncompliance process, EEOC has conducted program evaluations of DHS components. In 2013, EEOC initiated a program evaluation at TSA to determine the adequacy and appropriateness of the EEO complaint framework in all offices and directorates within TSA. EEOC reported in 2014 that TSA’s EEO complaint process was adequate and complied with its regulations. However, EEOC made eight recommendations to TSA, including one that called for TSA to modify the EEO information in its training materials and presentations. In response to EEOC’s report, TSA submitted a letter to EEOC stating that it planned to address all of the EEOC recommendations and had already taken steps to implement several of them. In addition, in 2018, EEOC conducted a multiagency program evaluation that included U.S. Customs and Border Protection (CBP). It found that CBP had no women serving in positions that involve intercepting prohibited commodities or persons, and that women comprised only 5 percent of its border patrol agents. EEOC stated that the report’s recommendations may help CBP with its hiring efforts. DHS’s EEO Program CRCL, through the Deputy Officer for EEO and Diversity, is responsible for processing complaints of discrimination; establishing and maintaining EEO programs; fulfilling reporting requirements as required by law, regulation, or executive order; evaluating the effectiveness of EEO programs; leading the department’s diversity management program; and preparing and submitting DHS’s annual MD-715 report to EEOC. According to EEOC policy, a second-level reporting component is one that enjoys autonomy from its parent agency, and has 1,000 or more employees. EEOC instructions require second-level reporting components submit MD-715 reports to their agency headquarters for inclusion in the agency-wide report in addition to submitting them directly to the EEOC. DHS’s Headquarters EEO Office, a part of CRCL, implements the EEO program for all headquarters employees and applicants. DHS has nine second-level reporting components, including DHS Headquarters, that are required to submit individual MD-715 reports to EEOC. Each component has an office headed by a director charged with implementing its EEO program. Figure 1 shows the officials who are primarily responsible for EEO at DHS. DHS’s MD-715 Reporting Process CRCL develops DHS’s annual MD-715 EEO program status report and submits it to EEOC. CRCL works with components to gather and analyze necessary data and information, and to perform the required MD-715 exercises that are ultimately used to complete the overall DHS MD-715 report. CRCL includes components’ identified deficiencies in the DHS- wide MD-715 report. The Secretary of Homeland Security (or its designee) and CRCL’s Deputy Officer for EEO and Diversity are to certify DHS’s MD-715 report before CRCL sends the report to EEOC. Figure 2 illustrates DHS’s MD-715 report development process. This process includes conducting a self-assessment checklist of DHS’s and its components’ efforts to achieve a model EEO program and barrier analysis to eliminate identified EEO barriers. DHS Has Plans to Address the EEO Barriers It Identified, but Could Better Measure Its Progress toward Eliminating the Barriers DHS Has Taken Steps to Follow EEOC Guidance to Identify and Address EEO Barriers DHS has taken steps to follow EEOC’s guidance by using and analyzing various information sources, investigating possible causes of potential barriers or triggers, and planning activities to address and eliminate barriers. EEOC MD-715 guidance calls for federal agencies to continually work towards preventing all forms of discrimination and eliminating barriers that may impede free and open competition in the workplace. Figure 3 shows the barrier identification and elimination steps under MD- 715. DHS generally uses the information sources that EEOC guidance recommends in addition to workforce data to help identify potential barriers. As directed by EEOC guidance, DHS analyzes its workforce data to help identify triggers or indicators of potential EEO barriers by comparing the racial, national origin, gender, and disability profiles of its total workforce, and for various occupational categories to relevant civilian labor workforce data. In fiscal year 2017, DHS analyzed all available workforce data including: Total Workforce – Distribution by Race/Ethnicity, Gender, and Participation Rates by Major Occupations – Distribution by Race/Ethnicity, Gender, and Disability, and Applicants and Hires by Major Occupations – Distribution by Race/Ethnicity, Gender, and Disability. In addition to analyzing workforce data, in each of the fiscal years 2014 through 2017, DHS utilized the U.S. Office of Personnel Management’s Federal Employee Viewpoint Survey (FEVS) and DHS’s employee exit survey results to help identify and address barriers. For example, CRCL, in DHS’s fiscal year 2017 MD-715 report, used FEVS and exit survey results to help investigate the possible causes of higher-than- expected nonretirement separations for white females and several other ethnic and racial groups. According to the report, possible causes included the lack of advancement opportunities, insufficient work/life programs, and the lack of alternate work schedules. During our small group discussions, DHS employee groups told us that through the MD-715 report development process, they helped identify and address triggers and barriers. For example, Special Emphasis Program Managers we spoke with told us that DHS components conduct climate surveys to obtain input from employees on workforce practices every 1 or 2 years. Further, several DHS components’ MD-715 reports referenced soliciting employee input, such as obtaining Disability Employment Program Managers’ input via quarterly disability employment advisory council meetings where they share best practices and discuss issues and topics including barriers. DHS Has Identified Workforce Demographics Data Anomalies That Indicate Potential EEO Barriers Our review of DHS’s MD-715 reports showed that DHS identified three department-wide triggers in fiscal years 2014 through 2017. The three triggers were (1) high rate of nonretirement separations for certain groups, particularly white women; (2) low participation rates of women and various ethnic and racial groups in the permanent workforce; and (3) low participation rates of individuals with disabilities and targeted disabilities. While Investigating Various EEO Anomalies, DHS Identified Three Barriers Subsequent to its trigger identification and department-wide barrier analysis, from fiscal years 2014 through 2017, DHS identified three barriers: (1) problems with supervision/management, lack of advancement opportunities, lack of alternate work schedules, insufficient work/life programs, and personal/family related reasons causing higher- than-expected nonretirement separations for white females and several ethnic and racial groups; (2) the geographic location of jobs which has contributed to the low hiring rates of racial groups in certain major occupations; and (3) medical and physical requirements of law enforcement positons, such as the ability to engage in moderate to arduous physical exertion, which limit the eligibility of some applicants with targeted disabilities. DHS identified these barriers by analyzing component and DHS level workforce data and reviewing DHS FEVS and exit survey results. DHS identified barriers in its MD-715 reports for fiscal years 2014 through 2017. However, EEOC noted that for fiscal years 2015 to 2017, DHS had not identified any policies, procedures, practices, or conditions causing (1) low hiring rates for women in certain major occupations, and (2) the high separation rate of employees with disabilities. DHS’s Planned Activities to Address the Identified EEO Barriers As stated in EEOC’s guidance, barrier elimination is a vital step to addressing identified barriers and working towards the goal of making the federal government a model employer. To address and eliminate identified barriers, EEOC’s instructions direct agencies to include in their MD-715 reports measurable objectives, an action plan that includes planned activities and completion dates, as well as officials responsible for overseeing the plan, and a summary of accomplishments. Since our 2009 recommendations, DHS has included interim milestones in its MD-715 reports. Our 2009 report showed that DHS had modified nearly all of its target completion dates. We recommended that DHS identify essential activities and establish interim milestones necessary for the completion of all planned activities to address identified barriers to EEO. In its fiscal year 2011 MD-715 report to EEOC, DHS identified essential activities and established interim milestones. Based on its MD- 715 reporting for fiscal years 2014 through 2017, DHS has continued to identify planned activities and establish interim milestones. Our review of DHS’s MD-715 reports from fiscal years 2014 through 2017 also shows that DHS has planned activities and targeted completion dates to address each identified barrier, and each trigger for a potential barrier. For example, to address the low participation rates of women and several ethnic and racial groups in DHS’s overall workforce, DHS’s planned activities included researching where to conduct outreach for the identified groups, and producing a plan to integrate data from the multiple applicant data-tracking systems used across DHS. DHS’s outreach activities included identifying colleges and universities with large populations of underrepresented groups, identifying relevant job fairs in selected service areas, and conducting focus group meetings with employees from underrepresented groups to determine how to improve recruitment and retention, among other events. These events were initiated in 2011, but are to be reviewed and updated annually. For example, DHS reported that it develops a “Top 25” list of annual outreach and recruitment activities that include law enforcement focused events. DHS also reported developing a framework in 2016 for applicant flow data analysis—important for identifying and addressing potential recruitment and outreach barriers. In 2017, activities included conducting more robust department-wide analysis of applicant data. Many of the activities were initiated in prior years and target dates for completion were met. To address the high nonretirement separation rate of certain groups, notably white women, DHS’s planned activities included updating and augmenting previously instituted exit survey methods, and identifying and implementing retention interventions. Further, in its fiscal year 2014-2017 reports, DHS has identified essential activities, established interim milestones, and met recurring interim milestones for its planned activities. For example, DHS reported that it planned to research where to conduct outreach for groups in occupations with underrepresentation. DHS components completed this outreach activity in 2012, and components and facilities are to annually identify (1) colleges with substantial populations of underrepresented groups, (2) relevant job fairs in the service area, and (3) relevant local affinity groups and community groups, among other outreach activities. Additionally, DHS’s Office of the Chief Human Capital Officer (OCHCO) has lead responsibility for implementing a multiyear plan for targeted recruitment of applicants from identified underrepresented groups. OCHCO completed its initial multiyear plan in 2012 and is to annually update its established goals for intern programs, job fairs, and local advertising. Selected DHS Components Took Steps to Conduct Barrier Analyses All four selected DHS components have taken steps to follow EEOC guidance to conduct barrier analyses. Of the components, Federal Law Enforcement Training Centers (FLETC), the U.S. Secret Service (Secret Service), the Transportation Security Administration (TSA), and U.S. Citizenship and Immigration Services (USCIS), only one, TSA, identified any EEO barriers. However, each of the components identified triggers and analyzed potential barriers by reviewing workforce data (i.e., data on total workforce, new hires, and mission critical occupations) and comparing the data to relevant benchmarks, reviewing various information sources to help identify possible barriers that may be resulting in the current condition highlighted by the analysis of workforce data, and reporting action plans and time frames for addressing potential or actual barriers. The Secret Service’s fiscal year 2017 MD-715 report showed that after analyzing demographic data to identify triggers, the Secret Service used FEVS data to identify potential barriers to the employment of individuals with disabilities in occupations where the triggers were identified. In addition, USCIS stated in its fiscal year 2017 MD-715 report that its review of exit survey data provided reasons that men, Hispanics, and whites left the agency, but data were inconclusive regarding the continuing underrepresentation of those groups. USCIS also reported that it would continue analyzing exit data in fiscal year 2018. In fiscal year 2017, TSA identified two barriers in its MD-715 report—(1) medical and physical restrictions limit opportunities for individuals with disabilities and individuals with targeted disabilities in Transportation Security Officer and Federal Air Marshal occupations, and (2) women are not applying to Transportation Security Officer or Federal Air Marshal positions at the same rate as men. TSA reported that it analyzed workforce data and policies, procedures, and practices related to recruiting, hiring, and promotions to try to determine what may be contributing to low participation rates for women and individuals with disabilities. TSA also interviewed employees involved in those processes, and conducted focus groups with supervisors and leadership at airports and field offices. TSA’s plans to address barriers include developing a communication plan to promote TSA programs that support persons with disabilities and with targeted disabilities; making sure training modules are accessible; conducting training to increase awareness of unconscious bias towards working with individuals with disabilities; and working with its human capital office and others to assist with recruiting and hiring to more effectively target women. Although FLETC, Secret Service, and USCIS did not identify EEO barriers in fiscal year 2017, they each developed action plans that identified activities designed to help address and correct undesired conditions, identified responsible officials, and set time frames for addressing the conditions. Examples of selected components’ plans and activities include: FLETC. To address low participation rates of persons with targeted disabilities in the permanent workforce, FLETC-planned activities include working with human resource specialists to identify data and timelines needed to create reports in its applicant data flow system that would help identify any barriers in the selection process, and working to resolve issues concerning applicant flow data in the applicant pool. Secret Service. To address low participation rates of certain groups in the general workforce and new hires, planned activities include quarterly tracking and reporting ethnicity, race, and gender data net changes, hires, resignations, and retirements. Other activities would involve working closely with the Office of Human Resources Talent and Employee Acquisition Management Division in recruitment activities. USCIS. To address the lower-than-expected participation rate of certain groups in the permanent workforce—for example, white males and females and Hispanic males—planned activities include conducting comprehensive applicant flow data analysis of the top five major occupational categorizes, and administering and analyzing a bi- annual EEO and Diversity Climate Survey. DHS has provided training for its components on how to conduct EEO barrier analysis. In 2016 and 2018, DHS trained DHS component EEO officials on methods for identifying the root of specific triggers in the workplace, as well as steps for eliminating identified barriers. According to DHS’s analysis of participant training evaluations, the majority of participants believed they would be able to apply what they learned from the training. In 2017, DHS provided a 2-day barrier analysis training to agency and component affirmative employment practitioners that introduced various barrier analysis methods. It included an exercise involving a hypothetical federal agency. Based on our review of participant evaluations, participants were satisfied with the training. DHS Reports Some Improvements in Employee Engagement and Representation of Minorities and Women, but Lacks Performance Metrics for Tracking Progress DHS reported improvements in EEO indicators in its MD-715 reports from fiscal years 2014 through 2017. DHS cited its higher FEVS scores under employee engagement. For example, although DHS’s employee engagement remained 7 percent below the government-wide average, it increased from 54 percent in 2014 to 60 percent in 2017. According to DHS, this score was largely driven by TSA and U.S. Customs and Border Protection employees, who accounted for 46.8 percent of DHS’s completed surveys. Our review of DHS’s workforce data from fiscal years 2014 through 2017 showed that every minority group as well as individuals with disabilities and individuals with targeted disabilities had been trending in a positive direction since fiscal year 2014. Further, DHS officials told us that minority representation was up 3 percent and female representation was up 2 percent since 2015. In addition, DHS has produced barrier analysis reports that address underrepresentation of women and various ethnic and racial groups. In 2018, DHS completed a barrier analysis report on Hispanic employment in General Schedule pay scale grades 12 and higher, as required by EEOC and the U.S. Office of Personnel Management. The report identified several potential triggers, such as Hispanic women separating from DHS, and related barriers, such as possible harassment of Hispanic employees and women, and glass walls. DHS also developed action plans focused on enhancing elder and family care programs, offering training on preventing harassment in the workforce, increasing recruitment into job series with substantial promotion opportunities, and ensuring interview panels were diverse and interviewers properly trained. Although DHS has reported positive trending in various underrepresented groups, DHS officials said they were unable to fully identify the barriers contributing to the underrepresentation of women in its workforce despite conducting the required barrier analysis. In 2014, DHS conducted a barrier analysis of women in law enforcement to help identify any barriers. While specific barriers were not identified, DHS’s report, Women in Law Enforcement Study, provided insight into why DHS employed lower rates of female law enforcement officers than federal government-wide. For example, study participants shared anecdotal instances of where they or their colleagues did not pursue promotional opportunities because they perceived their work environment made them choose between the job and family. The study also highlighted steps DHS could take to help address its underrepresentation of women, such as being more creative in its approach to attracting qualified women through use of social media, and by creating more family-friendly environments. According to EEOC, one important tool in examining the fairness and inclusiveness of the federal government’s recruitment efforts is applicant flow data. By reviewing the yield of an agency’s recruitment effort, the organization can reassess and improve its effort to reach all segments of the population. EEOC guidance states that having department-wide applicant flow data could aid in analyzing differences in selection rates among different groups for a particular job. In July 2017, EEOC informed DHS that the agency’s applicant flow data were incomplete. DHS has reported challenges in collecting department-wide data that could help identify potential barriers. EEOC found that DHS’s workforce data tables do not always contain all of the agency’s applicant flow data. According to EEOC, without such data, it becomes much more difficult to pinpoint the specific policies, procedures, or practices in which barriers might be embedded. DHS does not have a consolidated applicant flow data system. According to DHS, four of its components use one system (USA Staffing), while five other components use a different system (Monster Government Solutions). Office for Civil Rights and Civil Liberties (CRCL) officials told us DHS is developing a new system to integrate applicant flow data department-wide. However, the officials could not give us a time frame for when the system is expected to be completed. As a work-around, DHS explained that it obtains these data directly from each component that uses Monster Government Solutions. CRCL officials said they will report complete applicant flow data in fiscal year 2019. In addition to creating a model EEO environment, progress in eliminating EEO barriers can help DHS avoid costs related to workplace disputes. According to EEOC guidance, the elimination of barriers may help an agency avoid expensive costs, such as back pay awards, compensatory damages, and attorney’s fees, from findings of discrimination. EEOC found 81 instances of discrimination from fiscal years 2014 through 2017 resulting in DHS paying nearly $30 million to cover judgments, awards, and settlements for these EEO cases, or an average of $7.4 million per year. These expenses were nearly equal to the average annual cost of DHS’s EEO program, which DHS estimated at about $7.63 million in fiscal year 2019. DHS Lacks Metrics for Tracking Progress towards Eliminating Identified EEO Barriers DHS does not have complete performance metrics or mechanisms for tracking progress towards eliminating its identified EEO barriers. For example, CRCL does not maintain numerical objectives or goals for eliminating barriers involving certain EEO groups, such as workplace satisfaction of white females or the retention rate of women in law enforcement positions. According to CRCL officials, they are not required to establish performance metrics or mechanisms for tracking progress towards eliminating barriers beyond what is included in the department- wide MD-715 report. DHS reported one performance measure for its EEO program—the percent of timely merit Final Agency Decisions (FADs). Standards for Internal Control in the Federal Government states that management should establish specific and measureable objectives, and ways to assess progress including performance metrics and milestones. It also states that management should design control activities to achieve objectives and respond to risks. Such control activities may include the establishment and review of performance metrics. Further, EEOC guidance states that agencies are not prevented from establishing additional practices that exceed its requirements. DHS officials acknowledged that their EEO program performance measurement does not reflect all the work that they do. According to CRCL officials, CRCL has proposed additional performance measures for its MD-715 activities, but they were rejected by DHS’s Office for Policy because they were not directly related to national security or public safety. DHS’s Office for Policy is responsible for approving new performance measures. CRCL officials told us that adopting hiring goals for individuals with disabilities and individuals with targeted disabilities—which had previously been identified as potential barriers—has been beneficial in garnering support and commitment towards meeting them. They said that DHS incorporated these goals into its efforts and initiatives to increase the recruitment, hiring, advancement, and retention of individuals with disabilities. Implementing performance metrics could help DHS better assess its progress in eliminating EEO program barriers. DHS and Its Components Have Taken Steps to Identify EEO Program Deficiencies, but Lack Action to Fully Address Them DHS and Its Components Have Identified Various Deficiencies in Their EEO Programs As shown in table 1, our analysis of DHS’s MD-715 reports found that DHS did not meet about a quarter of the compliance measures for a model EEO program for each fiscal year from 2014 through 2017. Specifically, over this 4-year period, DHS did not meet 26 percent of its compliance measures (128 out of 487). The largest percentage of unmet measures occurred under the model EEO essential element D—which focuses on proactive steps taken by an agency to prevent unlawful discrimination—where about 53 percent or 21 of 40 measures were unmet. According to DHS officials, in Part G of its MD-715 report, DHS includes deficiencies identified and reported at the component level as well as deficiencies directly attributable to the department. For example, in each of the fiscal years 2015 through 2017, DHS reported that it did not meet a compliance measure under element D that senior managers successfully implement EEO action plans and incorporate EEO action plan objectives into agency strategic plans. Specifically, in fiscal years 2015 through 2017, DHS noted that USCIS had not met this measure, and in fiscal year 2017, the Federal Emergency Management Agency (FEMA) and DHS Headquarters did not meet this measure. Our analysis of components’ MD-715 reports showed that components did not meet 9 percent of the compliance measures for a model EEO program from fiscal years 2014 through 2017. Specifically, over this 4- year time frame, components had a combined total of 369 program deficiencies out of a total of 4,229 compliance measures. DHS Headquarters, one of the nine second-level reporting components, accounted for 36 percent of deficient measures (134 of 369), while the other eight components accounted for 64 percent (235 of 369) of deficient measures. Examples of DHS’s deficient measures included EEO directors not under the direct supervision of the agency head, and the lack of established timetables or schedules for the agency to review its employee development and training programs for systemic barriers that may be impeding full participation in training opportunities by all groups. DHS and Its Components Lack Action Plans to Address Some EEO Program Deficiencies DHS and its components did not have action plans to address some of their self-identified deficiencies from fiscal years 2014 through 2017. Specifically, DHS did not have action plans to address 56 percent, or 72 of the 128 reported deficiencies, and components did not have action plans to address nearly half, or about 179 of the 369 deficiencies reported by all of the components during the four year period. For example, in fiscal year 2017, four out of nine DHS components—U.S. Customs and Border Protection (CBP), DHS Headquarters, FEMA, and Federal Law Enforcement Training Centers (FLETC)—did not have action plans to ensure that their EEO directors report directly to their agency heads. EEOC guidance requires agencies to demonstrate meaningful progress toward the removal of deficiencies, and to develop action plans for how agencies will attain the essential elements of a model EEO program. Specifically, for each deficient measure, agencies are to develop an action plan for correcting the deficiency. The plan should identify and briefly describe the deficiency; provide a measurable objective, including the reason for the deficiency, and target date for completion; identify officials responsible for overseeing implementation of planned activities to accomplish the objective; and provide for a yearly update on status of activities until objective is completed (i.e., the deficiency is removed). In addition, Standards for Internal Control in the Federal Government states that management should design control activities to achieve objectives. Control activities, such as policies or procedures to enforce directives, can help identify if a required activity is not being achieved (e.g., action plan completion) or implemented. The four selected DHS components told us that they do not have standard operating procedures for completing their MD-715 reports, including review and assessment of deficiencies and action plans, but have various processes in place to review their reports for accuracy and completeness. For example, according to USCIS, its process for the preparation and review of its MD-715 report includes providing a self- assessment checklist to each of its subcomponents; a review of their responses for accuracy; and follow-up with subcomponents to address any questions. In addition, USCIS stated that its MD-715 report undergoes multiple levels of reviews by subject matter experts and managers that include collaboration with human capital and chief counsel, and obtaining review and approval from the Director and other agency officials. The other three components—FLETC, TSA, and Secret Service—also reported having MD-715 review processes in place, including report review and approval by senior management; however, none specifically cited a review and approval of action plans to address reported deficiencies. CRCL officials told us that DHS and its components’ MD-715 reports met EEOC requirements for action plans for fiscal years 2014 through 2017 by providing explanations for, or briefly stating plans to address, the majority of their deficiencies rather than developing action plans identifying how each deficiency would be addressed. During our review of the MD-715, we noted that the Part G self-assessment checklist form gave respondents the option of providing a brief explanation in a comment box on the form or completing an action plan for each deficiency in Part H of the MD-715 report. For fiscal year 2018, EEOC revised its MD-715 report form and instructions to clarify that a plan is required for each identified program deficiency. For example, one component responded to a measure that asks whether an agency implemented an adequate data collection and analysis system that permits tracking of the information required by MD-715 and these instructions by stating “selected offices were currently working on the initiative.” The same component responded to a measure that asks whether an agency tracked recruitment efforts and analyzed efforts to identify potential barriers in accordance with MD-715 standards by stating, “The inconsistencies with the reporting of applicant flow data will need to be addressed to help identify potential barriers.” Another component responded to this measure by stating that, while its participation in various events are tracked, a clear, concise, and efficient system to track and analyze recruitment efforts according to MD-715 standards is currently not in place. Neither component provided a plan for how these deficient measures would be addressed. EEOC continues to identify areas of noncompliance in DHS component EEO programs. For example, in fiscal year 2017, EEOC noted that three of four selected components had areas of noncompliance. The areas of noncompliance included (1) failure to timely issue FADs, and (2) not establishing timetables or schedules to review its merit program policies and procedures, employee recognition awards programs, and employee developmental training programs for potential barriers. Developing policies or procedures, in consultation with the Deputy Officer for EEO and Diversity, to help ensure component EEO programs have action plans with measurable objectives for addressing deficiencies could help DHS components better comply with EEOC requirements. DHS and Its Components Lack Adequate Staffing to Address EEO Program Deficiencies DHS continues to report insufficient staffing to support its EEO program. In 2009, we reported that, according to CRCL, DHS modified the target dates for planned activities to address identified barriers primarily because of staffing shortages in both CRCL and the Office of the Chief Human Capital Officer. We also reported that DHS had not conducted barrier analyses of policies, procedures, and practices that were established or used after fiscal year 2004 because of resource limitations, such as staffing and limited funding to contract for this activity. According to CRCL’s MD-715 and Notification and Federal Employee Antidiscrimination and Retaliation (No FEAR) Act reports from fiscal years 2014 through 2017, certain aspects of DHS’s EEO program did not have sufficient staffing. In addition, in fiscal years 2014 through 2017, DHS reported that staffing shortages contributed to it not meeting its target for the percent of timely decisions on discrimination complaints. In fiscal year 2017, DHS reported deficiencies for five out of seven staffing measures in its MD-715 report. In February 2019, CRCL officials told us they lacked staffing to issue timely decisions on discrimination complaints, to increase the number of mediators in the alternative dispute resolution program, and to provide them with training. From fiscal years 2014 through 2017, DHS and its components reported funding and staffing challenges in components’ EEO programs. As shown in table 2, DHS and its components reported that certain aspects of components’ EEO programs do not have sufficient funding or staffing from fiscal years 2014 through 2017. EEOC guidance states that an agency must provide its EEO program with sufficient budget and staffing to be able to successfully implement various activities, including (1) conducting a self-assessment of the agency for possible program deficiencies; (2) conducting a thorough barrier analysis of its workforce; and (3) ensuring timely, thorough, and fair processing of EEO complaints. CRCL and component EEO officials told us that they do not have formal staffing models to assess appropriate staffing of their EEO program sections. CRCL officials explained that each component EEO program section is unique with its own assessments and measures by the leaders in charge of their funding and staffing resources. Using these informal processes to identify staffing needs, CRCL and component EEO officials told us that they have requested additional staffing and funding to address some of their EEO program deficiencies from their top leadership. However, they said that additional staffing has not been granted. Our analysis of DHS’s congressional budget justifications show that DHS’s EEO program funding requests have decreased each year from nearly $8 million in fiscal year 2016 to nearly $7 million in fiscal year 2019. CRCL officials told us that DHS’s overall resources for the EEO program have not significantly increased. A staffing model could be a computer-based formula that estimates the number of staff needed to conduct varying numbers of EEO activities, such as processing a certain number of complaints or providing a certain number of training courses on an annual or ad hoc basis. As we have reported, a staffing model is a helpful tool that could better justify requests for resources to top leadership. Staffing models can identify resources required to enable program delivery to a sufficient degree and in a timely manner, or to adapt to changes in program delivery. According to DHS, the department has contracted support to help components develop models for Mission Support areas as part of a larger effort to ensure that all positions are eventually covered by a staffing model. Developing and utilizing a formal staffing model for its EEO program could help CRCL better identify, request, and obtain the staff it needs. Further, developing staffing models, in collaboration with the Deputy Officer for EEO and Diversity, would help components to better assess the staff they need. DHS Has Plans to Address the Nine Areas of EEOC Identified Noncompliance DHS did not respond timely to EEOC’s findings of noncompliance and EEOC did not follow up with DHS concerning the untimely response. In July 2017, in its most recent review of DHS compliance with EEOC requirements, EEOC reported nine areas of noncompliance in DHS’s EEO program. For example, EEOC found that DHS lacked resources to process EEO complaints and to conduct trend analyses of workforce data. EEOC stated in its feedback letter to DHS that it would initiate its noncompliance process if DHS did not submit a report explaining the agency’s progress in correcting its EEO program deficiencies by January 2018. However, according to DHS officials, due to an administrative oversight, DHS was unaware of EEOC’s July 2017 feedback letter until October 2018, when we asked about it. In February 2019, DHS submitted a report to EEOC that responded to each area of noncompliance. EEOC officials told us that it had not initiated its noncompliance process against DHS, but that it had placed DHS in its queue for agencies to be held in noncompliance. As discussed earlier, DHS Headquarters, a second-level reporting component, is required by EEOC to submit a separate MD-715 report to EEOC. However, DHS’s Headquarters EEO Office did not submit a separate MD-715 report to EEOC during fiscal years 2014 through 2017. DHS Headquarters EEO Office staff told us that the office had not submitted the required reports due to staff vacancies, including its EEO director position. They explained that the component’s EEO data and information were subsumed in DHS’s department-level MD-715 submission. In October 2018, CRCL filled its Headquarters EEO director position, which had been vacant for 8 months. DHS officials told us they plan to submit Headquarters’ fiscal year 2018 MD-715 report to EEOC by the due date. In February 2019, EEOC officials told us that DHS could be subject to EEOC’s noncompliance process if the report is not received. DHS’s EEO and Human Capital Offices Have Taken Steps to Oversee and Support Components, but Need to Strengthen Oversight over Components DHS’s EEO and Human Capital Offices Use a Variety of Means to Oversee and Support Components in Identifying and Addressing EEO Barriers As shown in figure 4, CRCL and the Strategic, Recruitment, Diversity, and Inclusion (SRDI) Office support and oversee components in their efforts to identify and address EEO barriers. For example, CRCL convenes an EEO council consisting of EEO directors from each component that meets monthly and, among other things, shares best practices for identifying and addressing barriers. In addition, CRCL hosts EEO and Diversity Training Conferences for EEO staff that includes barrier analysis training. Further, CRCL provides midyear feedback to component EEO officials on components’ planned action items and plans for inclusion in their respective MD-715 reports. For example, based on our review of the CRCL statistician’s notes, during feedback meetings with components in 2017, he suggested that components consider opportunities for improving their draft MD-715 reports. The notes show that at least two out of nine components—CBP and DHS Headquarters—were given feedback to conduct more robust barrier analyses. SRDI supports component EEO program efforts to address EEO barriers related to recruitment, hiring, veterans, and individuals with disabilities. For example, to increase the participation of women in law enforcement across the department, SRDI held a joint hiring event in Dallas based on its analysis that a large number of female veterans live in Texas. According to SRDI officials, SRDI also assists DHS components with their evaluations of their human capital policies, procedures, or practices that may represent EEO barriers, such as awards, promotions, and career development. For example, in fiscal year 2016, SRDI analyzed the representation of the DHS Senior Executive Service Candidate Development Program applicant pool by various ethnic and racial groups, and by actual selectee participation. When it found that the representation rate of women decreased from 32.5 percent in the application stage to 23.4 percent in the selection stage, SRDI stated that the results, among other things, triggered the need for further analysis. Two cohorts later in 2018, the representation rate of women increased from 23.3 percent to 41.4 percent in the selection/participant stage. Further, CRCL and SRDI officials said they collaborate on a number of EEO activities to identify and address EEO barriers. For example, SRDI works together with CRCL to provide input for completing MD-715 report sections that address human capital-related EEO barriers. In addition, SRDI and CRCL worked together to conduct a barrier analysis of Hispanic employee representation. DHS Components Are Generally Satisfied with CRCL’s Collaboration Practices to Identify and Address EEO Barriers DHS components told us that its collaboration practices are generally working well and provided examples. In our interviews of nine DHS components, they told us they are generally satisfied that DHS has: clearly defined its short- and long-term outcomes, bridged the organizational cultures of participating agencies, clearly defined roles and responsibilities for participating agencies, included all relevant DHS participants when identifying and addressing EEO barriers, funded and staffed its collaborative mechanisms, such as monthly EEO council meetings, and documented its agreements on how participating agencies will be collaborating in identifying and addressing barriers. All nine components told us that CRCL regularly meets with them and provides guidance on identifying and addressing barriers. Four components specifically stated that CRCL provided assistance for reviewing and processing EEO data. For example, USCIS officials said that CRCL’s statistician provided direction on analyzing workforce data when conducting barrier analysis. Components also said that they find the training and technical assistance provided by CRCL helpful, and specifically commented that DHS’s EEO and Diversity Training Conferences have helped improve their barrier analyses. While DHS components are generally satisfied with DHS’s collaboration practices, some components provided examples of collaboration practices that could be improved. Three components—CBP, the U.S. Secret Service (Secret Service), and USCIS—told us that collaboration on funding or staffing efforts could be improved. For example, USCIS officials said CRCL lacks sufficient staffing to provide needed training, tools, and assistance to components to meet new MD-715 reporting requirements. Three components—CBP, the Transportation Security Administration, and USCIS—cited the lack of written guidance and agreements regarding collaboration between CRCL and components as areas that could be improved. For example, USCIS officials said that its collaborative efforts with CRCL were guided by informal best practices, feedback, and guidance, but having formal written guidance and agreements could clarify roles and responsibilities for identifying and addressing component EEO triggers and barriers. CRCL Does Not Review DHS Components’ EEOC Feedback Reports and DHS’s Organizational Structure Does Not Ensure Its Components Comply with EEOC Requirements CRCL officials and component EEO officials stated that component EEO directors report directly to their respective component heads and not to CRCL. While CRCL requires components to meet to discuss midyear updates on their EEO efforts, CRCL officials explained that DHS components are responsible for developing, certifying, and submitting their own MD-715 reports to EEOC. They also said that if EEOC finds areas of noncompliance in DHS components’ EEO programs, EEOC requires DHS components to submit their compliance reports directly to EEOC. In fiscal year 2017, EEOC provided notice to six out of eight DHS components for having areas of noncompliance in their EEO programs. For five out of six DHS components, EEOC required components to establish plans to correct deficiencies, submit compliance reports explaining the agency’s progress in correcting these deficiencies, and showing meaningful progress in implementing its plans within 6 months. We found that three out of five DHS components—CBP, FEMA, and USCIS—did not submit timely compliance reports in response to EEOC’s feedback letters. Due to an administrative oversight, CBP officials explained that the component did not submit a compliance report that was due in February 2018 until we asked about it during our review. Although CBP submitted the report in March 2019, the report did not include plans to correct three out of seven areas of noncompliance. As of July 2019, CBP has taken steps to address the areas of noncompliance but has not yet responded to EEOC. As a result, CBP remains at risk of EEOC initiating the noncompliance process against it. FEMA also did not submit a compliance report that was due in February 2018 until we asked about it during our review. In June 2019, FEMA responded to EEOC’s feedback letter and included plans to correct three areas of noncompliance. FEMA’s response stated that the component would provide another update on its plans to correct these areas to EEOC in October 2019. FEMA’s response also stated that it would update its actions on 12 other areas of noncompliance in its fiscal year 2018 MD- 715 report. EEOC guidance states that an agency’s EEO Director ultimately is responsible for ensuring equal opportunity throughout the entire agency. In addition, Standards for Internal Controls in the Federal Government states that management should implement control activities through policies. According to CRCL officials, CRCL does not have policies and procedures to ensure that components have addressed EEOC’s feedback letters completely and timely. CRCL officials said CRCL does not have the authority to ensure components’ responses completely and timely address EEOC’s feedback letters. They explained that components interact directly with EEOC and are not required to discuss EEOC’s feedback with CRCL. CRCL officials further said that components may address EEOC’s feedback in their MD-715 reports instead of sending compliance reports to EEOC. For example, in response to the EEOC’s 2017 feedback letter, in its MD-715 report for fiscal year 2017, the U.S. Coast Guard discussed ways to assist DHS with improving its issuance of Final Agency Decisions. CRCL reported in its MD-715 reports from fiscal years 2015 through 2017 that it had authority for components’ EEO programs. A DHS delegation of authority order states that CRCL can recommend EEO program improvements to the component head before he or she responds to EEOC’s feedback letters. In addition, CRCL could use its existing practices to discuss EEOC’s feedback letter with components, such as midyear update meetings and monthly council meetings. However, CRCL officials stated they did not meet to discuss EEOC’s feedback letters with components in 2018. EEOC officials told us they send component feedback letters to both the component and CRCL, and invite CRCL officials to participate in component site visits. They also explained that DHS could be found noncompliant if a component’s EEO program does not comply with EEOC guidance. Developing policies and procedures for responding completely and timely to EEOC’s feedback letters may help the department comply with EEOC guidance. While DHS officials told us that ensuring DHS components’ compliance with MD-715 guidance is EEOC’s responsibility, EEOC officials explained that DHS’s responsibility equaled the responsibility that EEOC has to ensure DHS components’ compliance with MD-715 guidance. In addition, MD-715 guidance states that federal agencies, such as DHS, have the primary responsibility to ensure nondiscrimination in employment. Our prior work has found that an agency can benefit from periodically evaluating its organizational structure. Additionally, Standards for Internal Control in the Federal Government states that agency management should establish an organizational structure to achieve the agency’s objectives. According to these standards, an effective management practice for attaining this outcome includes periodically evaluating the organizational structure to ensure that it meets its objectives. As we previously discussed, EEOC found areas of noncompliance in the EEO programs of six out of eight DHS components, and two of the six components did not have plans to correct all of the areas of noncompliance until we asked about them during our review. While CRCL officials told us that they lack authority to certify that components’ MD-715 reports comply with MD-715 guidance, EEOC guidance states that an agency’s EEO Director ultimately is responsible for ensuring equal opportunity throughout the entire agency. EEOC guidance allows DHS components to report to either the Deputy Officer for EEO and Diversity or the Secretary of Homeland Security. However, DHS has not taken steps—in consultation with EEOC and other agencies as relevant—to analyze options to address EEO program management weaknesses, such as analyzing alternatives for granting additional authorities to the Deputy Officer for EEO and Diversity to ensure DHS components comply with MD-715 guidance, and assessing benefits and trade-offs of each alternative. In the absence of these steps, DHS may not be positioned to effectively manage its EEO program. Conclusions As the third largest U.S. government department, the challenges DHS has faced to fully implement effective EEO programs may result in widespread negative consequences, including monetary expenses borne by the agency in connection with workplace disputes and decreased morale and productivity resulting from the ineffective and inefficient use of human capital resources. MD-715 requires DHS and its components to report annually on the status of their EEO activities and include plans that set forth steps they will take to correct deficiencies or improve EEO efforts. From fiscal years 2014 through 2017, DHS and its components have reported deficiencies in their EEO programs and identified EEO barriers in their workforces. We found areas for improvement in DHS and its components’ EEO programs that could help ensure success and compliance with MD-715. Specifically, DHS does not have complete performance metrics for the department’s EEO program, including a mechanism for tracking progress towards eliminating barriers. Developing performance metrics for the department’s EEO program could help improve progress in eliminating identified EEO barriers. In addition, DHS and its components reported that they lack action plans for addressing deficiencies in their MD-715 reports. Developing policies and procedures could help DHS component EEO Directors correct deficiencies in their EEO programs. DHS and its components also reported that areas of their EEO programs do not have sufficient staffing to successfully implement EEO activities. Developing formal staffing models could help DHS and its components better assess their resource needs to correct their deficiencies and eliminate their barriers. Further, from fiscal years 2014 through 2017, EEOC found areas of noncompliance in DHS and its component EEO programs. Without developing policies and procedures for responding completely and timely to EEOC’s feedback letters, DHS components may not correct areas of noncompliance and remain at risk of financial penalties and lost employee potential. Finally, DHS has not taken steps to address the key EEO program management weaknesses. Analyzing options for granting additional authorities to the Deputy Officer for EEO and Diversity can help position DHS to ensure its components are complying with MD-715 guidance. The commitment of DHS’s leadership is essential to successfully addressing these issues. By focusing leadership attention on developing performance metrics, policies and procedures, and staffing models, DHS and its components can help improve their EEO programs by making progress towards eliminating barriers, obtaining sufficient staffing, and addressing areas of noncompliance. Recommendations We are making the following six recommendations to DHS: 1. The Secretary of Homeland Security should develop performance metrics for the department’s EEO program including a mechanism for tracking progress towards eliminating barriers. (Recommendation 1) 2. DHS component EEO Directors, in consultation with the Deputy Officer for EEO and Diversity, should develop policies and procedures to help ensure that their component EEO programs have action plans for addressing deficiencies in their MD-715 reports. (Recommendation 2) 3. The Deputy Officer for EEO and Diversity should develop a formal staffing model for its EEO program. (Recommendation 3) 4. DHS component EEO Directors, in collaboration with the Deputy Officer for EEO and Diversity, should develop component formal staffing models. (Recommendation 4) 5. The Deputy Officer for EEO and Diversity should develop policies and procedures for responding in a complete and timely manner to EEOC’s feedback letters. (Recommendation 5) 6. The Secretary of Homeland Security—in consultation with CRCL and EEOC, and other agencies and components, as relevant—should analyze options for granting additional authorities to the Deputy Officer for EEO and Diversity to ensure DHS components comply with MD-715 guidance, including the authority of the Deputy Officer for EEO and Diversity to certify components’ MD-715 reports. (Recommendation 6) Agency Comments We provided a draft of this report to DHS and to EEOC for review and comment. In its official comments, reproduced in appendix I, DHS agreed with all six of our recommendations, and DHS and EEOC provided separate technical comments to the draft of our report, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Acting Secretary of Homeland Security, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6806 or jonesy@gao.gov, or Christopher P. Currie at (404) 679-1875 or curriec@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. Appendix I: Comments from the Department of Homeland Security Appendix II: GAO Contacts and Staff Acknowledgments GAO Contacts Yvonne D. Jones, (202) 512-6806 or jonesy@gao.gov. Christopher P. Currie, (404) 679-1875 or curriec@gao.gov. Staff Acknowledgments In addition to the contacts named above, Clifton G. Douglas, Jr. (Assistant Director), Luis E. Rodriguez (Analyst-in-Charge), Adam J. Brooks, Carla D. Brown, Andrew Howard, Haley Klosky, and Steven Putansu made key contributions to this report.

Why GAO Did This Study EEOC's Management Directive 715 requires that, to attract and retain top talent, federal agencies are to identify EEO barriers in their workforces and deficiencies in their EEO programs, execute plans to address them, and report annually to EEOC. In 2009, GAO reported that DHS had opportunities to better identify and address barriers to EEO in its workforce, and made recommendations which DHS has taken action to address. GAO was asked to provide an update on DHS's efforts to identify and address barriers to EEO in its workforce. This report examines the steps DHS has taken to (1) identify and address barriers to EEO in its workforce, (2) identify and address EEO program deficiencies, (3) address areas of noncompliance in its EEO program identified by EEOC, and (4) oversee and support component EEO programs. GAO reviewed DHS's and its components' policies, procedures, practices, and reports for their EEO programs for fiscal years 2014 through 2018, interviewed DHS and its component EEO officials, and assessed DHS employee survey results. GAO also reviewed EEOC's feedback on DHS's and its components' EEO programs, and interviewed EEOC officials. What GAO Found The Department of Homeland Security (DHS) has identified barriers to equal employment opportunity (EEO) and has plans to address them, but lacks performance metrics for tracking its progress towards eliminating these barriers. DHS identified three barriers from fiscal years 2014 through 2017: (1) problems with supervision/management, lack of advancement opportunities, and lack of alternate work schedules, among other things, causing higher-than-expected nonretirement separations for white females and several ethnic and racial groups; (2) the geographic location of jobs, which has contributed to low hiring rates of racial groups in certain major occupations; and (3) the medical and physical requirements of various law enforcement positions, such as the ability to engage in moderate to arduous physical exertion, which limit the eligibility of some applicants with targeted disabilities. While DHS reports some improvements in employee engagement and representation of minorities and women, it does not have complete performance metrics, such as the retention rate of women in law enforcement positions. Implementing performance metrics could help DHS better assess its progress in eliminating barriers. DHS and its components have identified various deficiencies in their EEO programs, but lack policies and procedures for developing action plans and formal staffing models to address some deficiencies. DHS components did not have action plans to address nearly half (179 out of 369) of the deficiencies self-reported by all components from fiscal years 2014 through 2017. For example, in fiscal year 2017, four DHS components did not have action plans to ensure that their EEO directors report directly to their agency heads. Developing policies and procedures to help ensure components' EEO programs have action plans for addressing deficiencies could help DHS components better comply with Equal Employment Opportunity Commission (EEOC) requirements. Developing and utilizing formal staffing models for their EEO programs could help DHS and its components to better identify, request, and obtain the staff they need. For example, DHS and its components reported that staffing challenges contributed to some of their EEO program deficiencies, and acknowledged they lack formal models to use their existing staffing to address the deficiencies. DHS has plans to address the nine areas of noncompliance in its EEO program identified by EEOC. For example, in its most recent review of DHS compliance with EEOC requirements, EEOC identified that DHS did not provide complete demographic data on new hires and promotions in its report to EEOC in fiscal year 2016. DHS officials told us that the department plans to report the data by collecting complete data from DHS components in fiscal year 2019. DHS's EEO and human capital offices assist and support DHS components in identifying and addressing EEO barriers. However, the EEO office lacks policies and procedures to ensure components respond timely and completely to areas of noncompliance identified in EEOC feedback letters. Additionally, DHS EEO officials said they lack authority to ensure components' compliance with EEOC requirements. Without addressing these issues, DHS may not be effectively positioned to manage its EEO program. What GAO Recommends GAO is making six recommendations, including: develop performance metrics for the department's EEO program; develop DHS and component formal staffing models; and analyze options for granting additional authorities to the most senior official for EEO and Diversity. DHS concurred with our six recommendations and described actions the department plans to take to address them.

gao_GAO-19-717T

gao_GAO-19-717T_0

Background Education administers federal student aid programs, including the William D. Ford Federal Direct Loan (Direct Loan) program, through the Office of Federal Student Aid. Only Direct Loans are eligible for the PSLF program and the temporary expanded process. Under the Direct Loan program, Education issues and oversees federal loans provided to students, and contractors service these loans. Education currently contracts with nine loan servicers that each handle the billing and other services for a share of the over $1 trillion in outstanding student loans provided through the Direct Loan program. Borrowers interested in pursuing loan forgiveness under PSLF, or the temporary expanded process, must have their loans eventually transferred to Education’s sole PSLF loan servicer in order to proceed. This designated PSLF servicer handles day-to-day activities associated with the PSLF program and the temporary expanded process, which include responding to borrower inquiries, making preliminary determinations about whether borrowers’ employment and loans qualify for loan forgiveness, and processing loan forgiveness applications. The PSLF program and the temporary expanded process provide eligible borrowers with forgiveness on the remaining balance of their Direct Loans after they have met all program requirements. To receive forgiveness for a loan, borrowers are required to be employed in a qualifying public service job for 10 years when making 120 qualifying payments. Borrowers must also be employed in a qualifying public service job at the time they apply for forgiveness, and at the time they receive forgiveness for their loans. Although there are some differences in the eligibility requirements for PSLF and the temporary expanded process, borrowers are generally required to: Work full-time for at least 10 years at a public service organization, a government organization, agency, or entity at any level (federal, state, local, or Tribal); a nonprofit, tax exempt organization (under section 501(c)(3) of the Internal Revenue Code); or another private nonprofit organization that provides certain public services. Not be in default on their loans. Make 120 on-time monthly loan payments for the full amount due on their bill. These monthly payments do not need to be consecutive. Repay their loans through a qualifying repayment plan. The PSLF program generally requires borrowers to repay their loans through one of the eligible income-driven repayment plans or the Standard repayment plan. The temporary expanded process allows borrowers to qualify for loan forgiveness if they are participating in several additional types of repayment plans, including the Graduated repayment plan, Extended repayment plan, Consolidated Standard repayment plan, and Consolidated Graduated repayment plan. In addition, for the temporary expanded process, the following two payments generally must be at least as much as the borrower would have paid under an income-driven repayment plan: (1) the payment made immediately prior to applying for the temporary expanded process, and (2) the payment made 12 months prior to applying for the temporary expanded process. There are key differences in the processes for applying for loan forgiveness under the PSLF program versus the temporary expanded process (see table 1). Approximately 99 Percent of Borrower Applications for Loan Forgiveness through Both the PSLF Program and the Temporary Expanded Process Have Been Denied Despite broad borrower interest in the PSLF program and the temporary expanded process, very few borrowers have been granted loan forgiveness. A large number of borrowers are pursuing the PSLF program, but our 2018 analysis found that Education had denied about 99 percent of borrowers that applied for loan forgiveness through the PSLF program during the first 8 months that Education was accepting applications (September 2017 through April 2018), according to data from the PSLF servicer. According to Education’s most recent publically released PSLF program data through March 2019, PSLF program denial rates have continued to hover around 99 percent since our 2018 review. Of the 76,002 loan forgiveness applications that had been processed, the PSLF servicer had denied 75,138 (99 percent), as of March 2019. According to data as of March 2019, close to one-half of the PSLF program loan forgiveness applications the PSLF servicer had processed were denied because the borrower had not yet made 120 qualifying payments. The other most common reasons PSLF program applications were denied were because of missing information on the application or because the borrower did not have qualifying federal loans. For borrowers that have been approved, Education had forgiven almost $31 million in outstanding student loans, an average of more than $59,000 per approved borrower. Denial rates are also very high for the temporary expanded process. We recently reported that from May 2018 through May 2019, Education had denied 99 percent of the completed requests from about 40,000 borrowers (see fig. 1). The majority of requests borrowers submitted for the temporary expanded process were ineligible for consideration and were therefore denied because the borrower had not previously submitted an application for the PSLF program, according to data from the loan servicer. For the 1 percent of applications that were approved from May 2018 through May 2019, Education had granted almost $27 million in loan forgiveness under the temporary expanded process, totaling about 4 percent of the $700 million appropriated funds, according to our 2019 report. Borrowers received an average of about $41,000 in loan forgiveness. The high denial rates for the PSLF program and temporary expanded process suggest that many borrowers are confused by the requirements. In our 2018 report, we noted that officials from the PSLF servicer said that borrowers were frequently confused by the PSLF program requirements related to qualifying loans, employment, repayment plans, and payments. PSLF servicer officials also said that borrowers were sometimes unaware that they were not on a PSLF-qualifying repayment plan or that forbearance, deferment, and loan consolidation would affect their qualifying payments. For example, the Consumer Financial Protection Bureau has reported that borrowers have complained of spending years making payments, believing they were making progress towards PSLF loan forgiveness, and then learning that they were not eligible. Similarly, in our 2019 report on the temporary expanded process, we noted that officials from Education, the PSLF loan servicer, and representatives from selected organizations representing student borrowers all said that the requirement to submit an application for the PSLF program to be eligible for the temporary expanded process can confuse borrowers. Shortcomings in the Information Education Provides to the PSLF Servicer Increase the Risk of Administrative Errors We have previously reported on how shortcomings in the information Education provides to the PSLF servicer has resulted in uncertainty about PSLF program requirements and increased the risk of potential errors in borrower eligibility determinations. To address these issues, we have made three recommendations to Education to provide the servicer with comprehensive guidance and instructions, additional information on qualifying employers, and standardized prior payment information (see table 2). Education agreed with these recommendations and has taken some actions, but has not yet fully implemented them. Piecemeal Guidance and Instructions In our 2018 report, we found that Education does not have a comprehensive document or manual to provide the PSLF servicer with guidance and instructions. This made it difficult to effectively administer the PSLF program and provide consistent service to borrowers, according to PSLF servicer officials. We reported that Education’s guidance and instructions to the PSLF servicer are dispersed in a piecemeal manner across multiple documents, including Education’s original contract with the servicer, multiple updates to the contract, and hundreds of emails. As a result, PSLF servicer officials said that their staff were sometimes unaware of relevant PSLF program guidance and instructions in emails provided by Education, which creates a risk that some policy updates will be overlooked and not consistently implemented. The absence of a central, authoritative source of PSLF guidance and instructions creates a risk of differing interpretations and inconsistent implementation. It also makes it difficult to maintain program continuity in the event of staff turnover or if Education decides to contract with a new servicer to administer the PSLF program. Federal internal control standards state that agencies should communicate information to those who need it, in a form that enables them to carry out their responsibilities. Around the time our 2018 report was issued, Education officials told us they planned to develop a comprehensive PSLF servicing manual, but they did not have a timeline for completing it. In response, we recommended that Education develop a timeline for issuing a comprehensive guidance and instructions document for PSLF servicing. Education agreed with this recommendation and reported in September 2019 that it was continuing its efforts to improve and streamline guidance for the PSLF servicer. While Education said it is working on developing its comprehensive PSLF servicing manual, it does not yet have a timeline for how it will complete this manual and has pushed back the estimated implementation date for this recommendation to 2020. To help ensure that program requirements are applied consistently by the PSLF servicer, we continue to believe that Education should fully implement this recommendation. Limited Information on Qualifying Employers In 2018, we reported that Education had not provided the PSLF servicer with a definitive source of information for determining which employers qualify a borrower for PSLF loan forgiveness. Instead, Education had identified some data sources the PSLF servicer can use to determine whether borrowers are working for qualifying employers. However, we found that these sources were not comprehensive, and that PSLF servicer officials said they sometimes had to consult other sources that have significant limitations. For example, PSLF servicer officials told us they used an online directory of nursing home facilities to help determine if certain nursing homes were nonprofit employers. However, this website explicitly stated that it did not guarantee that the information it provided was accurate or current. Federal internal control standards state that agencies should communicate the necessary quality information to those who need it, and PSLF servicer officials said that having additional information would help them assess employers more quickly and minimize the risk of inaccurate decisions. Borrowers would also benefit from additional information about qualifying employers, according to PSLF servicer officials, in part because it would help them make better informed employment decisions. Our 2018 report recommended that Education provide additional information to the PSLF servicer and borrowers to enhance their ability to determine which employers qualify for PSLF. Education agreed with this recommendation, and said it planned to incorporate qualifying employer information into an online PSLF Help Tool. As of September 2019, Education reported that it had incorporated a feature into its online PSLF Help Tool to help borrowers determine if their employer fits within general eligibility criteria. However, Education said more specific information to help the PSLF servicer make employer eligibility determinations and an employer database will not be available until 2020. We believe that if Education fully implements this recommendation to provide the servicer with more definitive employer information, it would help reduce the risk of errors in assessing employer eligibility for PSLF. Inconsistencies in Prior Loan Payment Data In our 2018 report, we found that Education does not ensure that the agency’s other loan servicers provide the PSLF servicer consistent information on borrowers’ prior loan payments, which could increase the risk of qualifying payments being miscounted for the PSLF program. PSLF servicer officials said inconsistencies in the information provided by other loan servicers make it challenging to determine whether borrowers are on qualifying repayment plans or making qualifying payments. Officials with Education and the PSLF servicer said that these inconsistencies increase the risk of miscounting qualifying payments. This is contrary to federal internal control standards, which state that agencies should use quality information. Our 2018 report recommended that Education standardize the payment information that the PSLF servicer receives from other loan servicers to ensure the PSLF servicer obtains more consistent and accurate payment information. Education agreed with this recommendation and stated that efforts were underway to improve the consistency of payment information exchanged between servicers. As of September 2019, Education reported that it is planning to standardize this loan payment data by spring 2020. If Education implements this recommendation, we believe it would reduce the potential risk of qualifying PSLF payment count errors moving forward. Education Can Provide Better Service to Borrowers by Expanding Outreach, Streamlining Processes, and Sharing Critical Information with Borrowers We have previously reported on how unclear processes and a lack of information about the PSLF program and the temporary expanded process could contribute to borrower confusion and high denial rates. We have also reported that borrowers can face challenges detecting any errors in payment counts for the PSLF program and with contesting eligibility determinations for the temporary expanded process. To address these issues, we have made five recommendations to Education to improve service to borrowers by expanding outreach, streamlining processes, and providing information to help borrowers catch and resolve errors (see table 3). Education agreed with these recommendations, but has not yet taken sufficient actions to fully implement them. Borrower Outreach Education uses several methods to inform borrowers about the PSLF program and temporary expanded process requirements, including through its website and webinars. Congress also appropriated $4.6 million in 2018 for Education to conduct outreach to borrowers about PSLF, including the temporary expanded process. However, our recent work has found several areas in which the agency’s outreach activities related to the temporary expanded process are limited. While Education and PSLF loan servicer officials told us that they primarily direct borrowers to online sources to inform them about requirements for the temporary expanded process, we found that the agency does not include information about the temporary expanded process in key online sources. For example, according to agency officials, one of Education’s primary PSLF outreach mechanisms—the online PSLF Help Tool, which the agency launched in December 2018—does not include any information about the temporary expanded process. Officials from Education and the PSLF servicer stated that integrating information about the temporary expanded process into the online PSLF Help Tool would be beneficial for borrowers and would reduce confusion about the temporary expanded process. In addition, our 2019 report found that while all nine of Education’s loan servicers’ websites contain some information on the PSLF program, none of them (other than the PSLF loan servicer) included information about the temporary expanded process on their websites or provided a link to Education’s website specific to the temporary expanded process. Education officials told us that only the PSLF servicer is required to have information about the temporary expanded process on its website; however, other loan servicers may also serve borrowers who are potentially eligible but may be unaware of the temporary expanded process. This limited outreach to borrowers about the temporary expanded process reduces the likelihood that borrowers are able to take advantage of this opportunity. Further, federal internal control standards state that management should externally communicate the necessary quality information to achieve the entity’s objectives. To improve Education’s borrower outreach about the temporary expanded process, our September 2019 report recommended that Education include information about the temporary expanded process in its online PSLF Help Tool and that Education require all loan servicers to provide information about the temporary expanded process on their websites. Education agreed with both of these recommendations, and stated that it would take steps to address them. If Education implements these two recommendations, we believe it would help the department provide better service to borrowers by raising awareness of the temporary expanded process and requirements. Streamlining Processes In September 2019, we reported that Education’s process for requesting loan forgiveness through the temporary expanded process is not clear to borrowers and may contribute to high denial rates. In particular, the requirement that borrowers must have already submitted a separate PSLF application in order to be eligible for loan forgiveness through the temporary expanded process can confuse borrowers. Borrowers currently must submit a separate PSLF application, even if they know it will be denied, before Education will consider their request for forgiveness through the temporary expanded process. Education officials acknowledged that the majority of requests for the temporary expanded process come from borrowers who have not first submitted a PSLF application. Similarly, our September 2019 report found that 71 percent of the denied requests were denied because the borrower had not submitted a PSLF application. Officials from the PSLF loan servicer said that borrowers who called were frequently confused when they received a denial for the temporary expanded process based on the fact that they had not first submitted an application for the PSLF program. This lack of a borrower-friendly process complicates the path towards loan forgiveness and does not align with Education’s strategic plan objective to improve the quality of service to customers. To address this issue, our 2019 report recommended that Education streamline the process for borrowers to request loan forgiveness through the temporary expanded process by integrating the request for temporary expanded process consideration into the PSLF application, eliminating the need for borrowers to submit a separate PSLF application prior to consideration. Education agreed with this recommendation and stated that it will integrate requests for the temporary expanded process into the PSLF application as part of its ongoing initiative to overhaul its online portal for student loan borrowers. Implementation of this recommendation would improve service to borrowers by making the process easier and less confusing. Information to Help Borrowers Identify and Remedy Potential Errors In 2018, we reported that although Education and PSLF servicer officials acknowledged the risk of miscounting qualifying payments, the PSLF servicer did not provide borrowers with sufficient information to easily identify PSLF program errors. Officials with the PSLF servicer said they rely on borrowers to catch any payment counting errors resulting from issues with information provided by other loan servicers. As we reported, the PSLF servicer provided borrowers with aggregate counts of qualifying payments, which are useful for helping borrowers track their progress, but did not provide borrowers with enough detail to check the servicer’s counts and identify prior payments that the servicer may have missed. This is also contrary to federal internal control standards which call for communicating necessary information to external parties. Our 2018 report recommended that Education ensure that borrowers receive sufficiently detailed payment information from the PSLF servicer to be able to identify any errors in the servicer’s counts of qualifying payments. Education agreed with this recommendation and stated that efforts were underway to standardize the payment count information that is provided to borrowers. As of September 2019, Education reported that it is reviewing communications from the PSLF servicer to ensure that borrowers receive sufficiently detailed information regarding payment counts and payment history and that this review will be completed by September of 2020. To help borrowers detect potential payment counting errors that could ultimately affect their eligibility for the PSLF program, we believe Education should implement this recommendation and provide borrowers with more detailed qualifying payment information. Further, our 2019 report on the temporary expanded process found that Education does not provide complete information to borrowers about options they have to contest payment counts or other aspects of the eligibility determination process. An Education official told us that while there is no formal process for borrowers who are dissatisfied with their temporary expanded process determinations to contest the determination, borrowers do have additional options for addressing concerns, such as an additional review by the PSLF servicer, or a complaint to Education’s Federal Student Aid Feedback System or Ombudsman. Education officials told us that the agency does not provide information about these options in its denial letters or on its website for the temporary expanded process, noting that borrowers could find this information at the bottom of Education’s Federal Student Aid main website. However, borrowers may not know where to find this information should they choose to contest their temporary expanded process determination, because this information is not effectively communicated to them in accordance with federal internal control standards. To address this, our 2019 report recommended that Education provide borrowers with more information on the website for the temporary expanded process and in the servicer’s denial letters about options available to borrowers should they wish to contest the servicer’s decision. Education agreed with this recommendation and stated that it would add information about the options borrowers have to contest temporary expanded process decisions to relevant websites and denial letters. Implementing this recommendation will increase the likelihood that borrowers with valid concerns about the temporary expanded process will have them appropriately resolved. In conclusion, my statement has highlighted several actions Education could take to strengthen the PSLF program and the temporary expanded process to deliver on the promise the federal government has made to borrowers pursuing careers in public service. Education is responsible for establishing an administrative structure for the loan servicer, but more than 10 years after the PSLF program was first established, Education has not provided the loan servicer with a comprehensive source of guidance and instructions on how to operate the PSLF program, and could provide additional information to help ensure that eligibility determinations are being made correctly. Education is responsible for ensuring that borrowers are aware of and understand programmatic requirements. However, the high denial rates for the PSLF program and its temporary expanded process suggest that borrowers are still confused. It is also important for Education to maintain borrower confidence, but the department has not provided critical information to borrowers to help them remedy potential errors. Large numbers of borrowers have pursued careers in public service, sometimes at lower pay than in the private sector, with the hope of one day achieving loan forgiveness through the PSLF program. They have often had to navigate the PSLF program requirements with a lack of sufficient information from Education only to be denied 10 years later when they applied for loan forgiveness because their prior years of employment or loan payments did not qualify. In addition, some borrowers who were denied may not be aware that they might be eligible for loan forgiveness through the temporary expanded process, potentially missing out on this temporary opportunity. Education needs to take action to better serve these borrowers and help smooth their long road towards loan forgiveness. Education has not yet taken action to fully implement the eight recommendations discussed in this testimony. We continue to believe that implementing these eight recommendations would strengthen program administration, improve service to borrowers, and help to fulfill the original goal of encouraging individuals to enter and continue in public service employment. We will continue to monitor Education’s efforts in these areas. Chairwoman Davis, Senior Republican Smucker, and Members of the Subcommittee, this completes my prepared statement. I would be pleased to respond to any questions that you may have at this time. GAO Contact and Staff Acknowledgments If you or your staff have any questions about this testimony, please contact Melissa Emrey-Arras, Director of Education, Workforce, and Income Security, at (617) 788-0534 or emreyarrasm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this testimony include William Colvin (Assistant Director), Nora Boretti (Analyst-in-Charge), Linda Collins, and Aaron Karty. Additional support was provided by James Bennett, Deborah Bland, Alex Galuten, Lara Laufer, Sheila R. McCoy, and Jessica Orr. 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.

Why GAO Did This Study The PSLF program was established in 2007 and forgives borrowers' remaining federal student loan balances after they have made at least 10 years of qualifying loan payments while working in public service. Starting in September 2017, the first borrowers potentially became eligible for the PSLF program and began applying to have their loans forgiven. In 2018, Congress appropriated $700 million to temporarily expand the PSLF program for certain borrowers who initially did not qualify for the program. This statement—based on GAO's reports issued in September 2018 ( GAO-18-547 ) and September 2019 (GAO-19-595 )—discusses (1) the extent to which borrowers' requests for loan forgiveness through PSLF and the temporary expanded process have been approved or denied, (2) the extent to which Education provides the PSLF servicer with sufficient information to administer the program, and (3) opportunities for improving service to borrowers. What GAO Found A large number of borrowers are pursuing the Public Service Loan Forgiveness (PSLF) program, but the Department of Education (Education) has denied about 99 percent of loan forgiveness applications as of March 2019. Close to one-half of these applications were denied because the borrowers had not yet made the required 120 qualifying monthly loan payments. As of May 2019, Education has also denied 99 percent of loan forgiveness requests made through the temporary expanded process, which is intended for borrowers who did not initially qualify for the PSLF program. In its 2018 report, GAO found that shortcomings in the information Education provided to the loan servicer that administers the PSLF program increased the risk of administrative errors. For example, Education had not provided the PSLF servicer with a definitive source of information for determining which employers qualify. GAO made three recommendations to Education to address these issues (see table below). Education agreed with these recommendations and has taken some actions, but has not yet fully implemented them. In its 2018 and 2019 reports, GAO found that Education can provide better service to borrowers by expanding outreach, streamlining processes, and sharing critical information with borrowers. For example, GAO found that Education does not include information for borrowers about the temporary expanded process in key online sources. GAO made five recommendations to Education to address these issues with the PSLF program and the temporary expanded process (see table below). Education agreed with these recommendations, but has not yet fully implemented them. What GAO Recommends GAO has made eight recommendations to Education to improve its implementation of the PSLF program and its temporary expanded process. Education agreed with GAO's recommendations. As of September 2019, GAO continues to believe that actions are necessary to fully implement all of the recommendations discussed in this statement.

gao_GAO-20-418

gao_GAO-20-418_0

Background In June 2018, DOD’s Chief Management Officer established the Community Services Reform Task Force to perform a business case analysis to determine whether consolidating the defense resale organizations would result in efficiencies. The task force conducted its work from July 2018 through November 2018, during which it collected financial and other data from the four resale organizations and conducted workshops with subject matter experts from the resale organizations. In November 2018, the military departments were given an opportunity to review the business case analysis, provide comments, and indicate whether they concurred with the analysis. In its business case analysis, the task force recommended consolidating the four defense resale organizations into a single organization. The task force stated that consolidation would eliminate duplication that currently exists across the resale organizations and increase the competitiveness, or financial viability, of defense resale, which has seen sales declines in recent years. Specifically, the task force recommended that a single chief executive officer or director be responsible for leading the organization and report to a single board of directors. The task force also recommended that separate leadership positions for commissary operations and exchange operations be established, and that a chief administrative officer manage the business functions that are common to the current resale organizations, such as information technology (IT), human resources, marketing, and finances. Figure 1 shows the task force’s recommended organizational chart for the consolidated resale organizations. The task force estimated that the time frame for consolidating the four defense resale organizations would be 5 years for implementation, and that consolidation would result in “net savings” (i.e., estimated savings minus estimated costs) ranging from about $690 million to $1.3 billion during the first 5 years, followed by annual net savings of approximately $390 million to $670 million every year thereafter. Specifically: Estimated savings: The task force estimated that consolidating the four defense resale organizations would result in savings in three areas: (1) reduction of the cost of goods sold in the commissaries and the exchanges; (2) reduction of the cost of goods and services that are not sold but are necessary for operating stores (e.g., plastic shopping bags and custodial services); and (3) reduction of payroll costs by eliminating redundant personnel. According to the task force, consolidation would result in estimated savings of $1.4 billion to $2.1 billion over the first 5 years, followed by annual savings of $470 million to $750 million. Estimated costs: The task force estimated that consolidating the defense resale organizations would result in costs from four areas: (1) development of new, common IT systems; (2) severance pay for separating employees and retention bonuses to incentivize employees to remain; (3) operation of a transformation management office, supported by private contractors, to implement the consolidation; and (4) costs to convert DeCA to a non-appropriated fund organization. According to the task force, consolidation would result in estimated costs of $700 million to $810 million over the first 5 years, followed by annual costs of $80 million. DOD’s Chief Management Officer in March 2019 and the Deputy Secretary of Defense in August 2019 both approved the results of the business case analysis and directed that plans be made for consolidation, pending congressional action to remove the statutory prohibition on consolidating the commissary and exchange systems. DOD’s Business Case Analysis Supporting Defense Resale Consolidation May Not Provide Reliable Savings and Cost Estimates The task force may have overestimated the expected savings from reducing the “cost of goods sold” (i.e., the cost of purchasing products that are resold in commissaries or exchanges) and underestimated the expected costs from IT consolidation and headquarters relocation. DOD’s Task Force May Have Overestimated the Expected Savings to Be Achieved from Reducing the Cost of Goods Sold The task force estimated that most of the savings (i.e., about 70 percent annually) to be achieved from consolidating the four defense resale organizations would result from reducing the cost of goods sold. According to the task force’s business case analysis, retailers often pay different costs for identical products, and mergers are an opportunity for retailers to compare costs across a larger combined organization and make decisions that maximize savings. In the case of a consolidated defense resale organization, the task force stated in its business case analysis that savings could be achieved by implementing what the task force called category management reforms and by obtaining the lowest cost for identical products sold by both commissaries and exchanges. Task force officials added that one board of directors and one chief executive officer overseeing the consolidated resale organization would be more likely to achieve savings than the current, individual boards of directors and chief executive officers of the resale organizations. GAO-18-151SP. DeCA and one or more exchange organization, or Two or more of the exchange organizations (and not DeCA). In addition to overlap in identical products sold, task force officials told us that savings from category management reforms are dependent, in part, on the amount of overlap in vendors that sell products to the resale organizations. Specifically, task force officials stated that there are opportunities to reduce cost of goods sold through negotiations with vendors that sell items to both DeCA and the exchange organizations. For example, task force officials stated that the consolidated organization could negotiate better prices with a vendor that sells family-size items to DeCA and single-size items to the exchange organizations, even though those items are not identical. Industry Benchmarks The industry benchmarks used by the task force are based on proprietary data gathered and owned by Boston Consulting Group based on its experience working with mergers and category management reforms in the private sector retail industry. These benchmarks were presented as a percentage of cost of goods sold; specifically, the task force estimated that savings from obtaining the lowest cost for identical items were from 1 to 1.5 percent of the cost of goods sold, and savings from category management efforts were from 2.5 to 4 percent of the cost of goods sold. We did not review and evaluate the underlying data that were used to develop the proprietary benchmarks. Based on this information, the task force calculated the estimated savings that would result from reducing the cost of goods sold by multiplying the total cost of goods sold for all four resale organizations in fiscal year 2017 ($9.5 billion) by industry benchmarks developed by Boston Consulting Group (see sidebar for more information on these benchmarks). This calculation showed an estimated annual savings of $329 million to $517 million from reducing the cost of goods sold. However, additional information from the task force suggests this savings estimate may be overstated because there is limited overlap in the products DeCA sells (i.e., groceries and household goods) and the products the exchange organizations sell (i.e., goods and services similar to retail stores). According to the task force, about $2.2 billion of DeCA’s cost of goods sold in fiscal year 2017 were for products also sold by at least one of the exchange organizations, which is equivalent to about 23 percent of the total cost of goods sold for the four resale organizations. This differs from the data provided in the business case analysis, which stated that 62 percent of the total cost of goods sold was for identical products sold by two or more resale organizations; however, that figure also includes products sold by two or more exchange organizations and not DeCA (we further discuss product overlap among the exchange organizations below). Given the more limited product overlap between DeCA and the exchange organizations, it is unclear whether using the total cost of goods sold for all four resale organizations as the basis for estimating savings was appropriate. Additionally, the business case analysis did not fully identify the amount of vendor overlap that exists between DeCA and the three exchange organizations, but the data that were provided in the business case analysis suggest that limited vendor overlap exists. Specifically, the business case analysis provided data for 10 vendors that sell to the defense resale organizations, but those vendors represent less than 20 percent of the cost of goods sold to DeCA and the exchange organizations in fiscal year 2017. Further, only 5 of the 10 vendors identified in the business case analysis sold goods to both DeCA and the exchange organizations, and their cost of goods sold accounted for about 10 percent ($972 million) of the total cost of goods sold for the four resale organizations ($9.5 billion). Based on these data, the extent of vendor overlap between DeCA and the exchange organizations—and, as a result, how much can be saved through category management reforms by consolidating DeCA and the exchange organizations—is unclear. Although the task force stressed the importance of a conservative estimate in both its business case analysis and in meetings with us, our assessment of the assumptions and methodology for estimating savings from the cost of goods sold found that a more conservative approach could have been used to better ensure estimated savings were not overstated. For example, one method could have been to multiply the benchmarks by the cost of goods sold for just the three exchange organizations (about $5.5 billion in fiscal year 2017, per the task force), as data provided by the task force indicate that about 67 percent of the cost of goods sold for the exchange organizations in fiscal year 2017 were for identical products sold by at least two exchange organizations. Another method, which task force officials suggested after we shared our concerns about their methodology, could have been to multiply the benchmarks by the cost of goods sold for the exchange organizations, plus the portion of DeCA’s cost of goods sold that overlaps with at least one exchange organization (about $2.2 billion in fiscal year 2017, per the task force). Either method would be more conservative than the one adopted in the business case analysis and would yield a savings estimate that is about 20 to 40 percent lower, but would be more consistent with the task force’s assertion that consolidation savings are dependent on the amount of overlap among the merging organizations. DOD policy states that an economic analysis should base its analysis of benefits on facts and data whenever possible. Additionally, our Assessment Methodology for Economic Analysis states that an economic analysis should examine the effects of an action by considering relevant alternatives and justifying what the world would be like under each alternative; describe and justify the analytical choices, assumptions, and data used; and assess how plausible adjustments to each important analytical choice and assumption affect the estimates of savings. Ensuring that the estimates for cost of goods savings are accurate is particularly important, as they account for approximately 70 percent of the task force’s overall savings estimate from consolidation. However, the task force did not fully identify and analyze in its business case analysis how many identical products are sold by both DeCA and the exchange organizations or how many vendors sell products to both DeCA and the exchange organizations. According to task force officials, they did not provide data on product overlap between DeCA and the exchange organizations because it would not change their savings methodology or estimate, and they did not provide more information on vendor overlap because of the proprietary nature of that data. However, the amount of product and vendor overlap that exists across the four resale organizations will have a direct effect on the amount of savings to be achieved from consolidation, as acknowledged by the task force. Without the task force reassessing the approach it used to estimate savings from the cost of goods sold and, if necessary, making adjustments to those estimates, decision makers in DOD and Congress may lack confidence in the reliability of the task force’s savings estimates in the business case analysis and will not have complete information as they consider defense resale consolidation. DOD’s Task Force May Have Underestimated the Expected Costs of Consolidating the Four Defense Resale Organizations Based on our analysis of the business case analysis, we found that DOD’s task force may have underestimated the expected costs of consolidating the four defense resale organizations in two areas: (1) the development of new, common IT systems and (2) the location of a new headquarters for the consolidated organization. Task Force May Have Underestimated the Cost of Consolidating Defense Resale Organizations’ IT Systems The task force estimated in its business case analysis that most of the costs (i.e., about 50 percent annually) of consolidating the four defense resale organizations will result from developing new, common IT systems to support the consolidated organization. In the business case analysis, the task force stated that it worked with the four resale organizations to calculate a cost estimate of $292 million to $352 million for developing five types of IT systems that are needed for the consolidated organization: merchandising, store inventory management, financial management and general ledger, transportation and logistics, and ecommerce. According to the business case analysis, the task force’s cost estimates for developing new, common IT systems for the consolidated organization were to be based on data provided by the resale organizations on recent or projected costs for replacing similar systems or performing major upgrades to existing systems, when available. For example, the task force’s estimate for the merchandising system was approximately $115 million, which the business case analysis stated is based on a $35 million estimate provided by AAFES, a $23.5 million estimate provided by NEXCOM, a $15 million estimate provided by MCCS, and a $41 million estimate provided by DeCA. However, the task force’s cost estimate for IT consolidation may be understated because it is based, in part, on less expensive minor IT system upgrades and partial replacements, according to the resale organizations. Based on our analysis of information provided to us by the resale organizations, about $140 million (about 40 percent) of the overall IT cost estimate was based on what the resale organizations described as minor upgrades or partial replacements. Specifically, while MCCS confirmed the cost estimates attributed to them in the business case analysis were for total IT system replacement costs, the other three resale organizations—AAFES, NEXCOM, and DeCA— disagreed with the task force’s characterization that all the data used to calculate IT system estimates represented costs for replacements or major upgrades. AAFES told us that the cost estimates cited in the business case analysis for its merchandising, financial management and general ledger, and transportation and logistics systems reflected minor upgrades of specific modules within the overall systems, and the cost to replace or upgrade the entire system would be significantly higher. NEXCOM stated that the cost estimates for upgrading its merchandising, store inventory management, and financial management and general ledger systems were for minor upgrades, not replacements or major upgrades, as stated by the task force. DeCA told us that the estimate for replacing its store inventory management system only represented 1 year of costs, even though DeCA plans to incur replacement costs through at least 2022. According to the task force, the task force and the resale organizations agreed on the methodology for estimating IT costs, and the subject matter experts from the resale organizations provided the cost data used in the business case analysis. However, based on information provided by the resale organizations, it appears that the task force may not have always based its cost estimate on replacement or major upgrade costs— consistent with the key assumption that new IT systems would be developed for the consolidated resale organization—but, rather, used minor upgrade or partial replacement costs in some cases. Specifically, task force officials told us they believed the estimates provided by the resale organizations were too high to be minor upgrades or partial replacements, based on their understanding of IT requirements for resale operations. Further, task force officials stated that their overall IT cost estimate was likely overstated, not understated. For example, they stated that their estimate is higher than what is typically spent for a private sector consolidation of similar size. However, the task force stated that it did not use private sector IT cost estimates in its business case analysis because it determined that public sector IT costs would likely be higher than private sector IT costs. Additionally, task force officials told us that some planned spending on existing IT systems by the four resale organizations would not be necessary as a result of consolidation. However, the business case analysis does not quantify how much future spending could be reduced or factor those reductions into the IT cost estimate. According to the GAO Cost Estimating and Assessment Guide, cost estimates are developed based on assumptions that are defined to establish the baseline conditions the estimate will be built from. Additionally, our Assessment Methodology for Economic Analysis states that an economic analysis should define an appropriate baseline that represents the best assessment of what the world would be like under that alternative. Thus, estimating costs that reflect the baseline conditions is a key step in developing a sound cost estimate. Additionally, we have previously reported that federal IT investments frequently fail or incur cost overruns and schedule slippages. As such, high-quality data are imperative for ensuring proper management and oversight of IT investments. The task force’s IT cost estimate is particularly important, as it represents about 50 percent of the total estimated costs for defense resale consolidation. Until the task force consults with the resale organizations to reassess the methodology for estimating IT costs, decision makers in DOD and Congress may not have a reliable and complete understanding of the estimated costs for the implementation of new, common IT systems, which is information DOD and Congress need as they consider defense resale consolidation. DOD’s Task Force Did Not Provide a Cost Estimate for Relocating the Four Defense Resale Organizations to a New Headquarters Location According to the task force, there would be costs associated with relocating AAFES, NEXCOM, MCCS, and DeCA to a new headquarters location, to include relocating existing personnel, hiring new personnel, and obtaining real estate. Although no relocation options were presented in the business case analysis, task force officials told us there are multiple options for where to locate the headquarters of a consolidated resale organization. One option cited by the task force would be to create a new headquarters in the Washington, DC, area, which would be the most expensive option, as it would likely involve acquiring new real estate and hiring personnel in a high-cost region. Another option cited by the task force would be to locate all exchange operations and staff at the existing AAFES headquarters in Dallas, TX, and maintain commissary operations and staff at the existing DeCA headquarters at Fort Lee, VA. This option would likely be less expensive, as personnel and available real estate are already present at both locations. In January 2020, task force officials also told us that an even less expensive option they might consider is maintaining commissary and exchange headquarters staff at their current locations, but having personnel work for the consolidated organization, rather than for DeCA or the exchange organizations. Despite the potential for relocation costs, the task force did not include a range of cost estimates for different relocation options in its business case analysis. According to task force officials, relocation cost estimates were not included because the headquarters location has not been chosen, and costs will vary widely depending on the chosen location. While actual relocation costs will depend on the chosen headquarters location, this fact does not prevent the task force from presenting a range of cost estimates in advance of that decision being made. Task force officials also said that including relocation cost estimates would not have changed the conclusion of the business case analysis. However, without a range of relocation cost estimates, we were unable to assess the effect of relocation costs on the conclusion of the business case analysis. DOD policy states that an economic analysis should quantify the costs associated with each alternative under consideration whenever possible so that they may be included in the economic analysis calculation. Additionally, our Assessment Methodology for Economic Analysis states that an economic analysis should quantify the important costs, where feasible, to inform decision makers about the economic effects of a proposed action. Without developing and providing a range of relocation cost estimates from the least expensive option to the most expensive, decision makers in DOD and Congress will not be fully informed about the costs of consolidation, which is necessary information for deciding whether to consolidate the four defense resale organizations. Military Departments Officially Concurred with the Business Case Analysis, but DOD Did Not Share Their Accompanying Comments with Congress The military departments officially concurred with the task force’s business case analysis for consolidating the four defense resale organizations. However, the military departments also provided written comments that detailed concerns with fundamental aspects of the business case analysis, to include: the use of proprietary industry benchmarks; estimated savings, costs, and timeline of the consolidation; and the proposed governance structure for the new resale organization. In an April 2019 report to Congress that summarized the business case analysis, DOD stated that the military departments agreed with the consolidation. However, the report did not disclose the military departments’ comments and concerns on the business case analysis, which are relevant as Congress considers defense resale consolidation. In their written comments, the military departments either stated concerns about the consolidation or included critical comments from the exchange organizations—all of which opposed the consolidation. Specifically: The Army concurred with the business case analysis but noted that funding for morale, welfare, and recreation programs must be preserved or increased as a result of the consolidation. In addition, the Army’s comment letter included as an attachment written comments from AAFES, which expressed opposition to the consolidation and detailed concerns with the business case analysis. For example, AAFES stated that the business case analysis relied on unverifiable, proprietary industry benchmarks that overstated the benefits of consolidation, underestimated the costs and time to consolidate, and did not account for recent efforts by the resale organizations to reduce costs by collaborating on a purchasing alliance. The Air Force also concurred with the business case analysis, but noted in its comments that mergers and acquisitions have historically cost more, taken longer, and saved less than originally expected. As a result, the Air Force recommended that a phased implementation plan be followed to guard against financial risk. The Air Force also stated that morale, welfare, and recreation funding currently provided by the exchanges should be maintained while opportunities are examined to reduce the need for appropriated funding. The Navy initially non-concurred with the business case analysis in December 2018. In its comment letter, the Navy stated that the task force’s analysis was flawed beyond repair and included comments from NEXCOM and MCCS that also opposed consolidation. For example, the exchanges’ comments stated that the expected cost savings were overstated, that potential inefficiencies from consolidation were not discussed, and that the resale organizations could achieve cost savings through greater collaboration without the need for consolidation. NEXCOM and MCCS also stated concern that the task force’s savings projection relied heavily on unverifiable industry benchmarks. In addition, MCCS expressed concern that consolidation could result in unexpected costs from separating exchange operations from the rest of MCCS operations, which also include the Marine Corps’ morale, welfare, and recreation and family programs. In January 2019, the Navy changed its position to concur subject to several significant comments and clarifications, and attached a letter detailing comments and concerns similar to those it submitted with its original non-concurrence in December 2018. Officials from the resale organizations further articulated their concerns about the business case analysis when they met with us. For example, resale officials told us they are concerned that savings are overstated, that costs are understated, and that the proprietary benchmarks used by the task force are unverifiable or may not be applicable to the public sector. Exchange officials also stated that they are worried about the effect of consolidation on morale, welfare, and recreation funding generated by the exchanges. Specifically, exchange officials are concerned that exchange revenue currently used for morale, welfare, and recreation programs could be used to pay for consolidation expenses or to reduce the amount of appropriated funds allocated to the commissaries. Despite the concerns detailed in the comments from the military departments and resale organizations, DOD did not include them in its April 2019 report to Congress summarizing the results of the business case analysis. The National Defense Authorization Act for Fiscal Year 2019 required DOD to include in its report the recommendations of the Secretaries of the military departments regarding the plan to consolidate the defense resale organizations. When we asked the task force why DOD did not provide Congress with the comments and concerns cited by the military departments and the resale organizations, officials stated that they were advised by DOD’s Office of General Counsel not to include the comments because they contained information that may have disclosed DOD’s deliberative process. Task force officials also stated that the savings, cost, and timeline estimates in the business case analysis were conservative, and that the proprietary industry benchmarks are based on years of experience by Boston Consulting Group and similar to those cited by prior studies. Regarding the purchasing alliance formed by the resale organizations to reduce their cost of goods sold, task force officials stated they do not believe such efforts to reduce costs will be sustained without a single chief executive officer and board of directors to ensure those efforts continued, as recommended in the business case analysis. Finally, the task force stated in the business case analysis that any savings achieved from consolidation could be used to increase morale, welfare, and recreation funding or reduce appropriations used to fund DeCA, and that decisions on how to allocate savings will be made by the proposed board of directors. According to task force officials, some of the concerns articulated by the military departments and the exchanges could be motivated by a general opposition to consolidation. However, without a more complete reporting of the military departments’ perspectives on consolidation and the task force’s response to those comments, Congress may be unaware of the views various organizations within DOD have regarding the business case analysis, which is relevant information as Congress considers defense resale consolidation. Moreover, fully reporting the comments and concerns could strengthen trust and collaboration among the task force, military departments, and resale organizations on any future resale reforms. Conclusions Four defense resale organizations currently operate about 240 commissaries (operated by DeCA) and 2,500 exchange facilities (operated by AAFES, NEXCOM, and MCCS) worldwide to provide reduced-priced groceries and retail goods and services to DOD servicemembers, their families, and retirees. DeCA operations are funded in part by appropriations, which have totaled approximately $1.3 billion in recent years. By law the commissary and exchange organizations must be operated separately. In November 2018, a DOD task force completed a business case analysis and concluded that consolidating the four defense resale organizations into a single organization would result in several hundred million dollars in annual cost savings. However, we found that the task force’s projected savings from reducing the cost of goods sold may be overestimated, and that projected costs for IT development and headquarters relocation may be underestimated. Further, while the military departments concurred with the task force’s recommendation to consolidate, DOD did not fully share their comments and concerns about the business case analysis with Congress. DOD’s proposed consolidation will cost several hundred million dollars, take years to implement, and involve multiple DOD organizations. Given the cost and complexity of the proposed defense resale consolidation, DOD can ensure that Congress has the reliable information it needs to consider consolidation by reviewing and updating savings and cost estimates and sharing comments and concerns from the military departments. Recommendations for Executive Action We are making the following four recommendations to DOD. The Secretary of Defense should ensure that the DOD Chief Management Officer direct the task force to reassess its approach to estimating savings from cost of goods sold—to include reassessing its use of the cost of goods sold for all four defense resale organizations rather than, for example, just for the three exchange organizations—and make any necessary adjustments to its savings estimates for consolidation and provide that updated information to Congress. (Recommendation 1) The Secretary of Defense should ensure that the DOD Chief Management Officer direct the task force, in consultation with the resale organizations, to reassess its methodology for estimating IT costs of consolidation, and make any necessary adjustments to its range of IT cost estimates and provide that updated information to Congress. (Recommendation 2) The Secretary of Defense should ensure that the DOD Chief Management Officer direct the task force to develop a range of cost estimates for relocating the defense resale organizations, and adjust its range of cost estimates for consolidation and provide that updated information to Congress. (Recommendation 3) The Secretary of Defense should ensure that the DOD Chief Management Officer provide additional written information to Congress on the comments and concerns from the military departments and resale organizations on the task force’s November 2018 business case analysis, as well as the task force’s response to those comments and concerns. (Recommendation 4) Agency Comments and Our Evaluation We provided a draft of this product to DOD for review and comment. In its comments, reproduced in appendix III and summarized below, DOD concurred with the first three recommendations and did not concur with the fourth recommendation. DOD also provided technical comments, which we incorporated as appropriate. DOD stated in its letter that it continues to firmly believe that consolidation of above-store operations of DeCA and the military exchanges is the right path forward and that it intends to move forward with this effort. DOD also requested that we consider the first three recommendations as implemented, based on information provided in the letter and as detailed below. Regarding the first recommendation for the task force to reassess its approach to estimating savings from cost of goods sold—to include reassessing its use of cost of goods sold by all four defense resale organizations rather than, for example, just the three exchange organizations—and make any necessary adjustments to its savings estimates for consolidation and provide that updated information to Congress, DOD stated that it had reassessed its approach and found that there is significant overlap and, therefore, savings opportunity in products sold by DeCA and the exchanges. DOD also provided revised savings estimates that exclude DeCA’s cost of goods sold from its methodology that show net savings ranging from $309 million to $739 million in the first 5 years of consolidation, followed by $255 million to $457 million per year thereafter. These figures are about 44 percent to 55 percent lower than the business case analysis’s estimate for the first 5 years and about 32 to 35 percent lower per year thereafter. By providing these revised savings estimates, we believe that DOD has addressed the intent of the recommendation. With regard to the second recommendation for the task force to reassess its methodology for estimating IT costs of consolidation, and make any necessary adjustments to its range of IT cost estimates and provide that updated information to Congress, DOD stated that the task force followed up with AAFES, NEXCOM, and DeCA to get an update on the cost estimates these entities expressed concern about to us. However, according to DOD, those resale organizations were unable to provide alternate data to use in place of the numbers in the business case analysis. DOD further stated in its letter that because no alternative data were provided, the department will continue to use the estimate in the business case analysis and will reengage with the resale organizations to develop more detailed IT design plans and make any necessary updates to the IT cost estimates as integration planning moves forward. As DOD develops its more detailed IT design plans and associated cost estimates, we will follow up with the department, including the resale organizations, to determine whether this recommendation has been addressed. In commenting on the third recommendation for the task force to develop a range of cost estimates for relocating the defense resale organizations, and adjust its range of cost estimates for consolidation and provide that updated information to Congress, DOD provided three possible courses of action, along with corresponding cost estimates. These possible courses of action, from least expensive to most expensive, are: (1) maintain operations at all four existing locations (no cost); (2) maintain commissary operations at DeCA headquarters, perform all exchange functions at AAFES headquarters, and close the NEXCOM and MCCS headquarters (one-time costs of $5.5 million and recurring annual costs of $1.3 million); and (3) create a new headquarters to perform all commissary and exchange operations near Washington, D.C. (one-time costs of $19.6 million and recurring annual costs of $19.7 million). DOD stated that consolidation would still result in financial benefits, even if the department chooses the most costly of these courses of action. By providing these cost estimates, we believe that DOD has addressed the intent of the recommendation. While we have determined that DOD has met the intent of the first and third recommendations, we also note that, in its comment letter, the department questioned some aspects of our analysis and conclusions regarding the first three recommendations. We stand by our analysis and conclusions and offer the following response: DOD stated in its comment letter that modifying the business case analysis’s approach to cost of goods savings would result in an incorrect use of benchmarks and go against industry best practice. For example, DOD stated that estimating savings by using the cost of goods sold for just the three exchange organizations, or for the three exchange organizations plus a portion of DeCA, would be flawed. However, as noted in our report, the latter method was recommended to us by task force officials when we raised concerns about the accuracy of the task force’s savings estimates in the business case analysis. In addition, multiplying the benchmarks by the cost of goods sold for the exchange organizations, as opposed to for all four resale organizations, would be more consistent with the assertion in the business case analysis that consolidation savings are dependent on the overlap among the merging organizations. DOD questioned the accuracy of some of our figures in the report by providing different data on product and vendor overlap between DeCA and the exchange organizations. However, this information was not included in the task force’s business case analysis or offered to us during the course of our audit. In addition, when we asked for supporting documentation that would allow us to validate the new figures, DOD did not provide any. DOD stated in its comments that excluding all or including only a portion of DeCA’s cost of goods sold implies that there is no opportunity to achieve savings between DeCA and the exchanges. Our report does not make this assertion, but rather offers a methodology that would result in a more conservative savings estimate, consistent with the data presented in the business case analysis and provided by task force officials, to better ensure that estimated savings were not overstated. As noted above, DOD did not concur with the fourth recommendation for the department’s Chief Management Officer to provide additional written information to Congress on the comments and concerns from the military departments and resale organizations on the task force’s November 2018 business case analysis, as well as on the task force’s response to those comments and concerns. DOD stated in its written response to our report that the department considered all the comments submitted in its decision-making process and that all of the military department secretariats agreed with above-store consolidation, despite their comments on the business case analysis. DOD further stated that the military department comments regarding the business case analysis were shared with congressional committee professional staff, and DOD suggested in its letter that this recommendation be closed. However, DOD’s written response did not provide information on which comments were shared, whether those comments were communicated in writing or orally, or which committee or committees received information on the comments. In their written comments on the business case analysis, the military departments detailed concerns with fundamental aspects of the analysis, to include: the use of proprietary industry benchmarks; estimated savings, costs, and timeline of consolidation; and the proposed governance structure for the new resale organization. We continue to believe that implementing this recommendation would help ensure that Congress has the full information it needs as it considers defense resale consolidation and would also help strengthen trust and collaboration among the various DOD stakeholders involved in defense resale, particularly given their role in any consolidation, should one occur. We will follow up with DOD as part of our regular recommendation follow- up process. We are sending copies of this report to the appropriate congressional committees; the Secretary of Defense; and the Secretaries of the Army, Air Force, and Navy. In addition, the report is available at no charge on our website at https://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2775 or FieldE1@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Prior GAO Work on Commissaries and Exchanges The National Defense Authorization Act for Fiscal Year 2016 mandated that the Department of Defense (DOD) report on DOD’s plan to achieve budget neutrality for commissaries and exchanges—which DOD interpreted as ending the use of appropriated funding—and included a provision for us to assess DOD’s report. In November 2016, we found that DOD’s May 2016 report did not provide a plan for achieving budget neutrality. DOD reported that it would not be able to eliminate fully the use of appropriated funds for defense resale, but the department did not provide detailed information supporting that conclusion. Instead, the report stated that DOD expected to achieve $2 billion in cost savings over a 5-year period from fiscal year 2017 through fiscal year 2021. However, we found that the report did not include any assumptions, methodology, or specific time frames related to initiatives that would lead to these savings. We recommended that DOD provide information to Congress to support its conclusion about budget neutrality and develop a plan for achieving reductions to defense resale appropriations. DOD concurred with our recommendations, but as of February 2020 had not addressed them. In March 2017, we reported on DOD’s commissary operations, including the extent to which the Defense Commissary Agency (DeCA) had assurance that it was maintaining the desired savings rate for its customers. DeCA’s desired savings rate—which at the time of our March 2017 report was 30 percent and is now 23.7 percent—shows how much a customer can expect to save on grocery purchases at a commissary in comparison to purchases at other local grocery stores. We found that DeCA lacked reasonable assurance that it was maintaining its desired savings rate for commissary customers because of weaknesses in its methodology for calculating the savings rate. For example, the methodology did not use a random sample of overseas commissaries or account for seasonal and geographic variations in item prices. We also found that DeCA’s business model departed from practices generally employed by commercial grocery stores. For example, DeCA did not assess the contribution of the sale of each product to a given store’s total sales in determining which products to sell, and it had not conducted cost- benefit analyses for its use of stocking and custodial service contracts or product distribution options across all commissaries. We recommended that DOD (1) address limitations identified in its savings rate methodology; (2) develop a plan with objectives, goals, and time frames to improve efficiency in product management; and (3) conduct comprehensive cost-benefit analyses for service contracts and distribution options. As of February 2020, DOD had addressed the first two recommendations but had not addressed the third recommendation. Appendix II: Assessment of the Department of Defense’s (DOD) Business Case Analysis on Defense Resale Consolidation We assessed DOD’s business case analysis on consolidating the four defense resale organizations against the five key elements of an economic analysis, as described in our Assessment Methodology for Economic Analysis (see table 1). Appendix III: Comments from the Department of Defense Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments Elizabeth A. Field, (202) 512-2775 or fielde1@gao.gov In addition to the contact named above, Suzanne Perkins (Assistant Director), Geoffrey Peck (Analyst-in-Charge), Pedro Almoguera, Noah Gerber, Mae Jones, Matthew Kienzle, Amie Lesser, Felicia Lopez, and Jeanne Sung made key contributions to this report.

Why GAO Did This Study DOD operates about 240 commissaries and 2,500 exchanges that sell groceries and retail goods and services to servicemembers, their families, and retirees. Commissaries and exchanges are operated by four resale organizations, and in November 2018 a DOD task force completed a business case analysis on consolidating those organizations. The National Defense Authorization Act for Fiscal Year 2020 included a provision for GAO to review DOD's business case analysis. This report evaluates the extent to which (1) DOD's business case analysis for consolidating the four resale organizations provided reliable savings and cost estimates and (2) the military departments concurred with the business case analysis and DOD shared their accompanying comments with Congress. GAO evaluated the business case analysis against DOD- and GAO-identified key elements of economic analyses; reviewed comments on the business case analysis; and interviewed DOD officials. What GAO Found A Department of Defense (DOD) task force's business case analysis for consolidating the defense resale organizations—the Defense Commissary Agency (DeCA), the Army and Air Force Exchange Service, the Navy Exchange Service Command, and Marine Corps Community Services—may not provide reliable savings and cost estimates. These organizations sell groceries and retail goods to servicemembers, their families, and retirees. The task force recommended consolidating the four resale organizations into a single organization, estimating “net savings” (i.e., savings minus costs) of about $690 million to $1.3 billion during the first 5 years. However, the task force may have overestimated savings and underestimated costs. Savings from reducing the cost of goods sold. The task force estimated that DOD would save several hundred million dollars annually by reducing the cost of purchasing goods that are resold in stores. Specifically, the task force multiplied the fiscal year 2017 total cost of goods sold for all four resale organizations by industry benchmarks, reasoning that mergers lead to more savings when merging organizations sell a high amount of identical products. However, task force data show that DeCA and the exchange organizations have limited identical products; the overlap between DeCA products and those of at least one exchange organization amounts to less than one-third of the total cost of goods sold. Thus, multiplying the benchmarks by the total cost of goods sold for all four organizations may not have been appropriate. Information technology (IT) costs. The task force estimated the costs of developing new, common IT systems to operate a consolidated resale organization to be between $326 million and $401 million, about 50 percent of estimated consolidation costs. The task force stated that it based IT cost estimates on data resale organizations provided for major upgrades or system replacements. But GAO found that about 40 percent of the IT cost estimate was based on minor upgrades or partial replacements, not major upgrades or system replacements. Thus, the estimate may be understated. Headquarters relocation costs. According to the task force, there will be costs if DOD decides to relocate the four defense resale organizations to a new headquarters location. However, the task force did not include cost estimates for relocation in its business case analysis. According to federal law, the operation of the commissary and exchange systems may not be consolidated unless authorized by Congress. Until the task force reassesses and updates, as necessary, its savings and costs estimates, DOD and Congress will not have reliable information to consider resale consolidation. The military departments officially concurred with the business case analysis, but provided written comments detailing fundamental concerns with the analysis, such as the use of proprietary industry benchmarks and the estimated savings and costs. In April 2019, DOD reported to Congress that the military departments agreed with consolidation, but did not disclose the accompanying comments. Without more complete reporting of those comments, Congress has limited visibility of the views of the organizations involved in a potential consolidation. What GAO Recommends GAO is making four recommendations, including that DOD reassess and update as necessary its estimates for consolidation savings and costs, and provide additional information to Congress on the military departments' comments on the November 2018 business case analysis. DOD concurred with three recommendations and provided updated estimates. DOD did not concur with the last recommendation. GAO continues to believe providing such information is beneficial, as discussed in the report.

gao_GAO-19-544T

gao_GAO-19-544T_0

Top Leadership Commitment Is Critical to Sustained Progress in Strengthening Management Functions With top leadership support and commitment, DHS has made important progress in strengthening its management functions, but considerable work remains. As shown in figure 1, as of March 2019, DHS had met three out of five criteria for removal from our High Risk List—leadership commitment, action planning, and monitoring progress. DHS has partially met the remaining two criteria: capacity (i.e., people and other resources) and demonstrated, sustained progress. To address the criteria for capacity, DHS needs to make additional progress in identifying and allocating resources in certain areas—namely, acquisition, information technology, and financial management—to fully demonstrate its capacity. For the criteria for demonstrated, sustained progress, we reported in March 2019 that DHS had fully addressed 17 out of the 30 outcomes that are the basis for gauging DHS’s progress across management areas, as shown in table 1. To fully meet the criteria for demonstrated, sustained progress, DHS needs to continue implementing its Integrated Strategy for High-Risk Management and maintain engagement with us to show measurable, sustainable progress in implementing corrective actions and achieving outcomes. DHS can accomplish this by, among other things, maintaining a high level of top leadership support and sustained commitment to ensure continued progress in executing its corrective actions through completion, and increasing employee engagement and morale. Examples of important programs and remaining work in the key management functions include: In the key management function of human capital management, DHS leadership is needed to address skills gaps that have had a significant role in the DHS management high-risk area. For example, we have found that DHS lacks guidance on how to identify critical cybersecurity and acquisition skills needed to support its new information technology delivery model. We have also found that DHS has insufficient technical skills to support its biometric identification services program. Addressing these skill gaps could help DHS fully demonstrate its capacity to strengthen and integrate its management functions. Additionally, within human capital management, DHS has struggled with low employee morale scores since it began operations in 2003. DHS’s 2018 score ranked 20th among 20 large and very large federal agencies. Increasing employee engagement and morale is critical to strengthening DHS’s mission and management functions. DHS has continued to strengthen its employee engagement efforts by implementing our 2012 recommendation to establish metrics of success within components’ action plans for addressing its employee satisfaction problems. Further, DHS has conducted audits to better ensure components are basing hiring decisions and promotions on human capital competencies. In addition, OPM’s 2018 Federal Employee Viewpoint Survey data showed that in the past 2 years, DHS’s score on the Employee Engagement Index (EEI) increased by 4 points—from 56 in 2016 to 60 in 2018—which was 1 point more than the government-wide increase over the same period. While this improvement is notable, DHS’s current EEI score is 1 point below its EEI baseline score in 2010, suggesting that DHS is still working to regain lost ground after an 8 point drop between 2010 and 2015. In the key management function of financial management, DHS officials have faced challenges modernizing DHS components’ financial management systems and business processes that affect the department’s ability to have ready access to timely and reliable information for informed decision-making. Effectively modernizing financial management systems for the Coast Guard, FEMA, and ICE would help improve the reliability of their financial reporting. As we have reported, perhaps the single most important element of successful management improvement and transformation initiatives is the demonstrated commitment of top leaders, as shown by their personal involvement in reform efforts. With regard to leadership commitment, DHS’s top leadership, including leaders at the Secretary and Deputy Secretary level, has demonstrated exemplary commitment and support for addressing the department’s management challenges. They have also taken actions to institutionalize this commitment to help ensure the long- term success of the department’s efforts. One such effort is the Under Secretary for Management’s Integrated Priorities initiative to strengthen the integration of DHS’s business operations across the department. During monthly leadership meetings with the Under Secretary for Management, the department’s Chief Executive Officers have been providing status updates on their respective actions to address this high- risk designation. Furthermore, top DHS leaders, such as the Under Secretary for Management and the department’s Chief Executive Officers, routinely meet with our management to discuss progress on high-risk areas. Continued Leadership Commitment Is Critical to Addressing Priority Open Recommendations In April 2019, we sent a letter to the Acting Secretary of Homeland Security detailing 26 open recommendations that we deem highest priority for implementation. Priority recommendations are those that we believe warrant priority personal attention from heads of key departments or agencies. These 26 recommendations fall into six major areas— emergency preparedness and response, border security, transportation security, infrastructure and management, cybersecurity, and chemical and nuclear security. Many of these recommendations cut across DHS’s mission areas that are critical for national security. Given that these recommendations are often the most complex and difficult to implement, top DHS leadership will play a critical role in addressing them. Fourteen of the 26 priority open recommendations we identified in the April 2019 letter are directed to acting officials serving in vacant positions. We have issued 12 recommendations to the Secretary of Homeland Security who is currently an acting official. We have also issued two recommendations to FEMA which is currently operating under acting leadership. Committed and consistent leadership at the department and component levels will be critical for addressing our priority recommendations. For example: In September 2014, we recommended that the Secretary of Homeland Security work jointly with the Administrator of the General Services Administration to strengthen management of the ongoing acquisition project to develop the multi-billion dollar headquarters facilities at the St. Elizabeth’s campus in Washington, D.C. Leadership is critical in this effort, given the magnitude of the project and the impact of headquarters consolidation on DHS operations. In October 2008, we recommended actions that FEMA should take to improve its administration of the National Flood Insurance Program high-risk area. We also recommended in September 2012 that FEMA develop a methodology to better assess a jurisdiction's capability to respond to and recover from a disaster without federal assistance. In July 2015, we further recommended that the Mitigation Framework Leadership Group establish an investment strategy to identify, prioritize, and guide federal investments in disaster resilience. Implementing these actions could limit the federal government’s fiscal exposure and increase the nation’s resilience to extreme weather events as the costs and impacts of weather disasters resulting from floods, drought, and other events are expected to increase in significance as previously “rare” events become more common and intense. In July 2018, we recommended that U.S. Customs and Border Protection (CBP) analyze the costs associated with future barrier segments along the southwest border and include cost as a factor in the Impedance and Denial Prioritization Strategy. Obtaining this key information could help CBP evaluate designs and prioritize locations for future border barrier segments to deter cross-border illegal activity. In February 2017, we recommended that DHS establish metrics and methods by which to evaluate the performance of DHS’s National Cybersecurity and Communications Integration Center in relation to its statutorily-required cybersecurity functions. Until it develops metrics and methods to evaluate its performance, the center cannot ensure that it is effectively meeting its statutory requirements, while cyber- based intrusions and attacks on federal systems and systems supporting our nation’s critical infrastructure are becoming more numerous, damaging, and disruptive. We also recommended in February 2018 that DHS take steps to better manage and assess its cybersecurity workforce gaps and areas of critical need. Given its important role in the nation’s cybersecurity, taking steps to address these issues will be critical. We will continue to monitor DHS’s progress in strengthening management functions and addressing priority recommendations. We also plan to continue to meet quarterly with DHS management to gauge leadership commitment, discuss progress, and review DHS’s goals and corrective action plans in its Integrated Strategy for High-Risk Management, which DHS issues twice per year. Thank you, Chairman Thompson, Ranking Member Rogers, and Members of the Committee. This concludes my testimony. I would be pleased to answer any questions. GAO Contact and Staff Acknowledgments For further information on this testimony, please contact Christopher P. Currie at (404) 679-1875 or curriec@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Other individuals making key contributions to this work include Alana Finley, Assistant Director; Luis E. Rodriguez, Analyst-in-Charge; Karin Fangman; Andrew Howard; and Thomas Lombardi. Key contributors for the previous work that this is based on are listed in each product. 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.

Why GAO Did This Study In 2003, GAO designated Implementing and Transforming DHS as a high-risk area to the federal government. DHS has made considerable progress in transforming its original component agencies into a single cabinet-level department, and as a result, in 2013, GAO narrowed the scope of the high-risk area to focus on Strengthening DHS Management Functions . In addition, DHS leadership is responsible for implementing numerous recommendations that GAO has made to the department and its component agencies. Current vacancies in top leadership positions could pose a challenge to addressing high-risk areas and priority recommendations that span DHS's diverse missions, which include preventing terrorism and enhancing security, managing our borders, administering immigration laws, securing cyberspace, and responding to disasters. This testimony discusses the need for DHS leadership commitment to strengthen its management functions and address GAO's priority recommendations. This testimony is based on GAO's 2019 high-risk update and other reports issued from March 2006 through April 2019. What GAO Found With the support and commitment of top leadership, the Department of Homeland Security (DHS) has made important progress in strengthening its management functions; however, considerable work remains. As of March 2019, DHS had fully addressed 17 of the 30 outcomes related to its management functions (see table). DHS needs to continue to show sustained leadership commitment in implementing its Integrated Strategy for High-Risk Management to achieve the remaining outcomes. Leadership commitment is also pivotal in addressing other GAO high-risk areas where DHS has a role, such as ensuring the cybersecurity of the nation, the National Flood Insurance Program, and limiting the federal government's fiscal exposure by better managing climate change risks. Currently, DHS has acting officials serving in eight positions requiring Senate confirmation, including positions with responsibilities for implementing high-risk outcomes, such as the Secretary, Deputy Secretary, and Under Secretary for Management. a “Mostly addressed”: Progress is significant and a small amount of work remains. b “Partially addressed”: Progress is measurable, but significant work remains. c “Initiated”: Activities have been initiated to address the outcome, but it is too early to report progress. In April 2019, GAO sent a letter to the Acting Secretary of Homeland Security detailing 26 open recommendations that GAO believes warrant the highest priority personal attention from the department and its components. These 26 recommendations fall into six major areas—emergency preparedness and response, border security, transportation security, infrastructure and management, cybersecurity, and chemical and nuclear security. For example, GAO has recommended that DHS take steps to strengthen human capital management, such as better managing and assessing its cybersecurity workforce gaps and areas of critical need. Fourteen of the 26 recommendations have been issued to acting officials serving in vacant positions, including 12 to the Secretary of Homeland Security, and two to the Federal Emergency Management Agency which is currently operating under acting leadership. What GAO Recommends Since the creation of DHS, GAO has made approximately 2,800 recommendations to the department, and DHS has implemented more than 75 percent of them, strengthening program management and performance measurement, among other things. GAO will continue to monitor DHS's progress in strengthening management functions and addressing priority recommendations.

gao_GAO-20-253

gao_GAO-20-253_0

Background DOD is the largest U.S. government department and one of the most complex organizations in the world. In support of its military operations, the department manages many interdependent business functions, including logistics management, procurement, health care management, and financial management. DOD relies extensively on IT to support its business functions. The department’s IT budget organizes investments in four categories, called mission areas—enterprise information environment, business, warfighting, and intelligence. Figure 1 shows the amount of DOD’s total requested fiscal year 2020 IT budget (of $36.1 billion) that the department plans to spend on each of its mission areas (including the approximately $8.9 billion it plans to spend on developing, modernizing, operating, and maintaining its business systems). The department further organizes its IT budget by segments. In this regard, the business mission area segments are logistics/supply chain management, human resource management, health, financial management, acquisition, real property management, training and readiness, other business services, and defense security enterprise. Figure 2 shows the department’s projected fiscal year 2020 spending (of $8.9 billion) for each segment in the business mission area. U.S. Code Establishes Requirements for Managing Defense Business Systems As amended in 2016, the U.S. Code requires DOD to perform certain activities aimed at ensuring that its business system investments are managed efficiently and effectively. Specifically, the amendments established four sets of requirements for the department related to (1) issuing policy and guidance for managing defense business systems; (2) developing and maintaining a defense business enterprise architecture; (3) establishing a Defense Business Council to provide advice to the Secretary on managing defense business systems; and (4) obtaining approvals before systems proceed into development (or if no development is required, into production or fielding). Further, the amendments to the code established specific designations and thresholds that, among other things, provided additional details about the department’s requirements: Covered defense business systems. The code defines a covered defense business system as a system that is expected to have a total amount of budget authority of over $50 million over a period of 5 years or more. Priority defense business systems. The code establishes a category of system, called a priority defense business system. This refers to a system that is (1) expected to have a total amount of budget authority of over $250 million over the period of the current future-years defense program, or (2) designated by the DOD chief management officer (CMO) as a priority defense business system based on specific program analyses of factors including complexity, scope, and technical risk, and after notification to Congress of such designation. Thresholds and officials responsible for review and certification of defense business systems. The code states that, unless otherwise assigned by the Secretary of Defense, military department CMOs are to have approval authority for their covered defense business system investments of below $250 million over the future- years defense program. The CMO is to have approval authority for defense business systems owned by DOD components other than the military departments, systems that will support the business process of more than one military department or other component, and priority defense business systems. Certification requirements. The code requires that a defense business system program be reviewed and certified at least annually, on the basis of its compliance with the business enterprise architecture and appropriate business process reengineering. In addition, the code requires that the business system program be reviewed and certified on the basis of having valid, achievable requirements and a viable plan for implementing the requirements; having an acquisition strategy designed to eliminate or reduce the need to tailor commercial off-the-shelf systems; and being in compliance with the department’s auditability requirements. DOD Has Assigned Roles and Responsibilities for Managing Defense Business Systems DOD Instruction 5000.75: Business System Requirements and Acquisitions assigns roles and responsibilities for managing defense business system investments. Table 1 identifies the key entities and their responsibilities for managing defense business system investments. GAO Previously Has Reported on DOD Business Systems Modernization Efforts GAO designated the department’s business systems modernization efforts as high risk in 1995 and has continued to do so in the years since. In addition, since 2005, we have issued 12 reports in which we assessed DOD’s actions to respond to the business system investment management requirements contained in U.S. Code Title 10, Section 2222. These 12 reports, which are listed in appendix I, collectively contained 29 recommendations to help strengthen the department’s management of its business systems. For example, In 2014, we reported that DOD had taken steps to comply with key provisions in the NDAA for Fiscal Year 2005; however, the department continued to face challenges in fully complying with the provisions and modernizing its business systems environment. As a result, we recommended that the department improve its business system certification and approval process. DOD agreed with, and implemented this recommendation. In 2015, we reported that DOD had implemented 5 of the 16 recommendations made by GAO since June 2011 to address each of the overarching provisions for improving business systems management described in the NDAA for Fiscal Year 2005. The department had partially implemented the other 11 recommendations. We also reported that DOD’s business enterprise architecture and process reengineering efforts were not fully achieving the intended outcomes described in statute. Thus, we made a recommendation aimed at ensuring that the department better achieve business process reengineering and enterprise architecture outcomes and benefits. DOD agreed with this recommendation and took steps toward implementing it. In 2018, we reported that DOD had made progress in complying with most legislative provisions for managing its defense business systems, but that additional actions were needed. For example, the NDAA for Fiscal Year 2016 required DOD and the military departments to issue guidance to address five requirements for reviewing and certifying the department’s business systems. We reported that, while DOD had issued guidance addressing all of the requirements, the military departments had shown mixed progress. Accordingly, we recommended that the military departments issue guidance to address certifiying their business systems on the basis of the five requirements. While DOD partially agreed with these recommendations, the military departments implemented them. As of June 2019, the department had implemented 15 of the 29 recommendations contained in the 12 reports. In addition, we closed two of the recommendations as “not implemented” because the actions taken by the department did not sufficiently address the recommendations. The department had not yet taken actions to address the other 12 recommendations. Table 2 identifies the four sets of requirements for strengthening DOD’s management of defense business systems identified in the U.S. Code. In addition, the table identifies a fifth category associated with human capital, which supports the department’s execution of the other four sets of requirements. The table also identifies the 12 GAO recommendations that remained to be implemented as of June 2019. DOD Made Progress in Strengthening Its Management of Business Systems, but Had Not Addressed All Recommendations Relating to the Statutory Requirements DOD took actions toward addressing GAO’s recommendations related to business system requirements contained in the U.S. Code. In doing so, the department made progress in strengthening the management of its defense business system investments. Specifically, between June 2019 and November 2019, the department demonstrated that it had implemented four of the 12 remaining recommendations aimed at strengthening business systems management. As of November 2019, for example, the department had taken actions to implement a recommendation that helped the department comply with the code’s requirement to establish guidance for effectively managing its defense business system investments. In this regard, the Office of the CMO demonstrated that DOD had implemented our recommendation to improve the department’s policy to require full consideration of sustainability and technological refreshment requirements for its defense business systems investments. Specifically, the department demonstrated that its Instruction 5000.75, DOD Directive 5000.01, and DOD Financial Management Regulation Volume 2B guidance includes policy requiring consideration of sustainability and technological refreshment. The department also demonstrated that its DOD Directive 5000.01 guidance includes policy to ensure that best systems engineering practices are used in the procurement and deployment of commercial systems, modified commercial systems, and defense-unique systems. In addition, with regard to the requirement that DOD ensure that business systems are reviewed and certified in accordance with U.S. Code Title 10, Section 2222, the Army demonstrated that it had implemented our recommendation that the department improve its guidance for certifying defense business systems. Specifically, the Army issued guidance to require that the systems be certified on the basis of (1) having valid, achievable requirements and a viable plan to implement the requirements; (2) having an acquisition strategy designed to eliminate or reduce the need to tailor commercial off-the-shelf systems to meet unique requirements, incorporate unique requirements, or incorporate unique interfaces to the maximum extent practicable; and (3) being in compliance with DOD’s auditability requirements. The Air Force also demonstrated that it had implemented our recommendation that the department improve its guidance for certifying defense business systems. Specifically, the Air Force issued guidance to require that systems be certified on the basis of (1) having an acquisition strategy designed to eliminate or reduce the need to tailor commercial off-the-shelf systems to meet unique requirements, incorporate unique requirements, or incorporate unique interfaces to the maximum extent practicable; and (2) being in compliance with DOD’s auditability requirements. Further, with regard to this requirement, the Office of the CMO demonstrated that it had addressed the recommendation to implement and use business enterprise architecture and business process reengineering compliance assessments more effectively. Specifically, in September 2019, the Office of the CMO demonstrated that it had reviewed business enterprise architecture and business process reengineering compliance assessments and identified in investment decision memorandums which systems had assessments that required action. Even with the actions taken, however, more remained to be done to implement eight other recommendations relating to the code’s requirements that could help strengthen the department’s management of its business systems. Specifically, with regard to the requirement to ensure that business systems are reviewed and certified in accordance with the code, the department had not implemented two of our related recommendations. For example, it had not implemented our recommendation to ensure that portfolio assessments are conducted in key areas identified in GAO’s IT investment management framework, such as current schedule, project reporting, and risks. In addition, with regard to the requirement to develop and maintain a defense business enterprise architecture and IT enterprise architecture, in accordance with relevant laws and Office of Management and Budget policies and guidance, the department had developed a business enterprise architecture. However, it had not implemented our five related recommendations. For example, it had not implemented our recommendation to integrate its business and IT architectures. According to officials in the Office of the CIO, the office plans to finalize the first increment of version 3 of its DOD Information Enterprise Architecture (i.e., IT enterprise architecture) by the end of December 2019, and intends to integrate the business enterprise architecture into the Information Enterprise Architecture as a part of that effort. Further, with regard to human capital, which supports the other requirements, the department had not implemented our related recommendation to develop a skills inventory, needs assessment, gap analysis, and plan to address identified gaps as part of a strategic approach to human capital planning. In commenting on its status in addressing the recommendation in September 2019, the Office of the CMO stated that it intends to publish a Defense Business Operations Management Workforce Plan by December 31, 2019. According to office officials, the plan is to include skills requirements for both Office of the CMO permanent employees and DOD employees detailed to the Office of the CMO in support of the management and reform of defense business operations management. Table 3 summarizes the status of the 12 recommendations, as of November 2019, relative to the requirements established in the code. Further, appendix II provides additional information about the status of the recommendations, as of November 2019. By taking actions to implement four of the 12 remaining recommendations, DOD made important progress in its efforts to comply with the requirements established in the U.S. Code. Nevertheless, taking further actions to implement all of the recommendations is essential to helping the department comply with all of the requirements—and ultimately, to strengthen the management of its defense business system investments and efforts to effectively transform its business operations. Agency Comments and our Evaluation DOD provided written comments on a draft of this report, which are reprinted in appendix III. In its comments, the department stated that it appreciated GAO’s recognition of the progress that DOD has made to strengthen business systems management. The department agreed that more needs to be done to strengthen the management of investments in DOD business operations and stated that, in the coming months, it plans to execute a reform agenda intended to strengthen oversight and improve business performance. Further, while this report made no new recommendations, the department concurred with our assessment of the status of seven of the eight open recommendations that DOD has not yet implemented. In addition, the department suggested that we close one recommendation that it perform specific activities as part of a strategic approach to human capital planning for the Office of the CMO, based on the CMO’s submission of a human capital analysis report to Congress in early January 2020. The department subsequently provided us a copy of the CMO’s human capital analysis report on February 7, 2020. We plan to assess the report to determine if it meets the intent of our recommendation. We are sending copies of this report to the appropriate congressional committees, the Director of the Office of Management and Budget, the Secretary of Defense, and other interested parties. This report also is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions on matters discussed in this report, please contact me at (202) 512-4456 or harriscc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix IV. Appendix I: Prior GAO Reports on Department of Defense Business System Modernization Since 2005, GAO has issued 12 reports in which we assessed DOD’s actions to respond to business systems modernization requirements contained in U.S. Code, Title 10, Section 2222. These requirements specify how DOD is to manage its business system investments. The reports are listed below: Defense Business Systems: DOD Needs to Continue Improving Guidance and Plans for Effectively Managing Investments, GAO-18-130 (Washington, D.C.: April 16, 2018). DOD Business Systems Modernization: Additional Action Needed to Achieve Intended Outcomes, GAO-15-627 (Washington, D.C.: July 16, 2015). Defense Business Systems: Further Refinements Needed to Guide the Investment Management Process, GAO-14-486 (Washington, D.C. May 12, 2014). DOD Business Systems Modernization: Further Actions Needed to Address Challenges and Improve Accountability, GAO-13-557 (Washington, D.C.: May 17, 2013). DOD Business Systems Modernization: Governance Mechanisms for Implementing Management Controls Need to Be Improved, GAO-12-685 (Washington, D.C.: June 1, 2012). Department of Defense: Further Actions Needed to Institutionalize Key Business System Modernization Management Controls, GAO-11-684 (Washington, D.C.: June 29, 2011). Business Systems Modernization: Scope and Content of DOD’s Congressional Report and Executive Oversight of Investments Need to Improve, GAO-10-663 (Washington, D.C.: May 24, 2010). DOD Business Systems Modernization: Recent Slowdown in Institutionalizing Key Management Controls Needs to Be Addressed, GAO-09-586 (Washington, D.C.: May 18, 2009). DOD Business Systems Modernization: Progress in Establishing Corporate Management Controls Needs to Be Replicated Within Military Departments, GAO-08-705 (Washington, D.C.: May 15, 2008). DOD Business Systems Modernization: Progress Continues to Be Made in Establishing Corporate Management Controls, but Further Steps Are Needed, GAO-07-733 (Washington, D.C.: May 14, 2007). Business Systems Modernization: DOD Continues to Improve Institutional Approach, but Further Steps Needed, GAO-06-658 (Washington, D.C.: May 15, 2006). DOD Business Systems Modernization: Important Progress Made in Establishing Foundational Architecture Products and Investment Management Practices, but Much Work Remains, GAO-06-219 (Washington, D.C.: November 23, 2005). Appendix II: Status of GAO Recommendations Aimed at Strengthening DOD’s Approach to Managing Its Business System Investments, as of November 2019 Table 4 summarizes the status of recommendations made to the Department of Defense (DOD), based on GAO’s assessments of DOD’s actions to respond to defense business system investment management requirements contained in U.S. Code Title 10, Section 2222. The recommendations, included in GAO reports issued from 2005 through 2018, were aimed at helping the department strengthen its approach to managing its business system investments. As of June 2019 (when GAO began its current review of DOD’s actions to address the recommendations), the department had not yet addressed 12 of 29 total recommendations. Subsequently, between June 2019 and November 2019, the department implemented four of the recommendations, but did not implement eight other recommendations. Appendix III: Comments from the Department of Defense Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact above, individuals who made contributions to this report include Michael Holland (Assistant Director), Tyler Mountjoy (Analyst in Charge), Camille Chaires, Cheryl Dottermusch, William Hutchinson, Monica Perez-Nelson, Priscilla Smith, and Adam Vodraska.

Why GAO Did This Study DOD spends billions of dollars each year on systems to support its key business areas, such as personnel and logistics. For fiscal year 2020, DOD reported that its business system investments are expected to cost about $8.9 billion. GAO has made many recommendations to DOD aimed at strengthening defense business systems management. Further, U.S. Code Title 10, Section 2222 requires DOD to perform activities aimed at ensuring that these investments are managed efficiently and effectively. The National Defense Authorization Act for Fiscal Year 2016 included a provision for GAO to report on the extent to which DOD is complying with the code's requirements. Accordingly, the objective of this review was to assess the extent to which DOD has taken actions that comply with the code's requirements for ensuring that business system investments are managed efficiently and effectively. To do so, GAO selected 12 recommendations that DOD had not implemented as of June 2019, and assessed the department's subsequent actions on the recommendations (through November 2019) against the requirements in the code. GAO also analyzed DOD's business systems guidance and business enterprise architecture documentation, and interviewed relevant DOD officials. What GAO Found As of November 2019, the Department of Defense (DOD) had taken actions that addressed some, but not all, of the 12 prior GAO recommendations for strengthening defense business systems management. In doing so, the department made progress in complying with related business system investment management requirements contained in U.S. Code Title 10 Section 2222 (the code or U.S. Code). Specifically, as of November 2019, DOD had implemented four of the 12 recommendations (see table). For example, with respect to the requirement associated with investment management guidance, DOD had implemented the recommendation to issue policy requiring full consideration of sustainability and technological refreshment requirements for its business system investments. However, DOD had not yet implemented eight other recommendations relating to the code's requirements. The recommendations that had not been implemented relate to the department's actions to: integrate its business and information technology (IT) architectures, ensure that portfolio assessments are conducted in key areas identified in the GAO Information Technology Investment Management framework. develop a skills inventory, needs assessment, gap analysis, and plan to address identified gaps as part of a strategic approach to human capital planning, among other things. Taking further actions to implement all of the recommendations is essential to helping the department achieve compliance with all of the requirements—and, ultimately, further strengthen the management of its defense business system investments as well as its efforts to effectively transform its business operations. What GAO Recommends GAO is not making any new recommendations in this report. As of November 2019, DOD had not yet implemented eight of the 12 prior recommendations. GAO will continue to monitor DOD's actions to address the remaining recommendations.

gao_GAO-20-213

gao_GAO-20-213_0

Background DHS and its components invest billions of dollars each year to acquire IT and other capabilities to support the department’s critical functions. The department plans to spend approximately $2.3 billion on major IT investments in fiscal year 2020. However, DHS has faced long-standing challenges in acquiring and managing IT. We have highlighted the department’s IT management issues on our high-risk list since 2003 and have made numerous recommendations to improve its IT management practices. For example, in 2013, we testified that, out of 68 major IT investments that the department had in development, 21 had one or more subsidiary projects that were not meeting cost and/or schedule commitments due to technical issues in the development phase, changes in agency priorities, or a lack of understanding of user requirements, among other things. Overview of Incremental and Agile Software Development Many federal agencies, including DHS, are accustomed to using a waterfall software development model. This type of model typically consists of long, sequential phases, resulting in product delivery years after program initiation. With many federal IT investments in a development phase, it is important to ensure that agencies are making the most efficient use of their financial resources through effective management practices. However, as we have previously reported and testified, historically federal IT projects often fail—that is, even after exceeding their budgets by millions of dollars and delaying the schedules by years—and the results do not meet requirements. Recognizing the severity of challenges related to the government-wide management of IT, in December 2014, federal IT acquisition reform provisions (commonly referred to as FITARA) were enacted as a part of the Carl Levin and Howard P. ”Buck” McKeon National Defense Authorization Act for Fiscal Year 2015. One of the provisions requires that the Office of Management and Budget (OMB) require in its annual IT capital planning guidance that each covered agency’s chief information officer (CIO) certify that IT investments are adequately implementing incremental development, as defined in capital planning guidance issued by OMB. Agile software development—one form of incremental development— calls for the rapid delivery of software. Probably the most well-known feature of Agile software development is iterative product development and delivery; that is, development of software in segments that are continuously evaluated against requirements. This method is well suited for programs in which the final product is to include distinct features, some of which may be discovered during the process rather than planned at the beginning. These frequent iterations can effectively measure progress and allow developers to respond quickly to feedback from customers, thus reducing technical and programmatic risk. With its emphasis on early and continuous delivery of working software, Agile can be a valuable tool for agencies in mitigating schedule and budget risks. Figure 1 compares requirements, design, development, and testing using Agile software methods versus a traditional waterfall approach; illustrating how requirements, design, development, and testing are performed concurrently in smaller time-boxed iterations for Agile and sequentially in waterfall development. As a result, using an Agile framework should result in producing high-quality software with frequent reviews and customer feedback to ensure that the highest value requirements are being met. The figure assumes that planning for both Agile and waterfall development has already occurred. DHS Adopted Agile Software Development to Address IT Challenges In February 2016, the DHS Under Secretary for Management announced an effort to pilot the use of Agile development methodologies to improve the department’s execution and oversight of IT acquisitions. This resulted in five Agile pilot programs. Each pilot program was overseen by a component integrated program team. Collectively, the first pilot programs were also overseen and supported by a DHS integrated program team. In April 2016, the department issued an Agile instruction, which identified Agile software development as the preferred approach for all DHS programs and projects that are to deliver an IT, or embedded-IT, capability. The department also set an expectation for its component CIOs to develop plans to increase the use of Agile development and justify any major IT programs that did not intend to use Agile development practices. Many DHS programs were already using Agile or similar incremental development methods before the department identified it as the preferred approach. Roles and Responsibilities for Agile Programs The DHS CIO, as the individual delegated departmentwide responsibility for approving, managing, and overseeing all of the department’s IT programs, sets the policies and procedures to help ensure Agile practices meet the department’s goals and comply with acquisition management policy. The DHS CIO is supported in this effort by the heads of other major DHS lines of business, such as the Chief Procurement Officer. Table 1 describes the roles and responsibilities that support Agile development within the department. Additionally, DHS established a headquarters-level team—the ITPM COE—to collaborate across the department on improvements to policy, governance, and acquisition guidance. In April 2017, the ITPM COE assumed responsibilities for the department’s transition to Agile development. The Office of the Chief Technology Officer (OCTO) Strategic Technology Management (STM) division within the OCIO facilitates the ITPM COE and serves as the official liaison between other OCIO divisions, other partner headquarters directorate and management offices, and operational components as needed. GAO Previously Reported on Challenges in DHS’ Management of Agile Programs We have reported on various programmatic and technical challenges that were limiting DHS’ efforts on Agile programs. For example, In 2016, we reported that the U.S. Citizenship and Immigration Services Transformation program, which was using Agile software development to modernize citizenship and immigration benefits processing, needed to improve testing of its software code and ensure its approaches to interoperability and end user testing met leading practices. We made 12 recommendations to improve Transformation program management, including ensuring alignment among policy, guidance, and leading practices in areas such as Agile software development and systems integration and testing. DHS concurred with the recommendations and has thus far implemented eight of them. We reported in October 2017 that the Transportation Security Administration Technology Infrastructure Modernization program had not defined key roles and responsibilities, prioritized system requirements, or implemented automated capabilities that were essential to ensuring effective adoption of Agile. We made 14 recommendations including that DHS should prioritize requirements and obtain leadership consensus on oversight and governance changes. DHS concurred with the recommendations and to date has implemented 13 of them. In November 2018, we reported that the U.S. Secret Service OCIO did not fully measure post-deployment user satisfaction with one project supporting the Information Integration and Technology Transformation investment. We made 13 recommendations to the U.S. Secret Service including that the Secret Service establish a process that ensures the CIO reviews all IT contracts, as appropriate; and identify the skills needed for its IT workforce. DHS concurred with the recommendations but has not yet implemented them. We reported in April 2019 that the Federal Emergency Management Agency Grants Management Modernization program had not yet fully established plans for implementing new business processes or established completed traceability of IT requirements. We made eight recommendations to implement leading practices related to reengineering processes, managing requirements, scheduling, and implementing cybersecurity. DHS concurred with the recommendations and has thus far implemented two of them. Organizational Change Management According to the Project Management Institute, the practice of change management is a comprehensive, cyclic, structured approach for transitioning individuals, groups, and organizations from a current state to a future state with intended business benefits. It helps organizations to integrate and align people, processes, structures, culture, and strategy. The Project Management Institute and GAO have both described leading practices for effective organizational change management. DHS Has Made Progress in Implementing Leading Practices, but Has Not Fully Addressed Others Leading practices in organizational change management advise an agency to (1) plan for, (2) implement, and (3) measure the impact when undertaking a significant change, such as a transition from one software development approach to another. Since DHS committed to its transition to Agile software development in policy in April 2016, the department has fully developed plans to facilitate the transition. However, DHS has not fully implemented these plans and has experienced challenges in measuring progress against its intended goals. In addition, many of the plans are part of a larger effort to improve overall IT acquisitions rather than specific to a transition to Agile development, an approach that may delay DHS’s execution of these plans. Leading practices for Agile software development adoption advise an agency to focus on three organizational levels of adoption: (1) agency environment, (2) program processes, and (3) team activities and dynamics. DHS has partially adopted practices at all three organizational levels. For example, the agency activities fully supported Agile methods through actions such as developing policies and procedures that called for the alignment of software, program goals, and agency goals. However, the department’s culture can better support Agile methods by, among other things, demonstrating an incentives and rewards structure to incentivize Agile teams. DHS Has Made Progress in Implementing Nine Leading Practices for Agile Software Development Adoption, but Has Not Fully Implemented All Leading practices that we developed for Agile software development adoption are organized into three areas, called organizational levels: agency environment, program processes, and team activities and dynamics. The organizational levels are further divided into nine leading practices. Table 2 identifies the three organizational levels and nine leading practices associated with these levels (three practices within each area). A detailed assessment of DHS’s implementation of each of the nine leading practices can be found in appendixes III, IV, and V. We refer to the leading practices related to an agency’s processes, culture, and acquisition strategies as agency environment practices. For an agency to successfully transition from an agency that supports traditional development methods, it should ensure that its activities, culture, and acquisition policy and procedures support Agile methods. DHS partially implemented the agency environment practice level by fully implementing two leading practices and partially implementing the remaining one. A more detailed assessment of DHS’s agency environment leading practices can be found in appendix III. Agency activities support Agile methods–fully implemented. DHS established appropriate life cycle activities to support Agile methods. For example, the department has outlined its policies, procedures, and guidance in several documents to assist its components in the acquisition and implementation of Agile software development. The department also developed policies and procedures that called for the alignment of software, program goals, and agency goals. Agency culture supports Agile methods–partially implemented. DHS established an environment that supported Agile development, and senior stakeholders supported its development throughout the agency. However, DHS did not take sufficient steps to ensure that senior stakeholders serving as executive sponsors understood Agile development, as called for by leading practices that are described in further detail in appendix III. The Director of STM stated that Agile sponsors were considered to be chief executive officers (e.g. Executive Director of PARM and the Deputy Under Secretary for Management). These parties oversaw the actions of the ITPM COE and approved the Agile action plans in June 2017. In addition, the department did not require training for senior stakeholders serving as executive sponsors, as called for by leading practices. In a written response, the Office of the Chief Human Capital Officer said that there were no Agile training requirements for officials at this level. By training executive-level sponsors in Agile development, the department can mitigate the risk of setting expectations for programs and projects that do not align with the values and principles of Agile software development. DHS also did not demonstrate that it established an incentives and rewards structure to incentivize Agile teams, as called for by leading practices. Officials from the Office of the Chief Human Capital Officer stated that the department’s existing rewards structure allowed for incentivizing team and individual performance even though it was not focused specifically on Agile methods. These officials stated that they did not believe that additional policy, guidance, or modifications to their existing policy were necessary. The Director of STM within OCIO stated that rewarding Agile teams was not a topic the ITPM COE was currently considering, but that OCIO might be interested in pursuing the topic after completing existing, higher-priority activities. By considering modifications to policy and guidance governing the incentives and rewards structure to promote team performance, DHS could improve team productivity and output. Agency acquisition policies and procedures support Agile methods–fully implemented. DHS guidance for acquisition strategies supported the unique needs of Agile programs. For example, DHS offered guidance for preparing acquisition strategies through its Procurement Innovation Lab and published Agile guidance that discussed contracting and acquisition strategies. Program processes involve staff being appropriately trained in Agile methods, technical environments enabling Agile development, and project planning controls being compatible with Agile development. DHS partially implemented the program processes practice level by fully implementing one leading practice and partially implementing the remaining two. A more detailed assessment of DHS program process leading practices can be found in appendix IV. Staff are appropriately trained in Agile methods–partially implemented. DHS training policy and guidance called for some of the acquisition management program staff to be trained in Agile methods. DHS has also taken steps to incorporate Agile concepts into required training for members of the acquisitions workforce. In addition, DHS offered elective training covering Agile methods and guidance for Agile teams, including contractors, to have the appropriate technical expertise needed to perform their role. The department also took steps to identify the necessary competencies for Agile teams and individuals. In April 2019, the Strategic Workforce Planning team within OCIO published a white paper identifying 27 competencies necessary for teams and individuals to use and training courses associated with the competencies. The white paper also made recommendations to help DHS address challenges in implementing Agile methods, such as establishing communities of practice for Agile practitioners to identify best practices and provide workshops. According to a written response by OCIO, the Strategic Workforce Planning team will create an implementation and communication plan for any deliverables associated with the white paper. However, the department did not provide policy or guidance to ensure that all program staff were trained in Agile methods, as called for by leading practices described in further detail in appendix IV. Existing Agile training requirements covered only the acquisitions workforce. DHS did not establish training requirements for program staff outside of the acquisitions workforce—such as a product owner or other staff—who may be assigned to an Agile program. As a result, individual programs must independently decide on and enforce training requirements if they want to ensure that all staff receive the needed training. DHS officials stated that the department focuses on key acquisition career fields in part because those career fields are defined in policy and procedures. According to the Director of STM, the department also encourages programs to independently find coaching and training because the components are more likely to have funding. By providing policy or guidance to ensure that all personnel staffed to an Agile program or project receive appropriate training, the department can better prepare program staff to plan and execute appropriately, and increase the likelihood of achieving the expected outcomes of the transition to Agile. Technical environments enable Agile development–fully implemented. DHS guidance called for technical and project tools to be available to support Agile development. For example, DHS test and evaluation guidance stated that automated testing should be implemented where practical. In addition, DHS guidance called for system designs that will support iterative delivery. For example, DHS enterprise architecture guidance and supplementary design considerations for acquisition programs discussed loose coupling and different methods for establishing a modular system. Project planning controls are compatible with Agile development–partially implemented. DHS guidance called for defining and incorporating non-functional requirements and critical features throughout development. In addition, DHS provided guidance for establishing a sustainable development pace. For example, the Agile instruction manual identified the benefits of monitoring the amount of work completed by Agile teams across each iteration in order to monitor ongoing team progress. However, DHS was not tracking and monitoring the pace of Agile team development as called for by DHS guidance and described further in appendix IV. According to the Director of STM, programs were not consistently reporting the Agile core metrics associated with development team pace as required. The Director of STM stated that the department was taking steps to begin tracking and monitoring the pace of Agile teams. In addition, the Director stated that he allocated staff to assist programs with consistently reporting the Agile core metrics. According to the Director of STM, the department was in the process of updating the core metrics and intended to publish a new version of them in the future, which would include tracking the pace. Nevertheless, DHS did not provide assurance that the metrics associated with development pace would be included in this revised set of metrics or that programs would consistently report that information in order for the department to track and monitor the pace of Agile teams. Until the department consistently tracks and monitors Agile programs and projects, it will not have the information needed to help ensure the development pace is maintained. Practices at the team activities and dynamics level include team composition supporting Agile methods, work being prioritized to maximize value for the customer, and repeatable processes being in place. DHS partially implemented the team activities and dynamics practice level by fully implementing one leading practice and partially implementing the remaining two. A more detailed assessment of DHS team activity and dynamics leading practices can be found in appendix V. Team composition supports Agile methods—fully implemented. DHS established guidance that called for self-organizing teams and defined the role of a product owner. For example, the Agile instruction and Agile instruction manual both explain that collaborative, self- organizing, and cross-functional teams help achieve the flexibility needed for the iterative development that characterizes Agile development methods. In addition, the Agile instruction manual states that the product owner is responsible for representing stakeholders and should be available to the development team throughout the iteration to answer questions and clarify requirements on behalf of the stakeholders. Work is prioritized to maximize value for the customer—partially implemented. DHS guidance called for Agile teams to craft user stories to define work. The guidance also called for user stories to be prioritized in a backlog based on value. However, the guidance did not describe how Agile teams can estimate the relative complexity of the user stories as called for by leading practices and described in further detail in appendix V. The Director of STM stated that relative estimation is a basic exercise and that guidance on this topic can be found in a number of sources outside of DHS. However, without providing guidance or directing Agile teams to external sources for additional information on relative estimation, OCIO risks that teams supporting Agile projects will not appropriately estimate user stories relative to each other. By providing guidance on estimating the relative complexity of user stories, the department can help Agile teams to effectively commit to an appropriate amount of work during a given iteration. Repeatable processes are in place—partially implemented. DHS guidance addressed holding daily meetings to review progress and discuss impediments, using a demonstration for the acceptance of a user story and conducting a retrospective to evaluate progress. In addition, the department’s guidance called for Agile programs to employ continuous integration and emphasized the need for mechanisms to help ensure code quality. However, DHS did not set expectations for automated testing and code quality, as called for by DHS guidance and described further in appendix V. DHS’s Agile core metrics included a series of metrics that addressed automated testing and code quality. The core metrics included targets but the targets were notional and, therefore, not expectations that DHS required a program to meet. According to the Director of STM, the initial core metrics were intended to assess the level of DHS team achievement without imposing artificial industry- based target measures for each. The Director stated that, on receiving the metrics for a period of time, the department would then adjust the core metrics and begin to include target measures based on the results achieved. According to the Director, this effort is currently underway and an updated set of core metrics will be distributed in early fiscal year 2020. Moreover, the department did not track and monitor automated testing or code quality against expectations. As discussed under project planning controls, DHS intended to track and monitor Agile practices, such as automated testing and code quality, through the Agile core metrics. However, according to the Director of STM, programs and projects were not consistently reporting these core metrics and those that were reporting did not collect data or report on particular metrics. By setting expectations for automated testing and code quality and beginning to track and monitor project performance against these expectations, DHS can increase the likelihood that Agile programs and projects are delivered within cost, schedule, and performance estimates. Conclusions DHS has taken many positive steps in its transition to Agile software development. It has implemented activities and artifacts that support all levels of adoption, from the department and component offices to Agile programs, projects, and teams. These activities and artifacts include providing opportunities for Agile programs and projects to streamline acquisition and life cycle processes to allow for iterative delivery and exhibiting senior support for the transition to Agile. The department successfully planned for the transition to Agile software development and completed many of its intended implementation activities. However, because DHS did not assess the skills and resources needed to complete deferred activities, it risks continued delays in completing these. In addition, without identifying target measures tied to expected outcomes, the department is limited in determining whether the transition is achieving its desired outcomes. Moreover, until DHS can ensure that all programs are consistently reporting on Agile core metrics, the department will not be able to track programs’ development techniques. Further measuring and communicating the benefits of the transition can enable the department to know whether Agile programs are performing better than those used prior to the transition. DHS has demonstrated significant progress in implementing leading Agile practices. The department can further improve its performance through full execution of the remaining partially implemented practices. At the agency environment level, DHS can mitigate risk and improve productivity through executive level training and modifications to policy to incentivize Agile teams. For program level practices, addressing training requirements for all necessary staff and tracking and monitoring the pace of Agile team development can help ensure teams’ success. With respect to team-level practices, DHS has not established guidance for estimating the relative complexity of user stories. As a result, Agile teams are hampered in effectively committing to an appropriate amount of work during a given period of time. Finally, because DHS has not set expectations for performance metrics for monitoring and tracking the use of automated testing and code quality, DHS is at a greater risk for programs breaching their cost and schedule expectations. Recommendations for Executive Action We are making the following 10 recommendations to the Secretary of the Department of Homeland Security (DHS). The Secretary should ensure that the Director of Strategic Technology Management (STM), in collaboration with other members of the Information Technology Program Management Center of Excellence (ITPM COE), identifies the skills and resources needed to complete the work intended for the upcoming fiscal year, including the availability of supplementary staff, such as subject matter experts. (Recommendation 1) The Secretary should ensure that the Executive Steering Committee overseeing the activities of the ITPM COE establishes target measures for the department’s desired outcomes of its transition to Agile development. (Recommendation 2) The Secretary should ensure that the DHS Chief Information Officer (CIO) defines a process and associated set of controls to ensure that Agile programs and projects are reporting a set of core required performance metrics for monitoring and measuring Agile adoption. (Recommendation 3) The Secretary should ensure that the ITPM COE, in coordination with the CIO, begins measuring results associated with the transition to Agile and the success of the transition based on its impact on the department. (Recommendation 4) The Secretary should ensure that the CIO, in collaboration with the Chief Procurement Officer, through the Homeland Security Acquisition Institute, establish Agile training requirements for senior stakeholders. (Recommendation 5) The Secretary should ensure that the Chief Human Capital Officer, in collaboration with the CIO, consider modifications to the current employee recognition and performance management governance to ensure that teamwork and team performance of Agile programs and projects are incentivized. (Recommendation 6) The Secretary should ensure that the CIO, in collaboration with the Chief Procurement Officer, through the Homeland Security Acquisition Institute, establish Agile training requirements for staff outside of the acquisition workforce but assigned to Agile programs. (Recommendation 7) The Secretary should ensure that the CIO, upon establishing a set of core performance metrics, tracks and monitors the pace of Agile team development. (Recommendation 8) The Secretary should ensure that the CIO, in collaboration with the Executive Director of the Office of Program Accountability and Risk Management (PARM), update or develop new guidance on Agile methodologies to describe how Agile teams can estimate the relative complexity of user stories. (Recommendation 9) The Secretary should ensure that the CIO, upon establishing a set of core performance metrics, sets expectations for automated testing and code quality, and tracks and monitors against those expectations. (Recommendation 10) Agency Comments and Our Evaluation DHS provided written comments on a draft of this report. In its comments (reproduced in Appendix VI), the department agreed with our 10 recommendations and described actions that it had completed and planned to address them. Based on the actions DHS said it had taken, the department requested that we close the first three recommendations as implemented. For example, the department described steps it had taken to address our recommendation that it identify the skills and resources needed to complete the work intended for the upcoming fiscal year, including the availability of supplementary staff such as subject matter experts. In addition, the department stated that it had addressed our recommendation to define a process and controls to ensure that Agile programs and projects are reporting a set of core required performance metrics for monitoring and measuring Agile adoption. We plan to follow up with DHS to assess the sufficiency of its actions to address our recommendations. The department also described actions that it plans to take to address the other seven recommendations. For example, DHS stated that it will use the results of its Agile core metrics and Agile Software Delivery Maturity Model to measure the success of the transition to Agile and its impact on the department. According to the department, it expects this action to be completed by June 30, 2021. Further, DHS stated that it will identify Agile training requirements for staff in Agile programs, and will use that to establish Agile training requirements for staff outside of the acquisition workforce but assigned to Agile programs. Specifically, DHS stated that the DHS OCIO will gather requirements from components via its IT workforce planning integrated project team to identify training resources available across the department that also address the skill sets needed for Agile programs. The department added that the DHS OCIO will utilize information from the April 2019 white paper, titled “OCIO Agile White Paper” to inform proposed Agile program training requirements. The department estimated that these actions are to be completed by September 30, 2020. DHS also provided technical comments, which we have incorporated as appropriate. We are sending copies of this report to the Acting Secretary of Homeland Security and interested congressional committees. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-4456 or harriscc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix VII. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. Appendix I: Objective, Scope, and Methodology Our objective was to assess the extent to which the Department of Homeland Security (DHS) addressed selected leading practices for its transition to the use of Agile software development. To accomplish this objective, we assessed the extent to which the department adhered to leading practices in two specific areas: organizational change management and Agile software development adoption. With regard to organizational change management, we reviewed leading practices published by the Project Management Institute and GAO. Based on this review, we identified 15 leading practices. We then grouped these 15 practices in three broad organizational change management areas: planning, implementing, and measuring change. To determine the extent to which DHS addressed leading practices for organizational change management in its transition to Agile development, we assessed DHS policies, procedures, guidance, plans, and other working group artifacts and compared them against leading practices. In particular, we reviewed working group charters for the DHS headquarters Agile Acquisition Integrated Program Team and IT Program Management Center of Excellence (ITPM COE), and any plans developed by these working groups, including the DHS Agile Action Plans and associated implementation plans. We then reviewed working group meeting minutes, presentation slides, and status update charts to assess the progress of the transition to Agile, identified artifacts prepared to support the transition to Agile, and assessed the status of plans for the transition to Agile. We reviewed all Agile artifacts prepared by or supporting the Agile working groups, such as a preliminary software development maturity model, the DHS Agile Acquisition Software Delivery Core Metrics (Agile core metrics), and an updated test and evaluation master plan template for Agile, among other artifacts. We also interviewed officials from DHS headquarters line of business representatives explicitly identified in the Agile Development and Delivery for Information Technology instruction (Agile instruction). This included officials from the Office of the Chief Procurement Officer, Office of the Chief Financial Officer, Office of the Chief Information Officer (OCIO), Office of Program Accountability and Risk Management (PARM), and the Science and Technology Directorate, and offices of Test and Evaluation and Systems Engineering. Within OCIO, we interviewed officials from the Office of the Chief Technology Officer (OCTO) within the Strategic Technology Management (STM) division, among others, as STM is the entity tasked with facilitating the ITPM COE and serves as the official liaison between other OCIO divisions, other partner headquarters directorate and management offices, and operational components. We also interviewed representatives from groups participating in ITPM COE activities but not explicitly called out in the Agile instruction, including the Privacy Office and Joint Requirements Council. In addition, we interviewed representatives from other groups not represented on the ITPM COE but potentially impacted by the transition to Agile. This included officials from the Office of the Chief Readiness Support Officer and Office of the Chief Human Capital Officer. With regard to leading practices for Agile software development adoption, we reviewed work performed by GAO to develop generally accepted leading practices. In developing these leading practices, GAO reviewed information from a variety of sources related to Agile adoption and compiled a draft of leading practices commonly mentioned across these different sources. We then convened a working group of experts from the public and private sectors and academia. This working group met three times a year between August 2016 and August 2019 to review and discuss these leading practices. More than 200 experts participated in the meetings, including more than 20 officials from DHS. GAO received comments from some of these experts both during these meetings and by email after the meetings. Based on this work, GAO developed a set of nine leading practices for Agile adoption. GAO grouped these leading practices into three organizational levels: (1) agency environment, (2) program processes, and (3) team activities and dynamics. The leading practices were further described by a series of core elements and core element expectations that, collectively, can be used to assess the status of an agency’s implementation. To determine the extent to which the department had implemented the leading practices for the adoption of Agile development, we obtained and assessed DHS policies, procedures, guidance, plans, and other documentation and compared them against the nine leading practices. In particular, we reviewed department acquisition policy, procedures, and guidance, such as acquisition management directive 102-01; software engineering life cycle policy, procedures, and guidance, such as those published in the software engineering life cycle guidebook; requirements policy, procedures, and guidance, such as the Joint Requirements Integration and Management System and Requirements Engineering User’s Guide; testing policy, procedures, and guidance, such as the Test and Evaluation Master Plan template and Test and Evaluation Management Guide; technical assessment and enterprise architecture policy, procedures, and guidance; program health assessment policy, procedures, and guidance such as the Acquisition Program Health Assessment instruction and CIO Program Health Assessment Scoring Guideline; and Agile-specific policy, procedures, and guidance, such as the Agile instruction and the Agile Development and Delivery for Information Technology Instruction Manual (Agile instruction manual), among other policy, procedures, and guidance. In addition to reviewing the department policy, procedures, and guidance, we obtained and assessed supplementary Agile documentation. In particular, we reviewed training materials prepared by the Homeland Security Acquisition Institute for acquisition workforce certifications and webinars offered by the Procurement Innovation Lab; ITPM COE Agile artifacts discussed under our assessment of the implementation of organizational change management leading practices, such as the Agile core metrics; and Agile-specific technical review completion letters, such as the release planning review. We also interviewed officials from the components responsible for the associated policy, procedures, and guidance and those specifically cited in the Agile instruction. This included officials from the Office of the Chief Procurement Officer, Office of the Chief Financial Officer, OCIO, PARM, Science and Technology Directorate, offices of Test and Evaluation and Systems Engineering, the Joint Requirements Council, Office of the Chief Readiness Support Officer, and Office of the Chief Human Capital Officer. As with our assessment of DHS implementation of organizational change management practices, within OCIO, we interviewed officials from the OCTO STM division, among others. We assessed a core element as being “met” if the department provided supporting documentation that demonstrated it met all of the expectations associated with the core elements. We assessed a core element as being “partially met” if the department provided supporting documentation that demonstrated some, but not all, aspects of the underlying expectations. We assessed a core element as “not met” if the officials did not provide any supporting documentation for the core element, or if the documentation provided did not demonstrate any aspect of the underlying expectations. The expectations associated with each core element are described more fully in appendixes III, IV, and V. We assessed each leading practice and practice level as being “fully implemented” if DHS provided evidence that it had met all of the core elements. We assessed each leading practice and practice level as being “not implemented” if DHS did not provide evidence that it had met or partially met any of the core elements. We assessed each leading practice and practice level as being “partially implemented” if DHS provided evidence that it had not met all core elements and partially met at least one core element. To supplement our assessment of the department’s implementation of the leading practices for adopting Agile development, we also assessed selected projects’ implementation of selected program process and team activity and dynamics leading practices. We updated the core element test plans to include general control objectives, associated controls, and associated test steps in order to reach a determination on the extent to which these projects implemented a particular aspect of a leading practice. We identified potential case study projects based on data provided by DHS from the Investment Evaluation, Submission, & Tracking system. We determined that the data in the Investment Evaluation, Submission, & Tracking system was sufficiently reliable for our use in selecting projects for our case studies. We selected case study projects, rather than programs, because, according to DHS officials from OCIO, programs report software development life cycle data to the Investment Evaluation, Submission, & Tracking system at the project level only. We selected only the projects supporting programs on the Major Acquisition Oversight List because these programs are expected to comply with the Agile instruction and acquisition management policy. We then further limited the scope of projects to those within components where GAO has not previously assessed a program using Agile methods or was not in the process of assessing such a program. This excluded the U.S. Citizenship and Immigration Services, Federal Emergency Management Agency, Transportation Security Administration, and U.S. Secret Service. We then further refined our selection based on the following criteria: Software development life cycle methodology (iterative development only) Project completion date (in-progress only) DHS component (selection of only one project per component) We then selected a random sample of three projects, with no more than one project selected from a component. The three case study projects we selected were the U.S. Coast Guard (USCG) Command, Control, Communications, and Computers, Intelligence, Surveillance and Reconnaissance (C4ISR) program New Asset Acquisition Offshore Patrol Cutter project, with particular attention to the SeaWatch portion of this project; the U.S. Customs and Border Protection (CBP) Biometric Entry Exit (BEE) program Air Exit project, with particular attention to the Traveler Verification Services portion of this project; and the U.S. Immigration and Customs Enforcement (ICE) Student and Exchange Visitor Information System (SEVIS) program 8001 project, with particular attention to the SEVIS modernization portion of this effort. In preliminary interviews, we confirmed that these projects were applying Agile practices in order to validate data reported to the Investment Evaluation, Submission, and Tracking system. These case studies were used to supplement our findings from our program process and team activity and dynamics-level evaluations of the department’s implementation of leading practices for adopting Agile development. To evaluate case studies’ implementation of these leading practices, we reviewed artifacts from the selected projects. In particular, we reviewed artifacts demonstrating a project’s use of Agile including testing metrics, evidence of Agile ceremonies, the existence of user stories and a backlog, and the availability of Agile coaching and training. We then interviewed officials responsible for program and project management and representatives of groups responsible for software development for the three selected case study projects to discuss gaps we identified. We shared our initial assessment with DHS, USCG, CBP, and ICE and obtained feedback and additional supporting documentation. Regarding our analysis of project implementation of the program process and team activity and dynamics core elements, we followed the aforementioned process in assessing a core element as being “met”, “partially met”, or “not met”. These assessments were used to gain insight into the extent to which DHS policy, procedures, and guidance prepared programs and projects for the successful adoption of Agile leading practices. We did not evaluate the projects in order to make specific recommendations to the individual projects. We conducted this performance audit from December 2017 through April 2020, in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: DHS Agile Action Plans In June 2017, Department of Homeland Security (DHS) senior stakeholders endorsed Recommendations Action Plans: Agile Acquisition Pilots, developed by the Agile Acquisitions Working Group. These recommendations were an effort to sustain the success of the information technology (IT) acquisition and delivery pilot program. The action plans were developed in response to the February 18, 2016, Acquisition Decision Memorandum from the Under Secretary for Management, which recognized the expressed need for both components and headquarters directorates to continue driving organizational change and process improvement to DHS IT acquisitions and delivery. The action plans were intended to codify lessons learned and recommendations based on independent interviews and retrospective meetings with those who participated in the five acquisition pilots. These plans were organized by priority: 12 critical, three high, and three moderate. The recommendations were weighted against one another based on impact, level of difficulty, and alignment with the original five goals of the Agile acquisition pilot program charter: reduce risk, increase customer value, faster time to market, economic value, and increased accountability and oversight. Table 3 describes the DHS Agile action plans, including the associated goal, primary organization(s), level of difficulty, impact, and executive priority. Each DHS action plan included a problem statement and recommendation, as detailed in table 4. Appendix III: Leading Practices for Adopting Agile Development—Agency Environment This appendix describes in detail our evaluation of the three leading practices for agency environment when adopting Agile development, including a further explanation of expectations for each practice as well as some of the findings associated with each practice. We do not present any additional recommendations from these findings; this information is intended to assist the Department of Homeland Security (DHS) in implementing the recommendations described in our report. Agency activities support Agile methods Establish appropriate life cycle activities Agency activities should support Agile methods by allowing for incremental and iterative software delivery that is tailored to the cadence of Agile software development and by incorporating technical reviews that occur throughout the development process. These activities and supporting policy and guidance should allow for requirements to be changed during development and the requirements change approval process should not impede the cadence of iterative and incremental development. Life cycle activities should also be user-focused and call for collaboration between the development team and users. To manage its multi-billion dollar investments, DHS has established policies, procedures, and guidance for IT program management. These publications govern the complete life cycle of a system, from technology development through integration and testing and, finally, implementation and operations and maintenance. DHS has outlined its policies, procedures, and guidance in several documents to assist its components in the acquisition and implementation of software development. Policies for managing its major acquisition programs are primarily set forth in a directive and supporting instruction. These policies outline an acquisition life cycle framework (ALF) that includes a series of predetermined milestones—known as acquisition decision events—at which the Acquisition Decision Authority reviews a program to assess whether it is ready to proceed to the next phase of the ALF. DHS’s Under Secretary for Management serves as the Acquisition Decision Authority for the department’s major acquisition programs. A separate DHS instruction and associated guidebook outline a framework of major systems engineering activities and technical reviews, collectively considered the systems engineering life cycle (SELC), which should be conducted by all DHS programs, both major and non-major. The SELC helps to ensure that appropriate systems engineering activities are planned and implemented and that a program’s development effort is meeting business needs. The SELC consists of major activities and a set of related technical reviews and artifacts that fit within the acquisition life cycle. Figure 2 depicts the acquisition life cycle and associated technical reviews established in DHS acquisition management policy. DHS provided programs with flexibility in their SELC technical reviews. Within the ALF, Agile processes are applied primarily within the obtain phase, where design, development, testing, and implementation of a system takes place. Prior to entering the obtain phase, a program selects its software development approach, such as Agile. The agreed-upon approach is then codified in an SELC Tailoring Plan, which is approved at acquisition decision event 2A. The SELC Tailoring Plan identifies the technical reviews and artifacts that the program is responsible for completing based on its unique characteristics (e.g., scope, complexity, and risk). To assist in tailoring efforts and further guide the implementation of Agile, DHS published an Agile instruction that includes the scope, definitions, roles and responsibilities, and procedures for establishing an Agile framework for developing all DHS IT acquisitions. DHS supplemented this instruction with an Agile instruction manual and provided a template that Agile programs can follow to tailor their activities. For example, instead of holding a system definition review, an Agile program is encouraged to conduct a release planning review (which encompasses the development and release of a segment of software). This optional approach to tailoring a technical review is depicted in figure 3. Outside of technical reviews, DHS updated acquisition policy in February 2019 and associated guidance in May 2019 to allow programs greater flexibility in the larger ALF. The Director of Strategic Technology Management (STM) stated that, under the previous acquisition policy and guidance, IT programs were using in-house expertise due to limited funding to prepare for the 2B decision, when full program funding was received. He noted that, by the time a contract was awarded for development following a 2B decision, the contractor might or might not have been using planning artifacts developed by the program and instead might have recreated them, thereby rendering 2 to 3 years of work useless. The Director stated that programs were unable to fully flesh out the program architecture and other key aspects of the program because programs did not receive funding until the 2B decision and in-house expertise was limited. For example, if a program had not proven out its architecture prior to a 2B decision, it could continue to refine and modify the architecture during the course of development, thereby impacting productivity and quality. DHS updated acquisition management policy and guidance to modify the requirements for the acquisition decision events and addressed a related GAO recommendation. DHS policy and guidance also allowed for programs to modify requirements over the course of development. The traditional process for requirements may be modified as part of tailoring the SELC in order to allow for increased flexibility. The DHS Requirements Engineering User’s Guide detailed requirements engineering steps, activities, and methods for performing those steps. DHS developed this user guide to supplement SELC policy and guidance. One section of this guide focused on Agile development. According to the guide, requirements are broken down over the course of the ALF and commitments are made at different levels of specificity. Fundamental capability gaps are defined in the mission needs statement presented at acquisition decision event 1. Subsequently, the analyze/select phase would ultimately define the high-level features and functions of each required capability, define the fundamental performance of those high- level features and functions, and establish the business case to support approving the acquisition at an acquisition decision event 2A. Often, a preliminary concept of operations is developed and delivered with the mission needs statement. The guide also states that the activities to evaluate these potential alternatives will ultimately result in a preferred solution with defined business practices, methods, and processes that allow the development of business epics and associated architecture epics. Business epic is an Agile term that defines the high-level “stories” that describe a capability, or what the new system is required to perform. Architectural epic is an Agile term that defines the architecture the system will be incorporated into. In addition, the preferred solution would have defined high-level performance requirements (stated from the operational perspective) in terms of how well the solution must perform to be operationally effective and suitable. Key constraints such as security, Section 508 compliance, privacy, reliability, etc., should also be identified. These top-level requirements will be documented in the operational requirements document. According to the guide, Agile teams capture the capabilities and constraints (essentially the functional and non-functional requirements that reflect the business epic level of performance) in an artifact called the capabilities and constraints document. Requirements statements in this document should follow the standard “shall statement” format for ease of translation between the operational requirements document and the capabilities and constraints document. The capabilities and constraints document and its contents mature over time and, as the document matures, business and architectural epics decompose to features/functions or themes, and ultimately to user stories that reflect the specific tasks that users will perform. Officials within the DHS Joint Requirements Council noted that headquarters involvement occurred at this level to approve the high-level operating requirements. After headquarters oversight and approval, the program may then decompose requirements as part of planning for and executing technical reviews. If tailored into an Agile program, the capabilities and constraints document should drive the development of a backlog. The backlog is a list of all the user stories that describe what the system needs to do. The backlog should become more refined as the program decomposes the high-level features (a service that fulfills a user need) and functions down to specific stories that an individual software developer will code and test during a specific iteration. To prevent the backlog from becoming unmanageable, DHS guidance stated that backlogs may be established at different levels. For example, the business and architectural epics along with the associated operating requirements would constitute the “program backlog.” Sub-epics are usually broken down into “high-level features” with business epics broken down into business features and architectural epics broken down into architectural features. Features or functions are decomposed into detailed stories that are then allocated to a “release”. The list of user stories in a specific release constitutes the release backlog. This process of decomposing stories continues to the iteration backlog. DHS guidance places an emphasis on end user needs. The Requirements Engineering User’s Guide raised the importance of identifying stakeholders, including system users, and capturing the needs of those users via requirements or, in the case of Agile, user stories. The Agile instruction manual placed an emphasis on the importance of users to a program and articulated that the product owner represent the user community and was expected to continually seek ongoing feedback and elicit requirements from users. The Agile core metrics also strongly recommended the use of a net promoter score. This score was one mechanism for measuring customer satisfaction through asking users to rank how likely they would be to recommend the system or application to a friend or colleague, based on a score of 1 to 10. Clearly align goals and objectives Program goals should clearly reflect stakeholder needs and concerns based on input from stakeholders and stakeholder review and approval. Program goals should align with strategic IT objectives. Software-related goals should be defined and clearly aligned with program goals. The agency should collect objective measures that are well defined to track progress towards achieving software goals so the agency knows which features and capabilities have been achieved. The Requirements Engineering User’s Guide described program expectations for tracing from mission needs to operating and functional requirements, or user stories. The guide recognized that, as a program progressed through the ALF and SELC, it was important to trace requirements from the top-level mission needs or capabilities and/or business requirements down to the system/sub-system, component, or configuration item level that enabled those requirements to be met. This helped ensure continuity across various DHS artifacts, such as the program’s mission need statement, concept of operations, and operational requirements document, to vendor specifications (or applicable equivalent artifacts). Although an Agile program will modify the SELC to accommodate its needs, generally programs were expected to follow the same conceptual approach to the requirements of planning, development, and management. The user’s guide stated that collaboration among the various stakeholders was important and the program requirements team must continuously work to establish partnerships and networks. To do so, the guide stated that the program team must identify all individuals and organizations that may be impacted by their program and ensure those stakeholders were engaged throughout development to facilitate understanding of their perspectives and needs. The first step was to identify applicable stakeholders, which would include end users, program sponsors, developers, maintainers, trainers, and other affected individuals or organizations. The program requirements team then solicited input from these stakeholders to understand their needs, policies, processes, and operations to begin the requirements definition effort. It identified some ways a team might begin the process of eliciting requirements from the stakeholders. After collaborating with stakeholders, the stakeholder needs must be translated into the program requirements, or goals. The guide stated that the program requirements team should take the inputs from the various stakeholders and decompose, prioritize, de-conflict, and validate the needs identified. It clarified that a “good” requirement was achievable, testable, clear, concise, technology-independent, feasible, and able to stand alone. The guide grounded all of the requirements elicitation and development process in the overall contribution to the agency mission, recognizing the need for general strategic alignment. In particular, the guide noted that requirements were “mission need” driven as opposed to “solution” driven. Requirements were developed throughout the life of a program, with the first formal requirements being the operating requirements documented in the operational requirements document. To ensure that DHS’s mission or strategic goals were key inputs for decision making, DHS relied, in part, on its enterprise architecture process. DHS policy for enterprise architecture stated that the enterprise architecture program provided a vehicle to tie the strategic mission goals and objectives of DHS to the business processes, information resources, and technology investments necessary to reach key performance outcomes. This methodology was intended to integrate IT into the mission and strategic priorities of DHS, which provided the core foundation for all subsequent processes. DHS capital planning and investment control guidance reinforced this fact, stating that the Federal Acquisition Streamlining Act of 1994 required capital investments to align with mission and strategic goals. This included the framework within which the department formulated, managed, and maintained its portfolio of investments as critical assets for achieving success in the DHS mission and alignment to the DHS IT Strategic Plan and the DHS Strategic Plan. Agency culture supports Agile methods Cascading sponsorship for Agile software development Senior stakeholders should support and model the use of Agile, along with its values and principles, through explicit policy or guidance impacting the business and should take steps to complete responsibilities defined in agency Agile policy or guidance. Agile should also be supported in all relevant areas of the business impacting a software development project through the use of Agile sponsors. These sponsors should represent the lines of business in key agency decisions on Agile. Senior stakeholders at DHS demonstrated support for Agile through the publication of policy and guidance that established Agile development as the department’s preferred approach for software development. As discussed previously, the department published Instruction 102-01-004 Agile Development and Delivery for Information Technology (Agile instruction), which provided the scope, definitions, roles and responsibilities, and procedures to establish an Agile framework for the development of IT acquisitions at DHS. Specifically, the Agile instruction established responsibilities for the CIO, the Chief Procurement Officer, the Chief Financial Officer, the Director, Office of Test and Evaluation within the Science & Technology Directorate, and the Executive Director of PARM. Each of these five stakeholders and their associated components demonstrated their support for Agile development by taking steps to complete their responsibilities defined in the Agile instruction. For example, the Office of the Chief Information Officer (OCIO), the Office of the Chief Procurement Officer, and the Director, Office of Test and Evaluation within the Science and Technology Directorate all had responsibilities related to providing guidance for the implementation of Agile within their specific area of expertise. All three components had taken steps to execute these responsibilities, such as by publishing the Agile instruction manual, providing supplementary guidance for test and evaluation in an Agile environment, and offering elective training on contracting strategies for Agile services. Representatives from offices with a role in software development also supported Agile via membership in the IT Program Management Center of Excellence (ITPM COE). In addition to the stakeholder organizations identified in the Agile instruction, the ITPM COE membership included representatives from the Joint Requirements Council and the Chief Privacy Officer. According to the ITPM COE charter, the ITPM COE served as a cross-functional team to identify and promote best practices, provide tools, and coordinate assistance for programs and projects to maximize the successful management of DHS IT investments. This included making progress towards the 18 Agile action plans that resulted from the Agile acquisition pilots. The ITPM COE membership requirements called for representatives of the member organizations to be involved in key decisions regarding Agile. According to the Director of STM within OCTO, ITPM COE members were selected and approved by their organization’s executives. The ITPM COE charter stated that these representatives must be authorized to represent or make decisions on behalf of their officers or organizations. Officials from all ITPM COE member organizations expressed support for the ITPM COE and confirmed that their component was appropriately represented in decision making. This was represented, in part, by the fact that at least one representative for each ITPM COE member group attended at least half of the meetings. For example, at least one representative from the Science and Technology Directorate attended approximately 95% of the meetings. Sponsor understanding of Agile software development Sponsors should understand and communicate changes resulting from Agile development. Sponsors should attend training or receive coaching on Agile and the agency’s framework, the agency should monitor completion of training, and sponsors should transmit learning from training to staff. Sponsors should also commit to achieving those intended results and sponsor performance should be tied to achieving those intended results. The Director of STM stated that Agile sponsors were considered to be chief executive officers (e.g. Executive Director of PARM and the Deputy Under Secretary for Management). They oversaw the actions of the ITPM COE and approved the Agile action plans in June 2017. DHS did not ensure that Agile sponsors attended training or received coaching in Agile development. The department made training available for Agile, including courses such as those required for acquisition professionals. However, in a written response, the Office of the Chief Human Capital Officer stated, and the Director of STM confirmed, that the department did not administer mandatory training on Agile for Agile sponsors. The department also did not monitor the completion of sponsor training in Agile. Although DHS employees leveraged the Federal Acquisition Institute Training Application System to track their training and certifications, the department was not using this system to monitor sponsor training in Agile. According to a written response from the Office of the Chief Human Capital Officer, the department did not keep a record of whether sponsors completed training in Agile because the department did not require Agile training specifically for sponsors. DHS Agile sponsors exhibited support for achieving the intended results from the transition to Agile. Agile sponsors committed to achieving these results through an endorsement of the 18 Agile action plans and the associated implementation plans. However, DHS did not demonstrate that Agile sponsor performance was tied to achieving the intended results of the transition to Agile. According to a written response from the Office of the Chief Human Capital Officer, the department’s employee performance management policy did not specifically address Agile. This written response further stated that addressing Agile in these policies was unnecessary because the Office of the Chief Human Capital Officer incorporated goals derived from project plans in individual performance plans. DHS policy and guidance for performance management identified individual performance goals as a component of employee performance, but the department did not provide evidence that specific performance plans for the sponsors were linked to such goals. Establish an environment supportive of Agile software development Team dynamics should be facilitated through access to common team rooms and/or modern communication and social media methods and headquarters infrastructure operations should allow for communal spaces and co-location in program offices. A headquarters technical environment should allow access to tools by programs to foster distributed communication, and there should be a process for continuous feedback on the Agile environment and modifications to that process (e.g. communities of practice, routine working group sessions). Agency governance bodies should allow programs greater autonomy and flexibility within existing acquisition processes through the modification of gate reviews and other touchpoints in the acquisition process for Agile projects and increased transparency for governance bodies into project operations when necessary. DHS policy and guidance allowed for team dynamics to be facilitated through access to common team rooms and modern communication methods. In addition, department policy promoted and allowed program offices to support team dynamics through the use of communal spaces and co-location. Specifically, the Director of Systems and Information Integration within the Chief Readiness Support Office confirmed that DHS had modified policy related to infrastructure operations to allow any office to reorganize their space, citing the USCIS Transformation program as an example of this reorganization. The Director of Systems and Information Integration also noted that he was not aware of any restrictions to this practice in policy. With respect to facilitating access to modern methods of communication, DHS offered programs the option of using a suite of tools that included those for distributed communication. DHS took multiple steps to establish a process for continuous feedback related to the department’s Agile environment and process modifications. According to the Director of STM, OCIO built support for Agile through the Centers of Excellence, communities of interest, brown bag lunches, and public speaking engagements. The Director added that these sessions facilitated the discussion of Agile and could be used to compile feedback. The Director of STM explained that, as this feedback came in, it was either addressed immediately or put into a backlog. Efforts to further streamline the acquisition process were tracked via Agile action plan 6. The department’s governance bodies also increased transparency into project operations when necessary. The Agile Development and Delivery for Information Technology Instruction Manual (Agile instruction manual) stated that the program or project manager should coordinate with the various oversight bodies that govern IT development. These bodies varied depending on the level of investment, but, for major programs, executive steering committees were often established to oversee all aspects of program planning and execution between major acquisition decision events. In addition, PARM officials stated that DHS increased the frequency of acquisition review board reviews and modified the content presented at the reviews to allow it to be more actively involved with projects earlier in the acquisition life cycle. Specifically, PARM updated the Acquisition Review Board slide templates and informed us of its intent to update acquisition management policy to require Agile projects to hold Acquisition Review Board reviews once every six months, as opposed to once every 12 months. Align incentives and rewards to Agile methods The agency should establish an incentive and reward structure promoting team successes and the value of individuals within those teams. Management should establish agency goals to align incentives and rewards with Agile methods. Goals for incentives and rewards should align with the agency’s goal(s) and focus on team success. The agency should allocate incentives and rewards based on team success. DHS did not establish an incentives and rewards structure that promoted team successes and did not demonstrate that management had established agency goal(s) to align incentives and rewards with Agile methods. Furthermore, the department did not demonstrate that human resources and others were actively involved in setting goals for incentives and rewards alignment. DHS guidance specifically discussed contract incentives for Agile projects. For example, the Agile instruction manual suggested that consideration be given to address the duration of the base term and options, scalability, deliverables, and pricing with a mindset that contractors need appropriate incentives to encourage them to perform well. The manual also stated that contract award terms could provide a greater incentive for contractors working on longer-term Agile projects. Although the department made efforts to adapt incentives and rewards for contractors supporting Agile projects, it acknowledged that it did not update existing incentives and rewards for federal employees working on Agile projects. Officials within the Office of the Chief Human Capital Officer stated that existing human capital and performance plan policy allowed for rewarding and incentivizing Agile teams as well as individuals. These officials further noted that DHS had numerous opportunities to recognize and reward team or individual performance, regardless of the development methodology a program relied on. Specifically, these officials clarified that the Office of the Chief Human Capital Officer used project plans to set goals and included those goals in employee performance plans. As these officials felt the existing performance plan policy was sufficient, they did not believe additional guidance or modifications to existing policy were necessary. Agency acquisition policies and procedures support Agile methods Guidance is appropriate for Agile acquisition strategies Agency acquisition policy and guidance should support awarding contracts for the unique needs of an Agile program. Acquisition strategies should recognize the need for interim delivery of software, allow for close coordination between the contracting office and program office staff, and allow for changing requirements and contract oversight mechanisms to be tailored to support Agile development methods. DHS offered guidance for preparing acquisition strategies through its Procurement Innovation Lab. Webinars offered by the Procurement Innovation Lab on acquisition strategies for Agile programs discussed the need for interim delivery of software, close coordination between contractors and program office staff, contract oversight mechanisms that were tailored to support Agile development, and changing requirements. For example, the “Transportation Security Administration Agile Services Procurement” webinar discussed planning, executing, and de-briefing technical demonstrations used to select the contract recipient, paying particular attention to the value of transparency and modifying contract oversight mechanisms. Officials from the Office of the Chief Procurement Officer clarified that the webinars were available as needed and were not required training. DHS also published Agile guidance that discussed contracting and acquisition strategies. From an oversight perspective, according to the Agile instruction manual, DHS executive steering committees oversee all aspects of program planning and execution between acquisition decision events. This authority extends to assisting programs in developing acquisition strategies where appropriate. The manual included a section that specifically called out Agile contracting considerations that pointed back to Office of Management and Budget Contracting Guidance to Support Modular Development, the TechFAR handbook, the Digital Services playbook, and innovative contracting case studies. Among other useful information in the Agile instruction manual were key contracting considerations for an Agile program or project manager. These considerations included, among other things, frequent, iterative deliveries of software, an ability to monitor changes to maintain contract and project scope, flexibility to accommodate refinement of requirements, transparency and collaboration, and prior experience in the Agile methodology. The manual also highlighted goals for the acquisition to discuss with a contracting officer, such as rapid contracting processes to keep pace with Agile development, contracting to accommodate incompletely defined scope and requirements, and the ability to respond to requirements changes without requiring extensive change orders. According to officials within OCTO and the Office of the Chief Procurement Officer, the department also supported Agile programs in preparing acquisition strategies through the IT acquisition review process. This process was established to provide a mechanism for the DHS Chief Information Officer to review and guide agency IT expenditures. The process was intended to analyze IT acquisitions to ensure alignment with DHS missions, goals, policies, and guidelines. This process relied on subject matter experts to assist in the review of IT acquisitions, including one for Agile reviews. According to the IT Acquisition Review Essentials Guide, Agile reviews occurred where software was being developed to ensure development activities adhered to Agile best practices and DHS SELC guidance. The Agile subject matter expert was expected to review acquisition materials against an established set of criteria for both the acquisition plan and the requirements document. For example, when reviewing the acquisition plan for approval, the subject matter expert should consider if the statement of need adequately addresses Agile or iterative project-specific activities and/or deliverables. The Director of STM stated that there is one staff member in STM who actively participates in the IT acquisition review process and was responsible for ensuring Agile language was correctly implemented in contract statements of work. The Director also added that they were willing to help teams that were having trouble providing explanations of Agile processes in their statements of work. The Director of STM stated that there was no policy to guide his staff member reviewing Agile language in the statement of work, but that he asked his division to put together a checklist review to govern this process. The Director added that the department sent programs and projects requiring assistance with Agile contracting to the Procurement Innovation Lab by request to streamline the acquisition plan. According to the Leader of the Procurement Innovation Lab Team, the Office of the Chief Procurement Officer was primarily focused on supporting Agile pilot programs, such as the Federal Emergency Management Agency Grants Management Modernization program. The team leader noted that, while the procurement office supported these programs, it relied on the program offices to ensure accuracy. For example, the program management office ensures that the requirements are structured and delivered, which could be challenging for Agile programs. The team leader mentioned that a particular focus at the moment was defining the pricing for contract line item numbers in such a way as to afford the flexibility needed for Agile development while still holding contractors accountable. Appendix IV: Leading Practices for Adopting Agile Development—Program Processes This appendix describes in greater detail our evaluation of the three leading practices for program processes when adopting Agile development. It does not present new findings; rather, the information is intended to assist the Department of Homeland Security (DHS) in implementing the recommendations described in this report. Program processes refer to leading practices related to the program office and technical environment. For programs to successfully transition from processes used for traditional development projects, programs should ensure that staff are appropriately trained in Agile methods by ensuring Agile teams have the appropriate technical expertise needed to perform their roles technical environments enable Agile development through making technical and project support tools available, and designing a system that supports iterative delivery project planning controls are compatible with Agile methods by maintaining a sustainable development pace and tracking and defining and incorporating non-functional requirements in defining and incorporating critical features in development The department develops an environment that supports these processes. Within DHS, program management offices are responsible for planning and executing individual programs and implementing applicable Agile methodologies. In addition, the DHS Office of the Chief Information Officer (OCIO) is responsible for setting policies and procedures to ensure that programs leverage Agile development best practices to meet the department’s goals and are within acquisition policy. The DHS OCIO is also responsible for providing guidance for and reviewing the adoption and execution of Agile development. Staff are appropriately trained in Agile methods Train all program staff in Agile methods The agency should provide a training program in Agile for staff and track and monitor the training. All members supporting the team, not only the software development team, should be trained in the specific Agile framework they will be using. DHS required its acquisitions workforce to take training that incorporated Agile methods. DHS Instruction 102-01-006, Acquisition Program Management Staffing, established certifications for key acquisition career fields, which included training requirements. According to the Associate Director for Training from the Homeland Security Acquisitions Institute, the certification requirements included training that has been updated to incorporate Agile methods. Specifically, the department updated course content for AQN 101: DHS Fundamentals of Systems Acquisition to include Agile development concepts, such as small team management and Agile metrics, following the issuance of department policy governing Agile development. This course was required training for seven of the acquisition career fields, including program and project managers, systems engineers, and test and evaluation managers. DHS tracked and monitored the completion of training requirements for the acquisitions workforce. According to DHS Directive 064-04, Acquisition Professional Career Information, component acquisition executives were responsible for ensuring that acquisition personnel met the mandatory training requirements. Officials from the Homeland Security Acquisitions Institute within the Office of the Chief Procurement Officer stated that DHS employees leveraged the Federal Acquisition Institute’s training application system to track their training and certifications. According to the catalog of product services of the institute, members of the DHS acquisition workforce were required to attach copies of their training certificates to request certification of completion of the required training. Because the DHS acquisitions workforce may not cover all personnel staffed to Agile projects, some program staff may not be subject to training requirements that incorporate Agile methods. According to the Director of the Homeland Security Acquisition Institute, certain Agile team members, such as the product owner, were not necessarily classified as part of the acquisitions workforce. For example, according to the U.S. Immigration and Customs Enforcement (ICE) Student and Exchange Visitor Information System (SEVIS) program staffing plan, the product owner role was not part of the acquisitions workforce and did not require any certifications. To help address the Agile training needs of all staff, including those who are not part of the acquisitions workforce, DHS also provided elective training in Agile methods. The department offered commercial training through the Homeland Security Acquisition Institute, such as acquisition of Agile services and Agile requirements for creating user stories. The DHS Agile instruction manual also identified training offered by the U.S. Citizenship and Immigration Services Office of Information and Technology as another resource on Agile concepts, such as user stories and automated testing. In addition to elective training, the Agile Development and Delivery for Information Technology Instruction Manual (Agile instruction manual) encouraged program managers to seek out an Agile coach to help teams adopt Agile methods and supplement training. The instruction manual suggested that program managers should identify an Agile coach to serve as an embedded trainer, consultant, and team advisor. This Agile coach could help the team adapt Agile methods to their environment and work through challenges. An Agile coach could also help individual team members understand the responsibilities of their role on an Agile team. Although DHS did not provide coaches for Agile teams, the department offered resources that could help programs select and obtain an Agile coach. First, the department established a blanket purchase agreement for programs to acquire Agile development support in the form of hands- on coaching services for the design and use of Agile methods. According to Homeland Security Acquisition Institute officials, this agreement would enable programs and projects to acquire Agile coaching. Among other things, this agreement defined the scope of Agile coaching services and their pricing so that programs would not need to develop these terms on their own. Second, the Agile instruction manual included considerations to help program managers select a qualified Agile coach. For example, the instruction manual encouraged program managers to collaborate with contracting officials to identify an Agile coach who had demonstrated successful past performance on projects implementing similar technology and Agile methodologies. The U.S. Customs and Border Protection’s (CBP) Biometric Entry Exit (BEE) program’s Air Exit project provided informal training for new team members that included a discussion of Agile methods. According to the Air Exit project manager in the Office of Information and Technology, new team members received onboarding training that covered CBP’s approach to Agile methods. The project did not track attendance for this onboarding training, but an Air Exit project manager noted that team members were incentivized to attend the training in order to learn how to satisfy their responsibilities. The Scrum master for the Air Exit project stated that this training was also available to the team as a refresher course approximately every fiscal quarter. The BEE program also relied on an Agile coach to support the Agile team. According to the Agile coach supporting the Air Exit project, this role included training for the Agile team on basic Agile topics and working with the team on their use of a project management software tool. According to a project manager within the Office of Field Operations Air Exit project management office, the Agile coach that supported the project was instrumental in designing the CBP Office of Information and Technology’s Agile development program beyond the BEE program. Ensure Agile teams have the appropriate technical expertise needed to perform their roles The agency should have policy or guidance in place to help programs ensure Agile teams have the appropriate technical expertise. A program should also consider Agile-centric skills when forming teams. In addition, programs should define requirements for contractor proposals and evaluate contractor proposals for Agile services (e.g., source selection). DHS guidance provided programs with considerations for forming teams with Agile-centric skills. The DHS Agile instruction manual stated that a development team with experience in Agile practices can mitigate risks to on-time delivery. This experience included Agile processes as well as technical skills, such as automated testing. In the context of the Agile Scrum methodology in particular, the Agile instruction manual stated that teams needed to be cross-functional and have all of the skills required to deliver a project from conception to delivery. To enable teams to deliver a project from conception to delivery, the Agile instruction manual stated that program managers should seek team members with general skills. The manual advised that team members should contribute to routine development activities and possess cross- functional expertise that allows the team to achieve work without depending on individuals outside of the team. For example, in Agile development, testers are part of the development team and should therefore possess both testing and development skills. In addition, the instruction manual stated that, according to industry experts, program managers should seek some overlap in team member’s skillsets to mitigate risks associated with a key person becoming temporarily unavailable. DHS guidance further provided programs with considerations for defining requirements in solicitations for contract proposals for Agile services. For example, DHS supplemental guidance for incorporating testing and evaluation into contract requirements noted that contracts should specify government test and evaluation staff, as well as contractors, on the development team in order to access the test data they need. The DHS IT acquisition review process also helped to ensure that requirements were defined in solicitations for contractor proposals. According to the Information Technology Acquisition Review Essentials Guide, Agile subject matter experts in the department review proposed contracts to ensure that they will enable development activities that adhere to Agile best practices and DHS systems engineering life cycle (SELC) guidance. For example, Agile subject matter experts should assess whether contract requirements documents, such as the statement of work, are prepared in terms that will enable vendors to clearly understand the Agile requirements. The department also provided guidance to assist programs in evaluating contractor proposals for Agile services. The Agile instruction manual noted that programs can consider certifications in various Agile methodologies and recommended that programs coordinate with contracting officials to review vendors’ past performance in implementing Agile methods. In addition, the department established the Procurement Innovation Lab within the Office of the Chief Procurement Officer to help programs address challenges in procuring Agile services, such as validating contractor qualifications. According to a Procurement Innovation Lab team leader, the lab shares lessons learned from Agile services contracts via webinars, which are available to staff on an as-needed basis. Several of these webinars highlighted the value of using technical demonstrations to validate the qualifications of vendors. The ICE SEVIS program provided training for all team members, including contractors, to ensure they had the necessary Agile-centric skills and expertise. A team process agreement for one development module showed that the technical lead, development team, test engineer, and Scrum master roles were filled by contractors, while other positions such as the project manager, product owner, and test automation subject matter expert roles were filled by government employees. According to the ICE SEVIS program manager and Scrum master, the program provided training for contractors that covered Agile processes as well as technical and project management support tools. In addition, some government employees took role-specific training. For example, the program’s test automation subject matter expert completed training in continuous integration and test automation. To further ensure contractors on ICE SEVIS Agile teams had the necessary Agile-centric skills, the ICE SEVIS program defined the Agile methodology and necessary technical expertise for contractors in the contract requirements. For example, the performance work statement for one development module required contractors to use the program’s management software tool to track user stories. The performance work statement also required use of the program’s continuous integration and automated testing tools. The terms and conditions for this contract also identified the required experience for key personnel, such as proven experience working in an Agile environment. The ICE SEVIS program also evaluated contractor qualifications to ensure they had the necessary technical expertise. According to the program manager, contractor qualifications were evaluated in two stages; first, by assessing the contractor’s proposal, and second, by conducting a technical challenge to ensure that contractors could demonstrate the technical skills in the proposal. According to the instructions included in the request for contractor proposals, this technical challenge required the contractor to leverage Agile best practices to design, develop, and demonstrate working software that addressed user stories provided by the program. Although the instructions stated that contractors were required to follow Agile methods, the ICE SEVIS program manager stated that the primary goal of the technical challenge was to assess development skills rather than knowledge of Agile. Technical environments enable Agile development Agency policy or guidance should call for technical and project tools to be available to support Agile development and for system design that will support iterative delivery. Make technical and project support tools available Project management and technical support tools should be integrated into a program’s technical environment, where appropriate. The tools within this technical environment should be readily available to Agile teams. DHS policy and guidance called for Agile projects’ technical environments to support Agile methods. The department published guidance for standing up technical environments specifically for Agile projects. For example, the DHS Agile instruction manual identified the benefits of using program support tools for tracking program progress, reporting on that progress as part of program governance, and automating tests within an Agile technical environment. The manual stated that a program or project manager is responsible for fostering an environment that enables the Agile team to succeed, including obtaining the appropriate tools. To supplement this guidance, DHS offered a suite of tools that Agile programs could access. The suite of tools was referenced in a checklist of activities for program or project managers in the Agile instruction manual. According to an IT specialist from the Technical Architecture and Engineering division within the Office of the Chief Technology Officer (OCTO), the tools available included program management tools as well as technical tools. The specialist stated that OCTO provided programs with access to this suite of tools to build support for and familiarity with the tools, evaluating any requested plug-ins from programs and doing their best to accommodate them. The ICE SEVIS program defined the technical environment to include technical tools for automated testing and continuous integration. The team process agreement for one of the program’s development modules identified technical tools that supported continuous integration and testing within the program’s technical environment. This included Jenkins for continuous integration as well as MUnit and Soap UI for continuous testing. In addition, the ICE SEVIS Modernization Test and Evaluation Master Plan discussed that tools for helping to ensure code quality, such as an automated code analytics tool, should be used to identify test coverage of code and cybersecurity code vulnerabilities. The program also defined management support tools in the process agreement. Specifically, it identified support tools for tracking and knowledge management, such as JIRA and Confluence. The team process agreement stated that JIRA should be the main knowledge management tool and that all changes, discussion, and history should be tracked in each ticket. This process agreement also stated that JIRA should be the team’s tracking tool with Confluence used to provide transparency. Design a system that supports iterative delivery The agency should adopt policy or guidance that allows project designs to develop modular system components and the program should establish a loosely coupled architecture that allows for modular development. DHS guidance allowed project designs to develop modular system components through upfront architecture planning. The DHS Technical Review Guide advises stakeholders to discuss and approve the technical design of the system, including its top-level architecture, as part of the system definition review. This review should take place prior to development work. For Agile programs, DHS suggested that programs may elect to switch the system definition review with a release planning review. The SELC Tailoring Examples for Selected Types of DHS Acquisition Programs specified that this design discussion should take place as a part of release planning. The department referred to this design as an “architectural runway”, a description that should enable the team to conceptualize how the user stories will be implemented. In exiting the release planning review, the Technical Review Guide noted that programs should answer whether or not an architecture exists, if the architecture enables the deployment of the release, if architecture collaboration is explained and understood for this development process, and if the appropriate resources are available. In addition to transitioning to a release planning review, DHS guidance urged Agile programs to move away from traditional artifacts associated with a system definition review. In this shift from traditional artifacts, the department proposed that programs document software design within a system design document on a release-by-release basis. According to the Requirements Engineering Users Guide, in Agile methodologies detailed design occurs at the iteration level and, as such, the design is documented in an iterative fashion in the system design document. The guide further stated that the system design document allows the development team to communicate the design to others including customers, managers, and other developers and that industry best practice was to represent the design through a series of “design views.” Each software design stakeholder could have a distinct perspective on what are the essential aspects of a software design. Together, these views provide a comprehensive description of the design in a concise and usable form that simplifies information access and assimilation. DHS guidance did not discuss the system design document as a delivered artifact until after the sprint review and demo and a release readiness review had been discussed. At the end of each iteration, DHS guidance stated that the system design document should represent the design of the feature, function, and/or system as it existed at that moment. To facilitate communication between Agile teams and to ensure the most up-to-date description of the design is available, guidance called for the system design document to be developed and maintained in an electronic form using any number of programs or web tools that are available. The Requirements Engineering Users Guide noted that the system design document is to be considered complete when each identified design concern is the topic of at least one design view, all design constraints have been applied, and sufficient detail exists to be an authoritative and primary “code-to” artifact. The system design document should also provide traceability to the feature, epic, and operational requirements document “shall” statements. The SELC Tailoring Examples for Selected Types of DHS Acquisition Programs stated that, prior to releasing software to the production environment, a release readiness review should be conducted. As part of this guidance, the department stated that the intent of this release readiness review included ensuring that all elements of the release were complete, including a system design document. DHS guidance also discussed designing a loosely coupled architecture, another important aspect of project design that facilitates modular development. A member of the contractor support staff for the DHS OCIO stated that the Enterprise Architecture Team was expected to consider modularity and loose coupling generally through consideration of technical complexity. According to DHS Enterprise Architecture principles, technical complexity is to be mitigated in part by the implementation of loose coupling. According to the principles, DHS will incorporate loose coupling into architecture and systems design to minimize the risk resulting from changes within one system necessitating changes within an interoperable system. The BEE Air Exit project design document defined the planned design for the system and addressed design and architectural concerns that could affect the system’s operating environment. As part of this design consideration, the project established a loosely coupled architecture. This loosely coupled architecture was illustrated within the project’s system design document. This system design document defined the Traveler Verification Services software as consisting of two distinct components: 1) traveler verification services core and 2) traveler verification services matcher. The functionality and responsibility of these two components were distinguished throughout the document. Moreover, the document detailed how the Traveler Verification Services software would be delivered as a system of applications, combining an integration layer, business layer, data access layer, and data layer. Project planning controls are compatible with Agile development Agency policy or guidance should call for teams to maintain a sustainable development pace and track and monitor that pace and for non-functional requirements and critical features to be defined and incorporated in development. Maintain a sustainable development pace and track and monitor that development pace The agency should have policy or guidance that calls for Agile projects to establish a sustainable development pace. This guidance should be supplemented by tracking and monitoring the pace. The program should establish a sustainable pace for Agile projects and that pace should be tracked and monitored. DHS guidance called for Agile projects to manage the pace of the software development. The Agile instruction manual stated that Agile projects should consider velocity and burndown rates to track the overall project status and update the project plan to reflect this status. In a separate appendix, the Agile instruction manual also identified metrics for project and program managers and executives to consider in order to monitor how a project was progressing, how Agile was optimizing the use of team members and resources, and where the project stood in terms of key Agile measures. In the list of Agile metrics, DHS highlighted burndown rate and velocity, and offered a description and method of calculation for each. In addition to the Agile instruction manual, the department provided training that spoke to development team pace. For example, the curriculum for lesson six of course APM 350 on managing program execution included a section covering Agile development metrics. Among the metrics discussed were those associated with progress, including velocity and burndown charts. Progress metrics were also covered in other course offerings. However, DHS guidance and training materials did not cover the concept of ensuring a sustainable pace. In order to track and monitor the development team’s pace, the department incorporated several related measures into the Agile core metrics. Among others, programs executing Agile were expected to report on the following pace-related metrics after each iteration: story points planned to be completed, number of production deployment per quarter, and average product deployment lead time. These measures could provide programs and the department with an understanding of the development team’s pace and the extent to which it was or was not sustainable. However, the department was not tracking and monitoring development team pace as intended. The Agile instruction required Agile programs to submit Agile core metrics within six months of the instruction’s publication. However, according to the Director of STM, programs were not consistently reporting these core metrics. According to the Director of STM, the department was still working with programs to ensure they consistently reported the core metrics to the Investment Evaluation, Submission, & Tracking system. The SeaWatch project at the United States Coast Guard (USCG) demonstrated that it was monitoring development pace on a monthly basis. SeaWatch officials stated that they used TAIGA as a tool to manage the overall project and to auto-calculate pace. Additionally, SeaWatch officials stated that contractors delivered a monthly progress report, which contained the accomplishments of each team and a snapshot from the latest TAIGA report. For example, one monthly report for SeaWatch included a burndown chart for the SeaWatch project’s development backlog and the monthly output of user stories and associated story points by development effort that could be used to assess development pace over time. SeaWatch officials stated that the teams used velocity to help plan for the next iteration. Officials added that they tracked the collective velocity of all four teams as they were all working together on the same ship build. In the future, officials stated that this tracking of velocity could also be used to track individual team velocities as necessary. The project demonstrated that it was adapting in order to achieve a sustainable pace. According to the April 2018 monthly report, the team completed 55 user stories worth 500 story points. The following month, in the May 2018 monthly report, the number of user stories dropped from 55 user stories to 17, worth 130 story points. According to the June 2018 monthly report, the team completed a development effort of 31 user stories and 278 story points. According to the SeaWatch acquisition manager, development pace fluctuated because not all sprints were of equal difficulty. The acquisition manager added that the number of completed story points per sprint could also be inconsistent due to inaccurate user story estimates, changes in staff availability from sprint to sprint, and other external factors such as weather. Define and incorporate non-functional requirements in development The agency should have policy or guidance in place for incorporating non-functional requirements for Agile projects and the program should account for non-functional requirements, such as security and privacy, in the program strategy and throughout development. DHS guidance addressed the incorporation of non-functional requirements for Agile projects. According to the Technical Review Guide, non-functional requirements could be governed via a system definition review. According to the guide, this review was required at the end of the requirements definition phase to focus on the completeness of the requirements engineering activities, including the gathering, analysis, and documentation of functional and non-functional requirements. This review assessed the traceability of these requirements to the operational requirements document and concept of operations. In the case of Agile programs, DHS suggested replacing the system definition review with a release planning review. In place of traditional artifacts associated with a system definition review, DHS guidance stated that the capabilities and constraints document, backlogs, and the system design document, which are developed iteratively throughout the release, should document the requirements and provide traceability to the operational requirements document. These artifacts served the function and filled in for the functional requirements document and the system requirements document previously required for a system definition review. The Technical Review Guide noted that, as the capabilities and constraints document matures, business and architectural epics should decompose to features or themes, and, ultimately, user stories that reflect the specific tasks that users will perform. The Technical Review Guide cited as exit criteria that a program or project should answer whether the capabilities and constraints document identified the specific features and non-functional requirements to be addressed in the release. DHS requirements engineering guidance expanded on how Agile programs and projects could manage non-functional requirements. The guidance explained that there were various ways that the constraints or non-functional requirements such as security, Section 508 accessibility, privacy, or reliability could be translated down to the iteration level. It stated that some Agile teams may include these non-functional requirements in the backlog, while other teams may include them as part of acceptance criteria or in an artifact called the “definition of done”. According to officials from the Science and Technology Directorate Office of Systems Engineering, once defined, the day-to-day operations and testing for non-functional requirements were the responsibility of the operational test agent. DHS maintained some governance over non-functional requirements. According to the DHS acquisition management instruction, the operational requirements document should be approved by the Acquisition Decision Authority after validation by the Joint Requirements Council. The operational requirements document should include both the functional and non-functional requirements. Officials from the Office of the Director of Test and Evaluation said that they do not usually provide feedback on the decomposed functional or technical requirements for software development projects, focusing only on the operating requirements, because that is what directly impacts operations. The CBP BEE program’s functional requirements document outlined a series of non-functional requirements as the requirements used to define how the system is to behave as opposed to functional requirements that define what the system should do. The project included 10 non-functional requirements in the functional requirements document. For example, the biometric match service should have an overall availability of greater than or equal to 99%, which included both scheduled and unscheduled downtime. These ten non-functional requirements comprised five related to availability, three related to reliability, one related to scalability, and one related to security. All of these non-functional requirements were scheduled for release as part of the initial operating capability. CBP officials noted that non-functional requirements were also captured within the operational requirements document as measures of effectiveness. According to project officials, measures of effectiveness and other security-related parameters translated into the key performance parameters for the project. Officials noted that these key performance parameters were tracked on a daily basis and that information was fed into a monthly report. The operational requirements document stated that the program’s suitability requirements conformed to the DHS and CBP enterprise architectures and all DHS and CBP infrastructure policies and guidelines. Moreover, it noted that National Institute for Standards and Technology guidance and DHS guidance factored into the development of security related non-functional requirements. For example, system security controls should be compliant with National Institute of Standards and Technology and DHS sensitive system guidelines based on its Federal Information Processing Standard 199 rating for availability, integrity, and confidentiality. Define and incorporate critical features in development The agency should have policy or guidance in place for incorporating critical features for Agile projects. The program should ensure that its strategy considers all mission, architectural, and critical safety components, along with their dependencies, on a regular basis. DHS policy and guidance addressed the incorporation of critical features for Agile projects. As discussed in the non-functional requirements section, programs were expected to document functional requirements via the systems design review or, as recommended for Agile programs, a release planning review. Artifacts associated with these reviews served to capture the functional requirements for the program and should be evaluated as part of the entrance and exit criteria defined in the technical review guide. Additional guidance elaborated on the process for decomposing requirements. Unlike non-functional requirements, applicable exit criteria on critical features expanded into the solution engineering review. This criteria included questions devoted to critical features and how they tied back to performance measures (e.g. key performance parameters). According to the Director of STM, headquarters oversight of critical features was limited to the higher-level requirements defined in the operational requirements and concept of operations documents. The ICE SEVIS program captured critical features in documents required by department acquisition management policy and guidance. The ICE SEVIS Modernization Concept of Operations listed specific functional capabilities associated with mission and mission support scenarios. The ICE SEVIS Modernization Operational Requirements document expanded on these functional capabilities and identified the operational and program-level requirements. These requirements were necessary to achieve the performance goals and mission of the Student and Exchange Visitor Program and the Department of State, the primary sponsors for the program. In particular, the SEVIS Modernization Operational Requirements document identified business capabilities and key performance parameters that measured system capabilities. The core capabilities are long-term initiatives intended to span multiple contracts and deliver the major components necessary for SEVIS modernization. The SEVIS Modernization Operational Requirements document stated that these capabilities must be present for the SEVIS modernization to be considered a success. These business capabilities represented the core SEVIS functions needed to close outstanding SEVIS vulnerabilities. According to the ICE SEVIS Modernization SELC Tailoring Plan, there were 79 sub-capabilities supporting the eight core capabilities. The sub-capabilities generally fulfilled one or more stakeholder needs and were delivered within a release or series of releases. The SEVIS Modernization Operational Requirements document confirmed that the program should prioritize and sequence the capabilities for delivery during the release planning and delivery processes. The program provided a road map for one development module. This road map listed areas for development in the order they were intended to be developed and identified the associated business capabilities. The business capabilities identified in the road map aligned with the sub- capabilities listed in the SEVIS Modernization Operational Requirements document. Examples of business capabilities in the road map that were also sub-capabilities identified in the operational requirements document included: create nonimmigrant record (including supporting forms), align nonimmigrant eligibility information with unique nonimmigrant, update nonimmigrant biographical information, and add/update dependent information. Appendix V: Leading Practices for Adopting Agile Development—Team Activities and Dynamics This appendix describes in more detail our evaluation of the three leading practices for team activities and dynamics when adopting Agile development. It does not present new findings; rather, the information is intended to assist the Department of Homeland Security (DHS) in implementing the recommendations described in this report. For teams to successfully transition from processes using traditional software development methods to Agile methods, leading practices for team activities and dynamics recommend that the composition of the team supports Agile methods by defining the role of a product owner work is prioritized to maximize value for the customer through creating user stories to define work prioritizing requirements in a backlog based on value estimating the relative complexity of user stories repeatable processes are in place by meeting daily to review progress and discuss impediments ensuring the quality of code being developed Within DHS, program management offices are responsible for planning and executing individual programs and implementing applicable Agile methodologies. According to Office of the Chief Technology Officer (OCTO) officials, DHS contracts for Agile services, including development, rather than performing development in-house. As a result, Agile teams may be predominantly contractors rather than federal employees. In addition, DHS Office of the Chief Information Officer (OCIO) is responsible for setting the policies and procedures to ensure that programs and, in turn, the teams that make up those programs, leverage Agile development best practices to meet the department’s goals and are within acquisition policy. DHS OCIO is also responsible for providing guidance for and reviewing the adoption and execution of Agile development. Team composition supports Agile methods Agency policy or guidance should require individual, self-organizing Agile teams for each segment or iteration and define the role and responsibilities of the product owner. Agile teams should be self-organizing, meaning they are empowered to collectively control how to accomplish their work and the resulting product. An Agile team’s authority should include lower-level decision making and team formation and highlight the importance of team stability. The team’s composition should be cross-functional and consist of members who possess all the skills needed to produce working software, including, but not limited to, contract specialists, developers, and testers. DHS provided guidance to Agile teams on self-governance. The Agile Development and Delivery for Information Technology instruction (Agile instruction) and the Agile Development and Delivery for Information Technology Instruction Manual (Agile instruction manual) both explain that collaborative, self-organizing, and cross-functional teams help achieve the flexibility needed for the iterative development that characterizes Agile development methods. The Agile instruction manual notes that most Agile methodologies assume the dedicated involvement of all stakeholder, developer, and integration staff throughout the project. DHS guidance also discusses team formation. The Agile instruction manual recommends that the project team include the roles of the program or project manager, a product owner, a development team of approximately five to nine members, testers, and an Agile coach, and any additional expertise as needed. According to DHS guidance, a program or project manager is responsible for establishing the project team. The program or project manager is supported in this by the component acquisition executive and other component management. At DHS, U.S. Immigration and Customs Enforcement’s (ICE) Student and Exchange Visitor Information System (SEVIS) program had self-organizing teams that defined their own processes for completing work. ICE Agile teams, including those supporting the SEVIS program, were expected to document their processes in a team process agreement, where a team had the authority to define its own operational strategy and make decisions about the product, including when to consider the product completed according to the program’s “definition of done.” According to the ICE Agile principles instruction, a program chooses a baseline set of practices that are documented in a team process agreement and are adjusted over time. ICE SEVIS teams were self-managing and included the roles necessary to deliver what they committed to in a sprint. ICE’s Agile playbook suggested minimum levels of experience, knowledge, and certifications necessary for key personnel to support Agile methodologies. For instance, the playbook suggests that Scrum masters be certified and have a minimum of one year of experience. To help ensure that contractors have the requisite skills necessary, ICE SEVIS officials stated that vendors are required to demonstrate their ability to develop a small software application before a contract is awarded to them. Define the role of a product owner A product owner should understand the business and strategic values of the agency and its alignment with the vision of the product team and support Agile methods. A product owner’s responsibilities include availability to the team, authority for making programmatic decisions, general responsibilities as a member of the team, and the need to possess subject matter expertise related to the business needs. A product owner is an authoritative user who manages the requirements prioritization, communicates operational concepts, and provides continual feedback to the team. DHS provided guidance on the role and responsibilities of a product owner. According to the Agile instruction manual, the product owner is responsible for representing stakeholders. To do so, the product owner should be available to the development team throughout the iteration to answer questions and clarify requirements on behalf of the stakeholders. The manual stated that the product owner is also responsible for ensuring that the product meets user needs and delivers value. This includes, for example, prioritizing user stories in the backlog and serving as an acceptance authority for work completed by the team. The department also provided elective training on the role of a product owner. For example, the U.S. Citizenship and Immigration Services Office of Information Technology offered an elective product owner training course. The USCIS product owner training covered concepts such as the importance of the product owner’s availability to the team and the product owner’s authority for making programmatic decisions. ICE identified a product owner for SEVIS to represent two user communities. The program identified one product owner from ICE’s Student and Exchange Visitor Program and a second product owner from a stakeholder organization within the Department of State. Both product owners were identified in the ICE SEVIS staffing plan. According to a team process agreement for one development module, a product owner is responsible for, among other things: Prioritizing and deciding which user stories will be implemented in each iteration. Making an acceptance decision for each user story based on the story’s acceptance criteria. Ensuring that the intended value of the functionality is delivered. According to program officials, product owners for the ICE SEVIS program prioritized user stories during planning sessions. The Student and Exchange Visitor Program Agile Overview slides stated that the team, including the product owner, attends sprint planning to review the prioritized product backlog and to ensure a common understanding of the product owner’s immediate priorities. Product owners also exercised authority to validate acceptance criteria and subsequently close user stories. The program’s “definition of done” stated that the product owner must test and indicate acceptance of each user story in order for a user story to be considered complete. In a written response, ICE SEVIS officials stated that ICE SEVIS product owners indicated a user story had met the acceptance criteria and could be closed by changing the user story’s status to “closed” using the team’s program management software tool. In addition, product owners were available to the development team to ensure timely input. According to a team process agreement for a development module, the product owner should work closely with the development team to communicate the details of requirements and answer questions about user stories. In an interview, the ICE SEVIS product owner representing the Student and Exchange Visitor Program stated that the role was a full-time position and did not have any competing responsibilities. To ensure availability, the ICE SEVIS product owner representing the Student and Exchange Visitor Program stated that there was a designated backup who had the same authority and responsibilities as the full-time product owner. Work is prioritized to maximize value for the customer Agency policy or guidance should call for Agile teams to create user stories to define the work; prioritize requirements in a backlog based on value, including tracking and monitoring the value of work accomplished; and estimate the relative complexity of user stories. Individual Agile teams within the respective programs and projects should implement these aspects of Agile development. Create user stories to define work A user story is to reflect a small segment of work that can be completed in a single iteration. The agency should have policy or guidance in place for writing user stories for Agile projects. The product owner should determine the value of a user story in consultation with the development team, including the acceptance criteria and defining what “done” means. User story value should then be re-evaluated based on requirements to ensure the greatest return on investment. DHS provided guidance that Agile programs and projects could leverage when writing a user story. The Agile instruction manual, Homeland Security Acquisition Institute Agile lessons, such as “Managing Program Execution” and the Requirements Engineering User’s Guide provided a basic format for how to craft a user story. These resources noted that a user story defines where a “role” wants some “goal/desire” accomplished to result in a “benefit”. The Requirements Engineering User’s Guide also discusses the role of acceptance criteria and a definition of done in user story development. The guide highlighted that acceptance criteria defines the boundaries of a user story and confirms when a story has been completed and is working as intended. It specifies that acceptance criteria should be included in an Agile program or project’s capabilities and constraints document, a DHS artifact unique to Agile development and highlighted in the systems engineering life cycle (SELC) tailoring example for Agile. This guide added that the definition of done identifies all of the activities/artifacts besides working code that must be completed for a feature or sub-epic to be ready for deployment or release including testing, documentation, training material development, certifications, etc. The Agile instruction manual places much of the responsibility for defining a user story under the purview of the product owner. The Agile instruction manual stated that the product owner is the individual tasked with providing requirements to the development team and is responsible for determining the features necessary for the product release. The manual also emphasized that the product owner is only responsible for clarifying the user story requirements that would meet his or her needs and not responsible for clarifying how user stories should be implemented to meet those needs. The ICE SEVIS program developed user stories based on business capabilities and other requirements as determined by the product owner and the business stakeholders. The SEVIS Modernization Operational Requirements Document describes eight business capabilities that represent core SEVIS functions. According to ICE SEVIS officials, these business capabilities are addressed through user stories, so there is traceability in the backlog from user stories to epics to business capabilities/operating requirements. The team’s process agreement for one development module—Information Sharing—assigned responsibility for writing user stories to the product owner. This agreement also noted that acceptance criteria would be required for most stories. User stories for the program were managed through a program management software tool. An output of the backlog from the program management software tool for one development module—Managing Nonimmigrant Information—contained 525 user stories. These user stories generally followed DHS and ICE guidance for capturing what a user needs and why. Most of these user stories also included acceptance criteria. The program also developed a “definition of done” for all user stories in the team process agreement. According to the definition, a user story was “done” when the following steps were addressed: All code to meet the story’s needs was written according to the system’s development standards. Unit tests were written and run successfully. All code was checked in and the build completed successfully. All database changes (if required) were complete and checked in (a functional test could be run). The software had been deployed to the system test environment and passed system tests. The product owner agreed that the implementation met the acceptance criteria written in the story as appropriate. All documentation required to support the story was completed (test cases, interface updates, etc.) Prioritize requirements in a backlog based on value Agile teams should pull work from a prioritized backlog and provide frequent deliveries of software with immediate value to the user. The team should determine the value of the user stories, prioritize work in a product backlog, and provide an ongoing assessment of value expected versus value delivered. The value of the work accomplished by Agile projects should be tracked and monitored. DHS guidance called for prioritizing user stories in a backlog. The department published an example of a SELC tailoring plan for Agile development that encouraged programs and projects to prioritize user stories in a backlog as part of each release. To ensure that programs or projects took these steps, the Technical Review Guide exit criteria for the release planning review asks if programs or projects will have a process in place for prioritizing user stories prior to the development of features for each release. Planning sessions were one such process that programs and projects could use to prioritize user stories in the backlog. The DHS Agile instruction manual stated that, during sprint planning, the product owner meets with the development team in order to identify user stories from the backlog that should be prioritized for the upcoming sprint and that prioritization decisions should be made based on value to the users. In addition, the product owner should ensure that prioritization decisions maximize mission values. The Requirements Engineering User’s Guide also states that requirements should be prioritized based on continuous stakeholder input so that programs can prioritize what users need the most. DHS guidance also discussed how to determine the value of individual user stories. While the Director of STM said that the product owner is responsible for interpreting the concept of value as it applies to a user story and the relative prioritization of the backlog, Agile Requirements and Road Mapping Guidance for DHS includes a discussion on how a program can sequence its road map for learning, risk, and economic value. In this section, DHS offers models to consider to assist in user story prioritization decisions and considerations for the product owner, such as seeking to balance between business value and cost. The Director added that there were venues, such as Agile “chat and chews,” where program staff could ask questions and receive informal guidance. DHS modified acquisition procedures to allow for an ongoing assessment of progress, and indirectly the value of work accomplished, via the release road map. DHS guidance stated that the release road map is submitted to the Acquisition Review Board prior to acquisition decision event 2B, as required by the Agile instruction. The Technical Review Guide exit criteria for the release planning review and the release readiness review asked if the development team was following the release road map and making adjustments that supported the successful completion requirements defined at acquisition decision event 2B. Thereafter, programs submitted a road map to the Acquisition Review Board during regular program reviews. In addition to tracking and monitoring the value of work accomplished against a release road map, regular Acquisition Review Board program reviews allowed for the assessment of value expected versus value delivered. The presentation template for Acquisition Review Board program reviews included a slide for programs to report their progress toward planned features. For each review, programs identified a percentage of each capability that they planned to complete by the next review. In addition, programs reported on the percentage of each capability that they had completed since the last review. The U.S. Coast Guard (USCG) SeaWatch Agile teams prioritized requirements in a backlog based on the team’s ability to complete them within a sprint. According to the acquisition manager for the Command, Control, Communications, and Computers, Intelligence, Surveillance and Reconnaissance (C4ISR) program, the SeaWatch product owner for new development determined priorities for new requirements with stakeholders. The product owner then defined those requirements as an epic or as a user story. The C4ISR acquisition manager stated that the user stories were prioritized in the backlog during sprint planning primarily based on whether the Agile team could complete the work in the upcoming sprint rather than on the value assigned by the stakeholders. According to SeaWatch officials, user stories that could not be completed during the current sprint were marked as a priority item for the next sprint. Although the SeaWatch program assessed value to the user for some epics, this did affect how the epic or its associated user stories were prioritized in the backlog. The C4ISR acquisition manager stated that SeaWatch assigned a value (e.g. extra large, large, or medium) to an epic based on the epic’s value to the user. However, the acquisition manager noted that user stories were not typically prioritized by the value of the associated epic. User stories were instead prioritized based on the Agile team’s ability to complete the work within the current sprint. The project reported on its accomplishments via a road map. In May 2018, SeaWatch reported on progress toward milestones in its road map during an annual briefing for the Non-Major Acquisition Oversight Council. The program reported that it had installed SeaWatch v3.0 on 65 out of 70 in-service cutters. Estimate the relative complexity of user stories The agency should have policy or guidance in place for relative estimation practices for Agile projects. Teams should use relative estimation for sizing the effort of work required to satisfy a user story by estimating its complexity based on work of similar size and complexity. Relative estimation enables teams to maintain a sustainable software development pace and predict work commitments. The team should size user stories relative to one another, assess the complexity of the work, refine user stories and estimates over time, and use prior estimates to inform future estimates. The product owner and team should continually revisit the estimates as they learn more about the business priorities and as user stories rise in the order of priority. DHS did not provide policy or guidance for relative estimation. Although the Agile instruction manual identified estimating user story size as an integral part to sprint planning, it did not describe the specific techniques or processes for estimating the relative complexity of user stories. Instead, the Agile instruction manual discussed how programs could successfully apply traditional earned value management and cost estimating principles to Agile projects. DHS guidance noted that programs had largely moved from measuring story points to feature points to help programs quantify incremental progress The U.S. Customs and Border Protection’s (CBP) Biometric Entry Exit (BEE) program defined practices and guidelines for how the program expected to estimate user stories. For example, the Traveler Verification Services process definition document identified a formula for calculating story point values on the basis that one story point would equate to approximately four working hours. Moreover, the process definition document noted that story points must be reconciled to better reflect the level of effort and task completion at the end of a given sprint. However, it was not evident that the BEE program had implemented its own guidance on the estimation of story points. Although the process definition document outlined procedures for estimating user stories, only two of 358 user stories in the Air Exit project backlog were estimated using story points. Repeatable processes are in place Agency policy or guidance should call for Agile teams to meet daily to review progress and discuss impediments, and to observe end-iteration demonstrations and end-iteration retrospectives. In addition, agency policy or guidance should call for Agile projects to employ continuous integration and confirm mechanisms are in place to ensure the quality of code being developed. This includes setting expectations for automated testing and code quality and tracking and monitoring against these expectations. Responsibility for these aspects of Agile development should lie with the individual Agile teams. Meet daily to review progress and discuss impediments The agency should have policy or guidance in place for holding the daily stand-up and teams should hold daily meetings in order to stay on track to meet the iteration goals for Agile projects and adjust as necessary. DHS guidance defined the general procedure for holding a daily stand-up. The Agile instruction manual stated that teams should conduct a daily stand-up meeting for all team members. It can be conducted in person or via another method of communication (particularly for remote employees) for a brief, informal meeting every work day. According to the manual, all team members should discuss the work each has accomplished since the last daily stand-up, the work to be accomplished by the next daily stand- up, and highlight any impediments that are preventing the team members from completing their work. Additionally, the manual suggests that it is necessary to conduct the daily stand-up with strict discipline, so that the meetings stick to their allotted brief time and are consistently productive. The Agile instruction manual also highlighted the importance of the daily stand-up meeting to an Agile process. It called this meeting an essential collaboration event during which all team members were expected to participate and discuss their work. The manual suggested that holding these meetings allowed the team to practice discipline that would assist them in their work and foster mentoring and partnering relationships within the team that were reinforced through the constant communication of meeting every single day. The manual added that this activity allowed the team to hold its members accountable and be made aware of issues that may be mitigated through collaboration. The Traveler Verification Services team supporting the BEE program Air Exit project at CBP held daily stand-up meetings. According to project officials and supporting project artifacts, a daily stand-up meeting was held each day at 10:00 a.m. Project officials noted that the daily stand- ups included the entire 40-person team. The agency should have policy or guidance in place for holding demonstrations or other interactions for acceptance of user stories in Agile projects. Teams should hold frequent demonstrations to showcase features that have been implemented and obtain feedback for acceptance of user stories in Agile projects. DHS guidance defined the general procedure for holding an end-iteration demonstration or review. In the SELC Tailoring Plan example for Agile development, DHS recommended a sprint review and demo as one type of technical review at the end of each iteration. The purpose of the review was to demonstrate the working software to end users and other stakeholders and to obtain feedback that could result in additional items being added to the backlog. It stated that this review should also ensure that the software design was documented for inclusion in the system design document, a proposed DHS Agile-specific artifact. The tailoring example noted that this review should formally end the iteration’s work with no further development or testing occurring on any stories. The Agile instruction manual added that this demonstration should confirm the value of the incremental piece of software produced. DHS guidance also encouraged the use of demonstrations. The Agile instruction manual states that a demonstration or review could be used to reach a consensus on whether the work associated with a user story met expectations or not. The manual also recommended that program and project managers ensure that the functional software developed during each iteration was demonstrated to the stakeholder at an iteration review meeting. The ICE SEVIS program held end-iteration demonstrations. The ICE SEVIS Modernization Systems Engineering Lifecycle Tailoring Plan stated that sprint demonstrations were tailored into the program to replace other review activities, such as the preliminary design, critical design, and integration readiness review. The Test and Evaluation Master Plan for SEVIS Modernization stated that standard sprint testing results were to be reported at sprint reviews. According to program artifacts, the sprint demonstration was to be conducted at the completion of each sprint, every other Wednesday from 11:00 a.m. to 12:00 p.m. The agency should have policy or guidance in place for holding a retrospective to adapt and continuously improve on Agile projects. Teams should hold a retrospective at the end of each iteration to identify areas for improvement to adapt and continuously improve Agile practices. DHS guidance defined the general procedure for holding a retrospective. The program or project manager and team reviewed progress after each iteration and release. This included the use of a retrospective to discuss what went well, what didn’t go well, and to identify actions to correct problems. Guidance noted that the team should immediately incorporate feedback from the retrospective into future iterations. The DHS Agile instruction manual highlighted the importance of the retrospective. The manual stated that the end-iteration retrospective is a key part of ensuring that teams following Agile methodologies are able to identify problems and adapt to continuously improve for future sprints. Additionally, the manual stated that end-iteration retrospectives are useful in satisfying governance needs. For example, the Agile instruction manual stated that programs could tailor standard-format SELC artifacts (as codified in the SELC Tailoring Plan) to instead rely on assessment and performance data addressed in end iteration retrospectives. The Traveler Verification Services team supporting the BEE program’s Air Exit project held end-iteration retrospectives. According to the process definition for this team, a retrospective was to be held between the end- iteration review and the subsequent planning session for the upcoming sprint. The process definition defined the goal of the retrospective as obtaining an honest review of the process with a consensus on how to adapt it. In an interview, project officials noted that the team documented the results of retrospectives on a release-by-release basis in a project management software tool. The agency should have policy or guidance that defines and emphasizes the use of automated testing and continuous integration. This guidance should be supplemented by defining expectations for automated testing and tracking and monitoring against these expectations. Agile teams should adopt practices for continuous integration and automated testing to ensure that software handoffs are repeatable and dependable. Automated testing should be tracked and monitored based on established expectations. The DHS Agile instruction manual defined continuous integration as the practice where delivery teams frequently integrate their code into a shared master copy. It noted that these integrations are verified by an automated build process, which performs testing to detect any integration errors quickly and automatically. The manual stated that continuous integration in Agile projects should be planned and recorded on a release-by-release basis. The Agile instruction manual also emphasized the importance of continuous integration and automated testing. With regard to automated testing, the manual set an expectation for program or project managers and stakeholders to consider both automated testing tools and infrastructure support for the Agile software build and test processes as part of general project planning efforts. Moreover, the manual identified continuous integration, automated acceptance testing, and automated unit testing as key practices program or project managers can use for continuously monitoring and reporting project health. These practices could also help to identify opportunities for improving project team performance. DHS officials acknowledged that current DHS programs implemented testing and evaluation inconsistently and that the department’s existing guidance and policies did not effectively support modern best practices in automated testing and continuous integration. To address these gaps, DHS had an Agile action plan that set an expectation for updating DHS acquisition guidance, policy, and practices for testing and evaluation to enable modern best practices in automated testing and continuous integration. In lieu of more explicit guidance, DHS incorporated training as part of a curriculum geared toward test and evaluation managers that discussed both continuous integration and automated testing. According to the Deputy Director of Policy and Workforce Development in the Test and Evaluation Division of the Science and Technology Directorate, an alternative course containing content addressing Agile and continuous integration and automated testing was recently merged with a required test and evaluation course, creating a new course. According to the Deputy Director, the new course was piloted during fiscal year 2019 and will be standard in fiscal year 2020 as the required course for level II test and evaluation certification. In order to track and monitor automated testing, the department incorporated several measures into the Agile core metrics. Programs executing Agile were expected to report on the following testing-related metrics after each iteration: Percentage of unit test coverage, Percentage of automated tests, and Percentage of regression testing coverage. DHS had not established expectations for these Agile core metrics. The Agile core metrics included a target. For example, the department suggested a program strive for seventy percent of tests to be automated. However, the instructions accompanying the Agile core metrics stated that all targets were notional and not expected to be reached. According to the Director of STM, the initial core metrics were intended to assess the level of DHS team achievement without imposing artificial industry- based target measures for each. The Director stated that, on receiving the metrics for a period of time, the department would then adjust the core metrics and begin to include target measures based on the results achieved. According to the Director, this effort was underway and an updated set of core metrics would be distributed in early fiscal year 2020. Moreover, the department was not tracking and monitoring automated testing as intended. The CBP BEE program Air Exit project stood up a technical environment that allowed for continuous integration. This technical environment was outlined within the process definition of the Traveler Verification Services team that was developing software. The Traveler Verification Services process definition identified three operating environments: the development, test, and production environments. All development activities during the sprint were conducted within the development environment. Similarly, all testing activities in preparation for the release were conducted in the test environment. The final approved software would then be deployed to the production environment. CBP officials noted that the BEE program primarily used Jenkins to integrate code for both continuous builds and deployment. The Air Exit systems design document also mentioned the role of Jenkins in continuous integration and continuous deployment for the project. The Traveler Verification Services team incorporated JaCoCo and FindBugs automated tests as part of the continuous delivery process and they were run automatically when the code was checked in. Moreover, the project’s system design document noted that the Traveler Verification Services team integrated JaCoCo with the Eclipse Integrated Development Environment as a code coverage inspection tool for unit testing. Officials also noted that Selenium was used for automating the testing within the technical environment. Ensure the quality of the code being developed The agency should have policy or guidance for an Agile project on ensuring the quality of code being developed. This guidance should be supplemented by defining expectations for code quality and tracking and monitoring against these expectations. Agile teams should adopt practices for code quality, such as having a test-driven development, pair programming, and manual code reviews to supplement automated testing. Agile teams should incorporate refactoring into code quality practices and understand the importance of setting aside time for refactoring. DHS guidance recognizes the importance of ensuring code quality as part of the development and testing process. The SELC Guidebook set an expectation that code review and testing should be part of the software development environment. The guide recommended setting up servers where developers could test code and check whether the developed application runs successfully with that code. The guide suggested another level of tests on application reliability to help ensure that the application did not fail on the production server. The guide stated that the program manager should ensure that the team takes corrective action for any hardware and software deficiencies. In order to find deficiencies early, DHS guidance identified coding and testing practices that could help development teams. The Agile instruction manual cited pair programming as one practice where two programmers work simultaneously on a single task: one programmer observes and reviews each line of code as it is written. DHS guidance also identified test-driven development as a practice that could motivate developers to write effective code. The Supplemental Guidance for Test and Evaluation stated that this approach consists of writing test cases that define a desired improvement, then writing the code to meet the desired functionality, ensuring that the test passes, and refactoring the code as necessary. Refactoring, or re-coding, without changing the way the application functions, is an Agile practice that DHS guidance recommends for correcting deficiencies in the code. The Agile instruction manual stated that refactoring aims to improve code readability and reduce the complexity of previously delivered increments of software. It noted that refactoring is important because development teams are focused on adding the desired functionality with each release and may proceed with making improvements to the code. Refactoring was cited as one way to address this accumulation of needed improvements to the code, which are known as technical debt. The Agile instruction manual further emphasizes the importance of setting aside time for refactoring to address risks associated with technical debt. The manual states that refactoring a previously developed increment of software to improve code quality may force a change in the release schedule. However, if the team does not make these revisions in a timely manner, the effort required to correct them later tends to increase. The manual states that this increasing technical debt is a risk factor to be addressed as soon as feasible. If the technical debt is allowed to accumulate unchecked, or if the project team loses track of the scope of its technical debt, the project could suffer from schedule and performance problems. In order to track and monitor the quality of code being developed, the department incorporated several code quality and testing measures into the Agile core metrics. Among others, programs executing Agile were expected to report on the following quality-related metrics after each iteration: Number of critical or major defects fixed. Number of critical or major defects in the backlog. Number of technical debt issues completed. Number of technical debt issues in the backlog. However, the department was not tracking and monitoring code quality as intended. These measures could provide programs and the department with an understanding of the development team’s ability to address defects and technical debt. In addition to these metrics, programs are also expected to report quarterly on the number of outages requiring a rollback or patch after production deployment. The ICE SEVIS program used manual testing to ensure code quality. The definition of done for the program stated that new code should be peer reviewed to identify risk to the existing code, assess compliance with coding best practices, and evaluate refactoring. According to ICE SEVIS officials, an independent specialist provides internal code reviews and offers feedback on areas for improvement. The ICE SEVIS program also employed automated testing to ensure code quality. The definition of done required that unit tests cover a minimum of 85 percent of code. Program officials stated that vulnerabilities and bugs identified through this process were added to the backlog and classified as technical debt. The program refactored code to address technical debt, but did not set aside time for refactoring each sprint. According to ICE SEVIS officials, the development team refactored code as necessary to improve overall quality but did not set aside time for refactoring unless they were addressing a consistent issue. ICE SEVIS officials stated that the development team could propose refactoring code during sprint planning if there was a specific technical debt they had identified. However, according to the Scrum master for the program, addressing technical debt was additional work for the team to take on beyond the user stories they planned to complete and this additional work incentivized the development team to prevent the accumulation of technical debt. Although DHS allowed Agile programs to tailor the core metrics, ICE SEVIS submitted some of the code quality-related Agile metrics to the department. The program included Agile metrics in June 2018 presentation slides for the Acquisition Review Board. For this initial reporting period, the program reported no critical or major defects in the backlog and no technical debt issues in the backlog. It also provided a screenshot of the Agile core metrics reported to DHS via the Investment, Evaluation, Submission, & Tracking system in February 2019. This reporting period covered two iterations. The program reported that it fixed four critical or major defects during the first iteration and did not have any critical or major defects in the backlog for either iteration. The program also reported that it completed eight technical debt issues in the first iteration, out of 14 technical debt issues in the backlog. The program did not report on the number of outages after deployment as part of the Acquisition Review Board program review or as part of the metrics submitted via the Investment Evaluation, Submission, and Tracking system. Appendix VI: Comments from the Department of Homeland Security Appendix VII: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, the following staff made key contributions to this report: Michael Holland (assistant director), Mathew Bader (analyst in charge), Lamis Alabed, Jennifer Beddor, Christina Bixby, Hannah Brookhart, Chris Businsky, Alan Daigle, Aryn Ehlow, Nancy Glover, Gina Hoover, Anna Irvine, Hoyt Lacy, Jennifer Leotta, Alexis Olson, Zsaroq Powe, Martin Skorczynski, Natalie Smith, and Daniel Spence.

Why GAO Did This Study Many of DHS's major IT acquisition programs have taken longer than expected to develop or failed to deliver the desired value. In April 2016, to help improve the department's IT acquisition and management, DHS identified Agile software development as the preferred approach for all of its IT programs and projects. GAO was asked to examine DHS's adoption of Agile software development. The objective of this review was to assess the extent to which DHS has addressed selected leading practices for its transition to the use of Agile software development. GAO identified leading practices for planning, implementing, and measuring organizational change that apply to DHS's transition to Agile through its review of guidance published by the Project Management Institute and GAO. GAO also reviewed work it performed to develop leading practices for Agile software development adoption. GAO analyzed DHS documentation, such as policies, guidance, plans, and working group artifacts and assessed them against the selected leading practices. GAO also reviewed the implementation of selected practices within individual IT projects. Finally, GAO interviewed DHS officials to discuss any practices that were not fully implemented. What GAO Found The Department of Homeland Security (DHS) has taken steps to implement selected leading practices in its transition from waterfall, an approach that historically delivered useable software years after program initiation, to Agile software development, which is focused on incremental and rapid delivery of working software in small segments. As shown below, this quick, iterative approach is to deliver results faster and collect user feedback continuously. DHS has fully addressed one of three leading practice areas for organization change management and partially addressed the other two. Collectively, these practices advise an organization to plan for, implement, and measure the impact when undertaking a significant change. The department has fully defined plans for transitioning to Agile development. DHS has partially addressed implementation—the department completed 134 activities but deferred roughly 34 percent of planned activities to a later date. These deferred activities are in progress or have not been started. With respect to the third practice, DHS clarified expected outcomes for the transition, such as reduced risk of large, expensive IT failures. However, these outcomes are not tied to target measures. Without these, DHS will not know if the transition is achieving its desired results. DHS has also addressed four of the nine leading practices for adopting Agile software development. For example, the department has modified its acquisition policies to support Agile development methods. However, it needs to take additional steps to, among other things, ensure all staff are appropriately trained and establish expectations for tracking software code quality. By fully addressing leading practices, DHS can reduce the risk of continued problems in developing and acquiring current, as well as, future IT systems. What GAO Recommends GAO is making 10 recommendations to DHS to implement selected leading practices for its transition to Agile software development. DHS agreed with GAO's recommendations and described actions taken and planned to address them.

gao_GAO-20-354

gao_GAO-20-354_0

Background USPS is required to provide postal services “to bind the nation together through the personal, educational, literary, and business correspondence of the people prompt, reliable, and efficient services to patrons in all areas and postal services to all communities.” To help fulfill this mission, USPS has developed a network of facilities that provides access to retail services and supports postal collection, processing, transportation, and “last mile” delivery of mail—functions that are generally co-located. The retail portion of these facilities includes public areas where USPS provides customers with retail services. The public areas can include full service counters where employees assist customers, self-service areas, and post office boxes. The non-public portions include features such as workrooms, where mail processing occurs, and employee support areas, such as lunch tables and lockers. Although USPS data show that customer visits and transactions have declined over the past 5 years, the size of USPS’s retail network has remained largely unchanged during that period (see fig. 1). We have previously reported that USPS has made some efforts to reduce the number of its retail facilities to align with the significant decline in mail volume and address rising costs. Several factors, however, have limited USPS’s ability to make further reductions. For example, in 2011 USPS moved to evaluate the potential closure of almost 3,700 retail facilities but abandoned their effort due to stakeholder concerns. Instead, USPS explored options to adjust its retail service without closing offices. Legal restrictions also limit USPS’s ability to close retail facilities. For example, USPS cannot close a small post office solely for operating at a deficit. If USPS wishes to close a retail facility, among other steps, it must take into consideration not only the economic savings but also the effects on the communities served, its employees, and the services provided, and it must provide customers with at least 60 days’ notice before the proposed closure date. Federal statute defines the types of services that USPS may and may not provide. As previously noted, PAEA placed limitations on the nonpostal products and services USPS could provide. In particular, it allowed USPS to continue to provide nonpostal products and services that were both offered as of January 1, 2006, and were permitted by PRC to continue. While PAEA generally prohibits USPS from initiating new nonpostal services, USPS uses a separate statutory authority to provide such services to federal executive agencies. If a nonpostal service is to be provided to a federal executive agency, generally, USPS and the parties must specify the terms and conditions of their collaboration, including the activities to be performed by USPS and the terms of reimbursement, if applicable. USPS is currently not authorized to provide nonpostal services to state or local entities. USPS may also lease excess space at retail facilities—including parking, office space, and roof areas—to other entities at both the facilities that USPS owns and those it leases from other entities. USPS officials said some leases give them the right to sublease space at leased facilities, but the officials did not know how many facilities where USPS had such rights. USPS collaborates with the GSA, which is the nation’s largest public real-estate organization, to lease space to other federal government entities. USPS works with a real estate firm to lease space to private or other entities, such as to state and local government entities. USPS’s Retail Facilities Are Not a Major Contributor to Its Costs and Revenues but Provide Benefits Retail Facilities Account for a Relatively Small Percentage of Costs and Revenues and Are Profitable Overall According to USPS’s analysis, retail facilities accounted for a relatively small portion—about $5.17 billion, or 7.1 percent—of USPS’s modified operating costs in fiscal year 2018. (See fig. 2.) Personnel costs accounted for the majority of retail facility operating costs ($4.87 billion), including nearly $4.4 billion in employee compensation. Non-personnel costs amounted to about $0.30 billion, which included rent ($0.09 billion), utilities ($0.06 billion), and depreciation and amortization ($0.15 billion). In terms of revenue, according to USPS’s analysis, retail facilities accounted for a relatively small portion—about 15 percent ($10.5 billion)—of USPS’s total revenue ($70.6 billion) in fiscal year 2018. As shown in figure 3, postage meters and validation and walk-in stamp sales generated most of that revenue. In fiscal year 2018, the majority of retail facilities were profitable (see fig. 4), with some considerations. USPS’s analysis showed that the total retail revenue generated at retail facilities was more than double retail facility costs in fiscal year 2018. However, over a third of retail facilities did not generate enough revenue to cover USPS’s retail costs. In particular, over half of the rural facilities were unprofitable, while the overwhelming majority of suburban and urban facilities were profitable. Overall, we found that of the unprofitable facilities, 89 percent were located in rural areas. Among the 10 most unprofitable facilities, though, 5 were in urban areas and 3 were in suburban areas. USPS officials told us some of these unprofitable urban and suburban retail facilities were in areas where rent, utilities, and maintenance were very costly, and the revenue generated was not enough at those facilities to make them profitable. Finally, while the number of customer visits to retail facilities has declined by 25 percent since fiscal year 2009, customers are still using them. In fiscal year 2018, among the retail facilities for which USPS had data, the average number of customer visits was 46,624 annually, or an average of 154 visits per day. The most unprofitable facilities averaged 31,731 customer visits in fiscal year 2018, or about 105 customer visits per day. Stakeholders Cited Several Non-Revenue Benefits of Retail Facilities According to studies we reviewed and USPS officials and two consumer groups we interviewed, USPS’s retail facility network produces economic, social, environmental, and civic benefits (see table 1). Some postmasters who responded to our survey told us that their retail facilities generated economic and social benefits. For example, 90 percent (134 of 149) of postmasters managed a retail facility within walking distance of other businesses or community buildings. Of those, about 31 percent (41 of 134) of postmasters indicated that the retail facility increased patronage of nearby businesses and community buildings to a great or very great extent. Almost half (71 of 149) of the postmasters stated that their retail facility served as a place for residents to interact in person to a great or very great extent. The economic and social benefits may benefit rural communities more than urban and suburban areas, according to our survey of postmasters, as well as the reports we reviewed and stakeholder interviews. One survey respondent stated: “Offices in rural communities are extremely important to the area in which they serve. Postmasters are often town leaders and hold various positions on councils, boards, and non-profit organizations. The USPS is usually the only government office in the community, and it is recognized by many as their only connection to the outside world.” According to USPS OIG and representatives from the two consumer groups we interviewed, retail facilities serve as a gathering place and help build social identity and connectivity, especially in rural areas. According to USPS, small business owners depend on access to USPS retail facilities across the country and in rural areas since USPS facilities are often the only retail shipping provider. In addition, according to USPS, retail facilities play important roles in connecting rural communities. USPS’s Nonpostal Retail Activities Generate Some Revenue and Offer Other Benefits USPS’s Nonpostal Products and Services Account for a Small Percentage of Retail Revenue and Are Mostly Profitable USPS has offered a variety of nonpostal products and services since 2008 (see table 2). As previously described, PAEA permitted USPS to continue offering certain nonpostal products and services that were approved by PRC, and did not alter USPS’s statutory authority to provide nonpostal services to federal executive entities. The nonpostal products and services we identified were either permitted to continue by PRC pursuant to PAEA or are services USPS provided through partnerships with other federal government entities, such as those made through interagency agreements. USPS chooses where to offer its nonpostal products and services based on several factors, and as noted in table 2 above, USPS does not offer each nonpostal product and service at all retail facilities. USPS officials told us they determine where to offer nonpostal products and services based on several factors, such as customer demand and analysis of the potential to generate revenue. In some cases, USPS consults with other entities to make these determinations. For example, USPS officials told us that the company that supplies the greeting cards conducts market research to determine locations where there is demand. For passport services, USPS and U.S. Department of State officials determine locations based on demand and whether retail facilities have adequate staff to perform related functions, among other factors. USPS collects a fee for some, but not all, of the identified nonpostal products and services offered at retail facilities. According to data from USPS’s revenue and cost analysis, nonpostal products and services generated a small amount of the total revenue collected at retail facilities—about $431 million in fiscal year 2018, which accounted for 4.1 percent of total retail facility revenue, and 0.6 percent of USPS’s total revenue. As figure 5 shows, passport applications and photo services, as well as money orders, accounted for the greatest percentage of revenue. Nonpostal products and services may also generate revenue from additional transactions that are made by customers during their visits to obtain these offerings. For example, in 2016, USPS OIG estimated that USPS generated almost $6.6 million in fiscal year 2015 from individuals purchasing money orders to pay for passport-related services. In addition, USPS OIG reported that passport services increased foot traffic at retail facilities. A representative from a postal employee union also told us that nonpostal products and services offered at retail facilities can drive increased foot traffic in post office lobbies. USPS officials said that USPS incurs various costs for nonpostal products and services related to (1) the time it takes mail clerks to perform transactions; (2) equipment and materials, such as passport photo equipment for passport photo services; and (3) any needed physical changes to the facility. USPS is required to analyze whether revenues cover costs for some of the identified nonpostal products and services, but not all. USPS officials said they are not required to report on whether revenues cover costs for services provided through federal interagency agreements, such as passport application processing. In addition, they told us they do not track the costs for offerings that are considered to be “non-commercial, non-revenue generating services,” such as community bulletin boards. USPS officials also said some of the nonpostal products and services we identified do not incur any costs; for example, greeting and gift card displays are provided at the vendor’s expense. While USPS does not track costs for all offerings and make all costs publicly available, USPS reported that most of the nonpostal products and services we identified (as shown in table 2), and for those that USPS tracks costs, generated more revenue than costs in fiscal year 2018. Specifically, USPS reported that money orders earned almost $12 million and philatelic sales earned about $1.1 million in revenue above their costs in fiscal year 2018. USPS reported the only nonpostal product or service that did not cover its costs in 2018 was in-bound international money transfers. USPS Initiated Pilots to Provide Services for Other Federal Entities at Retail Facilities Based on our discussions with USPS officials, we identified three pilots USPS has conducted since 2008 to provide nonpostal services on behalf of other federal government entities at retail facilities; two of these pilots involved USPS mail clerks performing in-person identity proofing and biometric capture and one of the pilots involved sharing retail facilities’ lobby space. USPS and the U.S. Census Bureau (Census Bureau) conducted a pilot to evaluate the feasibility of having USPS assist the Census Bureau with nationwide hiring for the 2020 Census. For this pilot, which lasted from March 2015 to July 2015, USPS mail clerks at 12 retail facilities used Census Bureau equipment to conduct in-person identify proofing and other administrative processes to help hire temporary Census Bureau employees. USPS and the Federal Bureau of Investigation (FBI) began a pilot in September 2018 in which USPS mail clerks began scanning and sending the fingerprints of individuals participating in the FBI’s Identity History Summary Checks—a program that enables individuals to request their arrest and conviction records (see fig. 6). USPS and the Census Bureau conducted a pilot from April 2018 to July 2018 as part of the Census Bureau’s 2018 testing activities to determine whether interactive kiosks could be used at retail facilities to allow customers to fill out their Census questionnaire. USPS installed kiosks at 30 retail facilities in Providence County, Rhode Island, that offered Internet access limited to the Census Bureau’s online questionnaire. According to officials from the Census Bureau and FBI, the agencies benefited from the three pilots, to varying degrees, and USPS generated revenue from the two in-person proofing and biometric capture pilots. USPS officials told us that they received $125,000 from the 2015 in- person proofing pilot with the Census Bureau, and Census Bureau officials said this pilot provided their staff with convenient locations to meet with prospective applicants. Regarding the 2018 kiosk pilot, Census Bureau officials said this resulted in 111 completed questionnaires. However, after the pilots ended, the agencies did not enter into subsequent partnerships. Census Bureau officials told us they did not wish to extend the in-person proofing pilot due to limited funding, and they did not wish to extend the Census kiosk pilot because the number of completed tests did not justify the cost and effort. For the FBI pilot, USPS had generated almost $425,000 in revenue, as of December 2019, from the fees paid by participating customers since September 2018, according to USPS documentation. FBI officials told us this pilot has improved their customer experience and enabled them to reduce their response time for providing information to customers. In March 2019, USPS and the FBI expanded the pilot from two to 28 retail facilities, and this pilot was still ongoing at the time we published our report. USPS officials told us they also created the Digital Business Services Team in June 2019 in part to pursue additional revenue-generating partnerships with federal executive agencies. The officials said a major focus of the team was to expand USPS’s in-person identity proofing and fingerprinting services, and they estimated—with certain assumptions regarding acquiring partnerships—fingerprinting services could generate about $87 million in annual revenue after a 5-year rollout. USPS is currently discussing potential new partnerships for in-person proofing pilots with federal government entities. Stakeholders Reported Non-Revenue Benefits and Few Challenges Related to Current Nonpostal Products and Services Offered USPS officials, federal government entities, and stakeholders we interviewed and postmasters we surveyed told us that the identified nonpostal products and services (as shown in table 2) currently offered at retail facilities provided the following non-revenue benefits. Enhanced consumer benefits. Access to certain nonpostal products and services at retail facilities enhanced consumers’ convenience, according to USPS officials, and representatives from consumer groups and a postal employee union. For example, in September 2019, USPS analysis found that 32 percent of the FBI pilot customers selected USPS’s fingerprinting services over others who offer similar services due to USPS’s location and rated their satisfaction with USPS’s service highly. A representative from the postal employee union said that many customers are happy that they can purchase a greeting card and gift card when visiting a retail facility. In addition, a representative from one of the consumer groups we interviewed said that some nonpostal products and services, such as international money transfers and money orders, may be otherwise unavailable to certain populations, such as those who do not have access to a bank. This representative also told us that some low-income consumers only have internet access through their phone, which makes it difficult to fill out forms online; these consumers therefore could benefit from having certain forms, such as voter registration and selective service, available at a retail facility. Enhanced government benefits. Officials from five of the six federal government entities that had partnerships with USPS said their partnerships supported their ability to fulfill their missions, such as by efficiently using resources and increasing customer convenience. For example, officials from all six of these federal government entities said USPS’s extensive network of retail facilities helped them reach customers or users. Also, officials from three of these federal entities told us that USPS’s services cut the processing time for certain applications or services. Enhanced community benefits. Representatives from the two consumer groups told us that community services offered at retail facilities—such as food drives, school tours, and community bulletin boards—may help sustain communities and increase social connectedness. Postmasters we surveyed also reported ways their communities benefited from nonpostal services provided at their retail facilities. For example, one postmaster we surveyed reported that his or her retail facility collected eyeglasses for a local community organization. Another postmaster we surveyed reported that during the holiday season, his or her retail facility offers decorative rubber stamps, which have become a community tradition. According to the postmasters we surveyed, some of the nonpostal products provided significant nonrevenue value, although the degree to which these provided value depended on whether the retail facility was located in a rural, suburban, or urban area. We asked postmasters to identify whether certain nonpostal products and services were offered at the selected facility, and if so, how much value the product or service provided to the community. Overall, passport services were the most highly valued nonpostal product or service. About 95 percent (36 of 38) of postmasters at retail facilities that offered passport services said passports provided great or very great value to their communities. Money orders were the next most highly valued nonpostal product or service. These were offered at more of the retail facilities selected for our survey than passport services. For the retail facilities that offered money orders, about 78 percent (115 of 147) of postmasters said this product provided great or very great value to their communities. Burial flags were the third highly valued nonpostal product or service, for some types of locations. About 66 percent (21 of 32) of postmasters managing rural retail facilities and about 70 percent (35 of 50) of postmasters managing suburban retail facilities said burial flags provided great or very great value compared to about 43 percent (9 of 21) of postmasters overseeing urban retail facilities. Among retail facilities that offered international money transfers (SureMoney), selective service forms, philatelic products, and gift cards, around one-third or more of postmasters reported these as providing some value or little to no value in their communities. Representatives of one postal employee union, postmasters we surveyed, and officials from the six federal government entities that had partnered with USPS reported minimal challenges related to providing the identified nonpostal products and services at retail facilities. For example, the representative from the postal employee union told us that the only challenge for postal workers was when locations did not have adequate staff to handle passport services. Very few of the postmasters selected for our survey identified challenges related to offering the thirteen nonpostal products and services we asked about. Officials from only two of the six federal government entities mentioned challenges, and none of them were significant in nature. For example, officials from the Department of Veterans Affairs mentioned that the only challenge was ensuring adequate supplies of burial flags at retail facilities. Although the stakeholders we interviewed and postmasters we surveyed cited few challenges associated with nonpostal products and services, USPS OIG has reported that USPS could take actions to further increase the use of some of these offerings. In 2015, USPS OIG suggested that USPS conduct better-targeted marketing for its money orders or consider pricing changes to the fees charged for money transfers. In 2016, USPS OIG identified several areas in which USPS could improve customer experience for passport services, such as improving the clarity of information provided to customers and improving the accuracy of offerings on USPS’s website. USPS generally agreed with the findings but reported it had already implemented or had begun to implement changes to improve customer service issues raised in report. Some Revenue Results from Leasing Excess Space, but Opportunities Are Limited USPS currently leases some of its owned excess space—including space at its retail facilities, such as parking, office space, and roof areas—to other entities, generating additional revenue and other benefits. According to USPS as of January 2020, USPS was leasing space in 232 facilities (about 3 percent) of its 8,362 owned facilities to federal and local government and private entities. USPS generated about $29 million from its leases in fiscal year 2018. USPS officials told us they are currently researching the feasibility and benefits of leasing space to entities to place automated teller machines in retail facility lobbies and parking lots as a way to generate revenue. Stakeholders we interviewed and USPS OIG have said leasing space may also result in non-revenue benefits for USPS and consumers. For example, a postmaster who managed a retail facility said that leasing office space to two local government entities likely increased foot traffic in the facility and increased community access to government services provided by the tenants. USPS, however, has little additional vacant rentable space. As of September 2018, USPS reported that it had vacant rentable space available in 307 (about 4 percent) of its 8,362 owned facilities. In 2018, USPS OIG reported that USPS faced unique challenges in leasing such excess space, including poor condition and limited size, lack of handicap accessibility, limited parking, lack of accessibility without interfering with USPS operations, and lack of a separate restroom. According to GSA officials, some available space is small—USPS reported that about 9 percent of available space is less than 500 square feet—and may require significant investment from GSA or the potential tenant agency to be suitable for occupancy (see fig. 7). GSA officials said it has been difficult to find federal government entities willing to lease retail facility space from USPS. For example, officials from the Census Bureau told us that they leased some space at two USPS facilities to support 2010 Census activities, such as to support hiring personnel and a location for training. However, they were not able to lease as much as they would have liked because there was very little available space that met Census requirements, and the space that was available would have required costly modifications prior to use. In addition, USPS officials told us there can be costs related to leasing space, such as USPS’s making needed renovations. USPS does not track the extent to which space at retail facilities, such as lobbies and parking lots, is shared with other entities without any payment. However, USPS officials told us that while community groups have asked to use retail facility space and parking lots, these requests do not happen frequently. Only two of the 149 postmasters we surveyed (about 1 percent) indicated that the selected retail facility they oversaw shared space with other entities at no charge. Additional Nonpostal Efforts at Retail Facilities Could Offer Some Benefits but May Generate Little Revenue and Have Other Limitations Additional Nonpostal Products and Services Could Provide Non- Revenue Benefits Studies we reviewed and postal experts and stakeholders we interviewed have suggested that USPS may be well positioned to offer additional nonpostal products and services due to its trusted brand, vast retail facility network, and experience with other nonpostal efforts. Examples of such additional offerings are set out in the following table (see table 3). USPS officials, stakeholders, and studies we reviewed indicated these identified additional nonpostal activities—if USPS were authorized to offer them—could offer a variety of non-revenue benefits to consumers, government entities, and communities. Examples of some of the suggested types of non-revenue benefits are shown in table 4. However, most of these additional nonpostal products and services would not increase revenues or greatly benefit their communities, according to the postmasters we surveyed. Specifically, a majority of postmasters did not think any of the nine additional nonpostal products and services we asked about would increase revenues or benefit the community to a great or very great extent. However, postmasters indicated that some potential services were more promising than others. In particular, about 40 percent (59 of 149) of postmasters indicated that notary services would increase revenues or benefit the community to a great or very great extent, while 36 percent (53 of 149) of postmasters indicated that driver’s license and other state license services would increase revenues or benefit the community to a great or very great extent. Few postmasters in our survey indicated that a benefit of offering any of the nine additional nonpostal products and services would be that they would be providing a product or service that is not offered elsewhere in the community. Most Additional Nonpostal Products and Services Would Likely Have Low Net Revenue Potential USPS officials, officials from other federal government entities, and stakeholders told us that most of the nonpostal products and services identified above would likely have limited revenue potential. They, as well as studies we reviewed, indicated a variety of reasons why USPS might not generate significant net revenue from the additional nonpostal products and services we identified. Low potential for a significant market share. USPS officials and stakeholders told us USPS could face challenges gaining enough of a market share for some of the additional nonpostal products and services to make a profit. For example, USPS officials said if they offered notary services, they would likely gain only a small share of the market because other retailers, such as banks, already offer these services for free. Also, representatives from four financial associations said they believed consumer demand for financial products and services was already being met or would best be met by existing financial entities, and that many consumers may not likely obtain these services from USPS. In addition, representatives from two financial associations said serving underbanked populations would likely result in limited profits because these tend to be riskier customers who may default more often, and the services they use result in slim profits for current providers. High operational costs. USPS OIG has reported that USPS incurs low customer foot traffic and high labor costs compared to other retail facilities. According to USPS officials, these factors make it difficult to compete on a cost-per-transaction basis for nonpostal products and services and make leasing space in owned facilities and subleasing space at USPS’s leased retail facilities attractive, because leasing and subleasing would not incur personnel costs. Additionally, USPS officials told us that offering some of the additional nonpostal products and services could require significant investment costs—such as major technology investments and additional training for mail clerks—further reducing USPS’s ability to make a profit. For example, USPS OIG estimated in 2015 that USPS could generate $1.1 billion annually after a 5-year ramp up from expanding the financial products it already offered. USPS officials, however, said that expanding such offerings at retail facilities would likely require extensive investments in physical and information technology security and incur ongoing costs. Accounting for these sorts of costs would mean USPS would likely generate about $100 million to $200 million in net revenue as opposed to $1.1 billion. In addition, representatives from the American Association of Motor Vehicle Administrators and the Maryland Motor Vehicle Association told us that if USPS were to offer state driver’s license services, it would need to invest in equipment and training. Specifically, USPS would need to purchase secure computer systems that require multiple electronic interfaces and train mail clerks to handle complex document verification for issuing state driver’s licenses in order to meet requirements set by the REAL ID Act of 2005, among other concerns. Limitations on amounts charged. USPS officials and other stakeholders indicated that the fees charged for some of the additional nonpostal products and services would be too low to result in high revenues. For example, USPS officials found there would be little potential revenue from providing photocopying services or placing vending machines in retail facilities. Also, NPS officials and representatives from AFWA told us that current providers of National Parks and Federal Recreational Lands Passes and state hunting and fishing licenses generally do not generate much revenue due to the fee structure. In addition, USPS may or may not charge other entities a fee for sharing information at retail facilities. For example, USPS’s partnerships to share information on other government entities’ behalf, such as providing selective service registration forms and displaying information for DOJ’s National Crime Victims’ Rights week, do not generate revenue. Last, any effort to expand community services would not be intended to generate revenue. We have previously reported that foreign posts began offering nonpostal products and services to increase revenues, such as offering banking or financial services and making additional government services available in their retail facilities, but these efforts have had mixed results. Some foreign postal operators have expanded their financial offerings at retail facilities and have generated significant revenue from these efforts. For example: The United Kingdom’s postal retail operator—The Post Office—has an agreement with virtually all the retail banks in the United Kingdom that enables customers to use retail facilities to access their banking services. According to The Post Office, the financial services it offers generated €205 million in 2017 and €215 million in 2018. (Cost data were not available.) Officials from The Post Office told us this agreement has not only generated income, but also increased foot traffic to their retail locations. France’s postal operator, La Poste, established a bank in 2005 to provide a full range of banking products and services through its retail facilities. According to La Poste, its banking services generated net revenues of €5.5 billion in 2018, which was down from €5.6 billion in net revenue in 2017. However, USPS officials and representatives from banking associations cautioned that the financial and regulatory infrastructures of other countries are too different from those of the United States to suggest that USPS could achieve similar results. In addition, foreign postal operators are starting to sell their banks or have a franchise model with relatively lower-paid staff. Views Varied on Demand and Feasibility; Other Limitations Affect Viability of Nonpostal Services and Products There were mixed views on whether there would be demand for any or all of these nonpostal products and services. On the one hand, USPS officials, officials from other federal government entities, postmasters we surveyed, and stakeholders generally said that there was little demand for many of the additional nonpostal products and services USPS could offer at its facilities. In particular, postmasters did not indicate very high demand for any of the nine additional nonpostal products or services we asked about in our survey. The highest response for products and services in demand was only about 36 percent (53 of 149) of postmasters who said there was demand for notary services, and the next highest was about 33 percent (49 of 149) of postmasters who said there was demand for printing and photocopying services. Also, officials from the 12 federal government entities noted that while they were open to new partnerships in which mail clerks perform transactions on their behalf or in which the entities would have access to retail facility space, officials did not identify many specific examples of a need for such services. None of the postmasters we surveyed reported that they had been approached by community members in the last 10 years to share or lease space at their retail facility. On the other hand, based on studies we reviewed and interviews with representatives from consumer groups, there may be demand for certain offerings at retail facilities, such as check cashing and payday loan services, particularly if offered at a lower price than competitors. For example, in 2014, the Pew Charitable Trusts conducted a nationally representative survey of 1,626 adults and estimated that only around a quarter of American adults would be very likely or likely to use certain financial products, including check cashing, prepaid cards, bill pay, and small-dollar loans, if offered at USPS retail facilities. However, for those surveyed who were already using such services, respondents indicated they would likely obtain these at USPS retail facilities if offered at a lower price. Regarding feasibility, stakeholders and postmasters identified various issues related to what types of nonpostal products and services made sense to them. Postmasters indicated that displaying information about government programs and printing and photocopying services were the most feasible potential service, about 64 percent (95 of 149) and about 63 percent (96 of 149), respectively. In contrast, postmasters thought grocery pick-up and public wifi services were the least feasible, with about 78 percent (116 of 149) and 64 percent (96 of 149) of postmasters, respectively, indicating that they were not feasible. A representative from one of the employee unions told us that the success of additional nonpostal efforts would depend on the retail locations’ having adequate staffing levels. The representative also said that any nonpostal efforts should be designed to align with community needs and the work that retail facility employees already conduct. Officials from the Federal Communications Commission and representatives from a telecommunications association told us telecommunications companies may consider leasing space at USPS’s retail facilities that are suitable for wireless antennas to help build out 5G networks, subject to network design and business needs. In addition, representatives from one state motor vehicle administrator told us state motor vehicle administrations may consider partnering with USPS to provide state vehicle tag and title services on their behalf, such services, would not require substantial investment for USPS to undertake. Finally, USPS officials, federal government entities, and postmasters we surveyed, and stakeholders identified a variety of other limitations that would affect the viability of nonpostal products and services. We found that these limitations included limited interest for partnerships from other federal government entities, the size and unfavorable characteristics of retail facility space, complexities related to existing regulatory structures and entering into new markets, and personnel concerns. Examples of these limitations are described in table 5. Moreover, we also asked postmasters what challenges would prevent their selected retail facility from offering these additional nonpostal products and services. The most commonly cited challenges were insufficient staff or the need for additional staff training, particularly for notary services, driver’s license or other state license services, financial services, and banking services. As discussed above, given that these potential nonpostal products and services may have benefits but face concerns about their viability, USPS and policy makers need to consider the benefits, costs, and limitations of potential nonpostal efforts before introducing new efforts. In particular, though some efforts could create benefits like enhanced access for consumers, a variety of challenges may limit revenue generation in such a way that the potential offerings are unlikely to significantly improve USPS’s financial condition. Moreover, there are a number of limitations to be considered, including a potential lack of demand and factors affecting USPS’s ability to implement such offerings. Agency Comments We provided a draft of this report to the Census Bureau; Federal Communications Commission; GSA; U.S. Department of Veterans Affairs; U.S. Department of Homeland Security’s Federal Emergency Management Agency and Transportation Security Administration; U.S. Department of the Interior’s Fish and Wildlife Service and National Park Service; U.S. Internal Revenue Service; U.S. Department of Justice’s FBI and Office for Victims of Crime; U.S. Department of State; and USPS. The Census Bureau and GSA sent us technical comments, which we incorporated as appropriate. In its response, USPS reiterated its legal constraints but noted that there were other limitations affecting its ability to expand its offering of nonpostal products and services. These limitations include low net revenue potential, low potential for significant market share, high operational costs, and limits on amounts that could be charged. USPS noted, however, that it continues to explore partnerships with other federal agencies. We are sending copies of this report to the appropriate congressional committees, the Postmaster General, the Chairman of PRC, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2834 or rectanusl@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix IV. Appendix I: Objectives, Scope, and Methodology the costs, revenues, and non-revenue benefits associated with U.S. Postal Service’s (USPS) retail facilities; the nonpostal efforts USPS has conducted since 2008 to increase revenues and non-revenue benefits from its retail facilities, and the costs, revenues, non-revenue benefits, and challenges of such efforts; and the key considerations of additional nonpostal efforts that USPS could take to increase revenues and non-revenue benefits from its retail facilities. For this report, we focused on USPS-managed retail facilities, including USPS-operated post offices, postal stations, branches, and carrier annexes, as defined in USPS’s Annual Reports to Congress. We also used the term “nonpostal” to refer to activities that are not directly related to mail delivery. We selected 2008 to begin our analysis because that was when new restrictions on nonpostal services took effect. To describe the costs, revenues, and non-revenue benefits of USPS’s retail facilities, we reviewed USPS’s financial analyses of its retail network from fiscal years 2017 and 2018, the only years USPS conducted such analyses at the time we published this report. For these analyses, USPS identified the sources and amounts of retail facility costs (e.g., personnel, rent, and utilities) and revenues (e.g., stamp sales and post office box rental fees) for most of USPS’s retail facilities. Because some retail facility costs USPS identified (e.g., rent and personnel) also support non- retail functions, USPS used models to distribute such costs across each retail facility. USPS’s analyses also calculated the net revenues (revenues minus costs) for those retail facilities included in its analysis. Further, USPS examined characteristics of retail facilities, such as the facilities’ surrounding population densities, to describe factors that may be related to retail facilities that did or did not achieve positive net revenues. Based on interviews with USPS officials and USPS documents we reviewed, we determined the reliability of these analyses were sufficient to describe costs and revenues of retail facilities. We compared retail facility costs and revenues with USPS’s total costs and revenues to determine how much retail facilities contribute to total costs and revenues. In addition, we identified the number of retail facilities located in urban, suburban, or rural areas where the revenues did and did not cover costs. We defined these geographic categories using USPS’s definitions in its retail facility cost and revenue analysis. We also interviewed USPS officials to describe why revenue at retail facilities may not cover costs. To describe non-revenue benefits of USPS’s retail facility network, we reviewed relevant publications and studies, such as those conducted by USPS, USPS’s Office of Inspector General (OIG), and USPS’s oversight body, the Postal Regulatory Commission (PRC). We also examined 2018 data from USPS on the number of customer visits at the retail facilities for which data were available. Additionally, we interviewed representatives from two consumer groups, which we selected based on their ability to provide us with consumer viewpoints on retail facilities and their offerings, and two organizations representing USPS workers at retail facilities: American Postal Workers Union (APWU) and United Postmasters and Managers of America (UPMA). To examine the nonpostal efforts USPS has conducted to increase revenues and non-revenue benefits from its facilities since 2008 and their results, we first identified nonpostal products and services offered at USPS’s retail facilities through interviews with USPS officials and reviews of relevant publications, such as PRC’s report on USPS’s product offerings and USPS OIG reports. See table 6 for a complete list of USPS OIG reports we reviewed. We also identified the number of retail facilities that offered each identified nonpostal product and service, where data was available, using USPS’s facility data from November 2019, as well as information from USPS’s Inspection Service on facilities that participated in the Department of Justice’s Office of Victim’s National Crime Victims’ Rights week. Because USPS’s facility data included facilities without retail functions, we merged these with data from USPS’s fiscal year 2018 financial analysis of its retail network to ensure we included only facilities with retail functions. To describe the revenues of the nonpostal products and services we identified, we reviewed data from USPS’s fiscal year 2018 financial analysis of its retail network. We also reviewed USPS OIG reports to identify indirect ways that nonpostal products and services can contribute to revenue at retail facilities. To identify the costs of these activities, for which information was available, we reviewed USPS’s fiscal year 2018 Annual Compliance Report, PRC’s fiscal year 2018 Annual Compliance Determination Report, and non-public data provided to us by USPS. From these reports, we also determined whether the revenue USPS generated from these nonpostal efforts did or did not exceed costs. We reviewed only fiscal year 2018’s revenue and costs because we did not have revenue data prior to fiscal year 2017 and USPS was unable to provide information on trends. As one of the nonpostal efforts USPS can currently take at its retail facilities includes leasing space for revenue, we reviewed USPS’s data on the amount of vacant rentable space for fiscal years 2017 and 2018 (the only available years), on tenants at its facilities as of January 2020, and on the amount of revenue USPS collected from its leased space from 2018. Based on interviews with USPS officials and USPS documents we reviewed, we determined the reliability of these data were sufficient to describe USPS’s efforts and results of leasing excess space at retail facilities. To obtain stakeholder views on USPS’s nonpostal efforts, including the costs, revenues, non-revenue benefits, and challenges of these efforts, we interviewed officials from six federal government entities that partnered with USPS on initiatives as well as representatives from APWU, UPMA, and two consumer groups. See table 7 for a complete listing of the entities we interviewed. We also interviewed officials from the U.S. General Services Administration (GSA), which leases space on behalf of USPS to other federal government entities, and the Association of United States Postal Lessors, which represents entities that lease space to USPS. We also interviewed postal stakeholders at three retail facilities that had vacant leasable space, leased space to other entities, or offered nonpostal products and services. We selected these locations to obtain information on a variety of USPS’s current nonpostal efforts. To examine the benefits and key considerations of offering additional nonpostal efforts at USPS retail facilities, we interviewed USPS officials, postal stakeholders, and postal experts, selected based on prior work; reviewed prior GAO reports and relevant USPS OIG studies; and attended a forum exploring community use of USPS’s delivery infrastructure, including retail facilities. From the studies we reviewed and stakeholder suggestions, we selected and categorized examples of nonpostal efforts that were mentioned at least twice. To obtain stakeholder views on the potential benefits and limitations of such offerings, we interviewed representatives from consumer, industry, and state licensing groups. We selected these entities because of their potential to be affected by USPS offering additional nonpostal products and services. We also interviewed officials from the six federal entities that have partnered with USPS on initiatives and an additional six federal government entities that could potentially establish expanded or new partnerships with USPS. See table 7 above for a complete listing of the entities we interviewed. We also interviewed two foreign postal operators—France’s La Poste and the United Kingdom’s Post Office— that have experience with nonpostal products and services similar to those we reviewed and reviewed relevant documentation, such as their annual fiscal year 2018 financial reports. We selected these postal operators based on prior work and other studies and to provide us with insight into their experiences. Finally, we reviewed statutes, including the Postal Accountability and Enhancement Act (PAEA); regulations; and legal rulings, to evaluate USPS’s current legal authority to provide these services. The views presented in our report are not generalizable to those of all stakeholders. Further, we surveyed USPS postmasters to obtain additional perspectives on the benefits of USPS’s retail facilities, the nonpostal efforts offered at those facilities, and the key considerations of offering additional nonpostal products and services. Specifically, we conducted a non-generalizable, web-based survey of postmasters who managed retail facilities located in urban, suburban, and rural areas from August to September 2019. We defined these geographic categories using USPS’s definitions, as described above. Using the dataset of facilities from the USPS revenue study, we removed all facilities that were located in an overseas American territory, any facility missing a geographic category code, and certain kinds of facilities that were not relevant to the survey. Using these filters we identified a sample frame of 26,600 retail facilities. We randomly sorted the facilities within each of the three geographic categories and took the first 150 from each random sort. We then matched the selected USPS facilities with the postmaster responsible for them, using USPS’s postmaster data provided to us in May of 2019. Because postmasters may manage more than one retail facility, we capped the number of surveys an individual postmaster could receive at one. If a postmaster already selected in the random sort occurred again the selected facility was omitted from the sample. We restricted the total sample size to no more than 100 unique postmasters within each stratum. The sample is comprised of 83 postmasters who oversee an urban retail facility, 100 postmasters who oversee a suburban retail facility, and 100 postmasters who oversee a rural retail facility. Approximately 52 percent of our sample—or 146 postmasters—completed the survey. The survey questionnaire can be viewed in appendix II. In developing, administering, and analyzing the survey, we took steps to minimize non-sampling error that may result from differences in how a question is interpreted and the sources of information available to respondents. To help reduce measurement error, we consulted an experienced former postmaster for input on the development of the survey instrument, and also conducted pretests of the draft questionnaire with four postmasters drawn from the intended survey population. The questionnaire was modified throughout development and pretesting to improve clarity of the questions, and we removed questions when our modifications were unable to remedy observed difficulties in interpretation. To maximize survey response, we sent pre-notification letters by postal mail to the selected respondents prior to launching the web survey. After launching the survey, we sent multiple email reminders and extended the submission deadline, and also conducted follow-up phone calls. Appendix II: GAO’s Survey of U.S. Postal Service (USPS) Postmasters at Selected Retail Facilities GAO administered the survey questions shown in this appendix to learn more about USPS postmaster views related to the community benefits of USPS’s retail facilities, the nonpostal efforts offered at those facilities, and the key considerations of offering additional nonpostal products and services. The survey was divided into seven sections covering background, post office benefits to the community, current USPS nonpostal products and services, leasing space for revenue, sharing space, additional nonpostal products and services, and additional information. Open-ended follow-up questions were selectively included to allow respondents to provide more detail about their responses. This appendix accurately shows the content of the web- based survey, but the format of the questions and response options have been changed for readability in this report. For more information about our methodology for designing and administering the survey, see appendix I. Appendix IV: GAO Contacts and Staff Acknowledgments GAO Contact Lori Rectanus, (202) 512-2834 or rectanusl@gao.gov. Staff Acknowledgments In addition to the contact name above, Kyle Browning (Assistant Director); Anne Doré (Analyst in Charge); Isabelle Aboaf; Carl Barden; Karen Chen; Barbara El Osta; Geoff Hamilton; Serena Lo; Tina Paek; Samuel Portnow; Malika Rice; Matthew Valenta; and Lauren Voloder made key contributions to this report.

Why GAO Did This Study USPS manages over 31,000 retail facilities, which help it provide postal services throughout the country. However, USPS faces financial challenges. In general, USPS is prohibited by statute from providing nonpostal products and services (i.e., services not directly related to mail delivery) unless approved by the Postal Regulatory Commission. But given the ubiquity of the retail network, some stakeholders have suggested that offering additional nonpostal products and services could help USPS generate revenue and provide benefits for consumers and communities. GAO was asked to review opportunities to enhance the value of USPS's retail facilities. This report examines: (1) the costs, revenues, and other benefits associated with USPS's retail facilities; (2) USPS's nonpostal efforts since 2008 at retail facilities and the outcomes; and (3) considerations of new nonpostal efforts at retail facilities. GAO analyzed USPS retail facility costs and revenue data from fiscal years 2017 and 2018 (the only years available); reviewed relevant documents and reports from USPS and others; conducted a non-generalizable survey of USPS postmasters who managed rural, suburban, and urban retail facilities; and interviewed USPS officials, and stakeholders, including postal employee unions, industry and consumer groups, and federal agencies that partner with USPS to obtain views on current and potential nonpostal efforts. GAO is making no recommendations. USPS, in its comments, reiterated that it faces various constraints to new offerings at retail facilities. What GAO Found In 2018, U.S. Postal Service's (USPS) retail facilities, such as post offices, generated about $10.5 billion in revenue and cost approximately $5 billion to operate, making them profitable overall. While such facilities accounted for about 15 percent of USPS's total fiscal year 2018 revenues, and about 7 percent of its total costs, stakeholders identified other benefits that retail facilities provide for communities—particularly in rural areas—such as local access to government information and services. Since 2008, USPS has offered a variety of nonpostal products and services at its retail facilities that have generated some revenue and other benefits. USPS data show that the nonpostal products and services for which USPS captures revenue data, such as money orders, generated about $431 million in total revenue in fiscal year 2018 and were profitable overall. Stakeholders said many of these nonpostal products and services also provided other benefits, such as enhanced convenience for customers, and postmasters GAO surveyed said some offerings, such as passport services, were highly valued in their communities. Offering additional nonpostal products and services at USPS retail facilities could provide consumer, government, or community benefits, but viability may be limited. Stakeholders said new offerings, such as expanded financial products or government services could, for example, enhance consumers' access and government efficiencies. In particular, some noted that USPS could provide a viable banking alternative for those lacking banking services. However, USPS officials, postmasters GAO surveyed, and stakeholders GAO interviewed said that additional offerings may generate minimal revenue and that USPS may face factors limiting the viability of these offerings. For example, groups representing states' licensing agencies said offering state hunting and fishing licenses could be problematic given different state requirements. Also, stakeholders said USPS may not have the expertise nor the required capital to enter the market of some of these new offerings. Given such concerns, USPS and policy makers need to carefully weigh costs, benefits, and limitations of any new offerings.

gao_GAO-20-385

gao_GAO-20-385_0

Background USPS has the mission of providing prompt, reliable, and efficient universal postal service, and federal law requires USPS to “provide postal services to bind the Nation together through the personal, educational, literary, and business correspondence of the people.” USPS is required to serve, as nearly as practicable, the entire population of the United States. USPS has a number of key stakeholders, each with different interests in USPS and its operations (see fig. 1). USPS is a key part of the mailing industry, and over time, it has become both a competitor and partner to private companies that also operate in the broader mail and delivery industries. For example, although United Parcel Service (UPS) and FedEx both pay USPS to deliver packages that they enter into USPS’s system at local post offices where carriers pick up their mail, they also compete with USPS for end-to-end package delivery business, such as moving packages from the retailer to the purchaser. Similarly, FedEx is USPS’s largest contractor, providing air transportation for Priority Mail Express (formerly Express Mail), Priority Mail, and First-Class Mail. UPS is also one of USPS’s largest contractors, providing long-distance mail transportation. Over the years, legislation has changed key aspects of the business model used to provide the nation’s postal services. Until 1970, the federal government provided postal services via the U.S. Post Office Department, a government agency that received annual appropriations from Congress. At that time, Congress was involved in many aspects of the department’s operations, such as selecting postmasters and setting postal rates and wages. In addition, the President controlled the hiring and firing of Postmasters General, as it was a cabinet-level position. By the late 1960s, the department had several major problems including financial losses, management problems, service breakdowns, and low productivity. Because key postal business decisions were made by Congress through the legislative process, postal management had limited ability to plan and finance department operations and capital investments in accordance with postal needs. In order to improve and modernize postal services, the Postal Reorganization Act (PRA) was enacted in 1970 and replaced the U.S. Post Office Department with USPS, an independent establishment of the executive branch of the government of the United States. Congress designed USPS to be a self-sustaining, business-like entity headed by a Board of Governors that would cover its operating costs primarily with revenues generated through the sales of postage and postal-related products and services. However, by the early 2000s, USPS faced a bleak financial outlook that put its mission of providing universal postal service at risk, according to the 2003 Presidential Commission on the United States Postal Service. The Commission evaluated USPS’s business model and concluded that USPS must have greater flexibility to operate in a business-like fashion, but that this latitude required enhanced transparency to enable effective management and congressional oversight. The Postal Accountability and Enhancement Act (PAEA) was enacted in 2006. PAEA provided USPS additional pricing flexibility for mail products, but with provisions for increased transparency, oversight, and accountability, among other things. Specifically, PAEA gave USPS broader latitude to change postal rates in a more streamlined process that included review by the newly created Postal Regulatory Commission (PRC). The PRC, which replaced the former Postal Rate Commission, is an independent establishment of the executive branch responsible for regulating USPS. PRC is required to make annual determinations of USPS’s compliance with mail delivery standards and postal rate requirements. If PRC finds noncompliance, it is required to specify USPS actions to restore compliance. USPS Cannot Become Financially Self-Sustaining under Its Current Business Model due to Three Key Challenges USPS’s current business model is not financially sustainable due to declining mail volumes, increased compensation and benefits costs, and increased unfunded liabilities and debt. USPS’s costs continue to rise faster than its revenues, and although USPS has made changes over the years to address these challenges, its efforts have been limited by stakeholder opposition and statutory requirements. Declining Mail Volumes As online communication and payments have expanded, USPS continues to face decreases in mail volume, its primary revenue source. First-Class Mail volume has declined 44 percent since fiscal year 2006, the year that total mail volume peaked. The long-term decline of First-Class Mail volume, which USPS has stated was exacerbated by the Great Recession and expects to continue for the foreseeable future, has fundamental implications for USPS’s business model because First-Class Mail is USPS’s most profitable class of mail. USPS Marketing Mail— which comprises most other mail volume—declined 27 percent from fiscal year 2007 to fiscal year 2019, in part due to electronic advertising alternatives. The volume of USPS competitive products more than tripled since fiscal year 2007. This volume, however, began to decline in the second half of fiscal year 2019 due to growing competition for package delivery. USPS has taken steps to right size its operations in response to declining mail volumes. For example, in both 2009 and 2011, USPS announced plans to close several thousand USPS retail facilities. However, due to stakeholder opposition—including from members of Congress, postal unions, and local communities, among others—USPS instead closed a few hundred retail facilities. USPS also expanded the alternative options for customers to access retail postal products and services outside of USPS-operated postal facilities—such as self-service kiosks and partnerships with other retailers such as contract postal units. According to USPS, as a compromise effort to right size the retail network and due in part to USPS’s efforts to expand retail alternatives, USPS began reducing retail hours at selected post offices in 2012, ultimately decreasing retail hours at approximately 13,000 post offices. Another major cost-cutting effort was its 2011 Network Rationalization Initiative, a multi-part plan to consolidate its mail processing network. USPS consolidated more than 160 mail processing facilities, but did not fully implement this initiative following opposition from various stakeholders. In addition to stakeholder opposition to changing postal services, federal laws also factor into USPS’s limited ability to respond to declining mail volumes. For example, federal laws define the level of postal services USPS is to provide, postal products, and pricing. Postal services to be provided: USPS has limited ability to make changes in the postal services it provides. Specifically, USPS is required to provide 6-days-a-week delivery and to operate postal facilities across the country. Federal law requires USPS to provide the maximum degree of effective and regular postal services to rural areas, communities, and small towns where post offices are not self- sustaining. Federal law also limits USPS’s ability to close retail facilities. For example, USPS cannot close a small post office solely because it is unprofitable. As of the end of fiscal year 2019, there were approximately 34,600 postal retail outlets nationwide, including approximately 31,300 USPS-managed post offices, branches, and stations, and, as we recently reported, USPS’s analysis showed that about 36 percent of its retail facilities were unprofitable in fiscal year 2018. Postal products and pricing: USPS’s pricing flexibility is limited by a price cap on market-dominant products that generally limits rate increases for these products to a common measure of inflation. Each competitive product is required to cover its attributable costs; competitive products collectively are required to recover their attributable costs; and competitive products collectively are required to cover a PRC-specified minimum of USPS’s institutional costs. In addition, USPS is prohibited from providing new nonpostal products and services. Such requirements affect USPS’s ability to increase revenues. Increased Compensation and Benefits Costs While mail volumes have decreased, USPS’s compensation and benefits costs for current employees have been increasing since 2014, despite USPS’s efforts to control these costs. Although USPS reduced its total workforce (career and non-career employees) from 785,900 in fiscal year 2007, to 617,700 in fiscal year 2013, its workforce increased to about 630,000 in fiscal year 2019. Similarly, as we previously reported, recent trends show total work hours increased from a combination of new hiring and increased work hours for current employees. Specifically, we reported that from fiscal years 2014 through 2018, work hours increased by 5.4 percent. The number of work hours associated with higher costs— overtime and penalty overtime—have also been increasing. According to USPS, total compensation and benefits costs increased by almost $1 billion in fiscal year 2019 alone. USPS has implemented changes to help control employee compensation and benefits costs, including lowering pay for new career employees and increasing use of non-career employees. For example, as we previously reported, starting about 10 years ago, USPS’s collective bargaining agreements have included the ability to hire up to 20 percent of the workforce as non-career employees. Non-career employees are less costly because they generally have lower pay rates and are not entitled to the full federal benefits received by career employees. According to USPS officials, non-career employees are also “more flexible” because there are fewer restrictions on their tasks and schedules. We recently reported that our analysis estimated that USPS likely saved about $6.6 billion from fiscal years 2016 through 2018 from increased use of non- career employees. USPS has recognized trade-offs in increasing the use of non-career employees in entry-level positions, such as a high turnover rate, as would be expected for almost any entry-level position in the private sector. With respect to benefits costs for current postal employees, we have recently reported that USPS has also achieved savings by gradually decreasing its contribution percentage for employee health insurance premiums over the past decade, with corresponding increases in the contribution percentage paid by employees. These changes were negotiated with the four major postal labor unions and were included in successive collective bargaining agreements, each of which covered a multi-year period. We found that the reduction in USPS health insurance contributions generated estimated savings of about $1.4 billion for fiscal years 2016 through 2018. A number of restrictions limit USPS’s ability to control employee compensation and benefits costs. As we recently reported, USPS compensation and benefits costs—which represent about three-fourths of its total costs—are driven by a mix of USPS contracts and policies, including collective bargaining agreements negotiated with unions representing 92 percent of USPS employees, and statutory requirements governing USPS employee pay and benefits. When USPS and its unions are unable to agree, the parties are required to enter into binding arbitration by a third-party panel. USPS’s collective bargaining agreements with these labor unions, some of which were established through binding arbitration, have established salary increases and cost- of-living adjustments and, as mentioned above, have also capped the number of non-career employees at approximately 20 percent of the number of employees covered by the agreements. Federal law requires USPS to participate in the Federal Employees Health Benefits Program (FEHBP), which covers current employees and retirees, as well as federal pension and workers’ compensation programs. Further, USPS must provide fringe benefits that, as a whole, are no less favorable than those in effect when the Postal Reorganization Act of 1970 was enacted. Increased Unfunded Liabilities and Debt USPS’s unfunded liabilities and debt, which consist mostly of unfunded liabilities for retiree health and pension benefits, have become a significant financial burden, increasing from 99 percent of USPS’s annual revenues at the end of fiscal year 2007 to 226 percent of its fiscal year 2019 revenues. At the end of fiscal year 2019, USPS’s unfunded liabilities and debt totaled approximately $161 billion. However, it has begun paying down this debt in recent years, leaving a balance of $11 billion at the end of fiscal year 2019 (see fig. 2). Total unfunded liabilities have risen in part due to USPS not making payments to fund its retiree health and pension benefits. USPS has stated that it prioritizes its “primary universal service mission” when it is unable to fulfill all of its financial obligations, and that it therefore did not make payments to fund its postal retiree health benefits and pensions to minimize the risk of running out of cash. In doing so, USPS cited its precarious financial condition and the need to cover current and anticipated costs and any contingencies. It has not paid $55.4 billion in required payments for funding these benefits through fiscal year 2019, including $47.2 billion in missed funding payments for retiree health benefits since fiscal year 2010, and $8.2 billion for funding pension benefits since fiscal year 2014. In addition, for many years, USPS had been at its statutory debt limit of $15 billion; however, it has begun paying down this debt in recent years, leaving a balance of $11 billion at the end of fiscal year 2019. A number of federal laws define the requirements for USPS’s retiree health and pension benefits that comprise most of its unfunded liabilities. Retiree health benefits: Federal law establishes certain requirements for postal retiree health benefits, including basic requirements for coverage eligibility and contributions. In administering the FEHBP, the Office of Personnel Management (OPM) negotiates with the insurance providers to establish the level of benefits provided to beneficiaries. USPS is required to prefund its share of health benefits for its retirees. Under PAEA, the first 10 years of prefunding payments were fixed—ranging from $5.4 billion to $5.8 billion annually from fiscal years 2007 to 2016. From fiscal years 2007 through 2016, USPS was also required to continue “pay-as-you-go” payments for its share of premiums for current retirees. The permanent schedule for USPS payments to prefund postal retiree health benefits under PAEA required USPS to make annual payments starting in fiscal year 2017. Currently, USPS no longer makes payments for retiree health benefits premiums. Starting in 2016, these premiums are paid out of the RHB Fund until it is depleted, whereupon USPS will resume paying premiums on a pay-as-you-go basis. As we previously reported, survey data we reviewed indicated that most companies do not offer retiree health benefits and that the number of companies providing such benefits is decreasing over time. Many companies that have retained their retiree health benefits have done so by making changes to control costs, including tightening eligibility and restructuring benefits. However, all approaches we identified have different potential effects and would require congressional action. Pension benefits: Federal law also requires USPS to finance its pension benefits under the Federal Employees Retirement System (FERS) and the Civil Service Retirement System (CSRS) and contains specific provisions defining USPS’s required contribution level to fund these benefits. USPS’s payments consist of a rolling 30-year amortization schedule to address unfunded FERS liabilities, an amortization schedule to address unfunded CSRS liabilities by 2043, and the normal costs of FERS benefits for current employees. Selected Domestic Companies and Foreign Posts Took Key Steps to Address Major Challenges The large domestic companies we selected in the airline, auto, and railroad industries took actions over a number of years to address major business challenges. Airlines such as Delta, American, and United faced competition from low-cost airlines, downward pressure on airfares, and rising compensation, benefits, and volatile fuel costs. These challenges were exacerbated by the economic downturn that began in 2000, the terrorist attacks of September 11, 2001, and the Great Recession that began in December 2007, all of which temporarily depressed demand for airline travel. Similarly, General Motors (GM) and Ford Motor Company (Ford) faced competition from lower cost competitors, the Great Recession, a workforce and networks too large to be supported by smaller sales volumes, and other changes in the market. Likewise, large U.S. railroads competed for freight and passengers from other transportation modes, such as the trucking and airline industries that operated over publicly provided infrastructure, while railroads had to invest in their own infrastructure. Based on our review of the selected companies’ annual reports and statements to the SEC as well as selected federal laws, and GAO, CRS, and other organizations’ reports, books, and academic articles, among other sources, we found that selected companies made changes to (1) products and services, (2) financial self-sustainment, and (3) use of the bankruptcy process. While some of the selected businesses restructured through a bankruptcy proceeding, other businesses took similar actions outside of the bankruptcy process. Mergers also played an important role for the airlines and railroads. Actions Taken by Companies to Address Challenges Actions Regarding Products and Services The selected companies made multiple changes to their products and services. Specifically: Airlines: Selected airlines altered pricing by changing route structure to focus on more profitable routes and adding fees, such as for checked baggage. In addition, all three selected airlines merged with other major airlines, thereby broadening their routes and revenues. Automakers: Selected automakers focused on producing more profitable brands and models, discontinuing some models and introducing others. For example, during its financial difficulties about a decade ago, GM discontinued a number of unprofitable brands. In 2018, after years of declining car sales, Ford said it would eliminate some of its most well-known cars in North America, allowing it to devote more resources to sport utility vehicles and trucks. Railroads: Large railroads focused on more profitable routes and abandoned unprofitable routes or sold them to other railroads. For example, the federal government created a new freight railroad, Conrail, by merging several bankrupt railroads in the Northeast and Midwest. As we have reported previously, federal government deregulation of railroad pricing and contracts after 1980 also helped Conrail to reach profitability and increase capital investment. Actions Regarding Financial Self-Sustainment Cost reduction was a major theme for the selected businesses in the airline, automotive, and rail industries, particularly with respect to compensation, benefits, and infrastructure costs. Specifically: Airlines: The three selected airlines negotiated wage cuts and work rule changes with their unions; made workforce reductions, in part by outsourcing work; and cut pension and retiree health benefit programs. Wage cuts included all levels of employees, such as management, pilots, flight attendants, and mechanics. Benefit cuts involved reducing the level of pensions and retiree health benefits and transitioning pension programs from defined benefits plans to defined contribution plans that were structured to be less costly. Airlines also reduced infrastructure costs by eliminating some hubs, reducing the total number of aircraft, and changing the mix of aircraft in their fleet to save on maintenance and fuel costs. The airlines further cut costs by restructuring debt, reducing facility leasing costs, and renegotiating aircraft leases and vendor contracts. While in bankruptcy, the airlines took major actions to reduce their costs. For example, United implemented steep pay cuts, cut retiree health benefits, and terminated its defined benefit pension plans, resulting in the Pension Benefit Guaranty Corporation (PBGC) assuming responsibility for some of its pension payments, and a reduction in benefits for the plan’s participants. United also cut its workforce size by 31 percent, reduced the number of airplanes by 19 percent, and reduced the total number of flights by 13 percent. Delta and American also reduced pay and pension benefits while in bankruptcy, and the PBGC assumed responsibility for some of Delta’s pension liabilities. Automakers: The two selected automakers negotiated pay cuts, lower wages for entry-level employees, and changes to work rules designed to increase competitiveness; cut the workforce size in about half; made changes to employee benefits; closed many auto plants and dealerships; eliminated some vehicle brands and models; and changed the production process to increase efficiency. Specifically, The International Union, United Automobile, Aerospace and Agricultural Implement Workers of America (UAW) agreed to cuts in compensation for the automakers’ employees to levels paid by GM and Ford’s competitors. UAW also agreed to move retiree health care benefits into a private Voluntary Employee Beneficiary Association (VEBA) for current and former UAW-represented employees in 2007; the VEBA took over health benefits for retirees starting in 2010. The automakers also closed defined benefit pension plans to new participants and moved to defined contribution pension plans for eligible new employees. Railroads: The Staggers Rail Act of 1980 was enacted to improve the financial stability of the railroad system. Subsequently, railroads improved their financial health through, among other things, cost reduction measures such as reducing their workforce through layoffs and federal buyouts agreed to by unions, and abandoned or sold off unprofitable rail lines to reduce infrastructure and operating costs. Overall railroad employment fell greatly as railroads consolidated, reduced service, and changed work rules. For example, Conrail reduced its workforce from about 82,000 employees in 1977 to about 21,000 in 1996. Congress specifically facilitated Conrail’s downsizing by passing the Northeast Rail Service Act of 1981, which among other things, directed a $200 million a year reduction in labor costs and authorized Conrail to terminate employees. The U.S. government provided assistance to selected companies in various forms, including appropriated funds, loans, and other actions that helped enable companies to reduce their expenses. For example: Automakers: The federal government committed $49.5 billion in funding to help GM continue to operate while restructuring. After the government became the majority owner of the GM that emerged from bankruptcy, the Administration laid out core principles that included managing its ownership in a hands-off manner and voting as a shareholder only on core governance issues. Airlines: Under the 2001 Air Transportation Safety and System Stabilization Act, the federal government provided nearly $5 billion in compensation to airlines for losses due to the September 11, 2001 terrorist attacks, including $856 million for American, $668 million for Delta, and $782 million for United. Additionally, under the Emergency Wartime Supplemental Appropriations Act, $2.4 billion was appropriated to the Transportation Security Administration to compensate airlines for certain security expenses and fees, including $358 million to American, $411 million to Delta, and $300 million to United. Railroads: The federal government spent about $8 billion creating, subsidizing, and preparing Conrail for sale to the private sector. This funding included $7 billion through 1988 for purchasing properties of bankrupt railroads, operating subsidies, and capital improvements and employee buyouts. Amtrak, the national passenger railroad, took over money-losing intercity and commuter passenger rail services and funded federal payments of up to $25,000 for each laid-off employee as authorized by legislation. Use of Bankruptcy Some airlines, automakers, and railroads made changes through the bankruptcy process. Specifically: Airlines: All three selected airlines went through bankruptcy proceedings—United in December 2002, Delta in September 2005, and American in November 2011. The actions these airlines took to reduce costs while in bankruptcy are discussed above. Automakers: GM declared bankruptcy in June 2009 to implement its restructuring plan. The federal government became GM’s majority shareholder and continued to provide financial assistance while GM was in bankruptcy. The bankruptcy court approved the sale of substantially all of old GM’s assets to a newly formed company (“new GM”) in June 2009 as well the old GM’s amended bankruptcy plan in March 2011, and its assets and liabilities were transferred to liquidating trusts. These actions and the restructuring—which included major cost reductions described above—helped enable GM to report positive net income in every year from 2010 through 2019. Railroads: In 1970, the Penn Central Railroad, one of the largest in the country at the time, filed for bankruptcy. As Penn Central’s losses continued while in bankruptcy, the value of whatever assets that might have been available to satisfy its creditors’ claims was further eroded. Therefore, the bankruptcy proceeding was initiated to liquidate the railroad to meet the demands of its creditors. Faced with the potential cessation of railroad service for an entire section of the country, the federal government created Conrail to take over and operate specified portions of Penn Central as well as several other bankrupt railroads in the Northeast United States. Legislation was enacted that, combined with Conrail actions described above, enabled Conrail to become profitable. Foreign Governments and Foreign Posts Took Actions to Address Challenges In the countries we selected, foreign governments also took actions to develop goals for their postal operators that enabled changes in the postal operators’ institutional structures and actions to address competitive pressures, economic downturns, and market changes. Based on our review of the selected countries’ government reports, including summaries of postal reform legislation, annual reports from foreign posts, and interviews with foreign government officials and representatives of foreign posts, we found that changes were made to foreign posts’: (1) products and services; (2) financial self-sustainment; and (3) institutional structure. Some of these actions were authorized by legislation that changed the status and duties of the postal operators; others were taken over a lengthy period that predated passage of key legislation. Some foreign posts also diversified into nonpostal products and services; however, postal-specific challenges and changes are the focus of this discussion. Actions Regarding Products and Services The selected foreign posts made multiple changes to their products and services. We found that, when transforming their postal operations, selected countries determined the level of postal services they deemed necessary to adequately serve the public. The main product changes involved expansion of their package delivery business by enhancing service and investing in facilities and sorting equipment dedicated to handling packages. Large increases in their package volumes and revenues helped offset declining letter mail volume and revenues. Two of the selected foreign posts reduced service levels to help control costs. For example, according to New Zealand Post officials, in response to reduced mail volume, New Zealand Post reduced its required frequency of mail delivery in urban areas from 5 to 3 days while maintaining 5-day delivery in rural areas with the stated goal of ensuring that postal service remained viable without government subsidization. Australia Post revised the service standards it provided for delivery of letter mail in 2016, resulting in slower delivery of some mail. In addition, the main pricing changes have involved price increases for all mail, as well as the introduction of discounted postal rates for letter mail entered at processing facilities that generally were closer to the final destination of the mail. For example, Royal Mail raised postal rates above the rate of inflation after relaxation of its price cap. In France, the postal regulator established a price cap in 2015 of 3.5 percent per year (in addition to inflation) and established a new cap in 2018 of 5 percent annual rate increases (including inflation) for 2019 through 2022 for letter mail and packages considered part of universal postal service. The 2018 price cap allows La Poste more price flexibility than the previous one. Likewise, Australia Post implemented an above-inflation rate increase in January 2020, citing the need to generate revenues to offset growing financial pressures from declining letter mail volume. Actions Regarding Financial Self-Sustainment Cost reduction was a major theme for the selected foreign posts, particularly with respect to compensation, benefits, and infrastructure costs. These actions were stimulated in part by reductions in mail volume and the associated revenues and workload, and in part by legislative or regulatory changes that allowed greater competition and created incentives, such as privatization, which resulted in shareholder pressure to enhance or encourage organizational profitability and efficiency. Because postal operations are labor-intensive, actions to address workforce costs were particularly important to improving financial results. These often went hand in hand with outsourcing, network restructuring, reductions in service levels to better align service with demand, and other changes to increase productivity and achieve cost savings. The governments and the postal operators of selected countries also considered the effects on stakeholders when making postal reform decisions. Specific changes varied from one foreign post to another. Workforce: In Germany, Deutsche Post officials told us that its employees hired after 1990 were designated private sector employees with lower pay and benefits than postal employees who were previously hired as civil servants. In France, La Poste officials told us that La Poste likewise transitioned its workforce in the 1990s from civil servants to private employees and ended recruitment of civil servants in 2000. They also said that this transition provided La Poste with a more flexible workforce and reduced its pension liability. Australia Post closed its defined benefit pension plan to new employees in 2012, while Royal Mail is transitioning to a defined contribution pension plan that it introduced in 2018. Royal Mail and New Zealand Post also have reduced the size of their workforce in recent years. Infrastructure: Some selected foreign posts consolidated their mail processing networks to reduce costs. For example, following the reunification of Germany, Deutsche Post replaced more than 320 mail processing facilities with 82 such facilities. Royal Mail and La Poste reduced the number of mail sorting centers by about 40 percent over the past two decades. In addition, all five selected postal operators have made changes to reduce retail network costs. As we reported in 2011, some foreign posts reduced the number of postal operator-owned and -operated facilities and in some cases closed facilities in an effort to reduce costs. At the same time, some minimized this disruption by expanding retail access through alternatives such as Internet sales and partnerships with retail businesses such as grocery stores or pharmacies. We reported that these changes either reduced operating and labor costs or improved customer service, in some cases because the partner retail facility stays open longer, or both. This trend continues. Deutsche Post, Post Office Limited in the United Kingdom, La Poste, Australia Post, and New Zealand Post have outsourced or franchised most of their postal retail functions to private nonpostal operators. For example, Deutsche Post franchised its postal retail outlets to local businesses to not only reduce expenses, but also increase the availability of postal retail services nationwide by putting retail counters in stores that were open longer than traditional post offices. Deutsche Post representatives stated that while there was some initial resistance to these changes, these concerns abated after a few months as customers realized they received better service and longer hours. The representatives also said the number of retail outlets has increased in recent years in response to increased demand for e-commerce package returns. La Poste has a substantial and growing proportion of retail facilities operated by private providers (in partnership with small shops, especially in rural areas) or co- located in local government-owned buildings (in partnership with local town halls in rural areas). In addition, Australia Post combined its letter and parcel delivery networks in 2018 to obtain efficiencies. Productivity: All of the postal operators in the selected countries took actions to enhance productivity, such as improving automation of mail processing, modernizing and streamlining operations, and changing work processes. For example, Deutsche Post officials stated they had streamlined their parcel sorting process and went from 140 parcel sorting centers in 1990 to 34 in 2019. Deutsche Post officials stated that this streamlining improved service performance for parcels; previously, most parcels were delivered within 3 or 4 days; by 2019, 93 percent of parcels were delivered within 1 day. Royal Mail officials also stated that they increased their level of automation and introduced new methods of parcel delivery, such as new high-capacity equipment for mail carriers, to increase efficiency as well. Government Assistance: The governments of some of the selected countries provided assistance to their postal operators in various forms, including assuming pension costs, granting tax exemptions, and providing subsidies to postal retail operations. For example, the governments of France and the United Kingdom assumed costs of defined benefit pensions for postal employees who are civil servants, while the government of Germany assumed these costs to the extent that they exceed the costs of private sector pensions. While all new employees are employed as private sector employees, German audit officials stated that the government’s pension obligation for postal employees who are or were civil servants and their dependents will last until 2079 and cost the government about €306 billion. The assumption of these pension plans was important in facilitating the privatization of these postal operators, according to Royal Mail and Deutsche Post officials, because without the reduced unfunded pension liabilities, the stock offerings for the newly created companies would have been much less attractive to private investors. France also provides subsidies for certain postal activities. In addition, the United Kingdom split off the postal retail network from Royal Mail into Post Office Limited, a separate entity owned and subsidized by the national government. Actions Regarding Institutional Structure Each of the five foreign countries we selected changed their institutional structure following the development of goals for postal transformation that were tailored to national needs and priorities. Each of these countries had definitions of universal postal services including provisions for nationwide delivery and access to postal retail services. As the following examples illustrate, the national context of each country has been of central importance to shaping these goals. In addition, postal transformation in the three selected European countries—Germany, the United Kingdom, and France—also had an international context in the broader effort to create and promote a single European internal market. Germany: In the 1990s, the German government changed its postal operator, Deutsche Post, from a government agency to a government- owned corporation. In 2000, the government changed Deutsche Post to a privately owned company so it could raise capital, modernize, and create a sustainable infrastructure. Goals for the newly created Deutsche Post were to maintain the high level of postal services, increase efficiency, and enhance profit. The legislation that created Deutsche Post also gave it more flexibility to respond to changes in the market. Currently, Deutsche Post remains a private company with the government holding a minority of its shares. United Kingdom: In 1969, the government of the United Kingdom changed the Post Office, its postal operator, from a government department to Royal Mail, a government-owned corporation prior to changing it to a privately owned company. It began privatizing Royal Mail in 2013 so Royal Mail could become more modern and competitive by raising private capital, operating with more flexibility, and be subject to shareholder scrutiny to drive efficiency. As mentioned above, the government of the United Kingdom also split off postal retail units into a new government-owned entity called “Post Office Limited” that is separate from Royal Mail. France: In 1991, the government of France changed La Poste, the postal service of France, from a government department to a public industrial and commercial establishment. In 2010, the government of France converted La Poste into a state-owned public limited company. This step allowed La Poste to raise additional public capital for investments to maintain and modernize its network, build a European parcel and express network, allow acquisitions outside Europe, and add nonpostal products and services, such as expanding its banking services. Australia: In 1989, the government of Australia changed its postal operator, Australia Post, from a government department to a government-owned corporation. It is required to earn a reasonable rate of return on its assets, maintain its equity, pay a reasonable dividend to the government, and be liable for the same taxes and charges as its competitors. New Zealand: According to a recent report, New Zealand Post began as a government department and became a state-owned enterprise in 1987, when legislation (State-Owned Enterprises Act 1986) created several such entities to address challenges in the national economy. Such corporations are required to be as profitable and efficient as a comparable business not owned by the state. USPS’s Transformation Involves Reassessment of Three Critical Foundational Elements of Its Business Model Congress will face difficult choices in fundamentally reassessing the three critical foundational elements of USPS’s business model—level of universal postal service, financial sustainability, and institutional structure. These choices are likely to require changes in laws and will have differing effects on postal stakeholders. While the specific impacts will depend on the changes made, some or all of USPS’s stakeholders could be affected and these impacts should be considered as part of any reassessment. All three key areas are interrelated and significant changes in one area may affect another. For example, we have testified that Congress faces a tradeoff between the level of postal services the nation needs and the level of postal services the nation is willing to pay for. Level of Universal Postal Service Needed Based on our prior work, a starting point for a fundamental reassessment of USPS’s business model should be determining the level of postal services the nation needs. While mail volumes have declined since fiscal year 2006, businesses, governments, and households still pay USPS billions of dollars annually to deliver more than 140 billion of pieces of mail, demonstrating a continued nationwide demand for postal services. We and others—such as USPS, PRC, and USPS OIG—have called for a fundamental reexamination of what postal services the nation needs now and may need in the future. In particular, we have testified that USPS’s growing financial difficulties, combined with changing demand for postal services, have provided Congress with an opportunity to examine and potentially redefine what postal services should be provided on a universal basis and how they should be provided. As mentioned above, there are numerous federal laws and requirements related to the provision of universal postal service. For example, 6-day delivery has long been required by annual USPS appropriations acts. Over the past decade, legislation has been introduced, and USPS and others have proposed reducing the frequency of delivery. However, no legislation has been enacted that would allow USPS to reduce delivery frequency. There is also no consensus on the level of postal services the nation needs. Changes in service levels face opposition from some stakeholders, such as labor unions, affected communities, and the general public. Currently, legislation has been introduced that supports the preservation of both 6-day and door-to-door delivery for addresses that have it, and some mailer groups support one or both of these positions. Representatives from postal labor unions we spoke with stated that universal postal service is appropriate as currently defined and could be expanded to provide more products and services. Stakeholders have also expressed differing views on whether the frequency of delivery should be reduced to help USPS address its financial problems. USPS and PRC have estimated that eliminating Saturday delivery would reduce USPS’s costs but also would likely affect mail volume sent by business mailers, although USPS and PRC disagreed on the degree to which it would do so. USPS estimated that it could save $1.4 billion to $1.8 billion a year by reducing the frequency of mail delivery to 5 days while maintaining 7-day package delivery. To put these potential savings into context, delivery is USPS’s most costly operation. We reported, however, that USPS would face challenges in, among other things, how efficiently USPS would absorb the additional volume delivered in the remaining delivery days and its potential effect on mail volume. We also described potential trade-offs, such as possibly reducing the demand and value of USPS products if customers are not getting their delivery needs met. Further, key postal stakeholders hold opposing views on many other options that have been proposed. For example, to raise revenues, USPS and some postal labor unions favor eliminating or raising the price cap on market-dominant products, which would enable USPS to raise rates more than the rate of inflation but would require changing the current regulatory system. Mailers, however, have expressed opposition to increasing postage rates higher than the rate of inflation. Postal labor unions also favor increasing revenues by introducing new postal and nonpostal products and services. We have recently found, however, that USPS’s nonpostal revenues generated at postal retail facilities are small and that there are limited opportunities to generate revenues from nonpostal products and services from USPS’s delivery network. For example, we reported that nonpostal products and services offered through USPS’s postal retail facilities generated about $431 million in fiscal year 2018, accounting for less than 1 percent of USPS’s total revenue. In addition, we reported several potential limitations to USPS adding nonpostal services to USPS’s mail carrier activities, such as checking in on homebound and older residents and reporting signs of blighted properties. These limitations included, among other things, limited net revenue potential and a potential adverse effect on mail service delivery. Financial Sustainability A fundamental reassessment of USPS’s business model would include determining the degree to which USPS should be financially self- sustaining, i.e., the degree to which USPS’s operating costs and liabilities should be covered by ratepayers (such as businesses and individuals who pay USPS to send mail). If a reassessment concluded that USPS should be fully self-sustaining, past legislative proposals that would change elements of USPS’s costs and revenues may be worth congressional consideration. Stakeholders, however, have not reached a consensus on any of these proposals and none has been enacted. Another avenue is to focus on reducing costs. As we have also reported, compensation and benefits costs, which comprise about three-quarters of USPS’s operating costs, are driven by a mix of USPS contracts and policies, including collective bargaining agreements negotiated with unions representing 92 percent of USPS employees and statutory requirements governing USPS employee pay and benefits. USPS compensation and benefits costs for its active employees increased by almost $1 billion in fiscal year 2019 despite a slight decrease in the size of the workforce and declining workload from reduced mail volume. While USPS has been able to make some reductions in pay and benefits, its ability to control compensation costs is significantly inhibited by the collective bargaining process, which results in binding arbitration if an impasse is reached. According to USPS, all negotiations take place against the backdrop of binding arbitration (and the arbitrators have historically been reluctant to deviate from the status quo), resulting in only incremental changes. We have long supported changing the laws regarding collective bargaining to require that USPS’s financial condition be considered in binding arbitration. We have also reported that the collective bargaining structure, which was established many years ago, should be reexamined considering the dramatic changes in USPS’s competitive environment and rising personnel costs that have contributed to USPS’s losses. Multiple bills have been proposed changing the process and/or criteria for collective bargaining to a different standard. The 2018 report from the Task Force on the United States Postal System recommended that collective bargaining over compensation should be eliminated for postal employees. While eliminating or revising the collective bargaining process could potentially provide USPS greater flexibility in employee pay, there would be trade-offs. For example, we recently found that the potential annual cost savings associated with USPS implementing cuts for all current employee pay by 1 percent would be about $321 million; a 10 percent cut would potentially save $3.2 billion. However, we also reported that while USPS could reduce its compensation costs through efforts such as reducing mail delivery frequency, USPS would face challenges in realizing these savings, such as the extent to which workhours could be reduced. Furthermore, these savings could be offset by other factors including service or morale issues. With respect to benefits, we recently reported on a wide range of possible changes that would reduce or limit costs for postal retiree health benefits, nearly all of which would require a legislative change. Some approaches would shift costs to the federal government; some would reduce benefits or increase costs to postal retirees or employees; and some approaches would change how benefits are funded. Similar types of legislative changes could be considered with respect to postal pension benefits. In addition, if Congress decides that USPS should be financially self- sustaining but makes no changes to improve USPS’s financial condition, USPS will be unable to address unfunded liabilities for postal retiree health and pension benefits, an inability that could eventually translate into higher costs for future postal ratepayers. Ultimately, if USPS’s expenses continue to exceed its revenues, USPS is likely to continue to miss required payments, reduce operations, or seek federal appropriations through the annual appropriations process to cover its operating costs. If Congress determines that USPS should no longer be expected to be financially self-sustaining or if actions taken do not restore financial self- sustainability, Congress could provide financial assistance—not unlike what happened in other countries or for selected domestic business—to enable USPS to cover its costs, and to fulfill its obligation to provide federal health and pension benefits to postal employees and retirees. Federal financial assistance could be provided in various forms, such as: Appropriating funds to help cover USPS’s operating costs, essentially the same arrangement that was used to finance the former U.S. Post Office Department. Appropriating funds to supplement USPS’s payment of certain costs, such as to help fund its capital investments. For example, the federal government provides Amtrak, which is operated as a for-profit corporation with annual grants to operate and make capital investments in passenger rail service to supplement the revenues it generates. Assuming some or all of USPS’s unfunded liabilities for retiree health benefits. This could take different forms, such as direct assumption of responsibility for unfunded liabilities or, more indirectly, requiring postal retirees to participate in Medicare which would decrease USPS’s costs but increase Medicare’s costs. Assuming some or all of USPS’s unfunded liabilities for pension benefits. Writing off some or all of USPS’s debts to the U.S. Treasury. Options regarding the federal government providing ongoing financial assistance to USPS could have effects on both USPS and the federal government as a whole. Notably, this assistance would have to be funded in some way—either through offsetting reductions in federal expenditures in other areas, through tax increases, or through an increase in federal deficits. Moreover, reliance on federal funding could mean that USPS would be exposed to the uncertainty inherent in the annual appropriations process. In addition, access to annual appropriations to cover financial shortfalls could have an unintended consequence of reducing USPS’s incentives to become more cost-efficient. At present, there is no consensus on USPS’s level of financial self- sustainability should be. For example, representatives of labor unions we spoke with stated that Congress should address issues regarding postal retiree benefits before any reassessment of USPS’s financial self- sufficiency can occur. Increased federal financial support of USPS might also face political opposition, due to concerns about minimizing federal deficits and ensuring fair competition between USPS and the private sector. Institutional Structure The final area of consideration in any reassessment of USPS’s business model is identifying what institutional structure could best deliver the level of postal services at the level of financial sustainability that Congress has determined. As an independent establishment of the executive branch, USPS must provide universal postal service while being expected to be financially self-sustaining. Thus, there may be a tension between attempting to fulfill public service missions while operating in an efficient, business-like and financially self-sustaining manner. USPS officials told us that as an entity of the federal government, its primary purpose is the achievement of its statutory universal service mission, and it has no incentive to seek to maximize profits at the expense of achieving its public service mission over the long term. Therefore, according to USPS, if it were maintained as an independent establishment of the executive branch or converted into a more typical government agency, it could continue to prioritize this public service mission. Additionally, there is widespread support for USPS’s institutional status as an independent establishment of the executive branch. Congressional resolutions have been introduced stating that “Congress should take all appropriate measures to ensure that the United States Postal Service remains an independent establishment of the Federal Government and is not subject to privatization.” Likewise, all four of the largest USPS unions, both of its management organizations, and a number of mailer groups and mailers support keeping USPS an independent establishment of the executive branch. Nonetheless, considering the depth of USPS’s financial problems and its poor financial outlook, now may be an appropriate time for Congress to reconsider what institutional structure will be most appropriate for USPS in the 21st century. However, any substantial change to USPS’s institutional status would require changing federal law. Based on our past work and options identified by USPS and others, Congress has a range of options it could consider in reassessing USPS’s structure (see table 1). The potential advantages and disadvantages of placing USPS into alternative institutional structures for USPS have long been debated. Several options have been discussed: USPS could revert to a traditional federal agency. USPS and its governance would be more consistent with other federal activities that are dependent on federal appropriations provided through the annual appropriations process. Many postal stakeholders, however, do not support such a change. For example, USPS told us that if it became a typical government agency reliant on federal appropriations to fill any operating gap, the political constraints that typically apply to government agencies could reduce USPS’s adaptability. Furthermore, changing USPS to a typical government agency could reduce its incentives to increase revenues or reduce costs in response to changing communication technologies and patterns. Consistent with this point, the 1968 presidential commission found that when it operated as a federal agency, the former U.S. Post Office Department had a lack of innovation, cost-control, and capital investment with major managerial decisions made through the legislative process. These and other issues led to persistent operational deficits, low productivity, and poor mail service. USPS could remain an independent establishment of the federal government with additional authority—relative to the status quo—over certain aspects of its business model. For example, USPS could be provided more flexibility to raise postal rates, introduce new nonpostal products, and make various changes to reduce its costs such as reducing the frequency of delivery or further consolidating its retail, transportation, and processing networks. USPS has long advocated for additional flexibility under its current institutional structure—such as to eliminate the price cap on market-dominant products and have greater flexibility to offer nonpostal products. Representatives from postal unions also stated that USPS should be provided additional flexibility, such as to expand into nonpostal products, which some representatives stated could help preserve its public service mission to provide universal postal services. Consensus does not, however, exist as to what flexibility should be given to USPS. For example, some mailer groups favor keeping the price cap unchanged, stating the cap is sufficient and provides incentives for increased efficiency. In addition, some stakeholders have supported further limiting USPS’s flexibility to reduce service standards, close retail outlets, or consolidate processing facilities, while other stakeholders noted that greater flexibilities in these areas would reduce USPS’s costs and enhance its efficiency. If USPS were to become a government-owned corporation or a government-sponsored enterprise, USPS could be incentivized to increase efficiency as a federally chartered entity providing a public service with a predominantly business nature. For example, three of the four third-party experts we spoke with stated that USPS should retain its current mission of universal postal service but become more like a private company with greater freedom to operate in a business- like manner. The new structure could promote greater incentives toward cost control and financial success. Government-owned corporations are federally chartered entities that provide a public service with a predominantly business nature. These corporations can have a board of directors that is appointed by the President. Government-sponsored enterprises are federally chartered entities that are privately owned and, typically, have a board of directors appointed by private sector owners. If USPS were to be a private company it would become accountable to the shareholders of that company. USPS told us that as a private company, its primary incentive could be to maximize profits and that, in USPS’s view, private shareholders would be most focused on short-term financial outcomes. Thus, any such design of a private USPS would need to balance its profit motive with the nation’s needs for universal postal service and the affordability of that service. For example, while Royal Mail is a private corporation owned by shareholders, the government of the United Kingdom still mandates 6- day delivery for letter mail (and 5-day delivery for packages) with specified delivery standards, and some mail types are subject to price controls. In addition, the United Kingdom monitors the provision of universal postal service and can take enforcement actions regarding regulatory conditions and competition law. Similarly, Germany has legal instruments to enforce the provision of universal postal services, although according to German government officials these instruments have not been used. Although some of the domestic businesses we examined reduced their costs through bankruptcy, this is likely not an option for USPS. As detailed in its report (see appendix I), National Bankruptcy Conference (NBC) found that USPS is not eligible to become a “debtor” under chapters 11 or 9 of the current Bankruptcy Code. According to NBC, a court likely would deem USPS to be a “governmental unit”—meaning it could not file for relief under chapter 11—and a court would deem USPS not to be a “municipality”—meaning it could not file for relief under chapter 9. Therefore, legislation amending the Code would be required to make USPS eligible for relief. According to NBC, however, even if the Bankruptcy Code were amended to allow USPS to file as a chapter 11 or 9 debtor, the Code would still not currently authorize a bankruptcy court to discharge the ongoing statutory obligations that have led to USPS’s current financial situation, and amending the Code to authorize such court action could raise constitutional (separation of powers) concerns. Moreover, NBC noted the bankruptcy process is designed to address obligations that have already accrued, not to override or amend statutes that apply to a debtor’s post- bankruptcy operations and obligations. In NBC’s opinion, because USPS’s pension and health care obligations are imposed by statute instead of by contract as in most bankruptcy reorganization proceedings, the bankruptcy process is not an effective or appropriate mechanism to address USPS’s obligations or potential transformation. NBC thus concluded that “although the bankruptcy process and bankruptcy tools raise interesting ideas for restructuring USPS’s existing and future obligations…all roads for doing so lead back to Congress.” Implementation Considerations Any changes that Congress makes to USPS’s business model will take time to implement and will need to be reevaluated as market conditions evolve. We have reported that fully implementing major transformations of government agencies can take years, and we also found that to be the case for the selected domestic businesses and foreign posts noted in this report, regardless of the changes needed. For example, railroads in the Northeast, airlines, and automakers took many years to implement a series of changes to their businesses. It took Germany more than a decade to fully liberalize and then privatize its postal operator, and the United Kingdom’s effort to privatize its postal operator took about 5 years. All of these organizations continue to adapt as they address ongoing challenges in a changing and highly competitive business environment. For example, GM recently stated that years after exiting bankruptcy and restoring profitability, it is closing some factories and focusing on developing electric and self-driving cars. Several freight railroads facing a downturn in freight traffic have also decided to run longer trains less frequently to reduce labor costs and increase efficiencies. Similarly, changes in the use of postal services will continue for the foreseeable future, necessitating continued adaptation. Some of the countries we selected are anticipating the need to be prepared for possible future changes. For example: In August 2019, German government officials said they would consider reducing postal delivery frequency from 6 to 5 days a week as part of an ongoing review to adapt Germany’s 20-year old postal law to changing market conditions and customer demands. In the United Kingdom, the postal regulator assessed postal users’ needs in 2020 in light of the changes in the postal market and to prepare for its regulatory review, which is to be concluded by 2022. A 2018 consultant’s report to the European Union (EU) recommended that the EU relax its universal service obligations to accommodate future changes in the postal market. The EU is currently studying how postal users’ needs are changing to determine if it needs to change its framework to allow member states to change their definitions of universal service obligations. In November 2019, the Australian government ordered a review of Australia Post’s long-term strategy to operate as a sustainable postal service provider, considering market conditions such as e-commerce, the regulatory environment and changes in business and consumer service needs. The government of New Zealand is scheduled to revise its memorandum of understanding with New Zealand Post defining universal service obligations by 2021. GAO’s Calls for Congressional Action to Address USPS’s Solvency Remain Unaddressed PAEA required GAO to evaluate strategies and options for the long-term structural and operational reform of USPS by December 2011. As USPS continued to face financial challenges, we accelerated this evaluation, which we issued in April 2010. However, we found that USPS’s business model, which was to provide universal postal service through self-supporting, business-like operations as an independent establishment of the executive branch, was not viable due to USPS’s inability to reduce costs sufficiently to respond to continuing declines in mail volume and revenue. In particular, we identified strategies to reduce compensation and benefit costs, reduce other operations and network costs, improve efficiency, and generate revenues through product and pricing flexibility. We also stated that while USPS may be able to improve its financial viability if it took more aggressive action to reduce costs, it was unlikely that those actions alone would fully resolve USPS’s problems unless Congress also took action. Therefore, we stated that Congress should consider, among other things, any and all options available to reduce USPS’s costs. While bills on these issues were introduced and in some cases passed congressional committees, postal reform legislation to address these considerations has not been enacted. In addition, in our most recent update to our High Risk List in 2019, we reiterated the basic elements of our 2010 matter for congressional consideration by stating that Congress should consider various options to better align USPS’s costs with its revenues. We stated that Congress should consider addressing constraints and legal restrictions that limit USPS’s ability to reduce costs and improve efficiency through considering a comprehensive package of legislative actions. To date, such a legislative package has not been enacted. Furthermore, we reported in 2018 that the financial outlook for the Postal Service’s Retiree Health Benefits Fund was poor, as USPS had not made any payments into it since 2010. OPM then forecasted the fund would be depleted by 2030 if USPS continued to not make payments. Therefore, we stated that Congress should consider passing legislation to put postal retiree health benefits on a more sustainable financial footing. However, legislation has not yet been enacted to address this issue. Conclusions We have often reported over the past 10 years that USPS’s ability to take actions taken under its current authority is insufficient to fully address its financial situation. Absent congressional action on critical foundational elements of the USPS business model, USPS’s mission and financial solvency are increasingly in peril. USPS’s growing difficulties to provide universal postal service in a financially self-sustaining matter provide Congress with the need to consider fundamental reform of the entire framework of postal services in the United States. In so doing, we continue to believe that as we stated in 2010, Congress should consider any and all available options. Comprehensive postal reform has not taken place in part because of the difficulty in obtaining compromise among various stakeholders with divergent views. Comprehensive, effective, and successful reform cannot occur until there is leadership and clarity around: what services should be provided, whether USPS is to be fully financially self-sustaining or the extent of federal financial support, and what institutional structure best supports these changes. Congressional leadership is critical in transforming USPS because consensus on policy decisions involving value judgments, trade-offs, and effects on postal stakeholders will be difficult to achieve. In addressing these issues, while all stakeholders’ interests should be understood and taken into consideration, the fundamental needs of the nation must take precedence. Continued inaction will result in deepening financial problems—putting USPS’s mission to provide universal postal service at greater risk and minimizing the ability to make the most appropriate or sustainable policy decisions. Matters for Congressional Consideration We are making the following three matters for congressional consideration: Congress should consider reassessing and determining the level of universal postal service the nation requires. (Matter for Consideration 1) Congress should consider determining the extent to which USPS should be financially self-sustaining and what changes to law would be appropriate to enable USPS to meet this goal. (Matter for Consideration 2) Congress should consider determining the most appropriate institutional structure for USPS. (Matter for Consideration 3) Agency Comments and Our Evaluation We provided a draft of this report to USPS and PRC. USPS and PRC provided written responses which are reproduced in appendixes IV and V, respectively. In its response, USPS concurred with our first two matters to reassess and determine the level of universal postal services the nation requires and to determine the extent to which USPS should be financially self- sustaining. USPS noted that the recent COVID-19 pandemic has both highlighted USPS’s essential role in the nation’s infrastructure and has caused a significant and sudden decline in mail volume, leading to a short-term liquidity crisis. USPS stated that while action by Congress is critical to ensure its ability to operate in the short-term, its financial situation has long been unsustainable due to statutory and regulatory structures that limit their ability to increase revenues and decrease costs. USPS noted that these changes require Congress to adopt reforms to secure USPS’s long-term financial viability. In addition, USPS concurred with the National Bankruptcy Conference’s legal analysis that Federal bankruptcy laws do not apply to USPS and that all roads for USPS restructuring lead back to Congress. USPS generally agreed with our third matter, stating that determining the institutional structure could logically be a part of a comprehensive congressional examination of its business model. USPS stated that it does not believe that corporatization or privatization would unlock new efficiency potential in USPS and that sustainable postal service does not hinge on the provider’s institutional form. However, as we and USPS have stated, its current legal and regulatory structure does not provide flexibility in some key areas. While our report states that a corporate or privatized institutional structure could provide both the flexibility and a greater incentive to operate in a more business-like manner than USPS’s current structure, we also recognize there are advantages and disadvantages to any institutional structure. As a result, we are not recommending any particular institutional structure for USPS, but are urging that Congress identify what institutional structure could best deliver the level of postal services at the level of financial sustainability that Congress determines. In its response, PRC agreed with all of our matters for congressional consideration. Particularly, PRC noted that the matter to reassess and determine the level of universal postal service the nation requires must be addressed as soon as possible. The PRC noted that given USPS’s severe and worsening financial situation (even before the impacts of the current pandemic crisis), a clear and specific definition of universal postal service and how that obligation can be funded must be provided. The PRC stated that Congress may want to consider mandating that PRC define and update the universal service definition by regulation. Both USPS and PRC provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Postmaster General, the Chairman of PRC, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2834 or rectanusl@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 making key contributions to this report are listed in appendix VI. Appendix I: Reprint of National Bankruptcy Conference’s Report on U.S. Postal Service Bankruptcy Issues Appendix II: Selected Legal Requirements Applicable to the U.S. Postal Service (USPS) Appendix II: Selected Legal Requirements Applicable to the U.S. Postal Service (USPS) Mail delivery quality and frequency 39 U.S.C. § 101(e),(f); see, e.g., Pub. L. 116-6, 133 Stat. 180 (2019) Legal requirements USPS is required to provide prompt, reliable, and efficient services to patrons in all areas, render postal services to all communities, and serve as nearly as practicable the entire population of the United States. USPS is specifically required to receive, transmit, and deliver written and printed matter, parcels, and like matter throughout the United States, its territories and possessions, and pursuant to certain agreements, throughout the world. USPS is required to provide a maximum degree of effective and regular postal services to rural areas, communities, and small towns where post offices are not self-sustaining. No small post office can be closed solely for operating at a deficit, it being the specific intent of the Congress that effective postal services be insured to residents of both urban and rural communities. Statutory and regulatory requirements specify the process and criteria for post office closings, including appellate review by the Postal Regulatory Commission (PRC). In determining all policies for postal services, USPS is required to give the highest consideration to the requirement for the most expeditious collection, transportation, and delivery of important letter mail. In selecting modes of transportation, USPS is required to give the highest consideration to the prompt and economical delivery of all mail. Service standards 39 U.S.C. § 3691, 39 C.F.R. Pt. 121 Letter mail monopoly 18 U.S.C. §§ 1693-99; 39 U.S.C. §§ 601-06 Mailbox monopoly 18 U.S.C. § 1725 Collective bargaining 39 U.S.C. §§1004,1206-07 Benefit programs 39 U.S.C. § 1005; 5 U.S.C. §§ 8348(h), 8423, 8909a Level of benefits 39 U.S.C. § 1005(f) Comparability 39 U.S.C. §§ 101(c), 1003(a) Workers’ compensation 39 U.S.C. § 1005(c) For many years, provisions in annual appropriations acts have stated “hat 6-day delivery and rural delivery of mail shall continue at not less than the 1983 level.” USPS is required to establish modern service standards for each market-dominant product (e.g., delivery of First-Class Mail within the continental United States in 2-3 delivery days); these service standards are defined in the Code of Federal Regulations. USPS’s letter delivery monopoly is codified in criminal and civil laws known as the Private Express Statutes. These laws generally prohibit anyone from establishing, operating, or using a private company to carry letters for compensation on regular trips or at stated periods over postal routes or between places where mail regularly is carried. Restricts access to mailboxes by prohibiting anyone from knowingly and willingly placing mailable matter without postage in any mailbox, providing USPS exclusive access to mailboxes. USPS negotiates collective bargaining agreements with its labor unions. If the parties are unable to reach an agreement, binding arbitration by a third-party panel will ultimately be used to establish agreement. USPS is also required to consult with postal supervisory and managerial organizations concerning changes in pay, benefits, and other programs that affect their membership. USPS is required to participate in federal pension and health benefit programs, with specific provisions regarding the required level of USPS’s funding of these programs. For example, USPS is required to prefund both postal pension benefits and postal retiree health benefits, each with payments that fully cover USPS’s share of future benefit costs. The law requires USPS’s fringe benefits to be at least as favorable as those in effect when the Postal Reorganization Act of 1970 was enacted, unless variation of benefits is collectively bargained. Compensation for USPS officers and employees is required to be comparable to the rates and types of compensation paid in the private sector of the U.S. economy. USPS policy also is required to maintain compensation and benefits for all officers and employees on a standard of comparability to comparable levels of work in the private sector. USPS is required to participate in the federal workers’ compensation program, which covers postal and other federal employees and provides compensation to federal employees, as well as dependents, in the event of an employee’s death. Citation(s) Access to facilities 39 U.S.C. § 403(b) Legal requirements USPS is required to establish and maintain postal facilities of such character and in such locations, that postal patrons throughout the Nation will, consistent with reasonable economies of postal operations, have ready access to essential postal services. Generally, annual appropriations prohibit USPS from using funding to consolidate or close small rural or other small post offices. Appropriations restrictions See, e.g., Pub. L. 116-6, 133 Stat. 180 (2019) Processing/logistics facilities Pub. L. 109–435 § 302(c)(5), (2006), 120 Stat. 3219, codified at 39 U.S.C. § 3691 note Price cap 39 U.S.C. § 3622(d) The law requires USPS to provide public information and opportunities for public input and comment before closing or consolidating any mail processing or logistics facilities, and take comments into account when making a final decision. An inflation-based price cap generally limits rate increases for market-dominant products, including First-Class Mail, USPS Marketing Mail, Periodicals and Package Services such as Bound Printed Matter, Media Mail, and Library Mail. The PRC, an independent establishment of the executive branch, must review USPS proposals to change domestic postal rates and fees. Debt limits 39 U.S.C. § 2005 Restriction on nonpostal lines of business 39 U.S.C. §§ 404(e), 102(5) Investment of postal retiree funds 5 U.S.C. §§ 8348(c), 8909a(c) Whenever USPS proposes a change in the nature of postal services that will have an effect on a substantially nationwide basis, it must request an advisory opinion from the PRC on the proposal. USPS has the authority to borrow up to $15 billion from the U.S. Treasury. The annual net increase of obligations for capital improvements and defraying operating expenses is limited to $3 billion. USPS is limited to providing nonpostal services to those offered as of January 1, 2006 that PRC has authorized USPS to continue. Nonpostal service is defined to mean any service that is not a postal service. A postal service is defined as the delivery of letters, printed matter, or mailable packages, including acceptance, collection, sorting, transportation, or other function ancillary thereto. Funds set aside for postal pensions and retiree health benefits are required by law to be invested in U.S. Treasury securities. Appendix III: U.S. Postal Service Financial Information for Fiscal Years 1972 through 2019 Dollars in millions Fiscal year 1972 Appendix IV: Comments from the U.S. Postal Service Appendix V: Comments from the Postal Regulatory Commission Appendix VI: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments Lori Rectanus, (202) 512-2834 or rectanusl@gao.gov. In addition to the individual named above, Derrick Collins (Assistant Director); Greg Hanna (Analyst-in-Charge); Amy Abramowitz; Kenneth John; Hannah Laufe; Serena Lo; Michael Mgebroff; Joshua Ormond; Joshua Parr; Susan Sawtelle; Crystal Wesco; and Laurel Voloder made key contributions to this report.

Why GAO Did This Study An independent establishment of the executive branch, USPS is required to provide prompt, reliable, and efficient services to the public. While USPS is to be self-sustaining, it lost about $78 billion from fiscal years 2007 through 2019 due primarily to declining mail volumes and increased costs. Given USPS’s poor financial condition, in 2009 GAO identified USPS’s financial viability as a high-risk area, a designation it retains today. GAO was asked to explore issues related the transformation of USPS and potential implications for stakeholders. This report (1) examines major challenges facing USPS, (2) identifies how selected domestic businesses and foreign posts reportedly have addressed serious challenges, (3) examines critical foundational elements of USPS’s current business model, and (4) identifies key previously issued GAO matters for congressional consideration regarding USPS and actions taken in response. GAO reviewed its prior reports and related matters for congressional consideration, analyzed laws and regulations, and assessed USPS documents on financial and operational performance. It also reviewed reports by the USPS Office of Inspector General, the Postal Regulatory Commission, and other selected groups such as the 2018 Task Force on the United States Postal Service. To identify how domestic businesses and foreign posts addressed similar serious challenges, GAO selected for review (1) six domestic organizations in the airline, automobile, and railroad industries and (2) five foreign posts in five countries—Australia, France, Germany, New Zealand, and the United Kingdom. The businesses and countries had characteristics similar to USPS, such as large unionized work forces, and had reportedly made significant changes to their business models. For each of these businesses and countries, GAO analyzed public reports on financial and operational performance, as well as institutional structure and requirements. GAO also interviewed government and postal officials from three selected countries and officials from the National Audit Offices of two of the selected countries. Because questions were raised regarding the application of the U.S. Bankruptcy Code to USPS, GAO also requested the National Bankruptcy Conference to assess whether USPS could use bankruptcy or other restructuring processes. To examine critical USPS business model elements, GAO reviewed its prior reports and reports from numerous other organizations, and obtained the views of stakeholders. What GAO Found Since GAO's 2009 high-risk designation, the U.S. Postal Service's (USPS) financial viability has progressively worsened due to declining mail volume, increased employee compensation and benefit costs, and increased unfunded liabilities and debt. First-Class Mail volume has declined 44 percent since fiscal year 2006. Additionally, employee compensation and benefits costs have been increasing. Although USPS's work force declined from about 786,000 in fiscal year 2007 to about 617,000 in fiscal year 2013, USPS's work force increased to about 630,000 in fiscal year 2019. Finally, total unfunded liabilities and debt continue their steady upward trend (see figure). To address these challenges, USPS has taken a variety of actions such as providing increased self-service options and reducing facility hours. Statutory requirements, however, limit USPS's ability to make changes in areas such as certain service offerings, pricing, and its employee compensation and benefits. In confronting similar types of challenges that are facing USPS, GAO selected large domestic businesses (companies) and foreign postal entities (widely known as “foreign posts”) that have seen significant change in foundational elements of their business models. Specifically, according to GAO's analysis of publicly available reports and interviews of cognizant officials, these organizations have had major changes in services and products, financial self-sustainment, and institutional structure: Companies and foreign posts have modified services and products to focus on profitable offerings, and two countries’ posts reduced postal service levels. For example, New Zealand Post reduced its mail delivery’s frequency from 5 to 3 days per week in urban areas. Companies have reduced their workforce, infrastructure, and operational costs, and some accepted government financial assistance to help remain financially viable. Cost reduction has also been a priority for all countries’ posts, especially in compensation and benefits, while three countries’ governments provided financial assistance to their posts. Four of the selected companies declared bankruptcy leading to restructured corporations; some merged with other companies to increase their revenues. Two countries privatized their posts, and three others restructured their posts from government departments into government-owned corporations. Regarding USPS, reassessing its business model should start with the level of required postal services. For example, delivery is USPS’s most costly operation; USPS officials estimate annual savings of $1.4 billion to $1.8 billion if delivery of mail were reduced to 5 days rather than 6 days per week. Second, USPS is to function as a financially self-sustaining entity; however, it does not. A reassessment could include determining whether some of USPS’s costs and liabilities should be borne by taxpayers. Third, alternative institutional structures for USPS range from a federal agency to a private company. A bankruptcy proceeding is not an effective or appropriate means to address the issues associated with a potential USPS restructuring, according to the National Bankruptcy Conference. Prior GAO reports have included suggestions for Congress to address USPS’s financial viability. For example, GAO’s 2010 report identified strategies to reduce compensation, benefits, and operational costs. GAO stated that Congress, among other things, consider all options available to reduce costs. While bills in this area were introduced and in some cases passed congressional committees, legislation was not enacted. In 2018, GAO reported that the financial outlook for the Postal Service Retiree Health Benefits Fund was poor—the Office of Personnel Management forecasted the fund would be depleted by 2030 if USPS continued not making payments into it. Legislation has not been enacted to place postal retiree health benefits on a more sustainable financial footing. Postal reform legislation has not taken place in part because of the difficulty in obtaining compromise among various stakeholders with divergent views (see figure below). However, since GAO’s 2010 report, USPS’s financial condition has significantly worsened raising fundamental questions about key elements of USPS’s business model. Such questions warrant congressional action. What GAO Recommends Congress should consider reassessing and determining the (1) level of postal services the nation requires, (2) extent to which USPS should be financially self-sustaining, and (3) appropriate institutional structure for USPS. Both USPS and the Postal Regulatory Commission (PRC) generally concurred with the matters.

gao_GAO-19-392T

gao_GAO-19-392T_0

How We Rate High- Risk Areas Our experience has shown that the key elements needed to make progress in high-risk areas are top-level attention by the administration and agency leaders grounded in the five criteria for removal from the High-Risk List, as well as any needed congressional action. The five criteria for removal that we issued in November 2000 are as follows: Leadership commitment. Demonstrated strong commitment and top leadership support. Capacity. Agency has the capacity (i.e., people and resources) to resolve the risk(s). Action plan. A corrective action plan exists that defines the root cause, solutions, and provides for substantially completing corrective measures, including steps necessary to implement solutions we recommended. Monitoring. A program has been instituted to monitor and independently validate the effectiveness and sustainability of corrective measures. Demonstrated progress. Ability to demonstrate progress in implementing corrective measures and in resolving the high-risk area. Starting in our 2015 update, we added clarity and specificity to our assessments by rating each high-risk area’s progress on the five criteria and used the following definitions: Met. Actions have been taken that meet the criterion. There are no significant actions that need to be taken to further address this criterion. Partially met. Some, but not all, actions necessary to meet the criterion have been taken. Not met. Few, if any, actions towards meeting the criterion have been taken. Changes to the 2019 High-Risk List We are removing two areas—DOD Supply Chain Management and Mitigating Gaps in Weather Satellite Data—from the list due to the progress that was made in addressing the high-risk issues. As we have with areas previously removed from the High-Risk List, we will continue to monitor these areas to ensure that the improvements we have noted are sustained. If significant problems again arise, we will consider reapplying the high-risk designation. We added two areas to the High-Risk List since our 2017 update—Government-Wide Personnel Security Clearance Process and VA Acquisition Management. DOD Supply Chain Management Remo ved From the High-Risk List We are removing the area of DOD Supply Chain Management from the High-Risk List because, since 2017, DOD has addressed the remaining two criteria (monitoring and demonstrated progress) for the asset visibility and materiel distribution segments. Congressional attention, DOD leadership commitment, and our collaboration contributed to the successful outcome for this high-risk area, which had been on GAO’s High-Risk List since 1990. DOD’s actions for the asset visibility segment of this high-risk area included (1) providing guidance for the military components to consider key attributes of successful performance measures during metric development for their improvement initiatives; (2) incorporating into after- action reports, information relating to performance measures; and (3) demonstrating sustained progress by, for example, increasing its visibility of assets through radio-frequency identification (RFID), an automated data-capture technology that can be used to electronically identify, track, and store information contained on a tag. According to DOD, the use of RFID tags to provide visibility of sustainment cargo at the tactical leg (i.e., the last segment of the distribution system) resulted in $1.4 million annual cost savings. DOD’s actions for the materiel distribution segment of this high-risk area included (1) making progress in developing its suite of distribution performance metrics; (2) incorporating distribution metrics, as appropriate, on the performance of all legs of the distribution system, including the tactical leg; (3) making progress in refining its Materiel Distribution Improvement Plan and incorporating additional actions based on interim progress and results; and (4) improving its capability to comprehensively measure distribution performance, identifying distribution problems and root cause, and implementing solutions. According to DOD, initiatives focused on distribution process and operational improvements have resulted in at least $1.56 billion in distribution cost avoidances to date. As we have with areas previously removed from the High-Risk List, we will continue to monitor this area to ensure that the improvements we have noted are sustained. Appendix I provides additional information on this high-risk area. Mitigating Gaps in Weather Satellite Data Removed From the High- Risk List We are removing the area of Mitigating Gaps in Weather Satellite Data from the High-Risk List because—with strong congressional support and oversight—the National Oceanic and Atmospheric Administration (NOAA) and DOD have made significant progress since 2017 in establishing and implementing plans to mitigate potential gaps in weather satellite data. The United States relies on polar-orbiting satellites to provide a global perspective on weather every morning and afternoon. NOAA is responsible for the polar satellite program that crosses the equator in the afternoon while DOD is responsible for the polar satellite program that crosses the equator in the early morning orbit. NOAA’s actions for polar- orbiting weather satellites that addressed the remaining criteria of action plan and demonstrated progress included (1) issuing three updates to its gap mitigation plan between January 2016 and February 2017 to address shortfalls we had identified previously; and (2) successfully launching the NOAA-20 satellite in November 2017, which is currently operational and is being used to provide advanced weather data and forecasts. Moreover, NOAA is also working to build and launch the next satellites in the polar satellite program. DOD’s actions for polar-orbiting weather satellites, pursuant to statutes and accompanying congressional direction, included DOD leadership (1) developing and implementing plans to acquire satellites as part of a family of systems to replace its aging legacy weather satellites, including awarding a contract for its Weather System Follow-on–Microwave program, planned for launch in 2022; (2) establishing plans to meet its highest-priority weather monitoring data collection needs that will not be covered by the Weather System Follow-on–Microwave program, including by acquiring and launching the Electro-Optical/Infrared Weather Systems satellite in 2024; and (3) monitoring the Weather System Follow-on- Microwave satellite program’s progress toward addressing critical needs and assessing its operations and sustainment costs. As we have with areas previously removed from the High-Risk List, we will continue to monitor this area to ensure that the improvements we have noted are sustained. Appendix I provides additional information on this high-risk area. Government-wide Personnel Security Clearance Process Added to the High-Risk List Executive branch agencies are not meeting investigation timeliness objectives, and these processing delays have contributed to a significant backlog that the National Background Investigations Bureau (NBIB)—the agency responsible for personnel security clearance investigations— reported to be approximately 565,000 investigations as of February 2019. In addition, the executive branch has not finalized performance measures to ensure the quality of background investigations and some long- standing key reform initiatives remain incomplete. Further, information technology (IT) security concerns may delay planned milestones for the development of a new background investigation IT system. We included the DOD program on our High-Risk List in 2005 and removed it in 2011 because of improvements in the timeliness of investigations and adjudications, and steps toward measuring the quality of the process. We put the government-wide personnel security clearance process on our High-Risk List in January 2018 because of significant challenges related to the timely processing of security clearances and completing the development of quality measures. In addition, the government’s effort to reform the personnel security clearance process, starting with the enactment of the Intelligence Reform and Terrorism Prevention Act of 2004, has had mixed progress, and key reform efforts have not been implemented government-wide. Since adding this area to the High-Risk List, the Security Clearance, Suitability, and Credentialing Performance Accountability Council (PAC), including its four principal members—the Deputy Director for Management of the Office of Management and Budget (OMB), the Director of National Intelligence (DNI); the Under Secretary of Defense for Intelligence; and the Director of the Office of Personnel Management (OPM)—have not fully met the five criteria for high-risk removal. Several issues contribute to the risks facing the government-wide personnel security clearance process: Clearance processing delays. Executive branch agencies are not meeting most investigation timeliness objectives. The percentage of executive branch agencies meeting established timeliness objectives for initial secret clearances, initial top secret clearances, and periodic reinvestigations decreased each year from fiscal years 2012 through 2018. For example, 97 percent of the executive branch agencies we reviewed did not meet the timeliness objectives for initial secret clearance investigations in fiscal year 2018. Lack of quality measures. While the executive branch has taken steps to establish government-wide performance measures for the quality of background investigations—including establishing quality assessment standards and a quality assessment reporting tool—it is unclear when this effort will be completed. Security clearance reform delays. The executive branch has reformed many parts of the personnel security clearance process— such as updating adjudicative guidelines to establish common adjudicative criteria for security clearances; however, some long- standing key initiatives remain incomplete—such as completing plans to fully implement and monitor continuous evaluation. IT security. DOD is responsible for developing a new system to support background investigation processes, and DOD officials expressed concerns about the security of connecting to OPM’s legacy systems since a 2015 data breach compromised OPM’s background investigation systems and files for 21.5 million individuals. As of December 2018, OPM has not fully taken action on our priority recommendations to update its security plans, evaluate its security control assessments, and implement additional training opportunities. However, since we added this area to our High-Risk List, the PAC has demonstrated progress in some areas. For example, NBIB reported that the backlog of background investigations decreased from almost 715,000 cases in January 2018 to approximately 565,000 cases in February 2019. NBIB officials credit an Executive Memorandum—issued jointly in June 2018 by the DNI and the Director of OPM and containing measures to reduce the investigation backlog—as a driver in backlog reduction. Further, in response to a requirement in the Securely Expediting Clearances Through Reporting Transparency (SECRET) Act of 2018, in September 2018, NBIB reported to Congress, for each clearance level, (1) the size of the investigation backlog, (2) the average length of time to conduct an initial investigation and a periodic reinvestigation, and (3) a discussion of the factors contributing to investigation timeliness. The PAC is also reporting publicly on the progress of key reforms through www.performance.gov, and for fiscal year 2018, the website contains quarterly action plans and progress updates, which present figures on the average timeliness of initial investigations and periodic reinvestigations for the executive branch as a whole, investigation workload and backlog, and investigator headcounts. We have made numerous recommendations to PAC members to address risks associated with the personnel security clearance process between 2011—when we removed DOD’s personnel security clearance program from the High-Risk List, and 2018—when we placed the government-wide personnel security clearance process on the High-Risk List. We consider 27 of these recommendations key to addressing the high-risk designation. Eight recommendations key to the high-risk designation have been implemented, including three since January 2018. Nineteen of these key recommendations remain open—including recommendations that the principal members of the PAC (1) conduct an evidence-based review of investigation and adjudication timeliness objectives, (2) develop and report to Congress on investigation quality measures, (3) prioritize the timely completion of efforts to modernize and secure IT systems that affect clearance holders government-wide, and (4) develop and implement a comprehensive workforce plan that identifies the workforce needed to meet current and future demand for background investigations services and to reduce the investigations backlog. See page 170 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. VA Acquisition Management Added to the High-Risk List VA spends tens of billions of dollars to procure a wide range of goods and services—including medical supplies, IT, and construction of hospitals, clinics, and other facilities—to meet its mission of providing health care and other benefits to millions of veterans. VA has one of the most significant acquisition functions in the federal government, both in obligations and number of contract actions. The Veterans Health Administration (VHA) provides medical care to veterans and is by far the largest administration in the VA. Since we began focusing on VA’s acquisition management activities in 2015, we have reported numerous challenges in this area. Since 2015, we have made 31 recommendations, 21 of which remain open, that cover a range of areas to address challenges in VA’s acquisition management. In fiscal year 2019, VA received the largest discretionary budget in its history—$86.5 billion, about $20 billion higher than in 2015. About a third of VA’s discretionary budget in fiscal year 2017, or $26 billion, has been used to contract for goods and services. VA’s acquisition management continues to face challenges including (1) outdated acquisition regulations and policies; (2) lack of an effective medical supplies procurement strategy; (3) inadequate acquisition training; (4) contracting officer workload challenges; (5) lack of reliable data systems; (6) limited contract oversight and incomplete contract file documentation; and (7) leadership instability. In light of these challenges and given the significant taxpayer investment, it is imperative that VA show sustained leadership commitment to take steps to improve the performance of its procurement function so that it can use its funding in the most efficient manner possible to meet the needs of those who served our country. This area has been added to the High-Risk List for the following reasons in particular: Outdated acquisition regulations and policies. VA’s procurement policies have historically been outdated, disjointed, and difficult for contracting officers to use. In September 2016, we reported that the acquisition regulations contracting officers currently follow have not been fully updated since 2008 and that VA had been working on completing a comprehensive revision of its acquisition regulations since 2011. VA’s delay in updating this fundamental source of policy has impeded the ability of contracting officers to effectively carry out their duties. We recommended in September 2016 that VA identify measures to expedite the revision of its acquisition regulations and clarify what policies are currently in effect. VA concurred with this recommendation but has not yet fully implemented it. Lack of an effective medical supplies procurement strategy. VA’s Medical Surgical Prime Vendor-Next Generation (MSPV-NG) program for purchasing medical supplies to meet the needs of about 9 million veterans at 172 medical centers has not been effectively executed, nor is it in line with practices at leading hospitals that have launched similar programs. We reported in November 2017 that VA’s approach to developing its catalog of supplies was rushed and lacked key stakeholder involvement and buy-in. As a result, VA was not able to accomplish some of the key efficiencies the program was intended to achieve, such as streamlining the purchase of medical supplies and saving money. We recommended in November 2017 that VA develop, document, and communicate to stakeholders an overarching strategy for the program. VA concurred with this recommendation and reported that it would develop a new strategy by March 2019. Contracting officer workload challenges. The majority of our reviews since 2015 have highlighted workload as a contributing factor to the challenges that contracting officers face. Most recently, in September 2018, we reported that about 54 percent of surveyed VA contracting officers said their workload was not reasonable. In addition, in September 2016, we reported that VHA contracting officers processed a large number of emergency procurements of routine medical supplies, which accounted for approximately 20 percent of VHA’s overall contract actions in fiscal year 2016, with obligations totaling about $1.9 billion. Contracting officers told us that these frequent and urgent small-dollar transactions reduce contracting officers’ efficiency and ability to take a strategic view of procurement needs. We recommended in November 2017 that VHA network contracting offices work with medical centers to identify opportunities to more strategically purchase goods and services frequently purchased on an emergency basis. VA concurred with this recommendation and reported in December 2018 that it is utilizing a supply chain dashboard to track items purchased on an emergency basis and determine which of those items to include on the catalog. VA noted that it added 13,300 items to the catalog from June 2018 to December 2018, including items often purchased on an emergency basis. We requested documentation showing which items added to the catalog were previously purchased on an emergency basis, but as of January 2019, VA had not yet provided it. Among other things, VA should implement our 21 open recommendations and specifically needs to take the following steps to demonstrate greater leadership commitment and strategic planning to ensure efficient use of its acquisition funding and staffing resources: Prioritize completing the revision of its acquisition regulations, which has been in process since 2011. Develop, document, and communicate to stakeholders a strategy for the Medical Surgical Prime Vendor program to achieve overall program goals. Identify opportunities to strategically purchase goods and services that are frequently purchased on an emergency basis. See page 210 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Emerging Issue Requiring Close Attention: Federal Efforts to Prevent Drug Misuse In addition to specific areas that we have designated as high risk, other important challenges facing our nation merit continuing close attention. One of these is the use of illicit drugs and the misuse of prescription drugs and the ways they affect individuals, their families, and the communities in which they live. Over 70,000 people died from drug overdoses in 2017—about 191 people every day—according to the Centers for Disease Control and Prevention, with the largest portion of these deaths attributed to opioids. Further, drug overdoses are the leading cause of death due to injuries in the United States. They are currently at their highest ever recorded level and, since 2011, have outnumbered deaths by firearms, motor vehicle crashes, suicide, and homicide, according to the Drug Enforcement Administration. The Council of Economic Advisors estimates that in 2015, the economic cost of the opioid crisis alone was more than $500 billion when considering the value of lives lost due to opioid-related overdose. Federal drug control efforts spanning prevention, treatment, interdiction, international operations, and law enforcement represent a considerable federal investment. According to the President’s fiscal year 2019 budget, federal drug control funding for fiscal year 2017 was $28.8 billion. Multiple federal agencies have ongoing efforts to respond to this crisis, including efforts to reduce the supply and demand for illicit drugs, to prevent misuse of prescription drugs, and to treat substance use disorders. However, we previously found that many efforts lacked measures to gauge the success of the federal response. Further, we have long advocated an approach to decision-making based on risk management. Such an approach would (1) link agencies’ plans and budgets to achieving their strategic goals, (2) assess values and risks of various courses of actions to help set priorities and allocate resources, and (3) provide for the use of performance measures to assess progress. The Office of National Drug Control Policy (ONDCP) is responsible for overseeing and coordinating the implementation of U.S. drug policy, including developing the National Drug Control Strategy (Strategy). ONDCP released the 2019 Strategy on January 31, 2019. The Strategy focuses on approaches related to prevention, treatment and recovery, and steps to reduce the availability of illicit drugs in the United States. We will continue to monitor the extent to which ONDCP and other federal agencies are employing a risk management and coordinated approach to their efforts to limit drug misuse. In particular, we have ongoing and planned work to assess ONDCP’s operations, including its (1) leadership and coordination of efforts across the federal government; (2) the effects of the drug crisis on labor force participation and productivity and on people with disabilities and other vulnerable populations; (3) key federal efforts to reduce the availability of illicit drugs; and (4) agency efforts around drug education and prevention. We will determine whether this issue should be added to the High-Risk List once we have completed this ongoing and planned work. High-Risk Areas That Made Progress Agencies can show progress by addressing our five criteria for removal from the list: leadership commitment, capacity, action plan, monitoring, and demonstrated progress. As shown in table 1, 24 high-risk areas, or about two-thirds of all the areas, have met or partially met all five criteria for removal from our High-Risk List; 20 of these areas fully met at least one criterion. Compared with our last assessment, 7 high-risk areas showed progress in one or more of the five criteria without regressing in any of the criteria. Ten high-risk areas have neither met nor partially met one or more criteria. Two areas showed mixed progress by increasing in at least one criterion and also declining in at least one criterion. Three areas declined since 2017. These changes are indicated by the up and down arrows in table 1. Figure 1 shows that since our 2017 update, the most progress was made on the action plan criterion—four high-risk areas received higher ratings. We rated two areas lower on leadership commitment and two areas lower on monitoring. Leadership Attention Needed to Meet High-Risk Criteria Table 2 shows that 17 of the 34 high-risk areas we rated have met the leadership commitment criterion while two high-risk area ratings regressed on leadership commitment from met to partially met since our last report. Leadership commitment is the critical element for initiating and sustaining progress, and leaders provide needed support and accountability for managing risks. Leadership commitment is needed to make progress on the other four high-risk criteria. Table 2 shows that only three high-risk areas met the criterion for capacity, six met the criterion for action plan, and two met the criterion for demonstrated progress. One high-risk area—U.S. Government’s Environmental Liability—has partially met only one criterion since we added the area to our list in 2017 and the rest are not met. Progress in High-Risk Areas As noted, seven areas showed improvement in one or more criterion without regressing in any criteria. Two areas showed sufficient progress to be removed from the High-Risk List. The other five high-risk areas remaining on the 2019 list demonstrated improvement and are described below. Three of these five improving high-risk areas are the responsibility of the Department of Defense (DOD)—DOD Support Infrastructure Management, DOD Financial Management, and DOD Business Systems Modernization. The two other improving areas are Department of Energy’s (DOE's) Contract Management for the National Nuclear Security Administration and Office of Environmental Management, and Medicare Program & Improper Payments. DOD Support Infrastructure Management: DOD manages a portfolio of real property assets that, as of fiscal year 2017, reportedly included about 586,000 facilities—including barracks, maintenance depots, commissaries, and office buildings. The combined replacement value of this portfolio is almost $1.2 trillion and includes about 27 million acres of land at nearly 4,800 sites worldwide. This infrastructure is critical to maintaining military readiness, and the cost to build and maintain it represents a significant financial commitment. Since our 2017 High-Risk Report, DOD’s rating for two criteria—leadership commitment and action plan—improved from partially met to met. DOD has demonstrated leadership commitment by stating its commitment to addressing key recommendations we have made by, for example, (1) better forecasting the initial Base Realignment and Closure (BRAC) costs for military construction, IT, and relocating military personnel and equipment; (2) better aligning infrastructure to DOD force structure needs by, for example, improving the accuracy and sufficiency of its excess capacity estimates; and (3) pursuing an effort to consolidate and standardize leases, which includes analyzing whether it is feasible to relocate functions from commercial leased space to existing space on an installation, thereby reducing leases and better utilizing excess space. DOD has developed action plans to better identify excess infrastructure and thus be positioned to dispose of it. For example, in the 2017 High- Risk Report, we stated that DOD’s Real Property Efficiency Plan includes DOD’s goals for reducing the footprint of its real property inventory and metrics to gauge progress, to be implemented by the end of 2020. We also found in 2018 that DOD was achieving cost savings and cost avoidances as it had begun using intergovernmental support agreements between military installations and local governments to obtain installation services, such as waste removal, grounds maintenance, and stray animal control. As a result of these and other actions, DOD now meets the action plan criterion for this high-risk area. As of December 2018, 23 recommendations related to this high-risk area remain open. DOD continues to partially meet the criteria for capacity, monitoring, and demonstrated progress. See page 158 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. DOD Financial Management: Since our 2017 High-Risk Report, ratings for the DOD Financial Management high-risk area improved for the criteria of leadership commitment and monitoring. For the leadership commitment criterion, the high-risk area rating improved from partially met to met in 2019 due to several DOD leadership actions. For example, in 2018, DOD leadership met the goal of undergoing an agency-wide financial statement audit and established a process to remediate any audit findings—ultimately to improve the quality of financial information that is most valuable in managing the department’s day-to-day operations. In addition, according to a DOD official, audit remediation efforts have produced benefits in certain inventory processes that have led to operational improvements. DOD leadership demonstrated its commitment to making needed improvements by developing a database that tracks hundreds of findings and recommendations that came out of the audits. In addition, senior leadership has been meeting bimonthly with military services’ leadership for updates on the status of corrective action plans to address audit findings and recommendations, and the Under Secretary of Defense (Comptroller) has been meeting frequently with the Secretary of Defense to review the plans. These same DOD actions also led to the high-risk area’s rating for the criterion of monitoring to improve from not met to partially met. For example, the database mentioned above is intended to capture, prioritize, and assign responsibility for auditor findings and related corrective action plans, which are meant to be used to measure progress towards achieving a clean audit opinion. Further, DOD leadership has held frequent meetings to discuss the status of corrective action plans. In addition, DOD also established councils in certain areas (e.g., financial reporting) to review the status of audit remediation activities and challenges. All of these actions demonstrate an improvement in DOD’s monitoring activities for its financial management function. However, DOD’s efforts to improve its financial management continue to be impaired by long-standing issues—including its decentralized environment; cultural resistance to change; lack of skilled financial management staff; ineffective processes, systems, and controls; incomplete corrective action plans; and the need for more effective monitoring and reporting. DOD remains one of the few federal entities that cannot accurately account for and report on its spending or assets. As of December 2018, 53 recommendations for this high-risk area are open. The DOD Financial Management high-risk area continues to partially meet the capacity and action plan criteria and not meet the demonstrated progress criterion. See page 147 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. DOD Business Systems Modernization: DOD spends billions of dollars each year to acquire modernized systems, including systems that address key areas such as personnel, financial management, health care, and logistics. This high-risk area includes three critical challenges facing DOD: (1) improving business system acquisition management, (2) improving business system investment management, and (3) leveraging DOD’s federated business enterprise architecture. DOD’s capacity for modernizing its business systems has improved over time and, since our 2017 High-Risk Report, DOD’s overall rating for the criterion of action plan improved from not met to partially met in 2019. DOD established a plan for improving its federated business enterprise architecture (i.e., description of DOD’s current and future business environment and a plan for transitioning to the future environment). Specifically, the rating improved for DOD’s federated business enterprise architecture segment of the high-risk area because DOD’s assistant deputy chief management officer approved a business architecture improvement plan in January 2017. Since 2017, we have made 10 recommendations related to this high-risk issue. As of December 2018, 27 recommendations are open. The leadership, capacity, monitoring, and demonstrated progress criteria remain partially met as in 2017. See page 152 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. DOE's Contract Management for the National Nuclear Security Administration and Office of Environmental Management: DOE oversees a broad range of programs related to nuclear security, science, energy, and waste cleanup, among other areas. As the largest civilian contracting agency in the federal government, DOE relies primarily on contractors to carry out its programs. For instance, DOE spends about 90 percent of its annual budget on contracts and acquiring capital assets. In fiscal year 2018, DOE’s budget was $34.5 billion. The high-risk area focuses on contracts, as well as major projects—those with an estimated cost of $750 million or greater—managed by DOE’s National Nuclear Security Administration (NNSA) and Office of Environmental Management (EM). Since our 2017 High-Risk Report, DOE has made progress by improving from a not met to a partially met rating for the demonstrated progress criterion. Specifically, through its Office of Cost Estimating and Program Evaluation, NNSA has enhanced its capability to estimate costs and schedules, and to assess alternatives for programs and projects, among other things. NNSA also made progress by adopting best practices in several areas, such as those for estimating costs and schedules in nuclear weapons refurbishment activities and capital asset acquisitions. For example, we determined that DOE’s revised cost estimate of $17.2 billion to construct a Mixed Oxide Fuel Fabrication Facility to dispose of surplus, weapons-grade plutonium substantially met best practices— providing assurance that the estimated costs could be considered reliable. This finding contributed to DOE’s reevaluation of the project and ultimate termination, in October 2018, in favor of a potentially less costly disposal approach. Fifty-one of our recommendations were open as of December 2018; 15 recommendations were made since the last high-risk update in February 2017. DOE continues to meet the criterion of leadership commitment, partially meet the criteria for action plan and monitoring, and not meet the criterion for capacity. See page 217 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Medicare Program & Improper Payments: In calendar year 2017, Medicare, which is overseen by the Centers for Medicare & Medicaid Services (CMS), financed $702 billion worth of health services for approximately 58 million elderly and disabled beneficiaries. Medicare faces a significant risk with improper payments—payments that either were made in an incorrect amount or should not have been made at all— which reached an estimated $48 billion in fiscal year 2018. Since our 2017 High-Risk Report, estimated improper payment rates declined more than one percent across the Medicare program. In addition, CMS’ rating for the capacity criterion of the improper payments segment improved from partially met to met in 2019 due to several actions. First, the Center for Program Integrity’s (CPI) budget and resources have increased over time and the agency has established work groups and interagency collaborations to extend its capacity. For example, CMS allocated more staff to CPI after Congress provided additional funding. CPI’s full-time equivalent positions increased from 177 in 2011 to 419 in 2017. Additionally, in August 2017, we reported that CMS’s Fraud Prevention System, which analyzes claims to identify health care providers with suspect billing patterns, helped speed up certain fraud investigation processes. Further, the Healthcare Fraud Prevention Partnership helped improve information sharing among payers inside and outside of the government. Since 1990, when we added Medicare to our High-Risk List, we have made many recommendations related to the Medicare program, 28 of which were made since the last high-risk update in February 2017. As of December 2018, more than 80 recommendations remain open. CMS continues to meet the criterion of leadership commitment and to partially meet the remaining three criteria of action plan, monitoring, and demonstrated progress. See page 241 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Congressional Action Aided Progress on High- Risk Issues Congress enacted several laws since our last report in February 2017 to help make progress on high-risk issues. Table 3 lists selected examples of congressional actions taken on high-risk areas. Congressional oversight also plays a vital role in addressing high-risk issues. For example, at a May 2018 hearing, we testified that the Census Bureau’s (Bureau) cost estimate was not reliable, and that the actual cost could be higher than planned. Further, the Secretary of Commerce created a dedicated team to provide oversight and guidance to the Bureau on cost estimation. In addition to its instrumental role in supporting progress in individual high-risk areas, Congress also enacted the following statutes that, if implemented effectively, will help foster progress on high-risk issues government-wide: Fraud Reduction and Data Analytics Act of 2015 (FRDAA):FRDAA is intended to strengthen federal antifraud controls. OMFRDAA requires OMB to use our Fraud Risk Framework to create guidelines for federal agencies to identify and assess fraud risks, and then design and implement control activities to prevent, detect, and respond to fraud. Agencies, as part of their annual financial reports beginning in fiscal year 2017, are further required to report on their fraud risks and their implementation of fraud reduction strategies, which should help Congress monitor agencies’ progress in addressing and reducing fraud risks. To aid federal agencies in better analyzing fraud risks, FRDAA requires OMB to establish a working group tasked with developing a plan for creating an interagency library of data analytics and data sets to facilitate the detection of fraud and the recovery of improper payments. This working group and the library should help agencies coordinate their fraud detection efforts and improve their ability to use data analytics to monitor databases for potential improper payments. The billions of dollars in improper payments, some of which may be a result of fraud, are a central part of the Medicare Program, Medicaid Program, and Enforcement of Tax Laws (Earned Income Tax Credit) high-risk areas. We reported in 2018 that, among other things, OMB did not involve all agencies subject to the act as required by FRDAA or hold the required minimum number of working-group meetings in 2017. As shown in figure 2, a majority of the 72 agencies surveyed indicated a lack of involvement with and information from the working group as challenges in implementing FRDAA. We made three recommendations, including that OMB ensure the working group meets FRDAA’s requirements to involve all agencies that are subject to the act and ensure that mechanisms to share controls, best practices, and data-analytics techniques are in place. OMB did not concur with our recommendations. We continue to believe the recommendations are valid, as discussed in the 2018 report. IT Acquisition Reform, statutory provisions known as the Federal Information Technology Acquisition Reform Act (FITARA): FITARA, enacted in December 2014, was intended to improve how agencies acquire IT and better enable Congress to monitor agencies’ progress in reducing duplication and achieving cost savings. Since the enactment of these provisions, OMB and federal agencies have paid greater attention to IT acquisition and operation, resulting in improvements to the government-wide management of this significant annual investment. These efforts have been motivated in part by sustained congressional support for improving implementation of this law, as highlighted in agencies’ FITARA implementation scores issued biannually by the House Committee on Oversight and Reform. This continuing oversight has produced positive results. For example, in the committee’s December 2018 FITARA implementation scorecard, 18 of the 24 major federal agencies received the highest possible rating for their efforts to improve the management of software licenses, of which we have found there are thousands annually across the government. Seven months earlier, in the prior scorecard, only eight agencies had achieved this rating. Moreover, federal agencies have taken actions to address 106 of the 136 related recommendations that we have made in this area since 2014. FITARA includes specific requirements related to seven areas: the federal data center consolidation initiative, enhanced transparency and improved risk management, agency Chief Information Officer authority enhancements, portfolio review, expansion of training and use of IT acquisition cadres, government-wide software purchasing, and maximizing the benefit of the federal strategic sourcing initiative. In November 2017, Congress extended or removed the sunset dates of several of these statutory requirements that were originally to end in 2018 and 2019. While all of the 24 federal agencies covered by this law have developed FITARA implementation plans, the agencies need to effectively execute these plans. Successfully addressing FITARA requirements is central to making progress in Improving the Management of IT Acquisitions and Operations, which has been on our High-Risk List since 2015. Program Management Improvement Accountability Act (PMIAA): Enacted in December 2016, the act is intended to improve program and project management in certain larger federal agencies. Among other things, the act requires the Deputy Director for Management of OMB to adopt and oversee implementation of government-wide standards, policies, and guidelines for program and project management in executive agencies. The act also requires the Deputy Director to conduct portfolio reviews to address programs we identify as high-risk. It further creates a Program Management Policy Council to act as the principal interagency forum for improving practices related to program and project management. The council is to review programs identified as high-risk and make recommendations to the Deputy Director or designee. OMB has produced a general strategy for implementing the law through 2022 and met some initial milestones required by PMIAA. For example, in June 2018, OMB issued OMB Memorandum M- 18-19, which includes: (1) agency guidance for implementing PMIAA, (2) a five-year strategic outline for improving program and project management, and (3) initial program management standards and principles. Further, agencies have designated Program Management Improvement Officers to guide their implementation of PMIAA. According to OMB, it began implementing PMIAA’s requirement to conduct portfolio reviews on high-risk areas by requiring relevant agencies to provide several items for discussion during the 2018 Strategic Review meetings. These annual meetings are to consist primarily of a discussion of agency progress towards each of the strategic objectives outlined in their strategic plans, but also cover other management topics such as enterprise risk management and high-risk area progress. According to OMB documents, in advance of these meetings, OMB required agencies to provide a high-level summary of (1) any disagreements with our recommendations, (2) progress barriers, and (3) actions needed by OMB, other agencies, or Congress to help the agency achieve progress towards removal from our High-Risk List. OMB officials told us their 2018 Strategic Review meetings did not address each high-risk area but did address government-wide high-risk areas, such as cybersecurity, information technology, and strategic human capital as they related to the President’s Management Agenda. In the past, senior management officials from OMB, applicable agencies, and our agency have met to address areas where additional management attention could be beneficial to high-risk issues. These trilateral meetings, beginning in 2007 and pre- dating PMIAA’s 2016 enactment, have continued across administrations. However, OMB has organized only one of these high-risk meetings since the last high-risk update in 2017, on the Government-wide Personnel Security Clearance Process. In November 2018, OMB told us of plans to hold additional meetings on priority high-risk areas, including the 2020 Decennial Census, Strategic Human Capital Management, Ensuring the Cybersecurity of the Nation, National Aeronautics and Space Administration (NASA) Acquisition Management, and Managing Federal Real Property. Effective implementation of PMIAA provides an important opportunity to enhance progress on high-risk areas by focusing leadership attention through the portfolio reviews and trilateral meetings. Further, a number of high-risk areas have longstanding or significant program and project management concerns, including the acquisition-related high-risk areas for DOD, DOE, NASA, and VA. These and other programs can benefit from improving program and project management. In December 2019, we will report on OMB’s progress in implementing PMIAA, including what further steps it has taken to use the portfolio review process required in PMIAA to address issues on our High-Risk List. Executive Branch Action on Our Recommendations Aided Progress on High- Risk Issues Agency leaders took actions to implement our recommendations. These resulted in numerous improvements to programs and operation and improved service. Further, these actions to implement our recommendations resulted in significant financial benefits. Table 4 shows some examples of the financial benefits achieved since our last High-Risk Report. High-Risk Areas Needing Significant Attention In the 2 years since our last High-Risk Report, three areas—NASA Acquisition Management, Transforming EPA's Process for Assessing and Controlling Toxic Chemicals, and Limiting the Federal Government's Fiscal Exposure By Better Managing Climate Change Risks—have regressed in their ratings against our criteria for removal from the High- Risk List. In addition, while progress is needed across all high-risk areas, we have identified nine additional areas that require significant attention to address imminent, longstanding, or particularly broad issues affecting the nation. Three High-Risk Areas That Regressed NASA Acquisition Management NASA plans to invest billions of dollars in the coming years to explore space, improve its understanding of the Earth’s environment, and conduct aeronautics research, among other things. We designated NASA’s acquisition management as high risk in 1990 in view of NASA’s history of persistent cost growth and schedule delays in the majority of its major projects. Following several years of continuing a generally positive trend of limiting cost growth and schedule delays for its portfolio of major projects, we found that NASA’s average launch delay increased from 7 to 12 months between May 2017 and May 2018. Further, the overall development cost growth increased from 15.6 percent to at least 18.8 percent over the same time period. NASA’s largest science project, the James Webb Space Telescope, has experienced schedule delays of 81 months and cost growth of 95 percent since the project’s cost and schedule baseline was first established in 2009. NASA is at risk for continued cost growth and schedule delays in its portfolio of major projects. Since our 2017 high-risk update, we have lowered NASA acquisition management from meeting the rating to partially meeting the rating in two criteria—leadership commitment and monitoring. The other three criteria ratings remained the same as in 2017. Ratings for capacity and demonstrated progress remain partially met and the rating for action plan remains met. Over the next several years, NASA plans to add new, large, and complex projects to the portfolio, including a lunar Gateway—currently being discussed as a platform in a lunar orbit to mature deep space exploration capabilities. In addition, many of NASA’s current major projects, including some of the most expensive ones, are in the phase of their life cycles when cost growth and schedule delays are most likely. NASA acquisition management requires significant attention for the following reasons: NASA leadership has approved risky programmatic decisions for complex major projects, which compounded technical challenges. For example, leadership has approved some programs to proceed (1) with low cost and schedule reserves, (2) with overly aggressive schedules, and (3) without following best practices for establishing reliable cost and schedule baselines. NASA leadership has also not been transparent about cost and schedule estimates for some of its most expensive projects. Without transparency into these estimates, both NASA and Congress have limited data to inform decision making. NASA has not yet instituted a program for monitoring and independently validating the effectiveness and sustainability of the corrective action measures in its new action plan, which NASA finalized in December 2018. In addition, while NASA has taken some steps to build capacity to help reduce acquisition risk, including updating tools aimed at improving cost and schedule estimates, other areas still require attention. For example, we reported in May 2018 that several major NASA projects experienced workforce challenges, including not having enough staff or staff with the right skills. NASA has also identified capability gaps in areas such as scheduling, earned value management, and cost estimating, and has efforts underway to try to improve capacity in these areas. Since 2017, we have made 9 recommendations on this high-risk area, and as of December 2018, 15 recommendations remain open. These recommendations include that NASA needs to improve transparency of major project cost and schedule estimates, especially for its human spaceflight programs, as well as continue to build capacity to reduce acquisition risk. NASA will also need to implement its new action plan and track progress against it. See page 222 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Transforming EPA's Process for Assessing and Controlling Toxic Chemicals The Environmental Protection Agency’s (EPA’s) ability to effectively implement its mission of protecting public health and the environment is dependent on it assessing the risks posed by chemicals in a credible and timely manner. Such assessments are the cornerstone of scientifically sound environmental decisions, policies, and regulations under a variety of statutes. Based on our work since our 2017 High-Risk Report, the overall rating for leadership commitment decreased from met to partially met due to limited information for completing chemical assessments and proposed budget cuts in the Integrated Risk Information System (IRIS) Program. The ratings for the remaining four criteria remain unchanged and are partially met. The EPA Acting Administrator indicated his commitment to fulfill the agency’s obligations under the Toxic Substances Control Act (TSCA) as amended by the 2016 Frank R. Lautenberg Chemical Safety for the 21 Century Act (Lautenberg Act) and ensure chemicals in the marketplace are safe for human health and the environment. Nonetheless, EPA needs to give more attention to several areas to fully realize the benefits of the new law, and to demonstrate additional progress in the IRIS Program, such as: While EPA released a document in late December 2018 called the IRIS Program Outlook, the Outlook fails to list the projected date for most of the assessments and includes no information regarding assessment prioritization—including how these assessments will meet program and regional office needs. The Lautenberg Act increases both EPA’s responsibility for regulating chemicals and its workload. EPA recently issued a rule under the act to collect fees from certain companies to defray a portion of the implementation costs, but it is unclear whether the fees collected will be sufficient to support relevant parts of the program. EPA issued a First Year Implementation Plan in June 2016 noting that this document is intended to be a roadmap of major activities EPA will focus on during the initial year of implementation. As of mid-February 2019 the plan has not been updated, according to publically available information, although EPA had indicated that it is a living document that will be further developed over time. EPA needs to ensure that the people and resources dedicated to the IRIS Program and TSCA implementation are sufficient. Our March 2019 report on chemical assessments provides information on what remains to be done to address challenges in the IRIS program and implement the Lautenberg Act. Since we added this area to our High-Risk List in 2009, we have made 12 recommendations to EPA related to IRIS and TSCA. As of February 2019, seven recommendations remain open. See page 204 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Limiting the Federal Government's Fiscal Exposure by Better Managing Climate Change Risks Numerous studies have concluded that climate change poses risks to many environmental and economic systems and creates a significant fiscal risk to the federal government. The rising number of natural disasters and increasing reliance on the federal government for assistance is a key source of federal fiscal exposure. As of December 2018, total federal funding for disaster assistance since 2005 is approaching half a trillion dollars (about $430 billion), most recently for catastrophic hurricanes, flooding, wildfires, and other losses in 2017 and 2018. The costliness of disasters is projected to increase as extreme weather events become more frequent and intense due to climate change. There are five areas where government-wide action is needed to reduce federal fiscal exposure, including, but not limited to, the federal government’s role as (1) the insurer of property and crops; (2) the provider of disaster aid; (3) the owner or operator of infrastructure; (4) the leader of a strategic plan that coordinates federal efforts and informs state, local, and private-sector action; and (5) the provider of data and technical assistance to decision makers. Neither global efforts to mitigate climate change causes nor regional adaptation efforts currently approach the scales needed to avoid substantial damages to the U.S. economy, environment, and human health over the coming decades, according to the November 2018 Fourth National Climate Assessment. Government-wide action is needed to improve the nation’s resilience to natural hazards and reduce federal fiscal exposure to climate change impacts. Congress continues to show its commitment to progress on this high-risk issue by enacting legislation. For example, in October 2018, the Disaster Recovery Reform Act was enacted, which, among other things, allows the President to set aside, with respect to each major disaster, a percentage of certain grants to use for pre-disaster hazard mitigation. In addition, the National Defense Authorization Act of 2018, required, among other things, DOD to report on climate impacts to its installations. However, the federal government has not made measurable progress since 2017 to reduce its fiscal exposure to climate change, and in some cases, has revoked prior policies designed to do so. Specifically, since 2017, the ratings for four criteria remain unchanged—three at partially met and one at not met. The rating for one criterion—monitoring—regressed to not met. Limiting the federal government’s fiscal exposure to climate change requires significant attention because the federal government has revoked prior policies that had partially addressed this high-risk area and has not implemented several of our recommendations that could help reduce federal fiscal exposure. For example, since our 2017 high-risk update, the federal government: revoked Executive Order 13690, which had established a government-wide federal flood risk management standard to improve the resilience of communities and federal assets against the impacts of flooding. This action could increase federal fiscal exposure, as taxpayer-funded projects may not last as long as intended because they are not required to account for future changes in climate-related risk. rescinded its guidance directing agencies to consider climate change in their National Environmental Policy Act of 1969 reviews for certain types of federal projects. has not implemented our July 2015 recommendation to establish a comprehensive investment strategy identifying, prioritizing, and implementing federal disaster resilience investments that could reduce federal fiscal exposure to climate change. has not implemented our November 2015 recommendations to create a national climate information system providing authoritative, accessible information useful for state, local, and private-sector decision making. We have made 62 recommendations related to this high-risk area, 12 of which were made since our February 2017 high-risk update. As of December 2018, 25 remain open. The federal government needs a cohesive strategic approach with strong leadership and the authority to manage climate change risks across the entire range of federal activities. See page 110 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Additional High-Risk Areas That Need Significant Attention Ensuring the Cybersecurity of the Nation Federal agencies and the nation’s critical infrastructures—such as energy, transportation systems, communications, and financial services— are dependent on information technology systems to carry out operations. The security of these systems and the data they use is vital to public confidence and national security, prosperity, and well-being. The risks to systems underpinning the nation’s critical infrastructure are increasing as security threats evolve and become more sophisticated. We first designated information security as a government-wide high-risk area in 1997. This was expanded to include protecting cyber critical infrastructure in 2003 and protecting the privacy of personally identifiable information in 2015. In 2018, we updated this high-risk area to reflect the lack of a comprehensive cybersecurity strategy for the federal government. Since 2010, we have made over 3,000 recommendations to agencies aimed at addressing cybersecurity shortcomings, including protecting cyber critical infrastructure, managing the cybersecurity workforce, and responding to cybersecurity incidents. Of those 3,000 recommendations, 448 were made since our last high-risk update in February 2017. Although many recommendations have been addressed, about 700 have not yet been implemented. Despite the number of unimplemented recommendations, since our 2017 High-Risk Report, the administration has made progress in this high-risk area as it continues to meet the leadership commitment criterion through various actions. These include the President issuing (1) an executive order in May 2017 requiring federal agencies to take a variety of actions, including better managing their cybersecurity risks and coordinating to meet reporting requirements related to cybersecurity of federal networks and critical infrastructure and (2) a National Security Strategy in December 2017 citing cybersecurity as a national priority and identifying needed actions. Further, the administration issued a government-wide reform plan and reorganization recommendations in June 2018 with, among other things, proposals for solving the federal cybersecurity workforce shortage. Additionally, the administration released a National Cyber Strategy in September 2018 outlining activities such as securing critical infrastructure, federal networks, and associated information. However, additional actions are needed. We have identified four major cybersecurity challenges facing the nation: (1) establishing a comprehensive cybersecurity strategy and performing effective oversight, (2) securing federal systems and information, (3) protecting cyber critical infrastructure, and (4) protecting privacy and sensitive data. To address the four major cybersecurity challenges, we identified 10 critical actions the federal government and other entities need to take. These critical actions include, for example, developing and executing a more comprehensive federal strategy for national cybersecurity and global cyberspace; addressing cybersecurity workforce management challenges; and strengthening the federal role in protecting the cybersecurity of critical infrastructure (see figure 3). Until these shortcomings are addressed, federal agencies’ information and systems will be increasingly susceptible to the multitude of cyber- related threats that exist. See page 178 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Resolving the Federal Role in Housing Finance The expanded federal role in housing finance that began during the 2007–2009 financial crisis has substantially increased the government’s exposure to potential mortgage losses. Federally supported mortgages include those backed by the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation (Freddie Mac)—collectively, the enterprises—which the Federal Housing Finance Agency (FHFA) placed into government conservatorships in 2008. Federal support also occurs through Federal Housing Administration (FHA) mortgage insurance and Government National Mortgage Association (Ginnie Mae) guarantees on mortgage-backed securities. The substantial financial assistance the enterprises required during and after the crisis, coupled with the large fiscal exposure they and other federal mortgage entities represent today, underscore the need to reform the federal role in housing finance. Delay in resolving the federal role in housing finance poses considerable risks. Through the enterprises, FHA, and Ginnie Mae, the federal government is exposed to potential losses on several trillion dollars in mortgage debt. A severe economic downturn could trigger significant taxpayer assistance to one or more of these entities. Congress and federal agencies have taken some steps to facilitate the transition to a revised federal role, such as holding hearings, introducing legislation, issuing regulations, and developing market monitoring tools. For example, in 2013 and 2014, housing and regulatory agencies finalized rules designed to prevent a recurrence of risky practices in originating and securing mortgages that contributed to the financial crisis. Additionally, FHFA and the Consumer Financial Protection Bureau have developed a representative database of mortgage information that could be useful for examining the effect of mortgage market reforms. However, overall progress on resolving the federal role will be difficult to achieve until Congress provides further direction by enacting changes to the housing finance system. Several issues contribute to the risks facing federal housing finance, including the following: More than 10 years after entering federal conservatorships, the enterprises’ futures remain uncertain and billions of taxpayer dollars remain at risk. Under agreements with the Department of the Treasury (Treasury), the enterprises have received $191.4 billion in capital support as of the end of fiscal year 2018 and have paid dividends to the department exceeding that amount. If they were to incur major additional losses, they would draw required amounts from their remaining $254.1 billion in Treasury commitments. In addition, prolonged conservatorships could hinder development of the broader mortgage securities market by creating uncertainty and crowding out private investment. Nonbanks (lenders and loan servicers that are not depository institutions) have played an increasingly large role in the mortgage market in recent years. While nonbanks have helped provide access to mortgage credit, they also may pose additional risks, in part because they are not federally regulated for safety and soundness. However, FHFA lacks statutory authority to examine nonbank mortgage servicers and other third parties who do business with and pose potential risks to the enterprises. The statutory 2 percent capital requirement for FHA’s $1.26 trillion mortgage insurance fund is not based on a specified risk threshold, such as the economic conditions the fund would be expected to withstand. As a result, it may not provide an adequate financial cushion under scenarios in which Congress may anticipate the fund would be self-sufficient. During the last housing downturn, the fund’s capital ratio fell below the required level and remained there for 6 consecutive years. At the end of fiscal year 2013, the fund required supplemental funds—about $1.7 billion—for the first time in its history. Six of our federal housing recommendations remain open, including those we made in June 2015 on assessing the effects of mortgage reforms already in place. Further, as we previously recommended in November 2016 and January 2019, Congress should consider housing finance reform legislation that: establishes objectives for the future federal role in housing finance, including the role and structure of the enterprises within the housing finance system; provides a transition plan to a reformed system that enables the enterprises to exit federal conservatorship; and addresses all relevant federal entities, including FHA and Ginnie Mae. As we recommended in March 2016 and November 2017, respectively, Congress also should consider granting FHFA explicit authority to examine nonbank servicers and other third parties that do business with the enterprises, and specifying the economic conditions FHA’s insurance fund would be expected to withstand without a substantial risk of requiring supplemental funds. See page 95 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Due to the significance and risk associated with Resolving the Federal Role in Housing Finance, we are separating it from the high-risk area of Modernizing the U.S. Financial Regulatory System. These areas were combined in our 2017 High-Risk report. See page 95 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Pension Benefit Guaranty Corporation Insurance Programs The Pension Benefit Guaranty Corporation (PBGC) is responsible for insuring the defined benefit pension plans for nearly 37 million American workers and retirees, who participate in about 24,800 private sector plans. PBGC faces an uncertain financial future due, in part, to a long- term decline in the number of traditional defined benefit plans and the collective financial risk of the many underfunded pension plans that PBGC insures. PBGC’s financial portfolio is one of the largest of all federal government corporations. While PBGC’s single employer program had a net surplus of about $2.4 billion at the end of fiscal year 2018, its multiemployer program had a net deficit of about $54 billion—or a combined net accumulated financial deficit of over $51 billion. Its deficit has increased by nearly 45 percent since fiscal year 2013. PBGC has estimated that, without additional funding, its multiemployer insurance program will likely be exhausted by 2025 as a result of current and projected pension plan insolvencies. The agency’s single-employer insurance program is also at risk due to the continuing decline of traditional defined benefit pension plans, as well as premiums that are not well aligned to the financial risk presented by the plans it insures. While Congress and PBGC have taken significant and positive steps to strengthen the agency in the past 5 years, challenges related to PBGC’s funding and governance structure remain. Congress established a temporary Joint Select Committee on multiemployer pension plans in 2018—with the goal of improving the solvency of the multiemployer program. However, the committee did not release draft legislation. Addressing the significant financial risk and governance challenges that PBGC faces will require additional congressional action. Over the years since we added PBGC to the High-Risk List, we have suggested a number of matters for congressional consideration, including: (1) authorizing a redesign of PBGC’s single employer program premium structure to better align premium rates with sponsor risk; (2) adopting additional changes to PBGC’s governance structure—in particular, expanding the composition of its board of directors; (3) strengthening funding requirements for plan sponsors as appropriate given national economic conditions; (4) working with PBGC to develop a strategy for funding PBGC claims over the long term as the defined benefit pension system continues to decline; and (5) enacting additional structural reforms to reinforce and stabilize the multiemployer system, and balance the needs and potential sacrifices of contributing employers, participants, and the federal government. Absent additional steps to improve PBGC’s finances, the long-term financial stability of the agency remains uncertain, and the retirement benefits of millions of American workers and retirees could be at risk of dramatic reductions. See page 267 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Managing Risks and Improving VA Health Care VA operates one of the largest health care delivery systems in the nation through its Veterans Health Administration (VHA), with 172 medical centers and more than 1,000 outpatient facilities organized into regional networks. VA has faced a growing demand by veterans for its health care services—due, in part, to the needs of an aging veteran population—and that trend is expected to continue. The total number of veterans enrolled in VA’s health care system rose from 7.9 million to more than 9 million from fiscal year 2006 through fiscal year 2017. Over that same period, VHA’s total budgetary resources have more than doubled, from $37.8 billion in fiscal year 2006 to $92.3 billion in fiscal year 2017. Given the importance of VHA’s mission, coupled with its lack of progress in addressing its high-risk designation, we continue to be concerned about VHA’s ability to ensure its resources are being used effectively and efficiently to improve veterans’ timely access to safe and high-quality health care. We have identified five areas of concern: (1) ambiguous policies and inconsistent processes; (2) inadequate oversight and accountability; (3) IT challenges; (4) inadequate training for VA staff; and (5) unclear resource needs and allocation priorities. VHA has begun to address each of these areas but, prior to Secretary Robert Wilkie’s July 2018 confirmation, its efforts were impeded by leadership instability. Since taking office, Secretary Wilkie has demonstrated his commitment to addressing the department’s high-risk designation by, among other things, creating an office to direct an integrated, focused high-risk approach and communicating to VA leaders the importance of addressing our recommendations. While VHA completed root cause analyses for each area of concern and developed an action plan in response, the plan lacks milestones and metrics needed to effectively monitor its implementation and demonstrate progress made in addressing the high-risk designation. Additionally, many of VHA’s capacity-building initiatives are either in the initial stages of development or are lacking necessary funding and resources. As such, VHA has not made sufficient progress since our 2017 update to improve its overall ratings, as two high-risk criteria remain partially met and three criteria remain unmet. We remain concerned about VHA’s ability to oversee its programs, hold its workforce accountable, and avoid ambiguous policies and inconsistent processes that jeopardize its ability to provide safe, high-quality care to veterans: In November 2017, we reported that, due in part to misinterpretation or lack of awareness of VHA policy, VA medical center officials did not always document or conduct timely required reviews of providers when allegations were made against them. As a result, we concluded that VA medical center officials may have lacked necessary information to reasonably ensure that their providers were competent to provide safe, high-quality care to veterans and to grant approvals about these providers’ privileges to perform specific clinical services at VA medical centers. We made four recommendations related to this and other findings, all of which remain open. In June 2018, we reported that VHA could not systematically monitor the timeliness of veterans’ access to Veterans Choice Program (VCP) care because it lacked complete, reliable data to do so. We also found that veterans, who were referred to the VCP for routine care because health care services were not available in a timely manner, could potentially wait for care up to 70 calendar days if the maximum amount of time allowed by VA processes is used. This wait time exceeds the statutory requirement that veterans receive VCP care within 30 days of the dates their VA health care providers indicated they should receive appointments, or if no such date existed, within 30 days of the veteran’s preferred date. We made 10 recommendations related to this and other findings, all of which remain open. Similarly, in July 2018, we reported that VA collected data related to employee misconduct and disciplinary actions, but data fragmentation and reliability issues impeded department-wide analysis of those data. Additionally, we found that VA did not consistently ensure that allegations of misconduct involving senior officials were reviewed according to its investigative standards or ensure these officials were held accountable. We made 16 recommendations related to this and other findings, all of which remain open. In November 2018, we reported that VHA’s suicide prevention media outreach activities declined in recent years due to leadership turnover and reorganization. Additionally, we found that VHA did not assign key leadership responsibilities or establish clear lines of reporting for its suicide prevention media outreach campaign, which hindered its ability to oversee the campaign. Consequently, we concluded that VHA may not be maximizing its reach with suicide prevention media content to veterans, especially those who are at-risk. This is inconsistent with VHA’s efforts to reduce veteran suicides, which is VA’s highest clinical priority. We made two recommendations related to this and other findings, both of which remain open. VA needs to further develop its capacity-building initiatives and establish metrics to monitor and measure its progress addressing the high-risk areas of concern. It is also important that our recommendations continue to be implemented. The department has implemented 209 of the 353 recommendations related to VA health care that we made from January 1, 2010 through December 2018, but more than 125 recommendations remain open as of December 2018. This includes 17 that are older than 3 years. In addition to addressing our recommendations, VA needs to make systemic change to department management and oversight in order to fully address the high-risk issues and improve the health care provided to our nation’s veterans. See page 275 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Strategic Human Capital Management Mission-critical skills gaps both within federal agencies and across the federal workforce impede the government from cost-effectively serving the public and achieving results. For example, the difficulties in recruiting and retaining skilled health care providers and human resource staff at VHA’s medical centers make it difficult to meet the health care needs of more than 9 million veterans. As a result, VHA’s 168 medical centers have large staffing shortages, including physicians, registered nurses, physician assistants, psychologists, physical therapists, as well as human resource specialists and assistants. OPM continues to demonstrate top leadership commitment through its numerous efforts to assist agencies’ in addressing mission-critical skills gaps within their workforces. This includes providing guidance, training and on-going support for agencies on the use of comprehensive data analytic methods for identifying skills gaps and the development of strategies to address these gaps. However, since we first added strategic human capital management to our High-Risk List in 2001, we have reported on the need for agencies to address their workforce skills gaps. As of December 2018, OPM had not fully implemented 29 of our recommendations made since 2012 relating to this high-risk area. Staffing shortages and the lack of skills among current staff not only affect individual agencies but also cut across the entire federal workforce in areas such as cybersecurity and acquisition management. Skills gaps caused by insufficient number of staff, inadequate workforce planning, and a lack of training in critical skills are contributing to our designating other areas as high-risk. As table 5 shows, of the 34 other high-risk areas covered in this report, skills gaps played a significant role in 16 of the areas. Over the years since we added this area to our High-Risk List, in addition to recommendations to address critical skills gaps in individual high-risk areas, we have made numerous recommendations to OPM related to this high-risk issue, 29 of which remain open. Agencies also need to take action to address mission-critical skills gaps within their own workforces – a root cause of many high-risk areas. See page 75 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. 2020 Decennial Census The 2010 Census was the costliest in history at about $12.3 billion; as of October 2017, the 2020 Census is projected to cost about $15.6 billion, a 27 percent increase. For the 2020 Census, the U.S. Census Bureau (Bureau) plans to implement several innovations, including new IT systems. Implementing these innovations, along with other challenges, puts the Bureau’s ability to conduct a cost-effective census at risk. The decennial census is mandated by the U.S. Constitution and provides vital data for the nation. Census data are used, among other purposes, to apportion seats in the Congress and allocate billions of dollars in federal assistance to state and local governments. To ensure its success, this complicated and costly undertaking requires careful planning, risk management, and oversight. Census activities, some of which are new for the 2020 cycle, must be carried out on schedule to deliver the state apportionment counts to the President by December 31, 2020. The Bureau and the Department of Commerce (Commerce) have strengthened leadership commitment with executive-level oversight of the 2020 Census by holding regular meetings on the status of IT systems and other risk areas. In addition, in 2017 Commerce designated a team to assist senior Bureau management with cost estimation challenges. These examples demonstrate both the Bureau’s and Commerce’s strong leadership commitment to implementing the 2020 Census. One of the Bureau’s major challenges is to control any further cost growth and develop cost estimates that are reliable and reflect best practices for the 2020 Census. According to the Bureau, the total cost of the 2020 Census is now estimated to be approximately $15.6 billion, more than $3 billion higher than previously estimated by the Bureau. The higher estimated life-cycle cost is due, in part, to the Bureau’s failure to previously include all cost associated with the decennial census. The Bureau’s schedule for developing IT systems has experienced delays that have compressed the time available for system testing, integration testing, and security assessments. These schedule delays have contributed to systems experiencing problems after deployment, as well as cybersecurity challenges. For example, as of December 2018, the Bureau had identified nearly 1,100 system security weaknesses that needed to be addressed. Continued schedule management challenges may compress the time available for the remaining system testing and security assessments, and increase the risk that deployed systems will either not function as intended, have security vulnerabilities, or both. As of January 2019, 30 of our recommendations related to this high-risk area had not been implemented. To make continued progress, the Bureau needs to ensure that its approach to strategic planning, IT management, cybersecurity, human capital management, internal collaboration, knowledge sharing, as well as risk and change management are all aligned toward delivering more cost-effective outcomes. Among other things, the Bureau needs to ensure cost growth is controlled and that the development and testing of key systems is completed and fully integrated with all census operations before the 2020 Census. In addition, the Bureau needs to address cybersecurity weaknesses in a timely manner and ensure that security risks are at an acceptable level before systems are deployed. See page 134 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Medicare, Medicaid, and Earned Income Tax Credit Improper Payments An improper payment is any payment that should not have been made or that was made in an incorrect amount (including overpayments and underpayments) under statutory, contractual, administrative, or other legally applicable requirements. Reducing improper payments—such as payments to ineligible recipients or duplicate payments—is critical to safeguarding federal funds. However, the federal government has consistently been unable to determine the full extent of improper payments and reasonably assure that appropriate actions are taken to reduce them. Since 2003—when certain agencies were required by statute to begin reporting improper payments—cumulative improper payment estimates have totaled about $1.5 trillion. As shown in figure 4, for fiscal year 2018, federal entities estimated about $151 billion in improper payments. Medicare and Medicaid improper payments and the Earned Income Tax Credit (EITC) improper payments—a part of the Enforcement of Tax Laws high-risk area—accounted for about 68.5 percent of this total. Federal spending for Medicare programs and Medicaid is expected to significantly increase in the coming years, so it is especially critical to take appropriate measures to reduce improper payments in these programs. Internal Revenue Service estimates also show that the EITC has consistently had a high improper payment rate. OMB has designated Medicare programs, Medicaid, and EITC as high-priority programs for improper payments, indicating they are amongst the highest-risk programs where the government can achieve the greatest return on investment for the taxpayer by ensuring that improper payments are eliminated. Our work has identified a number of strategic and specific actions agencies can take to reduce improper payments, which could yield significant savings, and help ensure that taxpayer funds are adequately safeguarded. Continued agency attention is needed to (1) identify susceptible programs, (2) develop reliable methodologies for estimating improper payments, (3) report as required by statute, and (4) implement effective corrective actions based on root cause analysis. Absent such continued efforts, the federal government cannot be assured that taxpayer funds are adequately safeguarded. See pages 241, 250, and 235 of the report (respectively) for additional detail on the Medicare Program & Improper Payments, Strengthening Medicaid Program Integrity, and Enforcement of Tax Laws high-risk areas, including more details on actions that need to be taken. Enforcement of Tax Laws The Internal Revenue Service (IRS) continues to face two pressing challenges in enforcing tax laws: addressing the tax gap—amounting to hundreds of billions of dollars each year when some taxpayers fail to pay the taxes that they owe—and combatting identity theft (IDT) refund fraud. Enforcement of Tax Laws has been on GAO’s high risk list since 1990. IRS enforcement of tax laws helps fund the U.S. government by collecting revenue from noncompliant taxpayers and, perhaps more importantly, promoting voluntary compliance by giving taxpayers confidence that others are paying their fair share. In 2016, IRS estimated that the average annual net tax gap, the difference between taxes owed and taxes paid on time, was $406 billion, on average, for tax years 2008-2010. While IRS continues to demonstrate top leadership support to address the tax gap, IRS’s capacity to implement new initiatives and improve ongoing enforcement and taxpayer service programs remains a challenge. For example, IRS’s strategic plan includes a goal to facilitate voluntary compliance and deter noncompliance that could address the tax gap. However, IRS could do more to identify specific efforts for improving compliance in its strategic plan, measure the effects of compliance programs—such as those used for large partnerships—and develop specific quantitative goals to reduce the tax gap. Such efforts would help IRS make more effective use of its resources and gauge the success of its strategies. The second challenge facing IRS is IDT refund fraud, which occurs when an identity thief files a fraudulent tax return using a legitimate taxpayer’s identifying information and claims a refund. IRS estimates that at least $12.2 billion in individual IDT tax refund fraud was attempted in 2016, of which it prevented at least $10.5 billion (86 percent). Of the amount attempted, IRS estimated that at least $1.6 billion (14 percent) was paid. IRS’s ability to combat IDT fraud continues to be challenged as more personally identifiable information has become readily available as a result of large-scale cyberattacks on various entities. This makes it more difficult for IRS to distinguish between fraudsters and legitimate taxpayers. While IRS has demonstrated some progress by developing tools and programs to further detect and prevent IDT refund fraud, it has not completed updating its authentication procedures to be in compliance with new government standards. As a result, IRS may be missing an opportunity to implement the most secure, robust technologies to protect taxpayers. As of December 2018, 189 GAO recommendations related to this high- risk area had not been implemented. To make continued progress on closing the tax gap, IRS needs to re-establish goals for improving voluntary compliance and develop and document a strategy that outlines how it will use its data to help address this issue. Reducing the tax gap will also require targeted legislative actions, including additional third- party information reporting, enhanced electronic filing, expanded math error authority (also referred to as correctible error authority), and paid preparer regulation. To help stay on top of IDT refund fraud, IRS should develop a comprehensive process to evaluate alternative options for improving taxpayer authentication. Given that IDT refund fraud continues to be a challenge, targeted legislative action, such as requiring a scannable code on returns prepared electronically but filed on paper could help IRS address such fraud. See page 235 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Improving the Management of IT Acquisitions and Operations The federal government currently invests more than $90 billion annually in IT, and OMB has implemented several key initiatives intended to help better manage this investment. Additionally, enactment of FITARA, in conjunction with greater attention paid to the acquisition and operation of IT, has helped further improve the government-wide management of this significant annual investment. OMB’s current level of top leadership support and commitment to ensure that agencies successfully execute its guidance on implementing FITARA and related IT initiatives has helped this high-risk area meet the leadership commitment high-risk criteria. Additional positive government-wide actions have enabled this high-risk area to partially meet the four remaining high-risk criteria. For example, OMB has established an IT Dashboard—a public website that provides detailed information on major IT investments at 26 federal agencies—and agencies’ data center consolidation efforts have resulted in a total savings of slightly more than 80 percent of the agencies’ planned $5.7 billion in savings since 2011. However, major federal agencies have yet to fully address the requirements of FITARA and realize billions of dollars in planned or possible savings and improved government performance through more efficient budgeting and management of IT. As government-wide spending on IT increases every year, the need for appropriate stewardship of that investment increases as well. However, OMB and federal agencies have not made significant progress since 2017 in taking the steps needed to improve how these financial resources are budgeted and utilized. While OMB has continued to demonstrate its leadership commitment through guidance and sponsorship of key initiatives, agencies still have not fully implemented all requirements of FITARA, such as putting into place authorities the law requires for chief information officers (CIO). Additionally, while the President’s Management Agenda has a goal to improve IT spending transparency, agencies are underreporting IT contract obligations by billions of dollars. OMB and the agencies also have not yet implemented hundreds of our recommendations on improving shortcomings in IT acquisitions and operations. In an August 2018 review of the 24 federal agencies covered by FITARA, none had IT management policies that fully addressed the role of their CIOs consistent with federal laws and guidance. Specifically, the majority of the agencies only minimally addressed, or did not address, their CIO’s role in assessing agency IT workforce needs and developing strategies and plans for meeting those needs. Correspondingly, the majority of the 24 CIOs acknowledged that they were not fully effective at implementing IT management responsibilities, such as IT strategic planning and investment management. Further, in January 2018, we reported that the majority of 22 agencies did not identify all of their IT acquisition contracts, totaling about $4.5 billion in IT-related contract obligations beyond those reported by agencies. In addition, in November 2018 we reported that four selected agencies lacked quality assurance processes for ensuring that billions of dollars requested in their IT budgets were informed by reliable cost information. Until agencies properly identify IT contracts and establish processes for ensuring the quality of cost data used to inform their budgets, agency CIOs are at risk of not having appropriate oversight of IT acquisitions and may lack adequate transparency into IT spending to make informed budget decisions. As of December 2018, OMB and federal agencies had fully implemented only 59 percent of the recommendations we have made since fiscal year 2010 to address shortcomings in IT acquisitions and operations. OMB and agencies should work toward implementing our remaining 456 open recommendations related to this high-risk area. These remaining recommendations include 12 priority recommendations to agencies to, among other things, report all data center consolidation cost savings to OMB, plan to modernize or replace obsolete systems as needed, and improve their implementation of PortfolioStat—an initiative that is to consolidate and eliminate duplicative systems. OMB and agencies need to take additional actions to (1) implement at least 80 percent of our open recommendations related to the management of IT acquisitions and operations, (2) ensure that a minimum of 80 percent of the government’s major IT acquisitions deliver functionality every 12 months, and (3) achieve at least 80 percent of the over $6 billion in planned PortfolioStat savings. See page 123 of the report for additional detail on this high-risk area, including more details on actions that need to be taken. Our high-risk program continues to be a top priority at GAO and we will maintain our emphasis on identifying high-risk issues across government and on providing recommendations and sustained attention to help address them, by working collaboratively with Congress, agency leaders, and OMB. As part of this effort, we hope to continue to participate in regular meetings with the OMB Deputy Director for Management and with top agency leaders to discuss progress in addressing high-risk areas. Such efforts have been critical for the progress that has been made. This high-risk update is intended to help inform the oversight agenda for the 116th Congress and to guide efforts of the administration and agencies to improve government performance and reduce waste and risks. Thank you, Chairman Cummings, Ranking Member Jordan, and Members of the Committee. This concludes my testimony. I would be pleased to answer any questions. For further information on this testimony, please contact J. Christopher Mihm at (202) 512-6806 or MihmJ@gao.gov. Contact points for the individual high-risk areas are listed in the report and on our high-risk website. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Appendix I: Areas Removed From the High- Risk List The following pages provide overviews of the two areas removed from the High-Risk List. Each overview discusses (1) why the area was high risk, and (2) why the area is being removed from the list. Each of these high- risk areas is also described on our High-Risk List website, http://www.gao.gov/highrisk/overview. DOD Supply Chain Management Asset Visibility Since our 2017 High-Risk Report, DOD has continued to meet the criteria of leadership commitment, capacity, and action plan for asset visibility. Further, DOD has fully addressed the three remaining actions and outcomes we outlined in 2017 in order to mitigate or resolve long-standing weaknesses in asset visibility. Consequently, DOD has met the monitoring and demonstrated progress criteria for asset visibility to remove this area from our High-Risk List. Leadership commitment: met. Senior leaders have continued to demonstrate commitment through their involvement in groups such as the Supply Chain Executive Steering Committee—senior-level officials responsible for overseeing asset visibility improvement efforts—and through the Asset Visibility Working Group, which identifies opportunities for improvement and monitors the implementation of initiatives by issuing its Strategy for Improving DOD Asset Visibility (Strategy) in 2014, 2015, and 2017. Capacity: met. DOD continues to demonstrate that it has the capacity— personnel and resources—to improve asset visibility. For example, DOD’s 2015 and 2017 Strategies advise the components to consider items such as staffing, materiel, and sustainment costs when documenting cost estimates for the initiatives in the Strategy, as we recommended in January 2015. Action plan: met. A provision in the National Defense Authorization Act for Fiscal Year 2014 required DOD to submit to Congress a comprehensive strategy and implementation plans for improving asset tracking and in-transit visibility. In January 2014, DOD issued the Strategy and accompanying implementation plans, which outlined initiatives intended to improve asset visibility. DOD updated its 2014 Strategy in October 2015 and in August 2017. Importantly, since 2017 DOD addressed the three remaining actions and outcomes related to the monitoring and demonstrated progress criteria through updates to and implementation of the Strategies (see table 6). Monitoring: met. DOD provided guidance in its 2017 update to the Strategy for the military components to consider key attributes of successful performance measures during metric development for their improvement initiatives. As appropriate, the military components have followed the guidance and provided high-level summary metrics updates to the Asset Visibility Working Group. In addition, DOD has taken steps to monitor asset visibility by incorporating into after-action reports, as appropriate, information relating to performance measures. These after- action reports serve as closure documents and permanent records of each initiative’s accomplishments. Demonstrated progress: met. DOD has demonstrated sustained progress by completing 34 of the 39 initiatives to improve asset visibility and continues to monitor the remaining 5 initiatives. These initiatives have supported DOD’s goals and objectives, which include: (1) improving visibility efficiencies of physical inventories, receipt processing, cargo tracking, and unit moves; (2) ensuring asset visibility data are discoverable, accessible, and understandable to support informed decision-making across the enterprise; and (3) increasing efficiencies for delivery accuracy and cycle times. Also, the Asset Visibility Working Group meets regularly to identify opportunities to further improve asset visibility within DOD. DOD has taken the following actions to demonstrate sustained progress: (1) created an integrated single portal system providing 7,500 users access to near-real-time, in-transit visibility of eight million lines of items of supply and transportation data; and (2) increased its visibility of assets through radio-frequency identification (RFID), an automated data-capture technology that can be used to electronically identify, track, and store information contained on a tag. There are two main types of RFID tags, passive and active, which show whether assets are in-storage, in-transit, in-process, or in-use. Passive tags, such as mass transit passes, do not contain their own power source and cannot initiate communication with a reader; while active tags, such as an “E-Z pass,” contain a power source and a transmitter, and send a continuous signal over longer distances. DOD closed nine initiatives from its Strategies by implementing RFID technology. For example, the Marine Corps implemented long-range passive RFID for visibility and accountability of items, resulting in improvements that include an increased range for “reading” an item— from 30 feet to 240 feet—and reduced inventory cycle times from 12 days to 10 hours. Also, the Navy reported that the use of passive RFID technology to support the overhaul of its nuclear-powered attack submarines enabled the Navy to better track parts, resulting in 98 percent fewer missing components and an average cost avoidance of $1.3 million per boat. Additionally, according to DOD, the use of RFID tags to provide visibility of sustainment cargo at the tactical leg resulted in $1.4 million annual cost savings. Further, DOD reported that the migration of the active RFID enterprise from a proprietary communication standard to a competitive multivendor environment reduced the cost of active RFID tags by half, resulting in an estimated $5.7 million annual reduction in costs. Materiel Distribution Since our 2017 High-Risk Report, DOD has continued to meet the criteria of leadership commitment, capacity, and action plan for materiel distribution. Further, DOD has fully addressed the four remaining actions and outcomes we outlined in 2017 in order to mitigate or resolve long-standing weaknesses in materiel distribution. Consequently, DOD has met the monitoring and demonstrated progress criteria for materiel distribution to remove this area from our High-Risk List. Leadership commitment: met. Senior leaders continue to demonstrate commitment through their involvement in groups such as the Supply Chain Executive Steering Committee—senior-level officials responsible for overseeing materiel distribution corrective actions—and through the Distribution Working Group, which helped develop the Materiel Distribution Improvement Plan (Improvement Plan) in 2016. Capacity: met. DOD has continued to demonstrate that it has the personnel and resources, such as key organizations and the associated governance structure, to improve materiel distribution. The Improvement Plan recognizes that additional resources will be required to accomplish its corrective actions and close any identified performance gaps within the time frame specified. Action plan: met. In 2016, DOD developed its corrective action plan to address the department’s materiel distribution challenges. The Improvement Plan details specific goals and actions to better measure the end-to-end distribution process, ensure the accuracy of underlying data, and strengthen and integrate distribution policies and the governance structure. Importantly, since 2017, DOD has fully addressed the four remaining actions and outcomes related to monitoring and demonstrated progress to mitigate or resolve long-standing weaknesses in materiel distribution (see table 7). Monitoring: met. DOD has monitored materiel distribution by making progress in developing its suite of distribution performance metrics, improving the quality of their underlying data, and sharing metrics information with stakeholders. For example, in January 2017, DOD developed a suite of performance metrics that provides a comprehensive picture of the distribution process, including whether supplies are delivered on time and at sufficient quantity and quality. Also, DOD implemented checklists to assess the quality of data underlying each performance metric based on relevance, accuracy, comparability, and interpretability. The checklists and their standards assist in identifying root causes and addressing areas where performance data quality may be lacking. DOD has also incorporated internal control requirements in its supply chain management guidance to increase confidence in the performance data. Additionally, DOD has revised its policy documents to require stakeholders to routinely capture and share distribution performance metrics, including cost data, and the department maintains websites to provide current performance information to distribution stakeholders. DOD has also incorporated distribution metrics, as appropriate, on the performance of all legs of the distribution system, including the tactical leg (i.e., the last segment of the distribution system). We previously reported on DOD’s deficiencies to accurately assess its distribution performance at the tactical leg, such as missing delivery dates for shipments in Afghanistan. Since that time, the geographic combatant commands have been tracking metrics at the tactical leg, including required delivery dates, to determine the movement and causes of delays for shipments, and have been sharing distribution performance information with the U.S. Transportation Command (TRANSCOM) through their deployment and distribution operations centers. DOD is implementing a cost framework to incorporate transportation costs for all legs of the distribution system, which will provide an additional metric for distribution stakeholders to assess the efficiency of the system. The first phase of the cost framework began operating in August 2018 and is expected to be fully implemented in 2019. DOD is making progress in refining its Improvement Plan and is incorporating additional actions based on interim progress and results. Since DOD issued the Improvement Plan in September 2016, the agency has (1) documented the results and monitored the status of each corrective action, (2) revised completion dates as needed, and (3) periodically provided decision makers with summary action charts, plans, and milestones. DOD is also updating its instruction on management and oversight of the distribution enterprise to clarify the roles and responsibilities of all distribution stakeholders. DOD officials have not determined a date for when this instruction will be issued. Demonstrated progress: met. DOD has demonstrated sustained progress in improving its capability to comprehensively measure distribution performance, identify distribution problems and root causes, and implement solutions. DOD has implemented 10 of 18 corrective actions in its Improvement Plan and is on track to implement the remaining 8 by September 2019. Because of this progress, DOD’s monthly shipment reports have assessed performance against enhanced metrics across the distribution system. For example, in December 2017, TRANSCOM investigated performance standards for truck deliveries from its Defense Logistics Agency warehouses in Bahrain to customers in Kuwait due to frequent delays in shipments. TRANSCOM determined that inadequate time for clearing customs in Kuwait resulted in an unrealistic delivery standard. TRANSCOM, in coordination with distribution stakeholders, adjusted the delivery standard to adequately account for the in-theater customs process. In addition, TRANSCOM, in partnership with the Defense Logistics Agency and the General Services Administration, developed and implemented initiatives focused on distribution process and operational improvements to reduce costs and improve distribution services to the warfighter. According to DOD, these efforts have resulted in at least $1.56 billion in distribution cost avoidances to date. Monitoring After Removal DOD has demonstrated commendable, sustained progress improving its supply chain management. This does not mean DOD has addressed all risk within this area. It remains imperative that senior leaders continue their efforts to implement initiatives and corrective actions to maintain visibility of supplies, track cargo movements, meet delivery standards, and maintain delivery data for shipments. Continued oversight and attention are also warranted given the recent reorganization of the Office of the Under Secretary of Defense for Acquisition and Sustainment and the resulting change in the oversight structure of Supply Chain Management. We will therefore continue to conduct oversight of supply chain management at DOD. Related GAO Products Defense Logistics: Improved Performance Measures and Information Needed for Assessing Asset Visibility Initiatives. GAO-17-183. Washington, D.C.: Mar. 16, 2017. Defense Logistics: DOD Has Addressed Most Reporting Requirements and Continues to Refine its Asset Visibility Strategy. GAO-16-88. Washington, D.C.: Dec. 22, 2015. Defense Logistics: Improvements Needed to Accurately Assess the Performance of DOD’s Materiel Distribution Pipeline. GAO-15-226. Washington, D.C.: Feb. 26, 2015. Mitigating Gaps in Weather Satellite Data NOAA’s Polar- Orbiting Weather Satellites Since our last high-risk update in 2017, NOAA continues to meet the criteria of leadership commitment, capacity, and monitoring and now also meets the criteria of action plan and demonstrated progress. Leadership commitment: met. NOAA program officials met the leadership commitment criteria in 2015 and have continued to sustain their strong leadership commitment to mitigating potential satellite data gaps since that time. For example, NOAA issued and frequently updated its polar satellite gap mitigation plan, which identifies the specific technical, programmatic, and management steps the agency is taking to ensure that satellite mitigation options are viable. In addition, NOAA executives continue to oversee the acquisition of polar-orbiting satellites through monthly briefings on the cost, schedule, and risks affecting the satellites’ development. Capacity: met. NOAA continues to meet the criterion of improving its capacity to address the risk of a satellite data gap. In December 2014, we recommended that NOAA investigate ways to prioritize the gap mitigation projects with the greatest potential benefit to weather forecasting, such as by improving its high-performance computing capacity. NOAA agreed with this recommendation and implemented it. For example, NOAA upgraded its high-performance computers, which allowed the agency to move forward on multiple other mitigation activities, including experimenting with other data sources and assimilating these data into its weather models. Action plan: met. NOAA now meets the criterion for having a plan to address the risk of a polar satellite data gap, which is an increase over its rating in 2017. In June 2012, we reported that, while NOAA officials communicated publicly and often about the risk of a polar satellite data gap, the agency had not established plans to mitigate the gap. We recommended that NOAA establish a gap mitigation plan, and the agency did so in February 2014. However, in December 2014, we recommended that NOAA revise its plan to address shortfalls, including (1) adding recovery time objectives for key products, (2) identifying opportunities for accelerating the calibration and validation of satellite data products, (3) providing an assessment of available alternatives based on their costs and impacts, and (4) establishing a schedule with meaningful timelines and linkages among mitigation activities. mitigation plan between January 2016 and February 2017. With the last of the updates, the agency addressed the shortfalls we had identified. Monitoring: met. NOAA met this criterion in 2017, and continues to meet it now, by implementing our recommendations to more consistently and comprehensively monitor its progress on gap mitigation activities. For example, all three NOAA organizations responsible for gap mitigation projects regularly brief senior management on their progress. Demonstrated progress: met. NOAA now meets the criterion for demonstrated progress, which is an increase over its prior rating. In our 2017 High-Risk Report, we noted that NOAA had identified 35 different gap mitigation projects and was making progress in implementing them. These projects fell into three general categories: (1) understanding the likelihood and impact of a gap, (2) reducing the likelihood of a gap, and (3) reducing the impact of a gap. Nevertheless, one of the most important steps in reducing the likelihood of a gap—keeping the launch of the next polar satellite on schedule—had encountered problems. Specifically, agency officials decided to delay the launch due to challenges in developing the ground system and a critical instrument on the spacecraft. This delay exacerbated the probability of a satellite data gap. More recently, however, NOAA was able to demonstrate progress by successfully launching the satellite in November 2017. That satellite, now called NOAA-20, is currently operational and is being used to provide advanced weather data and forecasts. Moreover, the agency is also working to build and launch the next satellites in the polar satellite program. DOD’s Polar-Orbiting Weather Satellites Since our last high-risk update in 2017, DOD now meets all five high-risk criteria. Leadership commitment: met. With strong congressional oversight, DOD now meets this criterion. Pursuant to enactment of the Carl Levin and Howard P. ’Buck’ McKeon National Defense Authorization Act for Fiscal Year 2015 (NDAA for FY 2015), the National Defense Authorization Act for Fiscal Year 2016 (NDAA for FY 2016), and the Consolidated Appropriations Act, 2016, DOD leadership committed to developing and implementing plans to address its weather satellite requirements. For example, in late 2017, the department awarded a contract for its Weather System Follow-on— Microwave satellite to fulfill core weather requirements. Capacity: met. With strong congressional oversight, DOD now meets the capacity criterion. Specifically, the NDAA for FY 2015 restricted the availability of 50 percent of the FY 2015 funds authorized for the Weather Satellite Follow-on System (now called the Weather System Follow-on— Microwave satellite program) until DOD submitted to the congressional defense committees a plan to meet weather monitoring data collection requirements. In addition, the explanatory statement that accompanied the Consolidated Appropriations Act, 2016, recommended that the Air Force focus on ensuring that the next generation of weather satellites meet the full spectrum of requirements and work with civil stakeholders to leverage appropriate civil or international weather assets. As called for in the law and the explanatory statement, DOD established plans to meet weather monitoring data collection needs, including by acquiring satellites as part of a family of systems to replace its aging legacy weather satellites. Additionally, DOD formally coordinated with NOAA on weather monitoring data collection efforts. In January 2017, the Air Force and NOAA signed a memorandum of agreement, and in November 2017, signed an annex to that agreement, to allow for the exchange of information and collaboration on a plan for collecting weather monitoring data. The Air Force and NOAA are now developing plans to relocate a residual NOAA satellite over the Indian Ocean, an area of concern for cloud characterization and area-specific weather imagery coverage. Action plan: met. In our 2017 High-Risk Report, we reported that DOD was slow to establish plans for its Weather System Follow-on–Microwave program and had made little progress in determining how it would meet weather satellite requirements for cloud characterization and area-specific weather imagery. Pursuant to the NDAA for FY 2015, the NDAA for FY 2016, and the explanatory statement that accompanied the Consolidated Appropriations Act, 2016, the department developed and began implementing plans to address its weather satellite requirements. As mentioned above, in late 2017, the department awarded a contract for its Weather System Follow-on–Microwave satellite to fulfill core weather requirements. Under this program, the department may launch a demonstration satellite in 2021 and plans to launch an operational satellite in 2022. capabilities. DOD plans to launch Operationally Responsive Space-8 as early as 2022. Monitoring: met. DOD now meets the monitoring criterion as evidenced by its actions to initiate a major acquisition program, the Weather System Follow-on–Microwave, and award a contract for the first satellite. In addition, program officials stated that they plan to monitor the program’s progress toward addressing critical needs and assess its operations and sustainment costs. Demonstrated progress: met. DOD now meets the demonstrated progress criterion because it has developed plans and taken actions to address gaps in weather data through its plans to launch the Weather System Follow-on–Microwave satellite in 2022. The department also plans to launch the Electro-Optical/Infrared Weather Systems satellite in 2024 and provide interim capabilities beginning as early as 2022. By developing these plans, DOD has reduced the risk of a gap in weather satellite data and addressed the concerns about a lack of planning that we identified in our 2017 High-Risk Report. DOD’s effective implementation of its plans will be key to further reducing the risks of gaps in weather satellite data in the future. Monitoring After Removal Moving forward, we will continue to monitor both NOAA and DOD efforts to develop and launch the next satellites in their respective weather satellite programs. NOAA plans to launch its next geostationary weather satellite in 2021 and to launch its next polar weather satellite in 2022. DOD plans satellite launches in 2021 (potentially), 2022, and 2024. In addition, we will continue to monitor DOD’s efforts to develop long-term plans to meet its weather satellite requirements. Related GAO Products Weapon Systems Annual Assessment: Knowledge Gaps Pose Risks to Sustaining Recent Positive Trends. GAO-18-360SP. Washington, D.C.: Apr. 25, 2018. Satellite Acquisitions: Agencies May Recover a Limited Portion of Contract Value When Satellites Fail. GAO-17-490. Washington, D.C.: June 9, 2017. Defense Acquisitions: Assessments of Selected Weapon Programs. GAO-17-333SP. Washington, D.C.: Mar. 30, 2017. Defense Weather Satellites: DOD Faces Acquisition Challenges for Addressing Capability Needs. GAO-16-769T. Washington, D.C.: July 7, 2016. Polar Satellites: NOAA Faces Challenges and Uncertainties that Could Affect the Availability of Critical Weather Data. GAO-16-773T. Washington, D.C.: July 7, 2016. Polar Weather Satellites: NOAA Is Working to Ensure Continuity but Needs to Quickly Address Information Security Weaknesses and Future Program Uncertainties. GAO-16-359. Washington, D.C.: May 17, 2016. Defense Weather Satellites: Analysis of Alternatives Is Useful for Certain Capabilities, but Ineffective Coordination Limited Assessment of Two Capabilities. GAO-16-252R. Washington, D.C.: Mar. 10, 2016. 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.

Why GAO Did This Study The federal government is one of the world's largest and most complex entities; about $4.1 trillion in outlays in fiscal year 2018 funded a broad array of programs and operations. GAO's high-risk program identifies government operations with vulnerabilities to fraud, waste, abuse, and mismanagement, or in need of transformation to address economy, efficiency, or effectiveness challenges. This biennial update describes the status of high-risk areas, outlines actions that are still needed to assure further progress, and identifies two new high-risk areas needing attention by the executive branch and Congress. Solutions to high-risk problems save billions of dollars, improve service to the public, and would strengthen government performance and accountability. GAO uses five criteria to assess progress in addressing high-risk areas: (1) leadership commitment, (2) agency capacity, (3) an action plan, (4) monitoring efforts, and (5) demonstrated progress. What GAO Found The ratings for more than half of the 35 areas on the 2019 High-Risk List remain largely unchanged. Since GAO's last update in 2017, seven areas improved, three regressed, and two showed mixed progress by improving in some criteria but declining in others. Where there has been improvement in high-risk areas, congressional actions have been critical in spurring progress in addition to actions by executive agencies. GAO is removing two of the seven areas with improved ratings from the High-Risk List because they met all of GAO's five criteria for removal. The first area, Department of Defense (DOD) Supply Chain Management, made progress on seven actions and outcomes related to monitoring and demonstrated progress that GAO recommended for improving supply chain management. For example, DOD improved the visibility of physical inventories, receipt processing, cargo tracking, and unit moves. Improvements in asset visibility have saved millions of dollars and allow DOD to better meet mission needs by providing assets where and when needed. The second area, Mitigating Gaps in Weather Satellite Data, made significant progress in establishing and implementing plans to mitigate potential gaps. For example, the National Oceanic and Atmospheric Administration successfully launched a satellite, now called NOAA-20, in November 2017. NOAA-20 is operational and provides advanced weather data and forecasts. DOD developed plans and has taken actions to address gaps in weather data through its plans to launch the Weather System Follow-on–Microwave satellite in 2022. There are two new areas on the High-Risk List since 2017. Added in 2018 outside of GAO's biennial high-risk update cycle, the Government-Wide Personnel Security Clearance Process faces significant challenges related to processing clearances in a timely fashion, measuring investigation quality, and ensuring information technology security. The second area, added in 2019, is Department of Veterans Affairs (VA) Acquisition Management. VA has one of the most significant acquisition functions in the federal government, both in obligations and number of contract actions. GAO identified seven contracting challenges for VA, such as outdated acquisition regulations and policies, lack of an effective medical supplies procurement strategy, and inadequate acquisition training. Overall, 24 high-risk areas have either met or partially met all five criteria for removal from the list; 20 of these areas fully met at least one criterion. Ten high-risk areas have neither met nor partially met one or more criteria. While progress is needed across all high-risk areas, GAO has identified nine that need especially focused executive and congressional attention, including Ensuring the Cybersecurity of the Nation, Resolving the Federal Role in Housing Finance, addressing Pension Benefit Guaranty Corporation Insurance Programs, Managing Risks and Improving VA Health Care, and ensuring an effective 2020 Decennial Census. Beyond these specific areas, focused attention is needed to address mission-critical skills gaps in 16 high-risk areas, confront three high-risk areas concerning health care and tax law enforcement that include billions of dollars in improper payments each year, and focus on a yawning tax gap. What GAO Recommends This statement describes GAO's views on progress made and what remains to be done to bring about lasting solutions for each high-risk area. Substantial efforts are needed by the executive branch to achieve progress on high-risk areas. Addressing GAO's hundreds of open recommendations across the high-risk areas and continued congressional oversight and action are essential to achieving greater progress.

gao_GAO-19-711T

gao_GAO-19-711T_0

Background The Coast Guard owns or leases more than 20,000 facilities consisting of various types of buildings and structures. The Coast Guard’s shore infrastructure is organized into five product lines and 13 asset types, known as asset lines. For example, within its shore operations asset line, the Coast Guard maintains over 200 stations along U.S. coasts and inland waterways to carry out its search and rescue operations, as well as other missions such as maritime security. Much of the Coast Guard’s infrastructure is vulnerable to the effects of extreme weather and can be costly to repair or replace after major storms. From December 2005 through June 2019, the Coast Guard received about $2 billion in supplemental appropriation funds to, among other things, rebuild or relocate 15 facilities damaged by hurricanes. During this time, the Coast Guard relocated facilities further inland or to higher ground, upgraded facilities to be more resilient, and designed new facilities with features to protect them from natural disasters. For example, after being damaged by Hurricane Ike in 2008, the Coast Guard relocated a regional facility in Houston, Texas further inland to help protect the new facility from extreme weather. The facility was also designed to withstand wind speeds of up to 115 miles per hour. In February 2017, the Coast Guard’s Civil Engineering program also issued guidance intended to increase the likelihood that new or recapitalized buildings would be designed to withstand natural disasters, and to enable the Coast Guard to better manage risks to its operations and personnel, among other things. Almost Half of the Coast Guard’s Shore Infrastructure is Beyond Its Service Life, and Project Backlogs Will Cost at Least $2.6 Billion to Address We found in February 2019 that the condition of the Coast Guard’s shore infrastructure was deteriorating and almost half of it was past its service life—resulting in (1) recapitalization and new construction and (2) deferred maintenance backlogs of at least $2.6 billion as of 2018. In 2018, the Coast Guard graded its overall shore infrastructure condition as a C minus based on criteria it derived from standards developed by the American Society of Civil Engineers. Table 1 shows information about the number of assets, replacement value, service life of, and condition grades assigned by the Coast Guard for each of its asset lines for fiscal year 2018. The aging and deteriorating condition of the Coast Guard’s shore infrastructure has led to at least $2.6 billion in deferred construction projects and maintenance backlogs. With almost half of its infrastructure past its service life, and given recent Coast Guard funding requests for its shore infrastructure, it will take many years for the agency to address these backlogs. For example, in 2018 the Coast Guard estimated that it would take almost 400 years to address just the $1.774 billion recapitalization and new construction backlog—assuming an overall 65- year service life and that funding would continue at the fiscal year 2017 appropriations level. This time frame estimate excludes the Coast Guard’s $900 million deferred depot-level maintenance backlog. Table 2 provides information on the Coast Guard’s two shore infrastructure backlogs as of August 2018. Nevertheless, the size and estimated costs of the Coast Guard’s backlogs may be understated. We found in February 2019 that the Coast Guard’s estimated costs did not include hundreds—or the majority—of the projects on the recapitalization and new construction backlog. For example, we reported that there were 205 projects on the backlog without cost estimates. Officials explained that they had not prepared cost estimates for these projects because they were in the preliminary stages of development. Coast Guard Has Taken Initial Steps toward Improving Its Management of Its Shore Infrastructure Our previous reports have identified various steps the Coast Guard has taken to begin to improve how it manages its shore infrastructure. Some of the steps the Coast Guard has taken align with leading practices for managing public sector backlogs and key practices for managing risks to critical infrastructure, including identifying risks posed by the lack of timely investment, identifying mission-critical facilities, disposing of unneeded assets, and beginning an assessment of shore infrastructure vulnerabilities. Specifically, the Coast Guard has: Identified risks posed by lack of timely investment. In February 2019, we found that the Coast Guard had a process to identify, document, and report risks to its shore infrastructure in its annual shore infrastructure reports for fiscal years 2015 through 2018. These reports identified the types of risks the Coast Guard faces in not investing in its facilities, including financial risk, capability risk, and operational readiness risk. The Coast Guard met this leading practice to identify risk in general terms—for example, in terms of increased lifecycle costs, or risk to operations. Identified mission-critical and mission-supportive shore infrastructure. In February 2019, we found that since at least 2012, the Coast Guard had documented its process to classify all of its real property under a tier system and established minimum investment targets by tier as part of its central depot level maintenance expenditure decisions. These tiers—which range from mission- critical to mission-supportive assets—were incorporated into guidance that Coast Guard decision makers are to follow in their deliberations about project funding, and to help them determine how to target funding more effectively. For example, Coast Guard guidance for fiscal years 2019 through 2023 prioritized expenditures on shore infrastructure supporting front line operations, such as piers or runways, over shore infrastructure providing indirect support to front line operations, such as administrative buildings. Assessed selected buildings for vulnerabilities. We issued a report today that discusses the Coast Guard Civil Engineering program’s efforts to conduct a vulnerability assessment of its owned and occupied buildings, which the Coast Guard initiated in 2015 and aims to complete in 2025. The Coast Guard calls this infrastructure review the Shore Infrastructure Vulnerability Assessment. The focus of Phase I of this assessment, completed in 2019, was to determine the vulnerability of certain occupied buildings to 10 natural disasters. Further, the assessment results are intended to assist with contingency planning by identifying which Coast Guard facilities are likely to remain operational after a natural disaster. During Phase I of this assessment, completed in 2019, the Coast Guard analyzed 3,214 buildings, almost 16 percent of its infrastructure, for vulnerabilities to disasters such as floods, earthquakes, and hurricanes. The analysis identified Coast Guard- wide infrastructure vulnerabilities to coastal risks such as shoreline loss, coastal erosion and earthquakes, as well as tsunami risks on the West Coast of the United States, Alaska, Guam, and Hawaii, and immediate and serious flood risks in Puerto Rico and the Gulf and East Coasts. The Phase I report recommended that Coast Guard units and contingency planners consider these vulnerabilities when preparing contingency plans or making capital investments. The Coast Guard has also initiated a follow up effort involving structural analyses for buildings it believes to be more susceptible to damage from earthquakes and wind. Officials involved said their aim is to complete this effort in 2025. Coast Guard Has Not Fully Applied Leading Practices and Key Risk Management Steps in Managing its Shore Infrastructure The Coast Guard has taken actions to begin to improve its shore infrastructure management. However, as we previously reported, the Coast Guard has not fully applied leading practices and key risk management steps to improve its shore infrastructure management. Specifically, we found, among other things, that the following actions could help improve the Coast Guard’s shore infrastructure management efforts: Employ models for predicting the outcome of investments and analyzing tradeoffs. In February 2019, we found that a 2017 Coast Guard Aviation Pavement Study employed a model that found that the Coast Guard could more efficiently prioritize investment in aviation pavement. A subsequent Coast Guard aviation pavement plan recommended actions to use the study results and potentially save $13.8 million. However, we found that the Coast Guard had not fully implemented its own recommended actions to achieve the cost savings. Additionally, we found that while a similar analytical approach to efficiently prioritizing investments in aviation pavement could be applied to all of the shore infrastructure asset lines, the Coast Guard had not applied the approach to other asset lines. By not employing similar models across its asset lines for predicting the outcome of investments, analyzing tradeoffs, and optimizing decisions among competing investments, the Coast Guard is missing opportunities to potentially identify and achieve cost savings across other asset lines. We recommended that the Coast Guard employ models for its asset lines that would predict the investment outcomes, analyze tradeoffs, and optimize decisions among competing investments. The Coast Guard agreed with our recommendation but as of August 2019 had not addressed it. The Coast Guard stated that it plans to assess the use of modeling tools used by the Department of Defense as well as other alternatives to enhance its real property asset management capability. We will continue to monitor its actions. Dispose of unneeded assets. In October 2017, we found that disposing of unneeded assets, such as closing unnecessarily duplicative boat stations, based on a sound analytical process, could potentially generate $290 million in cost savings over 20 years. Specifically, the Coast Guard identified 18 unnecessarily duplicative boat stations with overlapping coverage that could be permanently closed without negatively affecting the Coast Guard’s ability to meet its mission requirements, including its 2-hour search and rescue response standard. In 2017, the Coast Guard affirmed that its leadership believes the study remains valid, but as of September 2019 it has not closed any stations. Figure 1 depicts the extent of the Coast Guard’s overlapping boat and air station search and rescue coverage, as identified by the Coast Guard, some of which the Coast Guard determined to be unnecessarily duplicative. In February 2019, we found that 5 of the 18 boat stations recommended for closure had projects listed on the Coast Guard’s current project backlog. For example, Station Shark River, in New Jersey, was recommended for recapitalization in fiscal year 2017, despite Coast Guard recommendations to close the station in 1988, 1996, 2007, and 2013. Notably, the Coast Guard has made multiple attempts in previous years to close such stations but was unable to due to congressional intervention, and subsequent legislation prohibiting closures. In October 2017, we recommended that the Coast Guard establish and implement a plan with target dates and milestones for closing boat stations that it has determined provide overlapping search and rescue coverage and are unnecessarily duplicative. In February 2019, we further recommended disposing of unneeded assets to more efficiently manage resources and better position the Coast Guard and Congress to address shore infrastructure challenges. The Coast Guard agreed with our recommendations. As of September 2019, the Coast Guard reported that it was considering changes in the operational status of several stations, such as closing the stations during the winter months when they conduct few, if any, search and rescue cases. The Coast Guard estimated that it will continue to consider changes until March 2020. These are positive steps, but we continue to believe that it is important for the Coast Guard to dispose of unneeded assets. Given the Coast Guard’s competing acquisition, operational, and maintenance needs, and its existing $1.774 billion project backlog of recapitalization and new construction projects, these actions may help to mitigate some of its resource challenges. We will continue to monitor the Coast Guard’s efforts to implement these recommendations. Report shore infrastructure project backlogs accurately. In February 2019, we found areas in which the Coast Guard could increase budget transparency for shore infrastructure by accurately reporting project backlogs and costs in Congressionally-required plans. Specifically, we found that the Coast Guard had not provided accurate information to Congress necessary to inform decision- makers of the risks posed by untimely investments in maintenance and repair backlogs. For example, the Coast Guard had not provided complete information to Congress in its Unfunded Priorities Lists of shore infrastructure projects, including information about tradeoffs among competing project alternatives, as well as the impacts on missions conducted from shore facilities in disrepair. We also found that Coast Guard budget requests related to shore infrastructure for fiscal years 2012 through 2019 generally did not identify funding to address any backlogs of deferred maintenance or recapitalization, except for one fiscal year—2012—when the Coast Guard requested $93 million to recapitalize deteriorated/obsolete facilities. We also found that the Coast Guard had not provided accurate information about its requirements-based budget targets for shore infrastructure in its budget requests. According to Coast Guard officials, a requirements-based budget is an estimate of the cost to operate and sustain its shore infrastructure portfolio of assets over the lifecycle of the asset, from initial construction or capital investment through divestiture or demolition. Further, we found that Coast Guard recapitalization targets showed a far greater need than was reflected in the appropriations it requested from fiscal years 2012 through 2019. Specifically, Coast Guard targets for recapitalization of shore assets indicated the Coast Guard needs $290 to $390 million annually for its recapitalization efforts. However, its budget requests for fiscal years 2012 through 2018 have ranged from about $5 million to about $99 million annually. We recommended that the Coast Guard include supporting details about competing project alternatives and report tradeoffs in Congressional budget requests and related reports. Without such information about the Coast Guard’s budgetary requirements, the Congress will lack critical information that could help to prioritize funding to address the Coast Guard’s shore infrastructure backlogs. While the Coast Guard agreed with our recommendation, in August 2019 officials reported that they will continue to develop budgets as the agency has done but will include additional information in future required reports to Congress. We will continue to monitor these actions. Fully implement DHS’s Critical Infrastructure Risk Management Framework. In September 2019, we found that the Coast Guard has taken some steps to improve the resilience of its shore infrastructure by rebuilding storm-damaged facilities and initiating a vulnerability assessment, but its processes to improve shore infrastructure resilience are not fully aligned with the five steps DHS has identified for critical infrastructure risk management (DHS Critical Infrastructure Risk Management Framework). The five steps include: (1) setting goals and objectives, (2) identifying critical infrastructure, (3) assessing and analyzing risks and costs, (4) implementing risk management activities, and (5) measuring the effectiveness of actions taken. We found that the Coast Guard is not positioned to provide decision makers with complete details of which infrastructure facilities are critical, and the type of information the DHS Critical Infrastructure Risk Management Framework recommends for making cost effective risk management decisions. The Coast Guard identified occupied buildings that may be important to operations and assessed their vulnerability through its Shore Infrastructure Vulnerability Assessment process, but this process did not identify all shore infrastructure assets that are critical to its missions—such as aircraft runways—or screen them for all vulnerabilities, such as flooding. Similarly, we found that while the Coast Guard identified almost 800 buildings that may be vulnerable to tornadoes and another 1,000 buildings vulnerable to hurricanes, it has not analyzed the potential consequences, such as economic losses, costs for rebuilding, and impact on mission, should this infrastructure suffer damage from those vulnerabilities. Without a complete understanding of both the vulnerabilities of its infrastructure and the consequences to its mission operations if its infrastructure is damaged, the Coast Guard risks questionable recapitalization investments for improving resilience when selecting projects to fund. Such an understanding is especially important given its existing project backlogs of at least $2.6 billion. The five steps of the DHS Critical Infrastructure Risk Management Framework are intended to guide decision making and prioritize actions to more effectively achieve desired outcomes. Therefore, in September 2019 we recommended that the Coast Guard implement risk management processes that more fully align with the five key steps outlined in DHS’s Critical Infrastructure Risk Management Framework to better guide its shore infrastructure investment decisions. The Coast Guard agreed with our recommendation. It stated that it plans to make progress towards implementing the recommendation while developing and implementing its Component Resilience Plan, in accordance with the recently mandated DHS Resilience Framework. It intends to complete these efforts by the end of 2021. The Coast Guard also intends to develop, by July 2020, goals and objectives for measuring the effectiveness of actions taken to identify resilience readiness gaps and resource needs. We will continue to monitor these efforts. Chairman Maloney, Ranking Member Gibbs, and Members of the Subcommittee, this completes my prepared statement. I would be happy to respond to any questions you may have at this time. GAO Contact and Staff Acknowledgments If you or your staff members have any questions about this testimony, please contact me at (202) 512-3841 or andersonn@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. In addition to the contact above, Dawn Hoff, Assistant Director; Andrew Curry, Analyst-in-Charge; Peter Haderlein; Landis Lindsey; Calaera Powroznik, and Molly Ryan made key contributions to this testimony. Other staff who made key contributions to the reports cited in the testimony are identified in the source products. 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.

Why GAO Did This Study The Coast Guard, within the Department of Homeland Security (DHS), owns or leases more than 20,000 shore facilities—such as piers, boat stations, air stations, runways, and housing units—at more than 2,700 locations, from which it carries out its missions. This shore infrastructure is often positioned along the nation's coastlines where it can be vulnerable to damage from extreme weather. This statement summarizes GAO findings related to the condition of Coast Guard shore infrastructure, actions the Coast Guard has taken to improve its management of its shore infrastructure, and additional actions it needs to take. This statement is based on three GAO products issued from October 2017 through September 2019, along with selected updates on actions the Coast Guard has taken to address GAO's recommendations from these reports. GAO analyzed relevant Coast Guard documents, management processes and decisions, and interviewed Coast Guard officials. To conduct updates, GAO also reviewed information on the Coast Guard's actions to implement its prior recommendations. What GAO Found In February 2019, GAO reported that the Coast Guard's $18 billion portfolio of shore infrastructure was deteriorating, and almost half of it was past its service life as of 2018. Coast Guard data showed that it would cost at least $2.6 billion to address its maintenance and recapitalization (major renovation) project backlogs at recent funding levels. Coast Guard data also showed that hundreds of projects had not been factored into the backlog costs. GAO's prior work has shown that the Coast Guard has taken initial steps toward improving how it manages its shore infrastructure, including conducting an initial assessment of shore infrastructure vulnerabilities. However, GAO also found that the Coast Guard had not fully applied leading practices and key risk management steps in managing its shore infrastructure, and needs to take the following actions: Employ models for predicting the outcome of investments and analyzing tradeoffs . In February 2019, GAO found that the Coast Guard had used a model to determine that it could more efficiently prioritize its investment in aviation pavement—one segment of an almost $3 billion portfolio of aviation shore infrastructure—and save about $13.8 million. However, as of February 2019, the agency had not implemented the aviation pavement study results. Moreover, according to Coast Guard officials, the agency could employ models to its entire portfolio of shore infrastructure. By not implementing the results of its aviation pavement model or employing similar models across its shore infrastructure assets, the Coast Guard is missing opportunities to potentially identify and achieve cost savings across other assets. Dispose of unneeded assets. In October 2017, GAO found that closing boat stations that the Coast Guard had found to be unnecessarily duplicative could potentially generate $290 million in cost savings over 20 years. However, in February 2019, GAO found that instead of closures, the Coast Guard was planning recapitalization projects at 5 of the 18 stations it had recommended for closure. Given the Coast Guard's competing shore infrastructure priorities and existing project backlogs, GAO recommended disposing of unneeded assets to more efficiently manage resources and better position the Coast Guard and Congress to address shore infrastructure challenges. Implement DHS's Critical Infrastructure Risk Management Framework. In September 2019, GAO found that DHS has recognized the importance of protecting critical infrastructure from extreme weather and other risks. However, the Coast Guard has not fully aligned its processes for improving shore infrastructure resilience with DHS's five key steps for critical infrastructure risk management. For example, when identifying and then assessing risks to its infrastructure—two of the steps in the DHS process—the Coast Guard did not identify all assets that are critical to its missions, such as aircraft runways, or screen them for all vulnerabilities, such as flooding. Aligning its processes with the DHS steps would provide greater assurance that the Coast Guard is investing its resources to minimize potential damage and expenses caused by future extreme weather events. What GAO Recommends In the three reports, GAO made 10 recommendations to improve the Coast Guard's asset management efforts, including employing models for predicting investment outcomes, disposing of unneeded assets, and implementing DHS's critical infrastructure risk management framework to guide shore infrastructure resilience decisions. DHS concurred and generally described planned actions to address these recommendations, but has not yet fully implemented them.

gao_GAO-20-132

gao_GAO-20-132_0

Background The FSS program is directed and managed by GSA and provides federal agencies with a simplified process for obtaining commercial supplies and services at prices associated with volume buying. Schedules are catalogs of related products and services, from pre-approved vendors, with established pricing that can be used by federal agencies to obtain goods and services, ranging from office furniture to medical equipment and supplies. Since 1960, GSA has delegated authority to VA to manage health care related schedules, currently totaling nine schedules. (Throughout this report, we use the term “FSS” to refer to VA’s FSS program, unless otherwise noted.) These nine VA schedules, as shown in figure 1, are designed to provide FSS users at VA and other agencies with a menu of items—including medical equipment, supplies, and services—they can order from in a streamlined manner. Sales to VA on the pharmaceutical schedule were $33.5 billion from fiscal years 2014 through 2018. We omitted the pharmaceutical schedule from our review because it differs substantially from the other eight schedules, particularly in its use of a prime vendor. How the VA’s FSS Program Is Managed VA’s FSS program is managed by the National Acquisition Center (NAC), a VA-wide contracting organization which is also responsible for procuring items like high-tech medical equipment for medical centers. NAC is part of VA’s Office of Procurement, Acquisition and Logistics, which is overseen by VA’s Office of Acquisition, Logistics, and Construction. Within NAC, the FSS Service, which is comprised of about 80 staff, is divided into teams of contracting staff who are responsible for individual schedules. Another team, NAC’s Program Management and Resource Support, manages functions such as issuing guidance and providing training. In 2016, NAC issued a Procedural Guideline that generally sets a goal for its contracting staff to complete their review of and award decision for vendor-submitted FSS offers within 180 calendar days. Like GSA’s FSS program, users of the VA FSS program are charged a fee on the price of their FSS purchases, called the Industrial Funding Fee (IFF). VA’s FSS fee is 1 percent for services and 0.5 percent for goods. Fees generated by VA FSS fund its operations and other VA procurement operations. NAC facilitates collection of the IFF from vendors that sell products or services under VA FSS contracts, and vendors remit the IFF to VA’s Supply Fund, a self-supporting revolving fund. The Supply Fund, in turn, is used to provide funding to NAC for the operation of the FSS Service, the office that manages VA’s FSS program. VA FSS Users VA’s schedules are used by organizations across the federal government, including the Department of Defense, Department of Health and Human Services, and the Department of Homeland Security. In this review, we focus on how the Veterans Health Administration (VHA) uses VA’s schedules. VHA, the only VA administration that uses the VA schedules, provides medical care to about 9 million veterans at 170 medical centers. These medical centers are organized into 18 Veterans Integrated Service Networks (VISN), organizations that manage medical centers and associated clinics across a given geographic area. Each VISN is served by a corresponding Network Contracting Office (NCO), which is responsible for awarding contracts for medical goods and services that the medical centers need. Two primary groups of VHA staff place FSS orders: 1. VHA contracting officers, who are authorized to enter into contracts on behalf of the government, may place orders against the schedules. They handle purchases over the micro-purchase threshold, which is generally $10,000. 2. Certain VHA medical center logistics staff are authorized to make smaller purchases at or below the micro-purchase threshold, including placing FSS orders. Many of these staff are in the medical centers’ logistics offices, which are responsible for managing the supply chain for VHA’s medical centers. Figure 2 provides an overview of VA’s procurement structure and FSS users. VA’s Medical-Surgical Prime Vendor Program In addition to purchasing goods and services through FSS, VHA logistics staff at VA’s medical centers can also buy them through the MSPV-NG program. In this program, VA medical centers use contractors called medical-surgical prime vendors to obtain many of the supplies they use on a daily basis, such as bandages and scalpels. These prime vendors operate local warehouses and deliver supplies ordered by medical centers. The prices for these medical supplies are established by separate contracts or agreements that are awarded by contracting officers within VA’s Strategic Acquisition Center (SAC). The MSPV-NG program is managed by SAC and VHA. As we reported in 2018, for over a decade, each medical center used VHA’s legacy MSPV program to order medical supplies. Many of those items were purchased using VA’s FSS, which provided medical centers with a great deal of flexibility to order from a catalog containing hundreds of thousands of items. However, this flexibility prevented VHA from standardizing the items used across its medical centers and also affected its ability to leverage its buying power to achieve greater cost avoidance. In December 2016, VHA transitioned to a new iteration of this program called MSPV-NG, which has a narrower catalog of medical supplies than the legacy program, which offered hundreds of thousands of items. As of September 2019, the catalog offers about 22,000 supply items to medical centers. Veterans First Contracting Program Requirements In June 2016, a Supreme Court decision clarified that VA must apply the Veterans First Contracting Program preference before contracting with a non-veteran-owned business, including purchases made through FSS. This program, referred to in this report as Veterans First, provides preference within VA for contracting with veteran-owned small businesses. Specifically, VA contracting officers must apply the “VA Rule of Two,” meaning they must conduct market research to determine whether there is a reasonable expectation that two or more veteran- owned small businesses will submit offers for a particular good or service at a fair and reasonable price that offers best value to the government. If two or more such businesses are found, contracting officers must set aside the procurement for the veteran-owned small businesses. Use of VA’s FSS Was Flat While Overall VHA Spending Rose, but NAC Lacks Controls to Verify Sales Data and Visibility into Small Business Participation and User Experience VHA used NAC’s FSS to purchase billions of dollars in medical supplies and services over the past 5 years. Sales for the eight non- pharmaceutical schedules, however, have been largely flat, as compared to the rise in VHA’s total health care spending. Though we found vendor- submitted sales reports to be sufficiently reliable for describing overall trends, we found that NAC does not have controls in place to ensure that vendors are providing complete data—used to calculate fees that finance the program. In an attempt to assess data completeness, NAC recently began comparing vendor data to Federal Procurement Data System-Next Generation (FPDS-NG) data for verification. However, because agencies do not report micro-purchases made via purchase card in FPDS-NG, this approach alone is not effective. We also found that NAC does not analyze existing data on the number of veteran-owned small businesses that hold FSS contracts, the types of goods and services they offer, or which schedules have the most or least participation by these businesses. This information is important because VHA contracting officers must apply the Veterans First preference to contracts. The existence or lack of veteran-owned small businesses on FSS affects whether these staff can use FSS to fulfill their needs. Finally, we found that NAC has limited visibility into the FSS user experience. Those insights could help NAC identify potential areas for improvement. Billions of Dollars in Medical Supplies and Services Are Purchased through VA FSS, but Sales Are Flat amid a Rise in Overall VHA Medical Spending VHA obligated $291 billion from fiscal years 2014 to 2018 for health care services provided at its medical facilities—$12 billion of which was for medical supplies and services obligated using the eight non- pharmaceutical VA schedules. In contrast to VHA obligations for health care at its medical centers, which increased nearly 20 percent during this 5-year period, VA FSS purchases on these schedules were flat. See figure 3. We found that the VHA sales trends varied among our three selected schedules during this period, as shown in figure 4. Specifically, VHA sales on the FSS for Medical Equipment and Supplies, the largest of the three, were generally flat. VHA sales on the FSS for Patient Mobility Devices, which includes items such as wheelchairs, increased nearly 50 percent, while sales on the Healthcare Staffing schedule fell by more than 30 percent. VA Lacks Controls to Ensure That Vendors Provide Complete Data NAC does not have controls in place to ensure that vendors provide complete data in their sales reports. These sales reports—required per an FSS contract clause—are NAC’s only means of tracking FSS sales and related fees that finance the FSS program. In fiscal year 2018, vendors on VA’s nine schedules remitted $82 million in IFF from customers to VA’s Supply Fund. Figure 5 provides an overview of key steps in VA’s FSS vendor sales report and IFF collection process. Like NAC, GSA also depends on vendor-reported data to track its FSS sales. However, we found that GSA takes additional steps for its FSS program to ensure the completeness of vendor-reported data. GSA has internal controls to ensure data completeness, including a staff of 43 Industrial Operations Analysts who implement GSA procedures to ensure that, among other things, vendors have sound sales data reporting processes. We did not evaluate GSA’s use of these analysts, but, according to GSA FSS officials, these analysts review vendor sales data, educate vendors about GSA’s requirements, and conduct checks on vendor internal controls and compliance with GSA policies. Having internal controls in place is essential to ensure completeness of vendor-reported sales data. In February 2019, NAC officials told us they had tried to use obligation data reported in FPDS-NG to verify the completeness of vendor sales data. However, they found that FPDS-NG did not contain a substantial portion of vendor sales. These officials stated that the difference between the vendor sales data and the obligations in FPDS-NG was likely due in large part to purchases under the micro-purchase threshold (currently $10,000) that agencies do not report to FPDS-NG. Because agencies do not report micro-purchases made via purchase cards to FPDS-NG, per the Federal Acquisition Regulation, this approach alone is not effective to ensure data completeness. To estimate what portion of vendor-reported sales were below the micro- purchase threshold, we compared sales data that vendors reported to VA to the obligations included in FPDS-NG data for fiscal year 2018 for the eight schedules we reviewed. We found that 54 percent of VA FSS sales as reported in vendor data were not included in FPDS-NG. VA procurement officials we interviewed told us that the 54 percent were likely micro-purchases made by medical center logistics staff using their government purchase cards. For a more detailed view, we also reviewed the percentage of fiscal year 2018 sales included in FPDS-NG for our three selected VA schedules, and found that 65 percent and 71 percent of FSS sales for the Medical Equipment and Supplies and the Patient Mobility Devices schedules, respectively, were not included in FPDS-NG, as shown in figure 6. Finally, we compared FPDS-NG data to FSS sales reports for non-VA agencies. We found instances where obligations in FPDS-NG reported by these agencies exceeded those reported by vendors in FSS sales reports—sometimes significantly. Specifically, from fiscal years 2014 to 2018, FPDS-NG reflected a cumulative $533 million more than the sales that vendors reported to NAC for non-VA agencies for the Healthcare Staffing schedule. This difference between reported sales and obligations indicates a risk that vendors are under-reporting VA FSS sales to other agencies. Standards for Internal Control in the Federal Government require that management have adequate controls for ensuring the quality of data. NAC FSS officials told us that they would like to do more to mitigate the risk that vendors may not be reporting complete VA FSS sales data, given the differences we found and that NAC also found between reported sales and obligations. Additional internal controls, similar to the process used by GSA, would help NAC to ensure vendor-reported sales are complete, and that the appropriate IFF sales fees that finance the FSS program are collected. NAC Lacks Visibility into the Extent of Veteran- Owned Small Business Participation in the VA FSS Program We found that NAC does not assess data on the participation of, and items and services offered by, veteran-owned small businesses in NAC’s FSS program. This information is important because VHA contracting staff must apply the “VA Rule of Two” preference before contracting with a non-veteran-owned business. This preference for veteran-owned small businesses under the Veterans First program—which VA implemented more expansively after the 2016 Supreme Court decision—has had a major impact on VA procurement, as we reported in 2018. Thus, the availability of information about goods and services offered by veteran- owned small businesses on FSS affects whether contracting officers can use FSS as a simplified means of making purchases. However, NAC officials do not track the types of goods and services offered by veteran- owned small businesses holding FSS contracts, or which schedules have the most or least participation by these businesses. In interviews for this review and from our prior work, 10 contracting officers told us they use FSS less often than in the past. They cited the Veterans First requirement and instances where their market research showed a lack of veteran- owned small businesses holding FSS contracts. Instead, these contracting officers said, they found the goods and services they needed from veteran-owned small businesses on the open market. In addition, officials with VA’s Office of Small and Disadvantaged Business Utilization told us that that they do not analyze existing data on veteran-owned small businesses to assess these businesses’ participation in the FSS program. To conduct our own analysis, we looked at March 2019 data for the three selected VA schedules and found that goods and services provided by veteran-owned small businesses ranged from 11 to 23 percent of all the line items offered on these schedules, as shown in figure 7. According to the Standards for Internal Control in the Federal Government, program officials need quality information on how well their programs are serving end users—in this case, VHA contracting officers. Without analyzing veteran-owned small business participation in its FSS program, NAC cannot assess whether its program is meeting the needs of its users in light of the Veterans First preference, which requires contracting officers to apply the VA Rule of Two before purchasing through a non-veteran-owned business. Because the number of veteran- owned small businesses available on FSS directly affects how often contracting officers are able to use FSS, taking steps to better understand these data would enable NAC to, if necessary, make adjustments to its program to ensure contracting officers can use FSS as a regular, reliable, and simplified source for obtaining goods and services. NAC Has Limited Visibility into User Experience with the FSS Program NAC is not consistently obtaining and analyzing feedback from FSS users on their experience with the FSS program, despite having some tools in place to gather such information. To provide support to users, NAC’s website provides links to the FSS Help Desk and to a customer survey, among other contact information. However, while these tools could be used to gather feedback on whether the products and services offered on VA’s schedules meet user needs, NAC has received minimal user feedback via these tools. FSS leadership acknowledged the importance of user feedback, and stated they would like to develop a more comprehensive feedback mechanism, such as email surveys sent to users on a periodic basis with questions specific to their FSS program experience. Without such a feedback mechanism, NAC officials lack information on users’ experience with the program that could provide insights on areas for improvement. These insights, including whether program improvements are needed, are especially important given that FSS sales did not keep pace with the increased VHA spending over the past 5 years. Standards for Internal Control in the Federal Government state that, in order to formulate a strategy and achieve program objectives, management needs quality information to make informed decisions and evaluate performance. NAC Faces Numerous Challenges Managing Its FSS Program The NAC FSS program office faces numerous challenges—some of which are VA-wide issues we have identified in prior reports—including inadequate training and leadership instability. For example, NAC FSS guidance and training for contracting staff is not comprehensive, which poses a risk of inefficient use of contracting staff. Further, limited collaboration between FSS leadership at both NAC and GSA has resulted in missed opportunities to share tools and practices. These and other challenges faced by NAC were further exacerbated by a 3-year leadership gap in the FSS program; these positions have since been filled. NAC FSS Program Guidance and Training for Contracting Staff Is Not Comprehensive NAC FSS Guidance Is Not Comprehensive The Federal Acquisition Regulation, along with GSA and VA’s FSS regulations and policies, form the basis for NAC’s management of the VA FSS program. NAC issues additional guidance to operationalize these higher-level policies into NAC’s FSS work processes. This internal guidance takes several forms, including Procedural Guidelines, FSS Bulletins, and Standard Operating Procedures. We found that NAC’s internal guidance does not provide contracting staff with a comprehensive overview of key aspects of their jobs, creating confusion for the staff that implements the guidance. For example, NAC contracting staff members we interviewed stated that, for offers from resellers and distributors without significant commercial sales, assessing price reasonableness was a challenge. They said NAC’s standard processes assume that commercial sales data would be available to form the foundation of price analysis, but no NAC guidance outlines how to approach this analysis for distributors and resellers that lack significant commercial sales. Additionally, several contracting staff we interviewed told us that FSS team chiefs and supervisors provide them guidance informally, which can create confusion and variation in applying requirements across the VA FSS teams. For example, one member of the contracting staff told us that some teams require vendors to submit new commercial sales data when exercising an option to extend an FSS contract. But he told us this is not the case across all of VA’s FSS teams. Standards for Internal Control in the Federal Government state that management should periodically review policies, procedures, and related control activities for continued relevance and effectiveness in achieving objectives or addressing related risks. Reviewing NAC FSS policies and procedures, including those given informally, will allow NAC FSS leadership to provide more comprehensive guidance to all FSS contracting staff on these basic, but critical, steps to help ensure an even application of the FSS offer review process. NAC FSS Training Is Not Comprehensive We found that training for NAC FSS contracting staff is not comprehensive, posing another challenge to NAC’s management of its FSS program; we also cited inadequate training when adding VA Acquisition Management to GAO’s High Risk list. NAC contracting staff attend training at VA’s Acquisition Academy, and several staff members we interviewed said they found it helpful. However, according to NAC officials, and based on our review of course materials and VA-wide training, the Academy does not provide FSS-specific training. There are differences between awarding and managing FSS contracts and other types of federal contracts. For instance, before awarding an FSS contract, NAC FSS contracting staff must take steps that in some cases are specific to the evaluation of FSS contract offers, such as seeking to obtain discounts from a vendor’s commercial pricelist that are equal to or greater than the discounts offered to the vendor’s most favored customer. In addition, while NAC has provided some FSS-specific training to its contracting staff, such training is not offered on a regular basis, and no overall FSS training program or curriculum exists to guide NAC training for contracting staff. According to the NAC FSS official responsible for training, the last comprehensive office-wide FSS training provided to staff was about 3 years ago (since then, 16 new contracting staff have joined NAC FSS). We interviewed 16 other NAC contracting staff, and six of them expressed the need for more extensive training on FSS-specific contracting. In mid-2018, the NAC FSS Director created Team Lead positions to help train and mentor FSS contracting staff. This effort is still in development. The NAC official responsible for training, among other things, stated that other training and mentoring efforts are underway since he joined NAC in November 2018, specifically among individual NAC schedule teams, which sometimes provide informal training to their contracting staff. However, some of these informal training efforts are not offered on a consistent basis. For example, a senior member from the contracting staff on the Medical Equipment and Supplies schedule team provided a series of training sessions to his team on evaluating offers from distributors and resellers without significant commercial sales to the general public. However, this training was not provided to all NAC FSS contracting staff, even though contracting staff working on most of the eight schedules must review offers from distributors and resellers. NAC officials are in the early stages of exploring ways to improve training for their contracting staff. In the summer of 2019, NAC officials told us they plan to post materials from all prior training on NAC’s intranet. They also developed new orientation training for the nine new contracting staff who joined NAC in July 2019, covering topics such as FSS policies and where to find them, as well as an introduction to contract systems. The NAC FSS official responsible for training emphasized that these topics were covered at a basic level and said he would like to develop more comprehensive training that would be offered on a consistent basis. Standards for Internal Controls in the Federal Government state that well- trained staff, among other things, are essential to effective program execution. In addition, GAO’s guide for assessing federal government training states that in order to ensure training is effective, training programs should be guided by an overall strategy, informed by assessing priorities and evaluating results. While NAC has taken some steps to improve training for FSS contracting staff, NAC has yet to implement a comprehensive and consistently offered FSS training curriculum. Doing so could enable NAC to provide its staff with the tools and clarity needed to perform their roles and increase efficiency. Collaboration between NAC and GSA Is Limited, and GSA Does Not Have Current Documentation of Its Delegation of Authority to VA We found that NAC faces challenges effectively collaborating with GSA, the agency that oversees all FSS for the federal government. GSA has longstanding processes and established tools—such as its use of analysts to review vendors’ internal controls, as well as its automated offer-intake system—stemming from its decades of experience running an FSS program that is larger than VA’s program. However, collaboration and knowledge-sharing between NAC and GSA is limited. For example, in the past, NAC and GSA held meetings quarterly, but since 2015, these meetings have been held on an ad hoc basis. Neither organization took action until recently to ensure that meetings continued at regular intervals. GSA officials told us that during 2016 through 2018, they met with NAC a number of times in response to questions from NAC FSS officials. However, these meetings covered general policy questions, and according to NAC, did not focus on discussing cross-agency roles and responsibilities or on sharing practices for managing the FSS program. Separately, VA Office of the Inspector General’s Office of Contract Review officials told us of a 2010 working group formed to collaboratively discuss revisions to GSA’s regulations, which included representatives from GSA, NAC, the VA Office of the Inspector General, and others. According to these VA Inspector General officials, this group was disbanded about a year after it began due to disagreements among the participants. Upon NAC’s FSS Director’s request, in February 2019, NAC and GSA resumed quarterly meetings. However, NAC officials noted instances where collaboration is still limited. For example, GSA did not provide NAC officials with advance notice about the publication of a final rule establishing changes to GSA’s FSS regulations that were relevant to NAC’s administration of its FSS program. NAC discovered the final rule had gone into effect after it was published, independent of any communication from GSA. NAC officials stated they would like to have additional opportunities for input on GSA regulations that affect the VA FSS program. Our prior work has found that clearly defining roles and responsibilities is a key practice for cross-agency collaboration. Without a clear and shared understanding of their respective responsibilities, and processes to ensure they share tools and practices, NAC will not have the opportunity to learn from GSA’s experience or have timely input on GSA actions that affect the VA FSS program. GSA and VA are also missing a document—namely, GSA’s updated delegation of authority to VA—that could guide their collaboration efforts. This delegation should state what authority is granted to VA, and cite the limitations on that authority. We found a January 2008 Federal Register notice mentioned a 2004 update to the delegation, but the GSA Director of Policy for the Federal Acquisition Service was unable to locate or provide this update. VA was also unable to locate a copy of the 2004 update. Instead, GSA and VA gave us a number of documents, including memorandums and other communications that spanned from the 1960s to the 1990s. The documents were fragmented and outdated. Further, many of these older documents referred to outdated laws, regulations, or organizations, raising questions about their current applicability. Standards for Internal Control in the Federal Government state that the roles of those responsible for carrying out programs should be clearly outlined in policy, and GAO has also reported that written guidance and agreements on collaboration are key features of successful cross-agency collaboration. The lack of current documentation related to GSA’s delegation to VA, alongside the limitations in NAC and GSA communication, undermine a firm foundation on which to build collaboration. Without a clear delineation of roles and responsibilities—and effective overall coordination—NAC and GSA risk misunderstandings and missed opportunities to share information and tools that could improve NAC’s management of the VA FSS program. Key FSS Leadership Vacancies Spanned 3 Years From 2015 until 2018, senior VA FSS leadership positions were vacant, which affected VA’s FSS program management and directly contributed to many of the challenges we identified above. Namely, the FSS program director position was vacant for over 2 years and the role of FSS Program Management and Resource Support team chief was vacant for about 19 months. During that time, chiefs of individual VA schedules held the Director or Chief positions on an acting and rotational basis. During these rotations, these chiefs were dual-hatted as they maintained responsibility for their primary job role. The Associate Executive Director of the NAC stated that he was reluctant to make long-term, strategic policy decisions while the FSS Director position was vacant. In late 2017 and late 2018, respectively, NAC permanently filled these two FSS program positions. However, by then, broader changes had taken place within VA contracting that affected the VA FSS program: namely, the Supreme Court ruled in 2016 that before VA may contract with a non-veteran-owned business, VA must apply the “VA Rule of Two,” including instances when VA makes purchases through FSS. Also in late 2016, VA launched the MSPV-NG program, which offers items similar to those items offered on two VA schedules. Figure 8 provides a timeline of these FSS leadership vacancies and events. Although both of these leadership positions have since been filled, the effect of the gaps is still evident in some cases. According to NAC FSS officials, hiring to fill open FSS contracting staff positions was slowed by the leadership gaps, which added to workload pressures; more contract offers were received than completed in fiscal years 2015 and 2016, creating a backlog. Contracting staff workload is a VA-wide issue we previously identified, and is one of the areas of concern we cited in adding VA Acquisition Management to GAO’s High Risk List in 2019. In late 2018, the FSS Director sought approval for 10 additional contracting staff; 9 of these positions were filled in July 2019. NAC leadership stated that these positions should help address some of the backlog faced by FSS contracting staff. NAC Rarely Meets Its Timeliness Goals for FSS Contract Awards NAC has experienced major delays in awarding vendor contracts and missed its timeliness goal for contract award 75 percent of the time from fiscal years 2014 through 2018. NAC’s inefficient offer intake system and fragmented vendor guidance likely contributed to these delays. NAC FSS leadership has acknowledged these challenges and is working to address some of them. Assessing the appropriateness of these timeliness goals and taking steps to comprehensively identify and address barriers to achieving them will better position NAC’s contracting workforce to improve contract award timeliness. NAC Has Not Awarded FSS Contracts in a Timely Manner Our analysis shows that from fiscal years 2014 through 2018, NAC did not meet its timeliness goals for 75 percent of its FSS contract awards. To do this analysis, we compared NAC FSS contract award data for the eight non-pharmaceutical schedules against the timeliness goal of 180 calendar days for contracting staff reviews and decisions on vendor contract awards, as set forth in a NAC Procedural Guideline. Specifically, we found that 319 of the 803 FSS contract awards took at least double the 180-day goal. In addition, 12 of them exceeded the goal by more than 1,080 days—six times the goal. During fiscal years 2015 and 2016, the program accumulated a backlog of FSS offers, which coincided with a vacancy in the FSS Director position starting in October 2015. While NAC staff made some progress on mitigating this backlog, timeliness remains an issue. For example, during fiscal year 2018, NAC missed its timeliness goal 73 percent of the time. Figure 9 portrays NAC’s timeliness of FSS contract awards over this 5-year period. We also analyzed the timeliness of awards for the three selected VA schedules from fiscal years 2014 through 2018, as shown in figure 10. This analysis shows that NAC consistently missed its timeliness goals across the three different schedules. However, for contract modifications—typically changes to contract items or prices—NAC met its timeliness goal—set at 60 calendar days—80 percent of the time over this 5-year period. We reviewed data on about 14,000 modifications executed by NAC for the eight non-pharmaceutical schedules from fiscal years 2014 to 2018. About 2,300 of these modifications were executed to add new items to existing contracts. For these, NAC met the timeliness goal only 54 percent of the time. The ability to quickly add new items to FSS contracts is important to ensure that agency users have access to up-to-date medical supplies and services through the VA FSS program. These timeliness goals apply across all of NAC’s eight non- pharmaceutical schedules, regardless of how complex a contract award or modification might be. Various factors can affect contracting staff’s ability to meet these goals, including the complexity of the award, staff’s workload, and whether or not vendor documentation is complete. NAC FSS leadership has acknowledged these challenges and is working to address some of them. However, NAC has not assessed if the current timeliness goals are appropriate, or performed a comprehensive assessment of the barriers that prevent FSS contracting staff from achieving timeliness goals. Standards for Internal Control in the Federal Government state the importance of management making well-informed decisions and conducting meaningful evaluations of their organization’s performance. Assessing the appropriateness of current timeliness goals and taking steps to comprehensively identify and address barriers to achieving them will better position NAC contracting staff to meet these goals. Moreover, timelier contract awards enable medical centers to obtain needed goods and services and, as a result, help FSS to remain useful to medical centers. Inefficient Offer-Intake Process and Fragmented Vendor Guidance Likely Contributed to Delays When seeking a VA schedule contract award, NAC processes require vendors to submit an offer and required documents. Because NAC’s offer intake system is not automated, NAC officials must manually check a general FSS email inbox for vendor submissions and manually review the vendor’s offer and required documents to determine if all information is included. Further, there are no automated checks for completeness of vendor documentation. We analyzed a non-generalizable sample of 26 selected FSS contracts awarded beginning in fiscal year 2014 through January 2019 on three schedules—Medical Equipment and Supplies, Patient Mobility, and Healthcare Staffing—and found that in 14 instances, VA contracting staff identified incomplete documentation and had to follow up with the vendors to receive revisions. NAC officials told us that tracking vendor offers and associated documents from email is cumbersome and time consuming because they have to sort through several separate vendor email messages to splice together vendor offer submissions, due to file size limitations. The inefficient offer intake process also led to delays in assigning offers to contracting staff. NAC did not always assign offers to contracting staff immediately after vendor submission, delaying the start of work. Our review of 26 selected VA FSS contract files identified 10 instances where NAC took more than 20 days after receipt to assign the offer to contracting staff. NAC officials told us that these delays were caused by both the non-automated offer-intake process as well as the team chiefs’ lack of time to assign these offers to contracting staff for their review. According to the FSS Director, in mid-2018 he created a team lead for each NAC schedule team to assign offers to contracting staff and monitor these offers to better ensure timeliness. Figure 11 summarizes key steps in NAC’s FSS manual offer intake and award process, as described by FSS contracting officials. In contrast to VA’s manual system, since 2004, GSA has used an online system called eOffer to manage its FSS offer intake process. This system includes automated system checks to ensure documentation is complete before it is submitted by vendors. Once offers are submitted, supervisors in GSA FSS offices review offers in the system and assign them to contracting staff for review. We have not evaluated whether the eOffer system increases efficiency or reduces errors in submitted offers, but, according to GSA officials, eOffer achieves efficiency and accuracy due to the automated checks that will not let vendors submit an incomplete offer package. NAC contracting staff told us that they could benefit from a more efficient system to accept offers from vendors. In late 2018, NAC and GSA discussed the possibility of adopting GSA’s eOffer online system, as well as its companion eMod, which is used to process modifications. In November 2018, GSA’s estimate to add VA to the system was about $9 million for the first year, and nearly $8 million annually thereafter, which, according to NAC’s FSS Director, is cost-prohibitive. GSA officials told us that to determine this cost, they compared the number of NAC FSS contracts to the number of GSA FSS contracts, and apportioned 10 percent of the overall system development and operation cost to VA. Despite the cost of GSA’s systems, if VA does not address limitations in its own manual offer-intake process, such as implementing a system that can provide automated checks for completeness, delays in assigning offers to contracting staff will continue. Further, FSS contracting staff will continue to spend additional resources and time to gather and complete offer documentation before they can determine whether to award the contract. We also reviewed NAC’s website and found that guidance for vendors was fragmented. Pieces of guidance were spread out among a number of documents as opposed to being located in one document or section of the website for vendors to easily locate. This also contributes to vendors submitting incomplete offer documentation, which, in turn, contributes to delays in contract awards. Incomplete documentation for pricing and sales data is particularly common—namely, information that enables contracting staff to compare the prices vendors offer the government and commercial customers for the same goods. As stated in Standards for Internal Control in the Federal Government, clear communication with outside parties, like vendors, is essential to ensuring that NAC is able to help achieve its program objectives. Clearer guidance would provide vendors with a reminder of program requirements that could reduce VA FSS contracting staff review time and improve their efficiency in reviewing contract offers. VA Has Not Assessed Whether FSS and MSPV-NG Program Duplication Is Necessary or Efficient Over the past few years, the FSS and MSPV programs have transitioned from functioning together to existing as separate programs serving similar VA medical center needs. However, VA leaders have not assessed if the overlapping offerings are a necessary and effective use of resources, or allow VA to fully leverage its buying power—a stated goal of both the MSPV-NG and FSS programs. As we reported in November 2017, for over a decade, VA’s medical centers used VHA’s legacy MSPV program to order medical supplies—many of which were purchased using NAC’s FSS program. When VHA transitioned to its MSPV-NG program in late 2016, it significantly narrowed the catalog to 6,000 items, and VHA contracting officials told us that they modified the contracting approach in March 2018 to have the prime vendor supply the items directly, separate from FSS. At the outset, VHA set goals for the MSPV-NG program, including standardization of requirements for supply items and cost avoidance by leveraging VA’s substantial buying power. However, the MSPV-NG program recently revised its goals from focusing on standardization to increasing the number of catalog items available for medical centers’ use—the catalog contains more than 20,000 items as of September 2019. We compared the MSPV-NG catalog to the VA Medical Equipment and Supplies schedule to determine whether they offered similar products, and found overlap. For example, we found that as of June 2019, about two-thirds (139 of 206) of the MSPV-NG catalog suppliers also offered items on the Medical Equipment and Supplies schedule. Also, in March 2019, NAC FSS leadership provided analysis to the MSPV-NG program office showing that 41 percent of items that the MSPV-NG program planned to include in an update of the MSPV catalog were already available under VA FSS contracts. This duplication could result in inefficiencies whereby different sets of contracting staff within the FSS and MSPV-NG programs award, modify, and manage contracts for the same or similar medical supplies for VA medical center use. The MSPV-NG program office is currently developing the next iteration of the program, called MSPV 2.0, which it plans to roll out in February 2021. In April of 2019, a senior VHA procurement official announced at a vendor forum that the VA FSS program would be used as a source for its MSPV 2.0 supply catalog. However, in June of 2019, MSPV-NG program officials and VHA procurement leadership told us they decided against using FSS for this purpose and provided several reasons for this decision. First, these officials stated that FSS was not comprehensive enough to fulfill the MSPV 2.0 catalog; as noted above, FSS could provide about 40 percent of needed items. These officials also stated that the effort needed to create new FSS contracts or add new items to existing VA FSS contracts to fulfill the remaining 60 percent of the required MSPV 2.0 catalog would be too time consuming. Further, these officials also stated that there were not enough veteran-owned small businesses that offer items on VA FSS to support the MSPV 2.0 requirements. They stated this could result in extra time and resources to solicit both within FSS and the open market to ensure that they meet the VA Rule of Two. NAC FSS leaders told us that they communicated their willingness to support the MSPV 2.0 program by offering to work with vendors to quickly add the needed items; however, the MSPV-NG program office did not involve them in their final decision not to use FSS as a source for the MSPV 2.0 program. VA procurement leaders have informally discussed the future of the FSS program, according to a senior VHA procurement official. According to this official, VA has not determined whether it will change the strategy for FSS, or if the duplication between the FSS and MSPV-NG programs is a necessary and efficient use of resources. VA’s Strategic Plan for Fiscal Years 2018-2024 calls for related efforts to be coordinated with each other to achieve cross-organizational unity of purpose. When adding VA Acquisition Management to our High Risk List in March of 2019, we reported that VA lacks an effective medical supplies procurement strategy. While this finding stemmed from our review of VA’s MSPV-NG program and was related to the recommendation that VA develop an overarching strategy for this program, the same applies for VA’s FSS program in that VA does not have a strategic approach for its procurement of medical supplies through these two programs. Further, we reviewed VHA’s Modernization Campaign Plan, dated March 2019, and VHA’s Modernization Plan briefing slides, dated October 2019, which describe several modernization initiatives. One of these initiatives is to transform the supply chain through modernization. This modernization plan includes the planned MSPV 2.0 program and VA’s planned changes to its supply chain management system; however, it does not include FSS. As VA is undertaking these efforts, it is unclear how and whether FSS fits into VA’s vision of a modernized supply chain. Taking steps to assess VA FSS and MSPV program duplication will allow VA to determine if it is efficiently using its contracting staff. Moreover, communicating its decision to managers of these two programs will allow these managers to focus and coordinate their resources accordingly. This assessment will also help VA determine if it is leveraging its buying power to improve the effectiveness and efficiency of services for veterans and their families. Conclusions The continued utility of parts of VA’s FSS program is in question amid flat sales in recent years and competing programs available to medical center staff for supplies, such as MSPV-NG. VA has the opportunity to improve its FSS program by ensuring that it has complete vendor sales data and better information on participation by veteran-owned small business and user experiences. Obtaining such information would enable NAC to ensure it is collecting all fees it is owed which support program operations, and ensure that FSS can remain a regular, reliable and simplified source for contracting officers to obtain goods and services on behalf of the medical centers. Other steps are necessary, however, to address challenges VA faces with its FSS program. Specifically, the FSS program needs to provide comprehensive guidance and training to its contracting officers, and assess timeliness goals and barriers to achieving these goals to ensure the program remains useful to customers—namely medical centers that rely on the goods and services provided by FSS. In working to improve its FSS program, NAC has the opportunity to gain insights and experience from GSA on how it manages its much larger schedules program. However, lack of collaboration between GSA and NAC has resulted in missing opportunities for such information sharing. Finally, both VA’s FSS and MSPV-NG programs support VA’s overall medical supply chain, yet VA has not assessed whether duplication between them is a necessary and effective use of resources. Without this assessment, VA could be missing opportunities to leverage buying power and improve efficiency in procuring goods and supplies for its medical centers. Recommendations for Executive Action We are making a total of 11 recommendations, including nine to VA and two to GSA: The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s NAC puts controls in place to better ensure the completeness of vendor FSS sales reporting. (Recommendation 1) The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s NAC assesses data on the participation of and items and services offered by veteran-owned small businesses in NAC’s FSS program, in order to determine whether their program is meeting the needs of VHA contracting officers who use it given the Veterans First requirements they must meet. (Recommendation 2) The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s NAC directs the FSS Director to develop a mechanism to consistently obtain and analyze VHA user feedback on the FSS program. (Recommendation 3) The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s NAC provides FSS contracting staff with comprehensive FSS guidance. (Recommendation 4) The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s NAC develops an FSS-specific training program to include essential skills and processes to meet ongoing training needs for new and existing contracting staff. (Recommendation 5) The Administrator of GSA should work with the Secretary of VA to develop a memorandum of understanding outlining the roles and responsibilities of GSA and NAC for collaborating under GSA’s delegation of authority to VA for the healthcare-related Federal Supply Schedules, including the processes through which the two organizations will coordinate and share useful tools and practices. (Recommendation 6) The Secretary of Veterans Affairs should work with the Administrator of GSA to develop a memorandum of understanding outlining the roles and responsibilities of GSA and NAC in collaborating under GSA’s delegation of authority to VA for the healthcare-related Federal Supply Schedules, including the processes through which the two organizations will coordinate and share useful tools and practices. (Recommendation 7) The Administrator of GSA should take steps to document its delegation of authority for the healthcare-related Federal Supply Schedules to VA. (Recommendation 8) The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s NAC assesses the appropriateness of NAC’s current timeliness goals for FSS contract awards and takes steps to comprehensively identify and address barriers to achieving them. (Recommendation 9) The Secretary of Veterans Affairs should ensure that the Associate Executive Director of VA’s National Acquisition Center takes measures to ensure greater efficiency in the offer-intake process, such as providing additional guidance for vendors or by adopting a system that includes checks for completeness of required vendor documentation. (Recommendation 10) The Secretary of Veterans Affairs should take steps to assess duplication between VA’s FSS and MSPV programs, to determine if this duplication is necessary or if efficiencies can be gained. (Recommendation 11) Agency Comments We provided a draft of this report to the Department of Veterans Affairs and to the General Services Administration for review and comment. In VA’s comments, reproduced in appendix II, it concurred with all of our nine recommendations. In GSA’s comments, reproduced in appendix III, it concurred with our two recommendations. We are sending copies of this report to the appropriate congressional committees. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Secretary of the Department of Veterans Affairs and the Administrator of the General Services Administration. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-4841 or by email at oakleys@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Objectives, Scope, and Methodology This report assesses: (1) what is known about the Department of Veterans Affairs’ (VA) use of its Federal Supply Schedule (FSS) program for fiscal years 2014 through 2018; (2) challenges the National Acquisition Center (NAC) faces in effectively managing the FSS program, (3) the extent to which NAC awarded FSS contracts in a timely manner from fiscal years 2014 through 2018, and (4) the extent to which FSS and the Medical-Surgical Prime Vendor-Next Generation (MSPV-NG) programs provide overlapping or duplicative offerings. To assess what is known about VA’s use of its FSS program from fiscal years 2014 through 2018, we analyzed quarterly vendor sales report data provided by NAC. To assess the reliability of these data, we reviewed documentation and other information on the internal controls of the data systems used by NAC to collect and verify VA FSS sales reports. We found these data to be sufficiently reliable for analyzing overall trends in sales on VA’s FSS for this time period. We also obtained contracting data from the Federal Procurement Data System-Next Generation (FPDS-NG) for fiscal years 2014 through 2018, as well as a listing of VA FSS contracts active during that period from NAC. We used these data to analyze contract obligations on VA FSS contracts over this time period. Further, we compared total obligations based on FPDS-NG data to vendor-reported sales data provided by NAC by schedule and fiscal year. We found the FPDS-NG data sufficiently reliable for our purpose of comparing reported obligations to vendor sales data. We also used information obtained from a prior GAO review of VA’s Veterans First Program when discussing FSS use and used Standards for Internal Control in the Federal Government as criteria to assess this use. We selected the Medical Equipment and Supply (65IIA), Patient Mobility (65IIF), and Healthcare Staffing (621I) schedules as the focus of our review, based on total number of active contracts; they collectively represented about two-thirds of the approximately 1,700 active VA FSS contracts at the time we began our review. We excluded the pharmaceutical schedule from our review because, unlike the other schedules, orders are placed almost exclusively through the pharmaceutical prime vendor, and participation is a statutory requirement. We also reviewed and compared VA, Veterans Health Administration (VHA), and General Services Administration (GSA) policies, guidance, and memorandums related to the program and interviewed VHA- and VA- wide procurement officials regarding factors that affect use of the VA schedules, including the Veterans First program. We also analyzed data on items offered by vendors on the three selected schedules to determine the percent of items offered by veteran-owned small businesses. We interviewed VHA contracting staff and supply chain logistics staff at VA medical centers—users of the program—on factors that affect their use of FSS. We conducted site visits at a non-generalizable selection of two Veterans Integrated Service Networks (VISNs), visiting one medical center within each. Additionally, we interviewed Network Contracting Office (NCO) officials within each selected VISN, either in person or via telephone: VISN 12: VA Great Lakes Health Care System Clement J. Zablocki VA Medical Center (Milwaukee, Wisc.) Great Lakes Acquisition Center, NCO 12 (Milwaukee, Wisc.) VISN 10: VA Healthcare System Cincinnati, Ohio VA Medical Center NCO 10 (via telephone) We selected VISNs and medical centers primarily based on geographical proximity to NAC and GAO offices, as well for higher total obligations in fiscal year 2018. At each selected medical center, we interviewed the Facility Chief Supply Chain Officer and other members of the logistics staff. At each selected NCO, we interviewed leadership, branch chiefs, and contracting officers on teams that cover goods and services included on the three VA schedules we selected. Separately, we also spoke with representatives of the Coalition for Government Procurement, a group representing a number of FSS vendors, and attended a conference for vendors organized by NAC. To assess challenges NAC faces in effectively managing the FSS program, we reviewed GSA and VA procurement regulations, policies, and guidance as well as NAC FSS guidance. We analyzed the content of training offered by the Veterans Affairs Acquisition Academy and by NAC FSS. We also reviewed systems and processes used by NAC FSS staff to accept and review FSS offers and award contracts. We also interviewed NAC FSS leadership, contracting staff, and other staff regarding management of the FSS program during a site visit to NAC. We obtained and analyzed information on NAC FSS staffing, including leadership vacancies. We obtained documentation on analogous GSA practices for managing its FSS program, as well as documents delegating management of healthcare-related schedules to VA. We also interviewed officials in GSA’s Federal Acquisition Service who are responsible for overseeing its FSS program. To determine the extent to which NAC met its timeliness goal for processing FSS offers and modifications, we analyzed timeliness data collected by NAC for fiscal years 2014 through 2018, for the eight non- pharmaceutical schedules; we also performed limited analysis of timeliness for the pharmaceutical schedule, and additional analysis for our three selected schedules. We focused our analysis of timeliness on offers resulting in a contract award, because these are the cases that are relevant to users of the FSS program. To provide context for overall workload, we also analyzed timeliness for offers that were withdrawn, or where contracting staff decided not to make an award. We excluded offers that were reviewed by the VA Inspector General, Office of Contract Review, from our overall timeliness analysis because NAC policy does not count the time required for these reviews against its timeliness goal. To assess the reliability of timeliness data, we collected information on the system and processes used to maintain the data, performed electronic testing, and compared reported dates to source documents for selected contracts. We found these data sufficiently reliable for the purpose of assessing overall performance and trends in NAC FSS timeliness. From the three selected schedules, we selected a non-generalizable sample of 26 NAC FSS contracts awarded in fiscal years 2014 through January 2019. Eight of the contracts were randomly selected from all active contracts on the three schedules as of January 2019, while the remaining 18 contracts were selected by stratified random sample of contracts awarded in fiscal year 2018, focusing on those which exceeded the 180-day timeliness goal and omitting those with few or no sales. Thirteen of the contracts were under the Medical Equipment and Supplies schedule, seven of the contracts were under the Healthcare Staffing schedule, and the remaining six were under the Patient Mobility schedule. For each selected contract, we reviewed documents in the contract file; we also interviewed cognizant members of the contracting staff for 16 of the contracts. We selected this non-generalizable sample to provide illustrative examples of process steps and factors affecting timeliness; it was not the sole source of our findings on factors contributing to timeliness, which also included analysis of policies, guidance, and data, and interviews with NAC officials. To assess the extent to which the FSS and MSPV-NG programs provide overlapping or duplicative offerings, we reviewed policy and guidance related to both programs and interviewed VHA- and VA-wide procurement leaders. To assess the extent of overlap between the MSPV-NG and VA FSS catalogs, we also analyzed data on the items available through MSPV-NG and VA FSS, as well as the vendors participating in each, to assess extent of duplication. We interviewed VA officials from NAC, the Office of Acquisition and Logistics, the Strategic Acquisition Center, and VHA regarding the relationship between MSPV-NG and the FSS program. We analyzed policies related to these programs that affect management and use of VA FSS, and interviewed VA officials about their impact. We also reviewed documents and interviews with MSPV program office staff from an ongoing GAO review of the MSPV program. We used information obtained from an ongoing GAO review as well as published GAO reports on VA’s MSPV-NG program. Finally, we reviewed documents, including VA’s 2018-2024 Strategic Plan and VHA supply chain modernization plans. We conducted this performance audit from November 2018 to January 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the Department of Veterans Affairs Appendix III: Comments from the General Services Administration Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Shelby S. Oakley, 202-512-4841 or oakleys@gao.gov. Staff Acknowledgments In addition to the individual named above, Lisa Gardner, Assistant Director; Teague Lyons, Analyst-in-Charge; Erin Stockdale; Sarah Amer; Maurice Robinson; Emily Bond; Rashmi Agarwal; Andrew Burton; Virginia Chanley; Matthew T. Crosby; Susan Ditto; Lori Fields; Gina Flacco; Suellen Foth; and Alyssa Weir made key contributions to this report.

Why GAO Did This Study Through the FSS program, VA manages nine healthcare-related schedules—groups of contracts used to order medical supplies and services—under authority delegated by GSA. VA's FSS program management, including the speed with which it adds new contracts, affects VA medical centers' ability to use it to easily obtain goods and services. Further, recent changes in VA's medical procurement have also raised questions about the future role of the program. GAO was asked to examine VA's management and use of its FSS program. This report assesses (1) what is known about VA use of its FSS program for fiscal years 2014-2018; (2) program management challenges faced by NAC; (3) the extent to which NAC awarded FSS contracts to vendors in a timely manner from fiscal years 2014-2018; and (4) the extent to which the FSS and MSPV-NG programs provide overlapping or duplicative offerings. GAO reviewed eight VA schedules, excluding pharmaceutical due to the use of a prime vendor, among other things. GAO also analyzed three of these schedules representing about two-thirds of VA's FSS contracts; analyzed policies, guidance, and processes; and interviewed senior VA procurement, contracting, and supply chain logistics staff at NAC and two medical centers. What GAO Found Over the past 5 years, Department of Veterans Affairs (VA) medical spending increased, but spending on its eight non-pharmaceutical Federal Supply Schedules (FSS) was flat. GAO found the vendor-submitted sales reports to be sufficiently reliable for describing these trends. However, GAO found that VA's National Acquisition Center (NAC)—the VA-wide contracting organization responsible for FSS—lacks controls to ensure the completeness of vendor sales data, which is used to calculate fees that finance the program. The FSS program faces numerous challenges. For instance, NAC FSS guidance and training are not comprehensive, posing a risk of inefficiency and uneven application of requirements by contracting staff. Limited collaboration between FSS leadership at both NAC and the General Services Administration (GSA) also resulted in missed opportunities to share tools and practices. A 3-year FSS leadership gap further exacerbated challenges; these positions are now filled. NAC also failed to meet its 180-day timeliness goal for 75 percent of the non-pharmaceutical FSS contracts it awarded from fiscal years 2014 through 2018 (see figure), though NAC met its goal for contract modifications 80 percent of the time. By assessing timeliness goals and identifying barriers to achieving them, NAC leadership can take steps to better enable its contracting workforce to provide an efficient and reliable means to obtain needed goods and services through FSS. Moreover, VA's procurement leaders have not assessed, and communicated to program managers, whether the duplication between FSS and the Medical Surgical Prime Vendor-Next Generation (MSPV-NG) program is a necessary and effective use of resources. These two programs feature many of the same items, and different contracting staff manage different contracts for the provision of the same or similar medical supplies for VA medical centers. Without assessing duplication between these two programs, VA is at risk of inefficient use of its contracting workforce, and may be unable to fully leverage its buying power. What GAO Recommends GAO is making 11 recommendations: nine to VA and two to GSA; including that VA provide comprehensive guidance and FSS-specific training, improve NAC and GSA collaboration, evaluate timeliness goals and barriers, and assess FSS and MSPV-NG program duplication. VA and GSA agreed with GAO's recommendations.

gao_GAO-19-575T

gao_GAO-19-575T_0

Background Government-wide Reform Plan Requirements In March 2017, the President issued an executive order requiring comprehensive reorganization plans for executive branch agencies (see fig. 1). In April 2017, OMB provided guidance to federal agencies for developing their respective reform plans. The government-wide reform plan was to have been based on the agency reform plans, OMB-coordinated crosscutting proposals, and public input. According to OMB’s M-17-22 guidance, OMB, in coordination with the President’s Management Council, was to establish a way to track the progress of the reforms. OMB’s guidance also stated that it would track progress of the reforms by leveraging the federal performance planning and reporting framework originally put into place by the Government Performance Results Act of 1993 (GPRA), and significantly enhanced by the GPRA Modernization Act of 2010 (GPRAMA), through the use of cross-agency priority (CAP) goals, agency priority goals, and Performance.gov. In March 2018, OMB released the President’s Management Agenda (PMA), which provided information on the preliminary status of government reorganization efforts and is connected with these reform efforts. The PMA also identified a set of CAP goals, required under GPRAMA, to target those areas where multiple agencies must collaborate to effect change and report progress in a manner the public can easily track. The PMA gave OPM a key role in fulfilling the administration’s human capital-related goals. Specifically, OPM, along with OMB and the Department of Defense (DOD), were tasked to “align and strategically manage the workforce to efficiently and effectively achieve the federal government’s mission.” OPM Reorganization Proposals The administration is planning to transfer OPM’s background investigations to DOD, policy and workforce strategy functions to the EOP, and all remaining functions to GSA (see fig. 2). These remaining functions include human resource solutions, information technology systems, healthcare and insurance, retirement services, merit system accountability and compliance, and IG functions. The President’s fiscal year 2020 budget proposal, which was issued in March 2019, states that the administration is planning to complete the reorganization of OPM by the end of fiscal year 2020. As such, the budget proposal provided no funds for OPM for fiscal year 2020. According to that budget proposal, “the Administration has been developing plans to execute transfers of OPM functions to GSA and the DOD using a combination of existing legal authority and legislation” since June 2018. The budget proposal also requested $50 million to transfer certain OPM functions to GSA, including an additional $1 million to cover costs associated with merging the OPM IG with the GSA IG. On May 16, 2019, the administration proposed new legislation requesting authority to fully implement its reorganization proposal. OMB, OPM, and GSA Have Not Fully Addressed Key Reform Practices in Reorganizing OPM Setting Goals and Measures, and Assessing Costs and Benefits As we previously reported, a critical first step in the reform and reorganization process is to define the benefits of the merger, and describe how the future will be both different from and better than the past. As of May 17, 2019, OMB, OPM, and GSA had not fully established outcome-oriented goals and performance measures for, or assessed the costs and benefits of, the administration’s proposal to reorganize OPM (see fig. 3). Specifically, GSA provided one document, a draft Qualitative Business Case and Value Proposition for the GSA/HRS Merger (October 2018), which includes some preliminary goals and measures, such as to improve customer satisfaction. However, this document focuses only on the goals and measures related to the transfer of human resources solutions from OPM to GSA, rather than on the entire reform proposal. In addition, that document explicitly states that it is not a cost-benefit analysis, and OMB staff have told us that they have not conducted a cost-benefit analysis of the reform. In our prior work on organizational mergers and transformations, we have found that establishing a coherent mission and integrated strategic goals to guide the transformation involves adopting leading practices for results-oriented strategic planning and reporting. Leadership Focus and Attention We have previously reported that organizational transformations should be led by a dedicated team of high-performing leaders within the agency, and GSA has provided some evidence of this leadership focus and attention, but OMB, OPM, and GSA have only partially addressed this key practice (see fig. 4). According to GSA officials and documents we reviewed, the agency designated a member of its Senior Executive Service as the leader of the reorganization within GSA, and has established a Project Management Office with dedicated staff and resources which will take on the responsibility of supporting the transfer of OPM’s functions to GSA. Also, GSA officials told us that OMB leads the reform by, for example, leading meetings under the Six Sigma management approach to manage progress on implementing the reorganization. However, as of May 17, 2019, OMB did not provide documents we requested about the role of these management meetings for the reorganization, and OPM did not provide relevant information or documents. Our past work has also found that leadership should articulate a succinct and compelling reason for the reform, as this helps build morale and commitment to the organizational changes. OMB provided the case for change in several public documents, such as the government-wide reform plan, which primarily state that the administration’s reason for moving OPM’s functions to GSA and the EOP is that these changes would create greater efficiencies and elevate the importance of human resources policy. However, sharing the case for change is only one key factor in successful reforms and reorganizations. As we stated above, illustrating what success looks like is also important, and involves articulating the specific goals and costs and benefits of the reform. Involving and Communicating with Congress, Employees and Key Stakeholders Our prior work has shown that it is important for agencies to directly and continuously involve their employees, Congress, and other key stakeholders in the development of any major reforms. OMB and GSA have taken some actions to involve and communicate with Congress, employees, and other key stakeholders, but these initiatives lack documentation (see fig. 5). For example, GSA officials told us that they have met with members of Congress, conducted town hall meetings in which they provided information to and answered questions from GSA officials, and established an email inbox for communication between GSA leaders and employees on the reform. However, as of May 17, 2019, GSA officials had not provided us with documentation of their meetings and communications with employees, and neither OMB nor OPM had provided relevant documents on employee outreach and inclusion. Developing and Communicating Implementation Plans with Milestones and Time Frames We have previously reported that organizational transformations must be carefully and closely managed by developing an implementation plan with key milestones and deliverables to track and communicate implementation progress, among other actions. However, as of May 17, 2019, OMB, OPM, and GSA had not developed an implementation plan or publicly reported on key milestones (see fig. 6). This is the case despite the fact that the President’s fiscal year 2020 budget states that the reform is underway in fiscal year 2019, and that all remaining portions of the reform would be completed in fiscal year 2020 through legislation. Moreover, these agencies have not ensured transparency of their efforts by publicly reporting on implementation progress. Addressing Existing Management Challenges Our prior work has shown that successful reorganizations seek to implement best practices in the systems and processes wherever they may be found, and guard against automatically adopting the approaches used by the largest or acquiring component. The risk is that the new organization may migrate less-than-fully efficient and effective systems and processes merely because those systems and processes are most often used. Accordingly, OPM’s proposed reorganization should address agency management challenges, such as those in our high-risk program, priority open recommendations, or those identified by agency IGs. OMB, OPM, and GSA are aware of our related prior work, including major management challenges, but have not demonstrated how the proposed reorganization will help address these challenges (see fig. 7). Based on a document released by the administration on May 15, 2019 discussing its rationale for the merger of OPM and GSA, the reorganization should better support human capital delivery across the federal government by centralizing the services provided by both agencies, and reducing duplication. The reform plan also acknowledges that federal human capital management remains a high-risk area due to mission-critical skills gaps within the federal workforce. The reform plan further states that OPM does not have the capacity to address the high- risk issues we have identified, and progress would be achieved more efficiently by transferring OPM’s responsibilities to other government entities, including GSA and the EOP. However, as of May 17, 2019, OMB, OPM and GSA had not provided any documentation or analysis to demonstrate how the proposed reorganization would help resolve high- risk issues. The reform plan also draws attention to the OPM security breach that occurred several years ago, and cites it as a reason for moving information technology systems to GSA. We have five open priority recommendations to OPM regarding information security, as we reported to OPM in April 2019. For example, in May 2016, we recommended that OPM update security plans to ensure controls specific to high-impact systems are addressed, provide and track training for individuals with significant security responsibilities, and ensure that security control assessments specific to high-impact systems are comprehensive. To fully implement these recommendations, we reported that OPM needs to complete its ongoing efforts in each of these areas by implementing an automated system for management of security controls and security plans, defining and completing its planned corrective actions on training, and reviewing completed security control assessments. It is unclear whether OMB, OPM, and GSA have fully considered how relevant major management challenges identified by OPM’s and GSA’s IGs may affect the proposed reorganization (see fig. 7). For example, the GSA IG’s 2018 report on management challenges contains a number of findings that call into question GSA’s capacity to take on certain responsibilities the administration proposes transferring to GSA as part of the reorganization. Specifically, the report discusses GSA’s challenges with managing internal controls, prioritizing cybersecurity, and managing human capital. By addressing major management challenges and adopting best practices and processes as part of the reorganization effort, the administration will be better positioned to successfully implement their proposal. More Information Needed to Fully Assess Legal Authorities to Reorganize OPM As of May 17, 2019, OMB, OPM and GSA had not provided documentation that they had identified specific actions that can be taken administratively versus those that will require legislative action to reorganize OPM. We asked OMB, OPM, and GSA for their views on what legal authority, including appropriations, they are relying on to reorganize OPM, including any additional authority that may be needed. As described earlier in this statement, these agencies have not provided implementation plans or other details on the reorganization. Similarly, they have not provided details on the statutory underpinnings for OPM’s reorganization. To the extent the administration identifies the legal authority it is it relying on to support this proposed reorganization, or the additional legal authority it needs, we will continue to assess it. OPM is statutorily created as “an independent establishment in the executive branch.” In addition, the Director of OPM is vested with certain functions by statute, and the Director (or OPM designee) is required to perform those functions, including executing, administering, and enforcing civil service requirements. While the Director of OPM may delegate selected human capital management functions to other agencies, OPM remains statutorily responsible for certain oversight activities, such as establishing standards that apply to such delegated activities and making written findings, where appropriate, if an agency to which OPM delegated human capital management functions acts contrary to law, rule, regulation, or standard, and requiring that the agency take corrective action, among other activities. OPM has various statutorily required responsibilities related to administering civil service retirement, insurance, health benefits, and life insurance programs, among others. OPM is funded primarily through its revolving fund—which is made up of fees or reimbursements provided by agencies for services OPM provides, such as background investigations and human resources services— transfers from OPM’s Earned Benefits Trust Funds for administrative services, and discretionary appropriations for OPM’s general activities and the Office of IG. To execute certain transfers of functions from OPM to GSA, the administration has acknowledged the need for additional statutory authority, but has also stated that it will rely on existing authority to move certain functions administratively. For example, the Analytical Perspectives accompanying the President’s fiscal year 2020 budget acknowledges that the transfer of OPM functions to GSA will be completed using a combination of existing legal authority and legislation. However, the administration does not identify which functions will require legislation and which OPM functions may be transferred administratively. In particular, OMB’s Deputy Director for Management stated, in July 2018, that many of the administration’s reorganization proposals can be implemented in whole or in part through existing administrative authorities. The conference report accompanying the 2019 Appropriations Act directed OPM to submit a report that included, among other things, the legal authority under which OPM proposed to transfer the human resources solutions function within the OPM revolving fund to GSA. OPM’s report stated that it and GSA, in consultation with OMB, continue to deliberate upon the application and use of administrative authorities to transfer the OPM functions to GSA. In addition, in April 2019, the General Counsel of OPM told us that the agency is unable to provide its legal analysis to us because it was still in progress and the agency was waiting for certain executive branch actions to be finalized. Without this information, we cannot assess the legal authorities the administration is relying on to implement the reorganization of OPM. Key Capacities Important for Effective Strategic Human Capital Management As Congress and the administration consider whether or how to restructure OPM, regardless of the eventual decision about the organizational arrangement, we believe that it will be important to retain the capacity to execute certain government-wide, strategic human capital functions. These include the capacity to (1) identify trends affecting the future of the federal workforce; (2) effectively collaborate and coordinate with key stakeholders to address these government-wide trends; (3) lead the design of government-wide solutions to shared human capital challenges; and (4) administer and enforce civil service laws and regulations. As noted in our prior work, these functions are desirable and appropriate because they generate broad consistency across federal agencies, which is critical for, among other things, ensuring that each federal employee has certain safeguards and protections regardless of where he or she works. They also produce certain efficiencies and economies of scale that come from central coordination, and help maintain a reasonably level playing field among federal agencies when competing for talent. This is particularly important because we continue to designate strategic human capital management as a high-risk area. While many day-to-day human capital responsibilities have been delegated from OPM to individual agencies over the years, OPM continues to play an important strategic role including in the creation, execution, oversight, and strengthening of human capital policies and programs. For example, OPM’s 2018-2022 strategic goals are to: Transform hiring, pay, and benefits across the federal government to attract and retain the best civilian workforce. Lead the establishment and modernization of human capital information technology and data management systems and solutions. Improve integration and communication of OPM services to federal agencies to meet emerging needs. Optimize agency performance. Moreover, OPM was given a key role in fulfilling the human capital-related goal of the most recent President’s Management Agenda, in which the administration noted its intention to partner with Congress on “overhauling the statutory and regulatory rules that have, over time, created an incomprehensible and unmanageable civil service system.” OPM, along with OMB and DOD was tasked with the goal of aligning and strategically managing the workforce to efficiently and effectively achieve the federal government’s mission. To carry out these government-wide, strategic responsibilities, the following capabilities, whether possessed by OPM or some other entity, will be essential for ensuring cost-effective leadership, management, and oversight of the federal workforce. The Capacity to Identify Trends Affecting the Future of the Federal Workforce In our March 2019 report, we noted that such trends as technological advances, an increased reliance on nonfederal partners, and changing demographics and shifting attitudes toward work, are affecting how federal work is done, and consequently the skills and competencies that workers need to accomplish agency missions. Moreover, recent publications by the administration and others have raised concerns about whether the government’s employment policies and practices are still relevant and desirable to the current and future workforce. As far back as 1989, we reported that OPM had not provided the leadership necessary to sustain attention to identifying and resolving critical human resource problems affecting government operations and preparing for the future. Although OPM has made progress in this area and provides a variety of services, its progress has been inconsistent and issues still remain. For example, in 2018, OPM issued its Federal Workforce Priorities Report, which identifies changes in the external environment that will likely affect federal human capital management, including the evolving role of workers, changes in technology, employee health, and shifting generational demographics. OPM has also hosted a series of symposia that provide human capital specialists insight on addressing workforce challenges of the future. While these and other efforts are all important steps in the right direction, more work is needed in other areas. For example, as discussed in our March 2019 report, over the years we have made a number of recommendations to OPM to help agencies better meet their missions in an era of changing technology, demographics, fiscal constraints, and other challenges. OPM agreed with most of these recommendations and has made some progress, but additional actions are needed. They include, for example, identifying existing skills and competencies, assessing gaps in existing and future skills and competencies, and monitoring progress toward closing skills gaps. Moreover, in our March 2019 High Risk report, we noted that OPM needs to fully address the recommendations in our January 2015 report. Our recommendation called on OPM to make more strategic use of government workforce data to build a predictive capacity for identifying and mitigating emerging skill gaps across government. The Capacity to Effectively Collaborate and Coordinate with Key Stakeholders Certain human capital issues, such as addressing mission critical skills gaps, are crosscutting in nature and require the coordinated efforts of multiple stakeholders. However, a key challenge we identified in our May 2014 report on strategies to help agencies meet their missions in an era of highly constrained resources was that the federal human capital community is highly fragmented, with multiple actors both inside and outside of government informing and executing human capital policies and initiatives in ways that are not always aligned with broader, government-wide human capital efforts. Within government, OPM, OMB, the Chief Human Capital Officers (CHCO) Council, and individual agencies create, implement, and oversee human capital initiatives. Those initiatives are shaped, in part, by input provided by labor unions and federal management councils such as the President’s Management Council. The federal chief human capital officers with whom we spoke noted that each of these actors possess its own mission, initiatives, agendas, chain of command, budgets, and oversight. While this is to be expected given their various roles and responsibilities, these same factors can create disincentives to collaborating to achieve common human capital goals. In response to this issue, we recommended in 2014 that OPM work with the CHCO Council to, among other actions, strengthen coordination and leadership on government-wide human capital issues. OPM agreed with our recommendation and issued a final regulation, effective in April 2017, requiring it and agencies take significant steps in identifying, prioritizing, and coordinating efforts to address critical human capital issues. We believe this final regulation represents an important step toward addressing fragmentation within the federal human capital community. Going forward, it will be important for OPM, or another entity, if reorganized, to work with the CHCO Council and other stakeholders to address our open recommendations concerning specific human capital functions. Indeed, many of our open recommendations, including those that require priority attention from OPM, call on OPM to work in conjunction with the CHCO Council. The Capacity to Lead and Design Government-Wide Solutions to Shared Human Capital Challenges Government-wide or “enterprise” solutions are important because they can integrate the efforts of multiple departments and agencies to address crosscutting human capital challenges more effectively by leveraging agencies’ expertise, experience, technology, and other resources. However, in our 2014 report, we found that while agencies have many common human capital challenges, they tend to address these issues independently without looking to enterprise-wide solutions that could resolve them more effectively. Across government, there are examples of agencies and OPM initiating enterprise solutions to address crosscutting issues, including the consolidation of federal payroll systems into shared-services centers. While these and other actions are important steps in the right direction, the CHCOs we spoke with in 2014 identified certain barriers to greater coordination to address common problems. For example, federal budgeting and account structures reinforce the prevailing tradition of controlling agency resources within a single agency. Moreover, agencies may be reluctant to contribute resources to a government-wide approach because they may not get an equitable return on their investment, or may get a product that does not fit their needs. According to the CHCOs in 2014, two areas that are ripe for greater government-wide collaboration are human resource information technology (HR IT), and strategic workforce planning. Specifically, the CHCOs said agencies could be missing cost-savings opportunities by not coordinating HR IT investments within and across agencies. They noted that agencies are individually procuring identical systems rather than leveraging the purchasing power of multiple agencies to negotiate better prices or services, or use shared service centers. Similarly, several CHCOs we spoke with said agencies are not consistently leveraging lessons learned or collaborating to address difficulties they encounter with workforce planning models. To further agencies’ use of government-wide approaches, we recommended that the Director of OPM, in conjunction with the CHCO Council, should explore the feasibility of expanded use of enterprise solutions to more efficiently and effectively address shared or government-wide human capital challenges. Such actions could include: (1) seeking cost savings and improved functionality through coordinated government-wide human resources information technology planning and acquisition; (2) seeking agency input to ensure OPM’s workforce planning tools provide effective guidance for agencies; and (3) sharing workforce planning lessons learned and successful models across the government. OPM agreed with the recommendation and in September 2018, it reported that in spring 2019, data will be available to indicate whether surveys and tools to address government-wide human capital challenges are meeting their intended goals. In March 2019, OPM told us that it was conducting Human Capital Reviews with relevant agencies. However, to fully implement the recommendation, OPM, or another entity, if reorganized, needs to demonstrate continued progress in addressing government-wide human capital challenges. The Capacity to Administer and Enforce Civil Service Laws and Regulations Broad consistency across federal agencies is important for ensuring that all federal employees have the same safeguards, rights, and protections regardless of where they work. These include, for example, merit principles; protection from prohibited human capital practices; the ability to organize, bargain collectively, and participate through labor organizations; and due process that is fair, fast, and final. OPM is responsible for executing, administering, and enforcing the civil service rules and regulations, and the laws governing the civil service. Additionally, OPM is required to establish and maintain oversight over delegated human capital activities, including delegated competitive examining activities, to ensure agencies are acting in accordance with the merit system principles and the relevant standards established by OPM, such as compliance with applicable laws, rules, regulations, executive orders, and OPM policies. OPM monitors overall implementation and identifies corrective actions when deficiencies are found. OPM conducts this oversight through three primary means: delegated examining unit audits, human resource management evaluations, and special studies. However, in our prior work, we have identified the need for more effective oversight in such areas as agencies’ use of hiring authorities, agencies’ classification programs, the conversion of political appointees to career positions, and the Senior Executive Service performance-based pay system. With respect to agencies’ use of hiring authorities, for example, to help strengthen the government’s ability to compete in the labor market for top talent, and to improve the federal hiring process, we recommended in 2016 that the Director of OPM, in conjunction with the CHCO Council, should determine whether opportunities exist to refine, consolidate, eliminate, or expand agency-specific authorities to other agencies and implement changes where OPM is authorized. OPM agreed with the recommendation and in December 2018, OPM said that it continues to research and examine streamlining opportunities, such as those identified in its July 2018 study on excepted service hiring authorities, as part of the broader initiative to modernize federal hiring practices under the President’s Management Agenda. However, OPM did not provide a time frame for implementation. In its March 2019 Congressional Justification for the fiscal year 2020 budget request, OPM included legislative proposals for new hiring authorities such as highly qualified experts and temporary appointments to help agencies meet critical needs as well as a change to the criteria for granting direct hire authority. While OPM has made some progress in this area, it will be important for the agency to follow through on its planned actions to streamline hiring authorities. To fully implement the recommendation, OPM or another entity, if reorganized, needs to complete these efforts and, as appropriate, develop legislative proposals in consultation with the CHCO Council. Thank you, Chairman Connolly, Ranking Member Meadows, and Members of the Subcommittee. This concludes my testimony. I would be pleased to answer any questions. GAO Contact and Staff Acknowledgments If you or your staff has any questions concerning this testimony, please contact Triana McNeil at (202) 512-6806 (McNeilT@gao.gov). Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. In addition to the contacts named above, Sarah Veale (Assistant Director), Peter Beck (Analyst-in- Charge), Colenn Berracasa, Robert Goldenkoff, Chelsa Gurkin, Shelby Kain, Steven Putansu, Janet Temko-Blinder, Peter Verchinski, and Alicia White made key contributions to the testimony. Other staff who made contributions to the reports cited in the testimony are identified in the source products. 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.

Why GAO Did This Study In June 2018, the administration proposed reorganizing OPM by devolving its responsibilities to other agencies and entities including GSA and the EOP; see the figure for details. OMB's role is to coordinate and oversee the reorganization proposal, with support from OPM and GSA. In June 2018, GAO reported on key practices to assess agency reform efforts. This testimony focuses on preliminary observations from GAO's ongoing work related to the transfer of functions from OPM to GSA and the EOP. Specifically, we evaluated (1) the extent to which OMB, OPM, and GSA have addressed key practices for effective reforms and reorganizations; (2) legal authorities that may affect the reorganization of OPM, and (3) key capacities important for effective strategic human capital management, which need to be in place regardless of how the leadership over federal human capital is organized. For the information in this testimony, as of May 17, 2019, GAO met with OMB staff, GSA officials, OPM's and GSA's Inspectors General staff, and analyzed documentation provided by GSA. GAO also reviewed its prior related work. What GAO Found The Office of Management and Budget (OMB), Office of Personnel Management (OPM), and General Services Administration (GSA) have generally not addressed key practices for agency reform efforts as they have moved forward with their proposal to reorganize OPM. They have not established outcome-oriented goals, developed a cost-benefit analysis or implementation plans, and have not fully involved or communicated their efforts with the Congress, employees, and other key stakeholders. OPM and GSA also have not shown how they will address management challenges that may affect their ability to successfully reorganize the government's central human capital functions. OMB, OPM and GSA have not identified specific actions, as of May 17, 2019, that can be taken administratively versus those that will require legislative action to reorganize OPM. The administration has acknowledged the need for additional statutory authority to execute certain transfers of functions from OPM to GSA and the Executive Offices of the President (EOP), but has also stated that it will rely on existing authority to move certain functions administratively. Without additional information from OMB and agencies, GAO cannot assess the legal authorities the administration is relying on to implement the reorganization. As the Congress and administration consider whether or how to restructure OPM, it will be important to retain the capacity to execute certain government-wide, strategic human capital functions, regardless of the decision made about the organizational arrangement. These capacities include an ability to identify future workforce trends and to effectively collaborate with stakeholders—for the purpose of creating, executing, and overseeing human capital policies and programs, and enforcing civil service laws and regulations. This is particularly important because GAO continues to designate strategic human capital management as a high-risk area.

gao_GAO-19-662T

gao_GAO-19-662T_0

Background FEMA’s Public Assistance Program FEMA’s Public Assistance program provides grant funding to state, territorial, local, and tribal governments as well as certain types of private nonprofit organizations to assist with responding to and recovering from presidentially-declared major disasters or emergencies. As shown in figure 3, Public Assistance grant funds are categorized broadly as “emergency work” or “permanent work.” Within these broad categories are separate subcategories. In addition to the emergency work and permanent work categories, the program includes category Z, which represents indirect costs, direct administrative costs, and any other administrative expenses associated with a specific project. FEMA’s Public Assistance program also provides grant funding for cost- effective hazard mitigation measures to reduce or eliminate the long-term risk to people and property from future natural and man-made disasters and their effects. For example, a community that had a fire station damaged by a disaster could use Public Assistance grant funding to repair the facility and incorporate additional measures such as installing hurricane shutters over the windows to mitigate the potential for future damage. FEMA, the state or territorial government (the recipient), and local or territorial entities (the subrecipient) work together to develop projects under the Public Assistance program. After a project has completed FEMA’s review process and is approved, FEMA obligates funding for the project by placing money into an account where the recipient has the authority to draw down—or withdraw—funding to pay the subrecipient for eligible work upon completion. The Public Assistance Alternative Procedures Program in Puerto Rico and the USVI The Sandy Recovery Improvement Act of 2013 authorized the use of alternative procedures in administering the Public Assistance program, thereby providing new flexibilities to FEMA, states, territories, and local governments for debris removal, infrastructure repair, and rebuilding projects using funds from this program. Unlike in the standard Public Assistance program where FEMA will fund the actual cost of a project, the Public Assistance alternative procedures allow awards for permanent work projects to be made on the basis of fixed-cost estimates to provide financial incentives for the timely and cost-effective completion of work. Under these procedures, if the actual cost of the project exceeds the fixed-cost estimate agreed upon by FEMA and the recipient, the recipient or subrecipient is responsible for the additional costs at the time of the close-out process. However, if the actual cost of completing eligible work for a project is below the estimate, the recipient may use the remaining funds for additional cost-effective hazard mitigation measures to increase the resilience of public infrastructure. In addition, these funds may also be used for activities that improve the recipient’s or subrecipient’s future Public Assistance operations or planning. In October 2017, Puerto Rico requested, and FEMA approved, the use of the alternative procedures program for all large-project funding for Public Assistance permanent work projects in categories C through G. Although FEMA had approved alternative procedure grants in 30 states as of April 2018, in these cases, alternative procedures were used on a project-by-project basis. Puerto Rico’s recovery from the 2017 hurricanes is the first recovery to use alternative procedures for all large permanent work projects. In addition, in July 2018, FEMA approved a request from the Governor of the USVI to transition to using the Public Assistance alternative procedures program for permanent work in the territory. Unlike in Puerto Rico, the USVI may pursue the alternative procedures on a project-by-project basis. FEMA Had Obligated $5.6 Billion and $1.8 Billion in Public Assistance Grant Funding in Puerto Rico and the USVI, Respectively, as of April 2019 As of April 2019, FEMA had obligated a total of about $7.4 billion in grant funds for Public Assistance projects in both Puerto Rico and the USVI. Specifically, as shown in figure 4, FEMA obligated approximately $5.6 billion for 1,264 Public Assistance projects in Puerto Rico, including approximately $5.1 billion (90 percent) for emergency work (categories A and B) and $377.7 million (7 percent) for permanent work in categories C through G). Puerto Rico had expended approximately $3.5 billion—about 61 percent of total Public Assistance grant obligations in Puerto Rico—as of April 2019. Ninety-six percent of the expended amount went toward emergency work projects in categories A and B while just over one percent went toward permanent work projects. The majority of FEMA’s obligations and the funding Puerto Rico expended as of April 2019 are for emergency work because these projects began soon after the disasters struck and focused on debris removal and providing assistance to address immediate threats to life and property. In contrast, permanent work projects take time to identify, develop, and ultimately complete as they represent the longer-term repair and restoration of public infrastructure. In the USVI, FEMA had obligated approximately $1.8 billion for 583 Public Assistance projects across the territory, as of April 2019. Similar to Public Assistance grant funding in Puerto Rico, the majority of funding FEMA obligated and the USVI expended was in emergency work categories A and B. Specifically, FEMA obligated approximately $1.1 billion (63 percent) for emergency work (categories A and B) and $587.3 million (33 percent) for permanent work (categories C through G) in the territory (see fig. 5). Of the $1.8 billion FEMA obligated for Public Assistance projects, the USVI had expended approximately $982.4 million as of April 2019. Specifically, the USVI had expended about $808.1 million (82 percent) for emergency work projects in categories A and B and $163.1 million (17 percent) for permanent work projects in categories C through G. Emergency work. As of April 2019, FEMA had obligated a total of approximately $6.2 billion for emergency work projects in Puerto Rico and the USVI—including about $5.1 billion in Puerto Rico and $1.1 billion in the USVI. These projects focused on debris removal activities and providing assistance to address immediate threats to life and property. For example, as of April 2019, FEMA had obligated $138.9 million for projects focused on debris removal activities in the USVI under category A. This included $45.9 million to the USVI Department of Public Works for USVI-wide debris removal efforts and $39.1 million to the USVI Water and Power Authority for these activities in St. Croix (see fig. 6). In another example, FEMA obligated more than $140.0 million to the Puerto Rico Aqueducts and Sewer Authority under category B to fund emergency protective measures, including using back-up generators to supply water to the island after Hurricane Maria, among other things. Further, as of April 2019, FEMA had obligated $1.1 billion in Puerto Rico and $278 million in the USVI to fund the Sheltering and Temporary Essential Power pilot program. This program, which is implemented as a subprogram under Public Assistance program category B, is intended to provide essential repairs or restore power to private residences to allow affected individuals to return or remain in their homes, thereby reducing the demand for other shelter options. We are continuing to assess this program as part of our ongoing work on recovery efforts in the USVI. Permanent work. As of April 2019, FEMA had obligated approximately $965.0 million for permanent work projects in Puerto Rico and the USVI— including about $377.7 million in Puerto Rico and $587.3 million in the USVI. These projects focused on the restoration of disaster-damaged infrastructure or systems. For example, under category C, FEMA obligated $137.6 million for projects in Puerto Rico focused on the permanent repair of roads and bridges, such as the severely damaged road shown in figure 7 below. In addition, under category E, FEMA obligated $39.2 million and $67.7 million for projects in Puerto Rico and the USVI, respectively, focused on repairing and rebuilding damaged public buildings and equipment, such as the schools shown in figure 8 below. Further, under category F, FEMA obligated $504.9 million for projects in the USVI to repair damaged utilities. Specifically, FEMA obligated $481.8 million—or 95 percent of this total—through the standard Public Assistance program for projects focused on territory-wide permanent electrical distribution system repairs. This includes replacing damaged wooden utility poles with more resilient composite fiberglass poles that can withstand 200 mile per hour winds as well as power transmission lines and transformers (see fig. 9). Puerto Rico and the USVI Have Established Recovery Offices to Oversee and Monitor Recovery Efforts As the recipients of federal disaster funding, Puerto Rico and the USVI are responsible for monitoring and overseeing the Public Assistance program to ensure it is implemented in compliance with applicable laws, regulations, and requirements as well as FEMA policies and guidance. To address these responsibilities, Puerto Rico and the USVI established recovery offices to manage recovery activities and funding, including through the Public Assistance program. Puerto Rico’s Central Office for Recovery, Reconstruction, and Resilience Has Developed Internal Controls to Oversee Recovery Funds In March 2019, we reported that Puerto Rico, in accordance with Amendment 5 to the President’s disaster declaration, established the Central Office for Recovery, Reconstruction, and Resilience (COR3) to oversee federal recovery funds. We also reported that COR3 was developing an internal controls plan to help ensure better management and accountability of the funds. According to FEMA officials, FEMA instituted a manual reimbursement process due to Puerto Rico’s financial situation, weaknesses in internal controls, and the large amount of recovery funds, among other things, to mitigate risk and help ensure financial accountability. However, from our ongoing work on Puerto Rico’s disaster recovery efforts, we have learned that, on April 1, 2019, FEMA removed the manual reimbursement process and began a transition to allow the central recovery office to take responsibility for the review and reimbursement approval of federal recovery funds. We have also learned from our ongoing work that, in March 2019, COR3 released the Disaster Recovery Federal Funds Management Guide. Among other things, the guide outlines COR3’s roles and responsibilities and the internal controls COR3 put in place to oversee the recovery. For example, COR3 will identify, procure, and administer all federal, territorial, and private resources available to Puerto Rico related to recovery. In addition, it will provide oversight of subrecipients using risk-based monitoring, offer technical assistance, and advise Puerto Rico’s governmental agencies and municipalities regarding any matter related to recovery. COR3 continues to update its online transparency portal intended to provide a breakdown of FEMA Public Assistance and other federal funding obligated for disaster recovery in Puerto Rico. The USVI Established the Office of Disaster Recovery to Monitor and Oversee Recovery Efforts According to our preliminary observations, in February 2019, the USVI established the new Office of Disaster Recovery. This office serves as the primary territorial agency responsible for overseeing all disaster recovery efforts and funding in the territory, and coordinates across all USVI governmental agencies and other pertinent entities. According to USVI officials, following the 2017 hurricanes, key USVI agencies did not have enough employees with the knowledge and expertise necessary to staff recovery-related positions and effectively manage the implementation of recovery efforts. To address this challenge in the short-term, the USVI government hired two contractors in December 2017—Witt O’Brien’s, LLC and Ernst & Young Puerto Rico, LLC—to assist the territory in planning, developing, implementing, and overseeing Public Assistance program projects, among other responsibilities. The Director of the Office of Disaster Recovery told us that while contractor personnel had been valuable in augmenting the USVI’s management capacity in the short term, the territory’s longer-term vision included the establishment of the Office of Disaster Recovery to centrally manage all aspects of federal recovery in the territory. Among other things, the Office of Disaster Recovery is responsible for taking on the USVI’s monitoring and oversight responsibilities for the Public Assistance program in the long term. This includes tracking and reporting on the progress of projects and overseeing reimbursement requests for completed work to ensure compliance with applicable laws and FEMA policies. As of March 2019, the Director of the Office of Disaster Recovery told us the priority is to quickly hire and train qualified individuals to staff the new agency. FEMA officials in the USVI stated that the establishment of the Office of Disaster Recovery and the USVI’s ongoing efforts to hire local residents into recovery-related positions represented a positive step forward in increasing the territory’s capacity to oversee recovery efforts. We will continue to review the monitoring and oversight of recovery efforts in Puerto Rico and the USVI in our ongoing work. Public Assistance Challenges Remain in Puerto Rico and USVI, However FEMA has Taken Some Actions to Improve Program Implementation Our prior and ongoing work highlight the challenges with implementing the Public Assistance program—and the alternative procedures—in Puerto Rico and the USVI. In particular, our prior and ongoing work have identified challenges related to (1) the clarity of FEMA’s guidance for the Public Assistance program, (2) the time and resources needed to transition to FEMA’s new Public Assistance program delivery model in Puerto Rico, (3) the implementation of flexibilities provided by the Bipartisan Budget Act of 2018, and (4) developing fixed-cost estimates. FEMA has taken some actions, including issuing additional guidance and developing specific training, among other things, to improve Public Assistance implementation in Puerto Rico and the USVI. However, it is too soon to assess their effectiveness in addressing these issues. Clarity of Guidance. In March 2019, we reported that officials from FEMA, COR3, and municipalities said they experienced initial challenges with the recovery process, including concerns about lack of experience and knowledge of the alternative procedures; and concerns about missing, incomplete, or conflicting guidance from FEMA on the alternative procedures. In addition, in our June 2019 testimony statement we continued to report on these challenges and preliminary observations from our ongoing work indicate that these challenges continue. For example, officials from Puerto Rico’s government agencies told us they did not feel they had sufficient guidance on the FEMA Public Assistance program and where they did, written and verbal FEMA guidance was inconsistent or conflicting. For instance, officials from one agency expressed their desire for more FEMA guidance communicated in writing as FEMA officials would frequently interpret existing guidance differently. Similarly, officials from two agencies described situations where they had initially been directed to follow one interpretation of a policy, only to be directed to follow a different, conflicting interpretation in the subsequent months. Puerto Rico agency officials also stated that the lack of sufficient instruction led to a “back and forth” with FEMA for clarifications, which led to delays in the phases of project development. For example, officials from one Puerto Rico government agency stated that conflicting verbal instructions from several FEMA officials contributed to delays in opening the bidding process for recovery-related contracts. FEMA officials in Puerto Rico stated that the agency has developed specific guidance for disaster recovery in Puerto Rico and that there are various ways, such as in-person meetings, where officials from Puerto Rico can obtain clarification. FEMA officials also reported that they developed additional training for new FEMA employees. We are continuing to examine this issue as part of our ongoing review of Puerto Rico’s recovery. FEMA’s new delivery model in Puerto Rico. In May 2019, FEMA’s Federal Disaster Recovery Coordinator for Puerto Rico announced that FEMA was transitioning to using the new Public Assistance program delivery model in Puerto Rico beginning on June 3, 2019. Among other things, the implementation of the new delivery model establishes a new Consolidated Resource Center in Puerto Rico to support grant development for disaster recovery across all recovery sectors and geographic branches. Following the hurricanes, FEMA implemented a program delivery model developed specifically for Puerto Rico which included, among other things, a sector-based approach which coordinated recovery resources across the federal interagency, private sector, and nongovernmental organizations to identify and complete proposed work. According to FEMA officials, the decision to transition from the initial delivery model to the new delivery model in Puerto Rico was due to improvements made since its nationwide deployment in 2017. In response, COR3 officials raised concerns about the scope of the changes and potential challenges with the amount of time and resources needed to transition to the new delivery model. The Bipartisan Budget Act of 2018. We reported in June 2019 that in both Puerto Rico and the USVI, FEMA and local officials have reported challenges with the implementation of the flexibilities authorized by section 20601 of the Bipartisan Budget Act. This section of the Act allows for the provision of assistance under the Public Assistance alternative procedures to restore disaster-damaged facilities or systems that provide critical services—such as medical and educational facilities— to an industry standard without regard to pre-disaster condition. Officials from Puerto Rico’s central government stated that they disagreed with FEMA’s interpretation of the types of damages covered by section 20601 of the Bipartisan Budget Act of 2018. In response, FEMA officials in Puerto Rico stated they held several briefings with Puerto Rico’s central recovery office to explain FEMA’s interpretation of the section. In addition, FEMA officials in the USVI told us that initially, they had difficulty obtaining clarification from FEMA headquarters regarding how to implement key components of section 20601 of the Act. Further, USVI officials stated that at times, the appropriate process for implementing components of the Act was not clear and that ensuring program participants understood its key components was difficult. However, FEMA officials in the USVI stated that they continue to move forward with developing alternative procedures projects. USVI officials also told us that FEMA had been responsive and helpful in identifying its options for using the new flexibilities the Act provides. Developing Fixed-Cost Estimates. Preliminary observations from our ongoing work indicate that as of May 2019, FEMA had obligated funding for four alternative procedures program projects in Puerto Rico and two projects in the USVI. FEMA officials in Puerto Rico and the USVI stated that the ongoing development of a “cost factor” for use in the fixed-cost estimating process has slowed the pace of FEMA obligations for permanent work projects. Specifically, these factors are intended to ensure that the costs associated with implementing projects in Puerto Rico and the USVI are sufficiently captured when developing the fixed- cost estimates for alternative procedures projects. Since incorporating the cost factor into the fixed-cost estimating process will increase the amount of funding obligated for any given permanent work project, FEMA officials explained that Puerto Rico and the USVI have an incentive to delay the obligation of individual projects until this factor is finalized. For example, FEMA officials in the USVI told us in May 2019 that obligations for permanent work projects in the territory were mostly on hold until the USVI-specific cost factor was finalized. As of June 2019, the cost factors for use in both Puerto Rico and the USVI had not yet been finalized. According to FEMA guidance, the Puerto Rico-specific cost factor is being developed by a third-party center of excellence comprising personnel selected by FEMA and Puerto Rico, through COR3. In March 2019, we reported that while FEMA had identified and chosen personnel, COR3 had not yet finalized its hiring of personnel to staff the center of excellence, which resulted in delaying the cost estimation process. Through our ongoing work we learned that, as of June 2019, COR3 had identified and hired personnel to staff the center; however, FEMA and COR3 have not come to agreement on a cost estimation approach. Further, according to FEMA officials, no timeline has been established for the completion of the center of excellence’s standard operating procedures for developing fixed-cost estimates for permanent work projects in Puerto Rico. In addition, according to FEMA officials, the USVI-specific factor is being developed by an independent contractor. FEMA officials told us that territorial officials disagreed with the initial cost factors this contractor proposed and contended the factors were insufficient in accurately capturing the unique circumstances that influence construction costs in the territory, such as the limited availability of local resources and the need to import materials and labor. As of June 2019, these officials told us the contractor was developing a third and final cost factor for potential incorporation into the fixed-cost estimation process in the USVI. Despite these delays, FEMA officials in the USVI stated that they continue to work with territorial officials to develop alternative procedures projects in the territory. They added that once the cost factor is finalized and incorporated into FEMA’s fixed-cost estimating process, FEMA and the USVI will be well positioned to quickly finalize these projects and obligate funding. However, we reported in June 2019 that the territory plans to take a cautious approach in pursuing permanent work projects using the Public Assistance alternative procedures program. Specifically, USVI officials we interviewed told us that developing fixed-cost estimates for alternative procedures projects that accurately incorporate the future impact of inflation and increases in materials and labor costs for certain projects was difficult. Further, these officials stated that since the territory is financially responsible for any costs that exceed these fixed-cost estimates, the USVI plans to pursue alternative procedures projects that do not include high levels of complexity or uncertainty to reduce the risk of cost overruns, especially given its already difficult financial situation. As established in FEMA guidance, Puerto Rico’s deadline for finalizing fixed-cost estimates for permanent work projects using the alternative procedures—and the Bipartisan Budget Act, as applicable—is October 2019. Since Puerto Rico must use the alternative procedures for all large permanent work, all fixed-cost estimates for Public Assistance program permanent work projects in Puerto Rico must be finalized by this date, or, according to FEMA officials, Puerto Rico must request that FEMA extend this deadline on a project-by-project basis. In contrast, the USVI has the flexibility to pursue either the alternative procedures or the standard procedures on a project-by-project basis. As the USVI’s deadline for finalizing these projects is in March 2020, it is too early gauge the extent to which the alternative procedures will play a role in the USVI’s long-term recovery strategy. We will continue to evaluate these identified challenges and any efforts to address them, as well as other aspects of recovery efforts in the USVI and Puerto Rico, and plan to report our findings in late 2019 and early 2020, respectively. Thank you, Chairman Payne, Ranking Member King, and Members of the Subcommittee. This concludes my prepared statement. I would be happy to respond to any question you may have at this time. GAO Contact and Staff Acknowledgments If you or your staff has any questions concerning this testimony, please contact Christopher P. Currie at (404) 679-1875 or curriec@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Key contributors to this statement were Joel Aldape (Assistant Director), Bryan Bourgault, Aaron Gluck, Taylor Hadfield, Brian Lipman, and Amanda Prichard. In addition, Eric Hauswirth, Susan Hsu, Tracey King, Taylor Matheson, Amanda Miller, Heidi Nielson, and Kevin Reeves made contributions to this statement. Key contributors for the previous work on which this statement is based are listed in each product. Appendix I: GAO Products Previously Issued on Emergency Management 2017 Disaster Relief Oversight: Strategy Needed to Ensure Agencies’ Internal Control Plans Provide Sufficient Information. GAO-19-479 (Washington, D.C.: June 28, 2019). Emergency Management: FEMA Has Made Progress, but Challenges and Future Risks Highlight Imperative for Further Improvements, GAO-19-617T (Washington, D.C.: June 25, 2019). Emergency Management: FEMA Has Made Progress, but Challenges and Future Risks Highlight the Imperative for Further Improvements, GAO-19-594T (Washington, D.C.: June 12, 2019). Disaster Assistance: FEMA Action Needed to Better Support Individuals Who Are Older or Have Disabilities. GAO-19-318 (Washington, D.C.: May 14, 2019). Disaster Contracting: Actions Needed to Improve the Use of Post- Disaster Contracts to Support Response and Recovery. GAO-19-281 (Washington, D.C.: April 24, 2019). 2017 Hurricane Season: Federal Support for Electricity Grid Restoration in the U.S. Virgin Islands and Puerto Rico. GAO-19-296 (Washington, D.C.: April 18, 2019). FEMA Grants Modernization: Improvements Needed to Strengthen Program Management and Cybersecurity. GAO-19-164 (Washington, D.C.: April 9, 2019). Disaster Recovery: Better Monitoring of Block Grant Funds Is Needed. GAO-19-232 (Washington, D.C.: March 25, 2019). Puerto Rico Hurricanes: Status of FEMA Funding, Oversight, and Recovery Challenges. GAO-19-256 (Washington, D.C.: March 14, 2019). Huracanes de Puerto Rico: Estado de Financiamiento de FEMA, Supervisión y Desafíos de Recuperación. GAO-19-331 (Washington, D.C.: March 14, 2019). High-Risk Series: Substantial Efforts Needed to Achieve Greater Progress on High-Risk Areas. GAO-19-157SP (Washington, D.C.: March 6, 2019). U.S. Virgin Islands Recovery: Status of FEMA Public Assistance Funding and Implementation. GAO-19-253 (Washington, D.C.: February 25, 2019). 2017 Disaster Contracting: Action Needed to Better Ensure More Effective Use and Management of Advance Contracts. GAO-19-93 (Washington, D.C.: December 6, 2018). Continuity of Operations: Actions Needed to Strengthen FEMA’s Oversight and Coordination of Executive Branch Readiness. GAO-19-18SU (Washington, D.C.: November 26, 2018). Homeland Security Grant Program: Additional Actions Could Further Enhance FEMA’s Risk-Based Grant Assessment Model. GAO-18-354 (Washington, D.C.: September 6, 2018). 2017 Hurricanes and Wildfires: Initial Observations on the Federal Response and Key Recovery Challenges. GAO-18-472 (Washington, D.C.: September 4, 2018). Federal Disaster Assistance: Individual Assistance Requests Often Granted but FEMA Could Better Document Factors Considered. GAO-18-366 (Washington, D.C.: May 31, 2018). 2017 Disaster Contracting: Observations on Federal Contracting for Response and Recovery Efforts. GAO-18-335 (Washington, D.C.: February 28, 2018). Disaster Recovery: Additional Actions Would Improve Data Quality and Timeliness of FEMA’s Public Assistance Appeals Processing. GAO-18-143 (Washington, D.C.: December 15, 2017). Disaster Assistance: Opportunities to Enhance Implementation of the Redesigned Public Assistance Grant Program. GAO-18-30 (Washington, D.C.: November 8, 2017). Climate Change: Information on Potential Economic Effects Could Help Guide Federal Efforts to Reduce Fiscal Exposure. GAO-17-720 (Washington, D.C.: September 28, 2017). Federal Disaster Assistance: Federal Departments and Agencies Obligated at Least $277.6 Billion during Fiscal Years 2005 through 2014. GAO-16-797 (Washington, D.C.: September 22, 2016). Disaster Recovery: FEMA Needs to Assess Its Effectiveness in Implementing the National Disaster Recovery Framework. GAO-16-476 (Washington, D.C.: May 26, 2016). Disaster Response: FEMA Has Made Progress Implementing Key Programs, but Opportunities for Improvement Exist. GAO-16-87 (Washington, D.C.: February 5, 2016). Hurricane Sandy: An Investment Strategy Could Help the Federal Government Enhance National Resilience for Future Disasters. GAO-15-515 (Washington, D.C.: July 30, 2015). Budgeting for Disasters: Approaches to Budgeting for Disasters in Selected States. GAO-15-424 (Washington, D.C.: March 26, 2015). High-Risk Series: An Update. GAO-15-290 (Washington, D.C.: February 11, 2015). Emergency Preparedness: Opportunities Exist to Strengthen Interagency Assessments and Accountability for Closing Capability Gaps. GAO-15-20 (Washington, D.C.: December 4, 2014). Fiscal Exposures: Improving Cost Recognition in the Federal Budget. GAO-14-28 (Washington, D.C.: October 29, 2013). Federal Disaster Assistance: Improved Criteria Needed to Assess a Jurisdiction’s Capability to Respond and Recover on Its Own. GAO-12-838 (Washington, D.C.: September 12, 2012). Opportunities to Reduce Potential Duplication in Government Programs, Save Tax Dollars, and Enhance Revenue. GAO-11-318SP (Washington, D.C.: March 1, 2011). Appendix II: Ongoing GAO Reviews on Emergency Management 1. Review of U.S. Virgin Islands recovery planning and progress; 2. Puerto Rico disaster recovery planning and progress; 3. 2017 wildfire response and recovery; 4. Puerto Rico electricity grid recovery after the 2017 hurricane season; 5. Mass care sheltering and feeding challenges during the 2017 6. Department of Transportation highway and transit emergency relief 7. Drinking water and wastewater utility resilience; 8. Review of disaster death count information in selected states and 9. Department of Health and Human Services disaster response efforts; 10. Disaster and climate change impacts on Superfund sites; 11. FEMA Public Assistance program fraud risk management efforts; 12. Wildland fire fuel reduction efforts; 13. Preparedness challenges and lessons learned from the 2017 14. FEMA workforce management and challenges; 15. Small Business Administration response to 2017 disasters; 16. Development of the GAO disaster resilience framework; 17. FEMA Individuals and Households Program operations and 18. National Flood Insurance Program post-flood enforcement; 19. Emergency alerting capabilities and progress; 20. National Flood Insurance Program buyouts and property acquisitions; 21. Economic costs of large-scale natural disasters and impacts on 22. Community Development Block Grants – disaster recovery; and 23. Disaster Housing Assistance Program. 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.

Why GAO Did This Study In September 2017, two major hurricanes—Irma and Maria—struck Puerto Rico and the USVI, causing billions of dollars in damage to infrastructure, housing, and the economy. FEMA—a component of the Department of Homeland Security—is the lead federal agency responsible for assisting Puerto Rico and the USVI to recover from these natural disasters. Among other responsibilities, FEMA is administering the Public Assistance program in partnership with the governments of Puerto Rico and the USVI, providing them grant funding for response and recovery activities, including debris removal efforts, life-saving emergency protective measures, and the repair, replacement, or restoration of public infrastructure. This statement describes (1) the status of FEMA's Public Assistance grant funding in Puerto Rico and the USVI in response to the 2017 hurricanes as of April 2019, (2) the establishment of recovery offices in Puerto Rico and the USVI, and (3) challenges in implementing the Public Assistance program and actions FEMA has taken to address them. This statement is based on GAO reports issued in February, March, and June 2019, and includes preliminary observations from ongoing GAO reviews of FEMA operations. For ongoing work, GAO analyzed program documents and data on obligations and expenditures; interviewed agency officials; and visited disaster-damaged areas in Puerto Rico and the USVI, where GAO also interviewed FEMA and local officials. What GAO Found GAO's prior and ongoing work found that the Federal Emergency Management Agency (FEMA) obligated about $7.4 billion in Public Assistance grant funding to Puerto Rico and the U.S. Virgin Islands (USVI) as of April 2019, in response to the 2017 hurricanes. FEMA obligated about $6.2 billion in Public Assistance grants for emergency work—debris removal activities, power restoration, and other emergency measures—and about $965 million in Public Assistance grants for permanent work—including the repair or replacement of public infrastructure such as roads, electrical utilities, and damaged buildings. Further, FEMA is continuing to work with Puerto Rico and the USVI to develop additional permanent work projects to repair damaged public infrastructure, such as schools and hospitals (see figure). In 2017, Puerto Rico established the Central Office for Recovery, Reconstruction, and Resilience and in 2019 the USVI established the Office of Disaster Recovery to coordinate and oversee federal recovery efforts. Among other things, these recovery offices are responsible for monitoring and overseeing the Public Assistance program and developing internal controls to ensure it is implemented in accordance with applicable laws, regulations, and FEMA requirements. GAO's prior and ongoing work highlighted challenges with the Public Assistance program including concerns about the clarity of FEMA's guidance, and the time and resources needed to transition to a new Public Assistance delivery model in Puerto Rico. Further, Puerto Rico and USVI officials reported difficulties understanding FEMA's implementation of new flexibilities authorized by law as well as delays in jointly developing cost estimates for long-term recovery projects such as the repair or replacement of hospitals, buildings, and other public infrastructure. FEMA has taken some actions to help address these issues, including developing additional guidance and specific training. However, it is too soon to determine the effectiveness of FEMA's actions. GAO will continue to evaluate the Public Assistance program in the USVI and Puerto Rico and plans to report its findings in late 2019 and early 2020, respectively.

gao_GAO-20-157

gao_GAO-20-157_0

Background States have flexibility within broad federal requirements to design and implement their Medicaid programs. For example, while states must cover certain mandatory groups and benefits, they have the option to cover certain other groups of individuals and benefits. States’ Medicaid plans outline the services provided, the populations covered by their programs, and how they implement and comply with other federal requirements. States share responsibility for oversight of Medicaid eligibility with CMS. Medicaid Eligibility Processes States are primarily responsible for assessing applicants’ eligibility for, and enrolling eligible individuals into, Medicaid. These responsibilities include verifying individuals’ eligibility at the time of application, performing redeterminations of eligibility, and promptly disenrolling individuals who are no longer eligible. In verifying individuals’ eligibility, states must assess specified financial and nonfinancial information. Financial: Individuals applying for Medicaid generally must have an income below a certain limit. PPACA requires states to calculate the income for most nondisabled, nonelderly applicants using a uniform method based on modified adjusted gross income (MAGI), which is derived from a federal tax-based definition of income. States have more flexibility in determining how to calculate incomes for individuals whose eligibility is determined on the basis of age or disability, because their income is not calculated using MAGI-based methods. For example, states may disregard certain types or amounts of income for these MAGI-exempt populations. Additionally, individuals eligible on the basis of age or disability generally must also have assets—cash or real or personal property that are owned and can be converted to cash—below specified standards that vary by state. Nonfinancial: Individuals applying for Medicaid must also satisfy certain nonfinancial criteria. For example, to be eligible for Medicaid individuals generally must be residents of the state in which they are applying and must be either citizens of the United States or certain noncitizens, such as lawful permanent residents. States generally have flexibility in the sources of information they use to verify applicants’ financial eligibility and citizenship or immigration status. However, to the extent practicable, states must use third party sources of data for these verifications prior to requesting documentation from the applicant. When data from reliable third party sources are inconsistent with information from an application, the state must have processes in place to resolve these inconsistencies, such as through requesting additional documentation or accepting the applicant’s attestation. Additionally, states may accept self-attestation for some eligibility criteria, such as residency in the state and household composition (which is used in determining if applicants’ income is below the limit). Once a state determines that an individual meets relevant financial and nonfinancial eligibility criteria, the state enrolls the individual into Medicaid under one basis of eligibility. Examples of bases of eligibility include those applicable to children, pregnant women, individuals eligible for Supplemental Security Income (SSI)—a program that provides cash assistance to low-income adults and children with disabilities—and other low-income adults under age 65 in states that expanded their Medicaid populations under PPACA. (See table 1.) Since individuals may meet the criteria for more than one category and eligibility group, they could have more than one basis of eligibility. For example, a child who is pregnant could meet the criteria applicable to children and those applicable to pregnant women. However, a state would enroll each individual under one basis of eligibility. CMS regulations specify that when states determine applicants eligible based on MAGI criteria, they must notify these individuals of the benefits and services available through any MAGI-exempt bases of eligibility for which they may qualify, in order to provide the individual information about whether to request a MAGI-exempt eligibility determination. However, CMS officials explained that they advise states that they do not need to inform applicants of benefits and services under other eligibility groups if there is no meaningful difference in the benefits or cost-sharing that the individual would receive under one basis compared to another. CMS officials also noted that they have provided further guidance to states on assigning bases of eligibility, including that if an individual meets the criteria for more than one basis, the state should enroll the person into the most beneficial coverage in terms of factors such as the benefit package and out-of-pocket costs. In 2014, CMS issued a framework based on federal rules for states to use in developing their systems to assess individuals’ bases of eligibility. The framework describes a hierarchy for states to use in developing their eligibility systems that begins with bases related to receipt of other federal benefits, such as SSI and federally funded foster care and adoption assistance, which often result in automatic eligibility for Medicaid. Following these bases of eligibility, states are to assess eligibility for bases subject to MAGI-based income rules, and should first evaluate for mandatory coverage before evaluating for optional coverage. Federal rules allow for some exceptions to this sequence, such as when an individual who may be eligible for bases subject to MAGI-based income rules requests consideration under a MAGI-exempt basis to access certain additional benefits, such as long-term services and supports. Federal and State Oversight of Eligibility Determinations CMS has historically operated two distinct, but complementary programs to oversee states’ eligibility determinations in the Medicaid program. The Medicaid Eligibility Quality Control (MEQC) program, which is implemented by states and overseen by CMS, was created in 1978 to monitor the accuracy and timeliness of Medicaid eligibility determinations in order to avoid inappropriate payments and eligibility decision delays. MEQC was also designed to identify methods to reduce and prevent errors related to incorrect eligibility determinations by having states review sample cases to independently verify eligibility criteria and then report the results to CMS. The Payment Error Rate Measurement (PERM) program is CMS’s process to estimate the national Medicaid improper payment rate in accordance with the Improper Payments Information Act of 2002, as amended, and Office of Management and Budget guidance. To calculate the Medicaid improper payment rate through PERM, CMS computes an annual rolling average of improper payment rates across all states based on a 3-year rotation cycle of 17 states each year. PERM is comprised of three components, including one that measures errors in state determinations of Medicaid eligibility. For fiscal years 2015 through 2018, CMS suspended MEQC and the eligibility component of PERM to provide states with time to adjust to eligibility process changes in PPACA; in its place, CMS required states to implement pilots to assess the accuracy of their eligibility determinations. As a result, CMS did not publish an updated national estimate of improper payments due to Medicaid eligibility errors for fiscal years 2015 through 2018. Eligibility reviews under PERM, which are conducted by a federal contractor, resumed in July 2017 for fiscal year 2019. In November 2019, CMS released an updated national estimate of Medicaid eligibility errors, which reflected results of the first 17 states reviewed under the new PERM process. Going forward, states are to resume MEQC reviews in the 2 years between their PERM reviews. The MEQC reviews will focus, in part, on specific areas of improvement for each state. For example, states might choose to focus on specific populations, such as whether pregnant women were assigned to the appropriate eligibility group, or specific processes, such as asset verification. The Medicaid statute includes a provision for CMS to recoup, or disallow, federal funds related to erroneous payments for ineligible individuals and overpayments for eligible individuals. The provision generally requires CMS to recoup funds from states for eligibility-related improper payments if the state’s eligibility error rate exceeds 3 percent. CMS has general authority to recoup funds from states when it determines that an expenditure of federal funding is not an allowable expense; according to CMS, however, this general authority does not apply to eligibility-related errors, given the separate specific statutory authority. Therefore, it is the view of the agency that CMS cannot recoup funds from states whose eligibility-related improper payment rate is below the 3 percent threshold. In addition to the PERM and MEQC oversight, state auditors review Medicaid eligibility determinations, including through audits conducted at the auditors’ initiative and as part of audits required by provisions of the Single Audit Act of 1984. To guide auditors in performing reviews under the Single Audit Act, the Office of Management and Budget issues a document referred to as the Compliance Supplement, which identifies important compliance requirements that the federal government expects to be considered as part of such an audit. Beginning in fiscal year 2014, the Compliance Supplement directed auditors to forgo review of eligibility determinations for individuals whose income is calculated based on MAGI. The supplement noted that testing was being performed under Medicaid eligibility review pilots, which would serve as CMS’s oversight during the initial years of PPACA implementation when the MEQC and the eligibility component of PERM were suspended. In June 2019, the Office of Management and Budget issued the 2019 Compliance Supplement that included changes to permit state auditors to test eligibility determinations of both MAGI and MAGI-exempt populations to ensure enrollees qualify for the Medicaid program and are in the appropriate enrollment category. Basis of Eligibility Decisions Can Vary Among Selected States despite Consideration of Similar Factors The five selected states in our review considered similar factors when ranking the bases of eligibility to which individuals are assigned—such as bases related to children, pregnant women, or disabled individuals—but the resulting basis of eligibility in which individuals were placed could vary. Each of the five states ranked bases of eligibility by comparing how beneficial they were for enrollees across several key factors, and ordered the bases into a hierarchy starting with the most beneficial, according to officials. The states’ eligibility systems were programmed to apply these hierarchies in deciding each individual’s basis of eligibility; when an individual was potentially eligible for more than one basis of eligibility, the system would assign them to the basis highest in the ranking. The key factors the selected states considered in ranking the bases of eligibility, according to state officials, included (1) whether eligibility was related to the receipt of benefits from other programs, (2) the services provided through the benefit package, and (3) the financial implications for the individual. Eligibility related to other programs. The selected states ranked bases of eligibility associated with enrollment in other federal and state assistance programs at or near the top of their hierarchies. For instance, eligibility associated with receipt of SSI was generally at the top of the states’ hierarchies, and eligibility associated with receipt of federal foster care and adoption assistance benefits was ranked above other bases for which a child might be eligible. Services included in the benefit package. Bases of eligibility that conveyed additional benefits, such as long-term services and supports, were ranked higher. Similarly, bases that offered limited benefits, such as only covering family planning services or assistance with cost-sharing for Medicare beneficiaries (i.e., the Medicare Savings Program) were ranked lower in the selected states’ hierarchies. Financial impact. The selected states ranked bases of eligibility lower if they were associated with additional financial requirements for the individual, such as asset tests as a condition of eligibility, or out- of-pocket costs once enrolled. For example, bases of eligibility that required applicants to make copayments to receive certain services, or to pay a monthly premium, were ranked lower than those without such costs. Although the selected states considered similar factors when deciding an individual’s basis of eligibility for Medicaid, a similarly situated individual could be enrolled under a different basis of eligibility in one state versus another state. Decisions varied across states, in part, because of differences in (1) how states factored in the length of the enrollment period; and (2) the degree to which states’ eligibility systems and processes were integrated. Length of the enrollment period. Officials in selected states considered the length of the enrollment period when deciding bases of eligibility for certain populations, such as pregnant mothers (pregnant women who were also eligible as caretakers of dependent children). Pregnant women who are eligible for Medicaid have continuous eligibility, which guarantees enrollment through at least 60 days postpartum regardless of income changes. For this reason, Oklahoma enrolled pregnant mothers under a basis of eligibility applicable to pregnant women. In contrast, Virginia enrolled pregnant mothers under a basis of eligibility applicable to caretakers, because it has a 12-month enrollment period. However, pregnant women have continuous eligibility through at least 60 days postpartum regardless of income changes or whether they are enrolled as caretakers or on some other basis. As such, if a pregnant woman enrolled as a caretaker no longer met the income standard for a caretaker, for example, she could still remain eligible through her postpartum period. Alternatively, a woman enrolled under a pregnancy-related basis of eligibility would be redetermined for eligibility at the end of her postpartum period and could continue enrollment as a caretaker if she continued to meet the financial and other eligibility criteria. CMS noted that such variations in eligibility policies are allowable and expected among state Medicaid programs. Eligibility system integration. Differences in the degree to which selected states integrated their eligibility systems affected how individuals were assessed for potential bases of eligibility and potentially resulted in different eligibility determinations. Officials in four of our five selected states—New Mexico, Oklahoma, Tennessee, and Virginia—reported operating unified or integrated eligibility systems through which individuals could be considered for both MAGI and MAGI-exempt bases of eligibility. The fifth state, Maryland, had separate eligibility systems for MAGI and MAGI-exempt bases of eligibility, so an individual would need to apply through both systems to have all potential bases of eligibility considered. As such, an individual who is over age 65 and a caretaker of a dependent child would have to submit two separate applications to be assessed for all potential bases of eligibility in Maryland. Depending on the system to which he or she applied, that individual could be enrolled in a less beneficial basis of eligibility or denied eligibility for Medicaid. For example, the individual might be determined ineligible for full Medicaid benefits and enrolled in a Medicare Savings Program, in which Medicaid covers out-of-pocket costs related to Medicare benefits. (See fig. 1.) Audits Identified Multiple Issues Related to the Accuracy of Eligibility Determinations; Selected States Had Processes Designed to Address Many Identified Issues Our review of 47 state and federal audits across 21 states identified multiple issues affecting the accuracy of states’ Medicaid eligibility determinations. The accuracy issues identified in the audits we reviewed generally fell into nine broad categories, such as eligibility determinations made with incorrect or incomplete income or asset information, unresolved discrepancies between what applicants reported as their income and electronic data sources, and unidentified or unaddressed changes in circumstances, such as changes in household income or size. (See table 2.) Within these nine broad categories, the audits identified several specific accuracy issues, including states that were not conducting income checks for individuals reporting no income; not terminating the enrollment of individuals who had moved out of enrolling individuals who did not provide required information (such as proof of citizenship) on a timely basis; months or years behind schedule in conducting required eligibility not acting on—or not having adequate systems in place to detect— changes in enrollees’ circumstances that could affect eligibility, such as changes in income or household composition. See table 3 for examples of audit findings related to each of the accuracy issue categories, and appendix I for an overview of the key findings for each audit we reviewed. In some cases, the accuracy issues identified by auditors resulted in errors in eligibility determinations, such as instances when applicants were determined eligible even though their incomes were above the applicable limit, or instances in which the state did not enroll eligible individuals. However, in other instances, the accuracy issues identified by auditors did not result in erroneous eligibility determinations. For example, in some cases the audit found that a state determined that an applicant was eligible based on incorrect or incomplete financial information; however, auditors found that the applicant would have still been eligible for Medicaid even after reviewing additional financial information. In other cases, auditors found that eligibility determinations complied with state policies and federal requirements, but that changes in state policies— such as using additional data sources or checking sources more frequently—could provide more information that could be used to improve eligibility determinations. For example, audits in three states found that the quarterly wage data the states used to verify income did not detect certain nonwage income; that income could have been identified had the states chosen to use state or federal tax data as a verification source. Auditors in one of these states (Louisiana) also found that checking income data during individuals’ coverage period, such as on a quarterly basis, could have saved the state tens of millions of dollars in managed care fees for individuals whose incomes exceeded eligibility thresholds during their enrollment period. The selected states we reviewed reported having processes in place that were designed to avoid or address many, but not all, of the accuracy issues identified. The following are examples of the states’ processes related to specific accuracy issues. Incorrect or incomplete income or asset information. All five selected states we reviewed reported checking electronic data sources to verify income, including for individuals who report $0 in income. Officials from some states noted, however, that the electronic sources they have chosen to use do not include all relevant types of income, such as self- employment income. The five states also reported having electronic asset verification systems to verify financial assets, such as bank and retirement accounts for applicants subject to asset limits. One state (New Mexico) reported that it recently implemented an asset verification system that includes information from financial institutions, property ownership records, and vehicle licensing. Eligibility redeterminations not made in a timely manner. To help ensure that redeterminations are made in a timely manner, all five selected states reported conducting automatic redeterminations for at least some MAGI enrollees using electronic data sources to confirm continued eligibility. The proportion of MAGI enrollees whose eligibility was automatically redetermined ranged from about 10 to 80 percent. Officials from Virginia, which was cited by auditors in 2015 as having significant delays in conducting redeterminations, reported that automatic redeterminations have helped improve timeliness. Where automatic eligibility redeterminations are not conducted—such as for enrollees whose incomes could not be confirmed through electronic sources or who are eligible on a MAGI-exempt basis—the five selected states reported having systems in place to generate a redetermination packet or notice to be sent to enrollees prior to the end of their eligibility period. Officials reported that enrollees who do not complete their redetermination would be disenrolled, with states varying in how quickly they would take such action. For example, Oklahoma officials reported that the state automatically terminates enrollment for individuals who do not reply with the required information by the end of their coverage period. In contrast, Virginia officials reported that redeterminations for which no response was provided are kept open, pending eligibility worker action; the state’s systems do not automatically terminate enrollment. Unresolved income discrepancies. Officials in the five selected states reported that their eligibility systems automatically identify income discrepancies. For example, Oklahoma officials indicated that if there is more than a 5 percent difference in the income reported on the application and the income from electronic data sources, their system either alerts eligibility workers or automatically sends a request for additional information to the enrollee. Individuals enrolled in incorrect basis of eligibility. According to state officials, their eligibility systems have automated checks to reassess the eligibility for individuals reaching certain milestones, such as the maximum age for their basis of eligibility (i.e., children reaching adulthood and adults reaching age 65) and pregnant women who are approaching the end of their 60-day postpartum period. For example, to help ensure individuals are correctly assigned to the appropriate basis of eligibility, officials in Maryland noted that they apply system edits that preclude individuals who are pregnant, age 65 or older, or enrolled in Medicare from being incorrectly assigned to the new adult group. Unidentified or unaddressed changes in circumstances. Officials from the five selected states indicated that they generally had systems in place to identify if an enrollee had died or moved out of state. For example, officials from the five selected states reported conducting periodic checks of residency through the Public Assistance Reporting Information System—a federal data source that identifies individuals receiving benefits in other states—and following up with identified enrollees to see if they still reside in the state. None of the selected states conducted regular reviews to identify changes in MAGI enrollees’ incomes during the enrollment period, although one state—Oklahoma— planned to implement interim checks of income in response to a recent change in state law. Oklahoma also reported that it conducted quarterly checks of wage data for MAGI-exempt enrollees. Use of incomplete or incorrect information on household composition. The selected states generally did not have processes in place to detect accuracy issues related to household composition, although officials in four of the five states—Maryland, New Mexico, Oklahoma, and Virginia—noted that eligibility information from other benefit programs may be compared with Medicaid files to detect changes or discrepancies in household membership. Previous Reviews CMS Used for Measuring Eligibility Errors Were Insufficient to Recoup Funds from States; New Procedures Are in Place for 2022 In 1983, CMS implemented its statutory requirement to recoup funds associated with Medicaid eligibility-related improper payments for states with an eligibility error rate above 3 percent through its MEQC program. The MEQC program required states to randomly sample Medicaid enrollees to verify eligibility. Claims related to enrollees determined ineligible were tallied and compared with total claims for the sample universe to calculate an error rate. Following federal validation, states were subject to recoupment of all or part of the federal funds expended related to erroneous state payments over the 3 percent error rate threshold. However, in 1992, HHS’s Departmental Appeals Board—the department’s final level of administrative review—concluded that the MEQC error rate was not sufficiently accurate to provide reliable evidence to support recoupment of funds due to the small sample size from which the error rate was calculated. Consequently, the appeals board stated that it was “impossible to conclude with a reasonable certainty that the States failed to meet their target rates….” As a result of this opinion, CMS provided states the option, beginning in 1994, to either continue operating a traditional MEQC program or to conduct what CMS referred to as “MEQC pilots,” which focused on prospective improvements in eligibility determinations, rather than calculation of error rates. Since the “MEQC pilots” did not produce an error rate, CMS could not recoup federal funds expended due to erroneous eligibility determinations for states participating in the pilots. Between 2012, the earliest year for which CMS has maintained records, and 2014 when CMS suspended the MEQC program, 39 states participated in these “MEQC pilots” exempting them from possible recoupment of funds due to eligibility errors. While the other 12 states that continued to operate traditional MEQC programs could still be subject to recoupment of funds, CMS officials reported that no recoupments related to eligibility errors had occurred since the 1992 appeals board ruling, because none of these states had an error rate exceeding the 3 percent threshold. Thus, CMS has not recouped federal funds due to eligibility errors in decades. However, the agency has introduced new procedures through which it can, under certain circumstances, begin to recoup funds based on eligibility errors in fiscal year 2022. Specifically, in July 2017, CMS issued new regulations that included changes to its PERM process to satisfy the statutory requirements for recouping funds that MEQC was previously designed to operationalize. Under the revised PERM rules, CMS calculated an eligibility error improper payment rate beginning with the cohort of states under review for the fiscal year 2019 reporting period. However, it will not recoup funding from states with error rates exceeding the 3 percent threshold until states have a second review under the revised PERM rules, which will occur for the first cohort of states in fiscal year 2022. This allows each state the opportunity to implement improvements based on its initial PERM review and the MEQC review it will conduct in the off-cycle years to reduce the error rate or demonstrate a “good faith effort” to do so. CMS officials recognize the benefits of using state and federal audits, such as audits we reviewed for this report, as part of a broader strategy to improve program integrity and oversee states’ eligibility determination processes. However, CMS officials told us they do not have the authority to recoup federal funds related to eligibility errors identified outside of the PERM process, such as through state single audits. According to CMS officials, this is because of the specific statutory instruction limiting recoupments to instances when eligibility-related errors exceed the 3 percent error rate threshold, and because PERM is the process that CMS uses to calculate that error rate. The President’s fiscal year 2020 budget request included a legislative proposal to expand HHS’s authority to issue disallowances for eligibility errors. Specifically, the proposal requests legislative authority to permit HHS to issue disallowances outside of PERM and allow HHS, including the HHS-OIG, to extrapolate findings on beneficiary eligibility to ensure federal recovery of incorrect eligibility determinations; and eliminate the current 3 percent threshold for states’ eligibility-related improper payments. In place of the current 3 percent disregard, HHS would issue rulemaking specifying criteria for the recoupment of funds, including limiting them to instances of monetary loss, such as cases in which ineligible individuals received benefits. Concluding Observations Determining whether individuals are eligible for Medicaid is a complex process that is vulnerable to error. The processes used to measure the extent of eligibility errors have been, and will continue to be, in a state of transition over the next several years as CMS implements its new PERM procedures and states implement improvements after their initial PERM reviews under these new procedures. Because CMS has not had a complete national estimate of improper payments due to eligibility errors since 2014, policymakers and other stakeholders have had an incomplete picture of the extent of eligibility errors in the Medicaid program nationally. This state of flux will make the findings from federal and state audits an even more important source of information on the accuracy of states’ eligibility determinations. As we have previously reported, oversight of the Medicaid program could be further improved through leveraging and coordinating program integrity efforts with state auditors to further improve the integrity of the Medicaid program. Agency Comments We provided a draft of this report to HHS for review and comment. HHS provided technical comments, which we incorporated as appropriate. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Secretary of HHS, the Administrator of the CMS, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact me at (202) 512-7114 or yocomc@gao.gov. Contact points for our Office of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. Appendix I: Summary of Federal and State Audits of the Accuracy of States’ Medicaid Eligibility Determinations Table 4 provides a summary of key findings from the 47 federal and state audits that discussed the accuracy of states’ Medicaid eligibility determinations, published from 2014 through 2018, which we identified and reviewed. Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Michelle Rosenberg (Assistant Director), Perry Parsons (Analyst-in-Charge), and Heather Tompkins made key contributions to this report. Also contributing were Drew Long, Vikki Porter, and Jenny Rudisill.

Why GAO Did This Study In fiscal year 2018, Medicaid covered approximately 75 million individuals at an estimated cost of $629 billion, $393 billion of which were federal funds. Medicaid eligibility is governed by a network of federal and state laws and regulations. In assessing eligibility for Medicaid, states must determine whether applicants meet eligibility criteria, such as financial and citizenship requirements. The accuracy of eligibility decisions has implications for federal and state spending. The Patient Protection and Affordable Care Act made significant changes to Medicaid eligibility rules beginning in 2014, including new ways of calculating income and new requirements related to electronically verifying applicants' information. Yet, little is known about the accuracy of states' Medicaid eligibility determinations since these changes were implemented. GAO was asked to review Medicaid eligibility determinations. This report describes, among other things, what is known about the accuracy of Medicaid eligibility determinations, and CMS's efforts to recoup funds related to eligibility errors. GAO reviewed 47 state and federal audits of Medicaid eligibility determinations across 21 states published between 2014 and 2018. GAO also reviewed relevant federal laws and regulations, and interviewed CMS officials. The Department of Health and Human Services provided technical comments on a draft of this report, which GAO incorporated as appropriate. What GAO Found States are responsible for determining applicants' eligibility for Medicaid, including verifying eligibility at application, redetermining eligibility, and disenrolling individuals who are no longer eligible. The Centers for Medicare & Medicaid Services (CMS) oversees states' Medicaid eligibility determinations. CMS did not publish an updated national Medicaid eligibility improper payment rate from 2015 through 2018 as states implemented the Patient Protection and Affordable Care Act. CMS released an updated rate in November 2019 that reflected new information on eligibility errors from 17 states. In lieu of complete and updated data, GAO reviewed 47 state and federal audits published between 2014 and 2018 related to 21 states' eligibility determinations. The identified accuracy issues did not always result in erroneous eligibility determinations. For example, some audits found applicants were determined eligible based on incomplete financial information, but when the audits reviewed additional information they found that the applicants still would have been eligible for Medicaid; and eligibility determinations complied with state policies and federal requirements, but noted that changes in state practices—such as using additional data sources to verify applicant information or checking sources more frequently—could improve eligibility determinations. While CMS is generally required to disallow, or recoup, federal funds from states for eligibility-related improper payments if the state's eligibility error rate exceeds 3 percent, it has not done so for decades, because the method it used for calculating eligibility error rates was found to be insufficient for that purpose. To address this, in July 2017, CMS issued revised procedures through which it can recoup funds for eligibility errors, beginning in fiscal year 2022. In addition, the President's fiscal year 2020 budget request includes a legislative proposal to expand the agency's authority to recoup funds related to eligibility errors. During this period of transition, federal and state audits will continue to provide important information about the accuracy of states' eligibility determinations.

gao_GAO-20-187

gao_GAO-20-187_0

Background Funding for university STEM research. The five federal agencies included in our analysis provide billions of dollars annually for university research in STEM fields, with HHS-NIH providing more than the other four agencies combined. Figure 1 details the total amount of basic and applied STEM research funding provided to universities by each agency in fiscal year 2018, these data are preliminary and the most current data available during our review. Sexual harassment. While sexual harassment is not a defined term in Title IX, it can constitute sex discrimination under Title IX in some circumstances if, among other things, the harassment is “so severe, pervasive, and objectively offensive that it effectively bars the victim’s access to an educational opportunity or benefit.” Some federal agencies and NASEM define sexual harassment based on specific behaviors. Specifically, according to NASEM, sexual harassment encompasses three types of behavior: Sexual coercion: Favorable treatment conditioned on sexual activity. Unwanted sexual attention: Verbal or physical sexual advances that are unwelcome, including sexual assault. Gender harassment: Sexist hostility and crude behavior. The most common form of sexual harassment is gender harassment, which generally involves behavior that conveys hostility, objectification, exclusion, or second-class status about a person’s gender. According to the Consultant Report on the University of Texas System Campus Climate Survey for NASEM, female medical and engineering students enrolled in the University of Texas System were more likely to experience sexual harassment by faculty or staff compared with students enrolled in other majors. Title IX enforcement responsibilities. Title IX of the Education Amendments of 1972 is the primary federal law that addresses sex discrimination under education programs or activities receiving federal financial assistance, including federally funded grant programs at educational institutions, such as universities. Under Title IX, federal agencies that award grants to educational institutions, including universities, have enforcement responsibilities to ensure such institutions do not discriminate based on sex. Enforcement responsibilities fall under four main areas: 1. Issuing regulations. Title IX requires that agencies promulgate regulations to provide guidance on Title IX enforcement to recipients of federal financial assistance who administer education programs or activities. 2. Obtaining assurance from university grantees that they are in compliance with Title IX. Most of the agencies we reviewed require grantees to submit an “assurance of compliance” form as part of their grant application or award to attest compliance with anti-discrimination laws, including Title IX. 3. Conducting periodic compliance reviews of funding recipients. Grant funding agencies are required to conduct periodic Title IX compliance reviews of university grantees. A Title IX compliance review is an agency’s assessment of whether a grantee is complying with the law. According to DOJ’s Title IX legal manual, federal agencies have broad discretion in determining which grantees to review for compliance. Federal agencies may conduct these reviews on-site at a university (grantee) or via a desk audit. In both types of review, agency officials, among other things, review documentation that indicates compliance, such as the universities’ policies and procedures for receiving, investigating, and resolving Title IX complaints. During on-site reviews, officials interview staff, faculty, and students about their awareness of Title IX and any issues of potential sex discrimination that they have encountered. 4. Investigating written complaints of sex discrimination against recipients in a timely way. Federal agencies are required to establish and publish procedures for the prompt processing and disposition of complaints. An individual alleging discrimination on the basis of sex by a university can file a discrimination complaint with multiple entities, including the university or one of the federal agencies that provides funding to the university, which could include Education or another funding agency. In 2015, we reported on six federal agencies’ grant making to women in STEM research, including their Title IX compliance activities. We found that the Departments of Defense, and Health and Human Services were not conducting required Title IX compliance reviews at universities they funded and recommended that they periodically do so. In response to our recommendation, HHS conducted three Title IX compliance reviews in 2018 and according to officials, the agency initiated additional Title IX compliance reviews in 2019 and 2020. Education and DOJ also have responsibilities for administering Title IX. Education plays a key role in ensuring compliance with Title IX as it provides funding to most universities in the United States. DOJ’s Civil Rights Division is responsible for enforcing federal statutes prohibiting discrimination of protected classes, including Title IX. Under Executive Order 12250, DOJ also has the responsibility for playing a leadership role in coordinating the “consistent and effective implementation” of several civil rights laws, including Title IX. In 2015, we reported that DOJ had no formal information-sharing process for federal agencies to exchange best practices on Title IX compliance activities, and we recommended that it establish such a process. In response to our recommendation, DOJ reconstituted the Quarterly Title IX STEM discussion group in February 2016 to facilitate information sharing across the six major STEM grant-making federal agencies. Recipients of federal assistance—in this case, university grantees—also have Title IX compliance responsibilities. Specifically, universities are responsible for ensuring Title IX compliance, designating an employee to coordinate compliance (e.g., a Title IX coordinator), establishing procedures to promptly and equitably resolve student and employee complaints of sex discrimination made against the university, and publishing a notice stating that they do not discriminate on the basis of sex. Figure 2 outlines the various compliance activities required under Title IX and the entity responsible for carrying out each activity. Offices and their responsibilities for Title IX and grant management. Among the federal agencies we reviewed, different offices handle various aspects of Title IX and grant compliance activities. Generally, each agency’s civil rights or diversity office conducts Title IX compliance reviews, develops policies and procedures for grantees, and investigates allegations and complaints involving university researchers supported by their agency’s federal STEM grants. All five agencies (DOE, HHS, NASA, NSF, and USDA-NIFA) primarily address Title IX complaints, including sexual harassment complaints, through their civil rights or diversity offices. However, these offices are responsible for more than just addressing complaints and preventing sexual harassment at grantees, including universities; these offices oversee a number of civil rights, diversity, and inclusion efforts for the entire agency. Moreover, most of these offices also address internal employee sexual harassment complaints and other discrimination issues. The agency office that awards grants generally creates and modifies grant terms and conditions for universities receiving funding from the agency. Table 1 outlines each agency’s Title IX and grant management responsibilities by office. All Five Agencies Conducted Compliance Reviews, and Two Published Promising Practices for Universities All five agencies conducted periodic Title IX compliance reviews, as required by federal laws and regulations, from fiscal years 2015 through 2019, and three completed joint compliance reviews. Two agencies publicized promising practices from Title IX compliance reviews on their websites and did so to assist all grantees with Title IX compliance. The other three agencies have not clearly publicized such practices from their Title IX reviews on their websites. All Five Agencies Conducted Required Title IX Compliance Reviews from Fiscal Year 2015 through 2019 The five agencies we reviewed conducted periodic Title IX compliance reviews, as required by federal laws and regulations. From fiscal year 2015 through 2019, DOE, HHS, NASA, NSF, and USDA-NIFA officials reported that their agencies met the requirement for conducting periodic reviews. During this period, the agencies conducted between 4 and 11 Title IX compliance reviews among hundreds of grantees. No agency completed more than three reviews in a fiscal year. Two agencies— DOE and NASA—have requirements to conduct a minimum of two Title IX compliance reviews annually. DOE and NASA meet their statutory requirements by starting at least two Title IX compliance reviews each year, according to officials. HHS, NSF, and USDA do not have an annual minimum requirement and are not required to have one, according to officials. Agencies conducted visits to universities to assess compliance and developed written compliance reports. In the compliance report, agencies can recommend a grantee take action to improve existing compliance efforts to prevent sex discrimination and may highlight promising practices by grantees. For example, NASA recommended in a written compliance report that a grantee provide more targeted Title IX training geared toward STEM students and faculty, noting that such training should focus on subtle forms of gender bias that pervade STEM programs as well as on more egregious examples of sexual harassment. Similarly, HHS made recommendations in three of its compliance reviews for grantees to notify students and faculty of their right to file a Title IX complaint with the HHS Office for Civil Rights. University grantees are not required to implement the agency’s recommendations, but they must take corrective actions to resolve findings of Title IX noncompliance, according to DOE, HHS-OCR, NASA, and NSF officials. Agencies are required by law to seek voluntary compliance for Title IX violations. If an agency finds that a grantee has violated Title IX (noncompliance), it first seeks to establish voluntary compliance through a resolution agreement—an agreement with the agency and grantee outlining corrective actions for the grantee. If the agency is unable to achieve voluntary compliance in a Title IX case, it may initiate proceedings to suspend or terminate federal funding, or refer the case to DOJ for possible litigation. According to officials, the five agencies we reviewed have not suspended or terminated funding to enforce Title IX, including sexual harassment. Instead, according to agency officials, their reviews have found that most grantees are in compliance with Title IX from fiscal year 2015 through 2019, except for one grantee, where the agency worked with the grantee to achieve voluntary compliance. To leverage limited resources, three of the five agencies—DOE, NASA, and NSF—conducted joint Title IX compliance reviews. These reviews occur when two agencies providing funding to the same grantee jointly assess whether the grantee is complying with the law. DOE and NSF conducted three joint compliance reviews in fiscal years 2015 and 2016, while NASA and NSF conducted a joint review in fiscal year 2019. These joint reviews helped agencies leverage resources. Two Agencies Publicized Promising Practices for Title IX Compliance to Assist All Grantees, but the Other Three Agencies Have Not Clearly Publicized Practices from Title IX Reviews on Their Websites NASA and NSF publicized on their websites a list of promising practices identified as part of their compliance reviews to assist grantees with Title IX compliance. Promising practices—grantee actions that have the potential to advance equal opportunities, diversity, and inclusiveness for program participants regardless of sex—may be considered, adopted, and replicated by other grantees, according to NASA and NSF officials. Some actions may go beyond meeting Title IX compliance requirements. NASA identifies promising practices to provide grantees with information and examples on practices they may wish to consider replicating to help enhance or supplement their equal opportunity efforts, according to officials. For example, NASA noted a promising practice in which a grantee presents campus training sessions on Title IX at which participants develop bystander behavior skills, discuss consent and sexual respect, and learn how to encourage and support reporting of sexual misconduct. In addition, this university grantee facilitates the workshop using clickers to allow real-time, anonymous audience response, enabling the facilitators to measure learning progress and see attitudinal shifts in real time. According to NASA, since the inception of its Title IX compliance program, the agency has followed a philosophy of providing meaningful technical assistance to universities, including identifying and reporting on promising practices of the universities that the agency reviews. NASA officials told us this approach mitigates the fact that the agency only has the resources to conduct compliance reviews at a few of its hundreds of grantees annually. NSF takes a similar approach. For example, NSF’s webpage for promising practices has a link to a university’s complaint resolution flow chart as an example for others to draw on. According to NSF officials, the agency values opportunities to learn about practices that have the potential to make significant and meaningful impacts on grantees’ efforts to create and maintain research environments that are safe and free from sexual and gender-based harassment. Moreover, according to officials, NSF grantees have expressed gratitude to the agency for sharing what other universities are doing that is working well. In contrast, while DOE, HHS, and USDA identified promising practices in some of their Title IX compliance reviews, they have not clearly publicized a list of these practices to the broader grantee community. DOE has posted reports of Title IX compliance reviews, but no list of promising practices. As a result, grantees who want to learn from other universities would need to review individual compliance reports and search for promising practices. DOE does plan to develop a publication that identifies promising practices and lessons learned from its Title IX compliance reviews in fiscal year 2020, according to officials. The agency did not provide any plans or timeframes because officials stated that DOE’s Office of Civil Rights is determining the best approach for this project. USDA-NIFA is planning to create mechanisms to publicize best practices, according to officials, but it has not yet done so. According to USDA officials, the agency is discussing and determining the best promising practices from compliance reviews to publicize; however, recent staff changes have delayed this effort. As a result, USDA did not provide further details about how and when it will publicize promising practices. In October 2019, HHS’s Office for Civil Rights (HHS-OCR) updated its Title IX webpage to include a section dedicated to sexual harassment, including links to resources, guidance, and effective practices (also called promising practices) from other agencies, as well as a written resolution agreement between HHS and a university grantee that resolved findings of sex-based harassment. While a dedicated webpage for sexual harassment is a positive step, HHS’s webpage includes promising practices from other federal agencies—Equal Employment Opportunity Commission and NSF—but not HHS. HHS- OCR officials told us that promising practices are similar across federal agencies. However, HHS Title IX compliance reviews cover grantees that may be different from other federal agencies, such as medical colleges, and these grantees may face unique challenges in complying with Title IX. For example, according to the 2018 NASEM report, women students, trainees, and faculty in academic medical centers experience sexual harassment by patients and patients’ families, in addition to the harassment they experience from colleagues and those in leadership positions. HHS-OCR officials told us that the resolution agreement lists corrective actions that may be considered promising practices. However, a grantee who wanted to learn about these practices would need to know that they exist in the agreement and then review the document to find them. The agency has already identified potential promising practices in some its completed Title IX reviews. Therefore, publishing a separate list of these practices and corrective actions from resolution agreements on its website would require few resources and could benefit grantees. According to Standards for Internal Control in the Federal Government, management should use quality information to achieve its objectives and externally communicate such information to achieve objectives. The vast majority of grantees are reviewed for Title IX compliance infrequently by the five agencies and therefore receive little to no information on such compliance from these agencies. Moreover, while grantees can access completed Title IX reviews on some agencies’ websites, this endeavor would still require grantees to review the written reports in detail to uncover any promising practices. Without clearly publicizing promising practices to the broader grantee community, such as a stand-alone list of practices, DOE, HHS-OCR, and USDA are missing an opportunity to provide quality information to grantees about how best to ensure compliance with Title IX requirements and reduce the likelihood of sexual harassment. Agencies Vary in their Efforts to Address Allegations of Sexual Harassment The five agencies we reviewed received Title IX complaints, but varied in their efforts to address sexual harassment allegations, including: 1) finalizing procedures for processing Title IX complaints, 2) communicating complete information about the complaint process to grantees, and 3) addressing allegations outside of the Title IX process. Four agencies received three or fewer formal Title IX complaints total from fiscal year 2015 through 2019. Two of the five agencies do not have written procedures for the prompt processing and disposition of Title IX complaints—including allegations of sexual harassment—as required by federal regulations. According to agency officials, all five agencies use their websites as the primary means of communicating Title IX complaint information to grantees and individuals at universities; however, one of the five agencies’ websites does not provide clear guidance for grantees on the basics of the complaint filing processes—such as who can file. Additionally, two agencies have gone beyond the formal Title IX complaint process and also review sex discrimination concerns—including sexual harassment—as a means of improving agency Title IX oversight of university grantees. Four Agencies Received Few Formal Title IX Complaints Title IX affords individuals the ability to file formal complaints of Title IX violations directly to the federal agency providing funding for the program. According to agency officials, the five agencies generally define formal complaints as those that: Are submitted in writing; Are filed within 180 days of the incident—or if ongoing, within 180 days of the last incident—to be considered timely; Provide the name and contact information of the person who is Provide a general description of the person or people injured by the alleged discriminatory act(s) (names of those injured are not required); and Provide a description of the alleged discriminatory act(s) in sufficient detail to enable the agency to understand what occurred, when it occurred, and the basis for the alleged discrimination (sex discrimination in the case of Title IX). All five agencies accept formal Title IX complaints in multiple ways— including at minimum through email and postal mail. From fiscal year 2015 through 2019, four agencies received three or fewer formal complaints (see table 2). Agency officials provided several reasons why they believe agencies receive few formal Title IX complaints: Complaints are more commonly filed with the university or with Education and are rarely directly reported to the agency; Individuals may be unaware of their right to file complaints directly with the agency or how to file such a complaint; and Individuals may fear retaliation or a negative impact on their scientific career (see sidebar). Retaliation At a state university, a graduate student reported her advisor for sexual harassment. The university substantiated her claim and the professor left the university. According to the victim, fellow students upset at the impact of the professor’s departure on their own research and academic careers, retaliated against her. Student-centered retaliation included taking her lab equipment without permission, ostracizing her from social events, and withholding critical information and resources necessary for her research. This retaliation caused her to move her workspace and lose progress on her own work. Officials from DOE, NASA, HHS-NIH, and NSF stated they usually learn about instances of sexual harassment from other sources (e.g. media reports) and rarely from voluntary reporting from universities or other federal agencies. Title IX officials at two universities we interviewed agreed with agency officials about why few formal complaints are filed with agencies. For example, one Title IX official stated that concerns about retaliation for filing a complaint are amplified when there is an agency involved due to concerns over risk to the funding. Officials from NSF, which received the most formal Title IX complaints of the five agencies from fiscal year 2015 through 2019, stated that complaints to their agency have increased in recent years. They could not state definitively the reason for the recent increases, but said it may stem from the increased publicity of sexual harassment cases in STEM—including a Twitter movement known as #MeTooSTEM—along with NSF’s revised grant terms and conditions. Two Agencies Lack or Have Outdated Title IX Complaint Procedures DOJ’s regulations provide that federal agencies must establish and publish procedures for the prompt processing and disposition of complaints. While all five agencies specify general requirements for a formal complaint, two do not have clear or updated written procedures for processing and disposing of formal Title IX complaints. Specifically: While DOE’s agency regulations stipulate that the agency will investigate allegations of discrimination under Title IX, agency officials stated that DOE does not currently have written Title IX complaint procedures. In November 2019, the agency provided a preliminary draft outline of its procedures, but officials stated that the agency does not have a timeline for when they may be finished. This is because the agency is devoting its resources to investigating current complaints, according to DOE officials. The website for USDA’s Assistant Secretary for Civil Rights—the office handling complaints across the agency—contains a summary of procedures used to process and investigate discrimination complaints, but a USDA official stated that the procedures need more clarity with regard to the university and research environment. In November 2019, USDA officials highlighted a 1999 USDA Departmental Regulation that addresses processing administrative complaints of discrimination filed against any program or activity receiving financial assistance from USDA. Officials stated that this regulation was revised in fiscal year 2019 and is currently under review for approval. The Departmental Regulations as they stand are outdated, referencing out-of-date organizational responsibilities and department names. Despite the absence of formal complaint procedures for DOE and outdated procedures for USDA, both agencies have evaluated formal Title IX complaints to determine if they meet the necessary criteria for investigation. Specifically, according to agency officials, USDA evaluated and investigated a formal Title IX complaint in fiscal year 2017, and DOE is currently evaluating a complaint to determine if it meets the criteria for a formal complaint. However, without clear and specific guidance for the processing and disposition of complaints, DOE is not complying with DOJ’s regulations—which require federal agencies to establish and publish complaint procedures—and may not be able to consistently and efficiently handle formal Title IX complaints. Moreover, under Standards for Internal Control in the Federal Government, management should implement control activities through policies. For example, management should periodically review procedures for continued relevance and effectiveness in achieving the entity’s objectives or addressing related risks. Without updated complaint procedures, USDA does not have procedures that are aligned with the current structure and oversight responsibilities of the agency. In contrast, NASA, NSF, and HHS-OCR have developed written procedures for how the agency receives, investigates, and resolves formal Title IX complaints. As seen in appendix I, the formal complaint process is extensive and complex, with evaluative benchmarks to be met prior to investigation. USDA’s Complaint Website Is Unclear on Who Can File Title IX Complaints According to officials, all five agencies use websites as the main mechanism for communicating information about Title IX complaints to individuals at university grantees. However, USDA’s website is not clear about who can file a Title IX complaint. NASA, HHS-OCR, NSF, and DOE each have a website intended to provide information about filing complaints specifically to individuals involved with agency-funded grants. On these websites, the agencies state that they accept Title IX complaints or sex discrimination complaints, among other types. In contrast, USDA communicates information on its complaints process via a general discrimination website that is not specific to Title IX complaints or to individuals on agency-funded grants, making it unclear who can file Title IX complaints. Specifically, USDA provides information about complaint resolution through its Office of Assistant Secretary for Civil Rights website, but grantees wishing to file a formal complaint would need to do so as a “customer,” a term that is not defined on the agency’s website and that individuals on agency-funded grants may not recognize as including them. A USDA official acknowledged that such individuals may not realize that they can file through the website. USDA officials told us that the agency would consider revising its website to make clear that individuals on USDA grants can file a formal discrimination complaint with the agency. If USDA does not revise its website, the lack of clarity about who is a customer that can submit a complaint may inhibit its ability to obtain information necessary for Title IX oversight. Under Standards for Internal Control in the Federal Government, management should externally communicate the necessary quality information to achieve the entity’s objectives. For example, management may communicate and receive information through established reporting lines, such as websites, from external parties that can help the agency achieve its objectives, such as oversight of Title IX. The lack of clear communication of quality information may reduce the effectiveness of USDA’s Title IX enforcement. Unlike USDA, both HHS-OCR and DOE recently took action to improve website clarity on who can file a Title IX complaint. Specifically: HHS-OCR’s website provides information on filing formal complaints for multiple forms of discrimination—such as race, age, and sex discrimination—and allows formal complaints from all who feel they have been discriminated against by a program or activity that receives funding from any part of HHS. In part due to issues raised during the course of our study, HHS-OCR published several new or updated websites in October 2019—including a Title IX page with university- based examples of entities covered under Title IX and a website on sex-based harassment outlining definitions and examples of what constitutes sex discrimination under Title IX. Before this update, it was not clear if individuals working on HHS-funded grants could file formal Title IX complaints via HHS-OCR’s website. In October 2019, DOE updated its Title IX website to include clear information on the multiple ways individuals can file a formal Title IX complaint with the agency, after we pointed out that this information was missing from DOE’s website, according to officials. The updated website specifies that individuals can notify the agency of a Title IX complaint in person, by email, fax, or mail. Prior to this update, the website only provided information on how to mail the agency a Title IX complaint. Two Agencies Communicate Other Reporting Options for Concerns In addition to investigating Title IX complaints as required by Title IX, two agencies—HHS-NIH and NSF—go beyond this requirement by also reviewing concerns of sex discrimination—including sexual harassment— and publicly communicating the option for individuals to notify the agency of such concerns outside of the formal Title IX complaint process. “Concerns” are generally defined as information from individuals seeking to inform or notify the agency that sex discrimination has occurred or is occurring, but information is not intended to be a formal Title IX complaint. For example, HHS-NIH established a website, email, and online portal specifically for concerns of sexual harassment, publicly communicating this effort not only on the website, but also in public presentations and official statements. NIH officials stated the agency began reviewing concerns to provide clear channels of communication to NIH. HHS-NIH also developed internal guidance, which is still evolving, for agency staff on how to process concerns from individuals at university grantees through coordination with the grantee. While formal Title IX complaint investigations are agency-led, investigations of sex discrimination concerns, including sexual harassment, filed with HHS-NIH are university-led, with HHS-NIH assessing the university grantee’s response to the allegation to ensure appropriate action is taken to ensure a safe research environment (see appendix II for more details). NSF also publicly communicates the option to notify the agency of concerns of sex discrimination via their Awardee Civil Rights website. In addition to providing details on who should file a formal complaint and how, NSF also provides information on how to notify the agency of concerns and what is done with this information. For example, NSF has initiated a Title IX compliance review for fiscal year 2020 based in part on information contained in a concern it received, according to officials. In fiscal year 2019, HHS-NIH received 93 concerns of sex discrimination and NSF received 47, according to officials from each respective agency (see table 3). In contrast, the remaining agencies—NASA, USDA, DOE, and HHS- OCR—do not publicly communicate the option to notify the agency of concerns of sex discrimination or sexual harassment. Although these agencies stated that they do review all information received—including information from those seeking to notify the agency of concerns—the review is primarily to determine if the information provided meets the agency’s criteria for a formal complaint. Those complaints meeting the criteria for a formal complaint are processed by the agency following the legally required Title IX complaint process. According to officials, the agencies may use the information from concerns to help select a site for a Title IX compliance review. As shown in table 3 above, officials from DOE, USDA, and NASA stated their respective agencies received no concerns of sex discrimination in fiscal year 2019, and HHS-OCR does not track concerns—referred to as communications—that the agency cannot investigate under Title IX, according to HHS-OCR officials. DOE officials stated that the agency received its first sex discrimination concern in fiscal year 2020 and therefore DOE was not aware individuals were looking to notify the agency of concerns. While these agencies accept concerns, they have received few or no concerns and have not publicly communicated that individuals may send concerns to them. The 2018 NASEM report, agency officials, and stakeholders we interviewed all noted the importance of informal ways for individuals to report concerns outside of formal complaint processes. The NASEM report states that formal reporting procedures can re-victimize targets of harassment, and informal procedures—including the acceptance of anonymous complaints—may let them bring concerns forward without fear of retaliation. A stakeholder we interviewed pointed out the arduous nature of agencies’ formal complaint processes, and multiple stakeholders highlighted the difficulty of meeting the federal standard for a Title IX violation. All five agencies agreed that informal information— such as concerns—is helpful in providing the agency with additional information about the research environment on campus. Of the 140 total concerns of sex discrimination received in fiscal year 2019 by the agencies in our study, all were filed with either NSF or NIH. A comparison of the number of concerns and formal complaints received by the agencies shows that the five agencies as a whole received more than three times as many concerns in 1 year as they did formal complaints in 5 years. Title IX specifies federal agencies’ Title IX oversight responsibilities— including enforcing Title IX compliance at the universities they fund. Under Standards for Internal Control in the Federal Government, management should externally communicate the necessary quality information to achieve the entity’s objectives. For example, management may communicate and receive information through established reporting lines from external parties—in this case, through formal complaints and concerns—which can help the agency achieve its objectives, such as oversight of Title IX. By publicly communicating to individuals that they may notify the agency of a concern of sex discrimination outside of the formal Title IX complaint process, NASA, USDA, DOE, and HHS-OCR could receive additional information necessary for appropriate Title IX oversight. According to NSF officials, concerns not only reveal potential issues with the climate at an awardee university, they also aid in Title IX oversight by alerting the agency to possible Title IX violations a university may need to notify the agency of under the grant terms and conditions. In addition to reviewing concerns of sex discrimination from individuals at university grantees, NSF receives notifications directly from university grantees. In 2018, NSF modified its grant terms and conditions to require university grantees to notify the agency if there is a finding of sexual harassment against a principal investigator (PI) or co-PI on an NSF-funded grant, or if administrative action was taken against a PI or co-PI due to an allegation or complaint of sexual harassment. In 2019, NSF established written procedures to review these notifications from university grantees to determine if the university handled the matter adequately and appropriately, and if further action is needed by NSF. NSF received 13 notifications from university grantees through the new grant terms and conditions in fiscal year 2019, according to agency officials. NASA, similar to NSF, proposed changes to its grant terms and conditions. NASA published its notice of the proposed change in July 2019. However, according to NASA officials, the Office of Science and Technology Policy requested that NASA consult with the National Science and Technology Council’s joint committee’s subcommittees— Safe and Inclusive Research Environments Subcommittee and Coordinating Administrative Requirements for Research Subcommittee— prior to moving forward with finalizing the terms and conditions. NASA consulted with the Office of Science and Technology Policy in December 2019 and received concurrence to move forward with finalizing the change to its terms and conditions, according to NASA officials. On March 10, 2020, NASA published a final notice of its new terms and conditions for grants. Upon implementation, the new terms and conditions requires, among other things, grantees to report to NASA any findings or determinations of sexual harassment, other forms of harassment, or sexual assault regarding a NASA funded PI or co-PI. The reporting requirement will be applied to all new NASA awards and funding amendments to existing awards made on or after the effective date—30 days from the date of the publicized notice. HHS Components Do Not Share Information on Complaints and Concerns HHS differs from the other four agencies in that formal complaints and concerns are handled by two different components, which do not communicate with each other regarding information on sexual harassment, according to officials from both HHS-OCR and HHS-NIH. HHS-OCR—the enforcement authority of the agency—has the authority to conduct Title IX compliance reviews and investigate formal Title IX complaints. However, as previously mentioned, HHS-NIH—the grant- making component—has independently developed its own avenue for receiving concerns of sex discrimination, including sexual harassment. The Standards for Internal Control in the Federal Government state that management should internally communicate the necessary quality information to achieve its objectives. This communication includes conveying information down and across reporting lines to allow staff to perform key roles in achieving objectives and addressing risks. There are no formal procedures within HHS for communicating information across the agency components regarding Title IX complaints, concerns of sex discrimination including sexual harassment, or Title IX compliance, according to officials from both HHS-NIH and the HHS-OCR. An official from HHS-OCR stated that the department already shares broad information about findings of Title IX violations and completed Title IX compliance reviews via a listserv to HHS employees and stakeholders who subscribe, but HHS-NIH officials stated they were not aware of this information. HHS-NIH officials also stated they do not share information with HHS-OCR regarding concerns of sex discrimination, including sexual harassment, received by HHS-NIH or actions taken in response to these concerns. According to HHS-NIH officials, grantees are expected to provide safe and healthy working conditions—a term and condition of the grant—and therefore if harassment threatens the research environment, this is a potential violation of grant terms and conditions and officials stated that HHS-NIH has the authority to handle it. Establishing procedures for communicating grantee sexual harassment findings could improve efforts by both HHS-NIH and HHS-OCR to prevent sexual harassment at universities. For example, HHS-OCR could use HHS-NIH data to aid in selecting grantees for Title IX compliance reviews. Additionally, HHS-NIH could use HHS-OCR’s compliance review findings to inform oversight of NIH grants—including modifying university grantees’ grant terms and conditions if there were findings of non- compliance. Officials from HHS-OCR agreed that information on concerns of sex discrimination, including sexual harassment, from HHS-NIH would be helpful. HHS-NIH officials also agreed that information sharing may be useful for cross agency awareness, but HHS-NIH officials asserted that a formal agreement would be necessary to ensure privacy when sharing information. HHS-NIH officials did not provide any further details on what should be included in a formal agreement for sharing information on sex discrimination concerns, including sexual harassment. Agencies Have Established and Communicated New Grantee Sexual Harassment Prevention Policies but Lack an Overall Plan All five agencies have taken additional steps beyond Title IX compliance requirements to address sexual harassment by university grantees. As we reported in June 2019, all five agencies have developed and communicated grantee sexual harassment prevention policies, with some providing more detailed guidance than others. All five agencies have also established grantee sexual harassment prevention efforts beyond those required by Title IX, to varying degrees. For example, as we noted above, HHS-NIH launched a website to receive concerns of sex discrimination including sexual harassment, and NSF and NASA have modified grant terms and conditions that require universities to report findings of sexual harassment. All of the agencies we reviewed established and communicated their sexual harassment prevention efforts to grantees within the last 3 fiscal years, and most of them have continued to update and communicate them since we last reported on their efforts in June 2019 (see sidebar for an agency example and appendix III for more information on agencies’ efforts). Agencies have taken steps to create goals for and evaluate some of their individual grantee sexual harassment prevention efforts. However, four of the five agencies have not created goals for all prevention efforts. In addition, none of the five agencies have a plan designed to assess progress toward achieving those goals, including methods to regularly monitor and evaluate their various grantee sexual harassment prevention efforts together—both those that are required by Title IX and those that go beyond these requirements. USDA created a poster for grantees that describes how the agency prohibits discrimination in all forms, including discrimination on the basis of sex. The poster also provides information on how to file a discrimination complaint with USDA by phone, mail, fax or email. USDA requires all grantees to prominently display the poster in all offices where there is a USDA presence and where it may be read by customers. USDA also requires that the poster be a specific size. NSF and USDA-NIFA do not have goals for all of their grantee sexual harassment prevention efforts, according to officials. In addition, NASA and DOE have or are in the process of establishing goals for some prevention efforts related to Title IX requirements, while HHS-NIH has created goals for all of their grantee sexual harassment prevention efforts. NASA and DOE have goals or plan to establish goals for sexual harassment prevention efforts required by Title IX—such as compliance reviews—but they lack goals that include all other sexual harassment prevention efforts for university grantees. For example, according to NASA officials, NASA’s strategic plan has goals for equal opportunity and diversity and inclusion for the NASA workforce and grantees, and it includes a goal for the agency to promote equal opportunity for grantees and to encourage them to use best practices identified by NASA. To measure progress toward this goal, NASA officials told us that the agency plans to establish a timeline to track the percentage of Title IX compliance activities completed by grantees. However, NASA has not established goals for its other grantee sexual harassment prevention initiatives. In addition, DOE officials told us that they are in the process of establishing a goal for the number of Title IX compliance reviews they conduct each year, but DOE does not have goals or a plan for evaluating other DOE grantee sexual harassment prevention initiatives. In contrast, HHS-NIH’s Working Group of the Advisory Committee to the NIH Director has created goals for HHS-NIH’s various grantee sexual harassment prevention efforts and steps to achieve them. These goals include assessing the current state of sexual harassment allegation investigation, reporting, remediation, and disciplinary procedures at NIH- funded organizations; advising grantees on oversight, accountability, and reporting measures that will encourage a reduction in, and prevention of sexual harassment; and developing strategies for encouraging research on anti-harassment policies, procedures and training, and measures and evaluations of their effectiveness. HHS-NIH developed recommendations for the steps needed to achieve these goals, including immediate, actionable efforts and longer-term efforts to change the culture within NIH and at universities to end sexual harassment. HHS-NIH officials published a final report and recommendations in December 2019. The report recommended that HHS-NIH establish a hotline and a web- based form for reporting sexual harassment and inappropriate behavior by any principal investigator or key personnel funded by HHS-NIH, and that HHS-NIH also conduct an analysis of the prevalence and antecedents of sexual and gender harassment in order to develop interventions that address goal-specified gaps, among others. Agencies Do Not Have an Overall Plan to Regularly Monitor and Evaluate All Grantee Sexual Harassment Prevention Efforts In addition to most of the agencies not having goals for all of their grantee sexual harassment prevention efforts, none of the five agencies have a plan to measure progress toward achieving those goals, including methods to regularly monitor and evaluate them all. Some of the agencies have taken steps toward conducting evaluations of some—but not all—of their grantee sexual harassment prevention efforts: Evaluations of policies and procedures. Three agencies—NASA, NSF, and DOE—have evaluated or are beginning to evaluate some of their sexual harassment policies and procedures for university grantees. NASA officials said they conduct evaluations every five years for all of their agency’s civil rights compliance and complaints procedures, including their Title IX compliance review procedures. NSF is also developing an evaluation plan for its new sexual harassment reporting requirements and how they have affected grantees. NSF officials said that they have an evaluation team in place, which has outlined an approach for evaluating the new grant terms and conditions and has begun gathering information from universities. In addition, DOE officials told us that they are currently reviewing other agencies’ policies and using them as a benchmark as they draft their own grantee policies. However, agencies have not periodically evaluated all of their own sexual harassment policies and procedures related to university grantees. Agencies provided examples of evaluations of grantee or employee prevention policies, rather than an evaluation of their own policies created for university grantees. It is unclear why agencies have not yet established methods to evaluate all of their sexual harassment prevention efforts for university grantees, and we recognize the challenge in doing so. Yet agencies have found ways to evaluate the policies of other entities. For example, officials from HHS-OCR told us that they use information from past compliance reviews to improve their compliance review and resolution requirements for grantees. However, compliance reviews are an evaluation of the university grantee’s sexual harassment prevention policies and procedures, not HHS’s. In addition, when asked whether HHS-NIH evaluates its grantee sexual harassment policies, HHS-NIH officials did not give any examples of evaluations of their own policies created for university grantees. Instead, they gave an example of a climate survey they administered to their employees for work-life climate and harassment. Title IX officials from two universities and one university system all said none of the five agencies had asked for their opinions on how effective the agencies’ sexual harassment prevention policies for grantees are. Nor had they requested suggestions for improvement, even though at least one of the five agencies had been in direct contact with two of these officials for a recent compliance review. Evaluations of communication mechanisms. None of the five agencies periodically evaluate the mechanisms for how they communicate their sexual harassment prevention policies and procedures to individuals at universities receiving federal grants. Instead, agencies rely on general efforts to evaluate their website or are developing plans for such an evaluation. As a result, the agencies do not know the extent to which their various communication mechanisms are working and whether students, researchers, faculty, and university officials are getting the information they need from these mechanisms. For example, as previously stated, all five agencies use their website as the main mechanism to communicate information about Title IX complaint procedures to individuals at universities receiving federal grants. While agencies have taken steps to add more information to their websites for individuals at universities receiving federal grants, we found some of the agencies’ websites difficult to navigate. Even when key content existed, it was sometimes spread across multiple sections of the website or buried in supplemental materials, or in one case, associated with an incorrect destination page. For example, NSF officials stated that they prefer formal Title IX complaints be filed via their online complaint portal. However, this tool is not linked to the tab of the website discussing how to file complaints. Additionally, HHS-OCR’s newly-developed Title IX and sex-based harassment websites are not referenced or linked to the information on laws and regulations enforced by HHS-OCR or the complaints page. Rather, from the HHS-OCR home page, one would have to know to click on the “Special Topics” link to find links to the two new websites or find the link to the “Sex-based Harassment” page embedded within the new Title IX website. All of the agencies acknowledged issues with their websites. For example, NSF officials acknowledged that their agency’s website may not be user-friendly to individuals at universities—such as students—and is in the process of revising the website to increase ease of use. None of the five agencies have periodically evaluated this key communication tool at this time. NASA officials said that they evaluate their website for grantees, but these evaluations are not systematic and have not specifically focused on their sexual harassment prevention efforts for grantees. HHS-NIH officials said that they receive some feedback on the agency’s webpages, and the agency keeps track of website user satisfaction to improve their quality in general, but this effort is not specific to evaluating how HHS-NIH communicates information on sexual harassment prevention to grantees. NSF officials said they are planning to conduct an evaluation on the effectiveness of their communication efforts with their grantee community and will include actions that result from that evaluation in NSF’s corrective action plan. As previously mentioned, HHS-OCR and DOE recently took action to improve website clarity on who can file a Title IX complaint, in part due to issues raised during the course of our study. Evaluating the effectiveness of their communication mechanisms is important, as agencies may not be clearly communicating their sexual harassment prevention policies and procedures to their intended audiences. Nor can they be sure these policies and procedures are reaching the right university officials. For example, Title IX officials from two universities and one university system said that they had not received any information from the five agencies on their sexual harassment prevention policies for grantees. Two Title IX officials stated that, even if this information is already provided to the university departments or offices conducting scientific research, it should also be given to the university’s Title IX office, with one official noting they are the part of the university responsible for overseeing compliance with sexual harassment policies and procedures under Title IX. Title IX university officials also told us that the federal agencies providing their grants had never provided them with information on agencies’ policies and procedures for how individuals at their institutions could file sexual harassment complaints. One Title IX university official described how they would not know how to tell someone to proceed if they wanted to file a complaint with the agency funding their research project. Officials from all five agencies acknowledged the value of evaluating their grantee sexual harassment prevention efforts and noted that they may be able to conduct such evaluations in the future. In addition, four of the agencies have a general goal to prevent sexual harassment by their university grantees, and all have recognized the need to move beyond their current grantee sexual harassment policies and procedures. As we reported in June 2019, their completed or planned actions include modifying current department- level or agency-wide policy statements to include more specific definitions and examples of sexual harassment and strengthening requirements for their university grantees to report on findings of sexual harassment. The 2018 NASEM report also noted that, while it is clear that the agencies are concerned about sexual harassment in STEM, it is not yet apparent whether and how actions such as their new policy statements will translate into meaningful action. Standards for Internal Control in the Federal Government state that management should define objectives clearly to enable the identification of risks and define risk tolerances; for example, in defining objectives, management may clearly define what is to be achieved, who is to achieve it, how it will be achieved, and the time frames for achievement. Standards for Internal Control in the Federal Government also state that federal agencies should establish and operate monitoring activities to evaluate results, select and periodically evaluate methods of communication, and define objectives in specific and measurable terms. We recognize that the agencies’ civil rights or diversity offices are responsible for enforcing and overseeing a number of other civil rights laws, Executive Orders, regulations, and policy directives for grantees and their own employees, and that sexual harassment prevention efforts for grantees are just some of their many discrimination prevention initiatives. We also recognize that most agencies’ grantee sexual harassment prevention efforts are new, and in some cases, still under development. However, establishing goals and an overall plan to assess progress toward achieving those goals—including methods to regularly monitor and evaluate sexual harassment prevention policies and communication mechanisms— would better position the agencies to effectively coordinate and integrate such activities. It would also help them holistically evaluate all their efforts—both those that are required by Title IX and those that go beyond these requirements—to help grantees prevent sexual harassment at universities, determine whether their policies and procedures are reaching the populations they intend to receive the information, and allow them to more effectively target outreach if they find that there are deficiencies. DOJ’s Quarterly Title IX STEM Discussion Group Aids Collaboration but Has Not Fully Implemented Two Key Leading Practices DOJ’s Quarterly Title IX STEM discussion group provides a forum for the five agencies to collaborate and share information on Title IX compliance. While DOJ has implemented four of the six relevant leading practices on federal interagency collaboration, it has not fully implemented two key practices—agreeing on agency roles and responsibilities and developing mechanisms for monitoring, evaluating, and reporting collaborative efforts. Outside of the discussion group, the five agencies have taken collaborative steps to address the culture and climate for women in STEM. DOJ’s Quarterly Title IX STEM Discussion Group Facilitates Information Sharing across Agencies, but DOJ Has Not Fully Implemented Two Key Leading Practices on Collaboration DOJ’s Quarterly Title IX STEM discussion group facilitates collaboration and shares best practices on Title IX enforcement across the five agencies. Collaboration can be broadly defined as any joint activity that is intended to produce more public value than could be produced when organizations act separately. Since February 2016, after reconstituting the Quarterly Title IX STEM discussion group, DOJ has held quarterly group meetings to share information on Title IX enforcement. According to DOJ officials, the agencies discuss several topics, including: Strategies for conducting Title IX compliance reviews, including joint Strategies for investigating Title IX complaints General discussion of Title IX complaints, including sexual Title IX court cases and case history Officials at the five agencies agreed the group is useful to coordinate and share information on Title IX, for example, by avoiding duplication in compliance reviews and complaint investigations. Indeed, when multiple agencies received the same Title IX complaint, the three agencies collaborated with DOJ to determine which one would handle the complaint, according to officials. DOJ also provides technical assistance and training on Title IX enforcement, according to officials. Some agency officials identified steps that could potentially improve collaboration within the group, including: Clarifying and documenting the group’s purpose, scope, and roles and responsibilities to ease transition of new agency staff and leadership Incorporating more specific topics related to sexual harassment in meeting agendas Involving all federal agencies that fund STEM research at universities. Although agencies are not required to follow leading practices for interagency collaboration, doing so can enhance and sustain such collaboration, thereby improving performance and results. DOJ officials told us the agency has adopted leading practices for interagency collaboration as part of the group. Based on information from DOJ, we found the agency’s actions were consistent with four out of six of the relevant leading practices we have identified for collaborating across agencies. We also found that DOJ could take additional steps to fully adopt the remaining two leading practices, as shown in table 4. Without implementing the two interagency collaboration leading practices, DOJ is missing an opportunity to enhance and sustain collaboration among the five agencies we reviewed as they continue to address the problem of sexual harassment at universities. Agencies Have Taken Steps to Address the Culture and Climate for Women in STEM All five agencies reported taking collaborative steps with universities and federal agencies to address the culture and climate for women in STEM. For example, in 2019, HHS-NIH established a working group with university experts to collaboratively assess the current state of procedures for sexual harassment allegation investigation, reporting, remediation, and discipline at NIH-funded organizations. In December 2019, the working group made several recommendations. For example, it recommended that NIH require each principal investigator and key personnel on an NIH grant attest that they have not been found to have violated their institution’s code of professional conduct, including having a finding of sexual harassment, for a determined period of time. The working group also recommended that NIH create a parallel process to treat professional misconduct, including sexual harassment, as seriously as research misconduct. In addition, DOE, NASA, NSF, and USDA collaborated with universities at conferences and meetings. According to DOE officials, the agency attends the annual conference of university Title IX administrators, where participants discuss issues related to Title IX, sexual harassment, and sexual assault. In 2016, NASA held a conference to help universities address sexual harassment and share best practices to increase participation of underrepresented populations in STEM education (see sidebar). According to NASA officials, attendees included university presidents, deans, and provosts, as well as NASA leaders. The agency is planning another conference in 2020, according to NASA officials. NSF presented information on Title IX and their policies and procedures at numerous conferences and meetings in 2018 and 2019, according to officials. USDA-NIFA served on the planning committee and participated in a conference with public land grant universities to discuss diversity and inclusion in 2018. According to officials, USDA is considering participation in future events. Such efforts to go beyond compliance reviews and to address the larger culture and climate of STEM research are consistent with the 2018 NASEM report, which states that “adherence to legal requirements is necessary but not sufficient to drive the change needed to address sexual harassment.” increase participation in NASA business and grant opportunities, and education programs. address important issues related to America’s research environment. The joint committee also established the Safe and Inclusive Research Environments Subcommittee, in which DOE, HHS (including NIH), NASA, and NSF participate, along with other federal agencies and offices. USDA-NIFA is a member of the joint committee but does not participate in the Safe and Inclusive Research Environments Subcommittee; instead, USDA’s Agricultural Research Service participates in the subcommittee. The goals and planned actions of the joint committee and subcommittee have not been determined yet, according to Office of Science and Technology Policy officials. Office of Science and Technology Policy officials told us in December 2019 that its work plan is complete, but there are no plans to release it publicly since it is a deliberative document. The subcommittee is broadly focused on preventing harassment in research environments. Conclusions Sexual harassment in higher education is degrading and illegal. In 2017 alone, the media covered over 97 allegations of sexual harassment at institutions of higher education with some of the most high-profile cases occurring in the fields of science, engineering and medicine, according to the National Academies of Sciences, Engineering, and Medicine. Federal agencies, in connection with the billions of dollars in research funding they provide to universities and other institutions each year, are required to enforce Title IX—prohibiting sex discrimination, including sexual harassment—at these universities. As part of their enforcement responsibilities, all five agencies have conducted the required Title IX compliance reviews, but three agencies—DOE, HHS-OCR, and USDA— are missing an opportunity to share promising practices from their Title IX compliance reviews with the broader grantee community. Given that these agencies conduct compliance reviews at only a handful of the hundreds of grantees they fund in any given year, the vast majority of grantees receive little to no information on Title IX compliance reviews from these agencies. Another tool federal agencies can use to address sexual harassment is the prompt processing and disposition of Title IX complaints from students and employees. Although all five selected agencies received Title IX complaints, DOE and USDA have not finalized and published complaint procedures, as required by DOJ’s regulations. Furthermore, USDA does not provide clear information about the complaint process on its website—its primary means of communicating information to individuals and grantees. As a result, the agency may be missing the opportunity to better serve individuals seeking relief from sexual harassment at universities. Federal agencies can also review information from individuals seeking to notify the agency of a concern related to sex discrimination—including sexual harassment—in an informal manner outside of the Title IX complaint process. However, only HHS-NIH and NSF communicate the option to submit concerns, and only HHS-NIH has a written process for reviewing such concerns. In a single year, these concerns outnumbered formal Title IX complaints received by all the agencies over 5 years. The 2018 NASEM report, agency officials, and stakeholders we interviewed noted the importance of informal ways for individuals to report concerns outside of formal complaint processes, which can protect an individual from retaliation, alert agencies to possible Title IX violations, and help agencies select sites for Title IX compliance reviews. Two HHS components—NIH and OCR—do not share sexual harassment complaint information with each other. This poses the risk that HHS-NIH will be unaware of situations in which HHS-OCR finds non-compliance with Title IX and may approve a STEM research grant for that university. It also raises the possibility that NIH will receive concerns about a university that may warrant a Title IX compliance review, but the Office for Civil Rights may be unaware of these concerns. Establishing clear goals and an overall plan can help agencies assess progress and manage change, including, in this case, the creation of new sexual harassment prevention efforts for grantees. Although all five agencies have established a variety of prevention efforts, they have done so without a plan, and without methods to evaluate their policies and how they communicate them. As a result, agencies do not have a way to measure progress toward preventing sexual harassment at their university grantees, including how or whether these efforts are helping university grantees and individuals who have been subject to harassment. Finally, interagency coordination can help improve the results of agency activities. DOJ has not fully adopted two key interagency collaboration leading practices for its interagency working group. Without doing so, the agency is missing an opportunity to enhance and sustain collaboration among agencies as they continue to address the problem of sexual harassment at universities. Recommendations We are making 17 recommendations, including four to DOE, one to DOJ, four to HHS, two to NASA, one to NSF, and five to USDA. Specifically: The Secretary of the Department of Agriculture should direct the Assistant Secretary for Civil Rights to publicize promising practices for Title IX compliance on its websites for their university grantees. (Recommendation 1) The Secretary of Energy should direct the Director of the Office of Economic Impact and Diversity to publicize promising practices for Title IX compliance on its websites for their university grantees. (Recommendation 2) The Secretary of the Department of Health and Human Services should direct the Director for the Office for Civil Rights to publicize a stand-alone list of promising practices for Title IX compliance on its websites for their university grantees. (Recommendation 3) The Secretary of Energy should direct the Director of the Office of Economic Impact and Diversity to finalize and publish Title IX complaint procedures, consistent with DOJ’s regulations. (Recommendation 4) The Secretary of the Department of Agriculture should direct the Assistant Secretary for Civil Rights to finalize and publish revised Title IX complaint procedures. (Recommendation 5) The Secretary of the Department of Agriculture should direct the Assistant Secretary for Civil Rights to clarify on its website that individuals on USDA-funded grants can file Title IX complaints through the Assistant Secretary for Civil Rights—including clarifying who is considered “customers.” (Recommendation 6) The Secretary of the Department of Health and Human Services should direct the Director for the Office for Civil Rights to assess the feasibility of receiving and reviewing concerns of sex discrimination— including sexual harassment—and communicating to individuals on agency-funded grants the option to notify the agency of these concerns, outside of the Title IX complaint process. (Recommendation 7) The Secretary of Energy should direct the Director of the Office of Economic Impact and Diversity to assess the feasibility of receiving and reviewing concerns of sex discrimination—including sexual harassment—and communicating to individuals on agency-funded grants the option to notify the agency of these concerns, outside of the Title IX complaint process. (Recommendation 8) The Administrator of NASA should assess the feasibility of receiving and reviewing concerns of sex discrimination—including sexual harassment—and communicating to individuals on agency-funded grants the option to notify the agency of these concerns, outside of the Title IX complaint process. (Recommendation 9) The Secretary of the Department of Agriculture should direct the Assistant Secretary for Civil Rights to assess the feasibility of receiving and reviewing concerns of sex discrimination—including sexual harassment—and communicating to individuals on agency- funded grants the option to notify the agency of these concerns, outside of the Title IX complaint process. (Recommendation 10) The Secretary of the Department of Health and Human Services should direct the Director for the Office for Civil Rights and the Director of NIH to develop and implement formal procedures for sharing relevant information about Title IX (compliance reviews, violations, and complaints) and sex discrimination concerns, including sexual harassment. For example, HHS components should internally share information regarding findings of Title IX violations, concerns of sex discrimination, including sexual harassment, and Title IX compliance review reports. (Recommendation 11) The Secretary of Energy should establish goals and an overall plan to assess all of the agency’s sexual harassment prevention efforts for their university grantees, including methods to regularly monitor and evaluate its sexual harassment prevention policies and communication mechanisms (e.g. Title IX or sex discrimination websites). (Recommendation 12) The Secretary of the Department of Health and Human Services should establish goals and an overall plan to assess all of the agency’s sexual harassment prevention efforts for their university grantees, including methods to regularly monitor and evaluate its sexual harassment prevention policies and communication mechanisms (e.g. Title IX or sex discrimination websites). (Recommendation 13) The Administrator of NASA should establish goals and an overall plan to assess all of the agency’s sexual harassment prevention efforts for their university grantees, including methods to regularly monitor and evaluate its sexual harassment prevention policies and communication mechanisms (e.g. Title IX or sex discrimination websites). (Recommendation 14) The Director of NSF should establish goals and an overall plan to assess all of the agency’s sexual harassment prevention efforts for their university grantees, including methods to regularly monitor and evaluate its sexual harassment prevention policies and communication mechanisms (e.g. Title IX or sex discrimination websites). (Recommendation 15) The Secretary of the Department of Agriculture should establish goals and an overall plan to assess all of the agency’s sexual harassment prevention efforts for their university grantees, including methods to regularly monitor and evaluate its sexual harassment prevention policies and communication mechanisms (e.g. Title IX or sex discrimination websites). (Recommendation 16) In consultation with DOE, HHS, NASA, NSF, and USDA, the Assistant Attorney General for the Department of Justice should direct the responsible Civil Rights Division sections to fully adopt two federal interagency leading practices—agree on agency’s roles and responsibilities and develop mechanisms to monitor, evaluate, and report results of collaborative efforts, for its Quarterly Title IX STEM discussion group. (Recommendation 17) Agency Comments and Our Evaluation We provided a draft this report to DOE, DOJ, Education, HHS, NASA, NSF, the Office of Science and Technology Policy, and USDA for review and comment. We received written comments from the Departments of Agriculture, Energy, Health and Human Services, Justice, as well as NASA and NSF that are reprinted in appendixes IV through IX, and summarized below. Education did not have comments on the draft report, but it provided technical comments, which we incorporated as appropriate. The Office of Science and Technology Policy stated that it did not have comments on the draft report. All six of the agencies and departments to which we made recommendations stated that they agreed with the recommendations and most provided technical comments, which we incorporated as appropriate. The agencies’ comments are summarized below: In the Department of Agriculture's written comments, reproduced in appendix IV, the department agreed with all five recommendations. USDA outlined actions for improving the complaint process and communication with university grantees. For example, the department stated that its regulation for processing complaints is currently in the clearance process for publication. In addition, USDA will reach out to other agencies within the quarterly Title IX STEM discussion group to assess best practices for monitoring and evaluating sexual harassment prevention policies and communication mechanisms. In DOE's written comments, reproduced in appendix V, the department agreed with all four recommendations. DOE plans to publicize a promising practice guide on its website, publish complaint procedures, and evaluate the feasibility of receiving and reviewing concerns of sex discrimination, including sexual harassment. The department estimates completion by the end of calendar year 2020. DOE will establish goals for prevention efforts and an overall plan by the end of January 2021 and August 2021, respectively. In HHS's comments, reproduced in appendix VI, the department agreed with all four recommendations. In response to one recommendation, HHS stated that HHS-OCR and HHS-NIH would review the current procedure, and develop and implement, as necessary, formal procedures for sharing relevant information about Title IX and sexual harassment concerns. However, the department also noted that it did not "share GAO's supposition that coordination of Title IX enforcement between HHS-OCR and HHS-NIH raises privacy concerns." Therefore, we removed this portion of our recommendation. As stated in the report, it was HHS-NIH officials who asserted that a formal agreement is needed to ensure privacy when sharing information, in particular sexual harassment concerns. For another recommendation, the department stated that HHS-NIH has established goals and will develop a plan to assess progress towards achieving these goals, and that HHS-OCR will also develop a plan for its Title IX enforcement and outreach efforts. In DOJ's written comments, reproduced in appendix VII, the department generally agreed with our recommendation. DOJ stated that the department is prepared to delineate the agencies' roles and responsibilities within the interagency group (quarterly Title IX STEM discussion group), as recommended. DOJ also plans to develop an enhanced process for evaluating, monitoring, and reporting on the group's collaborations in enforcing Title IX that is achievable within its current resource allocation, or if more resources become available. In NASA's comments, reproduced in appendix VIII, the agency concurred with our two recommendations. NASA stated that the agency plans to assess the feasibility of receiving and reviewing concerns of sex discrimination and harassment and estimates completion by September 20, 2020. Through the interagency process, NASA will also adopt the policies and procedures developed by the National Science and Technology Council, overseen by the Office of Science and Technology Policy. In NSF's written comments, reproduced in appendix IX, the agency agreed with our recommendation for goals and an overall plan to assess sexual harassment prevention efforts. NSF is embarking on an assessment process to improve its policies and practices continually in order to achieve the goal of safe and inclusive research environments. We are sending copies of this report to the appropriate congressional committees; the Secretaries of Agriculture, Education, Energy, and Health and Human Services; the Directors of the National Science Foundation and the Office of Science and Technology Policy; the Administrator of the National Aeronautics and Space Administration; the Attorney General for the Department of Justice; and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6888 or neumannj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix X. Appendix I: General Procedures for Evaluating and Investigating Formal Title IX Complaints at Five Agencies While Department of Justice (DOJ) regulations require federal agencies to establish and publish complaint procedures, according to the DOJ Title IX Legal Manual, agency regulations with respect to procedures for the investigation of complaints of discriminatory practices are typically brief and lack details as to the manner or timetable for such an inquiry. The National Science Foundation (NSF) and the National Aeronautics and Space Administration (NASA) have developed complaint manuals in addition to the agencies’ Title IX regulations. According to NSF and NASA officials, both agencies follow the same general processes as those published in DOJ guidance or the Department of Education’s Investigative Manual for the prompt processing and disposition of complaints. Figure 3 is a visualization of the general Title IX complaint process described in NASA’s and NSF’s complaint manuals. Formal complaints can conclude in one of four ways: 1) dismissed for a variety of reasons—such as untimeliness or lack of information; 2) referred to another agency based on jurisdiction or authority; 3) resolved through a voluntary resolution process; or 4) resolved via an investigation and formal finding—either supporting the allegation or not. While the formal complaint may be alleging a discriminatory act against an individual, agency investigations focus on the university grantee’s compliance or non-compliance with Title IX. According to NASA and NSF officials, if there is a finding of non-compliance with Title IX, the onus is on the university grantee to take actions to come into compliance—which may include disciplinary action against the harasser. Appendix II: Department of Health and Human Services, National Institutes of Health Process for Responding to Concerns In recent years, the National Institutes of Health within the Department of Health and Human Services (HHS-NIH) has publicly addressed the agency’s efforts to prevent sexual harassment in science and elevate the seriousness with which the agency takes this issue through action. While already receiving information of sex discrimination concerns, including sexual harassment, from relationships built with institutions—including universities, in March 2019, HHS-NIH launched an email address to receive concerns about sexual harassment directly from individuals involved in HHS-NIH funded projects at universities. Shortly after, in June 2019, HHS-NIH also created an online portal in response to user feedback requesting a method of anonymous reporting. HHS-NIH developed preliminary internal guidance for staff regarding the process for handling concerns (see figure 4). According to officials, in order to review a concern, HHS-NIH needs basic information about the allegation, including: First and last name of the person who may have committed Institution that employs that person Brief description of the incident HHS-NIH notifies universities of the concern and may request details on the allegation and the university’s response to the allegations, according to officials. As part of this process, HHS-NIH assesses the university grantee’s response to ensure it is taking appropriate actions to ensure a safe research environment—altering the grant terms and conditions if needed to remove or replace key grant personnel. For example, in 2018, HHS-NIH followed up on sexual harassment-related concerns at more than 20 universities. According to a 2019 HHS-NIH Director Statement, this follow-up resulted in the replacement of 14 principal investigators named on NIH grant awards, disciplinary actions taken by awardee universities against 21 principal investigators—including termination of employment—and removal of 14 individuals from peer review. According to HHS-NIH officials, in fiscal year 2019 HHS-NIH received 93 concerns regarding sexual harassment. HHS-NIH does not just review allegations against personnel already funded by HHS-NIH, but also assesses if the allegations are against applicants for HHS-NIH funding. If a principal investigator or co-principal investigator listed on an application for an HHS-NIH grant is named in an allegation, HHS-NIH works with the institution to gather more information about the allegation in the context of HHS-NIH funded research. While the institution is conducting an internal investigation into the allegations, they may request to change the principal investigator or remove a co-principal investigator listed on the application. This may be a temporary or permanent action depending on the circumstances and the institution’s findings. Appendix III: Five Agencies’ Sexual Harassment Prevention Efforts for Grantees This appendix contains a summary of the five agencies’ sexual harassment prevention efforts for university grantees or individuals at universities receiving federal grants as of December 2019. This summary indicates the implementation status for each agency’s efforts, and whether they were complete, in progress or partially implemented, or not reported. These efforts are grouped in three categories: 1) activities required by Title IX, 2) activities beyond those required by Title IX, and 3) evaluation activities (see figures 5, 6, and 7, respectively). Appendix IV: Comments from the Department of Agriculture Appendix V: Comments from the Department of Energy Appendix VI: Comments from the Department of Health and Human Services Appendix VII: Comments from the Department of Justice Appendix VIII: Comments from the National Aeronautics and Space Administration Appendix IX: Comments from the National Science Foundation Appendix X: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments John Neumann, (202) 512-6888 or neumannj@gao.gov In addition to the individual named above, Mark Gaffigan (Managing Director), Melissa Emrey-Arras (Director), Robert Marek (Assistant Director), Michelle St. Pierre (Assistant Director), Nkenge Gibson (Analyst-in-Charge), Nora Adkins, Caitlin Cusati, Cindy Gilbert, Kristy Kennedy, Anika McMillon, Kristen Pinnock, Amanda Postiglione, Janay Sam, and Benjamin Shouse made key contributions to this report.

Why GAO Did This Study Sexual harassment is degrading and illegal. Studies show it has a negative effect on the ability of women to engage in research at the same level as men. Title IX prohibits sexual harassment and other forms of sex discrimination in education programs that receive federal funding, and federal agencies are required to enforce the law at universities they fund. In fiscal year 2018, the most recent year for which data were available during GAO's review, U.S. universities were awarded about $27 billion in federal grants for STEM research. GAO was asked to review federal efforts to help prevent sexual harassment at universities that receive such grants. This report examines, among other things, (1) how selected federal agencies receive, investigate, and resolve Title IX complaints; (2) the extent to which selected agencies have established an overall plan for their sexual harassment prevention efforts for university grantees, including for communicating and evaluating these efforts and (3) the extent to which selected agencies collaborate on efforts to prevent sexual harassment at universities they fund for STEM research. GAO reviewed agencies' relevant regulations and documentation and interviewed agency officials. What GAO Found The five agencies GAO reviewed provided approximately 80 percent of federal science, technology, engineering, and mathematics (STEM) research grants since fiscal year 2015. From fiscal year 2015 through 2019, four of the five agencies received few complaints—including sexual harassment—under Title IX from individuals at universities. Inconsistent with federal regulations implementing Title IX, two of the agencies—the Departments of Energy (DOE) and Agriculture (USDA)—lack finalized procedures for complaints and thus cannot ensure they are consistently handling complaints. Sex-discrimination concerns—including sexual harassment—can also be raised by individuals outside of the Title IX complaint process (see table). However, only two agencies—the National Science Foundation (NSF) and Department of Health and Human Services (HHS)—publicly communicate the option to notify them of concerns. The other three—DOE, the National Aeronautics and Space Administration (NASA), and USDA—received no concerns in fiscal year 2019 and may be missing opportunities to obtain information for Title IX oversight. All five agencies have established grantee sexual harassment prevention efforts beyond those required by Title IX. However, none of the agencies have goals and plans for all of their efforts, and thus they lack clear ways to evaluate how well these efforts are working and to identify any needed improvements. They may also be missing opportunities to coordinate and integrate prevention activities. Additionally, the Department of Justice (DOJ) reconstituted an interagency discussion group on Title IX in 2016, where all five agencies share information about their activities. However, DOJ has not fully adopted two leading practices for collaboration: agreeing on agency roles and responsibilities and developing mechanisms to monitor, evaluate, and report collaborative efforts. Officials at one agency said clarifying agencies' roles and responsibilities would improve the group. Adopting leading practices would help enhance and sustain collaboration. What GAO Recommends GAO is making 17 recommendations to the five agencies funding STEM research and DOJ, including to finalize and publish complaint procedures, establish goals and an overall plan for prevention efforts, and fully adopt two collaboration leading practices. The agencies agreed with GAO's recommendations and identified actions they plan to take to address them.

gao_GAO-19-594T

gao_GAO-19-594T_0

FEMA Has Taken Steps to Strengthen Disaster Resilience and Preparedness, but Additional Steps are Needed to Fully Address Remaining Challenges We have previously reported on various aspects of national preparedness, including examining the extent to which FEMA programs encourage disaster resilience and identifying gaps in federal preparedness capabilities. We have found that when federal, state, and local efforts aligned to focus on improving disaster resilience and preparedness, there was a noticeable reduction in the effects of the disaster. However, our prior and ongoing work also highlight opportunities to improve disaster resilience and preparedness nationwide. Disaster Resilience Hazard mitigation is a key step in building resilience and preparedness against future disasters. In July 2015, we found that states and localities experienced challenges when trying to use federal funds to maximize resilient rebuilding in the wake of a disaster. In particular, they had difficulty navigating multiple federal grant programs and applying federal resources towards their most salient risks because of the fragmented and reactionary nature of the funding. In our 2015 report, we recommended that the Mitigation Framework Leadership group—an interagency body chaired by FEMA—create a National Mitigation Investment Strategy to help federal, state, and local officials plan for and prioritize disaster resilience. As of May 2019, according to FEMA officials, the Mitigation Framework Leadership group is on track to address the recommendation, and they expect the strategy to be published by July 2019. In September 2017, we reported that the methods used to estimate the potential economic effects of climate change in the United States—using linked climate science and economics models—could inform decision makers about significant potential damages in different U.S. sectors or regions, despite the limitations. For example, for 2020 through 2039, one study estimated between $4 billion and $6 billion in annual coastal property damages from sea level rise and more frequent and intense storms. We found that the federal government has not undertaken strategic government-wide planning on the potential economic effects of climate change to identify significant risks and craft appropriate federal responses. As a result, we recommended the Executive Office of the President, among others, should use information on the potential economic effects of climate change to help identify significant climate risks facing the federal government and craft appropriate federal responses, such as establishing a strategy to identify, prioritize, and guide federal investments to enhance resilience against future disasters; however, as of June 2019, officials have not taken action to address this recommendation. In November 2017, we found that FEMA had taken some actions to better promote hazard mitigation as part of its Public Assistance grant program. However, we also reported that more consistent planning for, and more specific performance measures related to, hazard mitigation could help ensure that mitigation is incorporated into recovery efforts. We recommended, among other things, that FEMA (1) standardize planning efforts for hazard mitigation after a disaster and (2) develop performance measures for the Public Assistance grant program to better align with FEMA’s strategic goal for hazard mitigation in the recovery process. FEMA concurred with our recommendations, and as of March 2019, officials have reported taking steps to increase coordination across its Public Assistance, mitigation, and field operations to ensure hazard mitigation efforts are standardized and integrated into the recovery process. Additionally, FEMA officials reported taking actions to begin developing disaster-specific mitigation performance measures. However, FEMA has yet to finalize these actions, such as by proposing performance measures to FEMA senior leadership. As such, we are continuing to monitor FEMA’s efforts to address these recommendations. Disaster Preparedness In March 2011, we reported that FEMA had not completed a comprehensive and measurable national preparedness assessment of capability gaps—for example the amount of resources required to save lives, protect property and the environment, and meet basic human needs after an incident has occurred. Developing such an assessment would help FEMA to identify what capability gaps exist and what level of resources are needed to close such gaps. Accordingly, we suggested that FEMA complete a national preparedness assessment to evaluate capability requirements and gaps at each level of government to enable FEMA to prioritize grant funding. As of December 2018, FEMA had efforts underway to assess urban area, state, territory, and tribal preparedness capabilities to inform the prioritization of grant funding; however, the agency had not yet completed a national preparedness assessment with clear, objective, and quantifiable capability requirements against which to assess preparedness. We are continuing to monitor FEMA’s efforts to complete such an assessment. Furthermore, in March 2015, we reviewed selected states’ approaches to budgeting for disaster costs to help inform congressional consideration of the balance between federal and state roles in funding disaster assistance. Specifically, we reported that none of the 10 states in our review maintained reserves dedicated solely for future disasters, and some state officials reported that they could cover disaster costs without dedicated disaster reserves because they generally relied on the federal government to fund most of the costs associated with disaster response and recovery. In response to the 2017 disasters, we also have ongoing work to review national preparedness capabilities to assist communities in responding to and recovering from disasters. Based on our preliminary observations, some states and localities we interviewed reported that while they are prepared to deal with immediate response issues in the aftermath of a disaster, gaps exist in their capacity to support longer term recovery. One reason for this, according to these state and local officials, is because federal preparedness grant funds are largely dedicated to maintaining response capabilities and sustaining personnel costs for local emergency management officials. While these preparedness grants fund critical elements of the national preparedness system, there are some limitations to using them. Specifically, some state and local officials told us that the preparedness grant activities are generally focused on terrorism issues rather than all-hazards. In addition, they reported that the preparedness grants are generally spent on maintaining response capabilities rather than to enhance their capacity for disaster recovery—such as additional training and exercises. In addition to the state, territory, and urban region assessments that FEMA is conducting, FEMA is currently in the process of developing the first national Threat and Hazard Identification and Risk Assessment. This national assessment may help FEMA and policymakers better understand how to target federal resources in a way that enhances the nation’s capacity to respond and recover from future catastrophic or sequential disasters. We are continuing to evaluate national preparedness efforts and plan to report on FEMA’s Threat and Hazard Identification and Risk Assessment process in January 2020. FEMA’s Response to the 2017 Disasters Highlighted Some Areas of Progress, But also Identified Significant Weaknesses FEMA’s Response to the 2017 Disasters In September 2018, we reported that the response to the 2017 hurricanes and wildfires in Texas, Florida, and California showed progress made since the 2005 federal response to Hurricane Katrina. We also found that FEMA coordinated closely with Texas, Florida, and California emergency management officials and other federal, local, and volunteer emergency partners to implement various emergency preparedness actions prior to the 2017 disasters in each state, and to respond to these disasters. According to FEMA and state officials, these actions helped officials begin addressing a number of challenges they faced such as meeting the demand for a sufficient and adequately-trained disaster workforce and complex issues related to removing debris in a timely manner after the hurricanes and wildfires. In contrast, we also reported in September 2018, that in Puerto Rico and the USVI a variety of challenges—such as the far distance of the territories from the U.S. mainland, limited local preparedness for a major hurricane, and outdated local infrastructure—complicated response efforts to hurricanes Irma and Maria. Many of the challenges we identified are also described in FEMA’s 2017 Hurricane Season FEMA After-Action Report, including: the sequential and overlapping timing of the three hurricanes—with Maria being the last of the three—caused staffing shortages and required FEMA to shift staff to the territories that were already deployed to other disasters; the far distance of both territories from the U.S. mainland complicated efforts to deploy federal resources and personnel quickly; and the incapacitation of local response functions due to widespread devastation and loss of power and communications, and limited preparedness by Puerto Rico and the USVI for a category 5 hurricane resulted in FEMA having to assume response functions that territories would usually perform themselves. We also reported that FEMA’s 2017 Hurricane Season FEMA After-Action Report noted that FEMA could have better leveraged information from preparedness exercises in the Caribbean, including a 2011 exercise after- action report for Puerto Rico which indicated that the territory would require extensive federal support during a large scale disaster in moving commodities from the mainland to the territory and to distribution points throughout. In our September 2018 report, we also found that FEMA’s efforts in Puerto Rico after Hurricane Maria were the largest and longest single response in the agency’s history. According to FEMA, the agency’s response included, among other things, bringing in approximately $1 billion in food and supplies; and distributing food, commodities, and medicine via approximately 1,400 flights, which constituted the longest sustained air operations in U.S. disaster history. FEMA officials explained that the agency essentially served as the first responder in the early response efforts in Puerto Rico, and many of services FEMA provided—such as power restoration, debris removal, and commodity distribution—were typically provided by territorial or local governments. We also reported in September 2018, that in the USVI, recent disaster training and the pre-positioning of supplies due to the anticipated impact of Hurricane Irma facilitated the response efforts for Hurricane Maria, which made landfall less than two weeks later. According to FEMA’s federal coordinating officer, the lead federal official in charge of response for the USVI, the federal government deployed assets, including urban search and rescue teams and medical assistance teams. In addition, due to the sequence of Hurricane Irma hitting the USVI immediately before Hurricane Maria, the Department of Defense (DOD) already had personnel and resources (i.e., ships) deployed to the area, which enabled DOD to respond to Hurricane Maria faster than it otherwise would have. Additional challenges we have reported on regarding response operations have included providing short-term housing and sheltering for disaster survivors. The Department of Homeland Security’s (DHS) 2017 National Preparedness Report states that providing effective and affordable short- term housing for disaster survivors has been a longstanding and continuing challenge. For example, following the California wildfires, local officials faced challenges identifying shelter for displaced survivors, in part due to a housing shortage that existed before the wildfires. Federal, state, and local officials formed housing task forces which facilitated a joint decision-making approach to address these challenges. While this approach has enabled the state to meet its most pressing short-term housing needs, according to FEMA officials, the state faces other challenges in the long term. For example, FEMA officials in the region covering California told us that because of the nature of damage following a wildfire and because of housing shortages in California, some of FEMA’s forms of housing assistance have been less relevant in the wake of the California wildfires than for other disasters. We will continue to evaluate these and other challenges and plan to report in fall 2019. We also have ongoing work to review efforts to provide mass care— which includes sheltering, feeding and providing emergency supplies— following the 2017 hurricanes. Our preliminary observations indicate that during and immediately following the hurricanes, the number of people seeking public shelters outpaced the capacity. In Texas and Florida, emergency managers we spoke with described having unprecedented numbers of residents needing shelters but not always enough staff initially to operate the shelters. In Texas, Puerto Rico, and the USVI, hurricanes Harvey, Irma, and Maria flooded or destroyed many buildings planned for use as shelters, according to emergency management and local government officials in these areas. As a result, some remaining shelters were at maximum capacity. In the USVI, residents of some public housing units that had sustained significant damages sought help at the territory’s Department of Human Services because there was no more space in the shelters, according to local government officials. While they were turned away from the shelters, these families were able to take refuge in the lobby of the Department of Human Services building. We will continue to evaluate these and other challenges and plan to report in summer 2019. FEMA Disaster Contracting In December 2018 and April 2019, we reported that, in response to hurricanes Harvey, Irma, and Maria, as well as the 2017 California wildfires, FEMA and other federal partners relied heavily on advance contracts—which are established before a disaster to provide for life- sustaining goods and services such as food, water and transportation typically needed immediately after a disaster—and post disaster contracts—which can be used for various goods and services, such as debris removal and installation of power transmission equipment. FEMA is required to coordinate with states and localities and encourage them to establish their own advance contracts with vendors. In December 2018, we reported on inconsistencies we found in that coordination and in the information FEMA used to coordinate with states and localities on advance contracts. As a result of this and other challenges identified, we made nine recommendations to FEMA, including that it update its strategy and guidance to clarify the use of advance contracts, improve the timeliness of its acquisition planning activities, revise its methodology for reporting disaster contracting actions to Congress, and provide more consistent guidance and information for contracting officers in coordinating with states and localities to establish advance contracts. FEMA concurred with all of these recommendations, and we are continuing to monitor its efforts to implement each recommendation. Furthermore, in April 2019, we reported on challenges that we found in the federal government’s use of post-disaster contracts. These challenges included a lack of transparency about contract actions, challenges with requirements development, and with interagency coordination. In our report, we found that FEMA had begun taking some steps to address the consistency of post-disaster contract requirements with contracting officers, but that inaccurate or untimely estimates in the contracts we reviewed sometimes resulted in delays meeting the needs of survivors. As a result of our findings in this report, we made 10 recommendations to FEMA and other federal agencies that use these post-disaster contracts related to improving the management of such contracts. FEMA and other agency officials concurred with nine of the recommendations and have reported taking actions to begin implementing them. We will continue to monitor FEMA’s progress in fully addressing these recommendations. FEMA Provides Long Term Disaster Recovery Support, but State and Local Officials Cited Continued Challenges Managing Complex Recovery Assistance Programs FEMA provides multiple forms of disaster recovery assistance after a major disaster has been declared, including Public Assistance and Individual Assistance. Through these grant programs, FEMA obligates billions of dollars to state, tribal, territorial, and local governments, certain nonprofit organizations, and individuals that have suffered injury or damages from major disaster or emergency incidents, such as hurricanes, tornados, or wildfires. In September 2016, we reported that, from fiscal years 2005 through 2014, FEMA obligated almost $46 billion for the Public Assistance program and over $25 billion for the Individual Assistance program. According to FEMA’s May 2019 Disaster Relief Fund report, total projected obligations through fiscal year 2019 for the Public Assistance and Individual Assistance programs for just the 2017 hurricanes—Harvey, Irma, and Maria—are roughly $16 billion and $7 billion, respectively. Given the high cost of these programs, it is imperative that FEMA continue to make progress on the challenges we have identified in our prior and ongoing work regarding its recovery efforts. FEMA Public Assistance Grants for Disaster Recovery FEMA’s Public Assistance program provides grants to state, tribal, territorial, and local governments for debris removal; emergency protective measures; and the repair, replacement, or restoration of disaster-damaged, publicly owned facilities. It is a complex and multistep program administered through a partnership among FEMA, the state, and local officials. Prior to implementing the Public Assistance program, FEMA determines a state, territorial or tribal government’s eligibility for the program using the per capita damage indicator. In our September 2018 report on federal response and recovery efforts for the 2017 hurricanes and wildfires, we reported on FEMA’s implementation of the Public Assistance program, which has recently undergone significant changes as a result of federal legislation and agency initiatives. Specifically, we reported on FEMA’s use of its redesigned delivery model for providing grants under the Public Assistance program, as well as the alternative procedures for administering or receiving such grant funds that FEMA allows states, territories, and local governments to use for their recovery. Our prior and ongoing work highlights both progress and challenges with FEMA’s Public Assistance program, including the agency’s methodology for determining program eligibility, the redesigned delivery model, and the program’s alternative procedures. Criteria for Determining Public Assistance Eligibility In September 2012, we found that FEMA primarily relied on a single criterion, the per capita damage indicator, to determine a jurisdiction’s eligibility for Public Assistance funding. However, because FEMA’s current per capita indicator, set at $1 in 1986, does not reflect the rise in (1) per capita personal income since it was created in 1986 or (2) inflation from 1986 to 1999, the indicator is artificially low. Our analysis of actual and projected obligations for 508 disaster declarations in which Public Assistance was awarded during fiscal years 2004 through 2011 showed that fewer disasters would have met either the personal income-adjusted or the inflation-adjusted Public Assistance per capita indicators for the years in which the disaster was declared. Thus, had the indicator been adjusted annually since 1986 for personal income or inflation, fewer jurisdictions would have met the eligibility criteria that FEMA primarily used to determine whether federal assistance should be provided, which would have likely resulted in fewer disaster declarations and lower federal costs. We recommended, among other things, that FEMA develop and implement a methodology that that more comprehensively assesses a jurisdiction’s capacity to respond to and recover from a disaster without federal assistance, including fiscal capacity and consideration of response and recovery capabilities. DHS concurred with our recommendation and, in January 2016, FEMA was considering establishing a disaster deductible, which would have required a predetermined level of financial or other commitment before FEMA would have provided assistance under the Public Assistance program. In August 2018, FEMA told us that it was no longer pursuing its proposed disaster deductible due to concerns about the complexity of the proposal. FEMA is considering options that leverage similar approaches, but does not have an estimated completion date for implementation. In addition, the DRRA requires FEMA to initiate rulemaking to (1) update the factors considered when evaluating requests for major disaster declarations, including reviewing how FEMA estimates the cost of major disaster assistance, and (2) consider other impacts on the capacity of a jurisdiction to respond to disasters, by October 2020. Until FEMA implements a new methodology, the agency will not have an accurate assessment of a jurisdiction’s capabilities and runs the risk of recommending that the President award Public Assistance to jurisdictions that have the capacity to respond and recover on their own. Redesigned Public Assistance Delivery Model Prior to our September 2018 report, we had previously reported on the Public Assistance program in November 2017. Specifically, we reported that FEMA redesigned the delivery model for providing grants under the Public Assistance program. As part of the redesign effort, FEMA developed a new, web-based case management system to address past challenges, such as difficulties in sharing grant documentation among FEMA, state, and local officials and tracking the status of Public Assistance projects. Both FEMA and state officials involved in testing of the redesigned delivery model stated that the new case management system’s capabilities could lead to greater transparency and efficiencies in the program. However, we found that FEMA had not fully addressed two key information technology management controls that are necessary to ensure systems work effectively and meet user needs. We recommended, among other things, that FEMA (1) establish controls for tracking the development of system requirements, and (2) establish system testing criteria, roles and responsibilities, and the sequence and schedule for integration of other relevant systems. FEMA concurred with these recommendations and has fully implemented the first recommendation. Regarding the second recommendation, FEMA has not yet finalized its decision on whether to integrate its new case management system with its current grants management system. As of March 2019, we are awaiting a final decision from officials to determine whether their actions fully address our recommendation. FEMA’s original intention was to implement the redesigned delivery model for all future disasters beginning in January 2018. However, in September 2017, FEMA expedited full implementation of the redesigned model shortly after Hurricane Harvey made landfall. In September 2018, we reported that local officials continued to experience challenges with using the new Public Assistance web-based, case management system following the 2017 disasters, such as not having sufficient guidance on how to use the new system and delays with FEMA’s processing of their projects. Public Assistance Alternative Procedures in the United States Virgin Islands and Puerto Rico In February 2019, we also reported that FEMA and the USVI were transitioning from using the standard Public Assistance program to using Public Assistance alternative procedures. FEMA and USVI officials stated that the alternative procedures will give the USVI more flexibility in determining when and how to fund projects and allow the territory to use any excess funds for cost-effective hazard mitigation measures, among other uses. Further, when using the alternative procedures, the Bipartisan Budget Act of 2018 allows FEMA, the USVI and Puerto Rico to repair and rebuild critical services infrastructure—such as medical and education facilities—so it meets industry standards without regard to pre-disaster condition (see Figure 1). Regarding the implementation of the Public Assistance program in Puerto Rico, in March 2019, we reported that Puerto Rico established a central recovery office to oversee federal recovery funds and was developing an internal controls plan to help ensure better management and accountability of the funds. In the interim, FEMA instituted a manual process for reviewing each reimbursement request before providing Public Assistance funds to mitigate risk and help ensure financial accountability. We also reported that officials we interviewed from FEMA, Puerto Rico’s central recovery office, and municipalities said they experienced initial challenges with the recovery process, including concerns about lack of experience and knowledge of the alternative procedures; concerns about missing, incomplete, or conflicting guidance on the alternative procedures; and concerns that municipalities had not been fully reimbursed for work already completed after the hurricanes, causing financial hardships in some municipalities. FEMA officials stated that the agency is taking actions to address reported recovery challenges, such as additional training for new FEMA employees and drafting supplemental guidance for the alternative procedures process. We continue to monitor FEMA’s efforts in our ongoing work. As part of our ongoing work, we are continuing to examine hurricane recovery efforts in the USVI and Puerto Rico. Our preliminary observations indicate that the USVI plans to take a cautious approach in pursuing permanent work projects using the Public Assistance alternative procedures program, which requires the use of fixed-cost estimates. Specifically, USVI officials we interviewed told us that developing such fixed-cost estimates that accurately incorporate the future impact of inflation and increases in materials and labor costs for certain projects was difficult. Further, these officials stated that since the territory is financially responsible for any costs that exceed these fixed-cost estimates, the USVI plans to pursue projects that do not include high levels of complexity or uncertainty to reduce the risk of cost overruns. From our ongoing work on Puerto Rico’s recovery efforts, we have learned that, in March 2019, Puerto Rico’s central recovery office released the Disaster Recovery Federal Funds Management Guide, including an internal controls plan for the operation of the recovery office. On April 1, 2019, FEMA removed the manual reimbursement process and began a transition to allow the central recovery office to take responsibility for review and reimbursement approval of federal recovery funds. We will review this transition process as a part of our ongoing work. Our preliminary observations also indicate that some of the challenges we reported in our March 2019 report continue. For example, officials from Puerto Rico’s central government agencies told us they did not feel they had sufficient guidance on the FEMA Public Assistance program and where they did, written and verbal FEMA guidance was inconsistent or conflicting. For example, officials from one agency expressed their desire for more FEMA guidance communicated in writing as it frequently happened that different FEMA officials would interpret existing guidance differently. Similarly, officials from two agencies described situations where they had initially been directed to follow one interpretation of a policy, only to be directed to follow a different, conflicting interpretation in the subsequent months. Puerto Rico agency officials also stated that the lack of sufficient instruction led to a “back and forth” with FEMA for clarifications, which led to delays in the phases of project development. FEMA officials in Puerto Rico stated that the agency has developed specific guidance for disaster recovery in Puerto Rico and that there are various ways, such as in-person meetings, where officials from Puerto Rico can obtain clarification. We are continuing to examine this issue as part of our ongoing review of Puerto Rico’s recovery. In addition, our preliminary observations from our ongoing work for both the USVI and Puerto Rico indicate that FEMA, USVI and Puerto Rico officials have reported challenges with the implementation of the flexibilities authorized by section 20601 of the Bipartisan Budget Act. This section of the Act allows for the provision of assistance under the Public Assistance alternative procedures to restore disaster-damaged facilities or systems that provide critical services to an industry standard without regard to pre-disaster condition. Officials from Puerto Rico’s central government stated that they disagreed with FEMA’s interpretation of the types of damages covered by section 20601 of the Bipartisan Budget Act of 2018. In response, FEMA officials in Puerto Rico stated they held several briefings with Puerto Rico’s central recovery office to explain FEMA’s interpretation of the section. Further, FEMA officials in the USVI told us that initially, they had difficulty obtaining clarification from FEMA headquarters regarding how to implement key components of section 20601 of the Act. As of May 2019, FEMA officials in the USVI stated that they continue to move forward with developing alternative procedures projects. USVI officials also told us that FEMA had been responsive and helpful in identifying its options for using the new authorities the Act provides. We will continue to evaluate these identified challenges and any efforts to address them, as well as other aspects of recovery efforts in the USVI and Puerto Rico, and plan to report our findings in late 2019 and early 2020, respectively. FEMA Individual Assistance The Individual Assistance program provides financial and direct assistance to disaster victims for expenses and needs that cannot be met through other means, such as insurance. In May 2019, we reported on FEMA’s effort to provide disaster assistance under the Individual Assistance program to older adults and people with disabilities following the 2017 hurricanes. We found that aspects of the application process for FEMA assistance were challenging for older individuals and those with disabilities. Further, according to stakeholders and FEMA officials, disability-related questions in the Individual Assistance registration materials were confusing and easily misinterpreted. While FEMA had made some efforts to help registrants interpret the questions, we recommended, among other things, that FEMA (1) implement new registration-intake questions that improve FEMA’s ability to identify and address survivors’ disability-related needs, and (2) improve communication of registrants’ disability-related information across FEMA programs. DHS concurred with the first recommendation and described steps FEMA plans to take, or is in the process of taking, to address it. However, DHS did not concur with the second recommendation, noting that it lacks specific funding to augment its legacy data systems. FEMA officials stated that they began a long-term data management improvement initiative in April 2017, which they expect will ease efforts to share and flag specific disability-related data. While we acknowledge FEMA’s concerns about changing legacy systems when it has existing plans to replace those systems, we continue to believe there are other cost-effective ways that are likely to improve communication of registrants’ disability-related information prior to implementing the system upgrades. For example, FEMA could revise its guidance to remind program officials to review the survivor case file notes to identify whether there is a record of any disability-related needs. We also have work underway to assess FEMA’s Individuals and Households Program, a component program of Individual Assistance. Through this program, as of April 2019, FEMA had awarded roughly $4.7 billion in assistance to almost 1.8 million individuals and households for federally-declared disasters occurring in 2017 and 2018. Specifically, we are analyzing Individuals and Households Program expenditures and registration data for recent years; reviewing FEMA’s processes, policies, and procedures for making eligibility and award determinations; and examining survivors’ reported experiences with this program, including any challenges, for major disaster declarations occurring in recent years. We plan to report our findings in early 2020. Longstanding Workforce Management and Information Technology Challenges Exacerbate Key Issues with Response and Recovery Operations FEMA Workforce Management Challenges FEMA’s experiences during the 2017 disasters highlight the importance of continuing to make progress on addressing the long-standing workforce management challenges we have previously reported on and continue to observe in our ongoing work. In September 2018, we reported that the 2017 disasters—hurricanes Harvey, Irma, and Maria, as well as the California wildfires—resulted in unprecedented FEMA workforce management challenges, including recruiting, maintaining, and deploying a sufficient and adequately-trained FEMA disaster workforce. FEMA’s available workforce was overwhelmed by the response needs caused by the sequential and overlapping timing of the three hurricanes. For example, at the height of FEMA workforce deployments in October 2017, 54 percent of staff were serving in a capacity in which they did not hold the title of “Qualified”—according to FEMA’s qualification system standards—a past challenge we identified. FEMA officials noted that staff shortages, and lack of trained personnel with program expertise led to complications in its response efforts, particularly after Hurricane Maria. In February 2016, we reported on, among other things, FEMA’s efforts to implement, assess, and improve its Incident Management Assistance Team program. We found that while FEMA used some leading practices in managing the program, it lacked a standardized plan to ensure that all national and regional Incident Management Assistance Team members received required training. Further, we found that the program had experienced high attrition since its implementation in fiscal year 2013. We recommended, among other things, that FEMA develop (1) a plan to ensure that Incident Management Assistance Teams receive required training, and (2) a workforce strategy for retaining Incident Management Assistance Team staff. DHS concurred with the recommendations. FEMA fully implemented our first recommendation by developing an Incident Management Assistance Team Training and Readiness Manual and providing a training schedule for fiscal year 2017. In response to the second recommendation, FEMA officials stated in July 2018 that they plan to develop policies that will provide guidance on a new workforce structure, incentives for Incident Management Assistance Team personnel, and pay-for-performance and all other human resource actions. We are continuing to monitor FEMA’s efforts to address this recommendation. In November and December 2017, we reported on staffing challenges in FEMA’s Public Assistance program. In November 2017, we reported on FEMA’s efforts to address past workforce management challenges through its redesigned Public Assistance delivery model. As part of the redesign effort, FEMA created consolidated resource centers to standardize and centralize Public Assistance staff responsible for managing grant applications, and new specialized positions to ensure more consistent guidance to applicants. However, we found that FEMA had not assessed the workforce needed to fully implement the redesigned model, such as the number of staff needed to fill certain new positions, or to achieve staffing goals. Further, in December 2017, we reported on FEMA’s management of its Public Assistance appeals process, including that FEMA increased staffing levels for the appeals process from 2015 to 2017. However, we found that FEMA continued to face a number of workforce challenges, such as staff vacancies, turnover, and delays in training, which contributed to processing delays. Based on our findings from our November and December 2017 reports, we recommended, among other things, that FEMA (1) complete workforce staffing assessments that identify the appropriate number of staff needed to implement the redesigned Public Assistance delivery model, and (2) document steps for hiring, training, and retaining key appeals staff, and address staff transitions resulting from deployments to disasters. FEMA concurred with our recommendations to address workforce management challenges in the Public Assistance program and have reported taking some actions in response. For example, to address the first recommendation, FEMA officials have developed preliminary models and estimates of staffing needs across various programs, including Public Assistance, and plan to reevaluate the appropriate number of staff needed and present recommendations to senior leadership by the end of June 2019. To address the second recommendation, FEMA has collected information on the amount of time regional appeals analysts spend on appeals, and the inventory and timeliness of different types of appeals. FEMA officials stated in September 2018 that they plan to assess this information to prepare a detailed regional workforce plan. As of June 2019, we are evaluating plans and documents provided by FEMA to determine whether they have fully addressed this recommendation. In our March 2019 report on the status of recovery efforts in Puerto Rico, we also reported Puerto Rico officials’ concerns about FEMA staff turnover and lack of knowledge among FEMA staff about how the Public Assistance alternative procedures are to be applied in Puerto Rico. As part of our ongoing work, we are continuing to examine recovery efforts in Puerto Rico. Our preliminary observations indicate that the concerns we reported on in our March 2019 report continue. For example, Puerto Rico agency officials said that the lack of continuity in FEMA personnel has been a challenge for communication and project development. Further, officials from all seven Puerto Rico government agencies we interviewed felt that the FEMA staff they interacted with did not have a complete understanding of FEMA processes and policies. We are continuing to evaluate FEMA’s recovery efforts in Puerto Rico and plan to issue our findings in late 2019. In April 2019, we reported on the federal government’s contracting efforts for preparedness, response, and recovery efforts related to the 2017 hurricanes and California wildfires. We found, among other things, that contracting workforce shortages continue to be a challenge for disaster response and recovery. Further, although FEMA’s 2017 after-action report recommended increasing contract support capacities, it did not provide a specific plan to do so. We also found that while FEMA evaluated its contracting workforce needs in a 2014 workforce analysis, it did not specifically consider contracting workforce needs in the regional offices or address Disaster Acquisition Response Team employees. In our April 2019 report, we recommended, among other things, that FEMA assess its workforce needs—including staffing levels, mission needs, and skill gaps—for contracting staff, to include regional offices and Disaster Acquisition Response Teams, and develop a plan, including timelines, to address any gaps. FEMA concurred with this recommendation and estimates that it will implement it in September 2019. In our May 2019 report on FEMA disaster assistance to older adults and people with disabilities following the 2017 hurricanes, we found that FEMA began implementing a new approach to assist individuals with disabilities in June 2018, which shifted the responsibility for directly assisting individuals with disabilities from Disability Integration Advisors— which are staff FEMA deploys specifically to identify and recommend actions needed to support survivors with disabilities—to all FEMA staff. To implement this new approach, FEMA planned to train all of the agency’s deployable staff and staff in programmatic offices on disability issues during response and recovery deployments. According to FEMA, a number of Disability Integration Advisors would also deploy to advise FEMA leadership in the field during disaster response and recovery. We found that while FEMA has taken some initial steps to provide training on the changes, it has not established a plan for delivering comprehensive disability-related training to all staff who will be directly interacting with individuals with disabilities. We recommended, among other things, that FEMA develop a plan for delivering training to FEMA staff that promotes competency in disability awareness and includes milestones and performance measures, and outlines how performance will be monitored. DHS concurred with this recommendation; however, officials stated that FEMA is developing a plan to include a disability integration competency in the guidance provided for all deployable staff, rather than through training. We will monitor FEMA’s efforts to develop this plan and fully address our recommendation. In addition to our prior work on FEMA’s workforce management challenges related to specific programs and functions, we are continuing to evaluate FEMA’s workforce capacity and training efforts during the 2017 and 2018 disaster seasons. Our preliminary observations indicate that there were challenges in FEMA’s ability to deploy staff with the right kinds of skills and training at the right time to best meet the needs of various disaster events. For example, according to FEMA field leadership we interviewed, for some of the functions FEMA performs in the field, FEMA had too few staff with the right technical skills to perform their missions—such as inspections of damaged properties—efficiently and effectively. For other functions, these managers also reported that they had too many staff in the early stages of the disaster, which created challenges with assigning duties and providing on-the-job training. For example, some managers reported that they were allocated more staff than needed in the initial phases of the disaster, but many lacked experience and were without someone to provide direction and mentoring to ensure they used their time efficiently and gained competence more quickly. Groups of FEMA field managers we interviewed told us that difficulties deploying the right mix of staff with the right skills led to challenges such as making purchases to support FEMA operations, problems with properly registering applicants for FEMA programs, or poor communication with nonfederal partners. Nonetheless, FEMA staff have noted that, despite any suboptimal circumstances during disaster response, they aimed to and have been able to find a way to deliver the mission. As part of this ongoing work, FEMA field leadership and managers also reported challenges using agency systems to ensure the availability of the right staff with the right skills in the right place and time. FEMA uses a system called the Deployment Tracking System to, among other things, help identify staff available to be deployed and activate and track deployments. To help gauge the experience level and training needs of its staff, the agency established the FEMA Qualification System (FQS), which is a set of processes and criteria to monitor staff experience in competently performing tasks and completing training that correspond to their job titles. According to the FQS guidance, staff who have been able to demonstrate proficient performance of all the relevant tasks and complete required training receive the designation “qualified,” and are expected to be ready and able to competently fulfill their responsibilities. Those who have not, receive the designation “trainee,” and can be expected to need additional guidance and on-the-job training. FQS designations feed into the Deployment Tracking System as one key variable in how the tracking system deploys staff. Among other challenges with FEMA’s Deployment Tracking System and Qualification System, FEMA managers and staff in the field told us an employee’s recorded qualification status was not a reliable indicator of the level at which deployed personnel would be capable of performing specific duties and responsibilities or their general proficiency in their positions, making it more difficult for managers to know the specialized skills or experience of staff and effectively build teams. We are continuing to assess these and other reported workforce challenges and plan to report our findings in January 2020. FEMA Information Technology Challenges In April 2019, we reported on FEMA’s Grants Management Modernization program, which is intended to replace the agency’s 10 legacy grants management systems and modernize and streamline the grants management environment. We found that, of six important leading practices for effective business process reengineering and information technology requirements management, FEMA fully implemented four and partially implemented two for the Grants Management Modernization program. The two partially implemented leading practices were (1) establishing plans for implementing new business processes and (2) establishing complete traceability of information technology requirements. In addition, we found that the program’s initial May 2017 cost estimate of about $251 million was generally consistent with leading practices for a reliable, high-quality estimate; however, it no longer reflected the current assumptions about the program at the time of our review. Moreover, the program’s schedule–specifically its final delivery date of September 2020—did not reflect leading practices for project schedules, as the date was not informed by a realistic assessment of development activities. Lastly, we found that FEMA fully addressed three and partially addressed two of five key cybersecurity practices. The two partially addressed practices were (1) assessing security controls, and (2) obtaining an authorization to operate the system. We made 8 recommendations to FEMA to implement leading practices related to reengineering processes, managing information technology requirements, scheduling system development activities, and implementing cybersecurity. DHS concurred with all of our recommendations and provided estimated completion dates for implementing each of them through July 2020. Thank you, Chairman Thompson, Ranking Member Rogers and Members of the Committee. This concludes my prepared statement. I would be happy to respond to any question you may have at this time. GAO Contact and Staff Acknowledgements If you or your staff has any questions concerning this testimony, please contact Christopher P. Currie at (404) 679-1875 or curriec@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Key contributors to this statement were Joel Aldape (Assistant Director), Amanda R. Parker (Analyst-in-Charge), Matthew T. Lowney, Rebecca Mendelsohn, and David (Ben) Nelson. In addition, Aditi Archer, Bryan Bourgault, Lorraine Ettaro, Aaron Gluck, Kathryn Godfrey, Taylor Hadfield, Eric Hauswirth, Robert (Denton) Herring, Adam Hoffman, Susan Hsu, Sara Kelly, Amy Moran Lowe, Heidi Nielson, Danielle Pakdaman, Sara Pelton, Amanda Prichard, and Johanna Wong made contributions to this statement. Key contributors for the previous work that this is based on are listed in each product. Enclosure I: Related GAO Products Previously Issued Opportunities to Reduce Potential Duplication in Government Programs, Save Tax Dollars, and Enhance Revenue. GAO-11-318SP, March 1, 2011. Federal Disaster Assistance: Improved Criteria Needed to Assess a Jurisdiction’s Capability to Respond and Recover on Its Own. GAO-12- 838, September 12, 2012. Fiscal Exposures: Improving Cost Recognition in the Federal Budget. GAO-14-28, October 29, 2013. Emergency Preparedness: Opportunities Exist to Strengthen Interagency Assessments and Accountability for Closing Capability Gaps. GAO-15-20, December 4, 2014. High-Risk Series: An Update. GAO-15-290, February 11, 2015. Budgeting for Disasters: Approaches to Budgeting for Disasters in Selected States. GAO-15-424, March 26, 2015. Hurricane Sandy: An Investment Strategy Could Help the Federal Government Enhance National Resilience for Future Disasters. GAO-15- 515, July 30, 2015. Disaster Response: FEMA Has Made Progress Implementing Key Programs, but Opportunities for Improvement Exist. GAO-16-87, February 5, 2016. Disaster Recovery: FEMA Needs to Assess Its Effectiveness in Implementing the National Disaster Recovery Framework. GAO-16-476, May 26, 2016. Federal Disaster Assistance: Federal Departments and Agencies Obligated at Least $277.6 Billion during Fiscal Years 2005 through 2014. GAO-16-797, September 22, 2016. Climate Change: Information on Potential Economic Effects Could Help Guide Federal Efforts to Reduce Fiscal Exposure. GAO-17-720, September 28, 2017. Disaster Assistance: Opportunities to Enhance Implementation of the Redesigned Public Assistance Grant Program. GAO-18-30, November 8, 2017. Disaster Recovery: Additional Actions Would Improve Data Quality and Timeliness of FEMA’s Public Assistance Appeals Processing. GAO-18- 143, December 15, 2017. 2017 Disaster Contracting: Observations on Federal Contracting for Response and Recovery Efforts. GAO-18-335, February 28, 2018. Federal Disaster Assistance: Individual Assistance Requests Often Granted but FEMA Could Better Document Factors Considered. GAO-18- 366, May 31, 2018. 2017 Hurricanes and Wildfires: Initial Observations on the Federal Response and Key Recovery Challenges. GAO-18-472, September 4, 2018. Homeland Security Grant Program: Additional Actions Could Further Enhance FEMA’s Risk-Based Grant Assessment Model. GAO-18-354, September 6, 2018. Continuity of Operations: Actions Needed to Strengthen FEMA’s Oversight and Coordination of Executive Branch Readiness. GAO-19- 18SU, November 26, 2018. 2017 Disaster Contracting: Action Needed to Better Ensure More Effective Use and Management of Advance Contracts. GAO-19-93, December 6, 2018. U.S. Virgin Islands Recovery: Status of FEMA Public Assistance Funding and Implementation. GAO-19-253, February 25, 2019. High-Risk Series: Substantial Efforts Needed to Achieve Greater Progress on High-Risk Areas. GAO-19-157SP, March 6, 2019. Puerto Rico Hurricanes: Status of FEMA Funding, Oversight, and Recovery Challenges. GAO-19-256, March 14, 2019. Huracanes de Puerto Rico: Estado de Financiamiento de FEMA, Supervisión y Desafíos de Recuperación. GAO-19-331, March 14, 2019. Disaster Recovery: Better Monitoring of Block Grant Funds Is Needed. GAO-19-232, March 25, 2019. FEMA Grants Modernization: Improvements Needed to Strengthen Program Management and Cybersecurity. GAO-19-164, April 9, 2019. 2017 Hurricane Season: Federal Support for Electricity Grid Restoration in the U.S. Virgin Islands and Puerto Rico. GAO-19-296, April 18, 2019. Disaster Contracting: Actions Needed to Improve the Use of Post- Disaster Contracts to Support Response and Recovery, GAO-19-281, April 24, 2019. Disaster Assistance: FEMA Action Needed to Better Support Individuals Who Are Older or Have Disabilities. GAO-19-318, May 14, 2019. Enclosure II: Ongoing GAO Reviews 1. Review of U.S. Virgin Islands recovery planning and progress; 2. Puerto Rico disaster recovery planning and progress; 3. 2017 wildfire response and recovery; 4. Federal internal control plans for disaster assistance funding; 5. Electricity grid restoration and resilience after the 2017 hurricane 6. Mass care sheltering and feeding challenges during the 2017 7. Department of Transportation highway and transit emergency relief 8. Drinking water and wastewater utility resilience; 9. Review of disaster death count information in selected states and 10. Department of Health and Human Services disaster response efforts; 11. Disaster and climate change impacts on Superfund sites; 12. FEMA Public Assistance program fraud risk management efforts; 13. Wildland fire collaboration on fuel reduction efforts; 14. Preparedness challenges and lessons learned from the 2017 15. FEMA workforce management and challenges; 16. Small Business Administration response to 2017 disasters; 17. Development of the GAO disaster resilience framework; 18. FEMA Individuals and Households Program operations and 19. National Flood Insurance Program post-flood enforcement; 20. Emergency alerting capabilities and progress; 21. National Flood Insurance Program buyouts and property acquisitions; 22. Economic costs of large-scale natural disasters and impacts on 23. Community Development Block Grants – disaster recovery; and 24. Disaster Housing Assistance Program. 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.

Why GAO Did This Study Recent hurricanes, wildfires, and flooding have highlighted the challenges the federal government faces in responding effectively to natural disasters. The 2017 and 2018 hurricanes and wildfires affected millions of individuals and caused billions of dollars in damages. In March 2019, the Midwest experienced historic flooding that affected millions of acres of agriculture and damaged significant infrastructure. Since 2005, federal funding for disaster assistance is at least $450 billion. Increasing reliance on federal help to address natural disasters is a key source of federal fiscal exposure, particularly as certain extreme weather events become more frequent and intense due to climate change. This statement discusses, among other things, FEMA's progress and challenges related to disaster resilience, response, recovery, and workforce management. This statement is based on GAO reports issued from March 2011 through May 2019, and also includes preliminary observations from ongoing GAO reviews of FEMA operations. For ongoing work, GAO reviewed federal laws; analyzed documents; interviewed agency officials; and visited disaster damaged areas in California, Florida, South Carolina, North Carolina, Puerto Rico, Texas, and the U.S. Virgin Islands, where GAO also interviewed FEMA and local officials. What GAO Found GAO's issued and ongoing work identified progress and challenges in the Federal Emergency Management Agency's (FEMA) disaster resilience, response, recovery, and workforce management efforts, as discussed below. Disaster Resilience. GAO found that federal and local efforts to improve resilience can reduce the effects and costs of future disasters. FEMA has made progress in this area, but in July 2015, GAO found that states and localities faced challenges using federal funds to maximize resilient rebuilding following a disaster. GAO recommended that the Mitigation Framework Leadership Group—an interagency body chaired by FEMA—create a national strategy to better plan for and invest in disaster resilience. FEMA is working to address this recommendation and plans to publish the strategy by July 2019. Response and Recovery. In September 2018, GAO reported that the response to the 2017 disasters in Texas, Florida, and California showed progress since Hurricane Katrina in 2005. Specifically, FEMA and state officials' pre-existing relationships and exercises aided the response and helped address various challenges. However, GAO and FEMA identified challenges that slowed and complicated FEMA's response to Hurricane Maria, particularly in Puerto Rico. GAO's issued and ongoing work also identified challenges in implementing FEMA Public Assistance grants. For example, FEMA and Puerto Rico officials identified challenges with Public Assistance policies and guidance that have complicated and slowed the recovery. GAO did not make recommendations, but continues to evaluate recovery efforts and will report its findings later this year. FEMA Workforce Management. GAO has previously reported on long-standing workforce management challenges, such as ensuring an adequately-staffed and trained workforce. For example, GAO reported in September 2018 that the 2017 disasters overwhelmed FEMA's workforce and a lack of trained personnel with program expertise led to complications in its response efforts, particularly after Hurricane Maria. While FEMA has taken actions to address several of GAO's workforce management-related recommendations since 2016, a number of recommendations remain open as the 2019 hurricane season begins. Also, GAO is currently reviewing FEMA's workforce management efforts and lessons learned from the 2017 disasters and will report its findings early next year. What GAO Recommends GAO has made numerous recommendations in its prior reports to FEMA designed to address the challenges discussed in this statement. As of May 2019, FEMA has addressed about half of these recommendations and GAO is monitoring FEMA's ongoing efforts.

gao_GAO-20-339

gao_GAO-20-339_0

Background DOD started the F-35 program in 2001 to develop a fifth-generation fighter aircraft intended to replace a range of aging aircraft in the U.S. military services’ inventories and to provide enhanced capabilities to warfighters that capitalized on technological innovations. Among other capabilities, the program designed the F-35 aircraft to be difficult to observe using radar and include sensors that can provide insights into potential targets and other warfighting information. The program is producing and delivering three variants of the F-35 aircraft: the F-35A conventional takeoff and landing variant for the Air Force, the F-35B short takeoff and vertical landing variant for the Marine Corps, and the F-35C carrier-suitable variant for the Marine Corps and the Navy. The characteristics of the services’ variants are similar, but each variant also has unique operating requirements. For example, the Marine Corps requires that the F-35B be capable of operating from aircraft carriers, amphibious ships, and main and austere operating bases alike, requiring the capability to conduct short takeoffs and vertical landings. Figure 1 shows an F-35B exercising this capability. While DOD plans to purchase 2,470 aircraft for the U.S. services, the F- 35 program is acquiring more than just aircraft. The complete F-35 air system has eight elements, including training and maintenance systems. Figure 2 shows the eight elements that make up the entire F-35 Air System and how they each support the aircraft. For example, the program intends for the Automated Logistics Information System (ALIS) to provide the necessary logistics tools to F-35 program participants as they operate and sustain the F-35 aircraft. To do this, ALIS consists of multiple software applications designed to support different squadron activities, such as supply chain management, maintenance, training management, and mission planning. For the F-35 aircraft to have full capability, each element of the air system has to be developed and fielded in sync with the aircraft. However, we found in March 2020 that problems with ALIS still pose significant challenges to day-to-day F-35 operations. According to DOD, it plans to replace ALIS with a new system named the Operational Data Integrated Network (ODIN). Furthermore, DOD reports that it is currently developing a strategy for ODIN, which will include key tasks, milestones and schedule, risks and opportunities, governance structure, and cost estimates. We concluded that, as DOD proceeds with replacing ALIS with ODIN, it will be important for the department to carefully consider and assess the key technical and programmatic uncertainties that we reported in March 2020. These include how much of ALIS will be incorporated in ODIN and the extent to which DOD has access to the data it needs to play a more active role in the management of the system. These issues are complex, and will require significant direction and leadership to resolve. Further, we reported in March 2020 that the F-35 program office was not able to provide us with historic costs showing how much the department had spent on ALIS over the years. Also, because DOD had not answered key questions about the future of the system, such as the extent to which the re-design will incorporate current ALIS software, DOD has not been able to develop accurate cost estimates for the ALIS re-design. We recommended that DOD develop and implement a strategy for the re- design of ALIS. The strategy should be detailed enough to clearly identify and assess the goals, key risks or uncertainties, and costs of re- designing the system. DOD concurred with the recommendation. Status of F-35 Program Development and Costs as of April 2019 DOD began development of the F-35 aircraft in 2001 without adequate knowledge of its critical technologies or a solid design, as we reported in March 2005. DOD’s acquisition strategy also called for high levels of concurrency between development and production—building aircraft while continuing to refine the designs of key components—which runs counter to GAO’s leading practices for major defense acquisition programs. In our prior work, we identified the F-35 program’s lack of adequate knowledge and high levels of concurrency as the major drivers of the program’s eventual significant cost and schedule growth, among other performance shortfalls. Since 2001, the program has been rebaselined with new cost and schedule estimates three times. DOD initiated the most recent restructuring in 2010 when the program’s cost estimates for each aircraft exceeded critical thresholds established by statute—a condition known as a Nunn-McCurdy breach. DOD then established a new acquisition program baseline that increased the program’s cost estimates by $162.7 billion and extended delivery schedules 5-6 years into the future. This last revision is the current program baseline, reflecting the cost and schedule estimates to deliver the aircraft and systems and to meet the original program requirements. From 2018 to 2019, the total cost estimate of the F-35 acquisition program increased by $22 billion, from $406 billion to over $428 billion. This increase was partially due to the addition of the estimated Block 4 modernization costs. Block 4 includes efforts to enhance and add capabilities—beyond the F-35 baseline program—through hardware and software upgrades. In April 2019, the F-35 program estimated that Block 4 development and procurement costs would add $13.9 billion to the program’s total baseline cost. Beyond this Block 4 increase, the F-35 program baseline costs also increased by $8 billion over the program’s 2018 estimate. Table 1 outlines the program’s baseline costs, the Block 4 modernization costs, and the sum total of the baseline and Block 4 cost estimates since 2001. In addition to the acquisition costs above, the program estimates that the sustainment costs to operate and maintain the F-35 fleet for its planned 66-year life cycle are $1.2 trillion, bringing the total cost of the F-35 program to over $1.6 trillion. Status of Testing, Production, and Reliability and Maintainability as of December 2019 The F-35 program office, in coordination with the Director of Operational Test and Evaluation (DOT&E), received approval to conduct some preliminary operational testing in January 2018. This included weapons, cybersecurity, and cold weather testing, among other things. The program’s formal operational testing (conducted by DOT&E) started in December 2018 and was ongoing in 2019. The purpose of operational testing is to assess the effectiveness, suitability, survivability, lethality, and mission capability of the F-35, including the information systems and the air vehicle, in an operationally representative environment. Operational testing includes cybersecurity assessments, some of which the program has conducted. The program plans for the remaining testing to take place through at least September 2020, while the program continues to produce and deliver aircraft. Through 2019, F-35 program test officials had identified over 3,200 deficiencies. Deficiencies represent specific instances where the weapon system either does not meet requirements or where the safety, suitability, or effectiveness of the weapon system could be affected. The test officials categorize deficiencies according to their potential impact on the aircraft’s performance. Category 1 deficiencies are critical and could jeopardize safety, security, or another requirement. Category 2 deficiencies are those that could impede or constrain successful mission accomplishment. In June 2018, we recommended that the program resolve all critical deficiencies before making a full-rate production decision, in part, to reduce the potential for additional concurrency costs stemming from continuing to produce aircraft while testing was ongoing. DOD concurred with our recommendation and stated that it would resolve critical deficiencies before full-rate production, currently planned to occur between September 2020 and March 2021. Production of the aircraft began one year after testing started in 2007, while development was in its early stages. Due to the concurrency of testing and production, according to an F-35 program official, as many as 550 aircraft delivered through 2020 will need retrofits to fix deficiencies and design issues found during testing. The program refers to the cost of these fixes as its concurrency cost, which the program estimates at $1.4 billion; this estimate did not change with the program’s last update in 2019. Until operational testing is complete, there is a risk that the program may identify additional deficiencies. As a result, as we have previously reported, the concurrency costs of retrofitting delivered aircraft could increase. In our June 2018 report, we found that the program was not on track to meet its reliability and maintainability (R&M) performance targets. R&M targets indicate how much time the aircraft will be in maintenance rather than operations. We concluded that the program was missing a prime opportunity to infuse affordability into the aircraft’s future with better R&M performance. As a result, we recommended that the F-35 program office identify what steps it needed to take to ensure the F-35 meets R&M requirements and update the R&M Improvement Program with these steps. DOD concurred with the recommendations, noting that the F-35 program office would update the R&M Improvement Program with the steps needed to ensure continued progress towards its goals. In April 2019, we found that F-35 R&M performance had shown some small improvements but that the program could take more actions to meet the R&M targets. We made additional recommendations to the Secretary of Defense, with which DOD concurred and has taken some actions to implement. Currently, the Office of the Under Secretary of Defense for Acquisition and Sustainment (OUSD (A&S)) is the acquisition decision authority for the F-35 program, and would direct the F-35 program office to take any further actions. In 2019, the program’s R&M performance generally remained unchanged. However, measurable improvements in R&M can take time to manifest. For example, fielded aircraft must be modified and flown for many hours before the program can measure improvements. For details about the R&M performance, see appendix IV. Block 4 Modernization’s Development Approach As we have previously reported, even though operational testing of the baseline program remains ongoing, the F-35 program office has turned its attention to Block 4 modernization activities using a different development approach. DOD refers to this approach as Continuous Capability Development and Delivery (C2D2). This method is loosely based on the Agile software development process. With this approach, the program plans to deliver capabilities to the warfighter faster than it did during the baseline development program. For example, rather than take years to develop and deliver all the required capabilities to the warfighter, the program intends to incrementally develop, test, and deliver small groups of capabilities every 6 months. In January 2018, to transition from the baseline development program to its Block 4 activities, the F-35 program started using the C2D2 approach to develop and test software updates to address deficiencies identified during testing. The planned $13.9 billion Block 4 effort exceeds the statutory and regulatory thresholds for what constitutes a major defense acquisition program, and Block 4 is more expensive than many of the other major weapon acquisitions already in DOD’s portfolio. To provide better oversight into Block 4 activities, in 2016, we recommended that the Secretary of Defense hold a milestone B review—a critical point in an acquisition program leading to the engineering and manufacturing development phase—and manage it as a separate major defense acquisition program. DOD did not concur with our recommendation, and it continues to manage Block 4 within the larger F-35 program. We maintain that DOD should manage the Block 4 activities as a separate program. Operational Testing Delays Provide More Time to Address Deficiencies before Full-Rate Production Decision Completion of Operational Testing Delayed by 9 Months In 2019, the F-35 program conducted a majority of its planned operational testing but added 9 months to the schedule to complete the remaining tests. Specifically, as of February 2020, according to test officials, the program completed 156 flight tests. The program must still conduct four open-air flight tests, the remaining cybersecurity tests of the air vehicle and mission systems, and 64 simulated flight tests. The 9-month delay needed to complete testing, however, also provides additional time for the program to address our June 2018 recommendation that it resolve critical deficiencies before making its full-rate production decision, currently planned to occur between September 2020 and March 2021. Figure 3 shows the test schedule as of 2019, the delay to the schedule into 2020, and the remaining tests events planned. The completion of operational testing hinges on three main tasks: (1) the final four open-air flight tests; (2) cybersecurity testing; and (3) the final development, integration, verification and validation of its simulator and 64 simulated flight tests. First, the program expects to complete the four remaining open-air tests between March and April 2020. To conduct these tests, the program must finish moving the Radar Signal Emulators—test assets that simulate long- range threat radars—from the Nevada Test and Training Range to the Point Mugu Sea Range in California. According to test officials, there is some risk with this move, such as damage to the sensitive test equipment. The test facilities will have to integrate the equipment into the testing infrastructure at Point Mugu. Second, while the program has conducted cybersecurity testing on several aspects of the F-35 aircraft and support systems, three air vehicle subsystems tests and two enterprise-level ALIS tests remain. The program expects to complete these by August 2020.The tests completed to date have identified multiple cybersecurity vulnerabilities. The F-35 program office has taken steps to address some identified vulnerabilities and is working to address the remainder. Test officials stated that some of the delays to cybersecurity testing of the aircraft are due to safety concerns and the risk of losing the use of a test aircraft before testing is complete. According to DOD policy, cybersecurity testing should be conducted as early in the operational test cycle as possible. Leaving this critical testing to the end of operational testing adds risk to the program because the program will not know the extent to which the aircraft may have cybersecurity vulnerabilities until near the expected decision to proceed to full-rate production. If the program cannot finish these tests by September 2020, officials stated that DOT&E could require that the cybersecurity testing be completed in follow-on testing and not hold up the full-rate production decision. Any additional cybersecurity vulnerabilities may require more time to develop and implement plans to address vulnerabilities in aircraft that have already been produced and those slated for production. Lastly, the program has not been able to complete the F-35 Joint Simulation Environment, which we refer to as the aircraft simulator, on time. The simulator runs the F-35’s mission systems software along with other software models (such as other weapons and modern threat systems) to provide complex test scenarios that the program cannot replicate in a real-world environment. We reported in April 2019 that the simulator’s development was behind schedule and was a risk to operational testing. Since then, the program has struggled to develop the complex software and functionality needed to complete the simulator. The difficulties stem, in part, from the program office’s original plan to have the contractor, Lockheed Martin, develop the simulator. However, in August 2017, program office officials decided that the contractor’s proposal was considered to be too expensive. To mitigate concerns over the cost of the proposal, the program decided to have the Navy complete the work. The program originally expected the Navy’s simulator to be ready for testing in 2017, but it is now 3 years behind schedule. According to program office officials, the simulator’s development effort has taken longer than expected to integrate F-35 aircraft and sensor data, in part because the contractor claimed the data as its own intellectual property. These issues were resolved by 2019 when the contractor provided the necessary data. Because of these delays, the program now expects that the simulator will be ready by August 2020, with the planned simulator testing expected to take about 3 weeks. According to test officials, there is increased risk that the completion of the simulator may face additional delays to correct deficiencies and add needed capabilities, but also stated that they can complete the tests by August 2020. Due to these delays to completing operational testing, the program has delayed its full-rate production decision by at least an additional 9 months. Though the program is working toward September 2020, the program has acknowledged this decision could be made as late as March 2021. Any additional delays due to challenges with moving the emulators, completing the simulator, or cybersecurity testing could further delay the end of operational testing and the program’s decision to enter into full-rate production. This delay, however, gives the program more time to complete two key steps consistent with statute and DOD policy. Complete operational testing, which is intended to demonstrate that the aircraft are operationally suitable. Resolve all deficiencies, which should be done prior to full-rate production, and is discussed below. Even with these delays, the program plans to have produced and delivered over 550 aircraft before operational testing is complete, adding to the risk of finding more deficiencies that will require retrofits—at additional cost—for the delivered aircraft. Statute and DOD policy states that the preliminary low-rate production quantities will be set at the development request for proposal decision point. If, at that time, low-rate initial production quantities are determined to be above 10 percent of the total quantity planned, the Secretary of Defense must explain the reasons for the increase in a report to Congress. When a program reaches the planned low-rate initial production quantity, and requires to exceed the quantity, the program may seek approval to produce quantities above that amount. The F-35 program will have delivered more than 10 percent of the total planned production quantities—due to the necessity to prevent a break in production—before operational testing and the full-rate production decision are complete. As noted above, this approach has contributed to the $1.4 billion in concurrency costs already incurred by the program. The program reports that none of the category 1 deficiencies is a safety of flight concern and all of them have operational workarounds. In 2019, the program split the category 1 deficiencies into two groups. Group A are deficiencies that may cause death, severe injury, severe occupational illness, or major loss or damage to equipment and has no workaround. The program has none of these deficiencies currently. Group B are deficiencies that may critically restrict the combat readiness capabilities or may result in adequate performance but not be able to accomplish the primary or alternate missions. All of the 9 category 1 deficiencies are in group B. F-35 Program Made Progress Resolving Deficiencies during 2019, but Many Remain Unresolved In 2019, the F-35 program resolved nearly 300 of the deficiencies it had identified in developmental and operational testing, but discovered even more over the same period. Specifically, 331 new deficiencies were identified in operational testing during 2019. As of December 2019, the F- 35 program had 870 open deficiencies. Of the 870 open deficiencies, the program characterizes nine as category 1 and 861 as category 2. The program reports that none of the category 1 deficiencies is a safety of flight concern and all of them have operational workarounds. In 2019, the program split the category 1 deficiencies into two groups. Group A are deficiencies that may cause death, severe injury, severe occupational illness, or major loss or damage to equipment and has no workaround. The program has none of these deficiencies currently. Group B are deficiencies that may critically restrict the combat readiness capabilities or may result in adequate performance but not be able to accomplish the primary or alternate missions. All of the 9 category 1 deficiencies are in group B. Of the 9 open category 1 deficiencies, the program reports all have operational workarounds—procedures that avoid encountering the deficiency. This represents four fewer open category 1 deficiencies than we reported in April 2019, reflecting the resolution of previously identified deficiencies and the addition of new ones, some of which were resolved. For example, the program fielded a software fix to a category 1 deficiency, which showed that the F-35’s cockpit display could falsely indicate its AIM-9X weapon—an air-to-air missile—selection status as “selected” though the weapon’s status is not selected. Figure 5 shows the F-35 firing an AIM-9X missile. The program office plans to continue to address the open deficiencies, but officials report that some will not be fully resolved for several years. Further, some deficiencies may not be resolved ever and some may be resolved well after the program has completed testing, and after it expects to have made a full-rate production decision. According to DOT&E, there are many significant deficiencies the program should address to ensure the F-35 baseline aircraft configuration is stable prior to adding all of the new capabilities planned in Block 4. As of December 2019, the program office and the contractor have resolved over 2,300 deficiencies and program office officials stated that they have a process in place to address the high priority ones. F-35 Aircraft Prices Decreased but Manufacturing and Supply Chain Risks Remain In 2019, the program reported continuing to negotiate lower unit prices across all F-35 aircraft variants and delivered more aircraft on time. However, officials also reported that the airframe and engine contractors demonstrated some declines in production performance, such as the number of labor hours to produce each aircraft, as production rates increased. We also identified other risk indicators that could affect the contractors’ future production performance. Specifically, the airframe contractor’s manufacturing processes do not meet all manufacturing leading practices that programs should meet before full-rate production. Additionally, parts shortages increased significantly in 2019 and Turkey’s suspension from the program will likely further complicate existing supply chain challenges. Aircraft Prices Decreased and More Aircraft Were Delivered on Time, but Other Measures of Production Performance Declined According to the program office, the negotiated prices for all F-35 variants have generally been decreasing with each production lot and as more aircraft are being procured in each lot. In April 2019, we reported that the program set a goal of reducing the negotiated unit price of an F-35A to less than $80 million by lot 13. According to a program official, in October 2019, the program finalized the contract action for lots 12-14 that met this goal. Specifically, with the most recent contract, the program agreed to purchase 351 F-35As, with unit costs declining to $73 million in lot 14. Figure 6 shows how the negotiated price for an F-35A has decreased since production began, as reported by the program office. According to the program office, it negotiated lower unit prices by working with the airframe contractor to leverage economic order quantity purchases and invest in cost reduction initiatives. Economic order quantities involve the contractor making large purchases of components that it will use across multiple procurement lots of aircraft to reduce production costs by achieving economies of scale. The program office estimates that the economic order quantity purchases for lots 12-14 will save the program about $225.5 million. In addition, the program office and prime contractors have continued to invest in various initiatives to lower production costs. Specifically, the program office spent $320 million in efforts to improve manufacturing processes that it estimates could result in up to $10.5 billion in savings over the life of the program. The airframe contractor told us that it has invested $170 million as of January 2019 to further lower its production costs. The engine contractor also told us that it spent $33 million to potentially realize over $4 billion of cost savings. Airframe Production Trends The airframe contractor—Lockheed Martin—delivered 43 more aircraft in 2019 than in 2018, and as of October 2019, there were 229 aircraft in various stages of assembly worldwide. The contractor also delivered more aircraft on time in 2019. According to contractor officials, the improved rates of on-time delivery are partially a result of the contractor’s efforts to obtain a performance incentive fee that was added to the lot 11 production contract. The program intended the incentive fee to focus the contractor on improving its performance in the final assembly phase of production, which was expected to improve its on-time deliveries. To earn the incentive fee, contractor representatives told us they took several steps to improve production rates. For example, because the F-35Cs were taking longer to produce and all variants had to move through the same final assembly area, the contractor made a separate final assembly line for the F-35Cs so work could proceed without delaying the other variants. This step, according to program office officials, allowed the contractor to improve on-time deliveries for F-35As and F-35Bs. Figure 7 highlights progress in the contractor’s aircraft deliveries since 2016. Other production metrics associated with the airframe, however, demonstrated varied performance over the last two years as production increased. For example, the average number of hours needed to build an aircraft decreased slightly for the F-35A but increased for the B and C variants. Defense Contract Management Agency officials told us the increase was partly attributable to new personnel. In particular, since January 2017, the contractor has hired and trained nearly 1,700 new personnel to accommodate increased production rates—nearly doubling its workforce. New personnel take time to train and gain experience on the production line. According to contractor representatives, as these new employees become more experienced and produce more aircraft, they expect the metric to improve. The contractor’s amount of rework needed was also mixed. During the course of production, the contractor may identify issues with a part or a process, which, in turn, may lead to scrap, rework, and repair to replace or fix the issue. Between 2016 and 2017, most F-35 variants realized improvements in the amount of scrap, rework, and repair needed. In 2018 and 2019, however, only the F-35A continued to show improvements. Figure 8 shows the average total hours for scrap, rework, and repair for each variant since 2016. According to the program office, the increased production rate posed a challenge, and because the contractor has not built as many F-35Cs, this has added to the increase in scrap, rework and repair. To improve performance in this regard, the contractor put teams in place to focus on addressing the main drivers of scrap, rework, and repair. Engine Production Trends Similarly, the engine contractor—Pratt & Whitney—increased its production rate by roughly 51 percent in 2019. However, engine on-time delivery performance has continued to decline which officials attribute to production quality issues and parts delays. Specifically, in 2019, 91 percent of engines delivered were late. In 2019, the airframe contractor was able to work around the late engine deliveries to deliver the entire aircraft on time. Figure 9 shows the engine contractor’s on-time and late deliveries since 2016. In addition, the average number of quality notifications per engine— production defects indicating a quality issue—has increased by 16 percent in 2019. Figure 10 highlights the engine contractor’s quality notifications per engine over the last 4 years. According to the Defense Contract Management Agency’s performance reports, engine test failures, among other quality issues, have affected engine deliveries. According to an official from this agency, there have been 18 engine test failures in 2019, which is eight more than in 2018, each requiring disassembly and rework. The engine contractor stopped deliveries due to the test failures, which has slowed engine acceptance and reduced on-time deliveries. These issues are affecting engines built at the engine contractor’s production facility in West Palm Beach, Florida, which opened in 2014. To address this issue, the engine contractor has developed new tooling for the assembly line and has established a team to identify characteristics leading to the test failures. Plans are also in place for additional training for employees. F-35 Program Has Not Met All Manufacturing Leading Practices, Indicating Risk to Future Production While F-35 aircraft have been in production since 2007 and have reached a high level of manufacturing readiness per DOD guidance, the program is not meeting two of eight manufacturing leading practices GAO has identified as indicators of a program’s readiness for full-rate production, or milestone C review. To date, the program is meeting or plans to meet six leading practices for this milestone: Demonstrating processes on a pilot production line. Building and testing production-representative prototypes to demonstrate product in intended environment. Collecting statistical process control data. Conducting an independent cost estimate. Conducting an independent program assessment. Conducting major milestone decision review to begin production. However, we also found that the production processes are not in control according to the Process Capability Index. This index is a tool to measure how closely the production steps result in a part or subsystem that meets predefined standards. According to the leading practices, meeting these standards provides greater confidence that the contractor can produce a high quality product consistently, to minimize variation which results in fewer defects or the need for rework. Additionally, the F- 35 aircraft have not achieved their reliability goals through testing of production representative prototypes. These two leading practices focus on gathering sufficient knowledge to determine the relative ease of manufacturing and whether the product is of high quality and sufficiently mature to move forward into full-rate production. Our analysis of contractor data shows that the airframe contractor’s production processes are in flux. The contractor continues to change some of its production processes, and in other cases, is not following its own established processes well, which has led to several quality issues over the years. For example, in 2018, we reported that the contractor had halted deliveries of aircraft after the Air Force identified corrosion between the aircraft’s surface panels and the airframe because the contractor did not apply a primer when it attached the panels. We reported in 2019 that the program office, the contractor, and the F-35 Program Executive Officer reached a mutual agreement on the cost to resolve this issue, the details of which they did not disclose publicly. In November 2019, a mechanic identified titanium fasteners installed in an area of the aircraft where the design calls for a fastener stronger than titanium. According to the program office, the incorrect fasteners were installed on most already- fielded F-35 aircraft. That same month, the contractor started implementing its corrective action plan. As of March 2020, the F-35 program office had reviewed and approved the contractors’ analysis as well as its durability and damage reports on the use of these fasteners. We describe other key F-35 technical risks in appendix V. Over the years, the airframe contractor has continued to change and refine production processes, aiming to improve efficiency amidst concurrent development and production. For example, the airframe contractor identified a particular process that installs wiring harnesses into the aircraft wings as a driver of one of its production quality issues. To address this issue, the prime contractor developed a new tool that helps the installer route the wires more consistently. While process changes like these can improve the quality of the product, they also indicate that the overall production process are not in control less than a year before the program’s planned full-rate production decision, or milestone C review. In 2019, according to our analysis, the total number of key F-35 manufacturing processes identified in the final assembly phase increased 70 percent, to a total of over 10,000 critical processes. Furthermore, of these critical processes, only 30 percent are currently able to produce a product within predefined design standards. According to manufacturing leading practices, critical processes should be repeatable, sustainable, and consistent in producing parts within quality standards. Meeting these practices provides confidence that the contractor can produce the product within cost, schedule, and quality targets. Without processes in control, the program could face continued quality issues that will add to the overall cost of the program. Figure 11 shows the F-35 aircraft in the final assembly phase of production where some of these processes take place. Another leading practice that should be met before making a full-rate production, or milestone C decision, is to demonstrate that a production representative prototype can meet the program’s R&M goals. The R&M goals lay out specific quantitative goals aimed at ensuring that an aircraft will be available for operations as opposed to out of service for maintenance. We reported in April 2019 that the F-35 aircraft in service around the world were still not meeting all of their R&M goals and recommended the program take actions to ensure that the aircraft would meet those goals. Despite some improvement in 2019, the program is not meeting half of its R&M goals. Until the program does so, the warfighter will continue to accept aircraft for delivery that are less reliable and more costly to maintain than originally planned. For details on the F- 35’s R&M performance, see appendix IV. The program has not met these two leading manufacturing practices, in part, due to the changes the airframe contractor made and continues to make to the production line and the program’s concurrent approach to acquisition. We have repeatedly found that DOD programs that moved into full-rate production carrying manufacturing risks experienced billions of dollars in cost growth in production, and nearly two-thirds reported increases in average procurement unit costs. With the risks the F-35 program still faces, it may realize additional cost and schedule growth if these production risks are not evaluated. Despite these risks, the program has continued to push forward with increased production rates and has not taken actions to determine the potential impact of not meeting these leading practices may have on future production and overall life-cycle costs. Furthermore, according to a program official, the F-35 program has not completed a comprehensive assessment of production risks and does not plan to ahead of its full-rate production decision. However, according to DOD officials, the F-35 program office and prime contractor convene a monthly Joint Risk Management Board, which identifies and manages overall program risk, and has completed an independent technical risk assessment to support the full-rate production decision, which identified production risks. Title 10 section 2366c of the U.S. Code requires the milestone decision authority for a major defense acquisition program to provide Congress with a report that includes, among other things, a summary of any manufacturing risks associated with the program; however, this summary is not required until 15 days after the authority grants approval for the program to enter the production and deployment phase. The program currently plans to obtain this approval between September 2020 and March 2021. In this case, however, the F-35 program has not met all of the manufacturing leading practices that should be met before the full-rate production decision. Furthermore, the underlying risks, such as not meeting R&M goals, have persisted for years and the program has yet to take steps to fully address these risks. If an evaluation of these risks is not provided ahead of the full-rate production decision, Congress will not be fully aware of the risks the program is taking by committing to increased production rates. F-35 Supply Chain Challenges Continue and May Be Exacerbated by Turkey’s Suspension from the Program According to program officials, some suppliers for the F-35 struggled to meet increased production demands in 2019 and, as a result, the program witnessed increased rates of late deliveries or parts shortages. In particular, the number of parts delivered late to the airframe contractor, as well as parts shortages, have grown steadily over the past 2 years. According to the Defense Contract Management Agency: Between August 2017 and July 2019, the number of parts delivered late increased from under 2,000 to more than 10,000. Between July 2018 and July 2019, the parts shortages per month increased from 875 to over 8,000. According to contractor representatives, roughly 60 percent of parts shortages are attributable to 20 suppliers. To mitigate late deliveries and parts shortages—and deliver more aircraft on time—the airframe contractor has utilized methods such as reconfiguring the assembly line and moving planned work between different stations along the assembly line. According to the program office, such steps can cause production to be less efficient, which, in turn, can increase the number of labor hours necessary to build each aircraft. Airframe contractor representatives and a program office official cited measures they are taking to improve supplier performance in light of the upcoming full-rate production decision. For example, the contractor instituted action plans to help problematic suppliers, sent task teams to struggling suppliers to help resolve issues, and, in some cases, is seeking alternative sources. Additionally, the program office has established joint meetings with the prime contractor to monitor progress on a weekly basis and holds a semiannual review to achieve executive-level coordination. While prime contractor representatives told us that they have been actively managing underperforming suppliers for several years, some of their efforts are new and will need time before results materialize. These supply chain risks may compound as the program continues to produce, deliver, operate, and maintain more aircraft each year. For example, in April 2019, we found that fielded, operational F-35 aircraft were not meeting warfighter requirements, largely due to spare parts shortages and difficulty in managing and moving parts around the world. We recommended that the program assess what actions it should take to meet warfighter requirements, which could include adjusting the amount of spare parts acquired. DOD concurred and is working toward addressing the recommendation to identify warfighter gaps with regard to the supply chain. However, with the aircraft in production also facing significant shortages, this problem could get worse as the program prepares to further increase the production rate from 141 aircraft in 2019 to 169 in 2022. We found that Turkey’s recent suspension from the F-35 program is likely to compound these existing supply chain issues. In July 2019, Turkey was suspended from the F-35 program. In particular, the Under Secretary of Defense for Acquisition and Sustainment directed that the F-35 program establish alternative sources and to stop placing orders from Turkish suppliers after March 2020. According to an official with that office, Turkish suppliers will provide parts through the end of lot 14 deliveries (scheduled to take place through 2022), in part, to avoid disruptions to aircraft deliveries and additional cost growth from standing up new suppliers. The F-35 program office identified that Turkish companies supplied 1,005 parts for the F-35 airframe and engine and some of these parts have been provided by only one supplier. As of December 2019, the program has identified new suppliers for all of these parts, but it still needs to bring roughly 15 parts currently produced in Turkey up to the current production rate. During our review, the program reported that production through lot 14 should not be adversely affected if it continues to accept parts from Turkey until lot 14 aircraft are delivered, but risk remains with the transition to alternate sources. However, lots 12-14 still face some risk receiving parts from Turkey. According to program officials, some of these new parts suppliers will not be producing at the rate required until next year, as roughly 10 percent are new to the F-35 program. Airframe contractor representatives stated it would take over a year to stand up these new suppliers, with lead times dependent on several factors, such as part complexity, quantity, and the supplier’s production maturity. In addition, these new suppliers are required to go through qualification and testing to ensure the design integrity for their parts. According to an official with the Under Secretary of Defense for Acquisition and Sustainment, by accepting parts from Turkish suppliers through lot 14, the program will have additional time to ensure new suppliers can meet demands for parts. Additionally, the program reported that it intends to utilize alternative sources for parts currently made in Turkey for aircraft delivered under lots 13 and 14 contracts. Furthermore, according to a program office official, it is also not clear how the prices for parts that will be obtained from new suppliers after Lot 14 will compare with the prices under the contracts with the suppliers from Turkey, but the official noted that alternative sources could be more costly. Block 4 Reporting Requirement Expires Before Completion Date, and Cost Estimate Does Not Fully Reflect Leading Practices In its May 2019 report to Congress, DOD outlined its plans for Block 4 with a development cost estimate of $10.6 billion for activities through fiscal year 2024. Since the 2019 report, we found the program office has increased its estimate by about 14 percent, to $12.1 billion, primarily due to schedule delays. The program now expects to extend the delivery of Block 4 capabilities by 2 additional years, through 2026. In the meantime, DOD’s Block 4 annual reporting requirement to Congress is scheduled to end in 2023, 3 years before development is complete. Additionally, most of the capabilities the F-35 program planned to deliver in 2019 were delayed. Furthermore, we found that the program’s cost estimate used to support its report to Congress does not fully meet cost estimating leading practices. Block 4 Modernization Will Not Be Completed before Reporting Requirements Expire, in Light of Schedule Delays The Block 4 development cost and schedule have grown considerably since DOD’s last report to Congress. In 2016, GAO recommended that DOD manage Block 4 as a major defense acquisition program with its own reporting requirements, separate from the original F-35 development program. DOD did not concur with our recommendation, citing the F-35 as DOD’s most closely managed system and its existing F-35 program oversight. The NDAA for Fiscal Year 2017 required DOD to report annually on elements of a Block 4 baseline, such as development and retrofit cost estimates, beginning no later than one year after the award of the development contract for follow on modernization, until March 31, 2023. At that time, we reported that DOD had requested funding for the development and delivery of Block 4 through the end of 2022. However, over the last year, the program has revised its Block 4 schedule and now expects to field Block 4 capabilities into fiscal year 2026. As a result, there is no requirement for DOD to report on Block 4 progress for at least 3 years even though those efforts will be ongoing. In its May 2019 Block 4 report to Congress, DOD reported that the total cost to develop 66 Block 4 capabilities—both hardware and software— would be $10.6 billion for activities planned from fiscal years 2018 to 2024. The report also included the F-35 program office estimate of an additional $6.4 billion in fiscal year 2018 through 2024 funding to retrofit aircraft from the baseline F-35 configuration to a full Block 4 configuration. The F-35 program based the costs in this report on its Block 4 development cost estimate from July 2018. However, we found that reported Block 4 costs did not include all Block 4 costs. In particular, the report did not include Block 4 costs the program incurred prior to 2018 or costs that the effort will incur after 2024. Because the F-35 program office is not managing the Block 4 effort as a separate program, it has chosen to exclude the past and future costs in the Block 4 cost estimate it reported to Congress. Instead, the program reported on Block 4 costs for the future years defense program—which is DOD’s projected spending for the current budget year and the next four years. By excluding any costs prior to 2018 and those that would be incurred after 2024, the program did not report on the total costs of Block 4. In May 2019, the program also updated its Block 4 development cost estimate, increasing both the time and cost to complete the work, but this updated estimate was not included in its May 2019 report to Congress. The updated cost estimate reflects that the program office will be fielding Block 4 capabilities into fiscal year 2026. This new schedule adds 2 years to the costs DOD reported to Congress in May 2019. Additionally, our analysis of DOD’s updated cost estimate indicates the total cost of Block 4 development grew by $1.5 billion to a total of $12.1 billion for activities in fiscal years 2018 through 2026. Furthermore, in addition to the Block 4 development costs, the program also estimates it will need another $2.9 billion to develop other capabilities, such as upgrades to ALIS. Program officials attributed this schedule and cost growth to having better insight into the scope of work to develop and test Block 4 capabilities and noted that they would continue to refine and update these costs annually as modernization efforts progress further into development. Once the existing statutory reporting requirement expires in 2023, DOD will no longer be required to provide Congress key information that would be useful in making informed decisions regarding the Block 4 effort— which now extends until 2026. Furthermore, without a complete cost estimate for Block 4, inclusive of costs already incurred and those not yet incurred but estimated through completion, Congress is left without a complete picture of what DOD intends to spend on the total Block 4 effort. Without a complete picture of these costs, the Congress’s ability to assess the program’s cost and schedule performance in the future will be hindered. Program Office Delayed Delivery of Most 2019 Block 4 Capabilities to the 2020s The airframe contractor did not deliver the Block 4 capabilities it planned to deliver in 2019. Specifically, according to the plan outlined in its May 2019 report to Congress, the F-35 program was going to deliver eight Block 4 capabilities in 2019. However, the program delivered only one—a software capability called the auto ground-collision avoidance system. This capability enables the aircraft to perform an automatic recovery when it predicts that the aircraft will strike the ground. This was ahead of schedule as the program had originally planned to deliver this capability after 2019. According to program officials, the development of the other capabilities is taking longer than planned and, as a result, the program pushed their delivery schedule into 2020. Development and delivery of the capabilities within the Block 4 effort are complex, and the program does not consider development complete until the products for all elements of the F-35 air system are ready. In particular, full capability delivery occurs when the contractor delivers all of the software and hardware needed for all of the F-35 air system elements to support the planned capability. Program officials stated they are still working to put the processes in place to synchronize the delivery of the late capabilities for all of the F-35 air system elements. For example, the airframe contractor had planned to deliver a capability called the interim full motion video for the Marine Corps in 2019. The contractor developed the software needed, but it is late in developing the hardware needed for the software to operate and, as a result, the contractor did not deliver the capability in 2019 as planned. DOD test officials we met with at Edwards Air Force Base stated that in 2019, using the C2D2 approach, the contractor delivered other, partial Block 4 capabilities to be tested. However, test officials told us those capabilities were delivered later than expected. Since the program could not fully test those capabilities on the aircraft, the program office deferred them to the next incremental update scheduled for 2020. Changes such as these have contributed to the Block 4 cost and schedule growth. The program is also discovering issues during Block 4 testing, causing the testing to take longer than anticipated. According to a DOT&E official, Block 4 software changes caused issues with functionality of F-35 baseline aircraft capabilities that worked before the program installed new Block 4 software onto the aircraft. The program discovered issues with each new software version during flight testing and has been working to fix these issues in subsequent software updates. Testing and DOD officials stated that the contractor had not performed adequate testing of the software before delivering it to the test fleet as the reason for these issues. Contractor representatives acknowledged these issues and stated that they will conduct additional lab testing for future software releases to avoid such problems going forward. Block 4 Development Cost Estimate Does Not Fully Meet Leading Practices, Which Limits Congress’ Understanding of Costs We found that the F-35 program office’s Block 4 cost estimate did not fully meet the four key characteristics of GAO’s cost estimating leading practices when projecting Block 4 development costs. Table 2 presents key points from our assessment, and appendix II provides additional detail on our rationale. As reflected in table 2, our assessment of the F-35 Block 4 development cost estimate identified a number of missing elements. Specifically, the estimate does not rely on a product-oriented work breakdown structure (WBS), it does not address cost risk and uncertainty, it does not take into account risk related to technology maturity, and it does not have an independent cost estimate, as leading practices reflect. While the program office updates its cost estimate regularly, officials told us that they do not intend to address some of these missing elements in future updates. Work breakdown structure. According to cost estimating leading practices, the program should base its cost estimate on a program- level, product-oriented WBS that allows a program to track cost and schedule by defined deliverables, such as hardware or software components. The WBS ensures that the program does not leave out any portions of the work and makes it easier to compare it to similar systems and programs. According to program officials, the Block 4 cost estimate does not rely on a single WBS; rather, multiple, contractor-derived WBSs exist for the program. Without its own, program-office-level WBS, the program lacks a framework to develop a schedule and cost plan that it can use to track progress and accomplishments. Risk and uncertainty analyses. The program did not perform cost risk and uncertainty analyses. Program officials said they do not plan to conduct a formal risk analysis. The program office works jointly with the contractor to identify and manage risks for the F-35 program. For example, there are monthly Joint Risk Management Boards attended by both program office and contractor leadership. However, overall program risk management is different from quantitative cost risk and uncertainty analyses in that program risk management is not specific to costs and it is not used to assess the cost variance of the cost estimate itself. When planning for funding decisions for a program of this scale, analyzing program-level risks alone is inadequate. Without a risk analysis, the cost estimate will not be fully accurate or credible because it will not account for the effects of potential schedule slips or other risks that the program could realize. Technology maturity. A program office official stated that in developing the cost estimate they did not consider that technologies would not be mature, but rather assumed that most technologies needed to deliver each Block 4 capability would be mature before the program begins development for that capability. The official stated that the complexity of design, development, and testing based on the baseline program experience was reflected in the estimate, but the cost estimate did not identify if there were specific costs associated with maturing these technologies. The official further noted that Block 4 costs would increase if a capability takes longer than planned to design, integrate, and test due to its immaturity. In 2019, we recommended that the Secretary of Defense ensure that the F-35 program office completes an independent technology readiness assessment, as part of its business case for the initial Block 4 capabilities, before initiating additional development work. DOD did not concur with our recommendation. According to a program official, as of December 2019, the program office had not completed any technology readiness assessments even though the contractor has started development of over half of the capabilities within Block 4. Going forward, the program is considering holding incremental technology readiness assessments as it plans for and develops a new set of capabilities, in accordance with the C2D2 schedule. Program officials told us that, going forward, as they update the Block 4 cost estimate, they will consider the results of future technology readiness assessments. Until the program office does so, management cannot determine a reasonable level of additional resources that might be necessary to cover increased costs resulting from unexpected design complexity, incomplete requirements, technology uncertainty, and other uncertainties. Independent cost estimate. In 2019, we also recommended the F- 35 program office include an independent cost estimate as part of its business case for Block 4. As noted in table 2, the Block 4 effort still lacks an independent cost estimate. The program is planning for the Office of the Secretary of Defense, Cost Assessment and Program Evaluation to have a draft independent cost estimate for an interim program review scheduled for March 2020 and to have a complete independent cost estimate in June 2020. This estimate will evaluate the entire F-35 program, including Block 4. With these pieces currently missing, the Block 4 development cost estimate does not present a full picture of Block 4’s cost. Ultimately, without a complete understanding of Block 4 costs, the program could face additional cost growth, which will be hard to track without a complete cost baseline. The lack of a complete cost baseline hinders insight and oversight into the program’s costs, plans, and progress to date and going forward. Moreover, if a cost estimate does not fully or substantially meet all four characteristics of cost estimating leading practices, it cannot be considered reliable. Conclusions DOD plans for the F-35 to be central to the warfighter prevailing in future conflicts. However, the program has been behind schedule and over cost almost since its inception. DOD is slated to move into full-rate production despite several key challenges in the production of aircraft. We acknowledge that the current F-35 program’s production rates are more commonly associated with programs already in full-rate production. However, the F-35 aircraft in the field have not met standards for reliability and maintainability, indicating that the program is not delivering aircraft at the level of quality expected. Additionally, the program’s concurrent approach and the contractor’s continual changes to the production line indicate that the production line processes are not in control. Leading practices indicate that mature production lines—production lines ready for full-rate production—should meet metrics for consistency. Furthermore, to minimize production risk and potential cost growth, suppliers should routinely meet quality and delivery schedules, although this is not yet true of the F-35 program. Not meeting these leading practices poses risks that DOD and the international partners will not routinely receive the F-35’s they specified and need. The long-standing challenges with receiving parts on time and efforts underway to replace Turkish suppliers of parts for the F-35 compound these production challenges and may raise additional risks. Unless the program office assesses and reports on these manufacturing risks ahead of the milestone C review, Congress may not have key insights into the risks that remain with the program and to the overall effort to deliver F-35s to the warfighter. Since the F-35 program is not managing the Block 4 effort as a separate program with traditional oversight tools, we are particularly concerned as Block 4 efforts proceed through development and testing. Specifically, because of the delays to the program, after 2023, DOD will not be required to provide Congress information on Block 4’s development efforts as the current reporting requirements will end. Furthermore, the program’s cost estimate, as presented in its report to Congress, does not fully present all incurred and future costs for Block 4. Without this information, Congress may not have the insight it needs to assess Block 4 cost and schedule progress as well as to make informed oversight and budgeting decisions. In addition, the Block 4 development cost estimate does not fully meet leading practices, lacking a full reflection of all costs. Specifically, the cost estimate does not have a program office level work breakdown structure, a risk and uncertainty analysis, and consideration of technology readiness. Without a comprehensive and credible cost estimate, DOD and Congress lack a sound basis for informed investment decision making, realistic budget formulation, meaningful progress measurement, proactive course correction when warranted, and program and contractor accountability for results. Matter for Congressional Consideration Congress should consider revising Section 224(d) of the National Defense Authorization Act for Fiscal Year 2017, Pub. L. No. 114-328, to extend DOD’s Block 4 reporting requirement until all Block 4 capabilities are fielded to ensure that Congress is aware of cost and schedule growth beyond 2023. (Matter for Consideration 1) Recommendations for Executive Action We are making the following five recommendations to the Secretary of Defense to direct the Undersecretary of Defense for Acquisition and Sustainment (OUSD (A&S)). The OUSD (A&S) should direct the F-35 program office to provide information that is similar to that which is statutorily required after the milestone C review to Congress ahead of the milestone C review (full-rate production decision). This submission should include an evaluation of the production risks associated with critical production processes that are not in control, reliability and maintainability (R&M) targets that are not met, and supplier readiness—particularly for those replacing Turkish suppliers, along with the steps it is taking to address those risks. (Recommendation 1) The OUSD (A&S) should direct the F-35 program office to establish a Block 4 cost estimate baseline that includes all Block 4 costs, including incurred costs and future costs in its reports to Congress as required by the NDAA for Fiscal Year 2017, so that Congress has a complete understanding of all Block 4 costs and can compare this baseline to future cost estimates and performance. (Recommendation 2) The OUSD (A&S) should direct the F-35 program office to complete a program office level, product-oriented work breakdown structure for the next update to its Block 4 cost estimate to ensure that the estimate meets the comprehensive leading practices. (Recommendation 3) The OUSD (A&S) should direct the F-35 program office to conduct risk and uncertainty analyses for the next update to its Block 4 cost estimate to ensure that the estimate meets the credible leading practices. (Recommendation 4) The OUSD (A&S) should direct the F-35 program office to consider the results of its future technology readiness assessment of all Block 4 technologies and incorporate the cost and schedule risks of developing those technologies in the next update to its Block 4 cost estimate to ensure that the estimate meets the comprehensive leading practices. (Recommendation 5) Agency Comments and Our Evaluation We provided a draft of this report to DOD for review and comment. DOD provided written comments, which we have reproduced in appendix VI. DOD concurred with three of the recommendations related to the Block 4 modernization effort (recommendations 2, 4, and 5 above). While DOD did not concur with the other two recommendations, it outlined planned actions that we believe, if implemented, would meet the intent of our recommendations. DOD also provided technical comments, which we incorporated as appropriate. We will continue to monitor the program and evaluate implementation of these recommendations. DOD officials did not concur with the first recommendation, which, in the draft report, was to evaluate production risks and provide a statutorily required report to Congress ahead of the program’s full-rate production decision. While DOD did not concur with the draft recommendation, it agreed to keep the Congress apprised of these matters in its quarterly briefings to the defense committees. To clarify the actions we intended DOD to take to address our findings, we revised the recommendation to indicate that DOD should provide information to Congress on the production risks we identified in our report, ahead of the milestone C review. If the DOD provides a substantive assessment highlighting these production risks, as well as the steps it will take to mitigate them, during its quarterly briefing to Congress ahead of the milestone C review, it would address the intent of our recommendation. DOD also did not concur with our third recommendation for the F-35 program office to complete a program-level, product-oriented work breakdown structure (WBS) for the next update to its Block 4 cost estimate. DOD noted that its next scheduled update was due in April 2020, after we provided our report for comment. While DOD noted it would be unable to complete a program-level WBS by the April 2020 update, it agreed to evaluate moving to a program-level, product-oriented WBS in 2021. If the F-35 program office utilizes a program-level, product- oriented WBS for this cost estimate update, it would meet the intent of our recommendation. We are sending copies of this report to the appropriate congressional committees; the Secretary of Defense; and the Under Secretary of Defense for Acquisition and Sustainment, the Secretary of the Air Force, the Acting Secretary of the Navy, and the Commandant of the Marine Corps. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-4841 or ludwigsonj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix VII. Appendix II: F-35 Block 4 Development Cost Estimate Analysis To assess the reliability of the F-35 Block 4 development cost estimate, we obtained and reviewed cost estimate documentation such as the Joint Program Office briefing on its May 2019 estimate, the Air System Procurement Playbook—a planning document for Block 4—and its cost estimate models. Additionally, we met with relevant staff in the F-35 program office and the Department of Defense’s Office of Cost Assessment and Program Evaluation. We analyzed this information and determined the extent to which the program office’s practices for developing the F-35 Block 4 development cost estimate were consistent with the leading practices identified in the GAO Cost Estimating and Assessment Guide. These practices have been found to be the basis for reliable cost estimates. We assessed each practice as being one of the following: Met—the agency provided data and documentation that satisfies the entire leading practice criterion. Substantially met—the agency provided data and documentation that satisfies a large portion of the leading practice criterion. Partially met—the agency provided data and documentation that satisfies about half of the leading practice criterion. Minimally met—the agency provided data and documentation that satisfies a small portion of the leading practice criterion. Not met—the agency provided data and documentation that does not satisfy any portion of the leading practice criterion. For our reporting needs, we collapsed GAO’s 18 leading practices into four general characteristics: comprehensive, well-documented, accurate, and credible. The assessment of each characteristic was based on an average of the F-35 program office’s scores for the leading practices included in that category. A second analyst verified the assessment and management reviewed the results. We determined the overall assessment rating by assigning each individual rating a number: Not met = 1, Minimally met = 2, Partially met = 3, Substantially met = 4, and Met = 5. Then, we took the average of the individual assessment ratings to determine the overall rating for each of the four characteristics. The resulting average becomes the Overall Assessment as follows: Not met = 1.0 to 1.4, Minimally met = 1.5 to 2.4, Partially met = 2.5 to 3.4, Substantially met = 3.5 to 4.4, and Met = 4.5 to 5.0. A cost estimate is considered reliable if the overall assessment ratings for each of the four characteristics are substantially or fully met. If any of the characteristics are not met, minimally met, or partially met, then the cost estimate does not fully reflect the characteristics of a high-quality estimate and cannot be considered reliable. See table 4 for a high level summary of each leading practice and the reasons for the overall scoring. Appendix III: Objectives, Scope, and Methodology This report fulfills two mandates. First, the National Defense Authorization Act for fiscal year 2015 included a provision for GAO to review the F-35 acquisition program annually until the program reaches full-rate production. This is the fifth report under that provision. Second, the National Defense Authorization Act for Fiscal Year 2020 includes a provision for GAO to review the program’s production and Block 4 progress annually through 2025. In this report, we (1) provide information on the program’s progress toward completing operational testing and resolving deficiencies found in testing; (2) assess the program’s production performance and manufacturing efficiency initiatives; and (3) assess the program’s modernization cost estimate and progress with Block 4 development efforts. To provide information on the program’s progress in operational testing and the resolution of deficiencies, we first reviewed the baseline program’s costs, schedule, and performance plans and compared the actual progress in each area with the goals established in its 2012 baseline to identify any significant trends. We reviewed progress on test events completed versus those that remain, test schedules, program briefings, and DOD briefings. We traveled to Edwards Air Force base to interview DOD test authorities and met with officials from the program office, DOD test authorities, and the contractor Lockheed Martin (the prime aircraft contractor), to discuss key aspects of F-35 development progress, including flight testing, future test plans, and recent findings from test events. Specifically, we obtained updates on key events that are required to complete testing according to the program office’s current schedule. We also interviewed the Director, Operational Test and Evaluation office and F-35 program developmental and operational test pilots. To provide information on the program’s progress resolving deficiencies, we interviewed the same officials mentioned above and discussed how the number of open and closed deficiencies changed in 2019. We reviewed program and contractor information on deficiency reports, mitigations, resolutions, and the deficiency resolution process. To assess the program’s production performance and manufacturing efficiency initiatives, we obtained and analyzed the production metrics from Lockheed Martin and Pratt & Whitney (the prime engine contractor) and their aircraft and engine delivery rates from 2012 through 2019. We reviewed metrics and briefings provided by the program office, Lockheed Martin, Pratt & Whitney, and the Defense Contract Management Agency to identify progress in improving manufacturing processes. We analyzed delivery dates for lot 11 aircraft delivered in 2019. We traveled to the production facility in Fort Worth, Texas to discuss reasons for any delivery delays and plans for improvements with officials from Lockheed Martin. We obtained cost investment and savings estimates and discussed cost and manufacturing efficiency initiatives, such as the economic order quantity purchases, with the contractors and program office officials to understand potential cost savings and plans. We collected and analyzed the extent to which the program has met leading practices identified by GAO for full-rate production. We also obtained and analyzed metrics on parts and aircraft quality through December 2019 and discussed steps taken to improve quality and deliveries with Lockheed Martin and Pratt & Whitney officials. We interviewed officials from Lockheed Martin, Pratt & Whitney, and Northrop Grumman (a key subcontractor) regarding the administration’s decision to suspend Turkey from the program and the implications of the suspension for the contractors. We determined that the contractors’ production metrics and delivery dates were sufficiently reliable for our purposes of determining production efficiency and deliveries. We collected and analyzed production and supply chain performance data from the program office, Lockheed Martin, and Pratt & Whitney. To assess the reliability of the May 2019 Block 4 development cost estimate, we evaluated documentation supporting the estimate, such as the cost estimating models, the F-35’s Air System Procurement Playbook, its updated acquisition strategy, the Decision Memorandum requirements document, and briefings provided to the DOD decision authority. We assessed the cost estimating methodologies, assumptions, and results against leading practices for developing a comprehensive, accurate, well- documented, and credible cost estimate, identified in GAO’s Cost Estimating and Assessment Guide. We also interviewed program officials responsible for developing and reviewing the cost estimate to understand their methodology, data, and approach for developing the estimate. We found that the cost estimate was not reliable. To assess progress with Block 4 development efforts, we interviewed DOD and program office officials, and contractor representatives regarding the program’s Block 4 planning, development, testing, and production activities to date. We reviewed other program documentation, such as the F-35’s fiscal year 2020 budget request, to identify costs associated with the Block 4 effort. We compared the program’s accomplishments in 2019 to its plans and identified what capabilities the program office delivered to the fleet. We reviewed the program office’s plans to develop and deliver additional Block 4 capabilities from 2020 through 2025. We conducted this performance audit from June 2019 to May 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix IV: The F-35’s Reliability and Maintainability Metrics The Joint Strike Fighter Operational Requirements Document, which outlines the requirements Department of Defense and the military services agreed the F-35 should meet, defines all eight reliability and maintainability (R&M) metrics. Table 5 shows each F-35 variants’ performance against these metrics’ targets, as of August 2019. Appendix V: Status of Selected F-35 Technical Risks Newly Identified Technical Risks F-35B Thrust Cutback: An F-35B aircraft can experience an unanticipated cutback in thrust during vertical landings (hover). The contractor put hover weight restrictions in place to mitigate the effect and has identified the root cause. The contractor is developing software and hardware fixes. F-35C Nose Landing Gear: During shipboard landings, the F-35C can experience bending stress, which causes cracking of the coating on a part in the nose landing gear. In the short term, this part will be inspected for damage every 400 flying hours. The contractor is also redesigning the part that is cracking and expects to test it between early 2020 to June 2021. F-35B Three Bearing Swivel Module: The module is mounted at the back of the aircraft and allows the thrust from the engine to be vectored from straight aft for conventional flight to straight down for short takeoff and vertical landing operations. In June 2019, an F-35B experienced a warning indicator in its short takeoff mode due to a hardware component. However, according to the contractor, this component should not cause a warning indicator or loss of functionality for the aircraft. The contractor has identified the root cause of the hardware issue and a gap in the software’s logic that led to the warning. As a result, the contractor is making manufacturing changes to the hardware and implementing software changes to address the issue. Technical Risks Identified in Our Previous Reports Canopy Coating Delaminations: The F-35 fleet has experienced over 50 incidents of the canopy transparencies delaminating after less than 100 flight hours since August 2017. This is over 30 more than we reported in 2019. The contractor tested solutions for the delaminations in 2019 and implemented a solution of adding a vent to the canopy’s frame. Since October 2019, the contractor has added a vent to 146 canopies with one subsequent delamination. Helmet Mounted Display: During low-light flights, the Helmet Mounted Display’s technology cannot display pure black images, instead presenting a green glow on the screen, which makes it difficult to see the full resolution of the night vision video feed. The contractor developed a new display to avoid this effect. According to F-35 program officials, they placed an initial order of 62 displays with 35 delivered by December 2019 to support U.S. Marine Corps and Navy F-35C fleet operations. Three F- 35C pilots completed initial day and night testing using the new display in July 2019 on a carrier. The contractor expects to have a fully qualified redesign by August 2021 and will incorporate it into the production of lot 12 aircraft. Appendix VII: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, the following staff members made key contributions to this report: Justin Jaynes and Alissa Czyz (Assistant Directors); Diana Maurer, Desirée E. Cunningham (Analyst-in- Charge), Jillena Roberts, Tim Moss, Rose Brister, Juaná Collymore, Emile Ettedgui, Jennifer Leotta, and Jeff Hubbard, Other staff who contributed include Leslie Ashton, Priyanka Sethi Bansal, Vinayak Balasubramanian, Julia DiPonio, Christine Pecora, Ralph Roffo, Roxanna Sun, Jessica Waselkow, and Alyssa Weir. Related GAO Products Weapon System Sustainment: DOD Needs a Strategy for Re-Designing the F-35’s Central Logistics System. GAO-20-316. Washington, D.C.: March 6, 2020. F-35 Aircraft Sustainment: DOD Faces Challenges in Sustaining a Growing Fleet. GAO-20-234T. Washington, D.C.: November 13, 2019. Defense Acquisitions: Observations on the F-35 and Air Force’s Advanced Battle Management System. GAO-19-456T: Washington, D.C.: May 2, 2019. F-35 Joint Strike Fighter: Action Needed to Improve Reliability and Prepare for Modernization Efforts. GAO-19-341. Washington, D.C.: April 29, 2019. F-35 Aircraft Sustainment: DOD Needs to Address Substantial Supply Chain Challenges. GAO-19-321. Washington, D.C.: April 25, 2019. F-35 Joint Strike Fighter: Development is Nearly Complete, but Deficiencies Found in Testing Need to Be Resolved. GAO-18-321. Washington, D.C.: June 5, 2018. Warfighter Support: DOD Needs to Share F-35 Operational Lessons Across the Military Services. GAO-18-464R. Washington, D.C.: April 25, 2018. F-35 Aircraft Sustainment: DOD Needs to Address Challenges Affecting Readiness and Cost Transparency. GAO-18-75. Washington, D.C.: October 26, 2017. F-35 Joint Strike Fighter: DOD’s Proposed Follow-on Modernization Acquisition Strategy Reflects an Incremental Approach Although Plans Are Not Yet Finalized. GAO-17-690R. Washington, D.C.: August 8, 2017. F-35 Joint Strike Fighter: DOD Needs to Complete Developmental Testing Before Making Significant New Investments. GAO-17-351. Washington, D.C.: April 24, 2017. F-35 Joint Strike Fighter: Continued Oversight Needed as Program Plans to Begin Development of New Capabilities. GAO-16-390. Washington, D.C.: April 14, 2016. F-35 Sustainment: DOD Needs a Plan to Address Risks Related to Its Central Logistics System. GAO-16-439. Washington, D.C.: April 14, 2016. F-35 Joint Strike Fighter: Preliminary Observations on Program Progress. GAO-16-489T. Washington, D.C.: March 23, 2016. F-35 Joint Strike Fighter: Assessment Needed to Address Affordability Challenges. GAO-15-364. Washington, D.C.: April 14, 2015. F-35 Sustainment: Need for Affordable Strategy, Greater Attention to Risks, and Improved Cost Estimates. GAO-14-778. Washington, D.C.: September 23, 2014. F-35 Joint Strike Fighter: Slower Than Expected Progress in Software Testing May Limit Initial Warfighting Capabilities. GAO-14-468T. Washington, D.C.: March 26, 2014. F-35 Joint Strike Fighter: Problems Completing Software Testing May Hinder Delivery of Expected Warfighting Capabilities. GAO-14-322. Washington, D.C.: March 24, 2014. F-35 Joint Strike Fighter: Restructuring Has Improved the Program, but Affordability Challenges and Other Risks Remain. GAO-13-690T. Washington, D.C.: June 19, 2013. F-35 Joint Strike Fighter: Current Outlook Is Improved, but Long-Term Affordability Is a Major Concern. GAO-13-309. Washington, D.C.: March 11, 2013. Fighter Aircraft: Better Cost Estimates Needed for Extending the Service Life of Selected F-16s and F/A-18s. GAO-13-51. Washington, D.C.: November 15, 2012.

Why GAO Did This Study The acquisition cost for the F-35 program increased substantially in 2019, partially due to the program's addition of estimated costs for modernization of hardware and software systems, referred to as its Block 4 efforts. This is the fifth report under the provision that Congress included in statute for GAO to review the F-35 program annually until the program reaches full-rate production. This is also the first report under another provision in statute to review the program's production and Block 4 progress annually through 2024. Among other objectives, this report assesses (1) the program's production performance and (2) the program's modernization cost estimate and development progress. GAO reviewed Department of Defense (DOD) and contractor documentation and interviewed DOD officials and contractor representatives. What GAO Found The F-35 program is at risk of missing its test schedule and not meeting manufacturing leading practices. In 2019, the F-35 program conducted much of its planned operational testing but extended the schedule by 9 months, which delays the program's full-rate production decision to between September 2020 and March 2021. Over that time, the program will continue to deliver aircraft. In addition, while the F-35 program has increased the production rate and negotiated lower aircraft prices, it is not meeting manufacturing leading practices identified by GAO. Specifically, only about 3,000 of the over 10,000 airframe contractor's manufacturing key processes meet predefined design standards for ensuring product quality. Also, the fielded aircraft, over 500 so far, do not meet the program's reliability and maintainability goals. Although the contractor is changing manufacturing processes to address problems and improve efficiency, more remains to be done. Unless the program office evaluates the risks of not meeting these leading practices, the military services and international partners are at risk of not receiving the quality aircraft they purchased. In addition, the July 2019 suspension of Turkey from the F-35 program—due to security concerns after its acquisition of Russian defense equipment—is likely to compound production risks. The program has identified new sources for 1,005 parts produced by Turkish suppliers, but the program is assessing the effect of 15 key parts not currently being produced at the needed production rate. In 2019, estimated development costs to modernize the F-35's hardware and software systems—known as Block 4—increased by over $1.5 billion. The cost increase puts estimated Block 4 development costs at $12.1 billion. However, the cost estimate did not fully adhere to leading practices, such as including all life cycle costs. In addition, while development will continue through 2026, reports on Block 4 that the program submits to Congress are slated to end in 2023. Without continued Block 4 reporting through the development phase, Congress will lack important oversight information. What GAO Recommends Congress should consider extending DOD's reporting requirement for Block 4 modernization beyond 2023. GAO is also making five recommendations to DOD. While DOD did not concur with two of these recommendations—including to evaluate production risks and update its Block 4 cost estimate with a program-level plan, it identified actions that, if implemented, will meet the intent of these recommendations. DOD concurred with GAO's three other recommendations.

gao_GAO-19-554

gao_GAO-19-554_0

Background Promotion of Human Rights Is a U.S. National Security Goal Promoting respect for human rights is a goal of U.S. foreign policy. The United States considers the advancement of human rights when providing security assistance to foreign countries. Providing training on human rights issues and international humanitarian law to foreign security forces can further U.S. credibility and interests. For example, such training could help maintain local populations’ cooperation with U.S. security efforts by curbing potential abuses by partner country forces. Human rights abuses by U.S.–backed forces can damage the local population’s support for the United States’ strategic aims, according to guidance from the U.S. Army. Human Rights Training Is Provided through Multiple Authorities The United States provides military equipment and training, including human rights training, to partner countries through a variety of security cooperation and assistance programs authorized by statutes, some of which are codified within Title 10 and Title 22 of the U.S. Code. Human rights training is incorporated into broader security cooperation and assistance efforts. DOD and State share responsibility for developing policy for, managing, and implementing human rights training. Title 10 programs are generally overseen by DOD. Title 22 programs primarily fall under State. According to DOD and State officials, most Title 22 human rights training is implemented by DOD. DOD integrates human rights concepts into various types of training and assistance, including “train and equip” programs and defense institution building. Train and equip programs provide training, equipment, and small–scale military construction activities intended to build the capacity of partner nations’ military forces. Defense institution building activities are security assistance programs intended to empower partner nation defense institutions to establish or re-orient their policies and structures to make their defense sector more accountable, effective, and responsive to civilian control, among other things. Some of the authorities under which DOD and State provide human rights training to partner countries require such training when security assistance is provided. For example, one of the more recent and significant changes to security assistance legislation was the 2017 NDAA, which enacted a new chapter in Title 10 of the U.S. Code containing authorities related to security cooperation. Among other things, the 2017 NDAA replaced multiple capacity building authorities with a new statute codified at 10 U.S.C. § 333 (Section 333). All Section 333 programs are required to include elements that promote observance of and respect for human rights and fundamental freedoms, rule of law, and the law of armed conflict, as well as respect for civilian control of the military. Prior to the 2017 NDAA, a similar requirement was mandated for security assistance delivered under the Global Train and Equip program (then codified at 10 U.S.C. § 2282), which required that U.S. assistance pursuant to this authority include “elements to promote observance of and respect for human rights and fundamental freedoms and respect for legitimate civilian authority.” Section 333 covers a greater range of security assistance programs—for example, counternarcotics assistance—than did Section 2282. Other authorities include human rights considerations in their authorizing language. For example, in 1976, Congress established the International Military Education and Training (IMET) program codified within Title 22. The program provides education and training to foreign military personnel with the objectives of professionalizing military forces and increasing respect for democratic values and human rights. In 1990, Congress expanded the objectives of the IMET program to include fostering greater understanding of and respect for civilian control of the military, contributing to responsible defense resource management, and improving military justice systems and procedures in accordance with internationally recognized human rights. State and DOD refer to the expanded IMET objectives as Expanded IMET (E-IMET). Table 1 lists key authorities through which DOD and State provide human rights training to foreign security forces. The United States Undertakes Additional Efforts to Further Human Rights Goals In addition to human rights training, U.S. agencies consider human rights records when providing certain assistance. The Foreign Assistance Act of 1961, as amended, prohibits assistance to a unit of a foreign government’s security forces if the Secretary of State has credible information that such unit has committed a gross violation of human rights. DOD–funded training programs are covered by a similar provision. These requirements are commonly referred to as Leahy laws. As we have previously reported, these laws and the corresponding policies developed to enforce and supplement these laws are intended to leverage U.S. assistance to encourage foreign governments to prevent their security forces from committing human rights violations and to hold their forces accountable when violations occur. To address requirements under both the State and DOD Leahy laws, State has established a process for vetting potential recipients of U.S. security assistance training. State’s Bureau of Democracy, Human Rights, and Labor (DRL) is responsible for overseeing this vetting process and for developing human rights vetting policies, among other duties. Human Rights Training Involves Multiple DOD and State Entities and Is Delivered by a Number of Training Providers Multiple DOD and State Entities Are Involved with Human Rights Training DOD incorporates human rights training as part of a wide range of assistance programs that involve a number of DOD entities in different capacities. (See table 2). State incorporates rule of law assistance and human rights training as part of a wide range of assistance programs that involve a number of State entities in different capacities. (See table 3). Numerous Training Providers Deliver Human Rights Training for Foreign Security Forces but a Few Deliver the Majority DOD operates a number of education facilities that provide training to foreign security forces and many include human rights–related material in their curriculum. However, there are a few training providers that deliver the majority of human rights training through courses explicitly focused on such topics as well as in courses and residential programs that include related material. In addition, State provides some human rights training through the International Law Enforcement Academies (ILEA). Defense Institute of International Legal Studies (DIILS): DIILS is housed under DSCA and is DOD’s lead resource for providing legal education and rule of law engagement training to foreign military personnel and civilian defense officials. DIILS delivers its training primarily through either in-residence courses—for which members of foreign security forces attend trainings at the DIILS campus—or through mobile education training that is delivered to foreign military forces overseas. DIILS provides three types of training: (1) core rule of law training in the United States and abroad, (2) defense institution building, and (3) mandated human rights training delivered under Section 333. DIILS is the only institute to provide the mandated human rights training delivered under Section 333. DOD officials said there are no plans for other facilities to be certified to meet these training requirements. Mandated Human Rights Training Provided by DIILS: In response to the increased demand for mandated human rights training, DIILS created a three–tiered training model to deliver mandated human rights training, according to DIILS officials, who also noted that DIILS is in the early stages of applying the model. The three–tiered training model categorizes mandated human rights training according to basic, intermediate, and advanced trainings. Basic training includes a 2-hour block of scripted coursework which is dedicated to general topics covering human rights and is appropriate when providing training to military units who are not dealing with a combat environment, for example. Military officials without legal training or nonattorney civilian personnel—including contractors— may conduct this training. Intermediate and advanced training is typically 8 or 16 hours of training, respectively, and instruction is provided by DIILS staff and other military attorneys. According to DIILS officials, each intermediate or advanced training is intended to be tailored for the recipient military unit based on an assessment of its duties and the lethality of any equipment provided through the security assistance. Western Hemisphere Institute for Security Cooperation (WHINSEC): WHINSEC, also operated by DOD, provides professional education and training, including human rights training, for military and law enforcement personnel from countries in the Western Hemisphere. The Institute’s Center for Human Rights and Democracy promotes human rights education and training through international programs and partnerships. Curriculum developed by the Center includes topics such as the lawful use of lethal force, due process under international human rights law, and violence against women and vulnerable groups. Examples of WHINSEC’s Human Rights Training: To meet its statutory requirement to provide human rights training, WHINSEC provides a mandatory, 10-hour training on human rights for every student. This training covers five objectives: (1) human rights, (2) the rule of law, (3) due process (4) civilian control of the military, and (5) the role of the military in a democratic society. Additionally, WHINSEC students are required to take an ethics course that builds on the material covered in the human rights and democracy classes. WHINSEC also includes human rights–related material in a number of other courses. For example, the Counter Transnational Threats course focuses on threat interdiction activities using simulated exercises and scenarios. WHINSEC officials explained that one such scenario involves students conducting a simulated raid of a drug lab. (See fig. 1). During the exercise, students encounter armed and unarmed criminals, along with civilians. The simulation is intended to create real–world human rights scenarios for students to assess and apply lessons learned from classroom–based human rights training. Center for Civil–Military Relations (CCMR): CCMR is a DOD organization within the Naval Postgraduate School. CCMR was designed to support the goals of E-IMET and strengthen civil–military relationships through a variety of education and training programs. Additionally, CCMR focuses on defense institution building activities provided under DOD’s Title 10 authority. Like DIILS, CCMR delivers in-residence programs and mobile education training. Examples of CCMR’s Human Rights Training: CCMR officials said that human rights–related material is included in many CCMR programs, although it is not always an explicit focus. For example, although the Maritime Security Program does not explicitly focus on human rights, CCMR staff said that human rights–related topics are integrated into various aspects of the program. One of the program’s modules focuses on how to apply the appropriate use of force when enforcing international and maritime law. CCMR staff said they use practical scenarios to prompt discussion among classroom participants on techniques to avoid use of lethal force. Participants might discuss how to respond if a potential suicide vessel is approaching a ship, including the use of barriers or other deterrents to prevent potential terrorist activity without use of lethal force. Additional DOD Training Providers: A number of other DOD facilities provide training to eligible foreign security forces that includes human rights–related material. These facilities include: Regional Centers: DOD operates five regional centers of strategic studies, whose main purpose is to engage senior leaders in partner countries. A common topic taught at Regional Centers includes civil– military relations, which generally contains information related to human rights. Judge Advocate General (JAG) schools: JAG schools train students on the rules of armed conflict and international humanitarian law; international students may attend these schools, according to DOD officials. Service War Colleges: The service war colleges educate representatives of foreign security forces at a general level about U.S. laws and policies. Human rights–related material may be included, although DOD officials acknowledged such material is peripheral to the main mission. Defense Institute of Security Cooperation Studies (DISCS): International partners who are interested in Foreign Military Sales management participate in human rights training at DISCS. According to DOD officials, DISCS trains hundreds of foreign partners each year on military sales. State’s Bureau of International Narcotics and Law Enforcement Affairs (INL): State INL funds human rights–related training that is delivered by ILEAs. The ILEAs are a global network of training centers with a mission to support emerging democracies; help promote U.S. interests through international cooperation; and promote social, political, and economic stability by combating crime. According to State, this mission is met through strengthening the rule of law and stressing respect for human dignity in law enforcement. ILEAs represent a major component of training provided to foreign law enforcement entities, but do not represent all human rights–related law enforcement training supported by State. DOD Does Not Systematically Track the Provision of Human Rights Training for Foreign Security Forces, but DOD and State Have Some Data on Funding DOD Does Not Systematically Track Human Rights Training for Foreign Security Forces DOD was unable to provide aggregate data on the extent of human rights training for foreign security forces. According to agency officials, DOD does not systematically track all human rights training in DOD systems. As a result, DOD officials noted they were unaware of the full scope of the agency’s human rights training. DOD officials said it is challenging to track human rights training because many courses and training activities might include human rights content. DOD training activities are tracked in the Security Cooperation Training Management System (SC-TMS). However, the tracking is focused on the training overall rather than on any one component of the training conducted, such as human rights. For example, a course at a Regional Center might include human rights– related topics in a civil–military relations class but DOD is not able to identify such a course in SC-TMS or elsewhere as one that could be considered human rights training. DOD officials noted that while DOD is not required to track all human rights training, DSCA and DIILS have systems in place intended to track the provision of human rights training mandated by Section 333, as described below. DSCA uses a case management system to track the mandated human rights training that DIILS provides under Section 333. However, limitations in the implementation of this system have led to questions about the completeness of the data. The case management system is used across DOD to track and manage a range of security assistance programs, in addition to DIILS training. The system is designed so that the implementing entity enters information into the case management system about the training or other security assistance programming provided. However, DOD has not designated DIILS as an implementing agency with authority to enter or edit data in the case management system. As a result, for many years DIILS has relied on a different entity to enter human rights training data into the system. DIILS officials said the U.S. Navy’s agent for international education and training acted as the implementing agency and entered data in the system for DIILS. Due to DIILS’ inability to enter data or make changes in the case management system, DIILS officials told us they have been unable to ensure that data on DIILS training are properly entered. In addition, although DSCA is the DOD entity with oversight responsibilities for ensuring that Section 333 human rights training is provided as appropriate, DSCA officials acknowledged that they did not consistently take steps to monitor the accuracy and completeness of data on the DIILS–provided Section 333 human rights training. DSCA officials said that most of the DIILS trainings likely were entered into DOD’s data system because policy and procedures for capturing training records require it, such as the requirements spelled out in DOD’s Security Assistance Management Manual. However, DSCA officials said they do not have assurance that all trainings were entered as a matter of practice because they lack a process to regularly review whether the training data were captured as required. DOD officials said as of fiscal year 2019, DSCA and DIILS are taking steps to enable DIILS to enter human rights training data directly into the case management system as an implementing agency, but this is still an ongoing effort and not yet operational. In addition, as part of broader changes implemented in 2019 related to how DIILS is funded, the Navy agent is no longer entering information into the case management system about training DIILS provides under Section 333. In the meantime, DIILS continues to track the provision of training using an internal spreadsheet, according to officials, and plans to enter training data into the case management system when they get access as an implementing agency. Federal standards for internal control state that management should use quality information and design appropriate types of control activities in the entity’s information systems to achieve objectives and ensure quality external reporting. In the case of human rights training, DOD officials acknowledged that they do not have a process to ensure that information on mandated human rights training is systematically and accurately entered into its tracking systems. Without such a process, DOD is limited in its ability to monitor compliance with the statutory requirement that Section 333 assistance include a human rights training component. DOD Has Some Data on Funding for Mandated Human Rights Training DOD tracks and reports funding for mandated human rights training at a global level, but not by country and program, although DOD is taking steps to do so. DSCA has published periodic reports that include global funding information for Section 333 activities, including the mandated human rights training. In 2016, Congress required the Director of DSCA to publish quarterly monitoring reports on the status of funding allocated for Section 333 activities. DSCA published three quarterly monitoring reports in fiscal year 2018, which identified the amount of unobligated funds, disbursements, and unliquidated obligations for Section 333 activities. According to the monitoring report from the third quarter of fiscal year 2018, year-to-date unobligated funds for human rights training totaled over $2 million dollars, disbursements totaled about $17,000, and unliquidated obligations totaled about $200,000. The funding data for human rights training is generally reported globally in these reports, not by a specific program or country. DOD could not provide the information we requested on funds obligated and disbursed for mandated human rights training, by program and country, for fiscal years 2015 through 2018. DSCA officials said they could not provide these data because it was challenging to pull this type of information from their systems in a usable way. Further, DOD officials noted that their previous accounting system made it challenging to obtain funding data easily. DSCA and DIILS transitioned to a new accounting system in 2017 which, according to DSCA officials, was expected to provide more detailed information on the status of funding for human rights training. However, DOD officials said that the transition to the new accounting system introduced errors in the data and DIILS staff are still working through a learning curve in adopting the new system. Under the new accounting system, DIILS is to enter information using a unique program and task- naming convention. DSCA officials said the new accounting system, when fully implemented, is expected to allow both DSCA and DIILS to track funds according to the specific recipient country and Section 333 security assistance program, which would better enable DOD and others to effectively monitor the status of funds dedicated to these efforts. State Relies on DOD to Track Training for Military Forces and Tracks Some Human Rights Training and Funding Data for Police State officials said they rely on DOD to track funding and information on the Title 22 authorities that DOD implements, including IMET, which State officials said is its most substantial source of human rights–related training for foreign military forces. DOD provided information on the funding for certified E-IMET courses in recent years. However, according to DOD officials, not all E-IMET courses are related to human rights. State INL maintains data on human rights–related training delivered by ILEAs, which is a major component of training provided to foreign law enforcement entities. In September 2018, we reported that while INL collects data for certain types of police training, such as training provided through the ILEA program, they do not have reliable information readily available on police trained through INL–funded projects. We recommended that State develop and implement a process to collect more reliable data on the number of police trained in El Salvador, Guatemala, and Honduras, the geographic focus of that review. State concurred with our recommendation and stated that it is in the process of developing specific indicators related to police training. According to our review of State data on human rights–related training delivered by ILEAs, State supported human rights training for over 5,400 law enforcement personnel from over 100 countries at ILEAs from fiscal years 2015 through 2017. (See fig. 2.) State identified 31 trainings provided by ILEAs that included human rights topics. (See table 4). According to State, the course that received the most funding—Law Enforcement and Leadership Development—is not expressly focused on human rights but is a 6-week long course that includes human rights concepts in different modules. State provided approximately $34.4 million for such training to foreign law enforcement entities at ILEAS from fiscal years 2015 through 2017. (See fig. 3.) DOD and State Have Not Evaluated the Effectiveness of Human Rights Training Neither DOD nor State Has Evaluated the Effectiveness of Human Rights Training Although officials at both agencies identified examples of past monitoring and evaluation (M&E)–related efforts for security assistance programs, DOD and State officials acknowledged that they have not assessed the effectiveness of human rights training for foreign security forces provided as part of such programs. DOD. DOD officials cited student surveys and after-action reports—which are summaries of the training events, training outcomes, challenges encountered, and further actions to be taken that are prepared by course facilitators—as examples of M&E–related efforts: At DIILS, course facilitators use surveys to solicit student feedback on courses and on the relevance of the course materials. They also use after-action reports, which, according to officials, provide continuity and capture lessons learned from human rights training in partner countries for DIILS facilitators who will be traveling to those countries in the future. At CCMR, according to CCMR officials, training facilitators prepare after-action reports for each course that involves human rights content. They also solicit input from the security cooperation officers in the country where the training took place. At the U.S. Africa Command, officials also said that they prepare after-action reports on DIILS–provided mandated human rights training, which they share with DIILS. Officials said these reports often discuss improvements needed with regard to logistics planning for human right training that DOD provides in African countries. State. Examples of related M&E efforts that State has conducted include a multi-year survey of IMET and evaluations of some security assistance programs. For example, State and DOD funded a survey of IMET graduates which DOD entities conducted and covered the period from 2007 through 2014. The multi-year survey measured, among other things, if graduates reported an improved understanding of internationally recognized human rights. DOD Is Developing an Approach for Monitoring and Evaluating Security Assistance Programs, but Has Not Established a Timeline for Assessing Human Rights Training According to DOD officials, DOD is beginning to develop a new M&E approach for DOD’s security assistance programs. However, DOD has not established a timeline for evaluating the effectiveness of human rights training for foreign security forces that is often included as part of such assistance. The 2017 NDAA, enacted in December 2016, requires DOD to conduct assessment, monitoring, and evaluation of its security assistance programs and activities. The steps DOD is taking to implement the 2017 NDAA M&E requirements include: Policy guidance: DOD issued Instruction 5132.14: Assessment, Monitoring, and Evaluation Policy for the Security Cooperation Enterprise in January 2017. The instruction states that M&E will foster accurate and transparent reporting to key stakeholders on the outcomes and sustainability of security cooperation and improve returns on DOD security cooperation investments. The new M&E requirements are intended to include centralized, independent, and rigorous evaluations of significant security cooperation initiatives to examine their relevance, effectiveness, and sustainability, among other things. DOD officials said that they planned to develop additional guidance to meet the mandated M&E requirements for security assistance, which includes human rights training. Security assistance guidelines: Based on new security assistance guidelines, DOD developed templates for documents that combatant commands are required to complete when planning security assistance activities. These templates for initial assessment and initiative design documents (including for rule of law and human rights training) incorporate M&E into design and planning of security assistance programs and activities. Geographic combatant commands are required to submit these documents to DSCA for projects that are developed in fiscal year 2019 and will be implemented beginning in fiscal year 2020. Draft evaluation agenda: In 2018, DOD prepared a draft evaluation agenda which outlines notional timeframes for evaluations. However, DOD officials could not specify when they plan to finalize the agenda, and as of April 2019 could not tell us when DOD planned to begin monitoring and evaluating human rights training for foreign security forces because they have not developed a timeline for doing so. According to DOD officials, DOD is in the initial phase of developing its overall approach to monitoring and evaluating security assistance, of which human rights training is a small part. The 2019 NDAA, enacted in 2018, requires, as a condition for expending 50 percent of DOD operations and maintenance funds made available for Section 333 assistance, that DOD establish a written plan describing, among other things, evaluation activities planned for security assistance programs for fiscal year 2019. In addition, according to the Office of Management and Budget’s monitoring and evaluation guidelines for the federal government entities providing foreign assistance, agencies should establish annual monitoring and evaluation objectives and timetables to plan and manage the process of monitoring, evaluating, analyzing progress, and applying learning toward achieving results. Developing a timeline for implementing its activities to monitor and evaluate the effectiveness of human rights training, which could be done as part of DOD’s monitoring and evaluation of its broader security assistance efforts, would provide greater assurance that DOD will complete M&E requirements. State Does Not Have a Plan with a Timeline for Evaluating Human Rights Training Provided under IMET According to State officials, they have not established a plan, with a clear timeline, for evaluating the effectiveness of human rights training provided as part of IMET. Officials from State’s Bureau of Political–Military Affairs (PM) acknowledged that State’s responsibilities for IMET include M&E of the program. According to these officials, PM is in the initial phase of developing M&E of its security assistance programs, including IMET. They stated that for this reason PM is not currently planning to evaluate human rights training provided under IMET. Although DOD implements IMET, PM has overall responsibility for the program. According to State’s January 2018 Guidance for the Design, Monitoring and Evaluation Policy at the Department of State, it is essential that bureaus and independent offices have comprehensive plans for monitoring and evaluating all their programs and projects, and the plans should include, among other things, an implementation schedule. An M&E plan with a clear timeline for human rights training provided under IMET will better position State and DOD to determine the effectiveness of a significant component of U.S. human rights training for foreign militaries and identify areas for improvement. Additionally, an evaluation of the effectiveness of the human rights training would provide other important stakeholders, including Congress, with evidence to better inform decisions about U.S.–funded human rights training provided under IMET. Such an evaluation could be done as part of State’s broader effort to evaluate IMET. DOD and State Officials and Experts Identified Challenges to Achieving Human Rights Objectives through Training According to DOD and State officials and outside experts we interviewed, there are several challenges to achieving human rights objectives—such as a decrease in human rights violations or promoting greater respect for human rights—through training alone. Such challenges include tailoring training to the partner nation, integrating it into operational training, and a lack of capabilities and accountability systems on the part of partner nations. Agency officials and outside experts we spoke with stated that it can be challenging to tailor human rights training to the partner nation, the unit receiving assistance, and, when appropriate, the type of equipment being provided. DIILS has developed a three–tiered training model to meet the requirements of Section 333, as discussed above, and DIILS officials stated that they work to tailor trainings to the extent possible, including by selecting trainers with experience relevant to the equipment that the U.S. government provides and adding additional training when needed. However, agency officials and experts stated that DIILS, as a small entity, has limited capacity to tailor human rights trainings for specific situations, especially since DIILS must cover certain material to meet the Section 333 requirements. In addition, DIILS’ ability to tailor training is limited because, according to agency officials, mandated human rights training—typically a classroom course—is generally added to a security assistance package for a partner nation once the planning process has been completed. Since the human rights training is not integrated when the security assistance is planned, it is not generally feasible to adjust the training after the fact to address a specific situation in a given partner country, according to DOD officials. DSCA officials acknowledged that most human rights training is not sufficiently tailored to the needs of the recipient countries and that they have not yet fully incorporated human rights training considerations into security assistance planning. These officials said more work remains to be done to ensure that assistance under the Section 333 authority include comprehensive human rights training designed to meet specific partner nation needs. Agency officials and outside experts we interviewed stated that it can be challenging to achieve human rights objectives through human rights training as currently delivered because mandated human rights training is typically delivered as a stand–alone course in a classroom setting, rather than integrated into operational training. Agency officials stated that integrated training can be more effective because it would expose participants to practical skills that could help them comply with human rights concepts and avoid human rights violations during military or law enforcement operations. For example, State officials said that operational training on how to run a checkpoint while respecting human rights principles is likely to be more effective than training slides that outline international treaties on human rights. Agency officials and outside experts also stated that partner nations may lack capabilities and accountability systems. A military justice system might not hold responsible soldiers who commit human rights violations. A partner nation may lack equipment, experienced personnel, and planning for precision targeting to avoid civilian casualties. Further, partner nations may lack the political will to focus on human rights, and poorly–resourced security forces might see human rights as a low priority. Agency officials and outside experts said that without defense institution building that would address some of these broader systemic issues, human rights training may be less likely to have an effect in some countries. Finally, agency officials noted that in some instances, competing priorities necessitate prioritizing U.S. national security interests when providing security assistance, with human rights receiving less emphasis. Conclusions Instilling respect for human rights in our foreign partners is important to achieving U.S. foreign policy goals. Human rights training that DOD and State provide is one means to do so, but DOD and State are unable to provide a comprehensive accounting of the full array of human rights training they support. With the demand for human rights training increasing as a result of Section 333, a process to ensure training information is systematically tracked would provide DOD greater assurance that it is complying with the statutory requirement to provide human rights training as a component of Section 333 assistance. Furthermore, DOD and State are not able to provide stakeholders, including Congress, with an evaluation of the effectiveness of human rights training the agencies support. Without monitoring and evaluation, decision–makers may be unable to identify whether human rights training provided through Section 333, IMET, and other authorities is achieving objectives and whether it could be adjusted for greater effectiveness. Recommendations for Executive Action We are making a total of three recommendations, including two to DOD and one to State. Specifically: The Secretary of Defense should direct the Director of the Defense Security Cooperation Agency to establish processes to ensure that information on the provision of Section 333 mandated human rights training is systematically and accurately entered into its tracking systems. (Recommendation 1) The Secretary of Defense should direct the Under Secretary of Defense for Policy to develop a timeline for implementing its activities to monitor and evaluate the effectiveness of human rights training for foreign security forces. (Recommendation 2) The Secretary of State, in consultation with the Secretary of Defense, should develop a plan with a clear timeline for monitoring and evaluating the effectiveness of human rights training for foreign security forces provided under IMET. (Recommendation 3) Agency Comments and Our Evaluation We provided a draft of this report for review and comment to DOD and State. DOD concurred with the two recommendations directed to the Secretary of Defense and identified actions it plans to take to address them. Regarding the recommendation to monitor and evaluate human rights training, DOD stated that it would do so as part of monitoring and evaluating its broader security assistance efforts. DOD’s written comments are reproduced in appendix II. State disagreed with the recommendation directed to the Secretary of State. State’s written comments are reproduced in appendix III. In its comments, State acknowledged that human rights training is a vital element of IMET programs and agreed with the need to monitor and evaluate the effectiveness of training—including human rights training— delivered through IMET. However, the department stated that it did not agree to separately conduct monitoring and evaluation of human rights training for IMET participants. Our recommendation for State to develop a plan with a timeline to evaluate the effectiveness of human rights training provided under IMET does not call for a separate evaluation. State could meet the intent of our recommendation through evaluating the effectiveness of human rights training as part of its broader efforts to monitor and evaluate IMET. We added a statement to the report to that effect. We are sending copies of this report to the appropriate congressional committees, the Secretaries of Defense and State, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have questions about this report, please contact Jennifer Grover at 202-512-7141 or groverj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. Appendix I: Objectives, Scope, and Methodology This report (1) describes the entities through which the Department of Defense (DOD) and the Department of State (State) provide training for foreign security forces on human rights and international humanitarian law; (2) assesses the extent to which DOD and State track the provision of and funding for the training; and (3) examines the extent to which DOD and State have evaluated the effectiveness of the training; and (4) provides DOD, State, and outside expert views on human rights training. To address these objectives, we reviewed laws, guidance, budget documents, course catalogs, and agency data. We also interviewed agency officials in Washington, D.C., and at DOD geographic combatant commands. In addition, we conducted site visits at three facilities that provide human rights training: the Center for Civil–Military Relations (CCMR) in Monterey, California; the Defense Institute of International Legal Studies (DIILS) in Newport, Rhode Island; and the Western Hemisphere Institute for Security Cooperation (WHINSEC) in Fort Benning, Georgia. To address the structures through which DOD and State provide training for foreign security forces on human rights and international humanitarian law, we also reviewed course catalogs and interviewed DOD officials from several DOD entities, including the Defense Security Cooperation Agency; the Office of the Undersecretary of Defense for Policy; U.S. Africa Command; U.S. Indo-Pacific Command; and CCMR, DIILS, and WHINSEC. At State, we interviewed officials from the Bureaus of Political–Military Affairs; Democracy, Human Rights, and Labor; and International Narcotics and Law Enforcement Affairs; and the Office of Foreign Assistance Resources. To address what is known about tracking and funding for the training, including whether and how DOD comprehensively tracks human rights training, we reviewed DOD guidance and interviewed DOD officials and training providers. With the 2017 National Defense Authorization Act (NDAA) consolidating authorities—codified at 10 U.S.C. § 333—and the resulting increase in demand for the human rights training DIILS provides under that authority, we then focused on the ways in which that training and its funding is tracked in DOD systems. We reviewed agency documents, including congressional notifications and quarterly monitoring reports, to review how the training data are reported. We also reviewed federal internal control standards to determine what responsibilities agencies have related to information collection and communication. To assess the extent to which DOD and State have evaluated the effectiveness of the training, we reviewed monitoring and evaluation (M&E) policy and guidance documents and other relevant documents. We interviewed DOD and State officials about their current and planned actions to monitor and evaluate human rights training as well as examples of M&E-related efforts for security assistance programs that include human rights training. We also reviewed legislation, including the 2017 and 2019 NDAAs, which outline M&E requirements for DOD’s security assistance. In addition, we reviewed State’s January 2018 Guidance for the Design, Monitoring and Evaluation Policy at the Department of State to determine M&E requirements for State. To collect information on DOD, State, and outside expert perspectives of human rights training provided to foreign security forces, we conducted individual semistructured interviews with selected stakeholders, including agency officials and outside experts, who consisted of former government officials and representatives of nongovernmental organizations and think tanks. To identify outside experts, we asked stakeholders, including current government officials, to recommend other stakeholders we should speak with (i.e., snowball sampling). In our interviews, we collected information on perspectives of factors that could potentially enhance the effectiveness of human rights training and challenges to achieving human rights objectives through such training. The information we obtained from these stakeholders cannot be generalized across all stakeholders. We conducted this performance audit from February 2018 to August 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the Department of Defense Appendix III: Comments from the Department of State Appendix IV: GAO Contacts and Staff Acknowledgments GAO Contact Jennifer A. Grover, 202-512-7141 or groverj@gao.gov. Staff Acknowledgments In addition to the contacts named above, Juan Gobel, Christina Werth, Emily Desai, Sada Aksartova, James McCully, David Payne, Neil Doherty, John Hussey, Mark Dowling, and Rachel Stoiko contributed to this report.

Why GAO Did This Study The U.S. government seeks to advance human rights when it provides security assistance to foreign countries. Such assistance includes DOD– and State–supported human rights and international humanitarian law training for foreign security forces. The NDAA for Fiscal Year 2017 consolidated multiple capacity building authorities, now codified at 10 U.S.C. § 333. DOD implements most U.S. human rights training for foreign security forces. Congress included a provision in the NDAA for Fiscal Year 2018 for GAO to review human rights training for foreign security forces. This report, among other objectives, (1) describes the entities through which DOD and State provide such training, (2) assesses the extent to which DOD and State track the provision of and funding for such training, and (3) examines the extent to which DOD and State have evaluated the effectiveness of the training. GAO reviewed laws, regulations, guidance, agency training and funding data, and course catalogs, and interviewed agency officials. What GAO Found Several entities within the Departments of Defense (DOD) and State (State) are involved in human rights training. DOD's Defense Security Cooperation Agency (DSCA) conducts program management for DOD's efforts to build the capacity of foreign security forces. The human rights training required by 10 U.S.C § 333 is provided exclusively by the Defense Institute of International Legal Studies (DIILS), a DOD entity. DOD operates a number of other educational entities that provide training to foreign security forces, and many include human rights–related material in their curriculum or through operational exercises. (See figure.) DOD does not systematically track human rights training and, as a result, only limited information is available on the provision of and funding for these activities. Without a process to ensure systematic and accurate tracking of human rights training data, DSCA is limited in its ability to monitor its compliance with the training–related provision of the National Defense Authorization Act (NDAA) for Fiscal Year 2017. State relies on DOD to track human rights training for military forces and tracks some training and funding data for police. DOD and State have not assessed the effectiveness of human rights training for foreign security forces, according to agency officials. The NDAA for Fiscal Year 2017 required DOD to conduct monitoring and evaluation of its security assistance programs. DOD has taken initial steps to develop monitoring and evaluation policies but officials stated that they have not yet determined when DOD will evaluate human rights training. State officials said they do not know when the agency will begin monitoring and evaluating human rights training provided under the International Military Education and Training program, a large source of funding for such training. Monitoring and evaluation would enable DOD and State to determine the effectiveness of U.S.–provided human rights training for foreign security forces. What GAO Recommends GAO is making three recommendations, including that the Secretary of Defense establish a process to systematically track mandated human rights training and develop a timeline for implementing monitoring and evaluation. DOD agreed. GAO also recommends that the Secretary of State develop a plan with a timeline to monitor and evaluate such training. State disagreed. GAO continues to believe the recommendation is valid as discussed in the report.

gao_GAO-20-40

gao_GAO-20-40_0

Background Opioid Use Disorder Opioids—such as hydrocodone, oxycodone, morphine, and methadone— can be prescribed to treat both acute and chronic pain. Many opioids have a high potential for abuse and may lead to severe psychological or physical dependence. OUD, which is a type of substance use disorder, is generally characterized by a loss of control of opioid use, risky opioid use, impaired social functioning, tolerance, and withdrawal. According to SAMHSA and the National Institute on Drug Abuse, OUD is a chronic, treatable illness. SAMHSA states that treatment for OUD should be individualized and can include a range of treatment options that include medication and behavioral health services. Specifically, services related to the treatment of OUD include the following: Screening can identify individuals who have OUD, are at risk for developing OUD, or have medical problems related to opioid use. Screening, brief intervention, and referral to treatment (SBIRT) is a specific type of screening that involves a health care provider educating individuals with a positive screen for opioid use and referring them to specialized treatment, as needed. Outpatient counseling and therapy includes counseling and treatment services individually or in a group. Inpatient hospital services include those that occur in a hospital, such as services for detoxification. Inpatient residential services include care in a 24-hour residential setting. Inpatient residential providers offer medical care in combination with housing, typically lasting from a week to several weeks or more. Medication-assisted treatment (MAT) combines the use of certain prescription medications (such as methadone, buprenorphine, and naltrexone) and behavioral therapy. Methadone and buprenorphine suppress withdrawal symptoms and control the craving for opioids, while naltrexone suppresses the euphoric effects of opioids. Research has shown that MAT for OUD reduces opioid use and increases the chance that patients will continue OUD treatment compared to abstinence-based treatment (where patients are treated without medication). Case management services include providing coordination and management of care across multiple health care providers. Crisis intervention includes immediate care intended to prevent harm. Peer recovery coaching includes recovery support through a certified peer specialist with experience of recovery from addiction. Medicaid Program Overview States administer their Medicaid programs within broad federal requirements and according to a state plan approved by CMS. The Medicaid program allows states to design and implement their programs within certain federal parameters, resulting in more than 50 distinct programs. A state’s approved Medicaid plan outlines the services provided and the groups of individuals covered. States also have the option of using waivers to expand services under the Medicaid program. As such, the types of services covered by Medicaid can vary across states. Historically, Medicaid eligibility has been limited to certain categories of low-income individuals, such as children, parents, pregnant women, persons with disabilities, and individuals aged 65 and older. The Patient Protection and Affordable Care Act (PPACA), enacted in 2010, allowed states to expand Medicaid coverage to nearly all individuals with incomes up to 138 percent of the FPL, regardless of eligibility category. As of October 2019, 33 states and the District of Columbia expanded Medicaid eligibility, and 17 states had not. Medicaid Services for Pregnant and Postpartum Women Under federal law, state Medicaid programs must provide coverage for health care services for certain pregnant women, including low-income pregnant women with incomes at or below 138 percent FPL. Most states opt to extend coverage to pregnant women with incomes above this threshold. By statute, states are permitted to limit the services covered for certain pregnant women, including low-income pregnant women, to services related to the pregnancy. Such coverage is referred to as “pregnancy-related coverage.” CMS defines pregnancy-related services as those services necessary for the health of the pregnant woman and fetus, or that have become necessary as a result of the woman having been pregnant, which includes prenatal, delivery, postpartum, and family planning services, as well as services for other conditions that may complicate the pregnancy. In contrast, states are required by statute to provide pregnant women who qualify for Medicaid on another basis, such as a disability, full Medicaid benefits. At a minimum, states must provide Medicaid coverage for pregnant enrollees through 60 days postpartum, though states may extend coverage further. Some women may qualify for continued Medicaid coverage after the 60-day postpartum period if they meet the requirements for another eligibility pathway, such as for parents, while others may transition to other programs or become uninsured. Medicaid Services for Children Medicaid’s Early and Periodic Screening, Diagnostic, and Treatment (EPSDT) benefit is the primary mechanism to help ensure the provision of appropriate health care services to children under Medicaid. Under the EPSDT benefit, states are required to cover comprehensive health screenings and preventive health services, such as those related to vision and oral health, and all other Medicaid coverable services that are necessary to correct or ameliorate any conditions discovered through screenings. States are required to follow a schedule of screenings, known as a periodicity schedule, that are recommended for children at specific ages and frequencies. States can develop their own schedule within federal parameters or follow an established schedule, such as from the American Academy of Pediatrics (AAP). States are required to report annually on the provision of certain EPSDT services to CMS. States must report information on the number of children provided health screening services, the total number of health screenings services provided, the number of children referred for corrective treatment, the number of children receiving dental services, and the state’s results in attaining EPSDT participation goals. States may also voluntarily report annually on the quality of health care services provided under EPSDT using a set of quality measures known as the Child Core Set. CMS plans to increasingly use the Child Core Set in the future and state reporting will be mandatory beginning with the state reports for fiscal year 2024. States also have flexibility in determining where EPSDT services can be delivered. As a means of improving access—particularly in underserved communities, such as rural areas—Medicaid programs may cover certain services delivered by health care providers in schools. There were approximately 130,000 schools across the United States as of 2016, according to data from the Department of Education. Some of these schools have health clinics. Additionally, there were approximately 2,600 school-based health centers as of 2017, some of which served children in underserved communities, according to the School-Based Health Alliance. Telehealth as a Modality to Provide Services Some state Medicaid programs allow for services to be delivered via telehealth, including in schools. Telehealth can be used to provide clinical care remotely, such as for screening, counseling, and therapy. Health care providers offer care to patients through remote technology, such as a live, two-way video call. Such services could be provided, for example, via a video conference on a desktop computer or laptop that connects a student in school with a provider in another location. States have flexibility to choose whether to cover services delivered via telehealth. Because the federal Medicaid statute does not recognize telehealth as a distinct service, CMS views telehealth as a service delivery mechanism. According to CMS, services delivered via telehealth are subject to the same Medicaid requirements as those services provided in person. Selected States Provide Medicaid and Other Coverage for a Range of Opioid Use Disorder Services for Eligible Pregnant and Postpartum Women Selected States Provide Medicaid Coverage of Most Opioid Use Disorder Services for Eligible Pregnant Women and Women through 60 Days Postpartum The Medicaid programs in our six selected states provided coverage of most OUD services during pregnancy and the first 60 days postpartum, as of January 1, 2019. Specifically, the selected states’ Medicaid manuals indicate these states provide coverage for at least seven of the eight OUD services we identified in our review, such as screenings, inpatient and outpatient services, and MAT. (See fig. 1.) In the six states we reviewed, we did not identify differences between the types of covered OUD services for pregnant women in expansion and non-expansion states. For example, the three expansion states— Arkansas, Colorado, and Massachusetts—and the three non-expansion states—Alabama, South Dakota, and Texas—each covered at least seven of the eight OUD services for pregnant women. Similarly, we did not identify differences in Medicaid coverage for OUD services between the selected states that limited coverage to pregnancy- related services and those that provided full benefits. Arkansas and South Dakota—the two states providing only pregnancy-related coverage— covered seven of the eight OUD services for pregnant women. According to Medicaid officials in these two states, the programs cover these OUD services because they are considered medically necessary. In contrast, neither state generally provides Medicaid coverage for peer recovery coaching for beneficiaries, including pregnant women. Three of the selected states—Alabama, Colorado, and Texas—covered certain OUD services for pregnant women that they do not cover for other beneficiaries under Medicaid. For example, in Alabama, screening services and inpatient residential services were covered only for pregnant women, but not other eligible, low-income women. In Colorado, pregnant women were the only group in the state for whom residential OUD services were covered under Medicaid. In Texas, pregnant women were the only group in the state for whom case management is a covered service under Medicaid. Selected States Provide Medicaid Coverage for Most Opioid Use Disorder Services If Women Are Able to Maintain Medicaid Eligibility beyond 60 Days Postpartum In all six selected states, once Medicaid coverage furnished on the basis of pregnancy ends after 60 days postpartum, women can continue to receive most OUD services under Medicaid if they qualify for Medicaid on another basis. For example, these women could qualify if their income is equal to or lower than the maximum allowable income for parents. However, in the six selected states, the maximum allowable income for eligible parents was generally lower than that for pregnant women, as of January 2019. (See fig. 2.) Women in the six selected states who are eligible to maintain Medicaid coverage after 60 days postpartum can continue most of the same OUD services that were covered during pregnancy. However, officials in two states said that the OUD services covered specifically for pregnant women under Medicaid would generally not be covered after the postpartum period ends. Four of the selected states provided estimates of the number of women who maintain Medicaid eligibility after the postpartum period ends. For example, officials in Massachusetts, an expansion state, estimated that in 2017 and 2018, approximately 99 percent of women with Medicaid coverage while pregnant maintained Medicaid coverage after the postpartum period ended. State officials in Colorado, also an expansion state, estimated that in 2015, 75 percent of women maintained coverage after the postpartum period ended. Additionally, in Arkansas, another expansion state, officials estimated that about 60 percent of women in 2017 and 2018 maintained Medicaid coverage after the postpartum period ended. Officials in Alabama, a non-expansion state, estimated that in 2017, about 43 percent of women maintained Medicaid coverage after the postpartum period ended. States may also obtain approval from CMS, such as under a waiver, to extend Medicaid coverage for women with OUD beyond 60 days postpartum, according to CMS officials. However, CMS officials were not aware of the number of states that have done so. In our review, we found that one of the six selected states, Colorado, used a section 1915(b) waiver to extend Medicaid eligibility for substance use services, including OUD services, for women beyond 60 days postpartum. Under the state’s “Special Connections” program, which was approved under the waiver, women who are eligible for Medicaid during their pregnancy can continue Medicaid coverage for OUD services for up to 12 months postpartum, including inpatient residential services, which would not otherwise be covered under Medicaid. A state official told us that the program began in 1991 to provide substance use disorder services to pregnant women and up to 60 days postpartum. In 2006, the state extended coverage under the program to provide substance use disorder services up to 12 months postpartum. The program aids in early identification and intervention for pregnant women with substance use disorders who are at risk of delivering low birth weight babies with health complications. Officials said the goal of the program is to improve the likelihood that the mother remains free from substance abuse. According to Colorado officials, 227 women participated in the program in 2018. Selected States Use Other Funding Sources to Provide Coverage of Opioid Use Disorder Services for Postpartum Women Not Eligible for Medicaid We also found that the six selected states use other funding sources to provide coverage of OUD services for women with incomes that exceed the state’s Medicaid eligibility thresholds. Officials from the six selected states reported that they received SAMHSA grants so each state could provide OUD services for pregnant and postpartum women that extend beyond 60 days. According to the SAMHSA officials we interviewed, grants have been used to increase access to MAT, expand recovery support for pregnant women, and provide enhanced services for women to access OUD treatment. Furthermore, SAMHSA officials said that pregnant and postpartum women are specifically identified as target populations for grants, such as the agency’s State Targeted Response to the Opioid Crisis grant and its State Opioid Response grant. For example, officials in Arkansas told us that by using SAMHSA grants, they are able to allow uninsured or underinsured women who are seeking treatment for OUD to continue MAT treatment after 60 days postpartum. In addition, officials in the six selected states told us that women beyond 60 days postpartum generally would not experience gaps in treatment for OUD when transitioning from Medicaid to SAMHSA grant-funded programs, as women can generally continue receiving the same services and seeing the same providers. For example, state officials in Alabama, South Dakota, and Texas told us the state Medicaid agency contracts with providers that agree to participate in both the state’s Medicaid program and SAMHSA’s grant programs to allow for continuity of eligible services. Officials in Texas also told us they used state funds to implement a program to provide OUD services for up to 18 months postpartum. State officials told us that in this program—called the Neonatal Abstinence Syndrome-Opioid Treatment Services program—when a woman’s Medicaid coverage ends, she transitions to state-funded treatment to continue the same OUD services with the same provider. According to state officials, this program, funded since 2015, expands treatment services to postpartum women who would typically lose Medicaid coverage and become unable to pay for MAT services, which officials say help reduce relapse, overdose, and maternal mortality risk. State officials added that there is flexibility to extend services for postpartum women participating in the program for up to 2 years, if needed. State officials told us that since 2016, 296 women have participated in the program. In addition to the efforts in the selected states, the federal government has planned efforts to help states combat the opioid crisis, specifically for pregnant and postpartum women with OUD. For example, CMS plans to offer up to 12 cooperative agreements to states under the Maternal Opioid Misuse model as a way to improve access to services under Medicaid to pregnant and postpartum women with OUD. The model will have a 5-year performance period, 2020 through 2024, to allow states to implement strategies to improve the quality of care for pregnant and postpartum women with OUD. According to CMS officials, the model does not require that states extend coverage beyond 60 days postpartum, but states could choose to do so. CMS published the funding opportunity for the model in February 2019, and plans to select states to participate by the fall of 2019. In July 2019, CMS issued guidance to states regarding Medicaid coverage of services such as counseling for postpartum women while their infant is receiving treatment for Neonatal Abstinence Syndrome. The Centers for Disease Control and Prevention also issued a publication in September 2019 summarizing an initiative on state strategies to address OUD among pregnant and postpartum women and infants prenatally exposed to substances. The initiative identified five focus areas, including access to and coordination of quality services, provider awareness and training, and financing and coverage. In addition, the SUPPORT Act includes a provision for HHS to issue guidance to improve care for postpartum women with substance use disorder by the fall of 2019. Selected States Provide Medicaid Coverage for Annual Screenings and Any Medically Necessary Services for Substance Use, Including Opioids, for Eligible Children The six selected states provide Medicaid coverage for annual screenings of eligible children for substance use, including opioids, as well as any medically necessary treatment for conditions identified through these screenings, as of 2019. This coverage is provided through the Medicaid EPSDT benefit. Based on our review of Medicaid state plans and EPSDT policies and periodicity schedules, we found that the six selected states established the following screening schedules at the time of our review: Arkansas’ Medicaid state plan and Colorado’s and South Dakota’s EPSDT policies specify that these states follow AAP’s screening schedule. AAP recommends annual substance use screening for all children beginning at age 11 until they reach the age of 21. Alabama’s EPSDT policy requires annual screening for all children ages 6 to 13. Massachusetts’ EPSDT policy requires providers to conduct an annual assessment of every child’s risk for substance use as part of a health history assessment during a child’s annual visit. This assessment can be conducted at any age. Texas’ EPSDT periodicity schedule recommends annual screening for all children ages 12 to 18. Additionally, Massachusetts and Texas Medicaid programs require behavioral health screening for all eligible children. The Medicaid programs in these states provide separate payment for behavioral health screening if the screening is conducted using an approved screening tool, some of which also screen for substance use, such as opioids. Texas officials reported that substance use screenings are considered part of the required overall mental health screening component of annual checkups. Similarly, Arkansas officials said that as part of a new EPSDT policy they are drafting, the state Medicaid program will require behavioral health screening for all children, which can include substance use screening if determined medically necessary by the provider. The six selected state Medicaid programs report data on the total number of screenings provided under EPSDT for children’s health care needs. However, officials from five of the selected states said that it is difficult or impossible to separate and thus track the number of the substance use screenings as distinct from other types of EPSDT screenings or visits that are recorded in Medicaid data. States are required to track and report the total number of EPSDT screenings provided, but not the number of substance use disorder screenings. Officials from all six selected states said that they conduct outreach and education to providers and parents to ensure awareness of the EPSDT benefits, as required. We found that the extent of information the states provide on these services varied among the six states. For example, outreach materials from three of the six selected states included information about the availability of substance use screening, and one of these three states, Alabama, also included information about services for opioid use. For all six selected states, officials emphasized that Medicaid’s EPSDT benefit requires states to cover any medically necessary treatment or service to address health conditions for a child, including opioid use. Officials from the six selected states also described a variety of initiatives to increase access to substance use disorder, including OUD, services for children. For example: Officials from Alabama said they recently began a program that offers more substance use disorder services in schools in a face-to-face capacity to help increase convenience and reduce stigma around these services for both the children and the parents. They explained that the Alabama Department of Mental Health added modifiers to ensure that their systems can capture data appropriately and analyze trends in providing school-based services, which are currently offered in over 40 individual schools. Officials added that Medicaid pays for covered services that are provided to Medicaid-eligible children in schools under this program. Officials from two states—Arkansas and Massachusetts—said they recently expanded the types of substance use disorder services covered for all Medicaid beneficiaries. Officials from Arkansas added that they are working to expand the number of providers who can offer substance use disorder treatment under Medicaid. Officials from two states—Massachusetts and Texas—said they had recently developed programs specifically aimed at serving children with substance use disorder, including OUD, using federal authority. Massachusetts received approval from CMS to conduct a Medicaid demonstration to establish OUD programs for children. Texas is using SAMHSA grant funding to support eight youth recovery centers that are intended to improve services and recovery supports for youth with substance use disorder, including OUD. Over 30 States Reported Medicaid Coverage of Opioid Use Disorder Services Delivered through Telehealth in Schools, but No Evidence Services are Being Used We conducted outreach to Medicaid officials from all 50 states and the District of Columbia between February and July 2019 to inquire about whether the state provided Medicaid coverage of OUD services delivered via telehealth in schools as a means of increasing access to these services for children. Officials from 31 states and the District of Columbia reported that they provide Medicaid coverage of OUD services delivered in schools via telehealth. Medicaid officials from some states reported that their Medicaid policies explicitly allow for coverage of OUD services provided in schools via telehealth, while others reported that they allow for Medicaid coverage of these services, but their policies do not explicitly address the issue. (See app. I for the state responses regarding Medicaid coverage of OUD services provided in schools via telehealth.) Officials from the remaining 18 states reported that their Medicaid policies do not allow for payment for OUD services delivered in schools via telehealth. Some of these state officials reported that they did not allow schools to serve as a location for patient services during a telehealth visit for Medicaid payment purposes. Other state officials reported that they allowed Medicaid payment for certain services provided in schools via telehealth, but OUD services were not among them. For example, officials from Texas said the state established a school-based telehealth program for behavioral health services. However, this program does not include services for the treatment of substance use disorder, including OUD. Officials added that the state has a requirement that substance use disorder services can only be delivered in certain approved facilities. While Medicaid officials from 31 states and the District of Columbia reported that they provide Medicaid coverage of OUD services in schools via telehealth, they also said they were not aware of any instances of these services being utilized. Medicaid officials from the 31 states and the District of Columbia reported that either these services were not being provided based on data or other information they had, or they were unaware if the services were being provided. Officials from seven states responded that they either reviewed Medicaid utilization data or asked school-based staff and determined that there was no utilization of these services. For example, officials from one state—Ohio—conducted a data query and found that in 2018 there had been two instances of substance use disorder services billed to Medicaid that were delivered via telehealth in a school. However, officials noted that these instances involved treatment for substances other than opioids. While not for OUD, these two instances were the only instances of Medicaid payment for substance use disorder services delivered via telehealth in schools that we identified in our review. As part of our outreach to states and background research, we did find that some states or localities have taken steps to facilitate the use of telehealth for delivery of substance use disorder services, including OUD, in schools. For example: Officials from one county in Maryland said they recently began using a telehealth smart device application to screen students in schools for substance use disorder, including OUD, and refer them to treatment. However, county officials said that the program was locally funded, and they had not considered seeking Medicaid payment. A South Dakota tribal reservation recently implemented a new school- based telehealth program for behavioral health. According to officials, this program could include OUD services delivered via telehealth in schools, and these services could be billed to Medicaid if the provider was already licensed to bill Medicaid; however, officials said that none of these services had been provided to date. Massachusetts recently expanded its Medicaid telehealth policy to allow for the payment of mental health and substance use disorder services provided in many locations, including schools. However, officials said the state was still building the telehealth infrastructure, and services had not yet been provided at the time of our review. Officials were unsure whether OUD services would be delivered via telehealth in schools under the new policy once implementation began. State officials and subject matter experts that we spoke with also reported a range of potential benefits and challenges associated with providing substance use disorder services, including OUD services, in schools via telehealth. (See table 1.) There have also been federal efforts to emphasize the use of telehealth to improve access to OUD services for children. For example, these efforts include the following: In June 2018, CMS issued guidance emphasizing the use of telehealth as a means of improving access to OUD services and noted that states need not necessarily submit a change to their state plan to begin delivering covered Medicaid services through telehealth. Similar to what we heard from experts, this guidance suggests that leveraging technology to provide such services might help with addressing provider shortages, particularly in rural areas. In July 2019, CMS and SAMHSA jointly issued guidance on addressing mental health and substance issues in schools. The guidance states that telehealth for mental health services in schools has been found to be effective. This guidance also emphasizes that telehealth can be helpful for ensuring that Medicaid services are provided to Medicaid beneficiaries who are in rural areas or in areas where qualified practitioners are scarce. HRSA officials we spoke with also described several different HRSA programs from which funds could be used to facilitate or deliver substance use disorder services via telehealth, including in some school-based health centers; however, the officials were not able to determine whether telehealth is being utilized to deliver OUD services in school-based health centers, specifically. The SUPPORT Act also includes a provision for HHS to issue guidance to states on Medicaid coverage of substance use disorder services delivered via telehealth, including in school-based health centers, by fall of 2019. Agency Comments We provided a draft of this report to HHS for review. HHS provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the Secretary of Health and Human Services, the Administrator of the Centers for Medicare & Medicaid Services, the Health Research and Services Administration, the Substance Abuse and Mental Health Services Administration, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact Jessica Farb at (202) 512-7114 or farbj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Major contributors to this report are listed in appendix II. Appendix I: State Reported Medicaid Coverage of Opioid Use Disorder Services Delivered through Telehealth in Schools Appendix II: GAO Contact and Staff Acknowledgements GAO Contact Staff Acknowledgments In addition to the contact named above, Rashmi Agarwal (Assistant Director), Kaitlin McConnell (Analyst-in-Charge), Arushi Kumar, Kimberly Lloyd Perrault, Jennifer Rudisill, and Emily J. Weisenberger made key contributions to this report. Also contributing were Drew Long and Ethiene Salgado-Rodriguez. Related GAO Products Medicaid: Additional CMS Data and Oversight Needed to Help Ensure Children Receive Recommended Screenings. GAO-19-481. Washington, D.C.: August 16, 2019. Opioid Crisis: Status of Public Health Emergency Authorities. GAO-18- 685R. Washington, D.C.: September 28, 2018. Adolescent and Young Adult Substance Use: Federal Grants for Prevention, Treatment, and Recovery Services and for Research. GAO- 18-606. Washington, D.C.: September 4, 2018. Newborn Health: Federal Action Needed to Address Neonatal Abstinence Syndrome. GAO-18-32. Washington, D.C.: October 4, 2017. Drug Control Policy: Information on Status of Federal Efforts and Key Issues for Preventing Illicit Drug Use. GAO-17-766T. Washington, D.C.: July 26, 2017. Medicaid Expansion: Behavioral Health Treatment Use in Selected States in 2014. GAO-17-529. Washington, D.C.: June 22, 2017. Health Care: Telehealth and Remote Patient Monitoring Use in Medicare and Selected Federal Programs. GAO-17-365. Washington, D.C.: April 14, 2017. School Based Health Centers: Available Information on Federal Funding. GAO-11-18R. Washington, D.C.: October 8, 2010.

Why GAO Did This Study The misuse of prescription opioid pain relievers and illicit opioids, such as heroin, has contributed to increases in OUD and overdose deaths. Pregnant women with OUD have an increased risk of overdose during the postpartum period. Opioids also caused over half of drug overdose deaths among youth in 2017. Medicaid plays a key role in covering services to treat OUD for low-income women and children. The SUPPORT Act includes a provision for GAO to study Medicaid coverage for pregnant and postpartum women with a substance use disorder, including OUD. The act also includes a provision for GAO to examine children's access to these services, such as through telehealth. This report describes Medicaid coverage of OUD services for (1) pregnant and postpartum women in selected states; (2) children in selected states; and (3) children delivered via telehealth in schools across all states, and utilization of these services. GAO reviewed documentation and interviewed officials from federal agencies within HHS to understand Medicaid coverage of OUD services for pregnant and postpartum women, as well as children. GAO also interviewed officials and reviewed documentation from six states selected for variation in opioid use rates, status of Medicaid expansion, and geographic variation, among other things. GAO also conducted outreach and received responses from 49 of 50 states and the District of Columbia about Medicaid coverage and use of OUD services delivered via telehealth in schools. HHS provided technical comments, which were incorporated as appropriate. What GAO Found All state Medicaid programs are required to provide coverage of health care services to pregnant women with incomes at or below 138 percent of the federal poverty level through 60 days postpartum. With regard to opioid use disorder (OUD), GAO found that six selected state Medicaid programs provide coverage of a range of services for eligible pregnant women with OUD. Specifically, the six states—Alabama, Arkansas, Colorado, Massachusetts, South Dakota, and Texas—covered OUD services, such as screening for opioid use, counseling, and medication-assisted treatment, which combines the use of medications with counseling. In the six selected states, women who are eligible for Medicaid coverage after 60 days postpartum can receive most of the same OUD services that were covered during pregnancy. Furthermore, GAO found that the six selected states also use other sources of funding, such as federal grants, to provide coverage of OUD services for postpartum women who are not eligible for Medicaid. GAO did not review how frequently the OUD services were actually provided to pregnant and postpartum women. GAO found that the state Medicaid programs in all six selected states cover annual screenings for substance use, which includes opioid use, for eligible children. This coverage is provided as part of Medicaid's Early and Periodic Screening, Diagnostic, and Treatment benefit, under which all states are required to cover certain screenings for eligible children under age 21. GAO also found that Medicaid programs in 31 states and the District of Columbia covered OUD services, including screenings, delivered through telehealth in schools. However, state Medicaid officials said they were not aware of any instances of these services being utilized through telehealth in schools. Telehealth can be used to provide clinical care remotely, such as for screening, counseling, and therapy. Such services could be provided, for example, via a video conference on a desktop computer or laptop that connects a student in school with a provider in another location. State officials and experts cited both benefits and challenges with providing OUD services through telehealth in schools. For example, benefits included addressing provider shortages, particularly in rural areas, as well as reducing the amount of time students spend outside of the classroom accessing services. Challenges included lack of needed infrastructure and provider discomfort with using telehealth. Agencies within the Department of Health and Human Services (HHS) have recently issued guidance emphasizing the use of telehealth for OUD services, particularly in schools.

gao_GAO-20-258

gao_GAO-20-258_0

Background When DOT issued regulations requiring accessible lavatories on twin- aisle aircraft in 1990, single-aisle aircraft were used primarily for shorter distances. However, technological advancements—such as the use of lighter, stronger composite materials—have enabled aircraft to fly longer distances with greater fuel efficiency. As a result, both Airbus and Boeing now offer single-aisle aircraft designs that can routinely fly 3,000 to almost 4,000 nautical miles—or easily from coast to coast in the continental U.S. as well as some overseas routes to and from the U.S. Of the eight U.S. carriers we interviewed for this review, five of them operate only single-aisle aircraft and the remaining three fly a mixture of single- and twin-aisle aircraft. As shown in figure 1, twin-aisle aircraft are rarely used for domestic flights by U.S. carriers. In 2018, 99 percent of U.S. aircraft departures for domestic flights occurred on single-aisle aircraft: 71.7 percent on the larger single-aisle aircraft; 26.61 percent on smaller, regional aircraft; and 0.68 percent on turbo jets. According to the 2010 U.S. Census, 57 million Americans (roughly 1 in 5) have a disability, of whom half have mobility issues that may require accommodations when flying. Furthermore, older Americans are representing a greater share of the U.S. population. By 2020, 16.5 percent of the nation’s population will be over age 65, and by 2030, 20 percent of the population will be over age 65, and the likelihood of this group’ needing assistance to access lavatories may increase as they age. As an indication of the number of people with reduced mobility flying, DOT’s monthly aviation travel data indicate that for the first 6 months of 2019, aviation passengers checked a total of 294,216 wheelchairs and scooters. Over the past 30 years, some efforts have been made to address the issue of ensuring non-discriminatory treatment of aircraft passengers, including access to aircraft lavatories. The enactment of the ACAA, which prohibits discrimination by airlines on the basis of disability, charged DOT with promulgating implementing regulations. DOT promulgated a final rule in 1990, in which it required aircraft with more than one aisle (twin-aisle aircraft) in which lavatories are provided to have at least one wheelchair accessible lavatory. DOT, however, deferred setting regulations for single-aisle aircraft, noting cost and feasibility concerns for carriers. Since that initial regulation, DOT has taken several steps to study the issue of accessible lavatories for single-aisle aircraft, but as of December 2019, none of these actions has resulted in a regulation for accessible lavatories in single-aisle aircraft. These steps include: DOT issued an advance notice of proposed rulemaking (ANPRM) to study, among other things, the issue of accessible lavatories on single-aisle aircraft that was issued in conjunction with its final rule mandating that twin-aisle aircraft must have a lavatory that is accessible to passengers who use wheelchairs. DOT created an advisory committee in 1992 to provide guidance to DOT concerning access to lavatories on single-aisle aircraft for persons with disabilities, including persons who rely on the aircraft’s onboard wheelchairs. In 1996, the committee reported to DOT that it would be feasible to provide accessible lavatories on single-aisle aircraft but acknowledged that there could be a cost to doing so. As part of a final rule that DOT issued in 2008 to amend the ACAA regulations to include foreign carriers that fly to the U.S., DOT acknowledged that requiring accessible lavatories on single-aisle aircraft was an ongoing issue. While the department noted that accessible lavatories on single-aisle aircraft would benefit passengers with disabilities, it also expressed concerns that revenue loss and other cost impacts could be too great for the carriers. The department said that it would continue to study the issue and review ongoing developments. DOT published a notice of intent in December 2015 to explore the feasibility of conducting a negotiated rulemaking concerning, among other things, accessible lavatories on single-aisle aircraft for travelers with disabilities. As a result of the 2015 notice of intent, DOT established the ACCESS Advisory Committee in 2016—composed of representatives from air carriers, aircraft manufacturers, disability groups, and other aviation stakeholders. Its charge was to negotiate and develop a proposed amendment to DOT regulations for DOT’s consideration concerning accommodations for air travelers with disabilities that would address whether to require accessible lavatories on new single-aisle aircraft, among other issues. The committee noted that the issue of requiring accessible lavatories on single-aisle aircraft merited exploration because of two developments: (1) the increased use of single-aisle aircraft on long flights, and (2) the availability of new accessible- lavatory designs for single-aisle aircraft. In late 2016, the ACCESS Advisory Committee agreed on proposed amendments that included short-term and long-term solutions to address the challenges persons with mobility impairments face when traveling on single-aisle aircraft. The committee, taking into account costs to industry, recommended accessible lavatories on new aircraft, did not recommend requiring the retrofit of existing aircraft, and proposed a multi-tiered approach to meet this goal. In 2019, DOT publicized its intent to issue notices of proposed rulemaking regarding accessible lavatories to address the ACCESS Advisory Committee’s final resolution, which we discuss in more detail below. Manufacturers of Single-Aisle Aircraft Offer Lavatories Designed to Better Accommodate Persons with Reduced Mobility, but Selected Carriers Rarely Choose to Acquire Them Aircraft Manufacturers Offer a Range of Lavatory Designs for Single-Aisle Aircraft, Including Designs to Accommodate Onboard Wheelchairs Both Airbus and Boeing offer their customers a range of standard lavatory designs. For example, both Airbus and Boeing offer a lavatory with a contoured design (see fig. 2). This design offers a smaller sink and different dimensions than previous lavatories and has a contoured or angled wall on the exterior allowing seats in the last row to recline into the bottom portion of the contour. Air carriers can also choose to move the last row of seats back into the contour and then add an extra row of seats after making other changes to the configuration of seating rows. Airbus and Boeing also offer flat-wall lavatory designs that are similar to the standard flat-wall lavatories that had previously been available on single- aisle aircraft for years but have slightly different interior dimensions. Compared to the contoured lavatory design, this current flat-wall lavatory design could offer a larger sink or more countertop space. According to measurements and diagrams the aircraft and lavatory manufacturers provided for these lavatory styles, some interior lavatory dimensions have decreased while other dimensions have increased. For example, changes in these two lavatory styles have resulted in increased interior space in some areas, such as the sitting knee space and diagonal shoulder width, and decreased space in other areas, such as the entry width and door height. These changes were to provide carriers options to help them meet their business strategies. In addition to making changes to the standard lavatories, since 2015, both Airbus and Boeing offered lavatory configurations for their single-aisle aircraft designed to provide greater access for passengers who rely on the use of onboard wheelchairs. According to officials for Airbus and Boeing, both manufacturers use a design that connects two adjacent lavatories with a retractable wall or partition. As shown in figure 3, when the folding partition is open, this configuration is designed to enable the person who relies on the aircraft-onboard-wheelchair to enter in one of the lavatories and then transfer or be transferred to the toilet in the other lavatory. While there are differences between the Boeing and Airbus models, they operate similarly. Both the Airbus and Boeing designs are for the rear of the aircraft and take up space in the area normally used for the galley where food and drink carts are located for flight attendants’ access. According to officials we interviewed from two carriers that have purchased aircraft with this design, a reduced galley area is less of a concern because their flights provide limited food and beverage service and do not need a full galley. They said that the space where the traditional lavatories were located could be used for other purposes, such as more seats. According to the manufacturers, the lavatory models that are designed for greater accessibility accommodate the onboard wheelchair to varying degrees. Airbus offers two designs to accommodate a passenger with an onboard wheelchair. The Space Flex version 1 design consists of two adjacent lavatories with a connecting retractable partition. This retractable partition can open to allow for a passenger who relies on the aircraft onboard-wheelchair to enter the lavatory with or without the help of an assistant. A representative from a disability organization was generally positive about the Space Flex version 1 and said it was a good design for both carriers and travelers with disabilities. Airbus also offers another lavatory design specifically for its A220 single-aisle aircraft model. Airbus officials told us that it is a single lavatory that is designed to accommodate a wheelchair but cannot accommodate both a passenger in an onboard wheelchair and an assistant. Boeing offers one lavatory designed to accommodate a passenger using an onboard wheelchair for single-aisle aircraft for its 737 aircraft family. This design, known as the Pax Plus, consists of two adjacent lavatories with a removable partition designed to enable a wheelchair and assistant to enter. In addition, officials from the eight selected air carriers told us that their crews are trained to assist passengers with reduced mobility to use lavatories. These officials from the eight air carriers stated that they provide their cabin crew with initial and, in some cases, recurrent training about how to assist passengers with reduced mobility, pursuant to DOT regulations. DOT regulations further stipulate that if there is an on-board wheelchair, the carrier must provide assistance to enable the passenger to move to and from the lavatory if, in general, such assistance is requested by or on behalf of a passenger with a disability. Onboard Wheelchair- Accommodating Lavatories Are Not a Common Feature on Selected U.S. Carriers’ Aircraft While aircraft manufacturers offer lavatories designed to accommodate passengers with mobility impairments, carriers do not often choose to acquire them. Of the eight U.S. carriers we interviewed, we found that four have some aircraft—all of which are Airbus aircraft—with lavatories that are designed to accommodate passengers with mobility impairments to some extent. Only one of these carriers is among the four with the largest number of aircraft in their fleet. Specifically, these four carriers have either the Space Flex version 1 or the Airbus A220 lavatory. Despite Boeing’s offering of the Pax Plus lavatories since 2017, Boeing officials told us that as of November 2019 no U.S. carriers have ordered these lavatories for their current or future single-aisle Boeing aircraft. Overall, about 4.5 percent of the combined single-aisle fleet of the eight selected carriers have lavatories designed to provide some measure of greater access to passengers with reduced mobility, including those who require the use of the onboard wheelchair (see fig.4). According to the carriers we interviewed, they consider the configuration of the aircraft among other factors, including their business strategy, when ordering lavatories for new aircraft. Providing a lavatory designed to accommodate onboard wheelchairs on single aisle aircraft may require financial tradeoffs for carriers, such as reducing the number of revenue generating seats in the aircraft cabin. According to airline officials, this reduction can result in higher costs for carriers that subsequently might be passed onto consumers through higher fares. Officials from all eight selected carriers, however, stated that all of their aircraft lavatories have features designed to increase access to certain lavatory functions, such as assist handles or grab bars, accessible call buttons, door locks, and faucets that passengers with disabilities can use. Carrier officials also stated that they need to make trade-offs between competing priorities; for example, taking into account how onboard wheelchair-accommodating lavatory designs may affect food service. According to officials from two carriers, an onboard wheelchair- accommodating lavatory can result in less galley space, and a full galley at the back of the aircraft is needed for the type of services they wish to provide to their customers without compromising customer seating capacity. Conversely, officials from two other carriers told us that trading galley space for onboard wheelchair-accommodating lavatories did not affect their food service, as they do not provide full meal service. For example, they said that because they did not need the full galley space, the Space Flex lavatory enabled them to add not only a lavatory that accommodates onboard wheelchairs but also an additional row of passenger seats. Lack of Onboard Wheelchair- Accommodating Lavatories Can Make Flying Difficult for Persons with Reduced Mobility, and DOT Is Considering Rulemaking for Accessible Lavatories on Single-Aisle Aircraft While Complaint Data Are Limited, Groups Advocating for Persons with Reduced Mobility Stated That the Lack of Accessible Lavatories Makes Flying Difficult Representatives from stakeholder groups we interviewed told us that the lack of accessible lavatories makes flying challenging for persons with reduced mobility. They described how some passengers with reduced mobility take precautionary measures to avoid the need to use an aircraft lavatory, such as severely limiting food and fluid intake in advance of the flight, risking dehydration; using a catheter; or wearing a protective undergarment. Some passengers with reduced mobility reportedly may avoid long flights altogether by purchasing flights with connections or layovers. However, according to one stakeholder group, these precautionary measures may not alleviate the fear and anxiety that passengers who rely on the onboard wheelchair to get to the lavatory may face during air travel as there is always the possibility of having to deal with circumstances beyond their control. For example, unforeseen events such as increased flight time or delays in getting to the gate can increase the time a passenger has to postpone attending to normal bodily functions. Finally, stakeholder groups report that passengers may choose not to travel at all, or to drive rather than fly, choices that may increase the cost and time of travel, particularly if it involves an overnight stay. Even when an aircraft has a lavatory that can accommodate an onboard wheelchair, which exists on about 4.5 percent of the combined fleet of single-aisle aircraft for the 8 airlines included in our review, passengers may have difficulty determining whether or not their flight has such a lavatory. According to officials of air carriers, passengers may call the carriers’ customer service department for this information, although not all phone representatives may have this information readily available. In addition, our review of selected carriers’ websites revealed that most do not have information about which flights or aircraft may have such a lavatory, although we found that two carriers include descriptions of aircraft amenities or diagrams denoting onboard wheelchair- accommodating lavatories. However, even if this information were made available, it may not guarantee that a passenger with a mobility impairment will be able to fly on an aircraft with this type of lavatory because air carriers sometimes switch aircraft at the last minute without notice, such as when, for example, an aircraft has a mechanical problem. While stakeholders described challenges, neither air carriers nor DOT receive a large number of complaints regarding the lack of lavatories designed to accommodate passengers who use onboard wheelchairs or lavatories in general. As we have previously reported, DOT receives and processes complaints from passengers and uses complaint data to help identify which carriers to inspect for consumer protection violations. From 2014 through 2018, DOT received 59,846 complaints about U.S. carriers. Of these, we reviewed 1,263 complaints related to accessibility, inadequate facilities, and flight delays and identified 69 complaints about lavatories in general and 5 about the accessibility of lavatories. Of the 69 lavatory complaints identified: 64 related to non-functioning lavatories (e.g., non-operational or unclean lavatories, sinks lacking running water, etc.); 5 related to lavatories being inaccessible by persons with disabilities (e.g., lavatory grab bars at an improper height, passenger using onboard wheelchair unable to enter lavatory); and 2 related to lavatory size (e.g., lavatory size has been reduced). We also discussed lavatory-related complaints with the eight selected air carriers, three of which reported that these complaints made up about 1 percent or less of the total passenger complaints they received in 2018. Four air carriers reported that lavatory complaints related to accessibility made up an even smaller portion—around 0.05 percent or less of their total passenger complaints in 2018. However, the small number of complaints related to lavatory accessibility does not necessarily indicate that individuals who use onboard wheelchairs are not affected by inaccessible aircraft lavatories, as some may choose not to fly, and others may take precautionary measures as described above to avoid having to use the aircraft lavatory. Furthermore, because accessible lavatories are not required on single-aisle aircraft and there is no expectation that the lavatory would be accessible, passengers may not see grounds to complain or may not take the time to submit a complaint. As we have previously reported, complaint data are inherently limited because, according to academic literature, a substantial portion of dissatisfied individuals never complain and are therefore not represented in the complaint data. Finally, when they do complain, their complaints may not be representative of other individuals. We also found that there were very few complaints about non-functioning lavatories. As noted above, DOT received 64 passenger complaints on non-functioning lavatories. Carrier officials also told us that they have received few complaints about non-functioning lavatories. According to the air carrier officials we interviewed, depending on the flight, some flights may operate with one or more lavatories not functioning. However, most carrier officials reported that according to data they collect, this occurred on less than 2 percent of flights. In such instances, some carrier officials stated they would notify passengers of nonoperational lavatories to give them the opportunity to use the airport lavatories prior to boarding. These officials also stated that if all lavatories are inoperable it is the responsibility of the pilot—in consultation with flight dispatchers—to decide if the aircraft will take off or, if lavatories become inoperable during a flight, to divert to an airport other than the destination. Carriers further noted that flights with no operational lavatories are extremely rare. DOT Has Introduced Rulemaking and Has Other Efforts Under Way to Help Address Difficulties Faced by Air Travelers with Disabilities As previously noted, in late 2016, the ACCESS Advisory Committee reached a consensus on proposed amendments that would require accessible lavatories on single-aisle aircraft. DOT announced in 2019 that it would address the issue in rulemaking. On December 16, 2019, DOT issued a notice of proposed rulemaking to solicit comments on short-term accessibility improvements on single-aisle aircraft through the installation of accessibility features within the lavatory, such as assist handles, call buttons, and lavatory controls, without changing the size of lavatories. In addition, DOT has announced its intention to issue an advanced notice of proposed rulemaking to address long-term accessibility improvements, also addressed by the Advisory Committee, and to solicit comments and gather information on the costs and benefits of requiring carriers to increase the size of the single-aisle lavatory on new aircraft models to enable passengers using an onboard wheelchair to enter and use the lavatory with an assistant, if necessary. In addition to the two rulemakings, DOT has recently established another advisory committee. The ACAA Advisory Committee was created in response to a requirement in the FAA Reauthorization Act of 2018, has a 2-year charter, and is required to report its findings to both DOT and Congress on current DOT regulations on barriers to persons with disabilities who want to travel by air. The ACAA Advisory Committee is also required to determine the extent to which DOT is addressing those barriers, recommend improvements to implement the ACAA, and improve the flying experience for travelers with disabilities. The committee— comprised of members representing aircraft manufacturers, national disability organizations, air carriers, and airports—plans to hold its first meeting in early 2020. According to DOT officials, although it is within the purview of this committee to consider issues regarding accessible lavatories, it does not plan to do so at the present time given that the two proposed rulemakings are proceeding and that there is a Congressional mandate for the committee to report on other issues within 6 months of the first meeting. Agency Comments We provided a draft of this report to DOT for review and comment. DOT provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, DOT, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have questions concerning this report, please contact me at (202) 512-2834, or vonaha@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix I. Appendix I: GAO Contact & Acknowledgements GAO Contact Staff Acknowledgments In addition to the individual named above, other key contributors to this report were Ed Laughlin, Assistant Director; Martha Chow, Analyst-in- Charge; James Geibel; Geoffrey Hamilton; Elke Kolodinski; Gail Marnik; Malika Rice; Amy Rosewarne; Travis Schwartz; Pamela Snedden; and Elizabeth Wood.

Why GAO Did This Study Flying can pose significant challenges for persons who rely on wheelchairs, including the lack of wheelchair accessible lavatories on most flights. In 1990, DOT required wheelchair accessible lavatories on twin-aisle aircraft used mainly for long flights. It did not require them for single-aisle aircraft, although DOT continued to study the issue. Since 1990, technological advances have enabled single-aisle aircraft to fly longer distances, and these aircraft now make 99 percent of domestic flights. In 2016, a DOT advisory committee recommended that DOT require accessible lavatories in certain single-aisle aircraft in the future. The Federal Aviation Administration (FAA) Reauthorization Act of 2018 included a provision that GAO examine the availability and designs of lavatories on commercial aircraft and the ability of passengers with disabilities to access them. This report describes (1) what is known about lavatory designs and accessibility for persons with reduced mobility and (2) the challenges wheelchair-bound passengers and others face while traveling on single-aisle aircraft without accessible or functional lavatories. GAO reviewed DOT's guidance and rulemaking and analyzed DOT's aircraft complaint data and fleet data for the eight largest U.S. air carriers. GAO interviewed officials from the eight largest mainline carriers and reviewed their fleet and lavatory data. GAO also interviewed officials from Airbus and Boeing and subsidiary lavatory manufacturers, as well as representatives from cabin crew labor associations and consumer groups representing persons with disabilities. What GAO Found Aircraft manufacturers offer lavatories that carriers can provide and that are designed to accommodate users of onboard wheelchairs, but carriers do not choose to acquire this option for their single-aisle aircraft. We found designs for lavatories that enable a passenger in an onboard wheelchair to use them, to varying degrees. In recent years, both Airbus and Boeing—makers of single-aisle aircraft—began offering similarly designed lavatories to provide greater access for these passengers. For example, one design consists of two adjacent lavatories located in the rear galley area with a connecting retractable wall to allow for a wheelchair-bound passenger to enter one lavatory and transfer or be transferred to the toilet in the other lavatory. Another design is a single lavatory large enough to accommodate a passenger using an onboard wheelchair. Four of the eight U.S. carriers—and only one of the four with the largest fleets—GAO interviewed have Airbus aircraft with an adjacent lavatory design (Space Flex version 1) or the single lavatory design found on the A220 aircraft, constituting about 4.5 percent of the carriers' combined single-aisle fleet (see figure). None of the eight U.S. carriers have purchased a similar lavatory for their Boeing's single-aisle aircraft. Carrier officials told GAO that they consider many factors when ordering lavatories, including financial and service tradeoffs such as the potential to lose seating spaces, or reduced food and beverage service for passengers. While the Department of Transportation (DOT) receives few complaints on lavatory inaccessibility, consumer groups told GAO that the lack of an accessible lavatory on single-aisle aircraft presents challenges for persons with reduced mobility. For example, some passengers take precautionary measures to avoid the need to use the aircraft lavatory and others avoid flying altogether. Additionally, although some aircraft have wheelchair-accommodating lavatories, they are not well advertised to passengers, making it difficult for passengers to know whether their flight may have such a lavatory. To address such challenges and the findings of its 2016 advisory committee, DOT issued, on December 16, 2019, a notice of proposed rulemaking to require carriers to install accessibility features without changing the size of the lavatories. DOT also expressed intent to study the costs and benefits of enlarging single-aisle aircraft lavatories to enable use by passengers using the onboard wheelchair.

gao_GAO-20-86

gao_GAO-20-86_0

Background Surface Ship Assignments to U.S. and Overseas Homeports The Navy bases the majority of its surface ships at homeports in the United States, and five regional maintenance centers manage their maintenance. At the time of our review, the Navy had 38 surface ships assigned to overseas homeports, as illustrated in figure 1. A homeport is where a ship is based and primarily managed and maintained. The Navy assigns all newly commissioned ships entering the fleet to a U.S. homeport, and the Navy may change a ship’s homeport throughout its service life. The Navy may move a ship to an overseas homeport to respond to strategic needs or to relieve another ship returning to a U.S. homeport. We found in May 2015 that basing ships at overseas homeports provides considerable additional time in strategic areas of operation and other benefits ranging from increased opportunities for collaboration with partners and allies to faster response time for emerging crises. However, we also found that the Navy’s high pace of operations for its overseas-homeported ships affected the material condition of these ships, and that they had experienced a worsening trend in overall ship readiness when compared to U.S.- homeported ships over the preceding 5 years. We also reported that the Navy generally intended ships to be homeported overseas for about 7 to 10 years, according to officials, but that some ships in Japan had been based there for longer than 10 years. In 2018 Congress instituted a 10- year cap on the length of time certain U.S. Navy ships may be based at overseas homeports. Maintenance Responsibilities for Surface Ships A number of organizations and commands within the Navy share responsibilities for setting maintenance policies and planning, scheduling, and executing ship maintenance, from the offices of the Secretary of the Navy and Chief of Naval Operations, to fleet commanders and ships’ crews. Key organizations include: Type Commanders. The Navy’s type commanders for surface ships— Commander, Naval Surface Force, U.S. Pacific Fleet, and Commander, Naval Surface Force, U.S. Atlantic Fleet—are responsible for maintaining, training, and ensuring the readiness of the surface ships assigned to each fleet. Naval Sea Systems Command (NAVSEA). NAVSEA, among other things, maintains surface ships to meet fleet requirements within defined cost and schedule parameters. These offices perform contract administration, program management, and planning for future maintenance periods informed by the historical maintenance needs of Navy ships. For example, the following NAVSEA organizations have certain responsibilities for overseas ship maintenance: NAVSEA’s Deputy Commander for Ship Maintenance and Modernization (NAVSEA 21). This office provides life-cycle management for surface ships and manages critical modernization, maintenance, training, and inactivation programs. NAVSEA’s Surface Maintenance Engineering Planning Program (SURFMEPP). SURFMEPP provides life-cycle management of maintenance requirements for surface ships, including providing centralized class maintenance and modernization planning and management of maintenance strategies. Commander, Navy Regional Maintenance Center (CNRMC). This office oversees the regional maintenance centers in the United States, as well as the Forward Deployed Regional Maintenance Center (FDRMC) headquarters in Italy, and its detachments in Rota, Spain, and Manama, Bahrain, that manage the maintenance for the U.S. Navy ships homeported there. NAVSEA’s Logistics, Maintenance, and Industrial Operations (NAVSEA 04). This office manages and oversees the naval shipyards and the Ship Repair Facility and Japan Regional Maintenance Center (SRF-JRMC) in Yokosuka, Japan, and its detachment in Sasebo, Japan. Surface Team One. This body of stakeholders from across the Navy’s surface ship maintenance, modernization, and sustainment organizations collaborates for the purpose of setting and developing surface ship maintenance and modernization priorities, conducting analyses, and improving surface ship maintenance performance. A Senior Flag Oversight Council comprised of Commander, Naval Surface Force Pacific, and Commander, NAVSEA, provides strategic vision and directs Surface Team One’s efforts, which may include knowledge-sharing networks, working groups, or deep-dive studies and business case analyses. Types of Ship Maintenance Periods The level of complexity of ship repair, maintenance, and modernization can affect the length of a maintenance period, which can range from a few weeks to 6 months or longer. The types of maintenance periods include the following: Chief of Naval Operations (CNO) maintenance. CNO maintenance periods are scheduled to accomplish industrial, depot-level maintenance and modernization—work that cannot be conducted by ship’s crews or goes beyond fleet capabilities. These depot-level maintenance periods can last 6 months or longer and the Navy generally schedules them every 2 to 3 years throughout a ship’s service life. This can include major repair, overhaul, or complete rebuilding of systems needed for ships to reach their expected service life, and involve complex structural, mechanical, and electrical repairs. For example, in certain types of depot-level maintenance, ships are taken out of the water and put into a dry dock to perform maintenance on below-water parts of the ship (see fig. 2 for a photo of a dry dock at SRF-JRMC in Yokosuka, Japan). To inform the planning of the work package for this maintenance period, Navy officials or contractor representatives typically perform one or more “ship checks” to assess the material condition of the ship in advance of the maintenance period. Continuous maintenance. Continuous maintenance periods are to conduct maintenance outside of the longer CNO maintenance periods that can be done in short periods typically scheduled to be 2 to 6 weeks in duration. According to Navy officials, the schedules of these periods can vary, and commanders can adjust, postpone, or cancel them based on operational demands. Voyage repair. Voyage repair maintenance periods are solely to accomplish corrective maintenance of a mission- or safety-essential nature necessary for a ship to deploy or continue its deployment. For example, ships based in the United States that are deployed overseas on a temporary basis schedule mid-deployment voyage repair to ensure they can continue their deployment. Planning Process for Surface Ship Depot-level Maintenance Periods The process for planning surface ship depot-level maintenance periods (i.e., CNO maintenance periods), whether the ship is based overseas or in the United States, is contained in the Navy’s Joint Fleet Maintenance Manual. In general, the Navy begins planning for a ship’s depot-level maintenance period 720 days—or roughly 2 years—before the planned start of the maintenance period. During this time, a variety of organizations within the Navy plan what will be repaired, how long it will take, where the work will be done, as well as select the contractors to perform the work, among other things. This process also includes activities to close out the maintenance period once it is complete, which overlap with the start of the planning cycle for the next maintenance period. For example, certain milestones serve both planning and closeout purposes—such as the Life-cycle Planning Conference Meeting, which is to both closeout a ship’s completed maintenance period and to begin planning for the next one by reviewing the maintenance requirements, deferred work, and planned schedules (see figure 3). NAVSEA 21, including SURFMEPP, is generally responsible for the advanced planning of maintenance periods, which includes setting the baseline requirements and early estimates of how long maintenance might take. In general, regional maintenance centers have overall responsibilities for meeting milestones approximately a year prior to the start of maintenance through execution and closeout, as illustrated in figure 3 above. Overseas, the responsible regional maintenance centers are the SRF-JRMC at the homeport in Yokosuka, Japan, and its detachment at the homeport in Sasebo, Japan, and the FDRMC detachments at the homeports in Rota, Spain, and Manama, Bahrain. Naval Supply Systems Command’s Fleet Logistics Centers offices overseas are responsible for soliciting and awarding maintenance contracts, for ships based overseas, among other things. The Navy’s Ship Maintenance Capacity and Approach Varies by Overseas Location The Navy has developed different maintenance capacity and approaches to maintain the 38 surface ships based in Japan, Spain, and Bahrain. The Navy maintains these ships through a mix of Navy, host government, and contractor industrial base facilities and resources that are different at each location. The Navy has tailored the maintenance approaches it uses at each homeport considering the available Navy and contractor capacity, as well as the number and type of ships, according to Navy documents and officials. Table 1 provides an overview of the Navy and contractor industrial base capacity for depot-level maintenance of surface ships based at the four main overseas homeports. U.S. Naval Ship Repair Facility and Japan Regional Maintenance Center (SRF-JRMC), Yokosuka, Japan. The Navy’s largest overseas maintenance facility, SRF-JRMC is located in Yokosuka and is responsible for the maintenance of 12 surface ships homeported there— including the most destroyers and the only cruisers based outside of the United States. According to Navy officials, SRF-JRMC in Yokosuka operates as a public shipyard would in the United States, with three on- base dry docks that fit all sizes of ships based there, as well as other smaller dry docks. SRF-JRMC employs a Japanese workforce that conducts the majority of the maintenance workload through a cost- sharing agreement between the United States and Japan. For example, in fiscal year 2018, SRF-JRMC directly conducted about two-thirds of the total ship maintenance workload, with about one-third conducted by local contractors, according to SRF-JRMC workload reporting documentation. For the contracted work, SRF-JRMC relies on one main contractor, Sumitomo Heavy Industries, for ship maintenance in Yokosuka, though additional smaller contractors are also used. Most contracted work also takes place at Navy facilities on base, according to SRF-JRMC officials. Ships in Yokosuka are able to receive deeper, more complex maintenance than other ships based overseas because of the Navy maintenance capacity at SRF-JRMC, according to NAVSEA officials. SRF-JRMC in Yokosuka also conducts detailed planning for maintenance periods that other regional maintenance centers do not, according to NAVSEA officials. Specifically, it plans all the individual maintenance and repair tasks to be conducted in each maintenance period, while other U.S. and overseas maintenance centers can rely on the contractors to plan the work they do. For additional information on SRF-JRMC in Yokosuka, Japan, see appendix III. SRF-JRMC Detachment, Sasebo, Japan. The Navy also operates its own shipyard with a Japanese workforce at the SRF-JRMC detachment in Sasebo, though it primarily relies on the local contractor base to conduct maintenance work. In fiscal year 2018, the SRF-JRMC detachment directly conducted about one-third of the total maintenance workload, with nearly two-thirds performed by contractors according to SRF-JRMC workload reporting documentation. For the contracted work, the Navy relies on about 14 smaller contractors, and while the SRF- JRMC detachment coordinates the work of the multiple contractors that may contribute to a single maintenance period, the contractors directly plan and manage their portion of the work, according to Navy officials. The SRF-JRMC detachment in Sasebo includes two Navy dry docks, though only one is used for depot-level maintenance periods. As a result, dry-dock maintenance and modernization can be conducted on ships based in Sasebo, but it is generally limited to the four MCM and two LSD ships. The other amphibious ships based in Sasebo receive depot-level maintenance that has been planned from about 2 to as long as nearly 9 months, but this does not include dry-dock maintenance. A unique maintenance consideration in Sasebo is the deployment schedule of the amphibious ships based there. These ships typically deploy three at a time with U.S. Marines based in Okinawa on board. As a result, there are times when all ships are in port and require maintenance, so the detachment tries to stagger the work with the MCMs and closely coordinate with contractors there in an effort to manage workload, according to SRF-JRMC officials. For additional information on the SRF- JRMC detachment in Sasebo, Japan, see appendix IV. Forward Deployed Regional Maintenance Center (FDRMC) Detachment, Rota, Spain. The FDRMC detachment and four destroyers are based in Rota, Spain, where a single state-owned contractor, Navantia, performs all depot-level maintenance on the ships. Beginning in 2014, the Navy deployed four destroyers to Spain to support the U.S. ballistic missile defense mission to the North Atlantic Treaty Organization. The Navy designed the maintenance approach for these ships with the understanding that they would not require access to Navy- or contractor- operated dry docks during the time they are based in Spain, according to Navy officials. The Navy initially expected these destroyers to be in Spain for about 6 years and to receive maintenance every 2 years. However, in 2015 the Navy updated its maintenance strategy for these ships to provide shorter, but more frequent maintenance periods to support a longer time basing them in Spain. Under the updated approach, the Navy plans for each destroyer to receive six maintenance periods during a roughly 8-year time period based in Spain. For additional information on the FDRMC detachment in Rota, Spain, see appendix V. FDRMC Detachment, Manama, Bahrain. The FDRMC detachment in Bahrain is responsible for the depot-level maintenance of the 10 patrol coastal and 4 mine countermeasures ships based there—the most ships based at an overseas homeport. While the Navy does not operate any dry docks or depot-level repair facilities in Bahrain, it relies on two main contractors, Bahrain Ship Repairing and Engineering Company and Arab Shipbuilding and Repair Yard, to conduct ship maintenance in Bahrain. The ships in Bahrain receive depot-level maintenance at contractor facilities there. Both contractors in Bahrain have dry docks or similar capacity to fit the MCMs and PCs based there, as well as some larger Navy ships. A unique capacity consideration for ships visiting Bahrain, according to officials there, is that the Navy does not have dedicated pier space for ships when they come into port. As a result, the Navy must rely on contractor space for maintenance, and on other pier space when visiting ships are at the homeport—which they share with others, such as commercial cruise lines. For additional information on FDRMC detachment in Manama, Bahrain, see appendix VI. In addition to the depot-level maintenance periods for the surface ships we reviewed, the Navy maintenance centers in Japan, Spain, and Bahrain, also support additional maintenance functions, such as voyage repairs or technical assistance for visiting U.S. ships; coordinating intermediate-level maintenance that may be conducted there; and providing additional maintenance support to overseas ships outside of scheduled depot-level periods. The Navy Did Not Complete the Majority of Maintenance on Time during Fiscal Years 2014 through 2018 for Ships Based Overseas, and Its Analysis on the Causes of Delays Is Limited The Navy did not complete the majority of the maintenance periods performed on ships based overseas on time during fiscal years 2014 through 2018. Navy officials identified a variety of factors that contribute to delays, such as the discovery of additional work requirements after maintenance has begun or staff shortages affecting management and oversight of maintenance. The Navy collects information on overseas maintenance at individual homeports, but its analysis of factors contributing to the delays is generally focused on the planning and execution of individual maintenance periods. The Navy Underestimated Time Required to Complete the Majority of Maintenance Periods for Ships Based Overseas The Navy underestimated the time needed to complete maintenance for 50 of the 71 maintenance periods—about 70 percent—started during fiscal years 2014 through 2018. Specifically, 21 maintenance periods ended early or on time and 50 maintenance periods ran beyond their planned schedules, as illustrated in figure 4. More than half of the maintenance periods that were completed late—29 of 50—went 31 or more days beyond the Navy’s planned schedule. As a result, from 2014 through 2018 there were 29 times when ships based overseas were unavailable for operational requirements, certain training, or other purposes for 31 or more unplanned days. During this time period, the Navy completed more maintenance periods a month or more later than planned than it completed early or on time. As a result of maintenance schedules not being completed on time, all four overseas Navy homeports with surface ships we analyzed— Yokosuka, Japan; Sasebo, Japan; Rota, Spain; and Manama, Bahrain— experienced a total of 3,475 days ships were in maintenance beyond their expected durations—referred to in this report as days of maintenance delay. As illustrated in figure 5, Manama, Bahrain, experienced the most days of maintenance delay during fiscal years 2014 through 2018, while Rota, Spain, experienced the least. We also analyzed delays at overseas homeports by calculating the days of delay experienced as a percentage of its total workload in terms of total days of maintenance conducted. Using this analysis, we found that ships in Bahrain experienced the highest rate of delay at 34 percent while ships based in Rota, Spain, experienced only a 2.2 percent rate of delay (as illustrated in figure 6). Taking workload into account illustrates some difference in the rate at which each of these homeports experiences ship maintenance delays. For example, ships in Sasebo and Yokosuka experienced a similar total number of days ship maintenance was delayed—1,001 days and 994 days over the 5-year time period, respectively. However, when port workload is taken into account, Sasebo’s rate of delay is higher. Specifically, ships based in Sasebo experienced a maintenance delay rate of 31.2 percent compared with 18.5 percent of the time for the surface ships in Yokosuka. Various Factors Contribute to Maintenance Delays for Ships Based Overseas According to Navy maintenance center officials and crewmembers from the ships we visited, and our analysis of Navy information, a number of interrelated factors and issues contribute to maintenance delays for the surface ships based overseas including: Discovery of additional, unplanned work after maintenance is underway. According to maintenance officials in Bahrain and Japan, the discovery of the need for additional maintenance and repair work after the work planned for the maintenance period has been finalized is a key driver of maintenance delays. This additional work can be in the form of growth in the magnitude of planned work, or identification of the need for new work that was not previously planned. For example, maintenance officials in Japan attributed maintenance delays they experienced on ships at both Yokosuka and Sasebo during fiscal years 2016 through 2018 to this growth in planned work or new work. Similarly, officials in Bahrain said that growth and new work is one of many contributing factors to maintenance delays for the aging MCMs and PCs based there. For example, officials from Commander, Naval Surface Squadron Five that track their ships’ depot maintenance identified that additional work to stern tubes on the USS Squall, which is homeported in Bahrain, resulted in the ship’s maintenance schedule being extended by 137 days. Navy officials also stated that the reason growth and new work is such a key driver of delays is that it can add further delays beyond that needed to complete the repair, due to time required for additional contract actions and ordering parts that are needed to conduct the added work. A number of factors can cause or further exacerbate growth and new work, according to Navy officials. For example, the Navy has made efforts to catch up on backlogs of deferred maintenance and improve the health and condition of the ship, so the Navy may decide to extend the maintenance period to ensure all identified maintenance has been completed rather than deferring it to a subsequent maintenance period. Additionally, officials pointed to ships’ complex propulsion, communication, and weapons systems that have complicated maintenance and modernization requirements that cannot always be fully anticipated. Missing or late maintenance planning milestones. The Joint Fleet Maintenance Manual emphasizes the importance of meeting planning milestones to identify, estimate, and schedule the work to be done in the maintenance period. These milestones include steps to guide advanced planning of initial maintenance requirements and schedules, and to further refine and develop the work, cost, and schedule estimates for each maintenance period. For example, these milestones include assessments of the ship’s condition and other ship checks to identify and validate planned work intended to minimize growth and new work; to identify and mitigate risks to planned schedules; and to provide deadlines for developing and awarding contracts to do the work. Adherence to these planning milestones becomes more critical as the planned start of the maintenance period approaches to ensure work can be contracted and begun on time. The final contract is awarded about 2 months prior to work beginning, and the Navy finalizes the planned duration and schedule of the maintenance period about a month before maintenance is scheduled to begin. According to Navy officials, missing or late planning milestones can contribute to maintenance delays. For example, NAVSEA and overseas maintenance officials emphasized that getting on board a ship at various points in the planning process to assess the ship’s condition and validate planned work is critical to developing accurate work scope, cost estimates, and schedules—otherwise, growth and new work or other issues can emerge once maintenance is underway. According to the Joint Fleet Maintenance Manual, ship checks are needed to inform specific planning milestones, to validate planned work, and should be done as early in the planning process as possible. The Navy requires this validation to ensure needed maintenance work is sufficiently defined, problems are accurately diagnosed, and feasible resolutions are recommended. However, even though ship condition assessments are important milestones to limit growth and new work, NAVSEA officials part of Surface Team One said that these assessments and other checks are regularly postponed, which can prevent work from being identified with sufficient time to plan for it. Similarly, maintenance officials in Japan said that, due to the operational tempo in Yokosuka and Sasebo, ships are often not available for required ship checks until the ship arrives in port at the start of its maintenance period. Though officials could not provide the frequency that such milestones are missed, they said missing assessments and other milestones can contribute to schedule delays and result in maintenance periods exceeding planned resources. For example, the Naval Inspector General found that the shortage of personnel at the FDRMC and Fleet Logistics Center in Bahrain resulted in contracting milestones being routinely missed for ships based there, and once these ships were in maintenance, the growth in work to be completed grew by an average of $830,000 for maintenance periods in fiscal years 2017 and 2018. Shortages of experienced and skilled personnel for planning, management, and oversight. According to NAVSEA and overseas maintenance center officials, shortages of U.S. personnel that perform maintenance planning, contracting, and oversight roles, particularly staff with critical skills and experience, can affect ship maintenance and contribute to delays. For example: Personnel shortages hinder staffing of project teams. FDRMC Bahrain officials said that due to personnel shortages, they are often unable to assign staff to the project teams until the maintenance period starts. According to the Joint Fleet Maintenance Manual, a project team is assigned to manage an individual maintenance period, and is composed of personnel with specific skills and responsibilities. Additionally, according to CNRMC Instruction 4790.4B, the project team is responsible for key maintenance planning and execution activities and related milestones from as early as a year before the maintenance begins. CNRMC Instruction 4790.4B also states that such maintenance planning milestones are to aid in developing project plans, identifying and mitigating risks, and tracking progress of planning. Project teams are also responsible for overseeing contracted maintenance work and ensuring it meets quality standards. For example, prior to the start of the maintenance period, project teams are responsible for identifying and mitigating risks to completing maintenance within the planned schedule and budget. However, officials in Bahrain stated that as a result of persistent staffing shortages, they have been unable to staff these project teams until the maintenance period begins, and have also been unable to provide sufficient oversight of the contractors’ performance during the maintenance period, which has resulted in maintenance delays. Shortages of personnel with relevant experience affect management and oversight of maintenance. Officials in Japan and Bahrain stated that insufficient numbers of personnel with ship maintenance experience can negatively affect maintenance timeliness. For example, the Fleet Logistics Center in Bahrain—which manages the contracting process for ships based there—had only eight of 18 authorized U.S. civilian contracting-related positions filled, as of March 2019, according to officials. Additionally, of the filled positions, only one contracting officer had prior experience with ship maintenance contracting, according to Fleet Logistics Center officials. Officials in Japan said that experience levels of U.S. civilians at SRF-JRMC have decreased as a result of high turnover in recent years with the average amount of work experience for U.S. civilians managing ship maintenance in Sasebo declining from over 5 years in 2014 to 3 years in 2017. Staff shortages on ships affect crews’ ability to conduct maintenance. Navy officials also emphasized the importance of ship crews in identifying and providing needed maintenance work, but noted that ship crew shortages negatively affect on-board ship maintenance. This can increase the amount of work required during depot-level maintenance periods. In May 2017, we reported that reduced crew sizes contributed to maintenance being deferred and increased maintenance costs, and Navy officials and ships’ crews we spoke to in Japan and Bahrain stated that ships there continue to experience manning shortages. For example, from September 2018 through February 2019, nearly 30 personnel from Bahrain-based Navy organizations were temporarily assigned to ships based in Bahrain to fill manning shortages, according to Navy officials and information, including for maintenance-specific positions. According to maintenance officials overseas and in the United States, other factors also can add to the complexity of maintenance planning and contribute to delays including the length of time it takes to obtain spare parts overseas, availability of obsolete parts, and other challenges associated with maintaining aging ships, such as the MCMs and PCs, which are at or beyond their original service lives. The Navy Collects Information on Overseas Maintenance at Individual Homeports, but Its Analysis of Factors Contributing to the Delays Is Limited The Navy uses a number of mechanisms to monitor the planning and execution of individual maintenance periods to track progress and mitigate possible risks. According to Navy documentation and officials, these mechanisms include: Individual homeports identify technical reasons for delays on individual maintenance periods. Maintenance centers overseas and in the United States monitor the planning and progress of individual maintenance periods. SRF-JRMC officials in Yokosuka, Japan, monitor ongoing and recently completed maintenance periods and may identify technical causes for ship delays. For example, new work was identified on the main reduction gear of the USS Barry that was not in the planned work package and led to delays, according to officials. Additionally, Commander, Naval Surface Squadron Five in Bahrain tracks instances of growth and new work during the depot- level maintenance periods for the PCs and MCMs based there, including tracking the specific number of delayed days attributed to certain issues. NAVSEA conducts regular meetings to report status of upcoming and ongoing maintenance. NAVSEA collects information on and monitors the progress of individual maintenance periods, including at overseas homeports, through a variety of regular meetings and briefings. For example, NAVSEA 04 and CNRMC each conduct biweekly meetings with their respective maintenance centers to monitor advanced planning of upcoming maintenance periods and the progress of ongoing maintenance periods for the ships under their responsibilities, according to officials. Information shared during these briefings can include tracking whether certain planning milestones are met and identifying risks to the on-time completion of individual ships’ maintenance periods. This information is then compiled into monthly briefings to the NAVSEA commander providing a snapshot of upcoming and ongoing maintenance periods and seeking approval for adjustments, according to officials. Collecting and sharing lessons learned throughout the planning process. According to the Navy’s maintenance manual and related guidance, the collection and sharing of lessons learned from individual maintenance periods is to be part of certain planning milestones, including to inform the maintenance schedule and work estimates. For example, CNRMC Instruction 4790.4B directs that maintenance completion conferences with relevant stakeholders are to provide a detailed review of the maintenance period, including lessons learned that can be used to plan future maintenance periods, such as to revise specific work items. According to CNRMC and NAVSEA 04 officials, lessons learned are collected at the end of each maintenance period and can be shared with other project teams. The Joint Fleet Maintenance Manual also states that while the lessons learned process is owned by the type commanders—for surface ships, these are Commander, Naval Surface Force, U.S. Pacific Fleet for ships in Japan and the western United States, or Commander, Naval Surface Force, U.S. Atlantic Fleet, for ships in Spain, Bahrain, and the eastern United States—the lessons learned process is part of the Surface Team One structure. However, Surface Team One officials noted that each of the milestones that include them is led by other Navy organizations, and its role in the lessons learned process is managed by a part-time contracted position. CNRMC tracks overall days of maintenance delay by fiscal year. CNRMC tracks and monitors the overall number of days individual ship maintenance periods are delayed and can perform analysis of overall delays, such as the number of days experienced by ship class and fiscal year. Additionally, CNRMC analysis has also identified specific ships that experience the longest delays, though it did not regularly include maintenance periods in Japan until 2018, according to officials. CNRMC tracks costs associated with growth and new work for individual maintenance periods. CNRMC tracks the costs associated with growth and new work discovered during maintenance periods by the regional maintenance centers it manages, including at overseas detachments in Bahrain and Spain. The costs that are tracked do not include information on any related delays, however, and do not include these costs for the ships in Japan. Other recent Navy efforts have begun to examine issues related to delays. According to Navy officials, several Navy entities are beginning efforts to improve the execution of surface ship maintenance. For example, in fiscal year 2019 the Navy began a broad effort to improve Navy surface ship, submarine, and aviation readiness, as well as public shipyards. This effort, called Performance to Plan, designates Commander, Naval Surface Forces, and Commander, NAVSEA, to improve performance of ship maintenance in private and public shipyards. According to Navy officials, the effort to improve surface ship maintenance consists of a pilot program examining how to better execute maintenance periods for destroyers, improve forecasts of maintenance period duration and assessments of ship condition, planning for growth and new work, and adherence to planning milestones. However, officials said this effort is still in the early stage and does not specifically assess maintenance delays for ships based overseas. NAVSEA’s SURFMEPP and Surface Team One also have recently begun related efforts. For example, SURFMEPP officials said they recently began an effort to examine and correct causes of growth and new work by analyzing changes to contracts or work items that result in more than $100,000 of additional cost. However, while officials said in July 2019 that this effort has been underway for about 9 months, they could not provide additional information on how it relates to delays. According to NAVSEA officials that co-chair Surface Team One, it has begun an effort to improve how adherence to key planning milestones is tracked across surface ship maintenance periods. To support this effort, in October 2018 the Navy updated the Joint Fleet Maintenance Manual to include additional requirements for meeting maintenance milestones and to document any changes, including reasons for those changes. However, according to officials, these efforts are in their early stages, and the Navy has not used the information to analyze maintenance delays for overseas ships. Although a number of different Navy entities conduct a variety of activities through which information on maintenance delays is collected and analyzed, these efforts are limited as the existing analysis is not comprehensive and systematic in nature. Specifically, the Navy has not positioned itself well to address the factors contributing to the maintenance delays because it has not (1) designated an individual entity responsible for conducting a single, comprehensive systematic analysis of overseas surface ship maintenance delays; and (2) developed a plan based on that analysis to address these delays. First, this is in part because the responsibilities for managing surface ship maintenance overseas is shared among NAVSEA 21, CNRMC, and NAVSEA 04, which use somewhat different processes for their work, according to officials. For example, NAVSEA 04 has responsibility for the maintenance of aircraft carriers and submarines at naval shipyards, while CNRMC focuses on surface ships. In addition, until SRF-JRMC was brought under control of NAVSEA in October 2018, CNRMC was not regularly including maintenance periods in Japan as part of its tracking and monitoring of days of maintenance delay. According to officials, an operating instruction to align roles, responsibilities, and processes for surface ship maintenance in Japan between CNRMC and NAVSEA 04 is being developed, but as of September 2019, this instruction had not yet been finalized. Further, CNRMC and NAVSEA 04 officials pointed to NAVSEA 21 or Surface Team One as more appropriate entities to conduct a comprehensive systematic analysis of ship maintenance delays given their broad, enterprise-wide roles for managing and improving surface ship maintenance. Surface Team One officials said that it could be an appropriate entity to conduct such analysis, and according to its charter, one of the entity’s purposes is to measure performance of the planning and execution of surface ship maintenance periods and to manage and improve schedule, cost, and quality. However, officials said they have not conducted such a systematic analysis of maintenance period performance or developed a comprehensive plan to address them, in part due to inconsistent organizational leadership and personnel turnover. According to officials, since its founding in 2009, Surface Team One has been re-chartered twice and is in the process of further reorganizing under a fourth version of its charter. Part of the reason for this reorganization, according to officials, is to resource and structure Surface Team One to conduct more systematic, enterprise-wide analyses of issues affecting surface ship maintenance, for which they hope to develop a plan by the end of 2019. However, officials said these efforts did not specifically include analysis of maintenance delays for ships based overseas. Additionally, while Navy officials said that Performance to Plan efforts could help inform overseas maintenance delays, this effort is in the early stages of a pilot effort looking only at destroyer maintenance, and does not specifically analyze maintenance delays for ships based overseas. Second, as a result of there being no single, comprehensive analysis of overseas surface ship maintenance delays, there is no plan for the Navy to improve the timeliness of its maintenance in a holistic way. Instead, individual organizations and maintenance centers have identified improvements for individual ships’ maintenance or have undertaken efforts to address certain contributing factors to delays. While these efforts are important, given the interrelated challenges related to maintenance across the Navy, and that the Navy is dependent upon synchronized and timely maintenance to provide ships for operations to meet national security needs, the Navy would benefit from a plan of action that was comprehensive in nature. Standards for Internal Control in the Federal Government state that management should assign responsibility to achieve objectives and remediate deficiencies; compare actual performance against planned performance; and evaluate deficiencies on both an individual basis and in the aggregate. Further, OMB Circular No. A-123, Management’s Responsibility for Enterprise Risk Management and Internal Control, emphasizes that when developing corrective actions, agencies should perform a root-cause analysis of the deficiency and ensure that subsequent strategies and plans address the root of the problem and not just the symptoms. Additionally, our past work on results-oriented management cites a number of key practices that can strengthen the use of performance information for process improvements. These practices include aligning agency-wide goals and measures, and building analytic capacity to use the information. Our past work has further shown this information should then be incorporated into improvement plans that include identifying analytically based goals; results-oriented metrics to measure progress; required resources, risks, and stakeholders to achieve those goals; and regularly reporting on progress. While several different Navy entities have a variety of efforts underway related to issues associated with ship maintenance delays, without designating an entity to conduct a comprehensive, systematic analysis to identify and understand the underlying causes maintenance periods grow beyond planned schedules, the Navy risks continuing to underestimate maintenance needs and the time and resources required to address them. Further, without conducting such an analysis to understand the underlying, interrelated causes of these delays, and incorporating this analysis into a comprehensive results-oriented plan to address them, the Navy cannot effectively target corrective actions to improve timely completion of ship maintenance to ensure ships are available for the critical training crews need and operations to support U.S. military and national security goals. Navy Has Not Assessed and Mitigated Risks That Challenges Pose to Implementing Its New Maintenance Approach or Included Overseas Maintenance in Its Plans to Grow Fleet The Navy is in the process of updating the maintenance approach for cruisers, destroyers, and amphibious ships based in Japan, but it has not assessed and mitigated risks that several challenges may pose to its successful implementation. Additionally, the Navy has not included assessments of overseas maintenance requirements in its long-range plans to support fleet growth to 355 ships. The Navy Has Developed a New Maintenance Approach for Surface Ships in Japan Based on the Approach Used in Spain The Navy has developed a new maintenance approach for the cruisers and destroyers in Yokosuka and the amphibious ships in Sasebo based on the approach developed for destroyers in Spain. Specifically, the Navy developed a new maintenance approach for the four destroyers it began to deploy to Rota, Spain, in 2014 and 2015 that includes generally shorter, but more frequent, maintenance periods. According to maintenance center and other Navy officials, the Navy developed this approach to avoid conducting dry-dock maintenance overseas so that the Navy could maximize the time the ships were available for operations. According to officials, the Navy tailored this approach to the specific ships, mission, and maintenance resources available in Rota. For example, the four destroyers in Rota conduct patrols two ships at a time with predictable patrol schedules. With such specific operational and maintenance schedules officials said there is little margin for changes, and adjustments or delays could affect the ships’ operational availability to support their ballistic missile defense mission. Under this approach, the Navy completed the majority of its maintenance on these four ships during fiscal years 2014 through 2018 on time—with only 20 total days of maintenance delay, equating to a relatively low overall delay rate of 2.2 percent. Navy officials said that the new maintenance approach in Rota has been successful because the Navy: selected four ships with a high degree of commonality; for example, the ships were of similar age, systems, and equipment configuration, which helped facilitate planning for and conducting maintenance; ensured the ships received all needed maintenance and modernization before being sent to Spain, and arrived from the United States in good condition, which reduced the likelihood that they would require unexpected maintenance while overseas; designed the maintenance center and its staffing around the maintenance approach for the four destroyers; and coordinated with the contractor in Spain to ensure it had sufficient workforce and resources, including capacity to surge resources if additional work is discovered so that it can be completed on time. Based on the performance of the maintenance approach for destroyers in Spain, officials stated that the Navy began to develop a similar approach in 2016 for its ships in Japan. NAVSEA officials identified that shorter, more frequent maintenance could help ensure that its ships based in Japan received the maintenance they need, while also meeting their high operational demands. The Navy finalized a new maintenance approach for cruisers and destroyers in Yokosuka in December 2018, and was in the process of finalizing the maintenance concept for the amphibious ships in Sasebo, according to NAVSEA officials in June 2019. For example, like in Spain, the Navy has adjusted the schedules for the planned periods in Yokosuka to be shorter, but more frequent. Planning documents show that under the new approach for the destroyers in Yokosuka, the Navy plans to provide them with eight maintenance periods over approximately 8 years overseas before rotating the ships back to the United States. Previously, the Navy planned for destroyers in Japan to receive eight maintenance periods over an estimated span of over 16 years overseas under the prior approach. Under the new approach, the surface ships in Japan are expected to receive all required maintenance, including completing most or all backlogged maintenance according to officials, in the United States before relocating the ships to Japan. Additionally, while officials expect ships in Yokosuka to receive some dry-dock maintenance during their rotation overseas, the amphibious ships in Sasebo generally will not—similar to the arrangement for destroyers in Spain. As a result, the new maintenance approach expects that ships in Sasebo will accrue maintenance backlogs that must be resolved upon return to the United States. Several Challenges Pose Risks to Successful Implementation of New Maintenance Approaches Overseas The Navy has decided to apply its new maintenance approach for cruisers, destroyers, and amphibious ships in Japan and in 2018 began initial implementation on certain ships already based there, but a number of challenges may pose risks to successful implementation of the strategy. Based on information from planning documents and officials, successful implementation relies on several planning assumptions that may be optimistic when compared to actual experience maintaining surface ships overseas and in the United States. Specifically, the new approach in Japan assumes that: Ships will receive robust, deep maintenance and modernization in the U.S. and meet their life-cycle health requirements prior to overseas assignment. Ships will receive and complete planned maintenance on time while overseas to maximize operational availability. Ships will rotate back to receive full maintenance in the United States after no longer than 9 years of overseas assignment. However, Navy officials and our analysis identified several challenges: (1) U.S. industrial base maintenance capacity limitations, (2) maintenance delays in the United States and overseas, (3) the ability of the overseas contractor industrial base to support future workload in Japan, and (4) differences in the operating environments between Spain and Japan. These challenges, which are discussed below in more detail, could pose risks to the successful implementation of the new maintenance approach. U.S. industrial base maintenance capacity limitations. Implementing the new maintenance approach in Japan assumes that the ships identified for deployment will receive all required maintenance and modernization in the United States prior to being based overseas. However, the Navy has been challenged to do this in the past due to limited domestic maintenance capacity. For example, the Navy deferred maintenance assessments of the condition of the USS Barry and USS Milius that were to take place in the United States before moving the ships to Japan. As a result, Navy officials said these assessments had to be done in Japan. Additionally, upon arriving in Japan in November 2017, the USS Barry had to begin immediate unscheduled maintenance to correct various issues, and as of our visit in February 2019, was still undergoing maintenance. According to U.S. Pacific Fleet and maintenance center officials, in fiscal year 2014 the USS Curtis Wilbur received modernization in Japan due to lack of capacity in the United States. Further, Navy planning documents identified U.S. commercial dry-dock capacity shortfalls that may hinder the Navy’s ability to support the future maintenance workload in the United States. For example, the Navy’s Long-Range Plan for the Maintenance and Modernization of Naval Vessels for Fiscal Year 2020 identified limited U.S. dry-dock capacity in the United States as posing a significant challenge to maintenance of U.S.-homeported ships and that this situation reduces the margin for schedule changes. According to the Navy’s analysis, demand for surface ship maintenance in the United States will exceed available maintenance resources for fiscal years 2019 through 2026. During this time, the Navy will be rotating ships based in the United States to exchange with those currently based in Japan and Spain. Navy officials said the capacity shortfall in the United States negatively affects ship condition and maintenance of ships sent to Japan. However, the Navy’s analysis does not account for the need to perform deep maintenance and modernization on ships in the United States before and after sending them to overseas homeports, as required by the new maintenance approach for ships bound for Japan, as well as Spain. Maintenance delays in the United States and overseas. Maintenance delays at both U.S. and overseas homeports may also affect the Navy’s implementation of its new maintenance approach. Successful implementation of the new approach depends in part on ships receiving all required maintenance on time prior to moving overseas, as well as receiving timely maintenance during their time based abroad. Our analysis of Navy surface ship maintenance periods that started in fiscal years 2014 through 2018 found that about 60 percent of maintenance periods in the United States ran 31 or more days beyond schedule. Additionally, our analysis shows that ships homeported at both U.S. and overseas locations experienced an average rate of delay of about 25 percent (see fig. 7). Additionally, rates of delay in Sasebo, where the Navy plans to implement one of its new maintenance approaches, exceed 30 percent. According to Navy officials, the new maintenance approach for ships in Japan is intended to provide more frequent maintenance periods, in an effort to improve ship maintenance and to maximize ships’ availability for operations. However, the approach also relies on most of these maintenance periods being shorter—and being completed on time. Given the Navy’s history of persistent maintenance delays in Japan, this could be a challenge. Further, Navy officials said that maintenance delays experienced in the United States could also affect the maintenance that ships bound for and returning from overseas homeports may receive, and pose a risk that maintenance will be deferred to overseas homeports. Challenges with overseas contractor industrial base meeting future workload in Japan. Navy maintenance officials in Spain said that successful implementation of the new maintenance approach there relied on sufficient contractor capacity overseas, and that the Navy involved the contractor in the development of the maintenance approach to ensure they could implement it. In contrast, Navy officials in Japan stated that current contractor capacity may not meet expected future workload. For example, Navy documentation shows that contractors performed almost two-thirds of ship maintenance in Sasebo in fiscal year 2018. Additionally, the documentation shows that maintenance planned for fiscal year 2020 is expected to increase beyond existing Navy and contractor capacity. Maintenance in Sasebo relies on a number of smaller contractors, and these contractors have experienced challenges planning for the unpredictable maintenance workload there, according to officials. Specifically, the amphibious ships based in Sasebo typically deploy as a group of three. As a result, Navy officials said the workload in Sasebo can be uneven. When all three ships return to port, they require maintenance at the same time. The Navy found that contractors have difficulty planning for this uneven workload, among other issues, which can deter contractors from bidding on work. For example, in fiscal year 2015, the Navy found that they were unable to award over 25 percent of work planned for contractors in Sasebo because no contractor bid on the work. The Navy plans to add a fifth amphibious ship in Sasebo in fiscal year 2020, in part to provide a more stable workload there, according to officials. The Navy expects the additional ship will also result in a forecasted increase in overall maintenance workload there. Navy officials also expressed concerns about the continuity of the existing industrial base in Yokosuka to be able to support future Navy needs. According to Navy documentation, in fiscal year 2018, about one-third of ship maintenance in Yokosuka was conducted by contractors, and, according to officials, the Navy relied on one main contractor to conduct much of this work. However, Navy maintenance center officials in Japan stated they have concerns about the continuity of the contractor to support this workload. The Navy has begun efforts to consider conducting maintenance at contractor facilities outside the ships’ homeports of Yokosuka and Sasebo. Specifically, the Navy has begun to consider using contractor facilities located outside the Yokosuka area, as far as 2 hours away from where the ships are currently based. For example, Navy officials told us that they conducted market research and outreach to potential contractors, and have awarded a small contract for a short continuous maintenance period to a new contractor about 30 minutes outside the Yokosuka area. However, maintenance and contracting officials stated these efforts face their own challenges. For example, conducting weeks or months of maintenance on a ship as far as 2 hours outside a ship’s homeport—where crews and families live—could require additional travel, housing, and other costs. Additionally, maintenance and contracting officials in Yokosuka stated that the substantial regulatory, legal, and Navy requirements that private companies must adopt to contract with the U.S. government might serve as disincentives for prospective Japanese contractors, and developing these contractors will take time. Differences in the operating environments in Japan and Spain. According to NAVSEA officials, the decision to apply the approach in Japan was based on its timely performance in Spain, but the ships, missions, and operating environment in Yokosuka and Sasebo differ substantially from the environment in Spain. For example: Greater diversity and number of ships in Japan. Navy officials told us that the four destroyers sent to Rota in 2014 and 2015 were specifically chosen with similar age, configuration, and condition, which made it easier to sustain the maintenance approach, since issues and lessons from one ship could be easily applied to the next. The ships in Japan in fiscal year 2019 consisted of a more diverse set of ships—eight destroyers and three cruisers in Yokosuka, and various classes of amphibious ships in Sasebo. According to officials, these ships are of different configurations and capabilities. Greater workload and staffing challenges in Japan. Navy officials have attributed the persistent maintenance delays experienced in Japan to insufficient U.S. maintenance prior to deployment, insufficient estimation of the maintenance work package, missed planning milestones, staffing challenges, and other causes, that are not currently being experienced in Rota. Less predictability in operational tempo in Japan. According to Navy officials in Rota, the four ships based there have the same mission, regular and predictable patrol schedules, and do not go above Navy deployment limits. Additionally, officials said the patrol schedules allow for some additional maintenance to be conducted when ships are in port, if needed. As a result, Navy officials in Rota said that they are able to meet key maintenance planning milestones such as conducting ship checks and other assessments. In Japan, however, Navy officials and operational commanders described operational tempo as more unpredictable, and that ships can be unavailable due to the operational demands of the varied missions with different timeframes for ships in Seventh Fleet’s area of responsibility. For example, according to Seventh Fleet officials, the cruisers and destroyers in Yokosuka are expected to serve a number of different missions, including conducting patrols around Japan or Guam; providing ship presence in the East China Sea; or escorting the carrier as part of the strike group. Additionally, according to Navy officials, operational tempo in Japan continues to be high, and in 2015 we reported that to meet increasing demands overseas, the Navy has extended deployments and increased operational tempo. Standards for Internal Control in the Federal Government state that it is a key responsibility of management to analyze and respond to identified changes and related risks, and to monitor program effectiveness. These standards also note that changing conditions often result in new risks or changes to existing risks that need to be assessed. Additionally, the April 2011 DOD Product Support Business Case Analysis Guidebook further states that each risk should be reviewed and assessed, and that effective mitigation plans may involve making tradeoffs in capabilities, schedule, and performance. However, NAVSEA officials said the Navy has not assessed the risks posed by these challenges to implementing its new maintenance approach in Japan. Instead, officials based the decision to implement the approach in Japan on the performance of the approach in Rota, Spain. Without a full assessment of the risks these challenges may pose to successful implementation of its new maintenance approach, and without identifying ways to mitigate any risks posed by these challenges, the Navy cannot ensure its overseas homeported ships complete all required maintenance as planned in support of fleet readiness needs. The Navy Plans to Grow Its Fleet but Has Not Included Overseas Ship Maintenance Requirements in Its Plans The Navy’s timeline for growing the fleet from 290 total ships (as of September 2019) to 355 ships shows that the largest increase will be in the number of surface ships. Specifically, the Navy plans to increase the number of surface ships in the fleet by a total of 48 ships in the next 15 years, or by 2034. However, the Navy’s long-range plans to grow its fleet do not consider the maintenance these ships will require while based or traveling overseas. The Navy’s Report to Congress on the Annual Long- Range Plan for Maintenance and Modernization of Naval Vessels for Fiscal Year 2020, which is intended to assess the maintenance and modernization requirements for the fleet as it grows, only assesses maintenance provided by private and public shipyards in the United States, not overseas. It does not identify or assess the maintenance requirements needed overseas—including those provided by Navy facilities or the contractor industrial base. Moreover, it does not identify overseas requirements, such as any expected changes in the number of ships based there or growth in the number of ships visiting overseas locations from the United States. For example: As the number of ships in the overall fleet grows, NAVSEA officials said they expect the number of ships based overseas to grow proportionally, and the number of U.S.-based ships conducting operations and exercises overseas to increase, thereby increasing overseas maintenance requirements. However, the expected increase in the fleet and associated maintenance requirements for ships based and visiting overseas were not included in the recent long-range plans. For example, according to officials, the Navy plans to base an additional amphibious ship in Sasebo, Japan, by fiscal year 2020, and the Navy is examining a possible increase to the number of destroyers in Rota, Spain. According to maintenance center officials in Rota, increasing the number of ships based in Rota would require additional planning to meet the Navy’s needs, such as negotiating with the Government of Spain to request additional capacity, such as pier space, for such future requirements. The Navy projects the number and type of ships based in Japan and Bahrain to change in the next few years. Specifically, the Navy plans to decommission the mine countermeasures (MCM) ships currently homeported in Japan and Bahrain by 2023 and replace them with littoral combat ships to perform the mine countermeasures missions. However, maintenance center officials in Bahrain stated that as of March 2019, plans for the overseas maintenance of littoral combat ships remained uncertain, even though officials expect the initial deployments of littoral combat ships to Bahrain to begin as early as 2020. Additionally, the Navy has not developed deployment timelines and overseas maintenance requirements for littoral combat ships in the Middle East and Western Pacific areas of operation, even though the USS Montgomery arrived in Singapore to begin its overseas rotational deployment in July 2019. According to Navy officials, the Navy expects long-term deployments of littoral combat ships to both areas of operation as the MCMs are decommissioned. Ships based in the United States also rely on voyage repair at overseas shipyards while conducting missions or patrols. For example, according to the Navy’s annual report to Congress listing all repairs and maintenance performed on Navy ships, in fiscal year 2018, the maintenance center in Bahrain conducted voyage repairs for the USS Monterey and USS Arleigh Burke, both based in Norfolk, Virginia, and the USS The Sullivans, based in Mayport, Florida. Additionally, voyage repairs were conducted in Japan for visiting Navy ships and submarines based in Washington and Hawaii. Standards for Internal Control in the Federal Government state that it is a key responsibility of management to consider changes within the environment and other factors, and analyze and respond to identified changes and related risks through methods such as strategic planning and other assessments. These standards also note that conditions affecting the organization and its environment continually change, and management can anticipate and plan for significant changes by using a forward-looking process. NAVSEA officials said that when planning for future growth, they have focused on analyzing U.S. industrial base issues and potential mitigations to increase capacity for U.S.-based ship maintenance as demand grows beyond existing dry docks and pier space. Officials said the Navy did not analyze overseas maintenance requirements or projected growth overseas to include in the long-range plan. According to NAVSEA officials, future iterations of long-term maintenance planning are to include analysis of the Navy’s overseas maintenance capacity, which Navy officials said could begin in March 2020. As the Navy continues its long-term maintenance planning, it will be important for the Navy to conduct and include analysis of anticipated overseas maintenance requirements given that substantial growth of surface ships is expected through 2034—including destroyers and amphibious ships, two types of ships currently based overseas. Without analyzing maintenance needs and requirements for ships based overseas, including any projected growth or other force changes, in its long-range plans, the Navy cannot ensure it is sufficiently planning for the total needs—and resulting readiness and health—of the future fleet. Conclusions The Navy bases and maintains 38 surface ships—such as destroyers, cruisers, and amphibious ships, among others—at homeports outside of the United States. The 2018 National Defense Strategy has prioritized military readiness, which depends in part on ships completing maintenance on time, to ensure that the United States is positioned to respond to events quickly all over the world. Ship maintenance is a complex process involving numerous Navy and private industry stakeholders that devote substantial time and effort to ensure that ships receive the maintenance they need. Yet we have previously reported on the persistent delays and other challenges the Navy faces in completing maintenance on time both for ships in the United States and overseas. While a number of entities in the Navy have different efforts underway to examine individual ship maintenance issues, a comprehensive, systematic understanding of the underlying and interrelated causes for these delays is essential to implementing corrective actions to ensure these strategically based ships are able and ready for operations when needed. The Navy has also taken steps to adjust its maintenance strategies to improve ship maintenance while balancing the high operational demands for ships based in Japan. Additionally, the Navy has begun planning to grow the fleet, but the expected increase in the fleet and associated maintenance requirements for ships based and visiting overseas were not included in the recent long-range plans. Also, the Navy’s plans to implement updated maintenance strategies overseas, as well as to grow the total fleet, were developed without accounting for risks that challenges may pose to these strategies, as well as analysis of the necessary overseas maintenance requirements to sustain the Navy’s strategically important ships homeported or visiting overseas locations. Ensuring the Navy’s maintenance plans and capacity for the total fleet align with its plans for substantial future fleet growth will enhance the Navy’s ability to conduct timely maintenance of its overseas surface fleet, which, in turn, is essential to the Navy achieving its strategic goals. Recommendations We are making a total of five recommendations to DOD. The Secretary of the Navy should assign responsibility to an entity to conduct a single, comprehensive systematic analysis of overseas surface ship maintenance delays. (Recommendation 1) The Secretary of the Navy should ensure the designated entity conducts a comprehensive, systematic analysis to identify the underlying, interrelated causes of overseas surface ship maintenance delays. (Recommendation 2) The Secretary of the Navy should use the results of the analysis to develop a plan to address surface ship maintenance delays overseas. Such a plan should incorporate results-oriented elements, including analytically based goals, identification of risks to achieving those goals, identification of required resources and stakeholders, metrics to measure progress, and regular reporting on progress. (Recommendation 3) The Secretary of the Navy should ensure that Naval Sea Systems Command assesses and mitigates risks posed by any challenges, such as persistent delays and capacity limitations, to successful implementation of its new maintenance approach in Japan. (Recommendation 4) The Secretary of the Navy should ensure that Naval Sea Systems Command conducts analysis to include overseas maintenance requirements as part of its long-term maintenance plan to support the planned growth and readiness of the fleet. (Recommendation 5) Agency Comments We provided a draft of this report to DOD for review and comment. In written comments provided by the Navy (reproduced in appendix VII), DOD concurred with our recommendations. The Navy also provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of Defense, the Secretary of the Navy, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have questions about this report, please contact me at maurerd@gao.gov or (202) 512-9627. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix VIII. Appendix I: Objectives, Scope, and Methodology This report (1) describes existing maintenance capacity and approaches the Navy uses for surface ships based overseas, (2) assesses the extent to which the Navy completed maintenance periods as scheduled in fiscal years 2014 through 2018 and analyzes factors contributing to any delays, and (3) evaluates the extent to which the Navy has assessed any challenges facing future overseas ship maintenance efforts. The scope of this review includes the regularly scheduled depot-level maintenance of surface ships based overseas, the maintenance of which is generally the responsibility of Naval Sea Systems Command (NAVSEA). These ships comprised 38 of the 40 ships based overseas during the time period we analyzed, and consisted of the following ship classes: guided-missile cruisers (CG 47 class), guided-missile destroyers (DDG 51 class), mine countermeasures ships (MCM 1 class), patrol coastal ships (PC 1 class), amphibious assault ships (LHD 1 class), amphibious transport dock ships (LPD 17 class), dock landing ships (LSD 41 class), and an amphibious command ship (LCC 19 class). These ships were based overseas at homeports located in Japan, Spain, and Bahrain as of the end of fiscal year 2018. For objective one, to describe existing capacity and maintenance approaches the Navy uses for the regularly scheduled depot-level maintenance periods for the 38 surface ships based overseas during the time of our review, we reviewed Navy documents and information on the Navy’s overseas maintenance centers’ physical capacity and authorized workforce, local contractor industrial base and capacity, and other Navy organizations and commands responsible for planning, managing, and overseeing the maintenance of these ships. To examine physical capacity, we analyzed Navy information on U.S. and contractor facilities and equipment such as dry docks and information on future planning or improvements. We reviewed NAVSEA information and data on ship maintenance periods, as well as information and documentation on historic and forecasted workloads at each homeport, including the number and type of ships that have received maintenance at those shipyards. We also reviewed Navy maintenance plans and guidance that document Navy maintenance approaches and organizations at overseas homeports. We conducted site visits to three overseas homeports— Yokosuka and Sasebo, Japan, and Manama, Bahrain—where the Navy bases a majority of the surface ships overseas. We observed the physical capacity and operations of the maintenance centers and shipyards there, as well as the Forward Deployed Regional Maintenance Center (FDRMC) headquarters in Naples, Italy. We interviewed cognizant officials at Navy commands, numbered fleets, and maintenance centers, including officials at all the overseas maintenance centers responsible for ships based overseas: the U.S. Naval Ship Repair Facility and Japan Regional Maintenance Center (SRF-JRMC) in Yokosuka, Japan, and its detachment in Sasebo, and the FDRMC headquarters in Naples, Italy, as well as its two detachments—in Rota, Spain, and Manama, Bahrain. For objective two, to assess the extent to which maintenance schedules are completed as planned, we analyzed Navy data on regularly scheduled, depot-level maintenance periods for surface ships—including those maintained at overseas homeports and in the United States. NAVSEA collects and manages data on these maintenance periods— known as Chief of Naval Operations maintenance availabilities—for surface ships, submarines, and aircraft carriers. We obtained the data on surface ship depot-level maintenance periods used by NAVSEA’s Surface Maintenance Engineering Planning Program and the Commander, Navy Regional Maintenance Center (SURFMEPP). We used Navy data to identify depot-level maintenance periods conducted at each homeport starting in fiscal years 2014 through 2018 and to assess the extent to which maintenance schedules for ships based overseas were executed as planned from fiscal year 2014 through 2018, and the delays experienced. To assess the reliability of this data, we interviewed cognizant NAVSEA officials to understand system operating procedures, organizational roles and responsibilities, and any data limitations. NAVSEA provided information based on our questions regarding data reliability, including an overview of the data, data-collection processes and procedures, data quality controls, and overall perceptions of data quality. NAVSEA also provided documentation of how the systems are structured and what written procedures are in place to help ensure that the appropriate information is collected and properly categorized. We interviewed officials from SURFMEPP and CNRMC to obtain further clarification on data reliability, discuss how the data were collected and reported, and explain how we planned to use the data. Some of these data were used in prior reports, and their reliability had previously been assessed. In addition, we also assessed the reliability of the data by checking: (1) for missing data entries, (2) for duplicate records, and (3) to ensure the data was formatted consistently. We determined that the data were sufficiently reliable for the purposes of summarizing surface ship maintenance periods and related information at homeports both overseas and in the United States, including reporting on the duration of maintenance periods and the number of days of maintenance delays. Because maintenance periods may cross over one or more fiscal years, to be able to report on days ships spent in maintenance periods from fiscal years 2014 through 2018, we analyzed data on maintenance periods that began in fiscal years 2012 through 2018 for all surface ships included in the data, including those based at overseas and U.S. homeports. Specifically, we used the dates of the planned and actual durations of the maintenance periods in our data set to determine the total number of days ships spent in maintenance in fiscal years 2014 through 2018 and by how many days the maintenance periods were extended beyond their planned number of days—which the Navy refers to “days of maintenance delay.” To determine the total number of days ships spent in maintenance in each fiscal year, we allocated the number of days spent in maintenance periods according to the fiscal year in which the maintenance days occurred. After we calculated the number of days each maintenance period went beyond the planned duration, we allocated these days of maintenance delay to the fiscal years in which they occurred. To compare ship maintenance delays experienced at different homeports while accounting for the varying workload at each, we calculated days of maintenance delay as a percentage of the total number of days ships spent in maintenance periods each location, which resulted in a rate of delay that we could compare across homeports. In addition, we analyzed the number of maintenance periods that were completed on or ahead of time or were completed later than planned, and we examined these maintenance durations by the fiscal year in which the maintenance periods started. We interviewed officials to understand the reasons they identified for delays. We reviewed the actions the Navy has taken to identify, evaluate, and resolve these delays, including information in Navy policies, guidance, and documentation on the planning, management, and oversight of overseas ship maintenance. This information included the Joint Fleet Maintenance Manual and related Navy instructions, documents establishing maintenance requirements. We also reviewed Navy guidance and documentation on the planning and execution of maintenance for ships based overseas and in the United States, including documentation such as status briefings, planning documents, and lessons learned information identifying certain reasons for maintenance delays of individual maintenance periods. We interviewed cognizant Navy officials responsible for planning, managing, and conducting oversight for surface ship maintenance in the United States and overseas to understand how they produce and use this information to improve maintenance planning and execution. We compared this information to standards for planning, scheduling, and monitoring events to correct deficiencies and identify process improvements, including Standards for Internal Control in the Federal Government, which includes principles pertaining to oversight responsibility, evaluating issues, and remediating deficiencies; our Schedule Assessment Guide; and OMB Circular No. A-123, Management’s Responsibility for Enterprise Risk Management and Internal Control, which includes guidance on conducting a root-cause analysis when developing corrective actions. We also compared this information with our past work identifying best practices for results- oriented performance management and planning. For objective three, to assess the extent to which the Navy has assessed and mitigated challenges that may affect overseas ship maintenance efforts, including new maintenance approaches and future maintenance requirements as the Navy seeks to grow the fleet, we analyzed Navy documentation, NAVSEA data, and available information documenting challenges that affect maintenance overseas, as well in the United States. We also analyzed Navy efforts to address these challenges, as well as Navy plans for future fleet growth and maintenance workload, including the long-range plans for shipbuilding and maintenance as the Navy seeks to grow its fleet, as well as other studies and analyses pertaining to these plans. We interviewed cognizant Navy officials who plan, execute, and oversee overseas shipyards and maintenance, as well as operational commanders, to obtain their perspectives on issues and challenges associated with execution of ship maintenance. We compared this information to government standards for planning and monitoring events to assess changes in risk, correct deficiencies, and identify process improvements, including Standards for Internal Control in the Federal Government and DOD Product Support Business Case Analysis Guidebook. Logistics, Maintenance, and Industrial Operations (NAVSEA 04) Deputy Commander for Ship Maintenance and Modernization (NAVSEA 21) Commander, Navy Regional Maintenance Center (CNRMC) Surface Maintenance Engineering Planning Program (SURFMEPP) U.S. Naval Ship Repair Facility and Japan Regional Maintenance Center (Yokosuka, Japan, and detachment in Sasebo, Japan) Forward Deployed Regional Maintenance Center in Naples, Italy, and its detachments in Rota, Spain, and Manama, Bahrain Naval Supply Systems Command Fleet Logistics Centers in Yokosuka, Japan, and Manama, Bahrain U.S. Naval Forces Central Command U.S. Naval Forces Europe-Africa Human Resources Office for Commander Navy Region Europe, Africa, Southwest Asia (CNREURAFSWA) We conducted this performance audit from August 2018 to February 2020 in accordance with generally accepted government auditing standards. Those standards require we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Naval Sea Systems Command (NAVSEA) Organizations with Responsibility for Surface Ship Maintenance Overseas History and Mission The Ship Repair Facility and Japan Regional Maintenance Center (SRF-JRMC) was originally created in 1947 as the Ship Repair Department, and in 1951 became the Ship Repair Facility. In 2004 it became the combined SRF-JRMC. Headquartered in Yokosuka, SRF- JRMC provides oversight and support to its detachment in Sasebo and is responsible for the maintenance for ships based in Yokosuka. SRF-JRMC also provides technical assistance and voyage repairs for Navy ships visiting Japan. U.S. Navy Surface Ship Maintenance Snapshot Surface Ships Maintained in Yokosuka as of Fiscal Year 2018 8 Guided-missile destroyers (DDG) 3 Guided-missile cruisers (CG) 1 Amphibious command ship (LCC) Fiscal Year 2018 Authorized SRF-JRMC Workforce U.S. Civilians: 272 U.S. Navy: 108 Japanese nationals: 2,341 (paid for by the Government of Japan) In 2018, the Navy finalized a new maintenance approach for cruisers and destroyers based in Yokosuka. This approach relies on deep maintenance in the United States prior to ships moving to Japan, and increases the frequency of maintenance periods while ships are in Japan. The Navy has begun efforts to identify additional private companies to conduct ship maintenance to meet future planned workload, according to Navy officials. History and Mission The Ship Repair Facility and Japan Regional Maintenance Center (SRF-JRMC) detachment in Sasebo was originally designated as the Sasebo Office to the Ship Repair Facility in Yokosuka in 1976, and made a detachment in 1984. The SRF-JRMC detachment is responsible for supporting the maintenance for the eight surface ships based in Sasebo, and can provide technical assistance and other maintenance to ships in and visiting Japan. U.S. Navy Surface Ship Maintenance Snapshot Surface Ships Maintained in Sasebo as of Fiscal Year 2018 1 Amphibious Assault Ship (LHD) 1 Amphibious Transport Dock (LPD) 2 Dock Landing Ships (LSD) 4 Mine Countermeasures Ships (MCM) Fiscal Year 2018 Authorized SRF-JRMC Workforce U.S. Civilians: 65 U.S. Navy: 40 Japanese nationals: 450 (paid for by the Government of Japan) U.S. Navy Surface Ship Maintenance Snapshot Surface Ships Maintained in Rota as of Fiscal Year 2018 4 Guided-missile destroyers (DDG) Fiscal Year 2018 Authorized FDRMC Workforce U.S. Civilians: 73 U.S. Navy: 8 Future Considerations The Navy plans to rotate the four current ships back to the United States beginning in 2020 through 2022. FDRMC officials said the next set of ships will not be as standardized as the first four. Additionally, the Senate Armed Services Committee has directed the Navy to assess the feasibility of increasing the number of guided-missile destroyers based in Rota from four to six. FDRMC officials said increasing the number of ships would require additional staff and physical infrastructure that would need to be negotiated with the Spanish government. History and Mission The Forward Deployed Regional Maintenance Center detachment in Bahrain (FDRMC Detachment Bahrain) was established in June 2014. FDRMC Detachment Bahrain manages the maintenance of ships based there, and can provide fleet technical assistance and coordinate voyage repairs for other ships in the U.S. Fifth Fleet area of operations including Military Sealift Command ships and visiting U.S. Navy ships. FDRMC Detachment Bahrain manages the maintenance for the most homeported ships of all overseas locations. U.S. Navy Surface Ship Maintenance Snapshot Surface Ships Maintained in Manama as of Fiscal Year 2018 10 Patrol Coastal Ships (PC) 4 Mine Countermeasures Ships (MCM) Fiscal Year 2018 Authorized FDRMC Workforce U.S. Civilians: 87 U.S. Navy: 29 Foreign nationals: 14 Future Considerations Beginning in fiscal year 2020, the Navy will decommission U.S.-based MCMs to provide spare parts to MCMs overseas. The Navy plans to decommission the MCMs and PCs in Bahrain in fiscal years 2023 and 2026, respectively. The Navy plans to replace the MCM mission with littoral combat ships but has not finalized plans for their deployment or maintenance, according to Navy officials. Appendix VII: Comments from the Department of Defense Appendix VIII: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition the contact named above, the following staff members made key contributions to this report: Suzanne Wren (Assistant Director), Sally Williamson (Analyst in Charge), David Ballard, Martin De Alteriis, Alexandra Gonzalez, Amie Lesser, Shahrzad Nikoo, Carol Petersen, Clarice Ransom, Rachel Schultz, and Samuel Woo. Related GAO Products Navy Maintenance: Persistent and Substantial Ship and Submarine Maintenance Delays Hinder Efforts to Rebuild Readiness. GAO-20-257T. Washington, D.C.: December 4, 2019. Naval Shipyards: Key Actions Remain to Improve Infrastructure to Better Support Navy Operations. GAO-20-64. Washington, D.C.: November 25, 2019. Navy Readiness: Actions Needed to Evaluate the Effectiveness of Changes to Surface Warfare Officer Training. GAO-20-154. Washington, D.C.: November 14, 2019. Military Depots: Actions Needed to Improve Poor Conditions of Facilities and Equipment that Affect Maintenance Timeliness and Efficiency. GAO- 19-242. Washington, D.C.: April 29, 2019. Military Personnel: Strategy Needed to Improve Retention for Experienced Air Force Aircraft Maintainers. GAO-19-160. Washington, D.C.: February 5, 2019. DOD Depot Workforce: Services Need to Assess the Effectiveness of Their Initiatives to Maintain Critical Skills. GAO-19-51. Washington, D.C.: December 14, 2018. Navy and Marine Corps: Rebuilding Ship, Submarine, and Aviation Readiness Will Require Time and Sustained Management Attention. GAO-19-225T. Washington, D.C.: December 12, 2018. Navy Readiness: Actions Needed to Address Costly Maintenance Delays Facing the Attack Submarine Fleet. GAO-19-229. Washington, D.C.: November 19, 2018. Military Readiness: Analysis of Maintenance Delays Needed to Improve Availability of Patriot Equipment for Training. GAO-18-447. Washington, D.C.: June 20, 2018. Navy Shipbuilding: Past Performance Provides Valuable Lessons for Future Investments. GAO-18-238SP. Washington, D.C.: June 6, 2018. Weapon Systems Annual Assessment: Knowledge Gaps Pose Risks to Sustaining Recent Positive Trends. GAO-18-360SP. Washington, D.C.: April 25, 2018. Military Readiness: Clear Policy and Reliable Data Would Help DOD Better Manage Service Members’ Time Away from Home. GAO-18-253. Washington, D.C.: April 25, 2018. Navy Readiness: Actions Needed to Address Persistent Maintenance, Training, and Other Challenges Affecting the Fleet. GAO-17-809T. Washington, D.C.: September 19, 2017. Naval Shipyards: Actions Needed to Improve Poor Conditions that Affect Operations. GAO-17-548. Washington, D.C.: September 12, 2017. Navy Readiness: Actions Needed to Address Persistent Maintenance, Training, and Other Challenges Facing the Fleet. GAO-17-798T. Washington, D.C.: September 7, 2017. Navy Shipbuilding: Policy Changes Needed to Improve the Post-Delivery Process and Ship Quality. GAO-17-418. Washington, D.C.: July 13, 2017. Department of Defense: Actions Needed to Address Five Key Mission Challenges. GAO-17-369. Washington, D.C.: June 13, 2017. Navy Force Structure: Actions Needed to Ensure Proper Size and Composition of Ship Crews. GAO-17-413. Washington, D.C.: May 18, 2017. Navy Ship Maintenance: Action Needed to Maximize New Contracting Strategy’s Potential Benefits. GAO-17-54. Washington, D.C.: November 21, 2016. Military Readiness: Progress and Challenges in Implementing the Navy’s Optimized Fleet Response Plan. GAO-16-466R. Washington, D.C.: May 2, 2016. Defensed Civilian Compensation: DOD and OPM Could Improve the Consistency of DOD’s Eligibility Determinations for Living Quarters Allowances. GAO-15-511. Washington, D.C.: June 16, 2015. Navy Force Structure: Sustainable Plan and Comprehensive Assessment Needed to Mitigate Long-Term Risks to Ships Assigned to Overseas Homeports. GAO-15-329. Washington, D.C.: May 29, 2015.

Why GAO Did This Study To meet operational demands, the Navy has doubled the number of ships based overseas since 2006. Navy ships based abroad represent about 14 percent of the total fleet and are there to provide presence, deter threats, quickly respond to crises, and build partnerships. Effective and timely maintenance is essential to meet strategic objectives, fulfill operational requirements, and ensure ships reach their expected service lives. House Report 115-676 included a provision that GAO assess maintenance for ships based overseas. This report: (1) describes existing maintenance capacity and approaches the Navy uses for surface ships based overseas, (2) assesses the extent to which the Navy completed maintenance periods as scheduled in fiscal years 2014 through 2018 and analyzes factors contributing to any delays, and (3) evaluates the extent to which the Navy has assessed any challenges facing future overseas maintenance efforts. To address these objectives, GAO analyzed Navy policies and maintenance data from fiscal years 2012 through 2018, and interviewed officials, including from Naval Sea Systems Command and overseas fleets and maintenance centers. What GAO Found The Navy maintains the 38 surface ships based in Japan, Spain, and Bahrain through a mix of Navy-operated facilities and private contractors. The Navy uses different maintenance approaches at each location depending on the number and type of ships based there and the Navy and private contractor industrial base available to provide maintenance support. For example, to support the 12 surface ships based in Yokosuka, Japan, the Navy uses both private contractors and its Ship Repair Facility and Japan Regional Maintenance Center, which is subsidized by the government of Japan. In Rota, Spain, the Navy relies on one Spanish contractor to maintain the four ships based at that location. Maintenance on surface ships based overseas took longer than planned for 50 of the 71 maintenance periods—or about 70 percent—started during fiscal years 2014 through 2018. More than half of these maintenance delays lasted a month or longer, which reduced the ships' availability for training and operations. Various factors contribute to delays, such as discovery that unanticipated additional repairs are needed, missed planning milestones, or shortages of key staff. However, the Navy's efforts to understand delays often solely focus on individual maintenance periods and result in steps to improve specific issues related to maintenance timeliness. The Navy has not conducted a comprehensive analysis of maintenance delays to systematically identify and address their root causes. Without such an analysis, the Navy cannot effectively target corrective actions, and risks continuing to underestimate maintenance needs and the time and resources required to address them. The Navy has developed a new maintenance approach for ships in Japan, but has not assessed the risks associated with this approach or analyzed the overseas maintenance requirements for a growing fleet. The new maintenance approach calls for ships to obtain all required maintenance in the United States before and after going overseas, among other things. The Navy decided to implement this approach in Japan based on use of the approach in Spain—where ships have experienced few maintenance delays. However, the Navy has not assessed the risks posed by differences between the operating environments in Spain and Japan, or by shortfalls in maintenance capacity at U.S. facilities. The Navy also plans to replace aging ships in Bahrain as it grows the fleet to 355 ships, but it did not analyze or include overseas maintenance requirements in its long-range plan. Without assessing the risks challenges may pose to the success of its new maintenance approach in Japan or analyzing the requirements of a growing fleet, the Navy could be hindered in its ability to ensure these ships are ready and available for operations. What GAO Recommends GAO is making five recommendations, including that the Navy comprehensively analyze and address maintenance delays, and assess the risks and analyze requirements of future overseas maintenance efforts. The Navy concurred with GAO's recommendations.

gao_GAO-20-83

gao_GAO-20-83_0

VA Established a Program Office and Issued an Action Plan to Advance Health Equity, but Lacks Mechanisms to Measure and Ensure Accountability for Implementation VA has taken several steps in recent years to reduce health care disparities for minority veterans and advance health equity, but lacks mechanisms to ensure accountability for advancing health equity using these or other actions. In particular, VA established a program office and a dedicated steering committee to draft VA’s first action plan designed to achieve health equity. However, this initial action plan lacked performance measures and clear lines of accountability; as such, the extent to which it has been implemented and the progress made in achieving its goals is unknown. VA Established a Dedicated Program Office to Lead VA’s Efforts to Achieve Health Equity In 2012, VA established OHE to lead the department’s efforts to advance health equity and reduce health disparities throughout VA’s health care system. At that time, the Principal Deputy Under Secretary for Health— who reports to the Under Secretary for Health—identified health equity as a cross-cutting issue of the highest priority, and as such, he required the Director of OHE to report directly to him. OHE is responsible for several efforts, including providing education, training, research, communications and information; promoting common awareness about health care disparities and working to improve health care outcomes; and representing VA and serving as a liaison to other governmental and non- governmental organizations working to achieve health equity. OHE was also tasked with developing and maintaining a comprehensive action plan to achieve health equity in VA medical centers and improve VA’s overall quality of care. Also in 2012, VA’s Principal Deputy Under Secretary for Health created the Health Equity Coalition (HEC). Chaired by the Director of OHE, the HEC is a VA-wide steering committee that is comprised of officials from several VA program offices dedicated to areas such as patient care services, communications, research and development, and minority veteran issues. The HEC advises and assists OHE in developing and implementing plans, sets milestones to review progress to ensure timely completion of initiatives, and ensures that program offices commit appropriate organizational resources needed to meet these goals. Since its establishment, OHE has experienced changes in budget and staff levels from year to year. OHE’s core budget, which VA officials told us is spent on salaries, travel, and supplies, fluctuated between 2013 and 2019; staffing levels began decreasing in 2014 and subsequently increased in 2019. (See fig. 1.) VA officials told us that staff reductions were due to transfers and reassignments of OHE staff to other offices and positions, which coincided with a shifted focus from equity issues to other issues such as veteran wait times and modernization efforts. In addition, OHE was also repositioned to a lower level within the organization from VHA’s Office of the Principal Deputy Under Secretary for Health to the Office of the Deputy Under Secretary for Health for Organizational Excellence. VA’s Health Equity Action Plan Set Goals to Eliminate Health Disparities, but Extent of Progress Is Unknown Due to a Lack of Performance Measures and Clear Lines of Accountability In February 2014, VA released its first Health Equity Action Plan (action plan), drafted by OHE and the HEC to document VA’s approach for eliminating health disparities and achieving health equity. The action plan included five focus areas, or goals, in which VA intended to direct its efforts to improve the overall quality of care for all veterans, including minority veterans: Awareness: increase awareness of the significance of health disparities, their impact on the nation, and the actions necessary to improve health care outcomes for racial, ethnic, and underserved populations. Leadership: strengthen and broaden leadership for addressing health disparities at all levels. Health system and life experience: improve health and health care outcomes for racial, ethnic, and underserved populations. Cultural and linguistic competency: improve cultural and linguistic competency, and the diversity of the health-related workforce. Data, research, and evaluation: improve data availability; and coordination, utilization, and diffusion of research and evaluation outcomes. OHE and the HEC included between two to 14 “implementation activities” in each of the five focus areas to describe specific plans and tasks VA could undertake to advance health equity. For example, under the “data, research, and evaluation” focus area, examples of implementation activities included, “identify limitations of existing data, barriers to access to data, and data collection methodologies that affect VA’s ability to describe disparities,” and “develop a strategy for prioritizing identified disparities.” In addition, all five focus areas also included “success criteria” to measure success, a list of resources needed in order implement the activities, and identified stakeholders. However, despite documenting elements needed to make improvements in the five focus areas, the extent of VA’s progress in implementing the action plan and advancing health equity is unknown because the action plan lacks performance measures and clear lines of accountability. Performance measures and clear lines of accountability are among the criteria identified in GAO’s body of work on effectively managing performance under GPRA. Without such mechanisms, VA cannot be assured that the plan has been implemented or will ultimately be effective in addressing health equity. In particular, we found that VA’s plan did not include the following mechanisms: Performance measures. Our past work on effectively managing performance shows that performance measures should demonstrate how well the organization is meeting its goals and should be linked directly to offices that have responsibility for the program or activity. As previously noted, although VA’s action plan included a list of “success criteria”, such criteria were not measurable, and were not directly linked to the specific implementation activities or to the responsible lead office for any of the five focus areas. For example, under the “leadership” focus area, the action plan identified “development of process tools for monitoring in FY 2014” as one of the success criteria, but it was not clearly linked to one of this focus area’s specific implementation activities and did not identify who among the list of lead offices and stakeholders was responsible for it. under the “data, research, and evaluation” focus area, the action plan identified “developed standards for collecting data used to understand disparities” and “improved on-going data sharing between programs” as two of the success criteria, but they were not linked to one of this focus area’s implementation activities, nor were they specifically assigned to one or more of the lead offices and stakeholders listed as responsible for achieving them. Clear lines of accountability: Our past work on effectively managing performance also shows that designating a lead official or office to be responsible for coordinating efforts to achieve results for each goal or action creates clear lines of accountability. This is critical to implementing change to achieve goals and marshaling resources needed to improve management. In contrast, VA’s action plan listed for each of the five focus areas: a broad group of lead offices and stakeholders responsible for the entire focus area, in general, (for example, HEC members and their respective offices, VISN officials, and VA medical center directors) but did not designate specific offices or officials responsible for coordinating efforts to complete specific implementation activities. vaguely described resources—such as leadership support, time, money, and travel—needed to execute all of the implementation activities under each of the five focus areas, but did not designate specific lead offices or stakeholders responsible for committing specific resources needed to implement each activity. Without such performance measures or lines of accountability, VA lacked the means to measure specific progress in implementing and achieving the action plan’s goals. Moreover, according to VA officials, following the release of the action plan and the reduction in number of OHE staff, the frequency of HEC meetings decreased and the last regular meeting before it reconvened in January 2019 occurred in early 2015. As such, VA officials told us that there was no formal involvement or oversight following the release of the action plan to ensure that coalition members were meeting their responsibilities, including committing the organizational resources needed to ensure implementation. In recent years, there have been several recommendations from stakeholders, urging VA to provide OHE with the resources needed to fully implement its action plan. Specifically, the 2016 Commission on Care report recommended that VA commit additional resources to address the causes of the problem and ensure the action plan is fully implemented. The Secretary of VA at the time concurred with the Commission’s recommendation and said that VA would identify health equity leaders and clinical champions in each VISN and VA medical center who could catalyze and monitor actions to implement the action plan and further advance the elimination of health disparities. More recently, VA has signaled renewed interest in supporting the advancement of health equity by increasing OHE’s budget and staffing levels in fiscal year 2019, and reconvening the HEC in January 2019. According to OHE officials, the reconvened HEC has held regular meetings and approved an updated action plan in September 2019. In October 2019, OHE officials told us that the action plan had been sent to VHA leadership for review, which they anticipated would be completed within the first fiscal quarter of 2020. VA Funds Research to Identify Disparities in Health Care Outcomes, but Weaknesses in Race and Ethnicity Data Impede Medical Centers’ Ability to Address Them Both OHE and other VA programs fund research conducted by VA clinicians and staff to identify disparities in health care outcomes. However, VA officials and researchers have noted problems with the completeness and accuracy of the data on veterans’ race and ethnicity. These weaknesses, in turn, limit VA’s ability to assess and address health care disparities at the VA medical center level. VA Funds Research Aimed at Identifying Disparities in Health Care Outcomes Involving Minority Veterans VA funds research aimed at identifying health care outcome disparities involving minority veterans. According to OHE officials, annually OHE receives a research budget separate from its core budget, and can apply monies from this separate budget to any appropriate research activities it wishes to support. As an example, in fiscal year 2019, OHE officials told us it provided funds to the Quality Enhancement Research Initiative, VA’s Center for Health Equity, Research, and Promotion, and two VISNs. In addition to OHE-funded research, VA’s Health Services Research & Development (HSR&D) has spent about $12 million to fund research studies related to identifying and reducing disparities in health care outcomes between minority and other veterans since 2014. This research has identified disparities in health care outcomes for minority veterans. Research funded by HSR&D includes the following studies: A 2017 report focused on whether changes in the way VA delivered primary care were effective in addressing racial and ethnic disparities in health care outcomes. Using VA data from 2009 and 2014, the study found lower rates of control of hypertension and diabetes among veterans who were African American, Hispanic, American Indian/Alaska Native, and Native Hawaiian and other Pacific Islanders compared with White veterans. A 2016 report examined why minority veterans with mental health and substance abuse disorders are less likely to use mental health and substance abuse services, and to complete mental health and substance abuse treatment. The study, which used 2013 data, found health disparities between White veterans and Black, Hispanic, and American Indian or Alaskan Native veterans with mental health and substance abuse disorders on several quality measures, including access to care. The study also found disparities by race and ethnicity in patients’ experiences communicating with providers and office staff. A 2017 systematic review of 351 studies published between 2006 and February 2016 examined the prevalence of disparities in health care outcomes experienced by veterans, including health disparities based on race and ethnicity. This systematic review concluded that a large proportion of the research conducted has focused on differences between Black or African American and White veterans and suggested that future targeted research is needed to capture the unique characteristics of American Indian or Alaska Natives and Native Hawaiian or other Pacific Islanders. Despite VA’s funding of numerous studies to identify health disparities and to explore interventions to potentially reduce or eliminate them, health disparities continue to persist among VA’s patient population. HSR&D officials told us that VA has faced difficulties translating research into practice in clinical settings, including their research findings about disparities in health care outcomes. HSR&D officials told us that they have recently undertaken new efforts aimed at implementing research findings, including those focused on disparities in health care outcomes. Among these efforts is the development of a new program to provide additional funding (for up to two years) for research projects that are completed or close to completion so that researchers can: 1) develop tool kits that others can adopt, and 2) implement research in additional VA medical centers in order to facilitate the sharing of information about successes and failures to make the impact of research more effective. VA Officials and Others Have Identified Weaknesses in Veterans’ Race and Ethnicity Data Generally, VA collects and records race and ethnicity information in veterans’ EHRs when they enroll in VA health care online, by mail, fax, or telephone applications, or through self-service touch-screen kiosks at VA medical centers. Intake clerks may also collect and record race and ethnicity information when assisting veterans with enrollment, as well as when checking a veteran in at a clinic for an outpatient appointment, or registering a veteran for an inpatient hospital admission. However, VA researchers and officials have identified weaknesses in the completeness and accuracy of VA’s patient data on race and ethnicity, which has raised data reliability concerns. (See fig. 2) VA cannot ensure that race and ethnicity information labeled in the EHR as self-reported is accurate. VA follows standards outlined by the Office of Management Budget, which state that self-reported information is the preferred method for obtaining an individual’s race and ethnicity, because it is more accurate than data collected by observation of a third party. However, a VA data expert with HSR&D’s Center for Health Equity Research and Promotion, and officials at one of the VISNs in our review told us that they are aware that intake clerks sometimes enter information based on observation, which may be inaccurate, because they feel uncomfortable asking veterans for their race and ethnicity information in case the veterans find it offensive. Adding further to potential inaccuracies, because VA’s EHR default setting automatically records all race and ethnicity information as self-reported, observational data are only accurately labeled as such if a clerk manually changes the default setting to ‘observational’. The VA data expert from HSR&D’s Center for Health Equity Research and Promotion told us that, based on her research, almost all of the information collected electronically in the EHR is automatically assigned as self-reported, the default setting, even when it is collected by observation of VA staff. This expert also told us that research efforts at VA medical centers have indicated that the default setting is rarely changed and that some clerks had never changed the setting because they do not know how. VA research has indicated that observational data is more accurate for Blacks or African Americans and Whites than other racial groups, and that studies focused on other racial groups may be especially vulnerable to misclassification bias. As such, VA lacks reasonable assurance that the identification of race and ethnicity as “self-reported” is accurate. Data on veterans’ race are often incomplete. Race and ethnicity information is collected as two separate categories in the EHR, and as previously stated, is generally obtained when a patient enrolls in VA health care, or seeks care at a VA medical center or clinic. Two VA researchers told us that ethnicity data—based on veterans’ designation of whether they are Hispanic or non-Hispanic—are often more complete than race data. They said that one reason for this is that the order in which the questions are asked may be problematic; specifically, the ethnicity question is asked first, followed by a second question to request a race designation. Veterans may self-report that they are “Hispanic” upon enrollment or check-in for an inpatient admission or outpatient medical visit, and then leave the race field empty because they believe that they have already provided this information. Missing data on race impedes VA’s ability to identify potential disparities in health care outcomes. Conflicting race and ethnicity information in a veteran’s medical records makes it difficult to determine which information is accurate. According to VA researchers we spoke with, because a patient’s race and ethnicity information is uploaded from his or her EHR after each inpatient admission and outpatient appointment, there can be multiple records for each patient’s race and ethnicity data in VA databases. These patient records often conflict with one another, and may result from the use of both observational and self-reported data. As such, officials stated that it can be difficult to determine which of the multiple race or ethnicity records are accurate. To account for the issues with completeness and accuracy, VA researchers have used various approaches. VA researchers we spoke with told us that while they use data entered into VA’s EHR, which are then uploaded into various databases, they also must use a variety of methods, often time-intensive, to enable the use of race and ethnicity data due to concerns about its completeness and accuracy. These methods include using veterans’ patient records that may be several years old and from multiple VA health care settings, and looking at patient race and ethnicity information captured across multiple years and VA facilities. In addition, researchers also described using multiple non-VA data sources to supplement VA’s race and ethnicity information, such as Medicare data, and data from the Department of Defense’s roster of veterans from recent military operations. VA officials and other stakeholders representing veterans’ interests have recognized the weaknesses in VA’s race and ethnicity data and the importance of improving those data in order to address disparities and improve health equity. For example, VA’s first action plan included a goal to improve data availability, supported by implementation activities to “identify limitations of existing data, barriers to access to data, and data collection methodologies that affect VA’s ability to describe disparities” and “identify strategies for capturing data on race, ethnicity, language…needed to stratify the results for all quality measures and to address disparities.” in 2016, the Commission on Care recommended that VA increase the availability, quality, and use of race, ethnicity, and language data to improve the health of minority veterans, as well as utilize systems that monitor trends in health status, patient satisfaction, and quality measures. in its 2015 annual report, the Advisory Committee on Minority Veterans recommended that VA enhance its existing data collection processes to include the reporting of race and ethnicity data for all benefits and utilization programs to ensure the identification of delivery gaps and potential disparate levels of service. Furthermore, in its 2017 annual report, the Committee again highlighted ongoing concerns with VA’s inconsistency in collecting race and ethnicity data and stated that it impedes VA’s ability to adequately identify health disparities and to ensure minority veterans are receiving quality care and services throughout VA’s system. VA is currently collaborating with the Department of Defense to implement a new EHR system. As yet, they have not yet addressed how the EHR will store race and ethnicity information. The new EHR system is to provide both departments with a common EHR platform that is intended to support the provision of seamless care and create a single health record for service members and veterans. VA officials from the Office of EHR Modernization told us that this collaboration is still in the very early stages and that while race and ethnicity information will be included in the new EHR system, the new EHR will take 10 years to fully implement. VA Officials Have Reported that Data Weaknesses Limit the Ability to Identify and Address Disparities in Health Care Outcomes at VA Medical Centers Data weaknesses, including incomplete and inaccurate data have limited VA’s ability to advance health equity and patient care in its medical centers, according to VA officials. Unlike VA researchers, who report being able to account for missing and inaccurate race and ethnicity data, most VA medical centers do not have the research staff and data specialists needed for these efforts, according to a VA researcher from HSR&D’s Center for Health Equity Research and Promotion and officials from a VISN included in our review. As a result, the inaccurate and missing data have limited the ability of VISN and VA medical center staff to identify and address disparities in health care outcomes by race and ethnicity at the medical center level. VISN officials we spoke to discussed the challenges they encountered when trying to obtain complete and accurate data on health care outcomes by race and ethnicity to identify disparities involving their minority veteran populations. For example, one VISN official told us that she began an effort to analyze disparities in health care outcomes by race and ethnicity in fiscal year 2018, but encountered challenges in obtaining complete and accurate data for minority veterans. She said she contacted OHE officials for assistance, who provided data for diabetes and hypertension by race and ethnicity, but these data were 2 years old and available only at the national level. According to the VISN official, complete and accurate health care outcomes data by race and ethnicity were not available for minority veterans that received care in her region. The official told us that she subsequently contacted both VISN-level and national data specialists, but was still unable to obtain the data to assess health care outcomes by veterans’ race and ethnicity at the regional or local level. other VISN officials we spoke to explained that they had a similar experience when they explored using race and ethnicity data to design a dashboard for a VISN-funded project to track efforts to address disparities in ambulatory care readmissions involving minority veterans. They also told us they too contacted regional and national data specialists, but were told that the readmissions data were missing and inaccurate by race and ethnicity and therefore not useable for their efforts. VA officials told us they are taking steps to provide VA medical centers with data on health care outcomes at their facilities. These officials told us that they are currently developing two health equity dashboards that will use VA’s data on race and ethnicity to provide information on health care outcomes, which would allow VA staff to identify any disparities in these outcomes at the VISN and VA medical center levels. The two health equity dashboards are in different stages of development and, as of September 2019, VA did not have a timeline for completion and implementation across VA medical centers for either dashboard. According to VA officials, the development of these dashboards will not address the accuracy and completeness of the race and ethnicity data in the VA’s EHR. In order to maximize the effectiveness of these dashboards, VA needs to ensure that underlying data weaknesses are addressed by ensuring that race and ethnicity data in the EHR are complete and accurate. VA Collects Patient Experience Feedback from Veterans, including Minority Veterans, through Surveys and VA Medical Center Patient Advocates VA collects patient experience feedback from veterans, including minority veterans, through the following surveys: The Survey of Healthcare Experiences of Patients (SHEP) is VA’s national, standardized, and publically reported patient experience survey that comprises up to 83 questions that are used to collect information about patients’ experiences in various inpatient care settings. The SHEP covers topics to assess patients’ perceptions of their experience using the Consumer Assessment of Health Providers and Systems Survey, which is the health care industry standard. According to VA officials, the response rate for the SHEP is just under 40 percent, and on average, 95 percent of respondents complete survey questions about their race and ethnicity. SHEP survey results are reported publically at the national, VISN, and VA medical center level. VA does not report survey data for specific racial and ethnic groups because, according to VA officials, the number of minority veterans responding to the SHEP is too small to report. In 2012, a memo establishing the OHE recommended that VA send the SHEP survey to a higher number of veterans from racial and ethnic minority groups so there are enough responses to report survey results by those minority groups. However, VA officials told us that they were not aware of this recommendation and had not addressed it. Currently, VA officials told us that VA staff can access SHEP data by race and ethnicity in four broad categories: Hispanic, White, African American or Black, and other, which includes American Indian or Alaskan Native, Asian, and Native Hawaiian or other Pacific Islander racial groups. VA officials told us these data can be accessed on VA’s intranet and are updated on a monthly basis. The Survey of Veteran Enrollees’ Health and Use of Health Care comprises questions about a range of issues, such as enrollee’s health status, insurance, VA and community health care use, and attitudes and perceptions of VA services. The survey is generally conducted on an annual basis, and achieved a 32 percent response rate in 2018. VA publically reports these survey results by race and ethnicity at the national level, and also provides survey results by race and ethnicity for each VISN. For example, one indicator in the 2018 survey results showed that Native Hawaiians were far less satisfied with their ability to get referrals compared to other minority groups. According to VA survey documentation, VA requires a minimum number of survey responses to draw conclusions across the VA enrollee population; the number of survey responses must be aggregated at the VISN level to meet this minimum number. Veterans Signals is a VA survey intended to collect immediate targeted feedback on veterans’ experiences with outpatient services on an ongoing basis. VA officials told us that about one to two million survey invitations are sent out via email each week to veterans who recently received outpatient services, and have provided their email addresses to VA. These short surveys include eight to nine questions and focus on a particular area related to veterans’ recent experiences with VA health care services, such as scheduling appointments, pharmacy wait times, and proficiency of provider communication about veteran concerns during appointments. VA officials told us that the surveys have a response rate of about 20 percent, and of the responses received, 44 percent of respondents provide their race and ethnicity information. VA officials told us that VA medical center staff have access to survey results in real-time and can review results by race and ethnicity for their individual medical centers. In addition to these surveys, VA collects patient experience feedback from veterans, including minority veterans, through its patient advocates located at its medical centers. Each of VA’s 172 VA medical centers is responsible for making at least one patient advocate available to respond to veterans’ feedback and for ensuring feedback is recorded in its Patient Advocate Tracking System (PATS)—an electronic system used to describe and track the resolution of veterans’ feedback across VA medical centers. Patient advocates enter veterans’ feedback in PATS and assign one or more issue codes that generally describe the nature of the feedback. Of the 21 patient advocates we interviewed across 12 VA medical centers, most said they generally do not include race and ethnicity information in PATS when filing a veteran’s complaint, but a few patient advocates said they will include such information if it pertains directly to the complaint. For example, a patient advocate told us she may include race and ethnicity information in PATS in the case of a concern that discrimination occurred. We found that some veteran complaints may not be consistently coded and reported under the correct PATS issue codes in a manner similar to inconsistencies we have identified in prior work. According to VA officials from the Office of Patient Advocacy, two PATS codes were created in 2017 that, in particular, may specifically apply to issues affecting minority veterans: (1) discrimination concerns, and (2) diversity concerns. Of the 21 patient advocates we interviewed about these two specific issue codes, nine were not familiar with or had never used them. VA officials told us that they expected to see patient advocates use these codes more often in 2019 as a result of updates to their patient advocate training curriculum, required for newly hired patient advocates and available to all others. Patient advocates that we interviewed often told us that they review PATS data to report systemic issues to their VA medical center leadership, and a few said they report on a weekly or monthly basis. Additionally, VISN patient advocate coordinators use the PATS data to determine whether there are any trends in PATS data across the medical centers in their networks. Conclusions As one of the nation’s largest health care systems, VA has a unique opportunity to gain a better understanding of the reasons that disparities in health care outcomes occur. VA signaled its commitment to reducing disparities for racial and ethnic minorities and achieving health equity by establishing a responsible program office, creating an action plan, and funding research toward this goal. As the number of minority veterans receiving VA health care services continues to increase, it is important that VA enhances and strengthens its efforts to identify and address disparities in health care outcomes to ensure that all veterans receive equitable care. Despite these efforts, however, without including performance measures or lines of accountability, VA lacks the means to ensure any action plan will be fully implemented to achieve its goals. Further, weaknesses in race and ethnicity data due to problems with the completeness and accuracy continue to limit VA’s ability to identify and address disparities in health care outcomes at the VA medical center level. Although VA is developing equity dashboards to provide health care outcomes data by race and ethnicity at the VA medical center level, these efforts will not improve the completeness and accuracy of the race and ethnicity data in VA’s EHR. Until VA resolves known weaknesses with the completeness and accuracy of its race and ethnicity data, it will be limited in its ability to assess health equity for veterans receiving care at its facilities. Recommendations for Executive Action We make the following two recommendations to VA: The Under Secretary for Health should ensure that any action plan for achieving health equity includes key elements for successful implementation by consistently applying criteria identified in GAO’s past work on effectively managing performance, including developing performance measures to assess progress and creating clear lines of accountability by designating specific offices or officials with responsibility for coordinating efforts to implement actions and committing resources necessary for achieving its goals and objectives. (Recommendation 1) To ensure the availability of information on health care outcomes by race and ethnicity throughout the VA health care system, the Secretary of Veterans Affairs should conduct an assessment to determine the completeness and accuracy of race and ethnicity data captured in VA’s electronic health record, and implement corrective actions as necessary to resolve any identified deficiencies. (Recommendation 2) Agency Comments We provided a draft of this report to the Department of Veterans Affairs and the Department of Defense for review and comment. The Department of Defense did not have comments. VA provided written comments, which are reprinted in appendix I. VA concurred with both of our recommendations—that any health equity action plan should include performance measures to assess progress and that VA should resolve weaknesses identified with the completeness and accuracy of race and ethnicity data. VA further provided information on how the agency intends to address our recommendations, with targeted completion dates of December 2020 and June 2021, respectively. We are sending copies of this report to the appropriate congressional committees, the Secretary of Veterans Affairs and the Under Secretary for Health, and the Secretary of the Department of Defense. In addition, the report is also available at no charge on GAO’s website at http://www.gao.gov. If you or your staff has any questions regarding this report, please contact me at (202) 512-7114 or draperd@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. Appendix I: Comments from the Department of Veterans Affairs Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Ann Tynan, Assistant Director; Michelle Paluga, Analyst-in-Charge; Jennie Apter, Romonda McKinney Bumpus, and Phil Steinberg made key contributions to this report. Also contributing were Kaitlin Farquharson and Ethiene Salgado-Rodriguez.

Why GAO Did This Study According to VA, in 2016, racial and ethnic minority veterans represented about 22 percent of the total veteran population of 18.6 million. VA projects racial and ethnic minority veterans will make up 36 percent of its total veteran population by 2040. VA has identified racial and ethnic disparities in its health care outcomes, mirroring trends seen across the United States. House Report 115-188 included a provision for GAO to review whether VHA provides quality, equitable care for minority veterans. GAO's report examines, among other issues, (1) the extent to which VA has taken steps to advance health equity for racial and ethnic minority veterans, and (2) VA's efforts to use race and ethnicity data to identify and address disparities in health care outcomes involving minority veterans. GAO reviewed relevant documents, such as strategic and operational plans and peer-reviewed research studies; assessed VA's health equity action plan against criteria identified in GAO's body of work on effectively managing performance; and interviewed VA officials familiar with VA's health equity efforts, as well as race and ethnicity data. What GAO Found The Department of Veterans Affairs (VA) has taken steps to reduce disparities in health care outcomes linked to race and ethnicity, but lacks mechanisms to measure progress and ensure accountability for results. In 2012, VA established the Office of Health Equity to identify and address health care outcome disparities and to develop an action plan to achieve health equity. This office issued an action plan in 2014 that identified activities to make improvements in five focus areas, such as increasing awareness of the significance of disparities and strengthening leadership for addressing them. However, GAO found that the extent of VA's progress in implementing the action plan and advancing health equity is unknown because the action plan lacked performance measures and clear lines of accountability for specific offices. For example, although VA's action plan included a list of “success criteria” for each of the five focus areas, these criteria were not measurable, and were not linked to specific activities or to offices responsible for implementation. VA funds research efforts that have identified disparities in health care outcomes involving minority veterans, but rely on data that VA officials and researchers noted have weaknesses in completeness and accuracy. One concern is that race and ethnicity information can be labeled incorrectly in VA patients' electronic health records as ”self-reported”, a highly reliable method of collection, when data were actually collected based on the less reliable method of VA staff observation. Other reported concerns include missing values on patients' race and conflicting race and ethnicity information. VA researchers told GAO they account for some of these concerns by using data from other sources, such as Medicare, but such work-arounds are time intensive. Further, VA officials reported that data weaknesses limit their ability to identify and address disparities in health care outcomes in their medical centers. Despite recognizing weaknesses related to the quality of race and ethnicity data, VA has not implemented corrective actions to address them. Without doing so, VA medical center officials cannot readily identify and address disparities in health care outcomes by race and ethnicity. Note: Concerns about the completeness and accuracy of race and ethnicity information were raised by officials from VA's Office of Health Equity, Veterans Experience Office, and Health Services Research & Development. What GAO Recommends GAO is making two recommendations to VA to (1) ensure that any health equity action plan includes performance measures to assess progress, and clear lines of accountability designating responsibility to specific offices, and (2) conduct an assessment to determine how to address weaknesses identified with the completeness and accuracy of race and ethnicity data in the electronic health record, and implement corrective actions as necessary. VA agreed with GAO's recommendations.

gao_GAO-20-325

gao_GAO-20-325_0

Background Phases of Technological Innovation In our prior work, we have found that technological innovation involves not only creating new ideas but also translating those ideas into a new product or service. Innovation, and the research driving it, is inherently risky because the likelihood that research can be translated into a product or service and the ultimate value of that product or service are unknown. The Department of Commerce’s National Institute of Standards and Technology describes the path from innovation to commercialization as comprised of three overarching stages: inventing, transitioning to making, and selling. (See fig. 1 for a description of the path from innovation to commercialization.) Federal Agencies’ Oversight of the Food Supply FDA and USDA have responsibility for overseeing the safety of the food supply. In general, FDA is responsible for ensuring the safety of virtually all domestic and imported food products except those regulated by USDA. USDA is responsible for ensuring the safety of meat, poultry, processed egg products, and catfish. FDA and USDA cooperate with states, tribes, and local food safety and public health agencies to carry out their federal responsibilities. FDA and USDA carry out their responsibilities in part through inspections of facilities where food is produced. The frequency of inspections the agencies conduct varies, as follows: FDA. FDA’s authority requires a risk-based approach, in which inspection rates vary depending on the level of risk associated with a food product. FDA conducts risk-based inspections of high-risk and non-high-risk food facilities. For example, the FDA Food Safety Modernization Act, signed into law in 2011, specified that FDA had to inspect all high-risk domestic facilities at least every 3 years. USDA. Depending on the type of facility, USDA conducts inspections at least once per operating shift or maintains a constant presence. Specifically, USDA conducts carcass-by-carcass inspection at all federally inspected meat and poultry slaughter facilities and verifies that these establishments follow all food safety and humane handling requirements. At facilities that process meat and poultry products, USDA conducts inspections at least once per production shift, following the agency’s longstanding interpretation of its statutes requiring it to do so. Among other things, the Federal Food, Drug, and Cosmetic Act requires that food additives be approved by FDA before they can be lawfully used in foods. Substances added to food are considered unsafe unless the agency establishes that the use of the food additive, under specific conditions for use, will be safe, or unless the substance is generally recognized as safe (GRAS) under the conditions of its intended use among qualified experts. As we reported in 2010, the Federal Food, Drug, and Cosmetic Act exempts GRAS substances from the act’s general requirement that companies obtain FDA approval before marketing food containing a new additive. GRAS substances include hundreds of spices and artificial flavors, emulsifiers and binders, vitamins and minerals, and preservatives that manufacturers add to enhance a food’s taste, texture, nutritional content, or shelf life. The GRAS exemption allows companies, without notice to or approval from FDA, to determine whether there is enough support to claim a substance is GRAS. For a company to claim a substance is GRAS, it must conclude that there is common knowledge about the safety of the substance among experts qualified by scientific training and experience to evaluate its safety. In addition, as part of their oversight of the food supply, FDA and USDA oversee food labeling of the products under their respective jurisdictions. USDA, by statute, is charged with assuring that products under its jurisdiction, including meat, poultry, and catfish, in interstate or foreign commerce are properly marked, labeled, and packaged. USDA develops and applies the labeling requirements for these products, and food manufacturers are responsible for complying with the USDA labeling rules and adhering to the process maintained by USDA for the evaluation and approval of these product labels. Consistent with its statutes, USDA requires preapproval of all labels before manufacturers can market their products. The Federal Food, Drug, and Cosmetic Act prohibits the misbranding of food, which includes food labeling that is false or misleading. Consistent with its statutes, FDA ensures that foods within its jurisdiction are not misbranded by focusing on the labels of products already in the market. FDA establishes regulations for the enforcement of these provisions and issues guidance. Food manufacturers are responsible for compliance with misbranding provisions in the Federal Food, Drug, and Cosmetic Act and its implementing regulations. Regulation of New Food Technologies From time to time, new technologies, such as those used to make cell- cultured meat, generate challenges for FDA’s and USDA’s regulatory structure. Other examples of new food technologies to which federal agencies have needed to adapt include the genetic modification of plants and irradiation of foods. In the case of genetically modified plants, there are no specific regulations addressing products resulting from the manipulation of the genetic material of living seeds. However, under FDA policy, new genetically engineered crop varieties are treated like other foods (including their conventional counterparts) under the Federal Food Drug and Cosmetic Act and may not contain either unapproved food additives or contaminants that would adulterate the food. In 1995, FDA established a voluntary pre-market consultation process through which companies are encouraged to notify the agency before marketing a food produced from a genetically modified crop and voluntarily submit a summary of the developer-performed safety assessment. FDA evaluates the safety assessment for any issues that need to be addressed and works with the developer to resolve those issues. In the case of irradiated foods, companies seeking approval for a source of radiation used to treat a food may submit a food additive petition to FDA demonstrating the safety of the proposed use. FDA grants approval only after agency scientists have determined that the proposed use is safe, then the process can be employed commercially. Specific Information about Cell-Cultured Meat Technology, Commercial Production Methods, and Final Product Composition Is Not Yet Known General information about the process of making cell-cultured meat is available, but specific information about the technology being used and the eventual commercial production methods as well as the final products is not yet known. While firms may vary in how they make cell-cultured meat, the general process they use can be described in five phases. However, the technology and methods to commercially produce cell- cultured meat are still in development, and producers, regulators, and consumers do not yet have clarity on what these will entail. The composition of the final product is also not yet known. The General Process for Making Cell-Cultured Meat Contains Five Phases The general process for making cell-cultured meat contains five phases: biopsy, cell banking, growth, harvest, and food processing. (See fig. 2.) The five-phase process is generally as follows: 1. Biopsy. A biopsy is taken by collecting rice-sized tissue samples from an animal, such as livestock, chicken, or fish. During this and subsequent phases, specific laboratory sanitation procedures are followed, and antibiotics may be used in order to avoid or minimize contamination from bacteria. Growth Media According to researchers and representatives from cell-cultured meat firms, the growth media for cell-cultured meat often contains fetal bovine serum, which is obtained from blood drawn from a bovine fetus at slaughter. However, researchers and representatives from cell-culturing firms we spoke with said they are working to develop growth media that do not contain fetal bovine serum. Representatives from some of these firms also told us that the composition of the growth media, including the exact ingredients and their proportions, can vary based on the specific needs of the cells and the variety of serum used. For example, cell-cultured seafood may have different growth media and environmental requirements than cell-cultured livestock and poultry. 2. Cell banking. Biopsied cells with the most desirable traits are selected and either used immediately for cell growth or frozen to create a cell bank for later use. These desirable traits can be obtained by either selecting existing cells or using genetic engineering methods to insert, delete, or edit the DNA to target desired traits in cells. Examples of desirable traits may include cells that divide quickly, cells that divide a greater number of times, cells that result in a reduced cholesterol or fat content or other desirable nutritional traits, or cells that are more resilient to environmental factors, such as temperature, than other cells. According to agency officials and representatives from cell-cultured meat firms, this phase represents an important opportunity to ensure that the source cells used to initiate commercial production are free of pathogens or other contaminants. 3. Growth. During the cell growth phase, cells are placed in a bioreactor and begin to divide and differentiate. A bioreactor is a container that creates an environment that can sustain the growth of cells and includes the ability to control factors such as temperature, pH, and oxygen and carbon dioxide concentrations. Bioreactors can vary in size, including microwave-sized and refrigerator-sized units, but could be as large as 20 to 30 feet tall in commercial production. Bioreactors contain a growth medium, which may include ingredients such as glucose, amino acids, hormones and other growth factors, and other basic nutrients that cells need to consume in order to thrive. In addition to the medium needed for growth, the cells may need to be attached to a structure, referred to as a scaffold, to properly develop into cell-cultured meat. 4. Harvest. Once the cells have divided to form a sufficiently large amount of cell-cultured meat, producers remove—or harvest—it from the growth medium and bioreactor. If a scaffold was used to provide a structure for cells to grow on, then the cell-cultured meat would either be separated from the scaffold during harvesting or left attached to an edible scaffold. 5. Food processing. The harvested cell-cultured meat is then prepared into a product such as meatballs or chicken nuggets. In the future, products similar to intact cuts of meat such as steak or chicken breast may be produced. The Technology to Commercially Produce Cell-Cultured Meat Is in Development, and the Eventual Commercial Production Methods and Aspects of a Final Product Are Not Yet Known The technology to produce cell-cultured meat at a commercial scale is still in development, and information about the methods to be used for commercial production and the composition of the final product are not yet known. In the continuum of moving a technology from innovation to commercialization, cell-cultured meat firms are in the middle stage of building and testing their prototypes, based on our discussions with representatives from these firms. Consequently, they have not finalized aspects of the technology and eventual commercial production methods to be used or the composition of the final product. As a result, certain information is not yet available to stakeholders—including cell-cultured meat firms themselves, regulators, and the public—about specific aspects of the technology and commercial production methods that will be used, such as the composition of the growth medium and of the final products. In addition to technology development, the scarcity of publicly available research on cell-cultured meat production limits information available to agency officials and the public. Each cell-cultured meat firm is developing detailed information on its own eventual commercial production methods for making cell-cultured meat. However, the firms, similar to other technology start-ups, are reluctant to disclose intellectual property and business-sensitive information due to concerns about competition. For example, one firm told us that they can reverse engineer parts of another company’s commercial production method by seeing pictures of the equipment the other company is using. In addition, cell-cultured meat firms compete with other firms for funding from sources such as venture capitalists, foreign governments, and conventional meat companies. This competition for funding contributes to firms being reluctant to share information they consider important intellectual property, such as parts of their production processes. As a result, agency officials and other stakeholders told us that they must largely rely on whatever information the cell-cultured meat firms are willing to provide to understand details of the companies’ prototype processes and products. This limitation can affect agencies’ ability to make regulatory and other decisions. Specifically, FDA and USDA officials said they have limited information on cell-cultured meat production methods and products and need more in order to regulate this new food. One USDA official explained that the agency cannot establish labeling requirements if the agency does not know the nutritional profile of the final product. For example, if the scaffold on which the cell-cultured meat is grown is not edible, the agencies may require firms to disclose certain aspects of their commercial production methods, such as how they removed the cell- cultured meat from the scaffold. However, if the scaffold is edible, it will affect the final composition of the product, which may require different labeling than a product that was developed without edible scaffolding. This lack of information results in unanswered questions about cell- cultured meat as it relates to the eventual technology and commercial production methods to be used and the composition of the final products. Among other things, this lack of information creates challenges for industry and federal regulatory agencies as cell-cultured meat nears commercialization. The sources we reviewed and stakeholders we talked to identified a number of open questions, including the following: Tissue collection. How often will producers need to collect biopsy samples from animals, and what animals will be used? Some stakeholders have stated concerns about whether, and how, regulators will ensure that biopsies are collected from healthy animals. For example, one cell-cultured meat firm stated that tissue samples would be taken from slaughtered donor animals that met federal standards for conventional processing at the time of slaughter. However, USDA and FDA have not indicated whether they would require cell-cultured meat firms to do so. Additionally, representatives from cell-cultured meat firms stated that they did not yet know how frequently they would need to collect biopsies from animals for commercial-level production. Additionally, according to researchers, there are too many unknowns to accurately estimate how much cell- cultured meat could be produced from a single biopsy of animal tissue. Genetic engineering. Will commercial production methods involve genetic engineering? Some stakeholders expressed concern that the use of genetic engineering in cell-cultured meat production could cause the product to experience a lengthy wait for regulatory approval, similar to that for genetically engineered salmon, which took approximately 20 years. One representative from a cell-cultured meat firm noted that uncertainty about pending government regulations could negatively affect firms’ ability to attract and retain investors. Representatives from some firms said understanding what regulatory requirements will look like might influence which scientific pathways they pursue as they continue to develop their commercial production methods. According to FDA officials and representatives from one cell-cultured meat firm, it is likely that some firms will use genetic engineering in their commercial cell-cultured meat production methods. However, representatives from two other cell-cultured meat firms told us they were undecided as to whether they would use genetic engineering in their commercial production methods. Antibiotics. Will antibiotics be used to make cell-cultured meat, and will residues be present in the final product? According to agency officials, the presence of antibiotics in commercial production and the potential for residues in the resulting product would represent a significant potential concern for food safety and public health. Officials stated that they would not expect antibiotics to be used past the cell- banking phase. Representatives from cell-cultured meat firms we spoke to differed on whether they planned to use antibiotics in their commercial production process, but they had not finalized their decisions. According to one firm, if antibiotics are used, the use would be limited both in quantity and duration. Growth medium. What type of growth medium will producers use, and how might variations in the media affect the final product? According to agency officials and other stakeholders, the ingredients used in the growth medium could affect the end product’s composition and raise potential safety concerns. For example, FDA officials stated that residual growth factors, such as hormones, in the final product would be something they would likely evaluate in premarket consultations. However, representatives from cell-cultured meat firms stated that their firms have not finalized the medium they plan to use. In addition, the formulation of the medium firms use could be an important piece of intellectual property or confidential business information. Scaffold. What type of scaffold will producers use, if any, and will it be edible or inedible? The use of edible or food-grade scaffolds, where they are used, will affect the composition of the product and may need to be evaluated by federal agencies for safety. According to USDA officials, the composition of edible scaffolding may also create labeling and jurisdictional concerns. For example, USDA officials stated that the addition of edible scaffolding may require significant additional aspects of production to be subject to USDA jurisdiction. Additionally, researchers have commented that a chemical separation technique needed to separate some inedible scaffolds may also need to be evaluated for potential safety concerns. Point of harvest. How will FDA and USDA define the point of harvest? The point of harvest is the point at which FDA will transfer oversight responsibilities, including inspections, to USDA. Stakeholders have raised concerns that not having a clear definition of the point of harvest could lead to challenges such as overlapping inspection requirements or a gap in inspection. Representatives from several cell-cultured meat firms we spoke to in the spring of 2019 said it was ambiguous how FDA and USDA intended to define the point of harvest. These representatives also said it is unclear how often each agency plans to conduct inspections during the phases for which it is responsible. Agency officials stated that they are working to develop a detailed process for the transfer of jurisdiction, including defining the point of harvest. Scaling up production. How will firms scale up production to commercial levels? One 2018 study conducted by researchers in the United Kingdom stated that to produce one pound of cell-cultured meat, firms would need bioreactors at least 2 1/2 times larger than what is currently available. Similarly, a senior FDA official stated that the capacity of existing production equipment is a challenge for firms seeking to produce cell-cultured meat products at a commercial scale. As a result, the firms themselves may have to develop the equipment or custom order such equipment. Representatives from one cell- cultured meat firm told us that they are interacting with equipment providers to identify commercial-scale production equipment. Production cost. How will firms sell their product at a price point that is both profitable to the firms and affordable to the consumer? Some studies and stakeholders we interviewed, including representatives from cell-cultured meat firms, said that the high production cost of cell- cultured meat is a key industry challenge. For example, in the last two years, one firm reported that it cost $600 to make a cell-cultured meat hamburger patty and reported that it cost about $1,200 to produce a single cell-cultured meatball. One of the biggest cost drivers in the production of cell-cultured meat is the growth medium, according to some studies and some cell-cultured meat firms. To address issues of cost and scale, some firms may develop their own, less expensive growth media. Safety considerations. Are potential safety hazards in commercial production methods for cell-cultured meat different from those for conventional meat, and how will eventual commercial production methods affect the overall safety of the product? According to agency officials, cell-cultured meat may present different safety challenges compared to conventional meat. For example, according to agency officials, residues and constituents in harvested cell-cultured meat would be expected to be different from those in conventional meat, depending on the details of the production process. Representatives from one cell-cultured meat firm told us that they likely will use food processing techniques similar to those used for conventional meat, abide by similar health and safety standards, and possibly share food processing facilities. However, because specific information about commercial production methods and final products is not yet known, it is unclear whether cell-cultured meat produced on a commercial scale will pose any hazards not present in conventional meat. Product composition. What will be the composition of any eventual products? Agency officials told us that without knowing the composition of a cell-cultured meat product, it is impossible to predict how food safety and labeling requirements will apply. According to representatives from some cell-cultured meat firms, initial cell-cultured meat products most likely will not be composed entirely of cell- cultured meat but, rather, a mixture of cell-cultured meat and other ingredients such as binding, flavoring ingredients, and plant-based materials used in conventional food products. Some firms have developed prototypes of cell-cultured meat products as part of their research and development. In April 2019, representatives from one firm told us that their prototype included about 90 percent plant-based ingredients and 10 percent cell-cultured meat. However, representatives from cell-cultured meat firms stated that they aim to produce products that contain more cell-cultured meat than other ingredients. For example, some cell-cultured meat firms have stated that a long-term goal is to commercially produce cell-cultured meat products that are similar to intact cuts of meat, such as steaks. As of December 2019, these firms had not provided regulators with specific information detailing the composition of their cell-cultured meat prototypes, according to FDA and USDA officials. Environmental, animal welfare, and health impacts. How will cell- cultured meat impact the environment, animal welfare, or human health, if at all? Cell-cultured meat firms and researchers have made various claims about the potential environmental, animal welfare, and health advantages of cell-cultured meat over conventionally produced meat. For example, some cell-cultured meat firms have claimed that cell-cultured meat production would use less water and emit less greenhouse gases than conventional meat production. Some cell- cultured meat firms have also claimed that cell-cultured meat will improve animal welfare because slaughter will be unnecessary. Additionally, some stakeholders stated that because there is less opportunity for contamination from animal feces—a potential source of contamination for conventional meat—cell-cultured meat would be less likely than conventional meat to contain foodborne pathogens. However, there are disagreements regarding the accuracy of these claims. Stakeholders told us that until commercial production methods and final products are established, these claims about impacts on the environment, animal welfare, and human health will remain unsubstantiated. Timeline to market. When will cell-cultured meat products reach consumers? As of December 2019, no cell-cultured meat products were available for retail sale in the United States. Stakeholders give varying estimates for when cell-cultured meat may be commercially available. Some estimates suggest that firms may be able to commercially produce some form of cell-cultured meat product as soon as 2020, while others estimate that such products may not be available for 2 to 4 years. Labeling. How will cell-cultured meat be labeled? Labeling was an area of concern for representatives from both conventional and cell- cultured meat firms who explained that the specific terminology, such as “clean meat” or “lab-grown meat,” can sometimes reflect bias for, or against, certain products, potentially affecting consumer acceptance of these products. Additionally, stakeholders, as well as agency officials, have emphasized the importance of labeling to ensure consumers have accurate information about what they are buying. For example, in February 2018 the United States Cattlemen’s Association submitted a petition to USDA requesting that the agency limit the term “beef” to products “born, raised, and harvested in a traditional manner” and “meat” to mean the “tissue or flesh of animals that have been harvested in the traditional manner.” USDA received over 6,000 comments on the petition, and the agency had not responded to the petition as of December 2019. However, according to agency officials, USDA has committed to a public process, likely rulemaking, for the development of labeling requirements for cell- cultured meat and poultry. In addition, in recent years, a number of states have passed laws that could affect the labeling of cell-cultured meat when it comes to market. For example, in 2018, Missouri enacted a law to prohibit plant-based products and cell-cultured meat from being labeled as “meat.” Consumer Acceptance How will consumers respond to cell-cultured meat? It remains unclear whether consumers will embrace and purchase cell-cultured meat products. Stakeholders we interviewed and studies we reviewed cited consumer acceptance as a challenge for commercializing cell-cultured meat. One study noted that consumers have both positive and negative views toward cell-cultured meat, which could impact their willingness to purchase and consume such products. FDA and USDA Have Begun to Collaborate on Oversight of Most Cell-Cultured Meat but Could More Fully Incorporate Leading Collaboration Practices FDA and USDA have established multiple mechanisms to collaborate on regulatory oversight of cell-cultured meat. Specifically, the agencies have collaborated through a joint public meeting, an interagency agreement, and three working groups. However, the interagency agreement and working groups, which are ongoing mechanisms, do not fully incorporate leading practices for interagency collaboration. In addition, FDA and USDA have not documented which agency will oversee cell-cultured seafood not covered by the interagency agreement. FDA and USDA Have Taken Initial Steps to Collaborate on Their Oversight of Cell-Cultured Meat through Several Mechanisms In 2018, FDA and USDA began taking steps to collaborate on the regulatory oversight of cell-cultured meat through several mechanisms: a joint public meeting, an interagency agreement, and three working groups. The agencies held the joint meeting in October 2018 to discuss the use of cell-culture technology to develop products derived from livestock and poultry, and topics included potential hazards, oversight considerations, and labeling. As part of this meeting, FDA and USDA held an open public comment period from September through December 2018, gathered 315 written comments, and offered interested parties the opportunity to offer comments in person. The agencies received public comments from members of the public, as well as from representatives from cell-cultured meat and conventional meat industries, food and consumer safety groups, animal welfare groups, and environmental organizations, among others. The written comments the agencies received focused on such topics as environmental considerations, labeling, potential health and safety implications, and potential regulatory and inspection processes. Stakeholders also presented multiple perspectives on these issues at the meeting. For example, stakeholders expressed different views as to whether cell-cultured meat should be regulated as a food additive, considered a GRAS substance, or whether new regulations were needed. In March 2019, FDA and USDA issued a formal interagency agreement that describes the intended roles and responsibilities of each agency in overseeing cell-cultured meat. The agreement establishes the following: Oversight. FDA will oversee the early phases of growing cell-cultured meat through the point of harvest. During harvest, FDA will work with USDA to transfer regulatory oversight to USDA. USDA will then assume oversight of cell-cultured meat through the food processing phase, including labeling, as shown in figure 3. Types of meat covered. The agreement covers cell-cultured meat derived from species overseen by USDA, such as livestock, poultry, and catfish. Future actions. The agreement also details future actions the agencies plan to take, such as developing a more detailed regulatory framework or standard operating procedures and developing joint principles for product labeling. Reviewing and updating the agreement. The agreement states that the agencies have the ability to modify it as needed and will review the agreement every 3 years to determine whether they should modify or terminate it. In June 2019, FDA and USDA created three working groups to carry out the terms of the interagency agreement. The working groups are comprised of FDA and USDA officials and operate independently, though some individuals are members of multiple groups. The groups are as follows: Pre-market assessment working group. Led by FDA, this group was created to clarify the process FDA will use for pre-market reviews of cell-cultured meat. Labeling working group. Led by USDA, this group will focus on developing joint principles for product labeling and claims. Transfer of jurisdiction working group. Co-led by FDA and USDA, this group will develop procedures for the transfer of inspection at harvest, among other things. According to agency officials, the working groups are still in the initial phases of development, though some have progressed further than others. For example, as of December 2019, the pre-market assessment and labeling groups had met and begun to address various areas, while the transfer of jurisdiction working group was still in discussions to outline the roles, responsibilities, and outcomes for the group and had not held a formal meeting. The Interagency Agreement and Working Groups on Cell-Cultured Meat Could More Fully Incorporate Leading Collaboration Practices FDA and USDA could more fully incorporate leading practices for collaboration in their interagency agreement and working groups. We have previously reported that interagency mechanisms or strategies to coordinate programs that address crosscutting issues may reduce potentially duplicative, overlapping, and fragmented efforts. In addition, while collaborative mechanisms may differ in complexity and scope, they all benefit from certain leading practices, which raise issues to consider when implementing these mechanisms. We compared the agencies’ interagency agreement and working groups with the seven leading practices to enhance and sustain interagency collaboration that we previously identified. These leading practices, and examples of the associated issues to consider, are as follows: Defining outcomes and monitoring accountability. Is there a way to track and monitor progress toward short-term and long-term outcomes? Do participating agencies have collaboration-related competencies or performance standards against which individual performance can be evaluated? Bridging organizational cultures. What are the commonalities between the participating agencies’ missions and cultures, and what are some potential challenges? Have participating agencies developed ways for operating across agency boundaries? Have participating agencies agreed on common terminology and definitions? Identifying and sustaining leadership. How will leadership be sustained over the long term? If leadership is shared, have roles and responsibilities been clearly identified and agreed upon? Clarifying roles and responsibilities. Have participating agencies clarified roles and responsibilities? Have participating agencies articulated and agreed to a process for making and enforcing decisions? Including relevant participants. Have all relevant participants been included? Do participants have appropriate knowledge, skills, and abilities to contribute? Identifying and leveraging resources. How will the collaborative mechanism be funded and staffed? Developing and updating written guidance and agreements. If appropriate, have the participating agencies documented their agreement regarding how they will collaborate? (A written document can incorporate agreements reached in any or all of the following areas: leadership, accountability, roles and responsibilities, and resources.) Have participating agencies developed ways to continually update or monitor written agreements? See appendix II for a full list of the associated issues to consider for each leading practice. Interagency Agreement Partially Incorporates All Seven Leading Practices We found that the interagency agreement for oversight of cell-cultured meat partially incorporates all seven leading practices for collaboration. For example: Defining outcomes and monitoring accountability. The interagency agreement partially incorporates the leading practice of defining outcomes and monitoring progress toward these outcomes. Specifically, the agreement identifies broad outcomes such as the development of labeling principles. However, the agreement does not describe how the agencies will track and monitor progress toward outcomes. Identifying and sustaining leadership. The agreement partially incorporates the leading practice of clarifying leadership structures. For example, it assigns each agency as the lead, or designates shared leadership, for different phases of the cell-cultured meat production process. However, the interagency agreement does not identify how the agencies will sustain leadership over the long term, including through succession planning. We have previously reported that given the importance of leadership to any collaborative effort, transitions and inconsistent leadership can weaken the effectiveness of any collaborative mechanism. Developing and updating written guidance and agreements. The agreement partially incorporates the leading practice of documenting how the agencies will collaborate. For example, the agreement includes a method for updating the document by including a provision that requires a review of the document every 3 years. This is consistent with our leading collaboration practice to continually update or monitor written agreements. However, the interagency agreement does not document how the agencies will track and monitor progress toward short-term and long-term outcomes. Table 1 provides more detail about the agencies’ incorporation of these leading collaboration practices in their interagency agreement. FDA and USDA officials told us that the interagency agreement was intended to be an initial, general outline for their collaboration. They also said that as the technology to produce cell-cultured meat develops and they implement the agreement, including developing the content of a regulatory program, they will consider incorporating leading practices for interagency collaboration. For example: Clarifying roles and responsibilities. FDA and USDA officials said in December 2019 that through the working groups the agencies would continue to explore and define the specific details of how they will manage their shared oversight responsibility. Including relevant participants. FDA officials said in December 2019 that the agency would like to engage many more stakeholders as it continues to develop its oversight of cell-cultured meat. Identifying and leveraging resources. As of December 2019, the pre-market assessment working group and the labeling working group were working to identify any human resources, physical, or financial resources they might need, according to FDA and USDA officials. The federal food safety system is on our High Risk List due to concerns about fragmentation, which we have reported has caused inconsistent oversight, ineffective coordination, and inefficient use of resources. As the agencies continue to collaborate on their shared oversight of cell- cultured meat, by more fully incorporating all seven leading practices for collaboration into their interagency agreement, they will be better positioned to address potential fragmentation in their efforts to ensure the safety of the food supply as cell-cultured meat products near commercialization and entry into the marketplace. Working Groups Partially Incorporate or Do Not Incorporate Leading Practices We found that the pre-market assessment, labeling, and transfer of jurisdiction working groups that FDA and USDA created to carry out the terms of the interagency agreement either partially incorporate or do not incorporate the seven leading practices for interagency collaboration. Specifically, all three working groups have partially incorporated three of the seven leading practices for collaboration, but none of the working groups have incorporated the four remaining leading practices. For example: Defining outcomes and monitoring accountability. The working groups have all defined and agreed upon their general purposes. However, FDA and USDA have not established methods, such as milestones and metrics, to evaluate the progress of any of the working groups. For example, FDA officials said in December 2019 that their next steps are to conduct a general and qualitative risk assessment of animal cell culture food technology to systematically identify particular areas of interest from a food safety perspective and prepare detailed procedural guidelines for cell-cultured meat firms to follow. However, the officials did not have time frames or a method to evaluate progress towards completing these actions. Including relevant participants. While the working groups have included relevant FDA and USDA officials, none of the groups have included state or tribal officials in initial discussions and planning. According to the state officials we spoke with, being excluded from these federal-level discussions may hinder their ability to align their safety and labeling requirements, among other things, with federal standards. Developing and updating written guidance and agreements. None of the working groups have documented how they will collaborate. For example, the working groups have not documented leadership, accountability, roles and responsibilities, or resources needed for working groups. Table 2 provides more detail about FDA and USDA’s incorporation of leading collaboration practices in the three working groups. In December 2019, FDA and USDA officials said that as they continued to stand up these working groups, they were considering leading practices for collaboration. For example: Defining outcomes and monitoring accountability. FDA and USDA officials said they were considering means to monitor, evaluate, or report on the results of the pre-market assessment working group. Including relevant participants. FDA and USDA officials said that they were working to determine what knowledge participants in the pre-market assessment working group and the labeling working group needed to perform the work of the working group. Developing and updating written guidance and agreements. FDA and USDA officials said they were considering documenting how they will collaborate in the pre-market assessment working group, including potentially creating a charter for the working group. We have previously reported that fragmentation has caused inconsistent oversight and inefficient use of resources in the federal food safety oversight system. The agencies’ 2019 agreement to share oversight of cell-cultured meat creates a new relationship between FDA and USDA, since the agencies will oversee different stages of the production of the same food and hand off oversight at a certain point in that production. These factors contribute to an already complicated system in which the two agencies must coordinate on food safety oversight. In this context, some industry representatives and other stakeholders have expressed concerns about potential fragmentation or overlap in oversight of cell-cultured meat, such as could occur during the harvest phase of cell-cultured meat production when FDA hands off its oversight to USDA. Additionally, representatives from one cell-cultured meat firm stated that avoiding overlap in federal oversight whenever possible was important to them. For example, representatives from one firm pointed to inspection, record-keeping requirements, and regulations as potential areas at risk of overlap. They stated that potential overlap would add unnecessary, burdensome requirements and create an uneven playing field with the conventional meat industry. By more fully incorporating all seven leading practices for interagency collaboration early in the development of the three working groups, FDA and USDA could proactively minimize potential fragmentation and overlap in their oversight of cell-cultured meat, ensure consistency and efficient use of resources, and provide clarity to key stakeholders. FDA and USDA Have Not Documented Which Agency Will Oversee Cell-Cultured Seafood While FDA and USDA officials told us they have decided who will oversee cell-cultured seafood, they have not formally announced or documented this decision, and some stakeholders have reported confusion or ambiguity about which agency will oversee cell-cultured seafood other than catfish. Specifically, FDA and USDA’s interagency agreement regarding cell-cultured meat states that it covers all cell-cultured meat derived from USDA-amenable species required to bear a USDA mark of inspection, which in the agreement includes livestock, poultry, and catfish. However, the agreement does not mention cell-cultured meat made from the cells of other fish, such as tuna and shellfish. FDA and USDA officials told us that FDA will have sole oversight responsibility for cell-cultured seafood other than catfish. According to FDA officials, they have verbally communicated this decision in various meetings with stakeholders. However, FDA and USDA officials told us that formally documenting FDA’s sole oversight of most cell- cultured seafood in their interagency agreement was unnecessary because FDA currently oversees most conventional seafood. According to cell-cultured meat firms, some firms are working on developing cell- cultured versions of seafood, such as bluefin tuna. However, stakeholders from two cell-cultured meat firms, including representatives of a cell- cultured seafood firm we spoke with in April 2019, stated that they did not know who in the federal government would oversee cell-cultured seafood. Representatives from one cell-cultured seafood firm said that not being able to rule out oversight by USDA prevented them from making key decisions regarding what direction to pursue in developing their commercial production method. While FDA and USDA officials told us they had agreed that FDA would oversee cell-cultured seafood other than catfish, as of December 2019, the agencies had not formally announced or documented this agreement. Developing and updating written guidance and agreements is a leading practice for collaboration, as we have previously reported. In addition, standards for internal control in the federal government state that agency management should externally communicate the necessary quality information to achieve its objectives and should select appropriate methods of communication, such as a written document or a face-to-face meeting. Management should also periodically evaluate the entity’s methods of communication so that the organization has the appropriate tools to communicate quality information throughout and outside of the entity on a timely basis. While FDA and USDA officials have informally communicated to some stakeholders that FDA will have sole oversight of most cell-cultured seafood, FDA has not communicated this information formally or in a method readily available to all relevant stakeholders, such as in their interagency agreement or other publicly available written document. FDA and USDA officials told us that they wanted to communicate this information through outreach to individual firms, but FDA or USDA officials said they did not think that revising their interagency agreement was necessary. By taking steps to document which agency will oversee cell-cultured seafood other than catfish, FDA and USDA will better ensure the public, including key stakeholders such as cell-cultured meat firms, have clarity about the agencies’ oversight responsibilities in this area. Conclusions Cell-cultured meat is a new food product that raises many questions. FDA and USDA’s shared oversight of cell-cultured meat poses various challenges for these agencies, as well as stakeholders such as industry. Compounding this challenge is that specific information about key aspects of cell-cultured meat, such as the technology and production methods to be used as well as the composition of the products, is not yet known. FDA and USDA have taken steps to collaborate on their shared regulatory oversight of cell-cultured meat, including establishing an interagency agreement and three working groups. However, the interagency agreement only partially incorporates the seven leading collaboration practices that can enhance and sustain agencies’ collaborative efforts, and the working groups either partially incorporate or do not incorporate these leading practices, which has raised concerns about potential fragmentation or overlap in oversight. By more fully incorporating all seven leading practices for collaboration into their interagency agreement, FDA and USDA could build on their existing efforts and be better positioned to sustain and enhance their collaborative efforts. Moreover, by more fully incorporating all seven leading practices for interagency collaboration early in the development of the working groups, FDA and USDA could proactively minimize potential fragmentation and overlap in their oversight of cell-cultured meat and ensure they are utilizing resources efficiently or effectively. Furthermore, the interagency agreement states that it covers USDA- amenable species required to bear a USDA mark of inspection, which in the agreement includes livestock, poultry, and catfish but does not include cell-cultured seafood other than catfish. FDA and USDA officials told us they have decided FDA will oversee most cell-cultured seafood, but the agencies have not formally documented this decision. By taking steps to document in their interagency agreement, or other publicly available document, which agency will oversee cell-cultured seafood other than catfish, FDA and USDA could better ensure that members of the public and other key stakeholders such as cell-cultured meat firms have clarity about the agencies’ oversight responsibilities in this area. Recommendations for Executive Action We are making a total of six recommendations, three to FDA and three to USDA: The Commissioner of the Food and Drug Administration, in coordination with the Secretary of Agriculture, should more fully incorporate the seven leading practices for effective collaboration in the agencies’ interagency agreement for the joint oversight of cell-cultured meat. (Recommendation 1) The Secretary of Agriculture, in coordination with the Commissioner of the Food and Drug Administration, should more fully incorporate the seven leading practices for effective collaboration in the agencies’ interagency agreement for the joint oversight of cell-cultured meat. (Recommendation 2) As the three cell-cultured meat working groups move forward, the Commissioner of the Food and Drug Administration, in coordination with the Secretary of Agriculture, should more fully incorporate the seven leading practices for effective collaboration, such as identifying specific outcomes and a way to monitor and evaluate progress toward outcomes. (Recommendation 3) As the three cell-cultured meat working groups move forward, the Secretary of Agriculture, in coordination with the Commissioner of the Food and Drug Administration, should more fully incorporate the seven leading practices for effective collaboration, such as identifying specific outcomes and a way to monitor and evaluate progress toward outcomes. (Recommendation 4) The Commissioner of the Food and Drug Administration, in coordination with the Secretary of Agriculture, should clearly document in their interagency agreement, or other publicly available document, which agency will oversee cell-cultured seafood other than catfish. (Recommendation 5) The Secretary of Agriculture, in coordination with the Commissioner of the Food and Drug Administration, should clearly document in their interagency agreement, or other publicly available document, which agency will oversee cell-cultured seafood other than catfish. (Recommendation 6) Agency Comments and Our Evaluation We provided a draft of this report to the Department of Health and Human Services’ (HHS) Food and Drug Administration (FDA) and the U.S. Department of Agriculture (USDA) for review and comment. In FDA’s comments, reproduced in appendix III, the agency stated that it values GAO’s recognition of the importance of collaborative mechanisms that facilitate coordination and affirmed its commitment to coordinate closely with USDA to ensure the regulatory framework for cell-cultured meat is clear and transparent to stakeholders. In USDA’s comments, reproduced in appendix IV, the department stated that the report put too much focus on best practices for interagency collaboration and not enough emphasis on industry’s role in providing the agencies with the information they need to move their processes forward to effectively regulate cell-cultured meat. USDA stated that it is difficult to review a developing technology and its future regulatory oversight when so little detailed information about the technology is known. We agree that the technology to produce cell-cultured meat is still in development and that information about the commercial production methods and composition of the final product are not yet known, as we state in our report. We also acknowledge in our report that having limited information can affect the agencies’ ability to make regulatory and other decisions. We recognize that cell-cultured meat is a new food product that raises many new questions and that specific information about key aspects of cell-cultured meat is not yet known. In light of this challenging context, it is all the more important that FDA and USDA more fully incorporate leading practices for collaboration into their joint efforts in order to ensure they are in the best possible position to oversee this new food product. FDA concurred with two recommendations and partially concurred with one. USDA also concurred with two recommendations and partially concurred with one. Specifically, both agencies agreed with our recommendations regarding (1) more fully incorporating the seven leading practices for effective collaboration in the three cell-cultured meat working groups as they move forward and (2) clearly documenting which agency will oversee cell-cultured seafood other than catfish. FDA and USDA partially concurred with our recommendation, directed to each agency, to more fully incorporate the seven leading practices for effective collaboration into the agencies’ interagency agreement for the joint oversight of cell-cultured meat. FDA stated that it concurred with the intent of incorporating the seven leading practices into the interagency agreement, and both agencies said that they are open to incorporating the practices into their development of the structure for joint oversight of cell-cultured meat. However, the agencies stated that they did not agree to revise the agreement at this time. FDA and USDA stated that the agreement is a general framework and that incorporating the leading practices would constitute an inappropriate level of detail. Instead, the agencies stated that they believe it would be most valuable to incorporate the leading practices into a more detailed joint framework or standard operating procedure they plan to issue. We appreciate the agencies’ willingness to incorporate the leading practices for effective collaboration into their efforts. The March 2019 interagency agreement states that the agencies have the ability to modify it as needed and will review the agreement every 3 years to determine whether they should modify or terminate it. Therefore, the agencies are due to revisit the agreement in March 2022, if not sooner. Regarding the agencies’ concern that incorporating the leading practices in the interagency agreement would add an inappropriate level of detail, we note that, as we state in our report, the existing agreement already partially incorporates each of the seven leading practices. We continue to believe that FDA and USDA should more fully incorporate the seven leading practices for effective collaboration into their interagency agreement for the joint oversight of cell-cultured meat. Developing a more detailed joint framework or standard operating procedure in accordance with the existing interagency agreement that incorporates those leading practices would meet the intent of our recommendation to improve the effectiveness of the agencies’ collaboration. FDA and USDA also provided technical comments, which we incorporated as appropriate. As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 30 days from its issue date. At that time, we will send copies of this report to the appropriate congressional committees, the Secretary of Health and Human Services, the Secretary of Agriculture, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions regarding this report, please contact me at (202) 512-3841 or morriss@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix V. Appendix I: Objectives, Scope, and Methodology Our report (1) describes what is known about methods for commercially producing cell-cultured meat and (2) examines the extent to which the Food and Drug Administration (FDA) and U.S. Department of Agriculture (USDA) are collaborating to provide regulatory oversight of cell-cultured meat. For both objectives, we conducted a literature review of journal and media articles from 2016 through 2019 to inform our understanding of cell- cultured meat, as well as regulatory activity related to cell-cultured meat in the United States and in other countries. Specifically, we conducted a review of scholarly and trade news from 2016 through July 2019 for specific terms related to cell-cultured meat and regulatory approaches. We conducted searches in more than 30 different academic and trade databases—such as SCOPUS, Foodline, and ProQuest’s Environmental Science Collection—and identified studies relevant to our research objectives. In addition to these formal literature searches, we also asked agency officials and stakeholders to refer us to research articles and publications on cell-cultured meat. We also reviewed documentation from FDA and USDA, including the 2019 interagency agreement, existing memoranda of understanding between the two agencies, Federal Register notices about relevant public meetings, and press releases. We also reviewed documentation such as letters to regulators, presentation slides, and information on organizations’ websites from the cell-cultured meat industry, conventional meat industry, and consumer safety groups, among others. We also interviewed officials from FDA and USDA and representatives of stakeholders from the cell-cultured meat industry and industry associations, conventional meat firms and industry associations, academia, food and consumer safety groups, and state and tribal public health associations, among others. We identified stakeholders to interview through consultation with agency officials and nonfederal stakeholders and through our review of literature. We conducted 17 interviews with representatives or researchers from: six cell-cultured meat firms or industry associations, four conventional meat firms or industry associations, two food and consumer safety groups, one state and tribal public health association, and one food law policy firm. Because this is a nongeneralizable sample, the results of these interviews do not represent the views of all stakeholders involved in or with an interest in the cell-cultured or conventional meat industries or federal regulation of cell-cultured meat. However, they illustrate the range of perspectives on these topics. We also attended public meetings and conferences and conducted site visits to several locations. Specifically, we attended FDA and USDA’s public meeting in October 2018 and four conferences in 2019 that included content pertaining to food safety or cell-cultured meat. We conducted site visits to two conventional meat-processing facilities in Georgia, three cell-cultured meat firms in California, an academic cell- culturing laboratory in California, and a medical cell-culturing facility in Maryland. We identified facilities and laboratories to visit through our literature review, online research, and the assistance of agency officials and stakeholders, such as representatives from the cell-cultured meat and conventional meat industry. To describe what is known about the process for producing cell-cultured meat and potential commercial production methods, we also reviewed two sets of public comments submitted to FDA and USDA in association with the two 2018 public meetings pertaining to cell-cultured meat. These meetings were “Foods Produced Using Animal Cell Culture Technology” in July 2018 and “Use of Cell Culture Technology to Develop Products Derived from Livestock and Poultry” in October 2018. Public comments were submitted by members of the public; representatives from cell- cultured meat firms and industry associations, conventional meat companies and industry associations, food and consumer safety groups, and animal welfare groups; and environmental organizations, among others. We reviewed and analyzed all comments submitted to (1) FDA related to the July 2018 meeting and (2) FDA and USDA related to the October 2018 meeting. We also attended the October 2018 meeting and listened to agency officials’ presentations and oral remarks made by stakeholders and members of the public. We shared our description of the process for making cell-cultured meat, and associated questions, with representatives from three cell-cultured meat firms and academic researchers at two universities for their technical review and incorporated revisions as appropriate. To examine the extent to which FDA and USDA are coordinating to provide regulatory oversight of cell-cultured meat, we identified actions they took to coordinate from July 2018 through April 2020. To identify these actions, we interviewed agency officials, emailed agency officials written questions, reviewed agency documentation and public announcements, and attended public events such as the October 2018 public meeting. We compared the agencies’ interagency agreement and working groups with seven leading practices to enhance and sustain interagency collaboration. Specifically, two independent GAO reviewers assessed the degree to which agencies’ actions incorporated these leading practices. A description of these leading practices and the associated issues to consider is in appendix II. We also assessed the agencies’ actions against standards for internal control in the federal government, including standards related to communicating quality information. In this report, and in our past work, we define collaboration as any joint activity that is intended to produce more public value than could be produced when organizations act alone. We use the terms “coordination” and “collaboration” interchangeably in this report. For the purposes of our report, we define cell-cultured meat as food derived from animal cells that were grown in a controlled environment outside of the animal. We define cell-cultured seafood as a subcategory of cell-cultured meat. When referencing conventional meat, we are referring to food produced from the traditional method of slaughtering an animal, such as a cow, hog, chicken, or fish. When referencing seafood, we are referring to shellfish, sea fish, and freshwater fish served as food. We conducted this performance audit from October 2018 to April 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Key Issues to Consider for Implementing Interagency Collaborative Mechanisms Appendix II: Key Issues to Consider for Implementing Interagency Collaborative Mechanisms Issues to consider Have short-term and long-term outcomes been clearly defined? Is there a way to track and monitor progress toward the short-term and long-term outcomes? Do participating agencies have collaboration-related competencies or performance standards against which individual performance can be evaluated? Do participating agencies have the means to recognize and reward accomplishments related to collaboration? What are the missions and organizational cultures of the participating agencies? What are the commonalities between the participating agencies’ missions and cultures and what are some potential challenges? Have participating agencies developed ways for operating across agency boundaries? Have participating agencies agreed on common terminology and definitions? Has a lead agency or individual been identified? If leadership will be shared between one or more agencies, have roles and responsibilities been clearly identified and agreed upon? How will leadership be sustained over the long term? Have participating agencies clarified the roles and responsibilities of the participants? Have participating agencies articulated and agreed to a process for making and enforcing decisions? Have all relevant participants been included? Do the participants have: Full knowledge of the relevant resources in their agency? The ability to commit these resources? The ability to regularly attend activities of the collaborative mechanism? The appropriate knowledge, skills, and abilities to contribute? Developing and updating written guidance and agreements How will the collaborative mechanism be funded? If interagency funding is needed, is it permitted? If interagency funding is needed and permitted, is there a means to track funds in a standardized manner? How will the collaborative mechanism be staffed? Are there incentives available to encourage staff or agencies to participate? If relevant, do agencies have compatible technological systems? Have participating agencies developed online tools or other resources that facilitate joint interactions? If appropriate, have the participating agencies documented their agreement regarding how they will be collaborating? A written document can incorporate agreements reached in any or all of the following areas: Leadership Accountability Roles and responsibilities Resources Have participating agencies developed ways to continually update or monitor written agreements? Appendix III: Comments from the U.S. Department of Health and Human Services Appendix IV: Comments from the U.S. Department of Agriculture Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments Steve D. Morris, (202) 512-3841 or morriss@gao.gov In addition to the contact named above, Nico Sloss (Assistant Director), Angela Miles (Analyst-in-Charge), Sahar Angadjivand, Tim Bober, Kevin Bray, Colleen Candrl, Pin En Annie Chou, Tara Congdon, Heather Dowey, Kim Gianopoulos, Gina Hoover, Hayden Huang, Robert Lepzler, Serena Lo, David Lysy, Marc Meyer, Michael Polak, Danny Royer, Sara Sullivan, and Sarah Veale made key contributions to this report.

Why GAO Did This Study Multiple firms have produced cell-cultured meat as part of their research and development. These products appear likely to become available to consumers in coming years. FDA and USDA are the primary agencies responsible for overseeing the safety of the nation's food supply. However, some stakeholders have expressed concern about the agencies' oversight of cell-cultured meat amidst a fragmented federal food safety oversight system. GAO was asked to review federal oversight of cell-cultured meat. This report (1) describes what is known about methods for commercially producing cell-cultured meat, and (2) examines the extent to which FDA and USDA are collaborating to provide regulatory oversight of cell-cultured meat. GAO conducted a literature review; reviewed documentation from FDA, USDA, and stakeholder groups; analyzed public comments submitted to the agencies; compared agency efforts with leading practices for interagency collaboration; and conducted site visits to selected cell-cultured meat firms. What GAO Found General information about the process of making cell-cultured meat—food products grown from the cells of livestock, poultry, and seafood—is available. However, no company is commercially producing cell-cultured meat. Specific information about the technology being used, eventual commercial production methods, and composition of the final products is not yet known. The general process contains five phases: biopsy, cell banking, growth, harvest, and food processing (see figure). The technology and methods to be used for commercial production are still in development, and producers, regulators, and consumers do not have clarity about many specifics about the process and final product. For example, it is unclear whether production methods and products will use or contain genetically-engineered cells or medications such as antibiotics. The Food and Drug Administration (FDA) and U.S. Department of Agriculture (USDA) have begun collaborating on regulatory oversight of cell-cultured meat. For example, in 2019, the agencies signed an interagency agreement and created three working groups to carry out the terms of the agreement. However, the agreement and working groups could more fully incorporate practices to enhance and sustain collaboration, such as defining outcomes. For example, the agreement identifies the development of labeling principles as an outcome, but does not describe how the agencies will track and monitor progress toward this outcome, and the working groups identify a lead agency but not members' roles. Also, agency officials said they decided FDA would oversee cell-cultured seafood other than catfish, but they have not formally announced or documented this decision. Developing and updating written guidance and agreements is also a leading practice for interagency collaboration. By fully incorporating leading practices into their efforts to collaborate, the agencies could minimize potential overlap and fragmentation, use resources in a more efficient manner, and better ensure the public and other key stakeholders have clarity about the agencies' oversight responsibilities. What GAO Recommends GAO recommends that FDA and USDA more fully incorporate leading practices for effective collaboration in the agencies' interagency agreement. FDA and USDA partially concurred and indicated a willingness to incorporate these practices in a more detailed agreement, which would also meet the intent of the recommendations. The agencies concurred with the four other recommendations.

gao_GAO-19-502T

gao_GAO-19-502T_0

The Army Can Take Steps to Improve the Way Army Futures Command Operates Army Futures Command Established to Lead Modernization Efforts In January 2019, we reported on the initial steps the Army has taken to consolidate all its modernization efforts under one authority. Establishing Army Futures Command is reported to be the most significant institutional change to the Army since it reorganized in 1973 after the Vietnam War. According to Army documentation, the intent of the new command is to provide unity of command, accountability, and modernization at the speed and scale required to prevail in future conflicts. The organization is led by a four-star general like its organizational peers: Army Materiel Command, Training and Doctrine Command, and Forces Command. The Army declared the commencement of operations for the command in July 2018, and has begun to define its organizational structures. Army Futures Command is expected to be fully operational by July 2019, meaning it will have sufficient staff with operational facilities, secure funding, and the ability to execute its assigned mission, roles, and responsibilities. Army Futures Command is headquartered in Austin, Texas. According to Army officials and documentation, the Army chose Austin because of its proximity to science, technology, engineering, and mathematics talent, as well as private sector innovators that officials believe will assist the command in achieving its modernization goals. According to senior Army leadership, the new command headquarters will have around 300 staff in place by July 2019, a workforce that may grow to as many as 500 employees—100 military and 400 civilians. Our analysis of Army’s plans for initial staffing at the Army Futures Command headquarters, based on data from July 1, 2018, found that about one-third of headquarters staff would be involved directly in modernization efforts, such as engineers and operations specialists, and the remaining two-thirds would consist of support staff, including legal counsel and contracting professionals. According to Army Futures Command officials and documentation, the new organization will be organized around three major components: Futures and Concepts Center is responsible for identifying and prioritizing capability and development needs and opportunities. This organization subsumed the Army Capabilities Integration Center on December 7, 2018. The center was formerly part of Army Training and Doctrine Command and is located at Fort Eustis, Virginia. Combat Capabilities Development Command is responsible for conceptualizing and developing solutions for identified needs and opportunities. This organization subsumed the Research, Development and Engineering Command on February 3, 2019 and is located at Aberdeen Proving Ground, Maryland. Combat Systems Directorate is responsible for refining, engineering, and producing new capabilities. This directorate will communicate with the program executive offices and program management offices reporting to the Assistant Secretary of the Army for Acquisition, Logistics and Technology. Combat Systems Directorate is in the process of being established and is located in Austin, Texas. Among other things, the reorganization is intended to establish Army Futures Command to oversee development of Army’s six modernization priorities. The Army’s then-Acting Secretary and the Chief of Staff in an October 3, 2017 memorandum identified these priorities to guide Army modernization: next generation combat vehicle, air and missile defense, and soldier lethality. Army Established Cross- Functional Teams to Improve How it Develops Capabilities As we reported in January 2019, to pursue the six priority areas, the Army established eight cross-functional teams. These teams were initially created as a pilot effort to increase the efficiency of requirements and technology development for modernization before the announcement of the new command. They were subsequently moved into Army Futures Command in 2018. These cross-functional teams are located throughout the country in areas of relevance to their mission. The eight cross- functional teams and the priority areas they address are outlined in table 1. These cross-functional teams are intended to: take steps toward achieving the six modernization priorities; leverage expertise from industry and academia; identify ways to use experimentation, prototyping, and demonstrations; and identify opportunities to improve the efficiency of requirements development and the overall defense systems acquisition process. The cross-functional team pilots were structured to help achieve these goals. Each cross-functional team consists of core staff and subject matter experts from across the Army. To facilitate the rapid approval of requirements, each cross-functional team is led by a general officer or a senior civilian official who could communicate directly with the highest levels of the Army. The goal of staffing these teams is to ensure that each team had individuals who specialized in acquisition, requirements, science and technology, test and evaluation, resourcing, contracting, cost analysis, sustainment, and military operations. The goal of bringing different experts together is to facilitate collaboration and immediate opportunities for stakeholders to provide input as opposed to the more traditional requirements development process, in which input has typically been provided separately. Officials told us that, while all of these subject matter experts may have provided input on the requirements development process in the past, placing them on a single team offers the promise of streamlining those efforts and could eliminate the need for multiple reviews. Figure 1 below compares the requirements development process under cross-functional teams to how the Army has traditionally developed requirements. Further Implementation of Leading Practices Could Reduce Risk for Army Futures Command In January 2019, we recommended that Army Futures Command incorporate leading practices for effective cross-functional teams. We determined that the documentation that established the cross-functional team pilots fully addressed four of our eight leading practices for effective teams, and at least partially addressed another four. The leading practices and their implementation by the cross-function teams are described in table 2 below. In addition to the practices listed above, the cross-functional team pilots generally applied leading practices for requirements development. One leading practice the teams generally applied was promoting communication between requirements developers, warfighters, and industry representatives. This enables the cross-functional teams to better match developer resources with end-user needs. While applying this practice, the cross-functional team pilots had initial progress in writing requirements documents more efficiently. According to cross-functional team officials, they were able to shorten the requirements development process for several capabilities. However, we found that Army Futures Command does not have a formal plan to identify and share lessons learned from cross-functional team pilots to incorporate or expand application of these leading practices. Doing so would allow Army Futures Command the opportunity to accelerate the progress these teams made and spread the benefits across all of the teams and a wider range of specific military capabilities they are pursuing. We recommended that the Army (1) incorporate cross- functional teams’ experiences in applying leading practices and (2) execute a process for identifying and incorporating lessons learned. The Department of Defense concurred with these recommendations, and stated that Army Futures Command expects to apply leading practices and capture lessons learned by the end of 2019. Our January 2019 report also identified leading practices for mergers and organizational transformations. These leading practices are listed in table 3 below. We found that the Army Futures Command had implemented some of these practices, particularly leadership’s dedication to the new command and the clear statement of its mission. However, we have previously reported that, according to federal internal controls standards, it is important to implement all of these practices in order to establish the organizational structure necessary to enable an entity to plan, execute, control, and assess the organization in achieving its objectives. Establishment of this structure is particularly important for the Army where leadership and its priorities can change frequently. Therefore, we recommended in January 2019 that Army Futures Command fully apply these leading practices. The Department of Defense concurred with the recommendation, and stated that it would start pilot processes in fiscal years 2019 and 2020. Army Futures Command Should Assess Availability of Key Acquisition Personnel Needed for Requirements Development In addition to further implementing leading practices, Army Futures Command can reduce risk to meeting its goals by fully assessing the workforce necessary to develop requirements—the testable and measurable characteristics necessary for the design of a proposed system. Historically, the Army has been unable to ensure that requirements for new capabilities are feasible due, in part, to a declining workforce for requirements development. In June 2017, we reported that the Army had prioritized combat readiness over resourcing its requirements development process to meet future readiness needs. We recommended that the Army assess the resources, particularly personnel, necessary for requirements development. The Army concurred with the recommendation, and has stated it would implement this recommendation once Army Futures Command is fully operational. As Army Futures Command centralizes and takes responsibility for requirements development, this recommendation is even more pertinent. Therefore, we recently elevated the status of the recommendation to a priority recommendation for the Secretary of the Army, as we believe it warrants greater attention from the Department of the Army. Army Futures Command Has Not Developed Formal Policies and Procedures for Coordination with Other Army Acquisition Entities As Army Futures Command approaches full operating status, it is important to define not only how the command functions, but how it works with other organizations. In our January 2019 report, we found that Army Futures Command had not yet established policies and procedures detailing how it will execute its responsibilities in coordination with other Army organizations that do not directly report to it. One such organization is the Office of the Assistant Secretary of the Army for Acquisition, Logistics, and Technology—the civilian authority responsible for the overall supervision of Army acquisition matters—and the acquisition offices it oversees. To mitigate concerns about coordination, the Army issued a directive in August 2018, signed by the Secretary of the Army, designating the military deputy to the Assistant Secretary as an advisor to Army Futures Command, and Army Futures Command officials have stated that the Assistant Secretary will retain full acquisition authorities as required by law. The command expects to continue to refine its coordination with the Office of the Assistant Secretary of the Army for Acquisition, Logistics, and Technology. The Army Is Funding Modernization Priorities, but Further Steps Can be Taken to Manage Risk Army Modernization Has Prioritized Near-Term Capability Gaps while Identifying and Beginning to Fund Long-Term Needs Since announcing the modernization efforts in 2017, the Army has directed more funding toward closing near-term capability gaps, focused on fiscal years 2019 through 2023. For example, as part of the planning for the fiscal year 2019 budget process, the Army identified 67 high- priority programs, such as the M-1 Abrams tank and the AH-64 Apache helicopter, which require further investment. To support these priorities, the Army identified a need for $16 billion in increased funding in fiscal years 2019 through 2023. The 2018 Army Modernization Strategy report identified the need for additional resources for near-term efforts, including plans to spend billions of dollars for acquisition of maneuverable short range air defense capabilities in fiscal years 2020 through 2024. In addition to the near-term capabilities the Army is pursuing, it has identified a number of long-term needs—those focused after fiscal year 2024—and begun to align research and development efforts with these needs. The Army identified long-term capabilities for all of the modernization priorities, as well as dates that science and technology efforts should transition to programs of record. As part of this overall effort, the Army has evaluated its science and technology portfolio to realign funding toward its six modernization priorities. In an October 2017 Army review, the eight cross-functional teams examined science and technology investments to identify which efforts contributed to the priorities and which did not. The review was performed for the Office of the Deputy Under Secretary of the Army. Based on that work, as of our January 2019 report, the Army had taken steps to realign over $1 billion from previous priorities and toward the new priorities for fiscal years 2019 through 2023. Army officials stated that they expect to undertake similar reviews annually. Tracking Near-Term Modernization Efforts and Costs Could Address Management Challenges The Army is executing near-term modernization programs, but could better manage how it evaluates them and estimate their costs. In September 2018, we reported that the Army used its six priority capabilities to identify key mission areas—such as long-range artillery, air and missile defense, brigade combat teams, and cyber and electronic warfare—that require near-term modernization investments. Based on its assessments, the Army prioritized and proposed several near-term solutions to address its critical capability gaps. These solutions included adding personnel—and different types of personnel—to combat forces, updating existing weapon systems, and investments in research and development. However, the Army had not established processes for evaluating whether its modernization efforts allow it to deter or defeat potential adversaries during a major conflict. We also found that the Army had not fully estimated the costs or sources of funding for its near-term modernization efforts. In particular, we found that the Army did not report in its modernization strategy the extent to which it relied on Overseas Contingency Operations appropriations. We recommended that the Army (1) develop a plan to finalize the processes for evaluating how its near-term investments contribute to the Army’s ability to decisively defeat a major adversary, and (2) finalize its cost analysis of near-term investments and report those costs to Congress in its fiscal year 2020 budget request. Army officials told us in April 2019 that the Army has taken steps to implement these recommendations. Addressing Past Challenges with Technology Development Could Help Address Long- Term Modernization Risks The most recent efforts to modernize follow several past efforts. Unfortunately, the Army has a history of failed, costly weapon system procurements to replace older weapons systems. These failures are due, in part, to requirements that could not be met and the immaturity of key technologies. Many of these programs failed to provide any capability to the warfighter despite the considerable time and funding expended. Some examples of these cancelled programs are listed in table 4 below. While the Army has dedicated significant funding towards its long-term modernization priorities, other changes may also be needed. Among them, we recommended in our January 2019 report, that Army Futures Command take steps to follow our leading practices to mature technology to a sufficiently high level prior to system development, which can reduce risk. There are indications that, in some cases, the Army plans to mature technology to a sufficiently high level prior to system development. For example, officials from the Future Vertical Lift cross-functional team told us they will complete technology demonstrations on two competitive prototypes before choosing to develop a design for the Future Attack Reconnaissance Aircraft. However, we found that the Army may continue its past practice of proceeding into system development with less mature technologies. In particular, we identified some plans to mature technologies in a relevant environment prior to authorizing the start of a new acquisition program, rather than the higher level of demonstrating them in an operational environment as recommended by our leading practices. This increases risk that new capabilities will require further maturation in system development, which could raise costs and extend timelines for delivery of equipment to the warfighter. We recommended in our January 2019 report that the Army should demonstrate technologies in an operational environment before starting a formal acquisition program. The Department of Defense concurred with the recommendation and stated that the Army Futures Command will execute a new development process that will include operational technology demonstrations. Pilot processes for this are expected to begin in 2019. In summary, we recognize that the Army is early in its modernization efforts but could make changes now that would be helpful. Army Futures Command should implement not only the leading practices we describe as well as the lessons learned by its own cross-functional teams. The Army should also increase the transparency of its efforts by clarifying how it evaluates its progress towards modernization goals and clearly stating the full costs of pursuing those goals. Finally, the Army can reduce the risk to the long-term modernization of its capabilities by ensuring that the technologies it uses in future weapon systems are fully mature. Chairman Norcross, Ranking Member Hartzler, and Members of the Subcommittee, this concludes my prepared statement. I would be pleased to answer any questions that you may have at this time. GAO Contact and Staff Acknowledgment If you or your staff have any questions about this testimony, please contact Jon Ludwigson, Acting Director, Contracting and National Security Acquisitions at (202) 512-4841, or ludwigsonj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this testimony are J. Kristopher Keener (Assistant Director), Joe E. Hunter (Analyst-in-Charge), Emily Bond, Matthew T. Crosby, Cale Jones, Kevin O’Neill, John Pendleton, and Roxanna Sun. 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.

Why GAO Did This Study The Army is investing in near- and long-term modernization efforts to maintain its technological edge over potential adversaries. It is doing this by upgrading and updating current weapon systems, developing new capabilities, and reshaping its doctrine, force structure, training, and leader development. This testimony is based on prior GAO work conducted 2016 through 2019 and addresses the Army's progress in: (1) establishing Army Futures Command, and (2) developing its near-term and long-term modernization strategies. It also highlights several actions recommended in prior reports related to Army modernization. To conduct this work, GAO assessed the Army's near- and long-term modernization efforts, application of leading practices to those efforts, budget documents, and the effectiveness of the process for developing requirements for major weapon systems. This statement includes updates to this information, as of April 2019. What GAO Found In January 2019, GAO reported on initial steps the Army has taken to consolidate its modernization efforts under one authority—Army Futures Command. Army officials call it their most significant institutional change since 1973, when the Army was reorganized after the Vietnam War. As a precursor to this new command, the Army established eight cross-functional teams as a pilot program to increase the efficiency of requirements and technology development in six key modernization areas. These areas are described in the table below. Since announcing the modernization efforts in 2017, the Army has directed more funding toward closing near-term capability gaps. For example, as part of the planning for the fiscal year 2019 budget process, the Army identified 67 high-priority programs that require a $16 billion investment between now and fiscal year 2023. In addition to the near-term capabilities the Army is pursuing, it has identified a number of long-term needs—those focused after fiscal year 2024—and taken steps to realign research and development efforts and funding with those needs. Over the past 2 years, GAO highlighted several steps Army should take to improve its modernization efforts, including: Apply leading practices to Army Futures Command's cross-functional teams, and capture their lessons learned. Assess the resources, particularly personnel, necessary to support its requirements development process. Increase the transparency of its efforts by clarifying how it evaluates whether its modernization efforts are achieving the Army's goals and clearly stating the full costs of pursuing those goals. Reduce risk by ensuring technologies are fully mature—such as demonstrating technologies in an operational environment before starting a formal acquisition program. By implementing these recommendations, Army Futures Command could better ensure its ability to deliver enhanced capabilities to the warfighter and decrease the risk of cost and schedule growth. What GAO Recommends Over the past 2 years, GAO has made recommendations related to this body of work. Department of Defense and Army concurred with all the recommendations and are working to implement them.

gao_GAO-20-297

gao_GAO-20-297_0

Background National Preparedness Goal Following Hurricane Katrina in 2005, the Post-Katrina Act required FEMA to develop a national preparedness system and assess preparedness capabilities to determine the nation’s disaster preparedness. In September 2011, DHS issued the National Preparedness Goal: a secure and resilient nation with the capabilities required across the whole community to prevent, protect against, mitigate, respond to, and recover from the threats and hazards that pose the greatest risk. The National Preparedness Goal also defined the “whole community” as individuals and communities, the private and nonprofit sectors, faith-based organizations, and all governments (local, regional/metropolitan, state, tribal, territorial, insular area, and federal). The National Preparedness Goal identifies and defines 32 core capabilities across five broad mission areas. These capabilities form the foundation for measuring overall national preparedness and assisting the nation in allocating resources to fill identified preparedness gaps. Three of the 32 core capabilities affect all mission areas and are considered to be “crosscutting” (see fig. 1). The five broad mission areas are: Prevention. Preventing an imminent threat, or actual act of terrorism. Protection. Protecting citizens, residents, visitors, and assets in a manner that allows interests, aspirations, and way of life to thrive. Mitigation. Mitigating the loss of life and property by lessening the impact of future disasters. Response. Responding quickly to save lives, protect property and the environment, and meet basic human needs in the immediate aftermath of an incident. Recovery. Recovering through a focus on the timely restoration, strengthening, and revitalization of infrastructure, housing, and a sustainable economy, as well as the health, social, cultural, historic, and environment fabric of communities affected by an incident. Since 2012, DHS has produced a National Preparedness Report annually, which assesses progress toward the National Preparedness Goal of achieving a secure and resilient nation. A key element of the National Preparedness Report is that it evaluates and measures the extent to which jurisdictions have strengthened their 32 core capabilities. From 2012 to 2017, all 50 states, District of Columbia, and 5 territories were required to assess the preparedness levels of their 32 capabilities by providing a rating of 1 to 5, with 5 being the highest preparedness rating. Emergency management capabilities with a rating of 1 or 2 are considered to have the largest capability gaps. FEMA used this assessment process to inform the National Preparedness Report by illustrating which threats and hazards occurred in the past and which capabilities have the largest gaps. FEMA’s National Preparedness Directorate, which includes the National Preparedness Assessment Division, is responsible for assisting communities in becoming more resilient by developing the capabilities needed to prevent, protect against, respond to, recover from, and mitigate against all threats and hazards. The Directorate provides guidance, programs, and processes to assist communities in completing the requirements associated with the National Preparedness System. Jurisdictions’ Emergency Management Capability Assessments To help jurisdictions more comprehensively assess their gaps, FEMA required they complete the Threat and Hazard Identification and Risk Assessment (THIRA) and Stakeholders Preparedness Review (SPR). The THIRA is conducted by jurisdictions every 3 years to, in part, identify threats and hazards that are both reasonably likely to affect the community and would most challenge the community’s ability to deliver one or more of its capabilities; and estimate and describe the potential impacts of those threats and hazards. The types of threats and hazards are defined as (1) natural hazards or acts of nature; (2) technological hazards that are accidents or failures of systems and structures; and (3) human-caused incidents resulting from intentional actions. Jurisdictions are to conduct the SPR annually to, among other things, identify capability gaps by assessing the capabilities against the types of threats and hazards identified in the THIRA. In 2012-2013, FEMA issued its initial guidance to jurisdictions to help them understand how to identify the threats and hazards through the THIRA, and assess their core capabilities. In 2018, FEMA issued new guidance for the THIRA and SPR requiring jurisdictions to change the methodology, moving away from proficiency-based ratings to a process that relies more on quantitative data to measure gaps across the core capabilities. In 2018, FEMA required jurisdictions to begin using the new methodology to assess the core capabilities within the response and recovery mission areas. Beginning in 2019, FEMA required jurisdictions to begin using the new methodology to assess the core capabilities across all mission areas. FEMA’s 10 regions provide technical assistance and training to help jurisdictions become more proficient in completing these capability assessments. FEMA also sponsors exercises with states, territories, tribes, and localities to help them assess their emergency management capabilities. In addition to the jurisdictions’ THIRAs and SPRs, in 2019, FEMA initiated an effort to assess the federal government’s emergency management capacity. According to FEMA, the effort is intended to provide a national THIRA and SPR that assesses the federal government’s capabilities against the nation’s threats and hazards. The Disaster Recovery Reform Act of 2018 (DRRA) requires FEMA, among other things, to provide congressional committees updates every 6 months on its progress in completing a national preparedness assessment until the assessment is complete. In July 2019, FEMA issued its 2019 National Threat and Hazard Identification and Risk Assessment (National THIRA): Overview and Methodology, describing its approach to completing a national-level risk assessment (i.e., a National THIRA). According to FEMA, the National THIRA was completed in 2020, and will be included in the 2020 National Preparedness Report. FEMA Preparedness Grants DHS, through FEMA, provides jurisdictions preparedness grants, which are used, in part, to strengthen the 32 core capabilities across the five mission areas. FEMA has traditionally provided three primary preparedness grants that jurisdictions can use to strengthen their emergency management core capabilities. Two of the three grants, the State Homeland Security Grant Program and the Urban Area Security Initiative, were established after the terrorist attacks of September 11, 2001. As established by federal law, these grants are intended to help states and localities prevent, prepare for, protect against, and respond to acts of terrorism. State Homeland Security Grant Program. Provides funding to assist state, local, and tribal governments in preventing, preparing for, protecting against, and responding to acts of terrorism. Helps support states’ implementation of homeland security strategies to address the identified planning, organization, equipment, training, and exercise needs at the state and local levels. In fiscal year 2019, the total funding available to all 50 states, District of Columbia and 5 territories was $415 million. Urban Area Security Initiative. Provides federal assistance to address the unique needs of high-threat, high-density urban areas, and assists the areas in building a capacity to prevent, prepare for, protect against, and respond to acts of terrorism. In fiscal year 2019, the total funding available to the 31 urban areas was $590 million. Emergency Management Performance Grant. Provides federal assistance to states to assist state, local, and tribal governments in preparing for all hazards. The program plays a valuable role in strengthening and sustaining the 32 core capabilities across the five mission areas. In fiscal year 2019, the total funding available to states, local governments, and tribes was $315 million. The National Preparedness System and Grants Have Helped Build Some Emergency Management Capabilities, but Gaps Remain The National Preparedness System and associated preparedness grants have helped jurisdictions strengthen and sustain their emergency management capabilities. More specifically, according to National Preparedness Reports since calendar year 2012, states and territories generally have rated their capabilities within the prevention and response mission areas, as well as their crosscutting capabilities—which involve all five mission areas, as having the highest preparedness levels. By contrast, states and territories generally have rated their capabilities in the recovery and protection mission areas as having lower preparedness levels, and these ratings showed little to no improvement from 2013 to 2017. Additionally, since 2013, jurisdictions have directed nearly 87 percent of their FEMA preparedness grants toward sustaining or strengthening capabilities in the crosscutting, prevention, and response mission areas, and around 13 percent on enhancing or sustaining capabilities in the protection, mitigation, and recovery mission areas. FEMA has encouraged jurisdictions to invest future preparedness grants to strengthen their capabilities that have lower preparedness ratings and to address emerging threats, such as cybersecurity. However, FEMA officials told us their efforts to help jurisdictions enhance their capabilities, including the distribution of existing preparedness grants, will likely not be sufficient to address the capability gaps that have been identified by jurisdictions. Preparedness Data Show Capabilities Are Strongest in the Crosscutting, Prevention, and Response Areas; Lowest in the Protection and Recovery Areas States and territories’ 2017 preparedness data showed that eight core capabilities in the response and crosscutting mission areas had the highest level of preparedness (a rating of 4 or 5 on a 5-point scale). For example, as shown in figure 2 below, over 50 percent of the assessment ratings by the states and territories identified crosscutting capabilities, such as public information and warning and operational coordination, in the highest category of preparedness. Similarly, 57 percent of the assessment ratings by states and territories identified on-scene security, protection, and law enforcement capabilities within the response mission area in the highest preparedness categories. In our discussions with local officials who were impacted by the 2017 and 2018 hurricanes, they told us that the operational coordination and public information and warning capabilities were effective during their response efforts. For example, Craven County, NC, officials told us that in response to the flooding from Hurricane Florence, their emergency operations center was instrumental in communicating with first responders. In doing so, they were able to keep county wells running while working with utility companies to prioritize areas that needed electrical power, such as hospitals and grocery stores. Additionally, Onslow County, NC officials said their emergency operations center was instrumental in communicating and coordinating the rescue operations of around 700 residents through the use of the county’s swift water rescue teams, with assistance from the U.S. Marine Corps, North Carolina’s National Guard, and the local fire and police departments. In addition, Brazoria County, TX, officials told us that in response to Hurricane Harvey they used videos and social media to get warning messages out to the residents and businesses about evacuation assistance as well as information on hurricane preparedness. Supply chain integrity and security Hurricane Florence caused significant flooding in and around New Hanover County, North Carolina. County officials told us the state’s National Guard high wheel clearance trucks had to be used to transport food, water, fuel, and generators throughout the flooded areas to isolated communities because the county did not have the capability to deliver these commodities. According to the North Carolina National Guard, the high-water vehicles were also used to evacuate citizens to shelters and transport essential civilian personnel such as nurses, doctors, and first responders. Preparedness data from 2017 show that almost 40 percent of jurisdictions’ ratings identified five capabilities in the recovery and protection mission areas in the lowest category of preparedness (a rating of a 1-2 on a 5-point scale). For example, within the recovery mission area, 51 percent of the ratings identified disaster housing in the lowest category of preparedness. Similarly, within the protection mission area, 46 percent of the assessment ratings identified cybersecurity in the lowest category. Additionally, these capabilities have been consistently rated in the lowest preparedness categories from 2013 through 2017 and have shown little-to-no change. For example, under the recovery mission area, 56 percent of the assessment ratings by states and territories identified disaster housing in the lowest category in 2013, with minimal changes through 2017. Some of the capabilities that had the lowest preparedness ratings in 2013 were: Economic recovery. The ability to return economic and business activities (including agricultural) to a state of health and develop new economic opportunities that result in a sustainable and economically viable community. Natural and cultural resources. The ability to preserve, conserve, rehabilitate, and restore historic property consistent with post-disaster community priorities and best practices and in compliance with environmental and historic preservation laws and executive orders. Disaster housing. The ability to address pre- and post-disaster housing issues and coordinate the delivery of federal resources and activities to assist local, state, tribal, territorial, and insular area governments as they rehabilitate and reconstruct destroyed and damaged housing. Supply chain integrity and security. The ability to secure and make resilient key nodes, methods of transport between nodes, and materials in transit between a supplier and consumer. Table 1 shows the percentages of the lowest-rated capabilities from 2013 through 2017. While the National Preparedness System may help jurisdictions assess their preparedness using emergency preparedness capability assessments, jurisdictional officials we spoke with told us that real-life disasters sometimes show jurisdictions to be less prepared than their capability assessments previously indicated. As a result, some states have lowered their preparedness ratings in subsequent capability assessments following a disaster. For example, after the 2017 and 2018 hurricanes, some states told us they lowered their preparedness rating in their 2019 assessments for disaster housing because they realized the capability gap was larger than they previously believed. Officials from the North Carolina Division of Emergency Management said that in 2018, they lowered their preparedness rating for housing because their housing capacity was not able to meet the needs of disaster victims who needed immediate housing assistance. From Fiscal Years 2013 through 2018, Jurisdictions Directed Almost 90 Percent of FEMA Preparedness Grant Funding to the Crosscutting, Prevention, and Response Mission Areas From fiscal years 2013 through 2018, jurisdictions received approximately $8.3 billion in preparedness grants funds primarily from the State Homeland Security Program, Urban Area Security Initiative, and the Emergency Management Performance Grant. Of this amount, jurisdictions directed about $7.3 billion—or about 87 percent of the funds—to capabilities in the crosscutting, prevention, and response mission areas, which constitute the highest-rated mission areas. For example, in California, $1.9 million in Urban Area Security Initiative grants were used to strengthen crosscutting and prevention capabilities by providing situational awareness to first responders and emergency managers working on active threats to infrastructure. Additionally, in Florida, up to $2.8 million of the State Homeland Security grant was used to create a system intended to strengthen crosscutting and prevention capabilities by enabling the state’s law enforcement agencies to more easily share information. Of the $8.3 billion in preparedness grant funding from fiscal years 2013 through 2018, about $1.1 billion—or about 13 percent—was directed to capabilities in the mitigation, protection, and recovery mission areas, which constitute the lowest-rated mission areas. During this time, jurisdictions directed the least amount of preparedness grant funds on the recovery mission area—$78 million, or about 1 percent (see fig. 3). Jurisdictions also directed about 5 percent of the $1.1 billion to capabilities within the mitigation mission area, though preparedness ratings in the mitigation mission area generally showed improvements each year. In 2017, 43 percent of the assessment ratings by states and territories rated three of the four mitigation-related capabilities in the highest category. Improvements in the mitigation mission area could be, in part, attributable to FEMA providing jurisdictions with grant funds other than preparedness grants, such as post-disaster grants, which include Hazard Mitigation Grant Program funds. Further, state and local decisions on how to prioritize preparedness grant awards resulted in about 1 percent—$78 million—being directed to capabilities within the recovery mission area between 2013 and 2018. As shown in figure 4, jurisdictions directed approximately 79 percent of the $78 million (about $62 million) in the recovery mission area to the infrastructure systems capability, which is intended to, in part, allow jurisdictions to re-establish critical infrastructure in disaster-impacted areas to support life sustainment activities, ongoing emergency response operations, and to help facilitate recovery efforts. Additionally, about 3 percent of the $78 million—about $2.4 million—was directed to disaster housing capabilities, such as implementing housing solutions that effectively support the temporary housing needs of an impacted jurisdiction. State officials from New York and North Carolina, as well as officials from five localities, said they often prioritize and use preparedness grants to maintain existing capabilities within the crosscutting, prevention, and response capabilities, rather than enhancing capabilities where gaps are known to exist, such as those in the recovery and protection areas. In addition, state officials from Texas, as well as officials from two localities told us that they need to use portions of their limited grant funds— especially from Emergency Management Performance Grant funds—to hire and retain local emergency management personnel, which leaves fewer funds for them to devote to enhancing lower-rated emergency management capabilities. For example, some county governments may not have the resources necessary to fund a single emergency manager, which requires them to use Emergency Management Performance Grant funds to hire and retain necessary staff. Another reason why jurisdictions do not use more of the grants toward lower-rated mission areas is because some view certain capabilities in the recovery and mitigation mission areas to be the responsibility of the federal government. Both FEMA and state officials told us that sometimes jurisdictions do not use these grants to strengthen capabilities such as housing because they consider the federal government responsible for filling the gaps. For example, preparedness data from 2013 to 2017 showed the percent of jurisdictions identifying the federal government as responsible for providing housing solutions to disaster survivors increased from 46 to 53 percent. According to state officials from North Carolina, it would not be a prudent use of grant funds for the state to purchase and store temporary housing units that may not be needed inside the borders of the state for several years. Following a major disaster declaration, FEMA coordinates with jurisdictions to provide disaster housing assistance to people displaced from their homes. For example, following Hurricane Florence, FEMA provided financial rental assistance and grants under its Individuals and Households program to help make repairs to damaged homes. In addition, FEMA, in coordination with the state of North Carolina, delivered travel trailers and manufactured housing units (i.e., mobile homes) to displaced disaster victims through FEMA’s Direct Temporary Housing Assistance program. FEMA Has Encouraged Jurisdictions to Invest More Grant Funds on the Largest Capability Gaps, and Has Proposed New Grants to Address Such Gaps FEMA has encouraged jurisdictions to make investments in core capabilities that have the largest preparedness gaps (i.e., the lowest preparedness scores). From 2013 to 2018, DHS identified investment priorities in its annual announcements of preparedness grant funding opportunities. The priorities focused on select capabilities where jurisdictions had reported lower preparedness scores, such as cybersecurity, disaster housing, economic recovery, natural and cultural resources, and supply chain integrity and security. Specifically, FEMA officials told us cybersecurity remains a high priority for all jurisdictions for 2020 and has identified areas from lessons learned where cybersecurity could be strengthened. Jurisdictions are considering investments in cybersecurity such as adding more information technology equipment and hiring personnel with cybersecurity expertise. However, according to state officials from New York and Texas, jurisdictions often lack the resources necessary to hire and retain personnel skilled enough to prepare for, respond to, and recover from cyberattacks. Preparedness grants, in general, are designed to allow jurisdictions discretion to spend the funds as they see fit on projects that meet eligibility requirements. While FEMA encourages jurisdictions to invest grant funds to address their capability gaps, it does not require or direct jurisdictions to spend grant funding in a certain area. In light of these challenges, FEMA has taken a number of other steps to try to address these capability gaps. FEMA proposed creating a new National Priorities Security Grant in the President’s 2019 and 2020 budget proposals, which could be used to address new and emerging threats and gaps, such as those in cybersecurity. FEMA proposed that the program’s priorities be assessed frequently and shift as needed to address emerging threats and capability gaps. In fiscal years 2019 and 2020, the President’s budget proposed $522 and $430 million respectively. The proposed grant program was not approved by Congress. In 2019, FEMA established the Regional Catastrophic Preparedness Grant Program to help jurisdictions address known capability gaps in disaster housing as well as logistics and supply chain management. In fiscal year 2019, FEMA awarded $10 million in these grants to eight local governments. In 2019, FEMA began implementing the Building Resilient Infrastructure and Communities (BRIC) program to provide jurisdictions with funding to make their infrastructure more resilient in future disasters. According to FEMA, grant recipients could use future funding to strengthen capability gaps in the recovery and mitigation mission areas. FEMA plans to issue a Notice of Funding Opportunity in the summer of 2020, followed by an application period. Based on historical disaster expenditures, FEMA anticipates BRIC will be funded between $300 million and $500 million per year on average. It is too early to assess the extent to which this program will help address capability gaps. While FEMA is taking steps to encourage jurisdictions to enhance their lower-rated capabilities, FEMA officials told us their efforts combined with existing preparedness grants, will likely not be sufficient to fully address jurisdictions’ capability gaps. Specifically, FEMA officials told us that the current suite of preparedness grants lacks the flexibility needed to address some of the long-standing capability gaps, in part, because the grants are required to be spent on capabilities that have a nexus to terrorism. In addition, as described earlier, one state official, and two local officials, suggested that the level of funding for the Emergency Management Performance Grant will likely not allow states and localities to hire and retain local emergency management personnel while also making the investments needed to address the capability gaps identified through the National Preparedness System. For example, one emergency management official from a county explained that without using the Emergency Management Performance Grant to offset his own salary, the county would not have an emergency management department with the capability to complete many of the FEMA requirements associated with receiving disaster assistance. In addition to the steps FEMA has already taken to attempt to address the capability gaps, FEMA has developed a new methodology for assessing national preparedness capabilities that uses more quantitative methods. According to FEMA, such methods could enable jurisdictions to more tangibly define what resources are needed to fill identified gaps. We describe this methodology in more detail below. FEMA Is Taking Steps to Strengthen Assessments of Federal and Jurisdiction Capabilities, but Opportunities Exist to Further Enhance National Preparedness FEMA Is Implementing a New Methodology to Strengthen How Jurisdictions Assess Preparedness FEMA has taken steps to enhance its methodology for assessing jurisdictions’ emergency management capabilities by requiring jurisdictions to collect more quantitative preparedness data to support their capability ratings. We reported in March 2011 that FEMA needed to improve its oversight of preparedness grants by establishing a framework with measurable performance objectives for assessing urban area, state, territory, and tribal capabilities to identify gaps and prioritize investments. Specifically, we recommended that FEMA complete a national preparedness assessment of capability gaps at each level of government based on tiered, capability-specific performance objectives to enable prioritization of grant funding. We also reported in March 2013 that FEMA has made some progress in assessing its preparedness capabilities, but continued to face challenges developing a national preparedness system that could assist FEMA in prioritizing preparedness grant funding. FEMA’s issuance of the 2020 National Preparedness Report could provide an assessment of capability gaps at each level of government—including an assessment of the federal government’s capabilities for the first time—and help FEMA address the intent of the 2011 recommendation. However, as discussed before, prioritizing jurisdictions’ preparedness grant funding alone may not effectively address the nation’s emergency management capability gaps. An assessment that also considers the federal government’s emergency management capabilities could help determine what capabilities federal agencies could provide to assist in the wake of disasters when jurisdictions’ capabilities become overwhelmed or are not otherwise available. Once the assessment is completed, FEMA and its federal budgeting stakeholders (i.e., Congress and the Office of Management and Budget) could use such an assessment to identify the potential costs of establishing and maintaining capabilities, not only at the jurisdictional level, but also at the federal level. FEMA has continued to take steps to implement the 2011 recommendation, but has not yet fully addressed it as of January 2020. For example, FEMA published new guidance in May 2018 to update the methodology for how jurisdictions are to evaluate their preparedness levels when completing THIRAs and SPRs. The intent was to allow communities to collect more specific, quantitative data to compare their capability targets to current capabilities, thereby more accurately defining their capability gaps. Beginning in 2018, jurisdictions used the new methodology to assess their capabilities in the crosscutting, response, and recovery mission areas. Beginning in 2019, jurisdictions were required to use the new methodology to assess the capabilities across all five mission areas: prevention, protection, mitigation, response, and recovery. According to FEMA, this new methodology improves on the prior one because the new methodology will allow jurisdictions to more accurately determine what amount of resources are needed to address specific threats and hazards. Specifically, as a result of using more quantitative data, such as the specific number of disaster victims able to be sheltered following a disaster, jurisdictions may be able to better define their capability gaps when compiling their SPRs. For example, if jurisdictions are able to understand that their current capability is less than their needed capability target, they will be able to define their capability gaps in quantitative terms. According to FEMA officials, the new methodology, if implemented successfully, will allow jurisdictions to know what additional resources and capabilities—beyond their own current capabilities—may be needed during future disasters. Table 2 shows an example of how FEMA’s updated methodology provides a more quantitative assessment to more accurately define their capabilities. FEMA Is Using Its New Methodology to Assess the Federal Government’s Emergency Management Capacity and Better Define the Nation’s Capability Gaps In 2019, FEMA began working on its first National Threat and Hazard Identification and Risk Assessment (National THIRA) to identify what federal government capabilities will be needed to address the greatest threats to the nation. According to FEMA, the results of this effort are expected to be included in FEMA’s annual National Preparedness Report in 2020, which is expected to be published late in calendar year 2020. FEMA’s effort is intended to provide a quantitative assessment of federal capabilities, which when combined with state, territory, urban area, and tribal THIRAs and SPRs, could provide a more meaningful assessment of the nation’s overall preparedness. Figure 5 below shows how national and jurisdiction risk assessments are intended to work together to provide a collective picture of overall capability gaps. As subsequent iterations of the National THIRA and National SPR are produced, FEMA intends to consolidate them with the THIRA and SPR assessments submitted by jurisdictions to provide a comprehensive overview of national preparedness. FEMA officials told us that they have begun to assess and measure the federal government’s capabilities in the crosscutting, response, and recovery mission areas. In conducting the 2019 National THIRA—that FEMA officials told us will be included in the 2020 National Preparedness Report—FEMA coordinated with over a dozen federal departments and agencies, as well as selected national laboratories and the White House to solicit feedback on the most challenging threats and hazards facing the nation. The 2019 National THIRA consists of nine catastrophic incident scenarios and 22 capability targets across the crosscutting, response, and recovery mission areas. For example, FEMA used catastrophic scenarios, such as a pandemic (see sidebar) or New Madrid Earthquake, to assess the nation’s emergency management capacity. Pandemic Scenario In early October, the Centers for Disease Control and Prevention (CDC) reports a new strain of influenza virus in the National Capital Region. Less than 2 weeks after the first confirmed case is identified at a local hospital, the illness causes hundreds of fatalities and thousands of people seeking medical attention. As the virus spreads, approximately 30 percent of the population across the United States and other countries becomes severely ill. Conventional flu vaccines are ineffective against the current strain, and the CDC estimates that a new vaccine could be months away from mass production. Because of the pandemic, social distancing is in widespread effect. Utilities, police, fire, government, and other essential services are disrupted due to social distancing and employee absenteeism. Businesses close, resulting in a large-scale loss of services across the region (e.g. banking, food stores, gas stations). There is a shortage of medical supplies, equipment, beds, and healthcare workers as hospitals are quickly overwhelmed, with up to millions of individuals seeking outpatient medical care and millions more requiring hospitalization. Civil disorder contributes to the high rate of absenteeism and the overcrowding of hospitals and medical centers. how quickly power service can be restored to customers; how quickly life-sustaining commodities can be delivered to people; how quickly emergency sheltering, food, and water can be provided to how quickly affected healthcare facilities can restore function. In the aftermath of the sequential 2017 disasters, FEMA’s 2017 Hurricane Season FEMA After-Action Report recognized the need to more effectively scale response efforts for concurrent, complex incidents. As a result, in addition to the nine catastrophic scenarios, the National THIRA considered the challenges associated with managing concurrent incidents. To examine the potential impacts of managing concurrent incidents, FEMA developed a set of “plausible concurrent operations.” FEMA acknowledged that the agency and its federal partners “will almost certainly be engaged in ongoing disaster operations at the time of any catastrophic-level incident” and gathered data from historical incidents from recent years, including the sequential disasters that took place in 2017; three large hurricanes and wildfires in California, among others. FEMA found that combining the impacts of a National THIRA scenario with the set of plausible concurrent operations more accurately reflects the challenges the nation would need to address should one of the threat scenarios identified in the National THIRA occur. However, given that FEMA has yet to finalize inclusion of the National THIRA into the 2020 National Preparedness Report, it is too early to determine the extent to which it helps FEMA more accurately define the nation’s emergency management capability gaps and results in the nation being better prepared for future catastrophic disasters. FEMA Has Yet to Determine What Steps Need to be Taken to Address Capability Gaps at the Federal and Jurisdictional Levels As discussed above, the National Preparedness System has identified gaps in jurisdictions’ emergency management capabilities since 2012. While jurisdictions have used preparedness grants to strengthen select capabilities, preparedness data shows that they have not used the grants to address capability gaps across all the mission areas. Furthermore, while FEMA has encouraged jurisdictions to use grant funding to address capabilities that have the largest capability gaps, such as those in the recovery and cybersecurity areas, they do not require that jurisdictions do this. However, if FEMA were to require jurisdictions to use their grant funds to address lower-rated capabilities, it could affect jurisdictions’ ability to sustain other core capabilities—or to fund emergency management personnel in select jurisdictions, some of which only have one full-time employee. As FEMA implements its new methodology and begins to more fully assess both federal and jurisdictional capabilities, the agency is expected to have better and more quantitative information on capability gaps in order to better prioritize grant funds and resources. According to FEMA, the agency and its partners will better understand the extent of the nation’s emergency management capability gaps when they issue the National Preparedness Report by December 2020. While these actions may allow FEMA to address our 2011 recommendation and better measure the nation’s overall preparedness, the agency has yet to determine what additional actions may be needed to close the remaining gaps once the 2020 National Preparedness Report is issued. Further, while FEMA has taken some steps to close the gaps jurisdictions have identified since 2012, such as proposing the National Priorities Security Grant, this program has not been approved by Congress, and thus, will not help to address the gaps. According to FEMA officials, preparedness grants alone are unlikely to address the gaps in an effective manner. In addition, the National Preparedness Goal states that analyzing current performance against intended capabilities allows the emergency management community the opportunity to determine necessary resource levels, inform resource allocation, and help guide federal investments in preparedness. Such information could help inform budget decisions across the preparedness enterprise and help prioritize limited resources. For example, determining what steps need to be taken, following the issuance of the 2020 National Preparedness Report, could help FEMA inform key decision makers, such as Congress and the Office of Management and Budget, about the necessary level of resources—including the allocation of resources—that are needed to address the nation’s capability gaps. Such an effort could be a significant step toward enhancing the capability gaps that have been identified since 2012 and help determine the nation’s overall preparedness levels, as called for in the Post-Katrina Act. FEMA Has Identified Some Areas for Improvement Following Disasters, but Could Strengthen After-Action Reviews and Follow-Up FEMA Has Policies and Processes for Using After- Action Reviews to Identify Areas for Improvement following a Major Disaster The Post-Katrina Act requires FEMA to analyze real-world events to identify and disseminate lessons learned and best practices, and to generate and disseminate, as appropriate, after-action reports to participants after real-world events. After major disasters occur, FEMA’s policy is to conduct an after-action review that identifies strengths, areas for improvement, and potential best practices identified during response and recovery efforts. Lessons learned from past disasters are to provide collective knowledge and diverse experiences for improving disaster response and recovery. Further, FEMA’s 2018-2022 Strategic Plan calls for sharing lessons learned from disasters and exercises with the whole community to help prioritize investments and anticipate known challenges during future disasters. In July 2018, FEMA published its 2017 Hurricane Season After-Action Report, which discussed findings and recommendations based on a review of the agency’s preparation for, immediate response to, and initial recovery operations for Hurricanes Harvey, Irma, and Maria. According to FEMA, the agency is implementing recommendations to address the challenges outlined in the after-action report, which include the following focus areas: scaling and staffing for concurrent complex incidents; improving logistics capabilities during response; improving response to long-term infrastructure outages; and, improving mass care to initial disaster housing operations based on innovations developed during the 2017 hurricane season. According to FEMA, the agency has taken a number of actions in response to this after-action report. For example, it increased its incident management workforce strength by 19 percent since Hurricane Harvey and updated hurricane plans, annexes, and procedures for the continental United States and for states and territories outside the continental United States, among other things. FEMA’s Continuous Improvement Program is responsible for collecting observations and conducting after-action reviews after disasters. The program is intended to consolidate feedback and information from regional, headquarters, and field operations staff and provide information to FEMA leadership and program offices to improve the efficiency and effectiveness of the agency’s disaster operations. The regional role in the Continuous Improvement Program is to identify lessons learned and best practices from disaster events in their regions, conduct after-action reviews, and track corrective actions and improvement plans applicable to the region through Continuous Improvement Working Groups. FEMA officials told us that after-action report findings that cannot be resolved at the regional level are elevated to headquarters for resolution. According to FEMA officials, FEMA headquarters reviews completed after-action reports to identify any areas for improvement that may need to be addressed through changes in policies and procedures. FEMA Conducts After- Action Reviews for Select Disasters, but Has Not Developed Guidance to Assist Regional Officials in Prioritizing Which Disasters Should Result in an After-Action Review Although FEMA’s policy requires after-action reviews be conducted after every presidentially-declared major disaster, we found that the agency does not consistently conduct after-action reviews after all major disasters and has not instituted time frames for following up on incomplete after- action reviews. As of January 2020 FEMA had completed after-action reviews for 29 percent of disasters since January 2017, with 43 percent pending or in the process of being completed, and 27 percent having been deferred (i.e., not completed or status unknown), as shown in figure 6. Our review of relevant policy indicates that FEMA does not specify time frames for when after-action reviews are to be completed. This is consistent with what we heard from FEMA officials who explained they do not have any time frames for when a certain region is to complete after- action reviews. FEMA has recently updated its Continuous Improvement Program. For example, in 2019, FEMA updated the Continuous Improvement Directive to formalize an annual Summary of Findings that consolidates the field, regional, and headquarters’ observations from the year’s incidents in order to identify the strengths, best practices, and lessons learned that should be addressed the following year. However, FEMA officials noted that this had only been done once in 2019, and would be completed in future years. Officials from FEMA’s Continuous Improvement Program in one region cited challenges with capacity, staffing, and the number of on-going after- action reviews as reasons for not being able to complete all of their after- action reports. According to FEMA officials, in 2017 each region was assigned one to two continuous improvement advisors who are responsible for developing the region’s after-action reviews. However, FEMA officials in one region said that in 2019, they faced challenges in having the staff resources necessary to operate the Continuous Improvement Program due to competing priorities, such as responding to active disasters. In addition, FEMA officials stated that due to limited staff, the regions have to prioritize which after-action reviews they can complete based on the severity and impact of the disaster. For example, in 2017, FEMA focused resources on reviewing the agency’s response and recovery for Hurricanes Harvey, Irma, and Maria. According to FEMA regional and headquarters officials, competing priorities, such as responding to active disasters, often result in staff being unavailable to conduct after-action reviews. While we acknowledge staffing is limited and that FEMA may need to prioritize completing some after-action reviews over others, FEMA officials have not established a process or framework by which regional offices are to prioritize after-action reviews. Based on our analysis of the after-action reviews since 2017 and discussions with FEMA headquarters and regional staff, we found that FEMA does not have a formal process to prioritize after-action reviews and has not established general time frames for how long following a disaster an after-action review should be completed, or followed-up on. FEMA officials agreed that having a formal process to prioritize after-action reviews, including establishing time frames for following up on incomplete after-action reviews, could provide the agency additional opportunities to improve response and recovery operations for future disasters. According to FEMA Regional officials, timely after-action reviews are useful. For example, as a result of the 2017 Hurricanes Season After-Action Report, Region II was able to update response plans for Puerto Rico, which could prove to be beneficial for future disasters. According to The Standards for Program Management, agencies should collect, measure, and disseminate performance information, analyze program trends, and point to areas in need of adjustment. In addition, leading practices for program management indicate that project schedules should be developed to define project milestones and identify and sequence activities in order to determine start and end dates for each activity. Additionally, in other branches of FEMA, the agency provides time frames for completing after-action reports. For example, states and territories are expected to submit after-action reports within 90 days of exercises that are funded by the Homeland Security Grant Program. . Similarly, FEMA policy requires Urban Search and Rescue teams to submit after-action reports 30 days after returning from deployment. Developing a process by which regional offices are to prioritize after- action reviews could help FEMA ensure that regions have a common framework to work from when determining what disasters should be prioritized for review and could help FEMA prioritize staff resources more effectively across the Continuous Improvement Program. Furthermore, establishing time frames for following up on incomplete after-action reviews could provide FEMA with greater assurance that the reviews will be conducted in a timely fashion, so that other FEMA Regions and key stakeholders can benefit from the lessons learned. FEMA Headquarters Lacks a Formal Mechanism to Document and Track Best Practices, Lessons Learned, and Corrective Actions Identified through After- Action Reviews As described earlier, FEMA regional offices Continuous Improvement Working Groups are responsible for developing and tracking, to the extent possible, corrective actions and best practices identified through after- action reviews. These working groups are to elevate to FEMA headquarters any issues that cannot be resolved at the regional-level. However, FEMA does not have a formal mechanism at the headquarters level for documenting and tracking best practices, lessons learned, and corrective actions that have been elevated from the regional working groups. According to FEMA, it has taken steps to track best practices and lessons learned through a serious of Microsoft Excel files, but it is not a long term or ideal operating solution due to its lack of accessibility, ease of use, and ability to be queried. In February 2016, we recommended that FEMA implement a process to document, track, and analyze recommendations and implement lessons learned after disaster deployments. FEMA concurred with this recommendation and implemented the recommendation by using the Department of Defense’s Joint Lessons Learned Information System as its primary system to capture and manage lessons learned data. However, according to FEMA officials, as of July 2019, it no longer uses the system to capture lessons learned data. FEMA officials also said the Joint Lessons Learned Information System was not user-friendly. FEMA officials stated that they hold a quarterly meeting, as required by FEMA Directive 107-1, with FEMA’s Associate Administrators to review national priorities and issues that have been elevated to headquarters for resolution. According to FEMA officials, this group performs the function that a Continuous Improvement Working Group does at the regional level by monitoring issues that need adjudication by senior management officials. While the quarterly meeting may be helpful, it does not serve as a mechanism, such as a data system, for documenting and tracking best practices, lessons learned, and corrective actions identified after a major disaster. Additionally, continuous improvement coordinators from the regions we interviewed stated that once a finding is elevated to FEMA’s headquarters, in general the region does not have visibility into what steps, if any, FEMA headquarters is taking or plans to take to address the issue. Having a mechanism, such as a database, to record after-action report findings, such as corrective actions or best practices, could help FEMA facilitate awareness across the agency about the status of FEMA’s efforts to address them. According to the Post-Katrina Act, FEMA should conduct remedial action tracking and long-term trend analysis. Furthermore, the National Response Framework specifies that evaluation and continual process improvement are cornerstones of effective preparedness. The framework notes that effective practices with continuity planning ensures the capabilities contained in the framework can continue to be executed regardless of the threat or hazard. Without a mechanism to document and track best practices, lessons learned, and corrective actions identified through after-action reviews across the regions and headquarters, FEMA may not be able to provide assurance that it is effectively leveraging best practices and lessons learned or taking corrective actions to improve its response and recovery programs. FEMA Lacks Guidance on Sharing After-Action Report Findings with External Stakeholders Following a Disaster As described earlier, the Post-Katrina Act requires FEMA to generate and disseminate, as appropriate, after-action reports to participants in exercises and real-world events. In addition, FEMA’s stated policy on knowledge sharing after disasters is to collaborate with public and private sector partners to share insights on critical issues facing emergency management, promote best practices, and discuss ways in which FEMA itself can improve. However, based on a query of FEMA’s website for after-action reports on disasters, since January 1, 2017, FEMA has placed on-line one after-action report on the 2017 hurricane season. In addition, state officials from Florida, as well as officials from ten localities told us that there has been no communication from FEMA specifically in regards to its 2017 Hurricane Season After-Action Report to ask jurisdictions to provide feedback on the final product or its findings. In addition to FEMA not communicating with jurisdictions about its final product or its findings, state and local officials we spoke with said that FEMA does not consistently share after-action reports with affected jurisdictions. For example, officials from the state of Florida and four localities told us that FEMA does not consistently share its reports after each disaster, while officials from the state of California stated that FEMA has regularly shared after-action reports from disasters. One FEMA regional official noted that it would be helpful to know who, when, and to what extent lessons learned should be shared with external partners. Further, according to FEMA, knowledge sharing allows communities impacted by disasters to prioritize investments and anticipate known challenges during disasters. According to The Standards for Program Management, agencies should collect, measure, and disseminate performance information and analyze program trends, and point to areas in need of adjustment. FEMA has guidance for sharing after-action reports internally within the agency, but according to FEMA officials has not developed guidance for when after-action reports, or findings from after-action reports, should be shared with external stakeholders. According to some state and local officials we spoke with, having access to disaster after-action reports could be useful to FEMA’s external stakeholders. For example, because FEMA’s 2017 Hurricanes Season After-Action Report was accessible, New York City officials said they were able to be proactive in areas that needed to be strengthened in the event of delayed federal assistance, such as providing disaster housing services. Lessons learned can be produced through after-action reports and are relevant to key stakeholders, such as state and local governments, which are instrumental in disaster preparedness, response, and recovery, and would play a key role in any future disasters. However, without guidance to help officials determine when it is appropriate to share after-action reports, FEMA may miss opportunities to share lessons learned. Further, FEMA’s Strategic Plan states that building a culture of preparedness requires continued learning, improvement, innovative ideas, and engagement of the whole community. As such, all sectors of society, including governments, nonprofit organizations, and the private sector, will need to be involved in preparedness for future disasters. The plan further states that insights can be gained through observations from after- action reports and through feedback from stakeholders. A FEMA official from one of the region’s Continuous Improvement Program agreed that developing guidance to determine when it is appropriate to share after- action reports, could help stakeholders better prepare for future disasters. By developing guidance for sharing after-action reports or their relevant findings—when appropriate—with key external stakeholders, FEMA could help communities better prepare for future disasters through knowledge sharing. Conclusions FEMA has taken numerous steps to continue to strengthen national preparedness, such as distributing grant funds. However, FEMA has not fully defined the capability gaps and determined what steps are needed to enhance capabilities across all levels of government. Informing key stakeholders, such as the Office of Management and Budget and Congress, about what resources will be necessary to address the gaps— across all levels of government—will be critical in addressing the nation’s emergency management capability gaps. In addition, opportunities exist to enhance FEMA’s after-action review process. More specifically, until FEMA prioritizes when—and for what disasters—after-action reviews should be completed and establishes time frames for following up on incomplete after-action reports, the agency will not be able to guarantee that FEMA and its stakeholders can leverage lessons learned from recent disasters and apply corrective actions before future disasters occur. Further, without a mechanism to document and track best practices, lessons learned, and corrective actions throughout the agency, FEMA may not be able to effectively leverage best practices and lessons learned or implement corrective actions to improve its response and recovery operations. By addressing areas needing improvement (i.e., corrective actions) once after-action reviews are completed, FEMA could improve response and recovery operations in the wake of future disasters. In addition, FEMA could help communities better prepare for future disasters by developing guidance to share its after- action reports or findings from its after-action reports—when appropriate—with key stakeholders, allowing them to provide feedback on the findings or adjust their own operational plans to be better prepared to work with FEMA during future disasters. Recommendations for Executive Action We are making the following four recommendations to the FEMA Administrator: Following the completion of the 2021 National Preparedness Report, determine what steps are needed to address the nation’s emergency management capability gaps across all levels of government and inform key stakeholders, such as the Office of Management and Budget and Congress, about what level of resources will be necessary to address the known gaps. (Recommendation 1) Develop guidance to help determine which after-action reviews should be prioritized based on factors, such as the severity of disasters and availability of staff and resources to conduct the review, and implement time frames for following up on incomplete after-action reports. (Recommendation 2) Develop a mechanism to consistently track best practices, lessons learned, and corrective actions that have been elevated to headquarters for resolution. (Recommendation 3) Develop guidance on sharing after-action reports and their relevant findings with external stakeholders, when appropriate. (Recommendation 4) Agency Comments and Our Evaluation We provided a draft of this report to the Department of Homeland Security (DHS) for their review and comment. DHS provided written comments, which are reproduced in appendix I. In its comments, DHS concurred with the four recommendations and described actions under way or planned to address them by March 31, 2022. DHS provided technical comments, which we incorporated as appropriate. DHS concurred with our first recommendation to determine what steps are needed to address the nation's emergency management capability gaps across all levels of government and inform key stakeholders about what level of resources will be necessary to address the known gaps. According to DHS, this recommendation is consistent with the requirements outlined in the Disaster Recovery Reform Act of 2018 (DRRA) noting that FEMA complete a national preparedness assessment of capability gaps at each level based on tiered, capability-specific performance objectives to enable prioritization of grant funding; and identify the potential costs for establishing and maintaining those capabilities at each level and determine what capabilities federal agencies should provide. DHS also stated that while the 2020 National Preparedness Report will include a nation-wide assessment of community capability against national capability targets to help understand gaps and inform grant investments, it will not include data on federal capabilities. The collection of that information, through the National Stakeholder Preparedness Report, was scheduled to begin in 2020 but was delayed due to response operations for the COVID-19 pandemic. According to DHS, this information will be incorporated into the 2021 National Preparedness Report, helping to form a more complete picture of national capabilities. FEMA stated that the costs to address the nation’s resource gaps cannot be estimated without first accounting for existing federal capabilities. According to DHS, the anticipated date for the 2020 National Preparedness Report, pending response operations to the COVID-19 pandemic, is October 30, 2020, and the 2021 National Preparedness Report is planned to be released in October 2021. DHS stated that once the 2021 National Preparedness Report is released, FEMA will develop and socialize a plan to work with the federal interagency to identify resources needed to address the national gaps identified in the 2021 National Preparedness Report. If implemented effectively, these actions combined with the steps taken to inform key stakeholders could meet the intent of our recommendation. Due to the impacts of the COVID-19 pandemic and the need to finalize the 2021 National Preparedness Report prior to being able to account for the federal government’s existing capabilities, we are adjusting the wording of this recommendation to follow the issuance of the 2021 National Preparedness Report. DHS estimates the expected completion date to be March 2022. DHS concurred with our second recommendation to develop guidance to help determine which after-action reviews should be prioritized and implement timeframes for following up on incomplete after-action reports. According to DHS, FEMA will address the prioritization of disaster after- action reports as the Continuous Improvement Program’s first priority for 2020. Additionally, FEMA plans to identify and develop timeframes for following up on after-action reports as part of a broader program evaluation effort in 2020. These actions, if implemented effectively, could meet the intent of our recommendation. While FEMA originally anticipated completing this guidance during 2020, the COVID-19 response extended this timeline. DHS estimates the expected completion date to be March 31, 2021. DHS concurred with our third recommendation to develop a formal mechanism to consistently track best practices, lessons learned, and corrective actions. DHS stated that FEMA, in December 2019, implemented an issue elevation and resolution system for tracking best practices, lessons learned, and corrective actions that are elevated to FEMA headquarters level for resolution, as appropriate. However, according to FEMA in April 2020, the agency has taken steps to track best practices and lessons learned through a serious of Microsoft Excel files, but this is not considered to be a long term or ideal operating solution due to its lack of accessibility, ease of use, and ability to be queried. Further, in April 2020, FEMA stated that it is working to identify resources to build an actual application that will be used for this purpose. These actions, if implemented effectively, could meet the intent of our recommendation. DHS concurred with our fourth recommendation that FEMA develop guidance on sharing after-action reports and their relevant findings with external stakeholders, when appropriate. According to DHS, FEMA is drafting program guidance for the Continuous Improvement Program to address the sharing of after action reports and their relevant findings with external stakeholders. These actions, if implemented effectively, could meet the intent of the recommendation. Due to the ongoing COIVD-19 pandemic, FEMA estimates its completion date to be March 31, 2021. We are sending copies of this report to the Secretary of Homeland Security, the FEMA Administrator, and the appropriate congressional committees. If you or your staff have any questions about this report, please contact me at (404) 679-1875 or curriec@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix II. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. Appendix I: Comments from the Department of Homeland Security Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Aditi Archer (Assistant Director), Robert Denton Herring (Analyst-in-Charge), Erin Guinn-Villareal, James Lawson, Ben Ayres, Eric Hauswirth, Tracey King, Amanda Miller, Kevin Reeves, and Minette Richardson made significant contributions to this report.

Why GAO Did This Study FEMA uses the National Preparedness System to help assess the nation's emergency management capabilities in preparing for disasters and, in part, to help prioritize federal preparedness grants it provides to state and local jurisdictions. Since 2002, FEMA has provided over $52 billion in such grants intended to enhance preparedness capabilities. GAO was asked to examine national preparedness. This report examines the extent to which: (1) FEMA's National Preparedness System and associated preparedness grants have assisted jurisdictions in preparing for disasters; (2) FEMA has strengthened the National Preparedness System and what steps remain; and (3) FEMA is using after-action reports to identify lessons learned and strengthen future preparedness. GAO evaluated agency guidance, analyzed 2013 to 2017 capability data—the most current available; conducted site visits to five states; and interviewed FEMA, state, and local emergency management officials. What GAO Found The Federal Emergency Management Agency's (FEMA) National Preparedness System and associated grants have helped build some emergency management capabilities, but gaps remain. Capabilities fall in five mission areas: (1) prevention—preventing imminent acts of terrorism, (2) protection—protecting citizens and assets, (3) mitigation—mitigating the loss of life and property, (4) response—responding quickly to save lives, and (5) recovery—timely restoration of infrastructure and housing, among other things. From fiscal years 2013 through 2018, jurisdictions directed almost 90 percent of FEMA preparedness grants ($7.3 of $8.3 billion) to capabilities in the crosscutting (i.e., benefit all five mission areas), response, and prevention areas (figure below). Jurisdictions reported a higher level of preparedness in these areas compared to capabilities in the other mission areas—recovery, mitigation, and protection. Jurisdictions have consistently rated select capabilities in these three mission areas—such as disaster housing and cybersecurity—in the lowest category since 2013. FEMA does not limit jurisdictions' use of preparedness grants for select capabilities, but it has encouraged jurisdictions to address the known gaps. FEMA is taking steps to strengthen the national preparedness system, but has yet to determine what steps are needed to address the nation's capability gaps across all levels of government. Specifically, FEMA is implementing a new methodology to collect more quantitative data on capabilities at the state, territory, and local levels—as GAO recommended in 2011—and also plans to begin assessing the federal government's capabilities. Including the federal government in such an assessment would enable FEMA and jurisdictions to assess national preparedness capabilities collectively. While these are positive steps that could meet the intent of the 2011 recommendation, FEMA has yet to determine what steps are needed to address the capability gaps once they are identified, including jurisdictions' capability gaps that have been known since 2012. By determining these steps and informing key stakeholders, such as Congress, about what resources will be needed across all levels of government, FEMA will be better positioned to address the nation's capability gaps. FEMA after-action reports have identified areas for improvement and lessons learned following disasters, but has completed after-action reviews for only 29 percent of disasters from 2017 through 2019. FEMA lacks a formal mechanism to track corrective actions and does not have guidance on sharing after-action reports with key external stakeholders, as appropriate. What GAO Recommends GAO is making four recommendations that FEMA (1) determine what steps are needed to address emergency management capability gaps, and communicate it to key stakeholders (2) prioritize completion of after-action reviews, 3) track corrective actions, and (4) develop guidance on sharing findings externally. The Department of Homeland Security concurred and FEMA is taking actions in response.

gao_GAO-19-241

gao_GAO-19-241_0

Background According to OMB, federal agencies reported that they operated 432 data centers in 1998, 2,094 in July 2010, and 9,995 in August 2016. Operating such a large number of centers has been, and continues to be, a significant cost to the agencies. For example, in 2007, EPA estimated that the annual cost for electricity to operate federal servers and data centers across the government was about $450 million. Further, according to the Department of Energy (Energy), a typical government data center has 100 to 200 times the energy use intensity of a commercial building. However, in 2009, OMB reported server utilization rates as low as 5 percent across the federal government’s estimated 150,000 servers. All of these factors contributed to OMB recognizing the need to establish a coordinated, government-wide effort to improve the efficiency, performance, and environmental footprint of federal data center activities. OMB and the Federal CIO Established FDCCI Concerned about the size of the federal data center inventory and the potential to improve the efficiency, performance, and environmental footprint of federal data center activities, OMB’s Federal CIO established FDCCI in February 2010. This initiative’s four high-level goals were to reduce the overall energy and real estate footprint of government data centers; reduce the cost of data center hardware, software, and operations; increase the overall IT security posture of the government; and shift IT investments to more efficient computing platforms and technologies. In February 2010, OMB required all of the agencies participating in the FDCCI to submit a data center inventory and a consolidation plan. In October 2010, OMB also clarified the definition of a data center and noted that, for the purposes of FDCCI, a data center was to be defined as any room used for the purpose of processing or storing data that is larger than 500 square feet and meets stringent availability requirements. Under this definition, OMB reported that agencies had identified 2,094 data centers as of July 2010. However, in 2011, the Federal CIO expanded the definition to include a facility of any size and OMB published its revised definition in March 2012. Based on the revised definition, OMB estimated that there were a total of 3,133 federal data centers in December 2011. In addition, its goal was to consolidate approximately 40 percent, or 1,253 of these data centers, for a savings of approximately $3 billion by the end of 2015. Figure 1 shows data center server racks at SSA’s National Support Center in 2017. The number of federal data centers reported by agencies has continued to grow since 2011. In May 2018, we reported that agencies had collectively identified a total of 12,062 data centers in their inventories as of August 2017—an increase of about 9,000 data centers compared to OMB’s October 2011 estimate. According to the Federal CIO, the increase in the number of data centers was primarily due to the expanded definition of a data center (discussed later in this report) and improved inventory reporting by the agencies. See figure 2 for a depiction of the increase in the number of data centers from 1998 through August 2018. Further, OMB placed greater emphasis on data center optimization to improve the efficiency of federal data centers when it issued memorandum M-13-09 in March 2013. Specifically, OMB stated that, to more effectively measure the efficiency of an agency’s data center assets, agencies would also be measured by the extent to which their primary data centers are optimized for total cost of ownership by incorporating metrics for data center energy, facility, labor, and storage, among other things. Subsequently, in May 2014, OMB issued memorandum M-14-08, which established a set of data center optimization metrics to measure agency progress, along with target values for each metric. All agencies were expected to achieve the target values by the end of fiscal year 2015. IT Acquisition Reform Law Enhanced Data Center Consolidation and Optimization Efforts Recognizing the importance of reforming the government-wide management of IT, Congress enacted FITARA in December 2014. Among other things, the law required agencies to: Submit to OMB a comprehensive inventory of the data centers owned, operated, or maintained by or on behalf of the agency. Submit, by the end of fiscal year 2016, a multi-year strategy to achieve the consolidation and optimization of the agency’s data centers. The strategy was to include performance metrics that were consistent with the government-wide data center consolidation and optimization metrics. Report progress toward meeting government-wide data center consolidation and optimization metrics on a quarterly basis to OMB’s Administrator of the Office of Electronic Government. In addition, according to FITARA, the Office of Electronic Government at OMB was to: Establish metrics applicable to the consolidation and optimization of data centers (including server efficiency), ensure that agencies’ progress toward meeting government-wide data center consolidation and optimization metrics is made publicly available, review agencies’ inventories and strategies to determine whether they are comprehensive and complete, and monitor the implementation of each agency’s strategy. Develop and make publicly available not later than December 19, 2015, a goal broken down by year for the amount of planned cost savings and optimization improvements to be achieved through FDCCI and, for each year thereafter until October 1, 2020, compare reported cost savings and optimization improvements against those goals. OMB Established DCOI to Address FITARA Data Center Provisions OMB issued memorandum M-16-19 in August 2016 to establish DCOI and included guidance on how to implement the data center consolidation and optimization provisions of FITARA. Among other things, the guidance required agencies to consolidate inefficient infrastructure, optimize existing facilities, improve their security posture, and achieve cost savings. For example, each agency was required to maintain a complete inventory of all data center facilities owned, operated, or maintained by or on its behalf, and measure progress toward defined optimization performance metrics on a quarterly basis as part of its data center inventory submission. OMB’s memorandum also directed each agency to develop a DCOI strategic plan that defined its data center strategy for fiscal years 2016 through 2018. Among other things, this strategy was to include a timeline for agency consolidation and optimization activities, with an emphasis on cost savings and optimization performance benchmarks that the agency could achieve between fiscal years 2016 and 2018. For example, each agency was required to develop cost savings targets due to consolidation and optimization actions and report any realized cost savings. OMB required each agency to publicly post its DCOI strategic plan to its agency-owned digital strategy website by September 30, 2016, and to post subsequent strategic plan updates by April 14, 2017 and April 13, 2018. Further, the memorandum stated that OMB was to maintain a public dashboard (referred to as the IT Dashboard) to display government-wide and agency-specific progress in areas such as planned and achieved data center closures, consolidation-related cost savings, and data center optimization performance information. In this regard, OMB began including data center consolidation and optimization progress information on the IT Dashboard in August 2016. OMB’s memorandum also provided new guidance for the classification of a physical data center and expanded the definition of a data center. According to the revised definition, a room with at least one server that provides services (whether in a production, test, staging, development, or any other environment) should be considered a data center, while a room containing only print servers, routing equipment, switches, security devices (such as firewalls), or other telecommunication components, was not to be considered a data center. In light of this new definition, OMB directed each agency to perform a comprehensive review of its data centers and maintain a complete and updated data center inventory. Further, OMB directed each agency to categorize each of its data centers as either a tiered data center or a non- tiered data center. OMB’s memorandum defined a tiered data center as one that uses each of the following: a separate physical space for IT infrastructure; an uninterruptible power supply; a dedicated cooling system or zone; and a backup power generator for a prolonged power outage. According to the memorandum, all other data centers were to be considered non-tiered. Moreover, OMB guidance included a series of performance metrics in the areas of data center closures, cost savings, and optimization progress. Data center closures: According to the guidance, agencies were to close at least 25 percent of tiered data centers government-wide, excluding those approved as inter-agency shared services providers, by the end of fiscal year 2018. Further, agencies were to close at least 60 percent of non-tiered data centers government-wide by the end of fiscal year 2018. OMB’s guidance further notes that, in the long term, all agencies should continually strive to close all non-tiered data centers, noting that server rooms and closets pose security risks and management challenges and are an inefficient use of resources. Cost savings: According to the guidance, agencies were to reduce government-wide annual costs attributable to physical data centers by at least 25 percent, resulting in savings of at least $2.7 billion for fiscal years 2016 through 2018. Data center optimization: According to the guidance, agencies were to measure progress against a series of new data center performance metrics in the areas of server utilization, energy metering, power usage, facility utilization, and virtualization. Further, OMB’s guidance established target values for each metric that agencies were to achieve by the end of fiscal year 2018. OMB’s guidance further noted that agency progress against these performance metrics was to be measured by OMB on a quarterly basis, using agencies’ data center inventory submissions and OMB-defined closures, cost savings, and optimization targets. OMB Published Proposed Changes to DCOI in November 2018 In November 2018, OMB published proposed changes to DCOI for public comment. The changes focus federal consolidation and optimization efforts on agencies’ larger, tiered data centers and also de-emphasize the consolidation of non-tiered facilities and other smaller spaces. The draft guidance also revises the classification of data centers and data center optimization metrics. The draft guidance redefines a data center as a purpose-built, physically separate, dedicated space that meets certain criteria. Similarly, OMB does not plan to continue to report on spaces not designed to be data centers. According to the draft, OMB also plans to work with agencies to set agency-specific goals for data center closures and cost savings and to update these targets from those set in OMB’s August 2016 memorandum to match agencies’ current status and progress. Additionally, the proposed changes to DCOI make several changes to the metrics currently used by agencies to monitor the performance of their data centers. Specifically, of the five metrics currently in use (described in detail later in this report), OMB proposes updating three, removing two, and adding one new metric. The draft guidance states that public comments will be collected through the end of December 2018, but does not provide a date for when the proposed changes will be finalized and implemented. However, the draft does state that the new guidance will sunset on September 30, 2020, a date that coincides with the extension of FITARA’s data center provisions. Agencies Have Taken Limited Action to Address Prior GAO Recommendations Since the enactment of FITARA in December 2014, we have reviewed and verified annually for the quality and completeness of each agency’s (covered by the law) inventory and DCOI strategy. We have also published reports documenting the findings from each of our reviews. In addition, we have examined and reported on agencies’ efforts to optimize their data centers, as well as the challenges encountered and successes achieved. In a report that we issued in March 2016, we noted that agencies had reported significant data center closures—totaling more than 3,100 through fiscal year 2015—with the Departments of Agriculture, Defense (Defense), the Interior (Interior), and the Treasury (Treasury) accounting for 84 percent of the total. Although the agencies fell short of OMB’s fiscal year 2015 consolidation goal, their plans identified about 2,100 additional centers planned for closure through fiscal year 2019. Agencies also reported significant consolidation cost savings and avoidances—totaling about $2.8 billion through fiscal year 2015 and expected to increase to over $8.0 billion in future years. The Departments of Commerce, Defense, Homeland Security (DHS), Transportation (Transportation), and the Treasury accounted for 96 percent of the total planned savings. However, we pointed out that many agencies lacked complete cost savings goals for the next several years despite having closures planned. In addition, we reported that 22 agencies had made limited progress against OMB’s fiscal year 2015 data center optimization performance metrics, such as the utilization of data center facilities. Accordingly, we recommended that the agencies take actions to complete their cost savings targets and improve optimization progress. As of December 2018, 18 of the 32 recommendations from this report had yet to be fully addressed. In May 2017, we reported that the agencies were reporting significant data center closures—totaling more than 4,300 through August 2016— with Agriculture, Defense, Interior, and the Treasury accounting for 84 percent of the total. The agencies’ plans for 2016 had identified more than 1,200 additional centers for closure through fiscal year 2019. Agencies also reported significant consolidation and optimization cost savings and avoidances, which totaled about $2.3 billion through August 2016. However, reductions in the amount of achieved savings reported to OMB, particularly by the Treasury, resulted in a net decrease of more than $400 million in these savings, compared to amounts we previously reported as planned in 2015. Further, our report noted that, as of December 2016, agencies’ total planned cost savings of about $656 million were more than $3.3 billion less compared to the amounts that we reported in 2015, and more than $2 billion less than OMB’s fiscal year 2018 cost savings goal of $2.7 billion. This reduction in planned savings was the result of eight agencies reporting less in planned cost savings and avoidances in their DCOI strategic plans compared to the savings amounts previously reported to us in November 2015. The reduction also reflected the absence of cost savings information for one agency (Defense) that did not submit its strategic plan in time for our review. In addition, our May 2017 report identified weaknesses in agencies’ DCOI strategic plans. Of the 23 agencies that had submitted their strategic plans at the time of our review, 7 agencies—Agriculture, the Department of Education (Education), DHS, and the Department of Housing and Urban Development (HUD); GSA; the National Science Foundation (NSF); and the Office of Personnel Management (OPM)—had addressed all five required elements of a strategic plan, as identified by OMB (such as providing information related to data center closures and cost savings metrics). The remaining 16 agencies that submitted their plans either partially met or did not meet the requirements. We also pointed out that there were inconsistencies in the reporting of cost savings in the strategic plans of 11 agencies. Given these findings, we recommended that OMB improve its oversight of agencies’ DCOI strategic plans and their reporting of cost savings and avoidances. We also recommended that 16 agencies and Defense (which did not submit a plan in time for our review) complete the missing elements in their strategic plans, and that 11 agencies ensure the reporting of consistent cost savings and avoidance information to OMB. As of December 2018, 10 of the 30 recommendations had not been fully addressed. In a subsequent report that we issued in August 2017, we noted that 22 of the 24 agencies required to participate in the OMB DCOI had reported (collectively) limited progress against OMB’s fiscal year 2018 performance targets for the five optimization metrics. The 2 remaining agencies, Education and HUD, did not own any data centers and, therefore, did not have a basis to report on progress. Specifically, for each of the five targets, no more than 5 agencies reported that they had met or exceeded that specific target. We reported that this limited progress against OMB’s optimization targets was due, in part, to agencies not fully addressing our prior recommendations in this area. In addition, we noted in the report that most agencies had not yet implemented automated monitoring tools to measure server utilization, as required by the end of fiscal year 2018. Specifically, 4 agencies reported that they had fully implemented such tools, 18 reported that they had not yet done so, and 2 did not have a basis to report on progress because they did not own any data centers. Accordingly, we recommended that OMB require that agencies include plans, as part of existing OMB reporting mechanisms, to implement automated monitoring tools at their agency-owned data centers. We also recommended that the 18 agencies that did not have fully documented plans take action, within existing OMB reporting mechanisms, to complete plans describing how they intended to achieve OMB’s requirement to implement automated monitoring tools at all agency-owned data centers by the end of fiscal year 2018. As of December 2018, none of our 19 recommendations had been fully addressed. Most recently, in May 2018, we noted that the 24 agencies participating in DCOI had reported mixed progress toward achieving OMB’s goals for closing their data centers by September 2018. Thirteen agencies reported that they had either already met, or planned to meet, all of their OMB- assigned goals by the deadline. However, 4 agencies reported that they did not have plans to meet all of their assigned goals and 2 agencies were working with OMB to establish revised targets. With regard to agencies’ progress in achieving cost savings, 20 agencies reported, as of August 2017, that they had achieved $1.04 billion in cost savings for fiscal years 2016 and 2017. In addition, the agencies’ DCOI strategic plans identified an additional $0.58 billion in planned savings— for a total of $1.62 billion for fiscal years 2016 through 2018. This total was approximately $1.12 billion less than OMB’s DCOI savings goal of $2.7 billion. This shortfall was the result of 12 agencies reporting less in planned cost savings and avoidances in their DCOI strategic plans, as compared to the savings targets established for them by OMB. In addition, the 24 agencies reported limited progress against OMB’s five data center optimization targets for server utilization and automated monitoring, energy metering, power usage effectiveness, facility utilization, and virtualization. As of August 2017, 1 agency had met four targets, 1 agency had met three targets, 6 agencies had met either one or two targets, and 14 agencies reported meeting none of the targets. Further, as of August 2017, most agencies were not planning to meet OMB’s fiscal year 2018 optimization targets. Specifically, 4 agencies reported plans to meet all of their applicable targets by the end of fiscal year 2018; 14 reported plans to meet some of the targets; and 4 reported that they did not plan to meet any targets. Because GAO had made a number of recommendations to OMB and the 24 DCOI agencies to help improve the reporting of data center-related cost savings and to achieve optimization targets, we did not make new recommendations and noted that, as of March 2018, 74 of the 81 prior recommendations had not been fully addressed. While agencies have made considerable progress, as of December 2018, 47 of the 81 recommendations had not been fully addressed. Agencies Reported Mixed Results in Efforts and Plans to Meet OMB’s Targets for Data Center Closures and Cost Savings According to OMB guidance, agencies were expected to close at least 25 percent of tiered data centers government-wide, by the end of fiscal year 2018. In addition, agencies were to close at least 60 percent of non-tiered data centers government-wide by this same deadline. Further, agencies were expected to reduce government-wide annual costs attributable to physical data centers by a least 25 percent by the end of fiscal year 2018, resulting in savings of at least $2.7 billion. About Half of the Agencies Planned to Meet OMB’s Targets for Data Center Closures The 24 agencies reported mixed results regarding their data center closure progress and plans, when compared with OMB’s goal for each agency to close at least 25 percent of their tiered data centers and at least 60 percent of their non-tiered centers. Specifically, as of August 2018, 13 agencies reported that they had already met the goal of closing 25 percent of their tiered data centers, another 3 agencies reported that they planned to meet the goal by the end of fiscal year 2018, and 6 agencies reported that they did not plan to meet the goal. Further, as of August 2018, 11 agencies reported that they had already met the goal for closing 60 percent of their non-tiered centers, 3 agencies reported that they planned to meet the goal by the end of fiscal year 2018, and 9 agencies reported that they did not plan to meet the goal by the end of fiscal year 2018. Table 1 displays a breakdown of the number of reported tiered and non-tiered data centers and completed and planned closures by agency, as of August 2018. As shown in the figure below, the 24 agencies reported a total of 6,250 data center closures as of August 2018, which represented about half of the total reported number of federal data centers. In addition, the agencies planned 1,009 closures by the end of fiscal year 2018, with an additional 191 closures planned through fiscal year 2023 for a total of 1,200 more closures. This would further reduce the number of open data centers to about 39 percent of the number reported in the agencies’ inventories. Figure 3 provides a summary breakdown of agencies’ data center inventories that were closed, planned for closure, or not planned for closure, as of August 2018. As noted, while about half of the agencies had met, or had planned to meet, their OMB targets as of August 2018, the other half planned to miss one or both of them. Officials from the 11 agencies that did not plan to meet one or both of their closure goals provided various reasons for why they had not planned to do so. For example, several agencies indicated that they were seeking revised closure goals because they viewed their goals as unattainable. Specifically, officials from Interior’s Office of the CIO stated that a number of the department’s non-tiered data centers were either mission-critical or not cost effective to close. Thus, the officials said Interior was working with OMB to establish a revised closure goal. Similarly, Transportation’s Director for IT Compliance stated that the department was working with OMB to establish a revised closure goal. The department reported having 186 tiered data centers in Federal Aviation Administration control towers that it believes should be excluded from its count of data centers when OMB sets the department’s goal for closures. In addition, officials in Defense’s Office of the CIO stated that the OMB closure targets for the department were based on including special purpose processing nodes that are mission critical and, therefore, are not subject to being closed. The officials noted that the department intends to continue operating its enterprise data centers, close its smaller data centers, and work with OMB to remove the special purpose processing nodes from DCOI consideration. When OMB launched DCOI in 2016, agencies originally had until the end of fiscal year 2018 to meet OMB’s stated time frame for closing their data centers. However, the extension of FITARA’s data center consolidation and optimization provisions through fiscal year 2020, and OMB’s planned revisions to DCOI goals, provide the 11 agencies that had not planned to meet one or both of OMB’s closure targets with additional time to meet their goals. Until these agencies take action to close enough data centers to meet OMB’s targets, they may not realize the efficiencies and cost savings that were expected from DCOI. Almost Two-thirds of Agencies Planned to Meet OMB-Assigned Savings Targets Since 2013, federal agencies have been required to report on data center cost savings, with guidance from OMB regarding how agencies were to report cost savings and avoidances. Specifically, the guidance required agencies to report both data center consolidation cost savings and avoidances, among other areas, as part of a quarterly reporting process. FITARA also called for each agency to submit a multi-year strategy for achieving the consolidation and optimization of data centers that included year-by-year quarterly calculations of investment and cost savings through fiscal year 2018, which has now been extended to 2020. In addition, in August 2016, OMB M-16-19 provided guidance on how agencies should implement the requirements of FITARA. Specifically, agencies were to develop a strategic plan that included information on historical cost savings and avoidances due to data center consolidation and optimization through fiscal year 2015. This guidance stated that agency strategic plans were also to include year-by-year calculations of target and actual agency-wide spending and cost savings on data centers from fiscal years 2016 through 2018. Further, the guidance established a DCOI combined cost savings goal of $2.7 billion for all federal agencies to achieve from fiscal years 2016 through 2018. This overall goal was then broken down into agency-specific targets on the IT Dashboard. In August 2018, 22 agencies reported through the quarterly reporting process that they had achieved $1.94 billion in cost savings for fiscal years 2016 through 2018, while 2 agencies reported that they had not achieved any savings. Further, 21 agencies identified an additional $0.42 billion planned through fiscal year 2018, for a total of $2.36 billion in planned savings from fiscal years 2016 through 2018. Nevertheless, this total is about $0.37 billion less than OMB’s goal of $2.7 billion for overall DCOI savings. Figure 4 compares the total achieved savings as reported by the 24 agencies for fiscal years 2016 through 2018 and the agencies’ additional planned savings through 2018 to OMB’s DCOI savings goal for fiscal years 2016 through 2018. The 24 participating DCOI agencies had achieved $1.94 billion in savings as of August 2018. In addition, agencies identified an additional $0.43 billion, for a difference of $0.37 billion between planned and achieved savings from fiscal years 2016 through 2018. Table 2 provides specific data related to each agency’s total planned savings, total achieved savings, and additional planned savings through 2018. As shown in table 2, 13 agencies reported that they had met or planned to meet or exceed their OMB targets, and 3 agencies that did not have an OMB target also identified achieved savings. In contrast, 5 agencies reported that they did not plan to meet their targets. Three agencies did not have a savings target and did not report any achieved savings. Agencies provided various reasons for why they did not plan to meet their savings targets. For example, the Department of Veterans Affairs (VA) reported that the implementation of its DCOI-related projects in fiscal years 2016 through 2018 was dependent on funding approval and might not result in cost savings and avoidances until later in the projects’ life cycles (i.e., fiscal year 2019 or later). In another example, GSA stated that the OMB target may be difficult for the agency to reach due, in part, to the methods OMB used to set the target for fiscal years 2016 through 2018. According to GSA, OMB used the data that GSA had reported on the IT Dashboard regarding the agency’s expenditure for data center infrastructure and reduced that amount by 25 percent. Agencies have now been working toward OMB’s DCOI savings goals since fiscal year 2016; however, almost half of the agencies are still not planning to meet OMB’s targets. Until agencies plan to meet and achieve OMB’s data center-related savings targets, they will likely not realize the expected financial benefits from DCOI. Most Agencies Continued to Report Limited Progress Toward Meeting Optimization Metrics Targets FITARA required OMB to establish metrics to measure the optimization of data centers, including server efficiency, and to ensure that agencies’ progress toward meeting the metrics is made public. Pursuant to FITARA, OMB established a set of five data center optimization metrics intended to measure agencies’ progress in the areas of server utilization and automated monitoring, energy metering, power usage effectiveness, facility utilization, and virtualization. According to OMB, while the server utilization and automated monitoring metric applied to agency-owned tiered and non-tiered data centers, the four remaining metrics applied only to agency-owned tiered centers. OMB’s memorandum also established a target value for each of the five metrics, which agencies were expected to achieve by the end of fiscal year 2018. OMB measures agencies’ progress against the optimization targets using the agencies’ quarterly data center inventory submission and publicly reports this information on its IT Dashboard. Table 3 provides a description of the five data center optimization metrics and target values. As of August 2018, most (22 of the 24) DCOI agencies continued to report limited progress in meeting OMB’s fiscal year 2018 data center optimization targets identified on the IT Dashboard. The remaining 2 agencies—Education and HUD—reported that they did not have any agency-owned data centers in their inventory and, therefore, did not have a basis to measure and report optimization progress. With regard to the data center optimization targets, agencies reported the greatest progress against two metrics: power usage effectiveness and virtualization metrics. Specifically, 8 agencies reported that they had met OMB’s target for power usage effectiveness and 6 agencies reported that they had met the target for virtualization. However, for the energy metering, facility utilization, and server utilization and automated monitoring metrics, no more than 3 agencies reported meeting each. Figure 5 summarizes the 24 agencies’ progress in meeting each optimization target, as of August 2018. As of August 2018, NSF, SSA, and EPA reported the most progress against OMB’s metrics among the 22 agencies with a basis to report— each met 3 targets. Nine agencies reported that they had met only one target, and 10 agencies reported they had not met any of the targets. Further, OMB began requiring the implementation of automated monitoring tools in August 2016; however, as of August 2018, of the 22 agencies with a basis to report, 5 reported that they had either not implemented the tools at any data centers, or had experienced shortcomings in their implementation. For example, the Department of State (State) reported that it had limited centralized monitoring capability and is installing automated monitoring tools in several phases. Thus, these 5 agencies were not able to report any progress against either or both of the server utilization or power usage effectiveness metrics because their data centers lacked the required monitoring tools to measure progress in these areas. The remaining 17 agencies reported that they had implemented the tools in at least one data center. Table 4 depicts the performance of the agencies in meeting OMB targets for data center optimization, as of August 2018. As of August 2018, multiple agencies had made changes to their data center inventory and operational environment, such as closing all agency- owned tiered data centers or implementing automated monitoring tools. These changes impacted which metrics were applicable or an agency’s ability to report on the status of its optimization metrics. For example, GSA reported that it no longer had any agency-owned tiered data centers, and, therefore, did not have a basis to report on four of the five optimization metrics. Additionally, NSF, which previously had only owned one non-tiered data center, migrated from the non-tiered center to a tiered data center as part of its headquarters relocation. Accordingly, NSF began reporting on the metrics applicable to its tiered facility. Further, the Nuclear Regulatory Commission (NRC) did not report on power usage effectiveness due to delays in awarding a new contract that was to include monitoring tools that would impact the ability to report on this metric. Agency officials stated that NRC plans to have the monitoring tools in place during fiscal year 2019. Overall, these changes since last year’s report have resulted in no significant changes to the progression of these agencies on their optimization metrics. In addition, agencies’ limited progress against OMB’s optimization targets was due, in part, to not fully addressing our prior recommendations in this area. As previously mentioned, in March 2016, we reported on weaknesses in agencies’ data center optimization efforts, including that 22 agencies did not meet OMB’s fiscal year 2015 optimization targets. We noted that this was partially due to the agencies having challenges in optimizing their data centers, including in their decentralized organizational structures that made consolidation and optimization difficult, and in competing priorities for resources. In addition, consolidating certain data centers was problematic because the volume or type of information involved required the data center to be close in proximity to the users. Accordingly, we recommended that the agencies take action to improve optimization progress, to include addressing any of the identified challenges. Most agencies agreed with our recommendations or had no comments. However, as of December 2018, only 4 of the 22 agencies had fully addressed them. The continuing shortcomings in data center optimization can also be attributed, in part, to agencies viewing OMB’s optimization metric targets as unrealistic. For example, Transportation stated in its DCOI strategic plan that it could not meet multiple optimization metrics due to funds not being available and competing priorities. In addition, Treasury indicated in its DCOI strategic plan that it struggles to report on automated monitoring because many of its data centers do not have the ability to centrally aggregate and report on central processing unit data. Further, DHS officials noted that it has 7 smaller tiered data centers where it has determined that it is not cost effective to equip those centers with the tools needed to report on metrics such as power usage effectiveness. Given these types of challenges, the targets for each optimization metric may not be realistic for every agency. Unless agencies take action to meet the applicable OMB optimization metrics, their data centers may not operate efficiently enough to provide expected cost savings. Only Two Agencies Planned to Meet OMB’s Fiscal Year 2018 Optimization Targets In addition to reporting current optimization progress on the IT Dashboard, OMB required agencies to include in their DCOI strategic plans planned performance levels for fiscal year 2018 for each optimization metric. However, according to the 24 agencies’ DCOI strategic plan information as of August 2018, only 2—Commerce and the U. S. Agency for International Development (USAID)—reported plans to fully meet their applicable targets by the end of fiscal year 2018. Of the remaining agencies, 14 reported plans to meet some, but not all, of the targets; 6 reported that they did not plan to meet any targets, and—as already discussed—Education and HUD did not have a basis to report planned optimization milestones because they did not report having any agency- owned data centers. Figure 6 summarizes agencies’ progress, as of August 2018, in meeting OMB’s optimization targets and planned progress to be achieved by September 2018. At the time of our review, only two agencies planned to meet all of their applicable targets, and it was doubtful that the agencies would be able to achieve OMB’s collective optimization target of at least $2.7 billion in cost savings by the end of fiscal year 2018. Until the remaining agencies take the steps necessary to meet their optimization targets, it is unlikely that these agencies will achieve the expected benefits of optimization and the resulting cost savings. Selected Agencies Highlighted Successful DCOI Practices As we noted previously in this report, many agencies have reported challenges that have hindered their efforts to meet OMB’s DCOI targets. However, a number of agencies have also reported success in meeting OMB’s targets ahead of DCOI’s end of fiscal year 2018 deadline. As noted in our methodology section, six agencies that were among the best performers in achieving data center closures, cost savings, and optimization performance reported a number of key practices that had contributed to their success. These practices were: obtaining executive leadership support for consolidation and using experiences and lessons learned to refine consolidation increasing the use of cloud and shared services to consolidate or optimize data center operations; emphasizing closing data centers to meet OMB targets and achieve increasing the use of virtualization to optimize data centers; and employing an organization-wide communications plan to facilitate adoption of consolidation and optimization activities. Obtaining Executive Leadership Support for Consolidation and Optimization Activities Five of the six agencies (Agriculture, Commerce, Justice, EPA, and GSA) reported that their success in consolidation and optimization activities was due to obtaining support from executive leadership for the agency’s consolidation efforts. Each agency obtained sponsorship and support from its executive leadership (e.g., Deputy Secretary or agency CIO), such as through a memorandum or policy that directed all agency offices to participate in, or comply with, the consolidation effort. For example, The Deputy Secretary for Agriculture issued a memorandum in 2017 that, among other things, declared the department’s intent to consolidate from 39 data centers down to 2 by the end of 2019. According to officials in the Office of the CIO, this memorandum from the Secretary’s office focused all data center owners on the same project task of reducing the data center inventory. The Commerce CIO and the department’s CIO Council provided overall governance through organizational policies, processes, and procedures for the department’s data center consolidation effort. Leveraging this departmental guidance, each component of Commerce developed its own consolidation plan that identified specific approaches and activities. Using these plans, the department and its components focused on reducing spending on redundant software, infrastructure, and data center operations. The Deputy Attorney General issued a memorandum in 2014 to the heads and CIOs of all components. This memorandum formally established Justice’s Data Center Transformation Initiative, established the Department Program Review Board to provide oversight for the initiative, and also directed the consolidation of all data centers into 3 enterprise facilities. In addition, Justice’s CIO issued a memorandum to component CIOs that provided additional details on how to execute planned activities and established further governance associated with the initiative. These memoranda provided clear leadership buy-in and support for the department’s data center consolidation and optimization activities that could be used to resolve any challenges or issues at the departmental level. EPA attributed much of its DCOI success to a top-down approach from its CIO office, saying that such support was critical to achieve data center closures. For example, EPA leadership decided to adopt and enforce geographical consolidation of data centers within major areas to minimize costs of consolidation while still meeting closure objectives. In doing so, the agency leadership provided clear direction and support for the agency’s consolidation effort by adopting the strategy to consolidate data centers within specific geographic regions. GSA reported that it obtained leadership commitment that made its data center consolidation and optimization activities a priority. The agency noted that having strong CIO and executive leadership was important for sponsoring technology modernization. As a result of the buy-in, the agency reported that it had minimized resistance to change and improved acceptance of its consolidation and optimization activities. Using Experiences and Lessons Learned to Refine Consolidation Planning Four agencies (Agriculture, Commerce, Justice, and SSA) reported that their success with consolidation and optimization activities was due to the use of a refined consolidation plan or process. Each of these agencies developed an initial consolidation plan or process for closing data centers, and then refined their procedures based on their experiences and lessons learned as data centers were closed. For example, Agriculture developed a set of streamlined processes to facilitate DCOI closures that were based on the experiences gained from successful data center closures under FDCCI. The set of processes consisted of 5 steps: The planning step included the discovery and documentation of all data center assets, including applications and IT hardware, in a given data center. In addition, this step involved identifying the necessary resources to move the applications and associated data to a target data center. The preparation step included identification of the target data center and development of a project schedule. The data migration step included moving both applications and data to the target data center or cloud-services, as planned. The testing step included ensuring the applications and data that were moved were integrated into the target data center, and functional testing to ensure that the applications worked and data was accessible. The application cutover step included putting the migrated applications and data into operation and closing the original data center. Using and refining this set of processes allowed the department to become more efficient in closing its data centers. After closing 46 data centers in fiscal years 2011 through 2014, the department closed 2,185 data centers over the next 2 years. In total, Agriculture reported that it had closed 2,253 data centers as of August 2018. Commerce established departmental guidance and then each departmental component leveraged that guidance to develop its own consolidation plan. The plans identified specific approaches and activities intended to achieve the stated goals and milestones. According to Commerce, the department and its components leveraged their IT planning processes and established IT governance to, among other things, reduce spending on redundant commodity software, infrastructure, and operations. Justice’s Office of the CIO developed a master plan for the department’s data center consolidation effort in June 2015. The plan included a planning framework, transformation approach, and a master schedule for data center moves and closures. It also included process steps similar to those used by Agriculture. Further, Justice’s plan noted that the department would use its initial closure efforts to gain experience and to refine its plans. Justice reported that it used the plan’s schedule and semi-monthly progress reports to ensure that consolidation activities stayed on schedule, or the department could make adjustments as needed. As a result, the department closed 84 of its 110 data centers and achieved more than $128 million in cost savings and avoidances as of August 2018. SSA used a project management framework process and controls that it believed efficiently addressed requirements, critical path, and risk management. In addition, SSA reported that it used an incremental development approach to its data center optimization plans, with each project expected to accomplish specific tasks that would lead to another project. Accordingly, SSA noted that the agency used a multi-year plan with many initiatives focused on specific goals. Using this approach, the agency successfully moved SSA’s operations and infrastructure from an older facility to the newly-built National Support Center. The agency reports that this facility is state-of-the-art and provides similar capabilities and efficiencies to major cloud service providers. Increasing the Use of Cloud and Shared Services to Consolidate or Optimize Data Center Operations Three agencies (Commerce, GSA, and SSA) also attributed their success in consolidation and optimization activities to increasing their agency’s use of cloud and shared services. In doing so, each agency emphasized the move of data center assets and systems to cloud services to optimize their data centers and reduce costs. For example, Commerce identified moving to cloud services and utilizing shared services as being most effective in closing data centers. As an example, the department cited the National Oceanic and Atmospheric Administration’s (NOAA) “cloud-first” policy that emphasized using cloud services rather than an agency-owned physical data center whenever feasible. The agency attributed its ability to handle increased traffic as an operational benefit of its increased use of cloud services. For example, NOAA did not have the capacity in its agency- owned facilities to meet the computing demands and requirements of a sudden increase in web traffic on the websites for NOAA and the National Hurricane Center, such as during Hurricanes Irma and Harvey in 2017. Commerce stated that using cloud services allowed NOAA to handle 4.7 billion page hits during Hurricane Harvey over a 6-day span, ensuring the websites were not adversely impacted by the increase in traffic. GSA reported that it focused on moving services from agency-owned tiered and non-tiered data centers to cloud services or to shared centers. As a result, GSA had closed 118 data centers as of August 2018, including all of the agency’s tiered centers. SSA developed an agency cloud initiative that encourages the adoption of cloud technologies as part of the agency’s infrastructure modernization. The agency reported that it is employing a hybrid cloud strategy that is comprised of both private cloud and public cloud services for the agency’s back office applications. By doing so, the agency will consolidate and standardize SSA’s IT infrastructure systems and software to simplify management of those resources and reduce costs. Emphasizing the Closure of Data Centers to Meet OMB Targets and Achieve Cost Savings Three agencies (Agriculture, Justice, and EPA) reported that their success in consolidation and optimization activities was due to focusing on the closure of data centers. In doing so, they emphasized the importance of closing data centers to reduce costs and achieve cost savings and avoidances. For example, Agriculture determined that the costs to improve DCOI performance metrics in its agency-owned data centers were prohibitive. Accordingly, the department decided that the only viable alternative was to close data centers to remove underperforming centers and improve optimization metrics performance and reduce costs. As a result, Agriculture reported that it had closed 2,253 data centers through August 2018. In addition, the department reported that it had improved its security posture, reduced its real estate footprint, and achieved realized cost savings and avoidance of $51.8 million from fiscal year 2012 through 2018. Justice reported that it took a practical approach to selecting the data centers that would remain as its enterprise facilities, considering factors such as the number of physical servers that could be eliminated, the efficiency of the remaining hardware, and potential labor savings. The department reported that it focused on retaining more efficient data centers (e.g., those with more efficient use of electricity or virtualization), rather than simply keeping its biggest existing data centers. As a result, Justice has closed 84 of its 110 data centers and achieved more than $128 million in cost savings and avoidances as of August 2018. EPA identified geographical consolidation as its best approach to meeting DCOI goals. Specifically, in its data center consolidation plan, the agency stated that, for geographic areas where it had multiple data centers, a single facility was identified into which data center IT assets would be consolidated. Using this approach, EPA had closed 43 of its 83 data centers as of August 2018. Increasing the Use of Virtualization to Optimize Data Centers Three agencies (Commerce, EPA, and SSA) reported that their success in consolidation and optimization activities also was due to focusing on the increased use of virtualization to run more software on the same or a reduced amount of servers. In doing so, the agencies expected to reduce costs by avoiding the purchase of additional servers to meet computing demands or eliminating unnecessary hardware and floor space in their data centers. For example, Commerce focused on moving systems from physical hardware to virtual servers, as part of its component offices’ plans to update technology and in cases where the systems did not require a specific type of server. Using this approach, the department reported that it had reduced the number of physical servers in its data centers, and was working to improve server utilization. The department also cited the ability to automatically increase or decrease computing capability through virtualization, such as when NOAA handled the increased traffic to its hurricane-related web pages during Hurricanes Irma and Harvey in 2017. EPA used the agency’s data center consolidation plan to implement an agency-wide “physical-to-virtual” policy that required offices to convert existing physical servers to virtual servers wherever possible. The agency also defined server and software standards for virtualized platforms. SSA reported that the agency’s goal, using its “Virtual 1st” policy, was to have failover capability within the data center, disaster recovery capability for both data centers, and balanced load capacity between data centers. The agency reported that it has continued to virtualize not only servers but storage and network applications, as well. For example, SSA stated that it has taken steps to virtualize as much storage as possible and used similar techniques to reduce the physical hardware footprint on the data center floor, as well as power, cooling, and network bandwidth requirements. Employing an Organization-wide Communications Plan to Facilitate Adoption of Consolidation and Optimization Activities Two agencies (Justice and GSA) reported that their success in consolidation and optimization activities was due to employing an organization-wide communications plan. In doing so, the agencies adopted a structured method for communicating with agency offices to improve acceptance and adoption of consolidation and optimization activities. This also facilitated conflict resolution. For example, Justice reported that it prioritized communications related to its Data Center Transformation Initiative and established an all-encompassing approach to initiative-related communications. To help communicate all related directives, strategies, plans, statuses, and accomplishments, the department used a variety of methods that included: regular meetings to share information, a dedicated email box to provide easy communication for answers or information, without the need to know specific individuals, an intranet web page that provided general information, instructions, templates, decisions, status information, and accomplishments related to the initiative; and email broadcasts on an as-needed basis. GSA reported that it communicated and collaborated frequently with business stakeholders to identify the best time frames to move systems, stagger transfers to minimize impact, and determine which systems could be virtualized. The agency indicated that these important factors required continuous communication between system owners, system administrators, and business leadership. As a result, the agency experienced minimal staff resistance to change and a commitment to reach a consensus on moving forward with the agency’s consolidation efforts. The aforementioned practices included elements of sound management techniques, such as gathering leadership support for a project and developing a communications plan to foster adoption of organizational changes. The practices also included activities that aligned with the core tenets of DCOI to consolidate inefficient infrastructure, optimize existing facilities, and achieve cost savings. Further, these practices each proved effective for multiple agencies and, while they were not the only practices that could be effective, they represent concepts that could provide the foundation for an effective data center consolidation and optimization program. Conclusions Federal data center consolidation efforts have been underway since 2010 and OMB’s fiscal year 2018 targets provided clear and transparent goals that helped define the tangible benefits that DCOI was expected to provide. However, most agencies continue to report mixed progress against those targets. Although agencies have taken action to close about half of the data centers in their combined inventories, 11 agencies did not plan to meet all of their closure targets. Further, the data center closures were expected to drive cost savings and avoidances and, to the agencies’ credit, the closures have led to more than $2.37 billion in planned and achieved cost savings and avoidances from fiscal years 2016 through 2018. However, five agencies did not plan to meet their cost savings targets. Until agencies consolidate the data centers required to meet their targets, as well as identify and report the associated cost savings, they will be challenged to realize expected efficiencies and the full benefits of DCOI will not be fully realized. Similarly, although OMB first established optimization metrics in May 2014, agencies continue to report only limited progress against the current performance targets. While two agencies do not have a basis to report any progress as they do not own any data centers, only two agencies reported that they planned to achieve all of DCOI’s fiscal year 2018 optimization targets. Ensuring the optimized performance of data centers is a key component to meeting OMB’s DCOI-wide savings goal and the 20 agencies that did not have plans to meet their targets call into question whether DCOI will realize its full potential savings. Although many agencies have struggled to meet their individual DCOI targets, other agencies have successfully met OMB’s goals for data center closures, savings, and optimization. Six such agencies that we identified reported on the importance of gathering leadership support, effective communication, and alignment with the core tenets of DCOI. Key practices such as these can play an important role in helping agencies better meet the overall goals and mission of DCOI. Recommendations We are making a total of 36 recommendations to 22 of the 24 agencies in our review. Specifically: The Secretary of Agriculture should take action to meet the data center optimization metric targets established by OMB under DCOI. (Recommendation 1) The Secretary of Commerce should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 2) The Secretary of Defense should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 3) The Secretary of Defense should identify additional savings opportunities to achieve the targets for data center-related cost savings established under DCOI by OMB. (Recommendation 4) The Secretary of Defense should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 5) The Secretary of Energy should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 6) The Secretary of Energy should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 7) The Secretary of the Department of Health and Human Services (HHS) should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 8) The Secretary of HHS should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 9) The Secretary of DHS should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 10) The Secretary of DHS should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 11) The Secretary of Interior should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 12) The Secretary of Interior should take action to meet the data center- related cost savings established under DCOI by OMB. (Recommendation 13) The Secretary of Interior should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 14) The Attorney General should take action to meet the data center optimization metric targets established for Justice under DCOI by OMB. (Recommendation 15) The Secretary of the Department of Labor (Labor) should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 16) The Secretary of State should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 17) The Secretary of State should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 18) The Secretary of Transportation should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 19) The Secretary of Transportation should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 20) The Secretary of Treasury should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 21) The Secretary of VA should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 22) The Secretary of VA should take action to meet the data center-related cost savings established under DCOI by OMB. (Recommendation 23) The Secretary of VA should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 24) The Administrator of EPA should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 25) The Administrator of EPA should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 26) The Administrator of GSA should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 27) The Administrator of the National Aeronautics and Space Administration (NASA) should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 28) The Director of NSF should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 29) The Chairman of NRC should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 30) The Director of OPM should take action to meet the data center-related cost savings established under DCOI by OMB. (Recommendation 31) The Director of OPM should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 32) The Administrator of the Small Business Administration (SBA) should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 33) The Commissioner of SSA should take action to meet the data center- related cost savings established under DCOI by OMB. (Recommendation 34) The Commissioner of SSA should take action to meet the data center optimization metric targets established under DCOI by OMB. (Recommendation 35) The Administrator of USAID should take action to meet the data center closure targets established under DCOI by OMB. (Recommendation 36) Agency Comments and Our Evaluation We requested comments on a draft of this report from OMB and the 24 agencies that we reviewed. Of the 22 agencies to which we made recommendations, 11 agencies agreed with our recommendations; three agencies agreed with some portion, but not all of the recommendations; one agency disagreed with our recommendations; and seven agencies did not state whether they agreed or disagreed with the recommendations. In addition, OMB and two agencies to which we did not make recommendations stated that they had no comments. Further, multiple agencies provided technical comments, which we have incorporated, as appropriate. The following 11 agencies agreed with our recommendations: In written comments from Commerce, State, NASA, SBA, and SSA, the agencies stated that they agreed with the recommendations and indicated their intent to address them. State also provided technical comments, which we have incorporated, as appropriate. The agencies’ comments are reprinted in appendices II through VI. In written comments, Energy agreed with our recommendations to meet its data center closure and optimization metric targets, and described actions that the department planned to take in order to address the recommendations. Energy initially estimated that it would complete these actions by March 1, 2019; however, the department subsequently revised its estimated completion date to April 15, 2019. Energy also provided technical comments, which we have incorporated, as appropriate. Energy’s comments are reprinted in appendix VII. In written comments, VA agreed with our recommendations to meet its data center closure, cost savings, and optimization metric targets. In addition, the department requested that we close our recommendation related to data center closures on the basis of its planned actions to implement a new inventory data collection tool and methodology to improve how the department collects data center inventory information, and a positive trend in its data center closures. The department estimated that its planned actions would be completed in March 2019 and reported that, as of November 2018, it had closed 78 data centers in fiscal year 2018, as compared with 24 in fiscal year 2017. However, as noted earlier in this report, we found that VA did not plan to meet the closure goal for either tiered or non-tiered data centers, which was the basis for our recommendation. While we acknowledge and encourage VA’s reported closure progress, the department still has not met its DCOI closure goals, as we recommended. Further, VA did not provide an update on the status of its planned actions in time for us to address them in this report. As such, we maintain that this recommendation is still appropriate. In addition, VA referred to OMB’s proposed changes to DCOI guidance when describing actions that it planned to take to meet the department’s cost savings and optimization metrics targets. However, OMB staff told us that the August 2016 DCOI guidance will remain in effect until the revised DCOI guidance is formally issued. Once OMB’s new DCOI guidance is finalized, we plan to assess agency progress against any revised targets, and we will continue to monitor the department’s efforts to address our recommendation. VA’s comments are reprinted in appendix VIII. We received emails from officials of Agriculture, Justice, Transportation, and OPM which stated that these agencies agreed with the recommendations we directed to them. In addition, three agencies agreed with some portion, but not all of our recommendations directed to them: In written comments, Defense stated that it agreed with our recommendation to meet its data center closure targets. However, the department partially agreed with our two other recommendations: to identify additional data center-related savings opportunities and to meet OMB’s data center optimization metric targets. In partially agreeing with our recommendation on data center savings, Defense asserted that it had already identified significant cost savings through activities such as the identification of system migration candidates and the use of cloud services, among others. The department further stated that, while it would continue to optimize its data centers, the need for IT would continue to grow, and this growth might ultimately lead to an increase in total data center costs, despite overall per unit cost reductions. However, the department’s planned savings of $205.46 million represented only 11 percent of its $1.8 billion savings goal by the end of fiscal year 2018 and, as such, this limited progress by the department formed the basis for our recommendation. As discussed in our report, OMB plans to revise DCOI guidance and work with agencies to set agency-specific targets. According to OMB staff, until the guidance is revised, the current guidance and its targets are still applicable. For these reasons, we maintain that our recommendation is still appropriate. Further, in partially agreeing with our recommendation to meet optimization metric targets, Defense stated that the department will continue to drive towards the achievement of data center optimization targets. It added, however, that it would not invest resources to improve the efficiency of data centers planned for closure and that, as a result, the composite view of Defense’s data center efficiency would fall short of meeting OMB’s targets. Our review found that Defense did not plan on meeting any of OMB’s five data center optimization metric targets by the end of fiscal year 2018. This finding was the basis for our recommendation. We acknowledge Defense’s position that investing resources into optimizing data centers that are already planned for closure would not be the best use of taxpayer dollars. We also noted in our report that OMB had proposed revising its optimization metrics, and that any such changes had not yet been finalized. Our recommendation is not intended to imply that an agency should meet a particular version of OMB targets but, rather, that the agency should meet any targets that are established by OMB. This would include any future changes to DCOI targets. Accordingly, we maintain that our recommendation is still appropriate and will continue to monitor the department’s efforts to address our recommendation. Defense’s comments are reprinted in appendix IX. In written comments, DHS stated that it agreed with our recommendation to meet its data center closure targets and disagreed with our recommendation to meet its data center optimization metric targets. Specifically, the department noted that it had met its tiered data center closure targets, and was reviewing the status of its remaining open non-tiered data centers. The department added that it expected to complete this activity by March 31, 2019. However, the department did not provide an update on its efforts in time to be included in this report. While we encourage DHS’s continued efforts to close its remaining non-tiered data centers, we note that the department’s letter cites an inventory of 18 open non-tiered facilities, which differs significantly from the 202 non-tiered centers counted in our draft report, and which DHS officials confirmed in November 2018. According to the department, this discrepancy is because OMB issued revised inventory reporting requirements in November 2018, and these revised requirements exempted certain types of facilities from DCOI reporting and resulted in the lower number. These changes in reporting requirements are similar to the proposed, but not yet finalized, revisions to the DCOI policy that are discussed earlier in this report. However, OMB staff told us that the August 2016 DCOI guidance will remain in effect until the revised DCOI guidance is formally issued. Once OMB’s new DCOI guidance is finalized, we plan to assess agency progress against any revised targets, and we will continue to monitor the department’s efforts to address our recommendation. Further, in disagreeing with our recommendation on meeting optimization metrics, the department stated that, while the recommendation was applicable under the original DCOI guidance that OMB issued in August 2016, OMB’s proposed changes to DCOI guidance would exempt most, if not all, DHS agency-owned data centers from the optimization metrics. Consequently, the department requested that our recommendation be closed. In our review, we found that the department did not plan on meeting any of OMB’s five data center optimization metric targets established under DCOI. This finding was the basis for our recommendation on meeting optimization metrics. Also, while OMB has proposed changes to its metrics, as we noted previously, it has not provided a date for when any such proposed changes will be finalized and implemented; and, according to OMB staff, until the changes to DCOI guidance are finalized, the current guidance is still applicable. Further, our recommendations do not specify that an agency should meet any particular version of OMB targets, but rather, that an agency should meet the targets established by OMB. This would include any future changes to DCOI targets. Accordingly, we maintain that our recommendation is still appropriate. DHS also provided technical comments, which we have incorporated, as appropriate. DHS’s comments are reprinted in appendix X. In written comments, Interior stated that it partially agreed with our recommendation to meet its data center closure targets and disagreed with our two recommendations to meet its data center-related cost savings target and its data center optimization metric targets. For all three recommendations, the department stated that OMB had proposed changes to DCOI guidance that would result in new targets for closures, cost savings, and optimization metrics and that Interior planned to adopt the new policy and work through OMB to establish its new targets. As noted in our report, Interior met its target for tiered data center closures, but did not plan to meet the closure goal for non-tiered data centers. Further, the department planned on achieving only $15.95 million of its $88.19 million savings target (18 percent) by the end of fiscal year 2018, and did not plan on meeting any of OMB’s five data center optimization metric targets. These three findings were the basis for our recommendations to the department. We also noted that, as part of OMB’s proposed changes to DCOI guidance, it planned to work with agencies to set agency-specific targets for data center closures and planned to modify the metrics currently used by agencies to monitor the performance of their data centers. However, as previously mentioned, OMB has not provided a date for when these proposed changes will be finalized and implemented and, according to OMB staff, until the changes to DCOI guidance are finalized, the 2016 guidance is still applicable. Furthermore, our recommendations do not specify that an agency should meet any particular version of OMB targets, but should meet any targets that are established by OMB. This would include any future changes to DCOI targets. As such, we maintain that our recommendations are appropriate. Interior’s comments are reprinted in appendix XI. One agency disagreed with all of our recommendations: In written comments, HHS disagreed with our two recommendations to meet its data center closure targets and data center optimization metric targets. In regard to both recommendations, the department disagreed with being held to what it termed “expired requirements” from DCOI guidance, pending the assignment of new targets being established by OMB. As noted in our report, HHS met its target for tiered data center closures, but did not plan to meet the closure target for non-tiered data centers. We also found that HHS did not meet any of OMB’s five optimization metric targets and had planned to meet only one of the five by end of fiscal year 2018. These findings were the basis for the two recommendations that we made to the department. We also noted that, as part of OMB’s proposed changes to DCOI guidance, OMB planned to work with agencies to set agency-specific targets for data center closures and planned to modify the metrics currently used by agencies to monitor the performance of their data centers. However, as previously mentioned, OMB did not provide a date for when these proposed changes will be finalized and implemented and, according to its staff, until the changes to DCOI guidance are finalized, the current guidance is still applicable. Further, our recommendations do not specify that an agency should meet any particular version of OMB targets, but rather, that the agency should meet the targets established by OMB. This would include any future changes to DCOI targets. Accordingly, we maintain that our recommendations are still appropriate. HHS’s comments are reprinted in appendix XII. Further, seven agencies did not agree or disagree with the recommendations: In written comments, EPA did not state whether it agreed or disagreed with our recommendations to meet its data center closure and data center optimization metrics targets. However, the agency requested that we close our recommendations, citing its reported progress in closing 21 of 34 targeted data centers and OMB’s proposed changes in its draft DCOI guidance that could result in revised closure targets and optimization metrics. As stated in our report, we found that EPA did not plan to meet its closure target for tiered or non-tiered data centers, nor did it plan to meet its data center optimization targets; these findings were the basis for our recommendations. We also noted that, as part of OMB’s proposed changes to DCOI, OMB planned to work with agencies to set agency-specific targets for data center closures and planned to modify the metrics currently used by agencies to monitor the performance of their data centers. However, OMB has not provided a date for when these proposed changes will be finalized and implemented and, according to OMB staff, until the changes to DCOI guidance are finalized, the current guidance is still applicable. Further, our recommendations do not specify that an agency should meet any particular version of OMB targets, but that it should meet the targets established by OMB. This would include any future changes to DCOI targets. Accordingly, we maintain that our recommendations are appropriate and should remain open. EPA also provided technical comments, which we have incorporated, as appropriate. The agency’s comments are reprinted in appendix XIII. In written comments, GSA did not state whether it agreed or disagreed with our recommendation to meet the agency’s data center optimization metrics targets. Specifically, the agency stated that it had complied with revised inventory reporting requirements, which OMB provided to agencies in November 2018 and which eliminated non- tiered data centers from the requirement to meet optimization targets. As a result, the agency noted that it no longer had a basis to measure and report on the one metric our report cited as applicable to GSA (i.e., server utilization and automated monitoring) and asked that we withdraw the recommendation. These changes in reporting requirements are similar to the proposed, but not yet finalized, revisions to the DCOI policy that are discussed earlier in the report. However, OMB staff told us that the August 2016 DCOI guidance is still in effect until the revised DCOI guidance is formally issued. Until OMB’s new DCOI guidance is finalized and agency progress against any revised targets can be evaluated, we maintain that our recommendation to meet the agency’s optimization metrics targets is appropriate, and we will continue to monitor the agency’s efforts to address it. GSA’s comments are reprinted in appendix XIV. In written comments, NSF did not state whether it agreed or disagreed with our recommendation. The agency’s comments are reprinted in appendix XV. In written comments, NRC agreed with the draft report, but did not state whether it agreed or disagreed with our recommendation. The agency’s comments are reprinted in appendix XVI. In written comments USAID did not state whether it agreed or disagreed with the draft report’s recommendation but agreed with our finding that the agency no longer had any tiered data centers. However, USAID stated that it had met DCOI’s closure targets for the agency by closing its 4 non-tiered data centers, and requested that we close our recommendation to meet those targets. While we encourage USAID’s continued efforts to close its remaining non-tiered data centers, we note that the agency’s letter cites an inventory of 4 non-tiered facilities, which differs significantly from the 83 non-tiered centers counted in our draft report, and which USAID officials confirmed in October 2018. As USAID communicated in subsequent emails, this discrepancy is because OMB issued revised inventory reporting requirements in November 2018 and these revised requirements exempted certain types of facilities from DCOI reporting, which resulted in the lower number. These changes in reporting requirements are similar to the proposed, but not yet finalized, revisions to the DCOI policy that are discussed earlier in the report. However, OMB staff told us that the August 2016 DCOI guidance is still in effect until the revised DCOI guidance is formally issued. Until OMB’s new DCOI guidance is finalized and agency progress against any revised targets can be evaluated, we maintain that our data center closure recommendation is appropriate, and we will continue to monitor the agency’s efforts to address it. The agency’s comments are reprinted in appendix XVII. In emails received from Labor’s GAO liaison in the department’s Office of the Assistant Secretary for Policy on January 8, 2019, and from an audit liaison in Treasury’s Office of the CIO on February 1, 2019, both departments did not state whether they agreed or disagreed with our respective recommendations. Finally, in emails received from a Management and Program Analyst in Education’s Office of the Secretary/Executive Secretariat on January 8, 2019; an audit liaison in HUD’s Office of the CIO, Audit Compliance Branch on February 15, 2019; and a GAO liaison in OMB’s Office of General Counsel on February 25, 2019, these agencies stated that they had no comments on the draft report. We are sending copies of this report to interested congressional committees, the Director of OMB, the secretaries and heads of the departments and agencies addressed in this report, and other interested parties. In addition, the report will be available at no charge on GAO’s website at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-4456 or harriscc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix XVIII. Appendix I: Objectives, Scope, and Methodology Our objectives for this engagement were to (1) determine agencies’ progress in data center closures and achievement in related savings to date and describe plans for future savings, (2) evaluate the agencies’ progress against OMB’s data center optimization targets, and (3) identify effective agency practices for achieving data center closures, cost savings, and optimization. To address the first objective, for data center closures, we obtained and analyzed August 2018 data center inventory documentation from the 24 departments and agencies (agencies) that participate in OMB’s Data Center Optimization Initiative (DCOI). To determine data center closures to date, we totaled their reported closures from fiscal year 2010 through August 2018 and to identify future closures, we totaled their reported planned closures through fiscal year 2018. We also compared agencies’ completed and planned closures to OMB’s fiscal year 2018 consolidation goals, as documented in its August 2016 memorandum (M-16-19). To verify the quality, completeness, and reliability of each agency’s data center inventory, we compared information on completed and planned data center closures to similar information reported on OMB’s IT Dashboard—a public website that provides information on federal agencies’ major IT investments. We also checked for missing data and other errors, such as missing closure status information. In some cases identified, we followed-up with agency officials to obtain further information. We determined that the data were sufficiently complete and reliable to report on agencies’ consolidation progress and planned closures. For cost savings and avoidance we obtained and analyzed documentation from the 24 DCOI agencies. This documentation is required by OMB’s March 2013 and August 2016 memorandums and included the agencies’ quarterly reports of cost savings and avoidances posted to their digital services websites and their DCOI strategic plans. To determine cost savings achieved, we totaled agencies’ reported savings and avoidances from the start of fiscal years 2012 through August 2018, as found in the August 2018 quarterly reports posted to the agencies’ digital services websites. To identify future planned savings, we totaled the agencies’ projected savings and avoidances from fiscal years 2016 through 2018, as reported in their DCOI strategic plans. To assess the quality, completeness, and reliability of each agency’s data center consolidation cost savings information, we used the latest version of each agency’s quarterly cost savings report and DCOI strategic plan, as of August 2018. We also reviewed the quarterly reports and DCOI strategic plans for missing data and other errors, such as missing cost- savings information. In addition, we compared agencies cost savings and avoidances with data from our most recent data center consolidation report. As a result, we determined that the data were sufficiently complete and reliable to report on agencies data center consolidation cost-savings information. For our second objective, we analyzed the August 2018 data center optimization progress information of the 24 DCOI agencies. This progress information was obtained from the IT Dashboard—an OMB public website that provides information on federal agencies’ major IT investments. To assess agencies’ planned optimization progress, we obtained the planned optimization performance from the 22 agencies’ DCOI strategic plans. We then compared the agencies’ current and planned optimization progress information to OMB’s fiscal year 2018 optimization targets, as documented in its August 2016 memorandum. Although OMB’s memorandum establishes a single optimization target value for the server utilization and automated monitoring metric, the IT Dashboard displays agencies’ progress for tiered and non-tiered data centers separately. To report consistently with OMB’s implementation memorandum, we combined the progress information for tiered and non-tiered data centers into a single assessment in this report. In addition, to assess the reliability of the planned optimization milestones in the DCOI strategic plans, we reviewed agencies’ documentation to identify any missing or erroneous data. We also compared the planned data center optimization milestones contained in agencies’ documentation against current optimization progress information obtained from the IT Dashboard; we then discussed any discrepancies or potential errors that we identified with agency officials to determine the causes or request additional information. As a result of these efforts, we were able to determine whether each agency’s strategic plan information was sufficiently reliable for reporting on plans to meet or not meet OMB’s fiscal year 2018 optimization targets. To assess the reliability of agencies’ optimization progress information on OMB’s IT Dashboard, we reviewed the information for errors or missing data, such as progress information that was not available for certain metrics. We also compared agencies’ optimization progress information across multiple reporting quarters to identify any inconsistencies in agencies’ reported progress. We discussed with staff from OMB’s Office of the Federal Chief Information Officer any discrepancies or potential errors identified to determine the causes. To identify effective agency practices for achieving data center closures, cost savings, and optimization progress, we selected two of the highest performing departments or agencies for each of those three data center areas that we reported on in our May 2018 report. For the data center inventory closures area, we selected the Departments of Agriculture (Agriculture) and Justice (Justice) from among the five agencies that had, as of August 2017, reached or exceeded both their tiered and non-tiered data center closure targets for the end of fiscal year 2018. For the cost savings area, we identified two departments and two small agencies reporting the highest cost savings DCOI to date, as of August 2017. From those, we selected one department (Commerce) and one small agency (the General Services Administration) to provide balance relative to agency size. For effective practices related to optimization performance, we reviewed agencies’ reported optimization performance as of August 2017 and selected the two highest-performing agencies in this area (the Social Security Administration and the Environmental Protection Agency), since they were the only two agencies reporting that they met more than half of OMB’s optimization targets. Selecting these agencies was designed to provide anecdotal information that could assist agencies struggling with DCOI implementation. The examples they provided are not findings nor should they be taken to be representative of all the agencies participating in DCOI. We asked each selected agency to identify practices that they found effective in implementing DCOI at their agency and in meeting OMB’s established targets in each of the areas, not just the area for which they were selected. We also solicited examples that demonstrated how those practices helped agency implementation or the benefits from implementing DCOI. Additionally, we considered information and examples that these agencies provided as part of our work to identify FITARA best practices. We analyzed the responses to determine the practices and reported those that were identified by at least two agencies. We conducted this performance audit from April 2018 to April 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the Department of Commerce Appendix III: Comments from the Department of State Appendix IV: Comments from the National Aeronautics and Space Administration Appendix V: Comments from the Small Business Administration Appendix VI: Comments from the Social Security Administration Appendix VII: Comments from the Department of Energy Appendix VIII: Comments from the Department of Veterans Affairs Appendix IX: Comments from the Department of Defense Appendix X: Comments from the Department of Homeland Security Appendix XI: Comments from the Department of the Interior Appendix XII: Comments from the Department of Health and Human Services Appendix XIII: Comments from the Environmental Protection Agency Appendix XIV: Comments from the General Services Administration Appendix XV: Comments from the National Science Foundation Appendix XVI: Comments from the Nuclear Regulatory Commission Appendix XVII: Comments from the U.S. Agency for International Development Appendix XVIII: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, individuals making contributions to this report included Dave Powner (director), Dave Hinchman (assistant director), Justin Booth (analyst-in-charge), Alexander Bennett, Chris Businsky, Nancy Glover, and Jonathan Wall.

Why GAO Did This Study In December 2014, Congress enacted federal IT acquisition reform legislation that included provisions related to ongoing federal data center consolidation efforts. OMB's Federal Chief Information Officer launched DCOI to build on prior data center consolidation efforts; improve federal data centers' performance; and establish goals for inventory closures, cost savings and avoidances, and optimizing performance. The 2014 legislation included a provision for GAO to annually review agencies' data center inventories and strategies. Accordingly, GAO's objectives were to (1) evaluate agencies' progress and plans for data center closures and cost savings; (2) assess agencies' progress against OMB's data center optimization targets; (3) and identify effective agency practices for achieving data center closures, cost savings, and optimization progress. To do so, GAO assessed the 24 DCOI agencies' data center inventories as of August 2018; reviewed their reported cost savings documentation; evaluated their data center optimization strategic plans; and assessed their progress against OMB's established optimization targets. GAO also solicited practices that selected agencies reported to be effective in meeting DCOI goals. What GAO Found The 24 agencies participating in the Office of Management and Budget's (OMB) Data Center Optimization Initiative (DCOI) reported mixed progress toward achieving OMB's goals for closing data centers and realizing the associated savings by September 2018. As of August 2018, 13 agencies reported that they had met, or had plans to meet, all of their OMB-assigned closure goals by the deadline. However, 11 agencies reported that they did not have plans to meet their goals. Further, 16 agencies reported that, as of August 2018, they had met, or planned to meet, their cost savings targets, for a total of $2.36 billion in cost savings for fiscal years 2016 through 2018. This is about $0.38 billion less than OMB's DCOI savings goal of $2.7 billion. This shortfall is the result of 5 agencies reporting less in planned cost savings and avoidances in their DCOI strategic plans, as compared to their savings targets established for them by OMB. Three agencies did not have a cost savings target and did not report any achieved savings. In addition, the 24 agencies reported limited progress against OMB's five data center optimization targets for server utilization and automated monitoring, energy metering, power usage effectiveness, facility utilization, and virtualization. As of August 2018, the agencies reported that 3 had met three targets, 9 had met one target, and 10 met none of the targets. Two agencies did not have a basis to report on progress as they do not own any data centers. Further, as of August 2018, 20 agencies did not plan to meet all of OMB's fiscal year 2018 optimization goals. Specifically, only 2 agencies reported plans to meet all applicable targets; 6 reported that they did not plan to meet any of the targets (see figure). We selected 6 agencies that had demonstrated success towards meeting their DCOI goals and those agencies reported a number of key practices that contributed to their efforts. The officials noted the importance of, among other things, obtaining executive leadership support for consolidation and optimization activities, employing an organization-wide communications plan, and focusing on data center closures. The officials also cited the use of past experience and lessons learned to inform improvements to future consolidation plans and processes. What GAO Recommends GAO is making 36 recommendations to 22 agencies to improve performance against established DCOI goals. Eleven agencies agreed with the recommendations, three did not fully agree, one disagreed, and seven neither agreed nor disagreed, as discussed in the report.

gao_GAO-20-108

gao_GAO-20-108_0

Background The 340B Program was created in 1992 following the creation of the Medicaid Drug Rebate Program and gives 340B covered entities—certain eligible hospitals, clinics, and other entities—discounts on covered outpatient drugs comparable to those made available to state Medicaid agencies. According to HRSA, which administers and oversees the 340B Program, the program’s purpose is to enable participating hospitals and other providers to stretch scarce federal resources to reach more eligible patients and provide more comprehensive services. In addition to realizing substantial savings through 340B Program price discounts— which HRSA estimates as 25 to 50 percent of the cost of drugs—covered entities can generate revenue through their participation in the 340B Program. For example, they can purchase covered outpatient drugs at the 340B Program price for all eligible patients regardless of the patients’ income or insurance status and generate revenue by receiving reimbursement from patients’ insurance that may exceed the 340B prices paid for the drugs. Covered Entities Entities are generally eligible for the 340B Program—that is, are covered entities—if they receive one of 10 federal grants or are one of six types of hospital. Hospitals must also meet additional requirements, such as being owned or operated by a state or local government, being formally granted governmental powers, or being nongovernmental. The 340B statute requires nongovernmental hospitals to be nonprofit and to have contracts with state or local governments to provide health care services to the 340B-specified low-income population. However, the requirement does not specify criteria for these contracts, such as the amount or type of services to be provided to these low-income individuals. Generally, hospitals must also meet other requirements to participate, such as treating a disproportionate number of low-income Medicare and Medicaid patients. Hospital participation in the 340B Program has more than tripled over the last decade, due, in part, to the enactment of the Patient Protection and Affordable Care Act in 2010, which expanded the types of hospitals that could qualify for the program. According to data from HRSA, in 2009, prior to the law’s enactment, there were more than 800 340B-participating hospitals, compared to more than 2,500 in 2019. The majority of participating hospitals are nongovernmental hospitals. Specifically, 1,690, or 67 percent, of the hospitals participating as of January 1, 2019 were nongovernmental hospitals. (See figure 1.) HRSA Oversight HRSA is responsible for verifying hospitals’ and other covered entities’ eligibility to participate in the 340B Program. HRSA reviews nongovernmental hospitals’ eligibility for the 340B Program at registration, recertification, and through audits. Registration. Prior to participation in the 340B Program, hospitals must register with HRSA, at which point they must self-attest to meeting the program’s eligibility requirements. Additionally, HRSA’s hospital registration instructions specify that, at the time of registration, a nongovernmental hospital must have documentation that shows it is nonprofit (such as copies of Internal Revenue Service documentation) and a copy of its contract with a state or local government to serve the 340B-specified low-income population. This documentation must be provided to HRSA upon request. During each quarterly registration period, HRSA conducts contract integrity checks for a random sample of 20 percent of newly registering nongovernmental hospitals. For the selected hospitals, HRSA requests a copy of the hospital’s contract with the state or local government, which it reviews to verify that the contract is signed by officials from both organizations, is in effect, and does not expire before program participation would begin. HRSA policy states that a hospital that cannot provide a state or local government contract when selected for a contract integrity check at registration will not be registered for the 340B Program. Recertification. To remain in the 340B Program, hospitals must annually recertify their eligibility. During recertification, hospitals are to ensure that their information (e.g. name, address, point of contact) is correct in HRSA’s internal 340B Program database and self-attest that the hospital still meets program requirements. HRSA collects documentation from the hospital if it reports changes to its name, classification (i.e., whether it is government owned or operated, delegated governmental powers, or nongovernmental), or nonprofit status. Audits. HRSA audits 200 covered entities—a combination of hospitals and federal grantees—per year. HRSA’s audits include covered entities (including hospitals) that are selected based on risk-based criteria (approximately 90 percent of the audits conducted each year), and entities that are targeted based on, for example, stakeholder allegations of noncompliance (10 percent of audits conducted). The criteria for risk- based audits include a covered entity’s changes in the volume of 340B Program drug purchases, time in the program, complexity of its program, and history of violations or allegations of noncompliance. Among other things, HRSA’s audits include assessments of each hospital’s 340B eligibility status. For a nongovernmental hospital, HRSA’s guidance indicates that auditors are expected to review the hospital’s contract with the state or local government to ensure that it is for serving the 340B-specified low-income population and is signed by both a hospital and state or local government official. Auditors are also expected to review the contract’s start and end dates to ensure that it is effective during a specific period of review. HRSA defines the audit’s period of review as the time frame beginning the first day of the audit’s sample period—a six-month period that predates and is not contiguous with the beginning of the onsite audit—and ending on the last day of the onsite audit. For example, a hospital with an onsite audit in March of 2017 may have a sample period from July 1, 2016 through December 31, 2016, which means that auditors should verify that the hospital’s contract was in effect from at least July 1, 2016 through the end of the March 2017 onsite audit. If HRSA identifies deficiencies in hospitals’ contracts, the agency may issue (1) findings of noncompliance, which are made public on HRSA’s website, or (2) areas for improvement, which are not made public. When an audit results in a finding of noncompliance, the hospital is required to submit a corrective action plan within 60 days of the audit report being finalized for HRSA’s approval. HRSA closes the audit once the hospital attests that the corrective action plan has been fully implemented, and any necessary repayments have been made to affected manufacturers. For example, if a nongovernmental hospital were unable to demonstrate that it had a contract with a state or local government when audited, HRSA policy states that the hospital would be issued a finding of noncompliance and may be subject to termination from the 340B Program for not meeting eligibility criteria. In addition, the hospital may be responsible for repayment to manufacturers for discounts it received during the period it lacked a contract. Contracts Reviewed Included Few Details on Nongovernmental Hospitals’ Obligations to Serve Low-Income Individuals Most of the contracts we reviewed between nongovernmental hospitals and state or local governments obligated the hospitals to provide health care services to low-income individuals, but they included few details about those obligations. The 340B statute requires participating nongovernmental hospitals to have state or local government contracts to provide health care services to the 340B-specified low-income population, but does not otherwise specify details for the content of these contracts. Of the 240 contracts we reviewed, 224 (93 percent) required the hospital to provide services to low-income individuals. Of these 224 contracts, 169 (75 percent) specifically mentioned providing services to the 340B-specified low-income population (low-income individuals not eligible for Medicaid or Medicare). 55 (25 percent) specified a more general obligation to provide services to individuals who are likely low-income, uninsured, or underinsured, such as enrollees in a county program for the medically indigent, inmates at a local detention center, or individuals receiving treatment through a county mental health program. Less than one-third of the contracts we reviewed defined “low-income” or included detailed requirements for the amount or type of services to be provided. Of the 224 contracts that contained an obligation to provide services to low-income individuals, 14 (6 percent) specified what was considered low income. Of these contracts, the specific income threshold varied, generally ranging from 100 percent to 400 percent of the federal poverty level. 71 (32 percent) specified the amount of services the hospitals were to provide to low-income individuals. The contracts generally defined the amount of services as a range in the cost of care the hospital was expected to provide; the amount varied by contract. For example, one contract specified that the hospital would provide $60,000 to $100,000 of services per year, while another included a range of $62 million to $85 million per year. Contracts that did not specify dollar amounts included, for example, provisions regarding the number of staff available to provide services or requirements to provide services at certain times. One such contract required a hospital to provide at least one full-time-equivalent behavioral health provider for specified sites, while another required a hospital to administer influenza vaccines at two clinics on two Fridays each year during influenza season. 53 (24 percent) identified specific types of services that hospitals were to provide, often specifying multiple categories of services. For example, one contract required a hospital, among other things, to provide inpatient and outpatient services, obstetrics, and cardiovascular surgery. Other contracts only identified a single category of service that the hospital was required to provide. For example, nine of the 53 contracts specified that the hospitals were required to provide emergency services, although hospitals that operate emergency departments are already required, as a condition of participating in Medicare, to screen, and if necessary stabilize patients who seek emergency care, regardless of their ability to pay. Additionally, four of the contracts reviewed required the provision of behavioral health services, two specified the provision of vaccinations, and one was for the evaluation and treatment of tuberculosis. 46 of the 224 contracts (21 percent) specified that state or local governments would pay hospitals for the services provided. In some cases, the contracts specified that the hospitals would be paid to provide care for low-income individuals at rates established under other programs–such as the state’s Medicaid program. Others established rates specifically for services provided to the population covered under the contract. Finally, approximately one-third of the contracts reviewed included provisions that would allow the state and local governments contracting with hospitals to ensure that the contractually required services are being provided. Specifically, 68 of the 224 contracts (30 percent) included provisions for reporting, monitoring or enforcement, as shown below in Figure 2; some contracts included more than one type of provision. Of the 68 contracts, 56 required hospitals to report information to the state or local government. Of these 56, 40 required reporting on the services provided under the contract, including types, dollar amounts, or number of services provided to certain populations, such as “medically indigent,” “uninsured persons,” and “underinsured persons.” For example, one contract required the hospital to provide the government with an annual report containing information about the value of free care provided to indigent persons, the total value of discounted care provided to uninsured patients, and the number of declined requests for free or discounted care. The remaining 16 contracts included more general reporting requirements, such as to provide copies of any reports requested by state or federal licensing, regulatory, or accrediting entities, to the state or local government. 29 contracts included provisions for governments to monitor the hospitals’ provision of care. Specifically, 10 contracts required regular reviews, with some of those at specific time intervals (e.g. annually, quarterly), while 19 contracts required that hospitals be available for periodic audits or to provide the government access, upon request, to medical records and documents which could be used to review or evaluate the services being provided under the contract. 34 contracts included enforcement mechanisms for the government to apply consequences if the hospital did not meet the terms of the agreement. For example, one contract allowed the state government to terminate the contract 90 days after providing notice of the state’s determination that the hospital was not providing sufficient services to low-income individuals. Contracts for eight hospitals provided for monetary fines or withholding of funds if hospitals were found to be in breach of the contract. HRSA’s Processes Do Not Provide Reasonable Assurance That Participating Nongovernmental Hospitals Meet 340B Program Eligibility Criteria HRSA uses self-reported data to determine whether hospitals are nonprofit without assessing whether the data are reliable for that purpose. Additionally, HRSA relies primarily on nongovernmental hospitals’ self- attestations to verify the existence of state or local government contracts, and weaknesses in the reviews of contracts it does conduct hamper the identification of potential eligibility issues. HRSA Uses Self-Reported Data That May Not Be Reliable to Assess Hospitals’ Nonprofit Status HRSA uses Medicare cost report data from CMS to determine whether hospitals are nonprofit, but these data may not be sufficiently reliable for this purpose. Specifically, HRSA relies on self-reported information from cost reports on whether hospitals operate as nonprofit, proprietary, or governmental organizations. HRSA reviews this information at registration to check that hospitals have indicated that they are nonprofit organizations. Additionally, in April 2019, HRSA began conducting quarterly checks of cost report data to identify hospitals that list themselves as proprietary for further review, as this designation could be used by for-profit, rather than nonprofit, hospitals, contrary to 340B Program eligibility requirements. HRSA officials told us that the agency has not independently evaluated the reliability of the cost report data for determining nonprofit status. Additionally, a CMS official responsible for oversight of Medicare cost reports told us that CMS does not have any formal processes to assess the reliability of the data on whether a hospital is nonprofit, proprietary, or governmental, because these data do not affect Medicare reimbursement. The official added that the question on the cost report used to collect these data was not intended to assess nonprofit status, is not clearly defined, and may not be reported accurately. For example, the cost report instructions do not include definitions of nonprofit and proprietary for providers to refer to when they are completing their cost reports. HRSA requires hospitals to maintain additional documentation, such as Internal Revenue Service forms for tax-exempt organizations or documents from the state, to demonstrate their nonprofit status, but does not collect or review this documentation if hospitals indicate that they are nonprofit on their cost reports. In August 2019, HRSA submitted a proposal to the Office of Management and Budget to require hospitals registering for the 340B Program to submit documentation supporting the hospital classification that they select during registration, which would include requiring nongovernmental hospitals to submit documentation of their nonprofit status. However, this requirement, if it goes into effect, would apply only to newly registering hospitals and would not affect the nearly 1,700 nongovernmental hospitals currently participating in the 340B Program. For those hospitals, HRSA would continue to rely on the Medicare cost report data. Relying on the self-reported data from Medicare cost reports is inconsistent with federal internal control standards related to information and communication, which state that management should use quality information to achieve the entity’s objectives, such as by obtaining relevant data, based on identified information requirements, that are reasonably free from error and bias, and that management should evaluate the data for reliability. Without ensuring that the information it uses on hospitals’ nonprofit status is reliable, HRSA cannot effectively determine if nongovernmental hospitals participating, or seeking to participate, in the 340B Program meet the statutory eligibility requirements, creating a risk that for-profit hospitals could receive discounted pricing for which they are not eligible. HRSA Primarily Relies on Hospitals’ Self-Attestations to Verify the Existence of Contracts with State or Local Governments HRSA primarily relies on self-attestations from nongovernmental hospitals to verify that they have contracts in place with state or local governments as required to participate in the 340B Program. Specifically, HRSA relies on the attestations that hospitals are required to make during registration and recertification that they meet the program’s eligibility requirements. Although HRSA requires nongovernmental hospitals to have copies of their contracts, and to provide them upon request, it does not require most hospitals to submit those contracts at either registration or recertification. Additionally, while HRSA previously required each nongovernmental hospital to submit a certification of contract form during registration that was signed by a government official and attested to the existence of a contract to serve the 340B-specified low-income population, officials said the agency stopped requiring submission of this form in July 2014. At that time, officials said HRSA initiated a process of contacting government officials directly through an online certification process to confirm that newly registering hospitals had contracts in place. However, that process was eliminated in September 2017, and HRSA no longer has a process that requires state and local government officials to confirm the existence of contracts with nongovernmental hospitals. HRSA does collect and review contracts with state or local governments for a sample of nongovernmental hospitals through its audit and contract integrity check processes, but these reviews are currently limited in number and scope. Specifically, in fiscal years 2017 and 2018, HRSA audited about 7 percent of nongovernmental hospitals per year (108 and 109 hospitals in fiscal years 2017 and 2018, respectively). Additionally, at the time of our review, HRSA conducted contract integrity checks for 20 percent of newly registering hospitals; this equated to 41 hospitals in calendar years 2017 and 2018 combined. HRSA’s August 2019 proposed information collection request, if approved, would require all newly registering nongovernmental hospitals to submit their state or local government contracts at registration. However, as previously mentioned, this new requirement would only affect newly registering hospitals and not those already participating. Consequently, for the large majority of nongovernmental hospitals already registered for the 340B Program, self-attestations made electronically at registration and recertification would remain HRSA’s sole method of verifying that hospitals have state or local government contracts as required by the 340B statute. Additionally, HRSA officials told us that when the agency does collect documents from nongovernmental hospitals through its audits or contract integrity checks, they do not review them to determine if they are contracts (i.e., mutually binding agreements to provide services or supplies in exchange for something of value). Based on our review of documentation submitted to HRSA from 258 hospitals, 18 hospitals submitted documents that did not appear to meet this common definition of a contract; examples included certification of contract forms without accompanying contracts, articles of incorporation, and descriptions of community programs. Nevertheless, these hospitals were permitted to participate in the 340B Program. HRSA’s reliance on hospitals to attest that the required contracts are in place is contrary to federal internal control standards related to information and communication, which state that management should use quality information to achieve the entity’s objectives, such as by obtaining relevant data from external sources in a timely manner based on the identified information requirements. Without a process to verify that all nongovernmental hospitals have contracts in place, HRSA does not have reasonable assurance that nongovernmental hospitals participating in, or seeking to participate in, the 340B Program have contracts with state and local governments. Consequently, this increases the risk that nongovernmental hospitals that do not have the statutorily required contracts and are thus ineligible may register for, and participate in, the program. Weaknesses in HRSA’s Contract Reviews Hamper the Identification of Potential Eligibility Issues In addition to not determining whether the documentation provided by nongovernmental hospitals during contract integrity checks and audits are contracts, weaknesses in HRSA’s reviews hamper its ability to identify and address issues that affect the hospitals’ eligibility for the 340B Program. Specifically, we identified three weaknesses: (1) contract integrity checks do not assess whether contracts require hospitals to serve the 340B-specified low-income population; (2) guidance for auditors’ review of contracts has not been consistently documented and lacks detail; and (3) HRSA allows hospitals to avoid audit findings by entering into new contracts with state and local governments while audits are being conducted. HRSA’s contract integrity checks for newly registering hospitals do not assess whether the contracts require the provision of services to the 340B-specified low-income population. HRSA’s contract integrity checks for newly registering nongovernmental hospitals are limited to verifying that contracts clearly list the names of the hospital and unit of government and have appropriate signatures and dates; procedures for conducting these checks do not instruct staff to review whether the contracts require hospitals to provide health care services to the 340B-specified low-income population, as required to participate in the 340B Program. Of the 38 contracts submitted to HRSA for contract integrity checks in 2017 and 2018, two (5 percent) did not appear to require the hospitals to serve the 340B-specified low-income population, yet HRSA allowed the hospitals to begin participating in the 340B Program. Specifically, one hospital submitted a contract with a state government that was limited to providing services to beneficiaries of the state’s Medicaid program, although nongovernmental hospitals are to have contracts to provide services to individuals who are not entitled to Medicaid benefits. The other hospital submitted an agreement with a nonprofit company for management services, including accounting and payroll services, for their hospital and nursing home facilities. To participate in the 340B Program, nongovernmental hospitals must have a contract with a state or local government to provide health care services to the 340B-specified low-income population. Thus, allowing hospitals to participate when the state or local government contracts they submitted for review do not require them to serve this population is inconsistent with HRSA’s responsibilities for oversight of the 340B Program, including ensuring that participating hospitals meet the statutory eligibility requirements. Without amending its contract integrity checks to include verifying that newly registering hospitals have contracts that meet statutory eligibility requirements, HRSA risks allowing hospitals that are not eligible, and which may not be providing services to the 340B- specified low-income population, to participate in the 340B Program. Guidance for contract reviews during audits has not been consistently documented over time and lacks detailed instructions. Although HRSA officials told us they have always expected auditors to look for a contract through which a nongovernmental hospital would be eligible for the 340B Program, we found that HRSA’s guidance for auditors has not clearly documented these expectations and lacks detailed instructions. HRSA did not document key elements to look for— signatures, dates, and a requirement to serve the 340B-specified low- income population—in its guidance for auditors until August 2018. Further, the agency has made frequent changes to its guidance and procedures. For example, between November 2017 and July 2019, HRSA modified its guidance for auditors at least six times. In addition, HRSA’s guidance states that auditors are expected to perform a “simple logic test” to determine whether contracts require the hospital to serve the 340B- specified low-income population, but HRSA has not provided any additional information about how auditors are expected to conduct such a test. The guidance also advises auditors not to “dive too deep” when reviewing contracts. Of the 202 contracts submitted by hospitals as part of HRSA’s audits that we reviewed, 11 contracts (5 percent) did not appear to require hospitals to provide care to the 340B-specified low- income population, yet HRSA allowed the hospitals to continue their participation in the program. One such contract was a consent order that stated that the state’s attorney general would defer enforcement action based on the hospital’s agreement to abide by certain medical debt collection practices, such as adopting a zero tolerance policy for abusive, harassing, oppressive, false, deceptive, or misleading language or collections conduct. Furthermore, HRSA’s procedures for audits do not require auditors to separately affirm and record their review of the dates, signatures, and services required in the contracts. Thus, HRSA has no way of knowing whether auditors have checked and verified each of these elements. In addition to the 11 contracts that did not appear to obligate the hospitals to provide health care services to the 340B-specified low-income population, our review of 202 contracts submitted to HRSA by audited hospitals found 16 contracts (8 percent) were missing one or both signatures; 15 contracts (7 percent) were missing effective dates or were expired; at least 8 contracts had dates that did not cover the audit’s period of review, which includes a 6-month sample period before the start of the audit. For at least some of these contracts, HRSA was unaware of the issues we identified; HRSA did not issue audit findings in response to any of these contracts. HRSA has taken steps to address expired contracts. Specifically, in May 2019, HRSA revised its procedures for hospital registration and contract integrity checks to include language specifying that a hospital should not be approved for registration unless a contract is currently in place and that the contract must not expire before the participation start date. In addition, HRSA officials told us that in January 2020 the agency plans to implement a quarterly check of its 340B database to identify hospitals with expired state or local government contracts. However, these efforts do not address other date-related issues such as missing effective dates, or the issues with signatures or contract service requirements. Federal internal control standards related to control activities and enforcing accountability state that agencies should (1) implement control activities through policies, such as by documenting policies in the appropriate level of detail to allow management to effectively monitor the control activity; and (2) evaluate performance and hold individuals accountable for their internal control responsibilities, such as by communicating with the service organizations contracted to perform roles about the agency’s objectives and related risks, assigned responsibilities and authorities, and the expectations of competence to enable the service organization to perform its responsibilities. Without more specific guidance for auditors’ review of contracts, and procedures requiring auditors to separately document their review of each contract element, HRSA lacks reasonable assurance that the audits are appropriately identifying deficiencies in nongovernmental hospitals’ contracts with state or local governments. As a result, some hospitals appear to be participating in the 340B Program based on contracts that are inconsistent with program requirements or HRSA’s guidance. HRSA allows audited hospitals to avoid audit findings by entering into new contracts with state and local governments while audits are being conducted. As previously noted, our review of contracts submitted to HRSA by audited hospitals found that eight hospitals provided contracts that did not appear to cover the audit’s period of review. Three of the eight hospitals entered into the contracts while the audit was ongoing. According to HRSA policy, a hospital that does not demonstrate that it had a contract for the entire audit period should be issued a finding of noncompliance and held responsible for repayment to manufacturers for any discounts received improperly during the period for which it did not have a contract. However, HRSA did not issue such findings or penalties for any of the hospitals we identified with contracts that did not cover the audit’s period of review. For example, in one case, officials said HRSA had included a finding in its draft audit report, but withdrew it when presented with a new contract with an effective date made retroactive to cover the audit’s entire period of review. in another case, a hospital that had been government-owned was sold to a private company in 2013, but did not switch its classification to nongovernmental until 2015, and did not sign a contract with a state or local government until it was audited in fiscal year 2018. The hospital’s contract, signed in 2018, included a retrospective attestation that the hospital had been providing care for the 340B-specified population since 2013. HRSA officials told us that they accept such retroactive documentation in conjunction with current, valid contracts on a case-by-case basis. As such, a hospital may avoid findings, and potential repayments to manufacturers, by asserting that it had been providing care even when a contract was not in place. To participate in the 340B Program, a nongovernmental hospital is required by statute and HRSA policy to have a contract with state or local government to serve the 340B-specified low-income population. Allowing hospitals to submit retroactive contracts after they have already begun participation in the program is inconsistent with HRSA’s responsibilities for oversight of the 340B Program, including ensuring that participating hospitals meet the statutory eligibility requirements. Further, allowing hospitals that are unable to demonstrate that they have contracts in place that cover the audit’s period of review to continue to participate in the 340B Program without consequence undermines the effectiveness of HRSA’s audit process and increases the risk that ineligible hospitals will receive discounts under the program. Conclusions The 340B Program allows hospitals and certain other providers to stretch federal resources to reach more eligible patients and provide more comprehensive services. Participation in the 340B Program also can be beneficial for hospitals and other covered entities as they can realize substantial savings on covered outpatient drugs and generate revenue on those drugs. Hospital participation in the 340B Program, and hospital purchases of discounted drugs through the 340B Program, has risen rapidly over time. However, HRSA’s current processes and procedures do not provide reasonable assurance that nongovernmental hospitals seeking to participate and benefit from the 340B Program meet the program’s eligibility requirements. Given the weaknesses in HRSA’s oversight, some hospitals that do not appear to meet the statutory requirements for program eligibility are participating in the 340B Program and receiving discounted prices for drugs for which they may not be eligible. Although HRSA has initiated some efforts to strengthen its processes for assessing hospitals’ eligibility, continued growth in the number of participating hospitals and 340B- purchased drugs highlights the need for HRSA to improve its oversight processes. This is critical to safeguarding the integrity of the 340B Program. Recommendations for Executive Action We are making the following six recommendations to HRSA: The Administrator of HRSA should ensure that the information it uses to verify nonprofit status for all nongovernmental hospitals that participate in the 340B Program is reliable—for example, by requiring and reviewing the submission of official documentation hospitals must already maintain or by ensuring the reliability of the data the agency uses. (Recommendation 1) The Administrator of HRSA should implement a process to verify that every nongovernmental hospital that participates in the 340B Program has a contract with a state or local government as required by statute. (Recommendation 2) The Administrator of HRSA should amend its contract integrity check procedures for the 340B Program to include a review of whether hospitals’ contracts with state and local governments require the provision of health care services to low-income individuals not eligible for Medicaid or Medicare as required by statute, and should provide guidance for staff to conduct these reviews. (Recommendation 3) The Administrator of HRSA should provide more specific guidance for 340B Program auditors on how to determine if nongovernmental hospitals’ contracts with state and local governments require the provision of health care services to low-income individuals not eligible for Medicaid or Medicare. (Recommendation 4) The Administrator of HRSA should revise its 340B Program audit procedures to require auditors to document their assessments of whether nongovernmental hospitals’ contracts with state and local governments are appropriately signed, cover the time periods under review, and require hospitals to serve low-income individuals not eligible for Medicaid or Medicare, such as by requiring auditors to separately affirm and record their review of each of these elements. (Recommendation 5) The Administrator of HRSA should require nongovernmental hospitals participating in the 340B Program to demonstrate that they have contracts with state or local governments in effect prior to the beginning of their audits’ periods of review and should apply consistent and appropriate consequences for hospitals that are unable to do so. (Recommendation 6) Agency Comments and Our Evaluation HHS provided written comments on a draft of this report, which are reproduced in appendix II, and technical comments, which we have incorporated as appropriate. In its written comments, HHS concurred with five of our six recommendations; it did not concur with one of them. In concurring with five of our recommendations, HHS stated that HRSA is evaluating its audit process and other program integrity efforts, and noted that HRSA has made improvements to strengthen its program integrity efforts that align with some of our recommendations. With respect to our recommendation to require auditors to document their assessments of the required elements of contracts, HHS concurred and noted that HRSA updated its audit procedures. Specifically, HRSA’s draft procedures for fiscal year 2020 audits require auditors to specify if the hospital provided a contract that includes the names and signatures for both the hospital and government agency, effective dates that cover the entire audit period, and that requires the provision of services to the 340B-specified low- income population. We are pleased that HRSA has already taken this step to implement our recommendation. To fully implement this recommendation, HRSA should incorporate these changes into its final audit procedures for fiscal year 2020. HHS also concurred with our recommendation to require nongovernmental hospitals to demonstrate that they have contracts in effect prior to the beginning of the audits’ periods of review, and to apply consistent and appropriate consequences if they do not. Also, as noted above, HRSA has updated its draft audit procedures to specify that auditors should look for effective dates that cover the entire audit period. While this is an important step, HRSA must also show that it has applied consistent and appropriate consequences when auditors find that nongovernmental hospitals did not have contracts in effect prior to the beginning of their audit periods. On a related issue, HHS expressed concern over and disagreed with our finding that HRSA allows hospitals to avoid audit findings by entering into new contracts while audits are being conducted, noting that HRSA assesses potential audit findings on a case-by-case basis to ensure that any necessary steps are taken to address issues. However, as we reported, HRSA officials have indicated that they accept retroactive contract documentation on a case-by-case basis; we continue to believe that this practice—accepting new contracts that are retroactive— effectively allows hospitals to avoid audit findings. In addition, while we agree that working with hospitals to address noncompliance is appropriate, we continue to believe that such efforts should be in addition to, not instead of, documenting noncompliance by issuing findings and applying appropriate consequences, in accordance with HRSA’s audit policies and procedures. To do otherwise undermines the integrity of HRSA’s audits, and increases the risk that ineligible hospitals will receive discounts under the program. HHS also concurred with our recommendation to ensure that the information HRSA uses to verify nonprofit status is reliable, but stated that HRSA believes that the information it uses from hospitals’ Medicare cost reports is reliable, because hospital administrators attest to the accuracy of their cost reports. However, as discussed in our report, neither HRSA nor CMS has evaluated the reliability of the cost report data for verifying nonprofit status, and a CMS official responsible for oversight of the cost reports told us that the question on the cost report is not clearly defined and may not be reported accurately. As such, we continue to believe that HRSA needs to assess the reliability of the Medicare cost report data should it continue to use those data for determining hospitals’ nonprofit status. Alternatively, HRSA could require hospitals to submit documentation of their nonprofit status, such as Internal Revenue Service documents, which HRSA acknowledged hospitals are required to maintain as part of their auditable records. HHS did not concur with our recommendation to implement a process to verify that every nongovernmental hospital that participates in the 340B Program has the statutorily required contract with a state or local government. HHS noted that it has requested authority to require hospitals registering for the 340B Program to submit documentation supporting the hospital classification that they select during registration. According to HHS, if approved, HRSA would begin collecting and reviewing contracts from all newly registering nongovernmental hospitals. However, HHS stated that HRSA does not have the resources to collect, review, and verify that every participating nongovernmental hospital has a contract with a state or local government. While we understand that verifying the existence of contracts for all participating nongovernmental hospitals would require additional effort on HRSA’s part, our review found that relying on hospitals’ attestations is not sufficient to ensure hospitals’ eligibility. Additionally, implementing a process to verify the existence of a contract does not necessarily require that HRSA collect and review contracts from every hospital. There are other potential options, such as obtaining confirmation from the state or local government that they indeed have a contract with the hospital to provide services to the 340B-specified low-income population. HHS also commented that implementing our recommendation would create a significant burden on covered entities. However, as we noted in our report, HRSA already requires hospitals to maintain copies of their state or local government contracts. Therefore, it is unclear how implementing a process to verify the existence of those contracts would represent a significant burden for nongovernmental hospitals already registered for the program. Ensuring the eligibility of covered entities that participate in the 340B Program is essential for program integrity. As such, we continue to believe that HRSA needs to take action, beyond relying on hospitals’ self-attestations, to verify that all participating nongovernmental hospitals have contracts with state or local governments that meet the statutory requirements of the program. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Secretary of HHS, the Administrator of HRSA, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or draperd@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Other major contributors to this report are listed in appendix III. Appendix I: Characteristics of the Contracts Reviewed Table 1 provides information about the 240 contracts between hospitals and state or local governments that were included in our review, including information about the type of hospital and the level of government that were parties to the contract. In at least two cases, the hospitals contracted with other health care providers who were themselves 340B Program participants, such as a community health center operated by a local health department. Officials signing on behalf of state and local governments included individuals with executive positions, such as the heads of state agencies, mayors, and county executives, but also included a city alderman, a vice-chancellor for finance at a state university health system, and a juvenile court judge. Appendix II: Comments from the Department of Health and Human Services Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Michelle Rosenberg, Assistant Director; Hannah Locke, Analyst-in-Charge; Jennie Apter; George Bogart; Kaitlin Farquharson; Matthew Green; Vikki Porter; Daniel Ries; Brienne Tierney; and William T. Woods made key contributions to this report.

Why GAO Did This Study The number of nongovernmental hospitals participating in the 340B Program has grown over time. HRSA requires participating hospitals to register and annually recertify their eligibility. HRSA also reviews hospitals' eligibility through audits of a small sample each year. GAO was asked to provide information on 340B-participating hospitals' contracts with state and local governments. This report (1) describes any obligations in selected nongovernmental hospitals' contracts to serve low-income individuals, and (2) examines HRSA's processes to assess nongovernmental hospitals' eligibility. GAO examined contract documentation from all 258 nongovernmental hospitals HRSA reviewed in 2017 and 2018; and HRSA's policies, procedures, and guidance related to 340B hospital eligibility. GAO also interviewed HRSA officials. What GAO Found Under the 340B Drug Pricing Program (340B Program), administered by the U.S. Department of Health and Human Services' (HHS) Health Resources and Services Administration (HRSA), drug manufacturers provide discounted prices on outpatient drugs to certain hospitals and other entities. About two-thirds of hospitals participating in the 340B Program (approximately 1,700) are nongovernmental hospitals (private, nonprofit hospitals), which qualify for the program, in part, based on having contracts with state or local governments to provide health care services to the 340B-specified low-income population—low-income individuals not eligible for Medicaid or Medicare. GAO's review of contract documentation for 258 nongovernmental hospitals found that most contracts obligated these hospitals to provide health care services to low-income individuals. However, few of the contracts reviewed included details about those obligations, such as the amount or type of care hospitals were required to provide. The statute does not require the contracts to contain such details. GAO found that HRSA's processes do not provide reasonable assurance that participating nongovernmental hospitals meet eligibility requirements. For example, HRSA primarily relies on hospitals' self-attestations to verify the existence of contracts with state and local governments. The agency reviewed contract documentation for less than 10 percent of nongovernmental hospitals per year in 2017 and 2018. GAO also identified several weaknesses in HRSA's review of the nongovernmental hospital contracts: HRSA does not conduct reviews to determine whether the documents submitted by nongovernmental hospitals are actual contracts, namely that they are mutually binding agreements to provide services or supplies in exchange for something of value. GAO found that 18 of the 258 hospitals reviewed submitted documents that did not appear to be contracts, such as descriptions of community programs, yet all of these hospitals were permitted to participate in the program. When audits have identified hospitals that did not have contracts in place throughout the audits' periods of review, HRSA has allowed hospitals to avoid audit findings by, for example, entering into new contracts with retroactive start dates. This practice undermines the integrity of HRSA's audits. HRSA's contract reviews do not always include assessments of whether contracts are consistent with the statutory requirement to provide health care services to the 340B-specified low-income population and HRSA's guidance for conducting such assessments, when required, lacks detailed instructions. As a result, GAO found that contracts for 13 hospitals reviewed did not appear to require hospitals to serve the 340B-specified low-income population. Despite this, these 13 hospitals were permitted to participate in the program. Given these weaknesses, some nongovernmental hospitals that do not appear to meet the statutory requirements for program eligibility are participating in the 340B Program and receiving discounted prices for drugs for which they may not be eligible. What GAO Recommends GAO is making six recommendations, including that HRSA implement a process to verify that all nongovernmental hospitals have contracts in place, including throughout hospitals' audit periods; amend its contract reviews to include an assessment of whether contracts meet statutory requirements; and provide better guidance on contract reviews. HHS concurred with all of the recommendations except the one to implement a process to verify that all nongovernmental hospitals have contracts. GAO continues to believe this action is needed to ensure that only eligible hospitals are allowed to participate in the 340B Program.

gao_GAO-20-348

gao_GAO-20-348_0

Background Eligibility for TRIA The purposes of TRIA are to (1) protect consumers by addressing market disruptions and ensuring the continued widespread availability and affordability of commercial property/casualty insurance for terrorism risk; and (2) allow for a transitional period for the private markets to stabilize, resume pricing of such insurance, and build capacity to absorb any future losses, while preserving state insurance regulation and consumer protections. TRIA only applies to certain commercial property/casualty lines of insurance (we refer to them as TRIA-eligible lines) and excludes lines such as health and life insurance. While the law requires insurers to make terrorism risk coverage available to commercial policyholders, commercial policyholders are not required to buy it. Additionally, the law requires an insurer to make coverage for terrorism losses available that does not differ materially from the terms, amounts, and other coverage limitations applicable to losses arising from events other than acts of terrorism. For example, an insurer offering $100 million in commercial property coverage must offer $100 million in coverage that is not materially different for property damage from a certified terrorist event. Insurers may charge a separate premium to cover their terrorism risk, although some include the coverage in their base rates for all-risk policies. The majority of terrorism risk insurance is purchased as part of these embedded policies. The remainder is purchased as stand-alone coverage. Neither insurers nor the federal government charge for the federal coverage of terrorism risk under TRIA, but the government must recoup at least some of its losses following a terrorist event. For eligible commercial lines of property/casualty insurance, TRIA covers insured losses resulting from an act of terrorism, which is defined, in part, as a “violent act or an act that is dangerous” to human life, property, or infrastructure. TRIA is silent about losses from attacks with nuclear, biological, chemical, or radiological (NBCR) weapons. Although TRIA and its reauthorizations do not specifically include cyber risk insurance as a TRIA-eligible line, Treasury issued guidance about such coverage in 2016. The guidance stated that TRIA provisions apply to cyber risk insurance written under an embedded or stand-alone policy in TRIA- eligible lines. A cyberterrorism event could cause minor-to-severe business disruption and physical damage to property. In this report, we include losses resulting from cyberterrorism events with conventional terrorism events. Public-Private Loss Sharing Any catastrophic terrorist event presents both explicit and implicit fiscal exposure for the federal government. TRIA alleviates some of the implicit exposure through loss sharing. Under TRIA, the federal government is legally required to make payments (reimbursements to insurers): this represents an explicit fiscal exposure. Even without a loss-sharing program, the federal government also faces implicit fiscal exposure through a potential expectation to provide policyholders or insurers assistance to address long-term effects after a terrorist event. We have defined implicit fiscal exposures as situations that create expectations for future federal spending based on current policy, past practices, or other factors. Under TRIA, the government and insurers share losses in the event of a certified act of terrorism with insured losses above the program trigger of $200 million and below the program cap of $100 billion. According to the statute, Treasury cannot certify an event as an act of terrorism under TRIA if the aggregate property/casualty “insurance losses” resulting from the event are less than $5 million. Additionally, TRIA is not triggered unless the aggregate property/casualty “insured losses” resulting from one or more certified acts in a particular calendar year reach $200 million. Annual coverage for losses is limited (capped) so that neither the private insurers nor the federal government are responsible for paying aggregate insured losses in excess of $100 billion. Specifically: “Insured losses” are defined in statute and regulation as any losses resulting from an act of terrorism (including an act of war in the case of worker’s compensation) generally occurring in the United States that are covered by primary or excess property/casualty insurance issued by an insurer. TRIA refers to insured losses in defining the program trigger and program cap. “Insurance losses” are not defined in statute or regulation, but TRIA refers to insurance losses in defining the event certification amount. TRIA’s loss-sharing structure requires that insurers pay claims on covered terrorism losses and that Treasury reimburse individual insurers for losses that exceed a specified amount. According to the coshare provision, Treasury reimburses the insurer for a certain percentage (80 percent) of its losses above the insurer deductible, and the insurer is responsible for the remaining portion (20 percent). The policyholder also may retain losses from a terrorist event in the form of an insurance deductible or self-insurance retention. According to Treasury, losses retained by the policyholder are not considered to be “insured losses” under TRIA and do not count toward losses included in the $200 million program trigger or $100 billion program cap. However, this retention may be counted toward the $5 million event certification threshold because it is calculated based on “insurance losses.” Recoupment TRIA provides for two types of recoupment—mandatory and discretionary—of the federal share of losses after a terrorism event. Figure 1 shows the claim and recoupment processes after a terrorism event resulting in losses covered by insurers and the government. Mandatory recoupment. TRIA requires recoupment of at least a portion of the federal share of losses if the aggregate sum of all insurers’ deductibles and coshares are below an amount prescribed by TRIA—known as the industry aggregate retention amount. Under mandatory recoupment, the insurers must impose and remit to Treasury a premium surcharge on all policies in TRIA-eligible lines until total industry payments reach 140 percent of any mandatory recoupment amount. TRIA specifies the collection time frame (from 1 year and 9 months to about 6.5 years, based on the date of the event). Treasury considers the collection time frame when establishing the amount of the mandatory recoupment surcharge. Discretionary recoupment. TRIA permits discretionary recoupment when the federal share of losses exceeds the mandatory recoupment amount. Under the discretionary recoupment provision, Treasury may consider a number of factors, such as economic conditions in the commercial marketplace, in determining the amount to recoup. To help in its decision-making, Treasury may issue a data call for insurer deductible and insured loss information. Treasury also sets the surcharge for discretionary recoupment, but the increase to TRIA- eligible premiums must not exceed 3 percent per calendar year. TRIA does not specify a collection time frame for discretionary recoupment. Treasury’s Data Calls Since the 2015 reauthorization, insurers have been required to submit information to Treasury about their coverage of terrorism risk, including the lines of insurance with exposure to such risk, the premiums earned on such coverage, and the participation rate for such coverage. Treasury’s 2017–2019 data calls included loss scenarios in which insurers estimate and report expected losses given a defined terrorist attack at a specified location, date, and time. Treasury’s defined loss scenarios were located in New York City in 2016, Chicago in 2017, and San Francisco in 2018. Treasury’s reporting requirements for insurers vary, based on the following categories: Small insurers. Insurance companies that had both policyholder surplus and prior year TRIA-eligible direct-earned premium of less than five times the program trigger. For example, TRIA’s program trigger in 2020 is $200 million. Thus, Treasury would categorize insurers with less than $1 billion in 2019 direct-earned premiums and less than $1 billion in policyholder surplus as small. Treasury does not require small insurers to report on a number of items, including the loss scenarios. Nonsmall insurers. Insurance companies with either policyholder surplus or prior year TRIA-eligible direct-earned premium greater than the small insurer thresholds. Captive insurers. Special-purpose insurance companies set up by commercial businesses to self-insure risks arising from the owners’ business activities. Alien surplus lines insurers. Insurance companies headquartered in a foreign country that have been qualified to do business in the United States through an NAIC-administered process, which assesses the financial stability and trustworthiness of the insurer. TRIA Reauthorizations Reduced Explicit Fiscal Exposure, but Treasury and Some Insurers Have Different Interpretations of Key Definitions Explicit Fiscal Exposure Decreased under TRIA Reauthorizations Each reauthorization of TRIA through 2015 reduced the magnitude of the government’s explicit fiscal exposure. Since 2003, changes to TRIA provisions have increased insurers’ share of losses and thus decreased explicit federal fiscal exposure in the event of certified acts of terrorism (see fig. 2). For example, the program trigger rose over time, from $5 million in 2003 to $200 million in 2020. These changes reduced explicit fiscal exposure because they increased the amount of insured losses required before the government would share in the losses. The insurer deductible increased from 7 percent in 2003 to 20 percent for 2020, also reducing the federal share of payments. The 2015 reauthorization required incremental reductions in the federal share of losses over 5 years. The 2019 reauthorization extended the program until 2027, but did not make any changes to the program parameters. See appendix II for more details of the changes in the reauthorizations. Currently, following a certified act of terrorism Treasury pays insurer claims for 80 percent of insurers’ losses above their individual deductibles once losses in a calendar year exceed the program trigger of $200 million (see fig. 3). For example, based on the scenario from Treasury’s 2019 data call, the federal government could have an explicit exposure of about $4.4 billion in reimbursements to insurers. Specifically, insurers estimated that a hypothetical 2018 terrorist event in San Francisco could generate $39.7 billion dollars in overall losses, of which insurers could pay about $17 billion in claims to policyholders and the government could pay about $4.4 billion in reimbursements to insurers after a policyholder retention amount. According to our analysis of Treasury data on insurer direct-earned premiums, federal losses following a terrorist event under the loss-sharing provision in effect in 2020 would be smaller than they would have been for a similar event under the loss-sharing provision in effect in 2015, across all event sizes and subsets of insurers. In addition, more of the federal losses would be recovered through mandatory recoupment (see fig. 4). As we found in 2017, the amount of federal losses depends on event size and how many and which insurers were affected. Additionally, the government share depends on the aggregate TRIA- eligible direct-earned premium of the insurers with losses. Specifically, the federal share of losses is smaller when losses are shared among insurers with larger aggregate premium bases. Insurers Adjusted to Program Changes but Some Were Unclear about How Treasury Would Calculate the Program Certification Threshold, Trigger, and Cap As the share of losses for which insurers are responsible has increased under TRIA, the ability of insurers to absorb the extra exposure also has increased. Insurers use risk-mitigation strategies to reduce or offset their exposures. These can include purchasing reinsurance—insurance for insurers—to cover their deductibles or coshare payments, or diversifying their portfolios (for instance, reducing concentrations of risk in certain locations or lines of insurance). Insurers told us that they considered the potential effect of program changes in each reauthorization and modified risk-mitigation strategies, as needed. Furthermore, other industry stakeholders, including a broker and an industry association, told us that because program changes have been gradual and expected, insurance companies have been able to adjust their coverage accordingly. Available evidence indicates that TRIA has been largely effective in meeting its statutory objectives of stabilizing the terrorism risk insurance market. First, terrorism risk insurance is available in the market for a relatively low cost and is purchased by the majority of commercial policyholders in the United States, according to industry reports. Second, private reinsurance capacity for terrorism risk insurance increased since the creation of the program, according to Treasury. Third, our analysis of Treasury data suggests there is market stability. Insurers in all of Treasury’s reporting categories largely remained in the market. Furthermore, the market share and number of insurers in the reporting categories generally remained stable. For example, using data on direct- earned premiums of insurers from 2016 to 2018, nonsmall insurers (92 insurers in 2018) held about 80 percent of the TRIA-eligible insurance market. Small insurers (186 insurers in 2018) held about 10 percent. Captive and alien surplus lines insurers (598 and 98 insurers, respectively, in 2018) each held 4 or 5 percent (see fig. 5). Our interviews indicate that some insurers’ interpretation of whether policyholder retention amounts count toward the program threshold, trigger, and cap may differ from Treasury’s. Some large policyholders may retain large amounts of loss in the form of a deductible or self- insurance retention following a terrorist event. Treasury officials said policyholder retention amounts are not counted toward the program’s $200 million trigger or its $100 billion cap, but could be counted toward the $5 million threshold for event certification. They stated that the law utilizes “insured losses” when referring to the program trigger and cap, and “insurance losses” when referring to the certification threshold. If Treasury counted policyholder retention amounts toward the program trigger, the program would be triggered and capped with a smaller amount of overall losses. For example: To illustrate, we use a hypothetical terrorist event resulting in $290 million in overall losses, of which $100 million would be retained by policyholders. Using Treasury’s interpretation that excludes policyholder retention amounts, “insured losses” would be $190 million, which is below the program trigger of $200 million. As a result, the government would not be required to pay insurers a coshare. In contrast, if Treasury included policyholder retention amounts, “insured losses” would be $290 million, which exceeds the program trigger. In this case, the government would pay $112 million to insurers in coshares. In either case, losses would not reach the program cap of $100 billion. We asked 12 industry stakeholders about their understanding of how Treasury would use policyholder retention amounts to calculate the program threshold, trigger, and cap. The distinction between “insured losses” and “insurance losses” in Treasury’s explanation was clear to one insurer. However, some aspect of this distinction was unclear to six industry stakeholders, including an insurer association and three insurers, potentially resulting in uncertainty about how Treasury would calculate losses in the aftermath of a terrorist event. For example, representatives of an insurer association and two insurers told us they interpreted insured losses as including the policyholder retention amounts because insurers could be responsible for paying this amount. This is because some insurers pay the entire claim, including all or a portion of the policyholder retention, up front and then seek reimbursement from the policyholder. In addition, if a policyholder cannot pay its retention, the insurance company is responsible for it. Differences in interpretation could lead to disputes between insurers and Treasury following a terrorist event. We previously found that insurers are concerned about the long-term consequences of disputes related to terrorist events. One purpose of TRIA is to stabilize the market following a terrorist event. Furthermore, federal internal control standards state that management should externally communicate the necessary quality information so that external parties can help the entity achieve its objectives and address related risks. Treasury’s program regulations define “insured losses” and do not define “insurance losses.” Furthermore, the regulations do not explain how such losses are calculated and therefore how the policyholder retention amount does or does not count toward the program threshold, trigger, or cap, as applicable. Treasury uses different methods to communicate program information and clarify program details to stakeholders, such as program regulations and interpretive letters, but has not clarified this issue using these or other methods because officials believe the distinction is understood in the industry. By closing the information gap of how it would calculate losses for the program threshold, trigger, and cap, Treasury would create a common understanding of a critical feature of the program. Furthermore, Treasury may prevent uncertainty in the insurance market and potential litigation following a terrorist event that could delay insurance payments and economic recovery. Implicit Fiscal Exposure Exists and in Some Situations Could Become Explicit TRIA explicitly limits federal exposure following a terrorist event, but the federal government could be expected to provide assistance beyond what is explicitly outlined in TRIA. Expectations for the government to provide assistance through its recoupment decisions and to policyholders and insurers, as described below, represent implicit fiscal exposures. Although the government may not act on these expectations, to the extent that it does, the implicit exposure would become an explicit exposure. Less-Than-Full Recoupment of All Funds Following a Terrorist Event In certain circumstances under mandatory and discretionary recoupment, such as potential effects on market stability, public expectation may lead the federal government to cancel recoupment or reduce the amount of funds recouped. Any portion of the federal coshare not recouped represents an implicit fiscal exposure. Mandatory Recoupment Some mandatory recoupment scenarios may or may not be perceived as burdensome to policyholders, prompting an expectation of federal assistance to ease the burden. Treasury determines mandatory recoupment surcharges based on the statutory deadlines for collecting mandatory recoupment. If a large terrorism act occurs in a year in which the statute requires the collection of mandatory recoupment in a short time frame, Treasury may need to set a high surcharge percentage on premiums for policies with TRIA-eligible lines. In this case, some policyholders may find it difficult to pay the surcharge, making collection of the mandatory recoupment amount burdensome. Large recoupment amounts or surcharges could prompt public expectation, and political will, for reducing or cancelling recoupment to alleviate this burden. Because mandatory recoupment time frames are based in statute, reducing or canceling this recoupment would require congressional action. One insurer told us that they are skeptical that Congress would allow Treasury to collect mandatory recoupment after a large event. Under current program provisions, the maximum mandatory recoupment amount will increase if the TRIA-eligible direct-earned premium increases. Industry stakeholders told us that, with this change, eventually all recoupment could be mandatory. While the amount to be recouped may increase, the recoupment time frame remains unchanged. Therefore, over time, surcharge amounts could increase, which may increase burden on policyholders and increase the expectation for Congress to cancel recoupment. If Congress were to cancel the collection of mandatory recoupment, the explicit fiscal exposure would include both the federal share of losses paid to insurers and the decreased corporate tax receipts from deductions for the recoupment charges policyholders may claim (such deductions otherwise were intended to be offset by the 140 percent recoupment). Changes in legislation following premium rate increases in the National Flood Insurance Program (NFIP) provide an example of Congress changing a law to ease policyholder burden. Congress enacted the Biggert-Waters Flood Insurance Reform Act of 2012, which was intended to strengthen the future financial solvency and administrative efficiency of NFIP by implementing provisions to reduce and eventually eliminate most subsidized premium rates. However, after public outcry claiming negative effects on home values, Congress enacted the Homeowner Flood Insurance Affordability Act of 2014, which repealed some of the premium rate increases in the 2012 act. Discretionary Recoupment Discretionary recoupment presents an implicit fiscal exposure because Treasury may decide not to collect the full discretionary recoupment amount. Under TRIA, Treasury decides whether and how much of the discretionary portion of the federal share of losses to recoup. Treasury may recoup some or all nonmandatory funds, or cancel discretionary recoupment. For any amount that Treasury chose not to collect, the fiscal exposure would be the dollar-for-dollar amount of the federal share of losses paid to insurers. As defined in statute, Treasury may consider several factors when determining whether to collect discretionary recoupment, in full or partially, or cancel recoupment. These factors include ultimate cost to taxpayers of no additional recoupment; the economic conditions of the commercial marketplace; the affordability of commercial insurance for small and medium-sized businesses; and other factors Treasury deems appropriate. According to agency officials, decisions regarding discretionary recoupment would be based on the parameters of the specific terrorism act, such as the size of the federal share of losses, location of the event, and length of the collection period. In our analysis of explicit exposure, we found that under some scenarios, the discretionary recoupment amount resulting from a terrorist event could exceed $50 billion. Depending on Treasury’s analysis of these factors, some or all of the discretionary recoupment amount may not be collected. For example, currently, much of the recoupment amount resulting from the most catastrophic losses would be considered discretionary under TRIA’s provisions. Because TRIA mandates an annual 3 percent cap on the increase of premium rates in TRIA-eligible lines for discretionary recoupment, in extreme cases Treasury might need to collect a premium surcharge for a protracted period of time to fully recoup the discretionary portion of losses. The effects of a protracted period of premium surcharges could be a factor in Treasury’s determination to cancel discretionary recoupment. Assistance for Uninsured or Underinsured Losses Following a Terrorist Event Based on previous federal action following natural disasters or financial market crises, there may be an expectation that the government would provide financial assistance to businesses for uninsured or underinsured losses related to a terrorist event, regardless of whether a loss-sharing program existed. For example, the federal government uses the Disaster Relief Fund to provide compensation for property damage or financial losses to victims of Presidentially declared major disasters and emergencies. In fiscal years 2005–2018, the federal government designated $138 billion in supplemental appropriations to this fund for extreme weather events. And following the financial crisis of 2007– 2009, the federal government provided financial assistance directly to General Motors Company and Chrysler Holdings to help stabilize the U.S. automobile industry and to avoid economic disruptions. Treasury officials and industry stakeholders described several terrorist event scenarios that could produce a large amount of uninsured or underinsured losses that affected businesses might not be able to absorb and that might lead to the expectation of federal assistance. Losses Resulting from an NBCR Event NBCR events present an implicit exposure. Historically, insurance coverage for losses related to a NBCR terrorist event has been limited or unavailable. Stakeholders told us that there likely would be an expectation of federal financial assistance for businesses with uninsured losses related to such an event. Treasury officials and stakeholders we interviewed agreed that primary and reinsurance coverage for NBCR events is limited, resulting in many businesses having limited or no coverage. Stakeholders also told us that, without TRIA, insurers would no longer offer the limited amount of NBCR coverage currently available. Stakeholders attribute the limitations to the potentially catastrophic losses associated with NBCR events and the difficulty in modeling and underwriting such events. Representatives of a policyholder association whose members purchase NBCR coverage stated that available coverage likely was insufficient to cover expected losses. Furthermore, they said some policyholders forgo NBCR coverage because of its limited availability, high cost, and the low perceived risk of a NBCR event. As a result, many businesses may be exposed to high loss. Treasury officials and industry stakeholders described possible NBCR terrorism events in which a significant amount of losses could be uninsured or underinsured. Treasury’s 2019 Small Insurers Report found that a NBCR terrorism event likely posed the greatest risk of total catastrophic terrorism losses, far outpacing a conventional attack. Although modeling these types of losses is difficult, NAIC’s Center for Insurance Policy and Research estimated, taking into account the program cap, that a NBCR event in New York City could generate nearly $60 billion of uninsured loss, 38 percent of the total loss. It also found that a larger NBCR event could create $850 billion in uninsured loss, or 90 percent of total losses. Furthermore, this research estimated large uninsured losses in other cities, such as Houston, where a large nuclear event was estimated to generate $67 billion in uninsured losses, or 40 percent of total losses. Such catastrophic losses could create a strong public expectation of federal financial assistance for uninsured losses. Losses above the Program Cap The expectation of financial assistance to policyholders if insured losses exceeded the program cap also creates implicit fiscal exposure. By law, insurers that have met their individual deductible and the federal government are not responsible for losses exceeding the TRIA program cap. However, industry stakeholders we interviewed expected that, in the event the program cap were exceeded, the federal government would provide some form of assistance to those who experienced loss. For example, losses from a NBCR event could be over $1 trillion, with TRIA- insured losses exceeding the $100 billion program cap. While a single conventional attack would be unlikely to exceed the program cap, according to stakeholders, a series of conventional attacks could. Although determining the frequency of terrorist events is difficult, one modeling firm with which we spoke estimates losses great enough to exceed the program cap in a conventional attack to have a 0.0005 percent (or 1/2,000) chance of occurring in a single year. This is less likely than severe natural catastrophes, such as Hurricanes Sandy and Harvey. As we found in a 2019 report on fiscal exposures, Congress demonstrated its willingness to fund the implicit exposure of policyholder claims that exceeded the amount NFIP was authorized to borrow from Treasury. In October 2017, when NFIP was about to exhaust its borrowing authority, Congress passed a supplemental appropriation, which the President signed into law, that cancelled $16 billion of NFIP debt to Treasury. Losses Retained by the Policyholder Implicit fiscal exposure also exists in the expectation that the federal government would assist policyholders unable to pay their retained losses. Policyholders with very large retained losses may face financial insolvency after a terrorist attack, which may create an expectation of government assistance. Stakeholders told us that policyholders may choose larger retention amounts to reduce premiums or because insurance for high-risk, high-value properties is unavailable. Insurers we interviewed said that most businesses have small deductibles, but some large businesses may choose higher retention amounts to reduce the high insurance cost in locations considered to be higher-risk and for high- profile properties. For example as shown in figure 6, Treasury’s 2017– 2019 data call scenarios explored estimated losses in locations with high- profile properties such as Rockefeller Center in New York City (2017), Willis Tower and O’Hare International Airport in Chicago (2018), and Embarcadero Center and Union Square in San Francisco (2019). Although the actual amounts may be lower than the estimates insurers reported, the aggregated policyholder retention could exceed losses paid through the program. This demonstrates the potential for large losses that could create an expectation of government assistance if policyholders with large retention amounts were unable to absorb the losses. Assistance to Insurers with Losses That Do Not Result in Program Loss Sharing An implicit fiscal exposure exists from the potential expectation that the government might help stabilize markets by assisting insurers with (1) losses that do not trigger the program’s loss sharing, or (2) losses from lines ineligible for TRIA. While these risks exist, the current market has some protections in place and stakeholders viewed this exposure as unlikely. Loss sharing not triggered. If the total losses from a certified act of terrorism were below the program trigger (currently $200 million), insurers with deductibles below the program trigger could sustain losses larger than their deductible without receiving any federal coshare. Because the amount of the program trigger has increased over time, more insurers potentially face this scenario. Stakeholders told us that small insurers and those that offer workers’ compensation insurance are most affected by changes to the program trigger. Our analysis of Treasury data shows about 97 percent of insurers have deductibles lower than the $200 million program trigger and thus could receive no coshare following a certified act of terrorism. This includes all small, captive, and alien surplus lines insurers, and more than half of the nonsmall insurers. Furthermore, our analysis of 2016–2018 Treasury data shows that these insurers are sometimes concentrated in certain insurance lines. For example, small insurers may be concentrated in commercial multiple peril lines. Additionally, market shares of small insurers and captive insurers increased in the aircraft (all perils) line. Such concentrations could destabilize specific insurance lines following a terrorist event. Although the market is currently stable, some insurers may leave the market to mitigate their risks if their losses are unlikely to trigger the program’s loss sharing, which could reduce the availability of insurance in certain markets. A reduction in the availability of insurance could lead to more uninsured losses in the event of a subsequent terrorist event and could result in an increased expectation for losses to be covered through federal assistance. In its 2019 report, Treasury recognized the potential for small insurers to not provide insurance in certain markets. Additionally, the report cautions that if the program trigger increased, the number of insurers that would face the possibility of a gap between their deductible and the program trigger also would increase. However, analysis of Treasury data indicate that, to date, insurers in this situation largely have not left the market. As previously noted, because the changes to program parameters were gradual, insurers have had time to adjust to the changes. Insurers facing this scenario also can mitigate their risk by purchasing reinsurance to cover the difference between their deductible and the program trigger. However, it is not clear that the reinsurance market can absorb all of this risk. For example, industry stakeholders told us that reinsurers are sensitive to accumulation of exposure and reinsurance is limited in perceived high-risk areas such as New York City, Washington D.C., Los Angeles, and Philadelphia where there are large concentrations of people and high-value properties. In addition, reinsurance premiums may be too costly for some insurers. According to state insurance commissioners, high reinsurance costs may not qualify as a reason for insurers to increase premiums in some states. As a result, insurers would be unable to pass reinsurance costs on to policyholders. Furthermore, limited availability and affordability of reinsurance for high-risk areas and NBCR could be exacerbated following a terrorist event. Alternative forms of reinsurance, such as catastrophe bonds, are not widely used for terrorism risk. Treasury officials said that widespread market instability and an expectation for federal assistance may be unlikely in the case of a certified act of terrorism that produces significant losses for insurers that do not reach the program trigger. They said any market effects would be localized because a smaller terrorist event likely would affect a small number of insurers with a gap between their deductibles and the program trigger. Industry stakeholders said that it is possible that such an event could cause insurers to reduce coverage offered, but other stakeholders said that small insurers were well aware of the risk and their reactions would not create market instability. Additionally, state guarantee funds may provide support to policyholders in case of insurer insolvency before expectations for federal financial assistance arose. Insurance lines not eligible for TRIA. Federal assistance may be expected for insurers if large terrorism losses occur in insurance lines— such as life or health insurance—that are not eligible for TRIA, according to industry stakeholders. A 2019 Insurance Information Institute report found life insurance losses resulting from the attacks of September 11, 2001, were $1.4 billion (about 3 percent of total insurance losses from the attacks), an amount that far exceeds the TRIA program trigger. The report notes that standard homeowner, condo or co-op, standard renters, automobile insurance, and travel insurance policies also could be affected by terrorism and are not covered under TRIA. Adding group life insurance coverage under TRIA has been proposed, but never passed into law. According to some perspectives in congressional debate, other insurance lines were not viewed as needing explicit federal assistance. Conclusions Insurers have adjusted to changing TRIA program parameters that increased the share of the losses for which they would be responsible. However, some insurers may not clearly understand whether Treasury would include policyholder retention amounts in calculating losses to certify a terrorist event, trigger loss sharing, or determine when the program cap has been reached. External communication to develop a clear understanding of how Treasury calculates “insured losses” and “insurance losses,” specifically as they relate to policyholder retention amounts, would help insurers understand when the program would be activated or capped ahead of any terrorist event. Furthermore, communicating this explanation could help Treasury alleviate uncertainty in the insurance market following a terrorist event. Recommendation for Executive Action The Director of the Federal Insurance Office should communicate to insurers in writing how it would utilize policyholder retention amounts in calculating “insurance losses” and “insured losses” in determining the program certification threshold, trigger, and cap, as applicable. (Recommendation 1) Agency Comments We provided a draft of this report to Treasury and NAIC for review and comment. Treasury provided written comments through the Federal Insurance Office, which are reproduced in appendix III. Additionally, Treasury provided technical comments, which we have incorporated, as appropriate. NAIC did not provide technical comments. In Treasury’s written response, the Federal Insurance Office agreed that limiting uncertainty in the insurance market following a certified act of terrorism was an important goal. The office accepted the recommendation and stated that it would work to implement it in the coming months. We are sending copies of this report to the appropriate congressional committees, the Secretary of the Treasury, the Chief Executive Officer of NAIC, and other interested parties. In addition, the report is available at no charge on the GAO website at https://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-8678 or garciadiazd@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Objectives, Scope, and Methodology The objectives of our report were to (1) examine the changes in explicit fiscal exposure under the Terrorism Risk Insurance Act (TRIA) and how insurers have adjusted to the changes; and (2) describe situations in which implicit fiscal exposures may arise and might become explicit. To address these objectives, we reviewed the Terrorism Risk Insurance Act of 2002, Terrorism Risk Insurance Extension Act of 2005, Terrorism Risk Insurance Program Reauthorization Acts of 2007, 2015, and 2019, implementing regulations, and congressional records. We also reviewed prior GAO work on TRIA and federal fiscal exposures. We reviewed reports from the Department of the Treasury (Treasury), the Congressional Budget Office, and Congressional Research Service. We also reviewed relevant reports from academic researchers and industry stakeholders. We used Treasury’s data calls from 2017 to 2019 for aggregated direct- earned premiums, numbers of insurers, and hypothetical loss scenarios throughout this report. We evaluated the reliability of the data by performing electronic tests and interviewing staff from Treasury and its data contractor, and industry stakeholders. We found the data sufficiently reliable for the purposes described below. We interviewed Treasury officials and representatives from the National Association of Insurance Commissioners (NAIC) and industry stakeholders, including insurers, insurance trade associations, a rating agency, risk modelers, and an insurance broker. We selected a nongeneralizable sample of five insurers to interview. These insurers were selected because they provide terrorism coverage to businesses and reflect a mix of sizes and types of insurance. In interviews, we asked about aspects of the program and the insurance market, and risks that could lead to implicit exposure. To describe the potential explicit fiscal exposure to the federal government, we reviewed the relevant laws and analyzed the changes made in each reauthorization. To quantify and compare the federal explicit exposure from potential terrorist events in 2015 and 2020, we estimated the TRIA-eligible direct-earned premiums for those years and used simulated loss scenarios similar to those we previously developed. To estimate the TRIA-eligible direct-earned premiums for 2015 and 2020, we used aggregated direct-earned premiums for 2016–2018 from Treasury’s data calls. Specifically, first we calculated the annual percentage change in direct-earned premiums for 2016–2017 and 2017–2018. Second, we found the average change to be an increase of about 0.4 percent. This estimate was smaller than the percentage change we used in the 2017 report, but we found the estimate reasonable because our current estimate was based on Treasury’s data, specific to TRIA-eligible lines of insurance. Our 2017 estimate was based on annual estimates of terrorism risk revenue. Third, we estimated the 2015 and 2020 direct-earned premiums using the 0.4 percent average annual change and the reported 2016 and 2018 direct-earned premiums, respectively. To check for reliability, we used the same method to estimate direct-earned premiums for 2014. Our 2014 estimate matched the 2014 data used in our 2017 report. We calculated the change in both nominal and real dollars and decided to use nominal dollars, which is consistent with how Treasury reports direct-earned premiums. We found the data to be sufficiently reliable for estimating the TRIA-eligible direct-earned premiums for 2015 and 2020. To compare events occurring under the program provisions in effect in 2015 and 2020, we analyzed how losses would be shared between the government and insurers by modeling terrorist events with insured losses of $5 billion, $25 billion, $50 billion, $75 billion, and the maximum terrorism event size ($100 billion in insured losses) in 2015 and 2020. In each case, we modeled the affected insurers to have an aggregate direct-earned premium base of 25, 35, 55, and 100 percent of all TRIA-eligible premiums. To demonstrate how the program trigger may affect insurers, we also modeled a smaller terrorist event with $290 million in losses and assumed that the affected insurers had an aggregate direct-earned premium base of $750 million. In all cases, we estimated the portion of federal losses that would be subject to mandatory and discretionary recoupment. To determine how insurers have adjusted to changes in TRIA and measure market stability, we reviewed Treasury and past GAO reports and relevant literature, analyzed Treasury data, and interviewed industry stakeholders. We conducted a literature search to determine how changes in the TRIA program parameters affected insurers and we summarized relevant findings from our review of the literature we identified. Additionally, we summarized Treasury and industry stakeholders’ views on insurers’ ability to cover their share of losses following an event and their willingness and ability to continue providing coverage after a large event without access to the federal share of losses to cover claims from the event. We also analyzed Treasury data to determine whether insurance lines have experienced changes in premiums or coverage availability since 2016. Specifically, we computed market shares by insurer category for calendar years 2016–2018, using direct-earned premiums from Treasury’s data calls. Using Treasury’s insurer categories (alien surplus lines, captive, nonsmall, and small), we computed market shares overall, and for TRIA-eligible lines of coverage in the U.S. terrorism risk insurance market, which includes all the states, the District of Columbia, U.S. territories, and other areas. We did not use number of policies because number of policies was not available in Treasury’s data for all insurer categories for all years. In the course of assessing data reliability for Treasury’s scenario data, we found some lack of clarity regarding two key terms under TRIA: “insured losses” and “insurance losses.” The external communication component of internal control—that management should externally communicate the necessary quality information to achieve the entity’s objectives—was significant to this objective, along with the related principle that management communicate quality information externally through reporting lines so that external parties can help the entity achieve its objectives and address related risks. We assessed Treasury’s external communications about these terms. Specifically, we reviewed the content of TRIA statutory language and program regulations and guidance. We obtained an interpretation of the terms from Treasury officials and also obtained industry stakeholders’ views on them. We determined Treasury’s scenario data to be reliable for the purpose of reporting the loss sharing as it was reported and for illustrating loss sharing examples. To identify potential situations in which implicit federal fiscal exposure may arise, we analyzed TRIA’s program design and reviewed our prior work for sources of implicit fiscal exposures, such as those faced by other disaster insurance programs. To ensure the reasonableness and completeness of our list of identified sources, we consulted with industry stakeholders and made modifications as appropriate. We grouped the sources of implicit fiscal exposures into three broad categories: (1) any unrecouped program expenditures; (2) federal assistance for uninsured or underinsured terrorism losses (including uninsured losses in a nuclear, biological, chemical, or radiological event, losses in excess of the $100 billion program cap, and policyholders’ retained losses); and (3) federal assistance to stabilize the insurance market for insurers that may be unable to access the loss-sharing feature of the program or for lines of insurance not included under TRIA. To describe the potential exposure resulting from not executing the recoupment of program expenditures, we summarized prior GAO reports and industry stakeholder views on the risks and challenges of collecting mandatory and discretionary recoupment funds, including the associated collection time frames. To describe the potential exposure resulting from federal assistance for uninsured or underinsured losses, we reviewed and summarized findings in prior GAO reports on past instances in which the federal government provided disaster assistance (such as following the September 11 terrorist attacks, the financial crisis, and large natural disasters). We summarized Treasury, NAIC, and industry stakeholder views on the likelihood of terrorist events reaching the program cap, their expectations for federal intervention in that case, and the importance of the cap to the terrorism risk insurance market. To describe the potential exposure resulting from federal assistance to stabilize the insurance market following a terrorist event, we interviewed industry stakeholders about scenarios that could produce an expectation for government assistance (implicit exposure), the stability of the terrorism insurance industry, how program details could affect different insurer groups, and insurers’ options for covering their share of losses. We used Treasury’s data to quantify the number and share of insurers that might not be able to access TRIA’s loss-sharing provision and to illustrate estimated loss sharing in Treasury’s scenarios of losses that would have occurred in three hypothetical events: New York City (2016), Chicago (2017), and San Francisco (2018). We also interviewed various insurer associations. We conducted this performance audit from July 2019 to April 2020, in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Changes in TRIA Reauthorizations As shown in table 1, Terrorism Risk Insurance Act (TRIA) reauthorizations through 2015 changed several loss-sharing provisions, which decreased the federal share and increased the insurer share of losses. The 2015 reauthorization required incremental decreases in the federal share of losses over 5 years. The 2019 reauthorization extended the program until the year 2027, but did not make any changes to the program parameters discussed below. Appendix III: Comments from the Department of the Treasury Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Jill Naamane (Assistant Director), Karen Jarzynka-Hernandez (Analyst in Charge), Rudy Chatlos, Giselle Cubillos-Moraga, Kaitlan Doying, Lijia Guo, John Karikari, Barbara Roesmann, Jessica Sandler, Jena Sinkfield, Frank Todisco, and Rachel Whitaker made key contributions to this report.

Why GAO Did This Study Congress enacted TRIA to help ensure the availability and affordability of commercial property/casualty insurance for terrorism risk and to address potential effects on the economy in the absence of such coverage. Under the TRIA program, which is set to expire December 31, 2027, the government and insurers share losses following a certified act of terrorism. TRIA creates explicit fiscal exposure because the government is legally required to make payments to insurers after such an event, but there also may be some implicit exposure from an expectation of federal spending. To date, Treasury has not certified any acts of terrorism. GAO was asked to examine federal fiscal exposure under the TRIA program. This report (1) examines changes in explicit fiscal exposure under TRIA and how insurers have adjusted to the changes, and (2) describes situations in which implicit fiscal exposures may arise and might become explicit. To conduct this work, GAO reviewed the TRIA statute and related studies, analyzed Treasury data, and interviewed a nongeneralizable sample of insurers of different sizes providing various types of insurance. What GAO Found Terrorism Risk Insurance Act (TRIA) reauthorizations through 2015 have decreased federal fiscal exposure, and insurers have adjusted by managing their increased risk. Changes in the TRIA program that the Department of the Treasury (Treasury) administers—particularly incremental changes since 2015—reduced the government's explicit fiscal exposure from a certified act of terrorism (see figure). For example, by increasing the program trigger—minimum amount of industry-insured losses needed to activate the program—Congress potentially reduced the number of events that qualify for federal payments. As explicit federal fiscal exposure has decreased, insurer exposure has increased. Nevertheless, the market for terrorism risk has remained stable. However, some insurers are uncertain how Treasury defines insured losses for the purposes of calculating whether the program's $200 million trigger or $100 billion cap have been reached. For example, some insurers interpreted insured losses to include the portion of losses policyholders retain, which was different from Treasury's interpretation. Differences in interpretations could lead to disputes between insurers and Treasury following a terrorist event. One purpose of TRIA is to stabilize the market following a terrorist event. Communicating how it would calculate losses toward these program amounts could help Treasury alleviate uncertainty in the insurance market following a terrorist event. The government also has implicit fiscal exposure following a terrorist event, arising from expectations based on current policy or past practices that it may provide assistance, even when it is not legally required to do so. Although the government may not act on these expectations, the government's implicit exposure might become explicit if it chooses not to recoup the full federal share of losses from property/casualty policies, as allowed under TRIA, to prevent further stresses on the insurance market after a major terrorist event; assists companies with uninsured or underinsured losses after a terrorist event or when losses exceed the program cap; covers uninsured losses from a nuclear, biological, chemical, or radiological terrorism event; or assists insurers with losses that did not meet TRIA's trigger for loss sharing, or that were incurred in excluded lines of coverage, such as life and health insurance. What GAO Recommends GAO recommends that Treasury communicate how it would calculate losses, as they relate to policyholder retention amounts, in determining the TRIA program trigger and cap. Treasury agreed with the recommendation.

gao_GAO-19-583T

gao_GAO-19-583T_0

Background Selected agencies’ funding for university STEM research. The five federal agencies included in our preliminary analysis provide billions of dollars annually for university research in STEM fields, with NIH providing more than the other four agencies combined. Table 1 details the total amount of research funding provided to universities by each agency in fiscal year 2017. Sexual harassment. As defined in the National Academies of Sciences, Engineering, and Medicine (NASEM) 2018 report, sexual harassment encompasses three types of behavior: Sexual coercion: Favorable treatment conditioned on sexual activity. Unwanted sexual attention: Verbal or physical unwelcome sexual advances, which can include assault. Gender harassment: Sexist hostility and crude behavior. The most common form of sexual harassment is gender harassment, which generally involves hostility, exclusion, or other discrimination based on a person’s gender. The 2018 report found that sexual harassment in academia is significantly more common among female students in engineering and medical majors than in non-STEM fields. According to the report, at least five factors create the conditions under which sexual harassment is likely to occur in STEM programs and departments in academia: Perceived tolerance for sexual harassment Environments where men outnumber women and leadership is male Environments in which the power structure of an organization is hierarchical with strong dependencies on those at higher levels or in which people are geographically isolated Increased focus on symbolic compliance with Title IX Uninformed leadership on campus Title IX of the Education Amendments of 1972. Title IX of the Education Amendments of 1972 is the primary federal law that addresses sex discrimination in all federally funded grant programs at educational institutions. Under Title IX, federal agencies that award grants to educational institutions have enforcement responsibilities to ensure such institutions do not discriminate based on sex. Enforcement responsibilities include issuing regulations, conducting periodic compliance reviews of funding recipients, and investigating timely written complaints of sex discrimination against recipients. DOJ and the Department of Education have other responsibilities for administering Title IX. DOJ is designated by Executive Order No. 12250 to coordinate Title IX compliance across federal agencies, including information sharing. Federal grant awards and grant life cycle. In general, federal agencies administer grants through a common administrative life cycle: pre-award, award, implementation, and closeout. During the pre-award stage, most of the agencies we reviewed require grantees to submit an “assurance of compliance” form as part of their grant application to attest compliance with anti-discrimination laws, including Title IX. For the award stage, the federal agency awarding the grant enters into an agreement with grantees stipulating the terms and conditions for the use of grant funds. During the implementation stage, among other things, the federal agency manages and oversees the grant, including any Title IX compliance reviews. A Title IX compliance review is an agency’s assessment of whether a grantee is complying with the law. Federal agencies may conduct these reviews onsite at an institution (grantee) or via a desk audit. In the closeout stage, the awarding federal agency and grantee bring the grant to its conclusion, once all the work associated with the grant agreement is complete, the grant end date has arrived, or both. Among the federal agencies we reviewed, different offices handle various aspects of grant compliance. Generally, each agency’s civil rights or diversity office conducts Title IX compliance reviews, develops policies and procedures for grantees, and investigates allegations and complaints involving university researchers supported by their agency’s federal STEM grants. The office that awards grants generally creates and modifies grant terms and conditions. Resources to Address Sexual Harassment Complaints Vary Our preliminary analysis indicates that the selected federal agencies’ staff and budget available to address sexual harassment complaints from individuals at grantee universities varies according to the duties and funding for the primary agency offices responsible for addressing the complaints, as well as with the number of complaints received from grantees. Duties and funding for offices responsible for addressing complaints. Our preliminary analysis shows that all five agencies (DOE, HHS, NASA, NSF, and USDA-NIFA) primarily address sexual harassment complaints through their civil rights or diversity offices. However, these offices are responsible for more than just addressing complaints and preventing sexual harassment at grantee universities; they also oversee a number of civil rights, diversity and inclusion efforts for the entire agency. Moreover, most of these offices also address internal employee sexual harassment complaints and other discrimination issues. For example, HHS officials described how staff in their Office for Civil Rights at headquarters and eight regional offices conduct compliance reviews and investigate all complaints alleging sexual harassment and other forms of discrimination against recipients of HHS federal financial assistance, including recipients of NIH grants. USDA- NIFA said their civil rights and diversity office staff are not always available when sexual harassment issues arise because they have other duties and also cover other discrimination issues. In addition, some agencies noted challenges in ensuring adequate staffing levels. For example, USDA-NIFA officials cited the need to fill vacant positions in their civil rights office, and NSF officials described a need to find staff with expertise in this complicated, specialized area. All five agencies fund their civil rights and diversity offices separately from their STEM research funding, and there is little relationship between the two budgets. For more information on selected agencies’ civil rights and diversity office staffing and budgets planned for fiscal year 2019, see table 2. Number of complaints received. Our preliminary analysis of sexual harassment complaint information indicates that four of the five selected agencies received three or fewer complaints from individuals at grantee universities from fiscal year 2015 through 2019. See table 3. Officials from DOE told us that because they receive so few sexual harassment complaints from individuals at grantee universities, they have enough resources to address those that are reported to their civil rights or diversity offices. In addition, officials from HHS told us that because they receive so few complaints, their civil rights office has used other oversight mechanisms, like Title IX compliance reviews, to examine whether sexual harassment is occurring at universities receiving HHS funds, including funds from NIH. However, as agencies continue to strengthen grantee policies or requirements, it may affect the number of complaints an agency receives from individuals at grantee universities, as well as the amount of resources an agency needs to address them. For example, NSF officials described how the number of sexual harassment complaints they receive has increased since the agency implemented new grant terms and conditions that require university grantees to report any sexual harassment findings involving a Principal Investigator or co-Principal Investigator for NSF-funded research. NSF officials also described an increased number of questions and calls about how to report incidents, requests for training and presentations, and meetings with program officers, awardee representatives and other stakeholders, among other items. Agencies Have Different Sexual Harassment Prevention Policies and Mechanisms for Communicating Them to Grantees Based on our preliminary review, all five of the selected agencies have established and communicated their own sexual harassment prevention policies to grantees within the last 2 fiscal years, but agency communication mechanisms and the content of these grantee policies vary. Specifically, our preliminary analysis shows that NASA, NIH, and NSF communicate their policies on sexual harassment in multiple forms, such as grantee policy manuals, best practices documents, and online FAQs. The result is that grantees receive a relatively high level of detail about preventing sexual harassment and mechanisms for reporting complaints. In contrast, Cabinet agencies DOE and USDA-NIFA provide fewer forms of guidance, either through their website or agency director and Secretary-level policy statements and documents, which focus more generally on the broader category of sex discrimination or provide different levels of information on sexual harassment prevention policies for grantees. See table 4 for more information. Regarding the content of the policies, our preliminary analysis shows that DOE, NIH, NSF, and USDA-NIFA updated their definitions of behaviors or actions that qualify as sexual harassment in their grantee policies and procedures, and NASA is in the process of doing so. The definitions are more specific than previous definitions; for example, they include descriptions of gender harassment, the most common form of sexual harassment. The increased specificity may make clear the behaviors or actions grantees are expected to address in their efforts to prevent sexual harassment. The agencies continue to develop and revise policies and communication mechanisms for grantees. Also, NSF and NASA have modified, or are taking steps to modify their grant terms and conditions to strengthen requirements for university grantees to report on findings of sexual harassment to the funding agency. Officials from both agencies told us these modifications will help hold grantees accountable for reporting sexual harassment; the NSF Director states on the agency’s website that these changes are “intended to provide targeted, serious consequences for harassers” while also providing “tools to make harassment stop without disturbing others’ careers and lives.” The requirement also supports the NASEM 2018 report recommendation for institutions to be transparent about reporting sexual harassment findings, which is intended to foster a culture and climate that does not tolerate sexual harassment at universities. Officials from cabinet agencies DOE, NIH (a component of HHS), and USDA-NIFA stated they would need to go through formal rulemaking processes to alter their grant terms and conditions in a similar manner. In addition, our preliminary analysis shows that two of the five selected agencies are taking steps to evaluate the effectiveness of their sexual harassment prevention policies and procedures for grantees. NSF officials told us that they are in the process of determining how best to evaluate the effectiveness of the new sexual harassment reporting requirements and how the requirements have affected grantees. Similarly, DOE is in the process of comparing its policies and procedures against other federal agencies’, according to officials. To date, the other three agencies have not yet evaluated the effectiveness of their policies, but officials at these agencies told us that they focus on ensuring grantees comply with Title IX regulations as a way to measure the effect of their policies. For example, NASA established a goal to promote compliance and encourage best practices among grantees, and the agency measures progress towards this goal through verifying grantee compliance with Title IX. USDA-NIFA officials are also in the process of creating a tool to provide a comprehensive blueprint for civil rights compliance—including Title IX compliance—and are planning to implement the tool in fiscal year 2020. We will continue to examine and assess the selected agencies’ sexual harassment prevention policies for university grantees and steps they are taking to evaluate them in our ongoing work. Agencies Have Taken Some Steps to Share Information and Collaborate Based on our preliminary review, all five selected agencies have taken some steps to promote information sharing and collaboration among agencies on Title IX compliance reviews through DOJ’s Title IX STEM working group. According to officials, the group discusses strategies for conducting joint Title IX compliance reviews to leverage limited agency resources and share best practices. For example, DOE and NSF have conducted three joint compliance reviews, and NASA and NSF told us that they are in the process of conducting a joint review. These joint reviews can be helpful, as the selected agencies conduct a small number of compliance reviews (two to four) annually relative to the number of university grantees who must comply with Title IX. Despite this collaboration, all five selected agencies reported challenges in obtaining and sharing information. For example, all five selected agencies told us they rarely discuss sexual harassment cases at DOJ’s Title IX STEM working group meetings unless they are directly related to an ongoing or planned compliance review. In addition, DOE, NASA, NIH, and NSF stated they rarely learn about instances of sexual harassment from voluntary reporting from universities or other federal agencies and instead must rely on other sources, such as news reports. This situation may change at NSF and NASA, which have taken steps to modify their grant terms and conditions to require reporting of sexual harassment findings by grantees. Challenges in obtaining and sharing information on sexual harassment cases may increase the risk of a situation known as “pass the harasser,” in which a researcher with substantiated findings of sexual harassment obtains employment at another university or grants from another funding agency without the university or funding agency being aware of the researcher’s history. Officials from all five selected agencies noted a willingness to participate in an interagency working group to address the culture of sexual harassment in STEM research that moves beyond conducting Title IX compliance reviews. The White House’s Office of Science and Technology Policy (OSTP) has taken steps to establish an interagency working group. In May 2019, OSTP established a joint committee under the National Science and Technology Council to address challenges in the research environment. OSTP, NIH, NSF, DOE, and the National Institute of Standards and Technology Directors were selected as joint committee chairs to engage with the academic and science community for policymaking insight and to convene interagency efforts. According to DOE officials, the committee will address several priorities, including the development of policies and procedures across the federal government regarding sexual harassment in the research environment. Three of the five selected agencies (NSF, NASA, and DOE) stated OSTP would be the appropriate entity to establish uniform sexual harassment policy guidelines to help provide consistency across the federal government. NSF and NASA officials suggested that DOJ or the Department of Education would be the appropriate agencies to collaborate with OSTP on the ongoing monitoring of sexual harassment policy guidelines. All five selected agencies reported taking collaborative steps with universities and federal agencies to address the culture and climate for women in STEM. For example, in 2019, NIH established a working group with university experts to collaborate with other federal agencies to assess the current state of sexual harassment allegation investigation, reporting, remediation, and disciplinary procedures at NIH-funded organizations and advise on oversight, accountability, and reporting measures for grantees, among other things. In addition, all five agencies provided examples of collaborative efforts that would help address the culture of sexual harassment in STEM research. For example, NASA officials told us that it would be helpful to conduct joint meetings with other university grantees across agencies to discuss sexual harassment in science. Lastly, efforts to improve information sharing and collaboration across agencies beyond conducting Title IX compliance reviews are consistent with findings in the 2018 NASEM report, which states, “adherence to legal requirements is necessary but not sufficient to drive the change needed to address sexual harassment.” We will continue to examine and assess selected agencies’ Title IX reviews and efforts to collaborate and share information in our ongoing work. In closing, I note that we are continuing our ongoing work on this topic. Sexual harassment is not only degrading to individual researchers, it undermines the quality and fairness of our nation’s research enterprise. It is therefore important that federal agencies ensure their grantees effectively prevent and address sexual harassment in STEM research. We look forward to continuing our work to determine whether additional federal actions may be warranted to promote this objective. Chairwoman Johnson, Ranking Member Lucas, and Members of the Committee, this completes my prepared statement. I would be pleased to respond to any questions that you may have at this time. GAO Contact and Staff Acknowledgments If you or your staff have any questions about this testimony, please contact John Neumann, Managing Director, Science, Technology Assessment, and Analytics, at (202) 512-6888 or NeumannJ@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Key contributors to this testimony include Rob Marek (Assistant Director), Michelle St. Pierre (Assistant Director), Kristy Kennedy (Analyst-In-Charge), Nora Adkins, Caitlin Cusati, Nkenge Gibson, Amanda Postiglione, Janay Sam, and Ben Shouse. 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.

Why GAO Did This Study In fiscal year 2017, U.S. universities were awarded over $15 billion in federal grant funding for STEM research. Federal agencies are required to enforce Title IX—a law prohibiting discrimination on the basis of sex in education programs receiving federal financial assistance—including at universities they fund. Sexual harassment is not only degrading and illegal, it has a negative effect on the ability of women to engage in research at the same level as men. GAO was asked to review federal efforts to help prevent sexual harassment by STEM research grantees. This testimony is based on ongoing GAO work and provides preliminary observations on selected agencies: (1) availability of staff and budget to address sexual harassment complaints at universities they fund for STEM research; (2) efforts to establish and communicate policies and procedures for university grantees on preventing sexual harassment; and (3) steps taken to promote information sharing and collaboration among agencies to prevent sexual harassment at universities they fund for STEM research. GAO selected five federal agencies that together funded approximately 80 percent of STEM research from fiscal year 2015 through 2017, the latest data available. GAO reviewed these agencies' relevant regulations and documentation. GAO also interviewed agency officials as part of GAO's ongoing work. What GAO Found Based on preliminary information, the availability of agency staff and budget varies across the five selected agencies for efforts to address sexual harassment complaints at universities that use federal funds for Science, Technology, Engineering, and Mathematics (STEM) research. While four of the five agencies received three or fewer sexual harassment complaints from individuals at grantee universities from 2015 through 2019, changes to agency grantee policies or requirements could impact the number of complaints an agency receives and the amount of resources an agency needs to address them. The five selected agencies have established and communicated sexual harassment prevention policies to university grantees to varying degrees. Agencies vary in how they have: Provided detailed policies to grantees on sexual harassment. Three agencies—the National Aeronautics and Space Administration (NASA), Health and Human Services (HHS) National Institutes of Health (NIH), and the National Science Foundation (NSF)—have communicated relatively detailed policies on sexual harassment by issuing multiple forms of guidance, such as grantee policy manuals and best practices documents. In contrast, the Department of Energy (DOE) and Department of Agriculture (USDA) National Institute of Food and Agriculture (NIFA) communicated through more general documents, including policy statements that do not specifically address grantees. Modified grant terms and conditions . Two agencies are modifying the terms and conditions of grants to require grantees to report sexual harassment. NSF now requires grantees to increase transparency by reporting findings of sexual harassment to NSF, and NASA plans to implement the same requirement. Evaluated effectiveness of grantee policies. To date, the five agencies have not evaluated the effectiveness of their grantee policies and procedures to prevent sexual harassment, although two agencies are in the process of planning such evaluations. Based on our preliminary analysis and interviews, all five selected agencies have taken some steps to promote information sharing and collaboration among agencies on the prevention of sexual harassment. But they also noted challenges to these efforts, such as the lack of information on sexual harassment cases. These challenges may increase the risk that universities or agencies are unknowingly funding researchers with a history of past sexual harassment findings. The White House's Office of Science and Technology Policy has taken steps to create an interagency working group by establishing a joint committee in May 2019 under the National Science and Technology Council with NIH, NSF, DOE, and the National Institute of Standards and Technology Directors. The committee plans to address challenges in the research environment, including the lack of uniform federal sexual harassment policies. What GAO Recommends GAO is not making recommendations at this time but will consider making them, as appropriate, as it finalizes its work.

gao_GAO-19-586

gao_GAO-19-586_0

Background DLA’s primary purpose is to meet the logistics requirements of the armed forces for food, clothing, fuel, spare parts, and other items. DLA’s major responsibilities are to buy and distribute about 5 million distinct consumable, expendable, and reparable items. In order to fulfill the logistics requirement of the armed forces, DLA provides more than $35 billion in goods and services annually. As part of its responsibility to provide spare parts, DLA officials stated that three of DLA’s major subordinate commands conduct reverse engineering—Aviation, Land and Maritime, and Troop Support. These commands are directly responsible for meeting the following military services’ needs. Specifically: Aviation provides aviation weapons systems spare parts, flight safety equipment, maps, environmental products, and industrial plant equipment. Land and Maritime provides for ground-based and maritime weapons systems spare parts, small arms parts, and fluid handling systems. Troop Support handles food, textiles, construction material, industrial hardware, and medical supplies and equipment, including pharmaceuticals. DLA Reverse Engineering Process Reverse engineering is the process of replicating a design by physically examining and measuring an existing item to develop the technical data necessary to reproduce the item functionally and dimensionally. In other words, it is the process of extracting information about an item from the item itself. According to DLA’s procedure, the intent of reverse engineering is either to develop an approved technical data package—the details needed to duplicate the item such as drawings or specifications, among other things—or approve a new supplier. The technical data package could be used for future competitive procurements when the item is needed for sustainment purposes. DLA uses reverse engineering for several reasons, such as identifying potential new sources for obsolete parts or those supplied by only one source, increasing competition, and achieving savings. Parts are identified for reverse engineering generally because the government does not have necessary legal rights to the drawings or lacks data needed to facilitate competitive procurements. However, reverse engineering is generally considered the least attractive alternative for replenishing parts because it is expensive and time-consuming. Therefore, according to DLA guidance, before starting reverse engineering DLA personnel should try to buy data from the original equipment manufacturer, previous manufacturing sources, or other data rights holder. DLA personnel also need to consider the costs of acquiring the data rights before initiating reverse engineering. DLA guidance for its reverse engineering processes is the same regardless of the size of the businesses involved. In order to start a reverse engineering project, DLA has established criteria, such as documenting a business case and showing that the part meets certain yearly purchase thresholds. Specifically, according to DLA guidance, procurements of the part over the past 2 years generally should exceed $10,000 in each year. Additionally, the engineering support activity at a military department (Army, Air Force, or Navy) responsible for the part generally approves a project before reverse engineering can begin. Once reverse engineering is completed, the technical data package must be approved by the appropriate engineering support activity. Figure 1 shows the general process from identifying a part as a potential candidate for reverse engineering to approving a technical data package or a new source of supply, and indicates whether DLA or a contractor is responsible at each step. DLA has two reverse engineering programs—the Replenishment Parts Purchase or Borrow program, funded by contractors, and the DLA internally funded efforts. Contractor-Funded Efforts Through the Replenishment Parts Purchase or Borrow program, contractors reverse engineer a part at their own expense. DLA’s goals for this program are to increase competition and achieve savings. After successful reverse engineering and military approval of a technical data package, the contractor becomes a new source of supply for the part. Contractors may identify parts for reverse engineering—mostly small hardware items and electrical components, such as antennas and cables—from a candidate list on a DLA website or through their own research. To conduct reverse engineering, contractors may purchase or borrow parts under agreement with the government subject to certain conditions, such as the part is not classified or considered a critical part— one that is crucial enough that a failure of that part would result in serious injury or impact the success of a mission—among other things. The government incurs minimum cost, if any, in this program. DLA-Funded Efforts The goal of DLA-funded reverse engineering is to develop a technical data package that will be used in future competitive procurements. Typically, DLA funds efforts for parts that are available from only one source, are obsolete, or have limited data rights. These efforts occur in several ways: DLA engineers conduct the reverse engineering. DLA funds the efforts through partnerships with other DOD entities. DLA awards contracts to companies to create drawings of parts that the government can then use in competitive procurements. The government obtains full use of these drawings. Intellectual Property: Data Rights and Patents Intellectual property derives from the work of the mind or intellect and is an application, right, or registration relating to property—such as an idea, invention, or process. It includes patents and proprietary information: Patents—grants an inventor the right to exclude others from making, using, or selling an invention in the United States, typically for a period of 20 years. The holder of a valid patent is the only authorized supplier of the patented item unless another supplier has acquired a license to manufacture and sell the item. Proprietary information—includes technical data which represents trade secrets usually developed at private expense, such as design, material composition, or manufacturing processes. The owner of a specific item does not make the information available to others without obligations concerning its confidentiality. This confidentiality of proprietary information does not protect it from discovery by reverse engineering. In order to share technical data, for example, to issue a solicitation for competitive procurement of an item, DLA needs to have sufficient data rights. Data rights are the government’s contractual license rights for technical data—recorded physical and material characteristics, such as item specifications, engineering drawings, or operating and maintenance manuals. If the government is entitled to and acquires unlimited data rights, it is allowed to use, reproduce, or disclose that technical data. When the government acquires limited rights the government may only use the data internally, such as for the operation of equipment, but may not disclose technical data for the purpose of procuring an item from another contractor. DLA procedures indicate the government should consider the cost of acquiring the data rights before initiating reverse engineering. Reverse Engineered Spare Parts Resulted in Some Lowered Prices and Increased Supplier Opportunities for Small Businesses Number of Reverse Engineered Spare Parts DLA’s Aviation, Land and Maritime, and Troop Support major subordinate commands initiated over 1,600 reverse engineering projects during fiscal years 2015 through 2018, according to DLA data. Table 1 describes the number of projects initiated in each fiscal year by DLA’s Aviation, Land and Maritime, and Troop Support for both the contractor- and DLA-funded efforts. Nearly two-thirds of all reverse engineer projects involved parts in the following five categories. 1. Hardware and abrasives. For example, screws, nuts, washers, and keys. 2. Vehicular equipment components. For example, floor mats, vehicle door hinges, and tailgates. 3. Electrical and electronic equipment components. For example, pressure switches, electrical assemblies, and antennae. 4. Electric wire, and power and distribution equipment. For example, batteries, wiring harnesses, and special purpose cable assemblies. 5. Aerospace craft components and accessories. For example, insulation blankets, filters, and door handles. Figure 2 shows examples of items in these categories. Some Reverse Engineering Projects Resulted in Lowered Prices Some of DLA’s projects resulted in lower prices for the reverse engineered parts in subsequent procurements. According to DLA data, the agency saved at least $22 million from reverse engineering projects initiated from fiscal years 2015 through 2018 as a result of lower prices paid. For example, in one project we reviewed, a small business successfully reverse engineered a retaining ring, which assists in securing parts in an aircraft engine, purchased by Troop Support and became a new approved source of supply. A subsequent purchase of this part resulted in a unit price that was almost $70 lower per unit compared to the most recent purchase before reverse engineering. We found that Troop Support saved over $11,000 through this project. We found that 141—or less than 10 percent—of all projects initiated from fiscal years 2015 through 2018 were successfully completed. Table 2 describes the number of projects that were completed, in-process, and canceled as of December 2018, according to DLA. DLA officials told us that less than 10 percent of projects have been completed for several reasons. DLA officials stated that the engineering support activities at the military services sometimes take a long time to respond to requests for approval, which adds time to reverse engineering projects. For contractor-funded projects under the Replenishment Parts Purchase or Borrow program, reasons include the following: DLA officials explained that contractors sometimes decide not to complete reverse engineering because the contractors determine the effort would be more labor intensive than originally anticipated or their priorities shift. DLA officials also stated that DLA does not necessarily hear back from a contractor after it purchases a part to begin the reverse engineering process. Representatives from contractors we spoke with told us that they experience delays in obtaining responses from DLA and military service engineers regarding approvals for their projects. In addition, DLA officials explained that the completion rate for DLA- funded projects is low for the following reasons: More urgent priorities arise after a project begins. Personnel who work on reverse engineering typically have other duties in addition to these efforts and if more urgent priorities emerge after a reverse engineering project has started, the effort may need to be postponed or abandoned to attend to the more pressing priority. The need no longer exists for some parts that were identified as reverse engineering candidates. Most Businesses Conducting Reverse Engineering Were Small Businesses We found that DLA’s reverse engineering programs created opportunities for contractors—particularly small businesses—to become new suppliers. According to DLA data, 124 contractors participated—or had the opportunity to become new sources of supply for parts—in the contractor- funded reverse engineering program from fiscal years 2015 through 2018. Of these, we determined that 103 were small businesses. In addition, while DLA performed reverse engineering work for most of the DLA- funded projects, DLA awarded contracts to 6 companies to conduct reverse engineering, all of which were small businesses. Roughly one-third of the 124 contractors worked on only one spare part during the time frame we reviewed, while the others initiated reverse engineering for multiple parts. For example, one small business requested to reverse engineer more than 50 parts during fiscal years 2015 through 2018. A representative for this contractor told us they use a business analytics system that queries publicly available information in order to identify opportunities for reverse engineering. While most companies conducting reverse engineering were small businesses, almost half the contractors that supplied parts prior to reverse engineering were not small businesses. Specifically, for fiscal years 2015 through 2018, DLA identified 74 contractors whose parts had been successfully reverse engineered, of which we determined that 34 were not small businesses and 26 were. In general, representatives at the small businesses we spoke with stated that the reverse engineering programs help small businesses. For example, representatives of one contractor that requested to reverse engineer a part told us that DLA’s reverse engineering program as a whole benefits small businesses because most of the parts being reverse engineered are originally from other than small businesses, and the program is a path that allows small businesses to become approved suppliers. In addition, representatives from one contractor that reverse engineered multiple projects stated that working with DLA has allowed their business to establish past performance ratings, which will help in future government procurements. Officials from DLA competition advocate and small business offices stated that reverse engineering is generally beneficial for small businesses. They stated that small businesses have not registered complaints about DLA’s reverse engineering program. Rather, small businesses seek opportunities for additional business with DLA, which reverse engineering can help provide. In addition, industry associations we spoke with stated that reverse engineering is a way to involve small business. They stated that reverse engineering projects provide small businesses opportunities to become qualified suppliers and compete for future DLA contracts. DLA Processes Generally Safeguard Intellectual Property in Reverse Engineering Efforts We found that DLA has processes to protect intellectual property, such as patented designs and proprietary information during reverse engineering. Of the 10 small businesses we spoke with that were involved with our 19 selected projects, none of the representatives identified concerns with DLA’s practices for protecting intellectual property. Although DLA’s standard operating procedure applies to both DLA-funded efforts and contractor-funded efforts, the provisions that safeguard against patent infringement are specific to the contractor-funded program. According to these provisions, patented materials should not be approved or shared for reverse engineering. Aviation and Land and Maritime officials stated that they physically review parts for patent marks before reverse engineering can take place for contractor-funded efforts under the Replenishment Parts Purchase or Borrow program. However, Troop Support officials said they do not check for patents under this program because they are supplying parts for legacy systems—systems that are typically 20 years or older—which means any potential patents would have expired. They also stated that they conducted reverse engineering to a limited extent through the Replenishment Parts Purchase or Borrow program. We found Troop Support initiated 30 of the 617 Replenishment Parts Purchase or Borrow efforts from fiscal years 2015 through 2018. For DLA-funded efforts, Aviation and Land and Maritime officials stated that they physically review parts for patents, and the project is stopped if one is found. For example, officials from Land and Maritime canceled a project for an eye guard because officials found drawings marked with a patent number. The officials could not determine if the patent had expired so they canceled the project. Aviation officials also stated that they rarely encounter patent markings; they have seen four or five in the last 2 years. Troop Support officials stated they do not check for patent markings and they rely on the engineering support activities to check for patents. Figure 3 shows an example of a patent marking. DLA’s guidance for reverse engineering also does not allow the release of limited rights data for the contractor-funded projects under the Replenishment Parts Purchase or Borrow program. DLA officials at all three major subordinate commands told us that they do not release drawings that have limited data rights to contractors conducting reverse engineering under this program. DLA guidance does not cover whether data can be used under DLA-funded efforts. Officials at Aviation, Land and Maritime, and Troop Support all stated that proprietary or limited release drawings or technical data related to the part cannot be used by engineers who work on reverse engineering efforts. Aviation officials noted that the tracking system records those who have seen propriety drawings, and this information is used to ensure that these individuals do not work on reverse engineering projects related to those drawings. Land and Maritime officials stated that the tracking system used for reverse engineering projects also identifies drawings marked as proprietary and controls who has access to drawings. In general, the small businesses we met with did not express any concerns about how DLA handles intellectual property. We spoke with four small businesses that supplied parts that had been reverse engineered by other businesses. Representatives of these four small businesses confirmed that they did not hold patents on the parts that were reverse engineered. Further, the small businesses that conducted reverse engineering stated that DLA adequately protected intellectual property. In one case, a representative from a small business that participates in the Replenishment Parts Purchase or Borrow program stated that DLA has never released a part that had controlled data to them and DLA takes protection of proprietary data seriously. Another contractor who performs DLA funded reverse engineering efforts stated drawings are not released and sometimes the function of the part is not even shared with contractors. DLA command officials stated they had not heard of any concerns from small businesses about their intellectual property being used inappropriately. In addition, officials from the DLA small business office stated they had not heard concerns from any small businesses. Further, one official from the small business office noted that the parts DLA purchases are not new innovations and so do not necessarily have protected intellectual property. Competition advocates stated they received no complaints from small businesses. The industry associations we met with asked their small business membership if there were any specific concerns regarding DLA’s protection of intellectual property. Officials from the Aerospace Industries Association, the National Defense Industrial Association, and the National Association of Manufacturers stated there were no complaints from the small businesses represented by their groups about the businesses’ experiences working with DLA. Agency Comments We provided a draft of this report to DOD for review. DOD had no comments. We are sending copies of this report to the Secretary of Defense; the Director, Defense Logistics Agency; appropriate congressional committees; and other interested parties. This report will also be available at no charge on GAO’s website at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-4841 or by e-mail at makm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix II. Appendix I: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Penny Berrier, Assistant Director; Brandon Booth, Alexandra Dew Silva, Stephanie Gustafson, Victoria Klepacz, Jean McSween, Ralph Roffo, Roxanna Sun, Anne Louise Taylor, and Alyssa Weir made key contributions to this report.

Why GAO Did This Study The Department of Defense spends tens of billions of dollars annually to sustain military assets including aircraft, ships, and missiles. In support of this effort, DLA strives to maintain a competitive supplier base through reverse engineering—the process of examining an item, such as a spare part, with the intent of replicating its design. Contractors consider intellectual property, such as their technical data and patented material, essential to their success. DLA also takes steps to safeguard contractors' intellectual property during reverse engineering. The Senate Armed Services Committee report accompanying a bill for the fiscal year 2018 National Defense Authorization Act included a provision for GAO to review DLA's reverse engineering efforts, including the protection of small businesses' intellectual property. This report describes (1) DLA's reverse engineering programs and the extent to which small businesses participated in these programs from fiscal years 2015 through 2018; and (2) how DLA safeguards certain intellectual property within its reverse engineering efforts. GAO analyzed data from three DLA commands—Aviation, Land and Maritime, and Troop Support, those that conduct reverse engineering—from fiscal years 2015 through 2018. GAO reviewed a nongeneralizable sample of 19 reverse engineering projects involving 13 parts, selected to include a variety of characteristics, such as the size of the contractors involved. GAO reviewed DLA's guidance and interviewed DLA officials and representatives from small businesses about safeguarding intellectual property as part of reverse engineering. What GAO Found The Defense Logistics Agency (DLA) is responsible for providing logistics support to the warfighter, including spare parts for military assets. From fiscal years 2015 through 2018, DLA initiated over 1,600 reverse engineering projects for spare parts at three of its commands—Aviation, Land and Maritime, and Troop Support. DLA uses reverse engineering to identify potential new sources for spare parts that are available from only one source and to achieve savings. DLA funded about 1,000 of the reverse engineering projects, while contractors funded the remaining 600 projects. Nearly two-thirds of all reverse engineering projects involved parts in five categories, with examples of the three largest categories illustrated in the figure. GAO found that the majority of contractors conducting reverse engineering for DLA were small businesses. Specifically, DLA identified 124 contractors that conducted reverse engineering projects from fiscal year 2015 through 2018, 103 of which GAO determined were small businesses. According to small business representatives and DLA officials, reverse engineering is beneficial for small businesses and can help provide opportunities for additional business with DLA. GAO found that the three DLA commands had processes to safeguard certain intellectual property in their reverse engineering efforts. Specifically: Officials from all three commands stated they do not release drawings with limited data rights to contractors interested in reverse engineering parts. Aviation and Land and Maritime officials stated that they check for patent markings on parts to ensure patented parts are not reverse engineered. Troop Support officials stated they do not check for patent marks because the parts they supply are often too old to have valid patents. The small businesses GAO met with did not identify concerns with how DLA handles intellectual property. Further, DLA officials stated that they had not received any complaints from small businesses about their intellectual property being used inappropriately.

gao_GAO-19-648T

gao_GAO-19-648T_0

Background The FSM and RMI are independent countries about 3,000 miles southwest of Hawaii. The FSM is a federation of four semiautonomous states—Chuuk, Kosrae, Pohnpei, and Yap—whose population and income vary widely. Chuuk, the largest state by population, has the lowest per-capita gross domestic product (GDP). Overall, the FSM had a 2016 population of approximately 102,000 and a GDP per capita of about $3,200. The RMI’s 2016 population was approximately 54,000 with a GDP per capita of about $3,600. The RMI’s most recent census, in 2011, found that approximately three-quarters of the population lived in Majuro, the nation’s capital, and on the island of Ebeye in the Kwajalein Atoll. Table 1 shows the FSM’s, FSM states’, and RMI’s estimated populations and annual GDP per capita in fiscal year 2016. Compact of Free Association (1986–2003) U.S. relations with the FSM and the RMI began during World War II, when the United States ended Japanese occupation of the region. Starting in 1947, the United States administered the region under a United Nations trusteeship. In 1986, after a period of negotiations, the United States entered into a compact of free association with the FSM and RMI that provided for economic assistance to the two countries, secured U.S. defense rights, and allowed FSM and RMI citizens to migrate to the United States. Amended Compacts of Free Association (2004– Present) In 2003, after a period of negotiations, the United States approved separate amended compacts with the FSM and the RMI that went into effect on June 25, 2004, and May 1, 2004, respectively. Compact Grants and Trust Fund Contributions The amended compacts’ implementing legislation authorized and appropriated direct financial assistance to the FSM and the RMI in fiscal years 2004 through 2023, with the base amounts decreasing in most years. The legislation also provided for partial inflation adjustment of the base amount of compact sector grants and trust fund contributions each year. As the base amount of compact sector grants decreases, the trust fund contributions generally increase by an equivalent amount. Because the annual inflation adjustment is less than full inflation, the value of compact sector grants declines in real terms. Figure 1 shows the amount of compact sector grants and trust fund contributions each fiscal year from 2004 through 2023. The amended compacts and associated fiscal procedures agreements require that compact sector grants support the countries in six core sectors—education, health, infrastructure, environment, private sector development, and public sector capacity building—with priority given to the education and health sectors. These grants are described in section 211(a) of each compact and are referred to as compact sector grants or 211(a) grants. Section 211(b) of the RMI compact further states that the RMI must target a specified amount of grants to Ebeye and other Marshallese communities within Kwajalein Atoll. The RMI military use and operating rights agreement (MUORA) states that the Kwajalein- related funds provided to the RMI in the compacts shall be provided through fiscal year 2023 and thereafter for as long as this agreement remains in effect. Compact Trust Fund Management and Implementation The amended compacts and their subsidiary trust fund agreements provided that each trust fund is to be managed by a compact trust fund committee. Each compact trust fund committee includes representatives from both the United States and the respective country, but the United States is required by the terms of the trust fund agreements to hold the majority of votes on each committee. The Director of Interior’s Office of Insular Affairs serves as the chair of each committee. Trust fund committee responsibilities include overseeing fund operation, supervision, and management; investing and distributing the fund’s resources; and concluding agreements with any other contributors and other organizations. As part of this oversight, the committees are to establish an investment and distribution policy. The committees are also to determine fiscal procedures to be used in implementing the trust fund agreements on the basis of the fiscal procedures used for compact grant administration, unless otherwise agreed by the parties to the agreement. The U.S.–FSM and U.S.–RMI trust fund agreements allow for the agreements to be amended in writing at any time, with mutual consent of the governments. However, the U.S. legislation implementing the amended compacts requires that any amendment, change, or termination of all, or any part, of the compact trust fund agreements shall not enter into force until incorporated into an act of Congress. Compact Trust Fund Structure The compact trust fund agreements state that no funds, other than specified trust fund administrative expenses, may be distributed from the funds before October 1, 2023. From fiscal year 2024 onward, the maximum allowed disbursement from each compact trust fund is the amount of the fiscal year 2023 annual grant assistance, as defined by the trust fund agreement, with full inflation adjustment. In addition, the trust fund committees may approve additional amounts for special needs. The RMI compact trust fund agreement excludes Kwajalein-related assistance, defined in section 211(b) of the RMI compact, from the calculation of the allowed disbursement. Although the compact trust fund agreements state the maximum allowable disbursement level, they do not establish or guarantee a minimum disbursement level. Each country’s compact trust fund consists of three interrelated accounts: the “A” account, the “B” account, and the “C” account. The A account is the trust fund’s corpus and contains the initial, and any additional, U.S. and FSM or RMI contributions; contributions from other countries; and investment earnings. No funds, other than specified trust fund administrative expenses, may be disbursed from the A account. The B account is the trust fund’s disbursement account and becomes active in fiscal year 2023. All income earned in 2023 will be deposited in the B account for possible disbursement in 2024. Each subsequent year’s investment income will similarly be deposited in the B account for possible disbursement the following year. If there is no investment income, no funds will be deposited in the B account for possible disbursement the following year. The C account is the trust fund’s buffer account. Through 2022, any annual income exceeding 6 percent of the fund balance is deposited in the C account. The size of the C account is capped at three times the amount of the estimated annual grant assistance in 2023, including estimated inflation. From 2023 onward, if annual income from the A account is less than the previous year’s disbursement, adjusted for inflation, the C account may be tapped to address the shortfall. After 2023, any funds in the B account in excess of the amount approved for disbursement the following fiscal year are to be used to replenish the C account as needed, up to the maximum size of the account. If there are no funds in the C account and no prior-year investment income in the B account, no funds will be available for disbursement to the countries the following year. Figure 2 shows the compact trust fund account structure and associated rules. According to the U.S. trust fund agreements with the FSM and the RMI, contributions from other donors are permitted. In May 2005, Taiwan and the RMI reached an agreement that Taiwan would contribute a total of $40 million to the RMI’s compact trust fund A account between 2004 and 2023. A “D” account may also be established to hold any contributions by the FSM and the RMI governments of revenue or income from unanticipated sources. According to the trust fund agreements, the D account must be a separate account, not mixed with the rest of the trust fund. Only the RMI has a D account, governed in part by the agreement between Taiwan and the RMI. Programs and Services Provided in Compact-Related Agreements The amended compacts’ implementing legislation incorporates, by reference, related agreements extending programs and services to the FSM and RMI. The programs and services agreement with each country identifies the following programs and services as being available to each country: U.S. postal services, weather services, civil aviation, disaster preparedness and response, and telecommunications. Each programs and services agreement extends for 20 years from the compact’s entry into force. The agreement with the FSM ends on June 24, 2024, and the agreement with the RMI ends on April 30, 2024. Programs Authorized by U.S. Legislation The amended compacts’ implementing legislation (Pub. L. No. 108-188) and other U.S. legislation authorize other U.S. grants, programs, and services for the FSM and RMI. Pub. L. No. 108-188 authorized an annual supplemental education grant (SEG) for the FSM and RMI in fiscal years 2005 through 2023, to be awarded in place of grants formerly awarded to the countries under several U.S. education, health, and labor programs. The FSM and RMI are not eligible for the programs replaced by the SEG during these years. Unlike the compact sector grants, the amended compacts’ implementing legislation authorized the SEG but did not appropriate funds for it. Funding for the SEG is appropriated annually to the U.S. Department of Education (Education) and is transferred to Interior for disbursement. Other provisions of the amended compacts’ implementing legislation, as well as other U.S. law, make the FSM and RMI eligible for a number of additional programs. The FSM and RMI Rely on U.S. Grants and Programs That End in 2023 As of fiscal year 2016, compact sector grants and the SEG, each of which end in 2023, supported a substantial portion of government expenditures in the FSM and RMI. Compact sector grants and the SEG supported about one-third of all FSM government expenditures. The four FSM states relied on these grants to a greater extent than the FSM national government does. In the RMI, compact sector grants and the SEG supported about one-quarter of all government expenditures. The end of the compacts’ programs and services agreements in 2024 would also require the FSM and RMI to bear additional costs to provide services currently provided by the United States under the agreements. U.S. Compact Grants and Other Grants Provide Substantial Support to the FSM and RMI Budgets U.S. Grants Scheduled to End in 2023 Supported About One- Third of Total FSM Government Expenditures in Fiscal Year 2016 Compact sector grants, the SEG, and other U.S. grants supported almost half of FSM national and state government expenditures in fiscal year 2016. Compact sector and supplemental education grants that end in 2023 supported approximately one-third of total FSM national and state government expenditures in fiscal year 2016, while other U.S. grants supported an additional 15 percent of total FSM government expenditures (see fig. 3). While the supplemental education grant ends in 2023, the FSM would be eligible for some of the programs that the supplemental education grant replaced after 2023. A small number of other U.S. grants also end in 2023. See GAO-19-648T, app. I, for a discussion of grants and programs that do and do not end in 2023. FSM States Relied on U.S. Grants Scheduled to End in 2023 to a Greater Extent Than the National Government Did in Fiscal Year 2016 In fiscal year 2016, compact sector and supplemental education grants that end in 2023 supported a larger proportion of FSM state governments’ expenditures than of the FSM national government’s expenditures. Compact sector grants and the SEG supported 8 percent of national government expenditures but supported 50 percent or more of each state’s government expenditures. Among the FSM states, Chuuk, which has both the largest population and the lowest per-capita income in the FSM, had the highest percentage of expenditures supported by U.S. grants. (See table 2 for a summary of FSM national and state government expenditures supported by compact sector grants and the SEG and by other U.S. grants.) U.S. Grants Scheduled to End in 2023 Supported About One Quarter of RMI Government Expenditures in Fiscal Year 2016 Compact sector and supplemental education grants that end in 2023 supported approximately 25 percent of the RMI’s $123.5 million in government expenditures in fiscal year 2016, while other U.S. grants supported an additional 8 percent. Kwajalein-related compact grants that do not end in 2023 supported an additional 3 percent (see fig. 4). While the supplemental education grant ends in 2023, the RMI would be eligible for some of the programs that the supplemental education grant replaced after 2023. A small number of other U.S. grants also end in 2023. See GAO-19-648T, app. I, for a discussion of grants and programs that do and do not end in 2023. FSM and RMI Eligibility for Some U.S. Grants, Programs, and Services Will Change after 2023 FSM and RMI budgets would be further affected if the countries assumed responsibility for providing programs and services currently provided by the United States. The following describes the status after 2023 of U.S. grants, programs, and services in the FSM and RMI under current law: Compact sector grants are scheduled to end in 2023, but the RMI MUORA extends the time frame of Kwajalein-related compact grants for as long as the MUORA is in effect. The SEG and additional grants identified in the amended compacts’ implementing legislation are scheduled to end in 2023. Also, after fiscal year 2023, the FSM and RMI will no longer be eligible for some programs that the SEG replaced, including Head Start (early childhood education, health, and nutrition services for low-income children and their families). The compact-related programs and services agreements with each country will end in 2024. However, some U.S. agencies, such as the National Weather Service, Federal Aviation Administration, and U.S. Agency for International Development, may continue to provide programs and services similar to those provided in the agreement under other authorities. The FSM and RMI will generally remain eligible for other programs identified in the amended compacts’ implementing legislation. These programs include U.S. Department of Agriculture (USDA) Rural Utilities Service grant and loan programs and U.S. Department of Education Pell grants for higher education and grants under Part B of the Individuals with Disabilities Education Act for children with disabilities. The FSM and RMI will remain eligible for additional programs we identified that have been provided under other current U.S. laws. Examples of these programs include USDA housing assistance programs and multiple public health, medical, and disease control and prevention grants provided by the U.S. Department of Health and Human Services. See appendix I for more information about the status after 2023 of U.S. grants, programs, and services in the FSM and RMI under current law. Compact Trust Funds Face Risks to Future Disbursements Our May 2018 projections for the compact trust funds showed that after fiscal year 2023, the funds are unlikely to provide maximum annual disbursements and may provide no disbursements at all in some years. The risk of disbursements below the maximum and the risk of zero disbursements increase over time for both funds. Potential strategies we analyzed in our May 2018 report would reduce or eliminate the risk of the compact trust funds’ experiencing years of zero disbursement. However, all of the potential strategies would require the countries to exchange a near-term reduction in resources for more-predictable and more- sustainable disbursements in the longer term. Projections Show Risks to Compact Trust Fund Disbursements Our May 2018 projections for the FSM and RMI compact trust funds after 2023 indicated that, given their balance at the end of fiscal year 2017 and current compact trust fund rules—the baseline scenario—the funds will be unable to provide maximum disbursements (equal to the inflation- adjusted amount of annual grant assistance in 2023) in some years and unable to provide any disbursement at all in some years, with the likelihood of zero disbursement in a given year increasing over time. The compact trust funds’ C account—designed as a buffer to protect disbursements from the B account in years when the funds do not earn enough to fund the disbursement—could be exhausted by a series of years with low or negative annual returns. Since current rules do not allow disbursements from the compact trust fund corpus (the A account), exhaustion of the C account would result in zero disbursement in years when fund returns are zero or negative. Thus, there may be no funds available to disburse even if the funds’ A accounts have a balance. As a result of low or zero disbursements, the countries could face economic and fiscal shocks and significant challenges in planning programs and budgets. Since we published our May 2018 report, an additional year of compact trust fund performance data and updated estimates of future inflation have become available; however, the updated information does not alter the conclusions we presented in May 2018. The updated data and inflation estimates change our model’s assumptions about the current compact trust fund balance, size of future U.S. contributions to the FSM and RMI compact trust funds, annual grant assistance in fiscal year 2023, and C account balance—each of which are relevant variables for our analysis. However; the updated variables would result in only slight changes to our 2018 report’s projections of future compact trust fund performance presented in this testimony and do not alter our broader conclusions about future risks to the compact trust funds. FSM compact trust fund projections. In May 2018, our model projected that, given the baseline scenario and a 6 percent net return, the FSM compact trust fund will experience declining disbursements relative to the maximum allowable disbursements and an increasing chance of zero disbursements. (See app. I of GAO-18-415 for a full description of our methodology, and see app. V of GAO-18-415 for the baseline results with alternative net returns.) Projected disbursements. We projected that the FSM compact trust fund will, on average, be able to provide disbursements equal to 82 percent of the maximum allowable disbursement—the inflation- adjusted amount of 2023 annual grant assistance—in its first decade of disbursements. The likely average disbursement falls to 49 percent of the maximum in the next decade and falls further in subsequent decades. In addition, the amount available for disbursement may fluctuate substantially from year to year. Depending on the compact trust fund’s performance in the previous year, disbursements may be higher or lower than the average amount if the balance in the C account is not sufficient to provide additional disbursements. Likelihood of providing zero disbursement. We projected a 41 percent likelihood that the FSM compact trust fund will be unable to disburse any funds in 1 or more years during the first decade of trust fund disbursements. This likelihood increases over time, rising to 92 percent in fiscal years 2054 through 2063. Figure 5 shows our May 2018 projections of the FSM compact trust fund’s average disbursements as a percentage of maximum disbursement and the likelihood of 1 or more years of zero disbursement, given the baseline scenario and a 6 percent net return. RMI compact trust fund projections. In May 2018, our model projected that, given the baseline scenario and a 6 percent net return, the RMI compact trust fund will experience declining disbursements relative to the maximum allowable disbursements and an increasing chance of zero disbursements. Projected disbursements. We projected that in its first decade of disbursements, the RMI compact trust fund will, on average, be able to provide disbursements nearly equal to the inflation-adjusted amount of 2023 annual grant assistance as defined by the trust fund agreement—the maximum allowable. However, in each subsequent decade, the projected disbursements as a percentage of the maximum disbursements decline by about 10 percentage points. In addition, from year to year, the amount available to disburse may fluctuate substantially. Depending on the compact trust fund’s performance in the previous year, disbursements may be higher or lower than the average amount if the balance in the C account is not sufficient to provide additional disbursements. Likelihood of providing zero disbursement. We projected a 15 percent likelihood that the RMI compact trust fund will be unable to disburse any funds in 1 or more years during the first decade of trust fund disbursements. This likelihood increases over time, rising to 56 percent in fiscal years 2054 through 2063. Figure 6 shows our May 2018 projections of the RMI compact trust fund’s average disbursements as a percentage of maximum disbursement and its likelihood of 1 or more years of zero disbursement, given the baseline scenario and a 6 percent net return. For our May 2018 report, we conducted a series of simulations to determine the likely effects of potential strategies for improving the outlook of the FSM and RMI compact trust funds. For example, we developed and analyzed potential strategies in which: annual disbursements are reduced below the maximum allowable additional annual contributions are made to the trust fund prior to the end of fiscal year 2023, and the trust fund agreement disbursement policies are modified to limit the annual disbursement to a fixed percentage of the fund’s moving average balance over the previous 3 years, up to the maximum disbursement amount defined by the current trust fund agreement. All of the potential strategies we analyzed would reduce or eliminate the risk of the compact trust funds experiencing years of zero disbursement. However, some of the potential strategies may require changing the trust fund agreements and all of the potential strategies would require the countries to exchange a near-term reduction in resources for more- predictable and more-sustainable disbursements in the longer term. (See app. VII of our May 2018 report for detailed results of our analysis.) Compact Trust Fund Committees Have Not Addressed Issues Related to Distribution Policies, Fiscal Procedures, and Disbursement Timing The trust fund committees have not taken the actions we recommended in 2018 to prepare for the 2023 transition to trust fund income. The compact trust fund committees have not yet prepared distribution policies, required by the trust fund agreements, which could assist the countries in planning for the transition to trust fund income. In addition, the committees have not established fiscal procedures for oversight of compact trust fund disbursements as required by the trust fund agreements. Further, the committees have not yet addressed a potential misalignment between the timing of their annual calculation of the amounts available to disburse and the FSM’s and RMI’s budget timelines, potentially complicating each country’s planning and management. Trust Fund Committees Have Not Developed Distribution Policies Required by the Compact Trust Fund Agreements The compact trust fund committees have not yet developed, as the compact trust fund agreements require, policies to guide disbursements from the trust funds after fiscal year 2023. Under the agreements, each trust fund committee must develop a distribution policy, with the intent that compact trust fund disbursements will provide an annual source of revenue to the FSM and RMI after the scheduled end of compact grant assistance. The trust fund committees could use distribution policies to address risks to each fund’s sustainability. For example, the committees have the discretion to disburse an amount below the established maximum. Our analysis of potential strategies for improving the funds’ outlook shows that reducing the size of disbursements would improve each compact trust fund’s long-term sustainability. Without a distribution policy that provides information about the size of expected disbursements, the FSM and RMI are hampered in their current and ongoing efforts to plan for the potential reduction in U.S. compact assistance after 2023. Trust Fund Committees Have Not Established Fiscal Procedures Required by Compact Trust Fund Agreements The compact trust fund committees have not yet established fiscal procedures for compact trust fund disbursements after fiscal year 2023. Each trust fund agreement requires the respective committee to determine the fiscal procedures to be used in implementing the trust fund agreement. The committees are to base their procedures on the compact fiscal procedures agreements, unless the parties to the trust fund agreement agree to adopt different fiscal procedures. No compact trust fund disbursements are to be made unless the committee has established such trust fund fiscal procedures. Without fiscal procedures in place, the trust fund committees will not be able to provide disbursements and the United States, the FSM, and the RMI will lack clear guidance to ensure oversight for trust fund disbursements. Trust Fund Committees Have Not Addressed Issues Related to Disbursement Timing The timing for the trust fund committees’ calculation of the amounts available for annual disbursement to the FSM and the RMI after fiscal year 2023 does not align with the countries’ budget and planning timelines. The amounts available for disbursement in a given fiscal year cannot be determined until each fund’s returns have been determined at the end of the prior year. Further, if the disbursement amounts are calculated from audited fund returns as determined by annual audits required by the trust fund agreements, the amounts may not be determined until as late as March 31, 6 months into the fiscal year for which the disbursement is to be provided. However, both the FSM and the RMI government budget cycles are completed before the annual amounts available for disbursement will be known. As a result, the FSM and RMI would have to budget without knowing the amount to be disbursed, complicating their annual budget and planning processes. Trust Fund Committees Have Discussed Potential Actions to Address our Recommendations The trust fund committees, chaired by Interior, have discussed potential actions to address the recommendations in our May 2018 report. In May 2018, we made six recommendations to Interior—three parallel recommendations regarding each country’s trust fund. We recommended that the Secretary of the Interior ensure that the Director of the Office of Insular Affairs work with other members of the trust fund committees to: develop distribution policies, develop the fiscal procedures required by the compact trust fund address the timing of the calculation of compact trust fund disbursements. Interior concurred with our recommendations and has stated that it plans to implement them before the FSM and RMI transition to trust fund income in 2023. The FSM and RMI also concurred with our recommendations to Interior. According to the Trust Fund Administrator and Interior officials, the distribution policy was discussed at subsequent trust fund committee meetings—including the most recent, in May 2019. According to the trust fund administrator, trust fund representatives met with FSM and RMI representatives in January 2019 to discuss the status of the trust fund and future scenarios for its management. Interior officials further stated that discussions about trust fund policies and controls were frequent and ongoing among committee members and staffers as well as the trust fund manager and investment advisers. The FSM’s and RMI’s transition to relying on income from the compact trust funds will likely require significant budgetary choices. However, the lack of trust fund distribution policies, and the lack of alignment between the trust fund committees’ annual disbursement calculations and the countries’ budget cycles hamper the countries’ ability to plan for the transition. In addition, without the required fiscal procedures governing trust fund actions after 2023, the trust fund committees will be unable to make disbursements and the United States, the FSM, and the RMI will not have assurance of necessary oversight. However, as of June 2019, Interior had not implemented our recommendations to address these issues. Further, while Interior has continued to discuss possible actions to address our recommendations with the trust fund committees, it targeted implementation of our recommendations for 2023. Chairwoman Murkowski, Ranking Member Manchin, and Members of the Committee, this concludes my statement. I would be pleased to respond to any questions you may have. GAO Contact and Staff Acknowledgments If you or your staff have any questions about this testimony, please contact David Gootnick, Director, International Affairs and Trade, at (202) 512-3149 or gootnickd@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this testimony are Emil Friberg (Assistant Director), Ming Chen, Neil Doherty, Mark Dowling, Christopher Keblitis, Reid Lowe, Moon Parks, and Michael Simon. Appendix I: Status of U.S. Grants and Programs in the FSM and RMI After 2023 The amended compacts, compact–related agreements, the amended compacts’ implementing legislation, and other U.S. laws provide grants or eligibility for U.S. programs and services for the Federated States of Micronesia (FSM) and Republic of the Marshall Islands (RMI). The amended compacts provide compact sector, Kwajalein-related, and audit grants. Under current law, compact sector and audit grants are each scheduled to end in 2023, but the RMI military use and operating rights agreement (MUORA) extends the time frame of Kwajalein-related compact grants for as long as the agreement is in effect. The amended compacts’ implementing legislation provides additional grants, including authorizing a supplemental education grant (SEG), and identifies several specific U.S. programs as available to the FSM and RMI. Under current law, the additional grants end in 2023, but the statutory authorizations for some programs identified in Pub. L. No. 108-188 provide for the continued eligibility of the FSM and RMI to receive benefits under the programs. However, after fiscal year 2023, the FSM and RMI will no longer be eligible under current U.S. law for some programs that the SEG replaced. The compact-related programs and services agreements with each country identify additional programs and services that the United States makes available to the FSM and RMI. While these agreements will end in 2024, under current law, some U.S. agencies may continue to provide programs and services similar to those provided in the agreement under other authorities. Based on the status of current law, the FSM’s and RMI’s eligibility for other programs we identified that have been provided under other current U.S. laws will not change after fiscal year 2023. Compact Sector and Audit Grants End in 2023, but Kwajalein-Related Grants for the RMI Will Continue Under current law, compact sector grants provided to the FSM and the RMI under section 211(a) of the amended compacts are scheduled to end in 2023. However, the RMI is scheduled to continue to receive $7.2 million, partially inflation adjusted, related to the U.S. military base in Kwajalein Atoll and provided under section 211(b) of its compact. Under the terms of the RMI MUORA, the United States agreed to provide these Kwajalein-related grants for as long as the MUORA is in effect. The MUORA continues until 2066 and may be extended at the discretion of the United States until 2086. The amended RMI compact provides for $18 million, partially inflation adjusted, in annual payments to the RMI government to compensate for impacts from the U.S. Army Garrison– Kwajalein Atoll. These payments will continue for as long as the MUORA is in effect. Annual compact grants of up to $500,000 (not inflation adjusted) to each country to pay for required annual audits of compact grants are scheduled to end in 2023. See table 3 for a summary of compact sector, Kwajalein-related, and audit grants. FSM and RMI Are No Longer Eligible for Many Programs Replaced by the Supplemental Education Grant The supplemental education grant (SEG) authorized by the amended compacts’ implementing legislation is scheduled to end in fiscal year 2023 and, under current law, FSM and RMI eligibility for most programs that the SEG replaced will not resume after fiscal year 2023. Absent changes to current law, the FSM and RMI will not be eligible after fiscal year 2023 for the following programs that the SEG replaced during fiscal years 2005 through 2023: U.S. elementary and secondary education grant programs, adult education and literacy programs, career and technical education programs, job training programs, and Head Start early education programs. However, under other provisions of current law, qualifying individuals in the FSM and RMI will be eligible after fiscal year 2023 for undergraduate education grants and work-study programs that the SEG replaced. See table 4. Some Programs and Services in the Programs and Services Agreement Will End, while Others May Continue under Other Authorities Although the programs and services agreements with the FSM and RMI will end in fiscal year 2024, current U.S. law enables U.S. agencies to continue providing some programs and services now provided under the agreements. After the agreements end, no current provisions of U.S. law will enable the Federal Emergency Management Agency (FEMA) to provide disaster response funding, enable the Federal Deposit Insurance Corporation to provide deposit insurance, or enable the U.S. Postal Service to provide the services that it currently provides to the FSM and RMI. However, the National Weather Service, the U.S. Department of Transportation’s (DOT) Federal Aviation Administration (FAA), and the U.S. Agency for International Development (USAID) could, under other legal authorities, provide services similar to those they now provide under the programs and services agreements. National Weather Service. The programs and services agreements authorize the National Weather Service to fund the operations of weather stations in the FSM and RMI, which it can continue to fund after the end of the agreements under other authorities, according to Department of Commerce officials. Federal Aviation Administration. The programs and services agreements authorize DOT’s FAA to provide technical assistance in the FSM and RMI, which it can continue to provide after the end of the agreements under other provisions of current U.S. law. However, DOT officials stated that FAA would require new bilateral agreements with the FSM and the RMI in order for the countries to continue to receive the civil aviation safety services that FAA currently provides under the programs and services agreements. The FAA would also seek reimbursement for any technical assistance it provides to the FSM and RMI. With regard to the civil aviation economic services provided under the programs and services agreements, DOT officials stated that, while the FSM and RMI could voluntarily decide to allow U.S. air carriers to continue operations in the FSM and RMI, new bilateral agreements would be needed to assure that result. U.S. Agency for International Development. Following a U.S. presidential disaster declaration, FEMA provides the funding for disaster relief and reconstruction, which is programmed through USAID. Under current law, FEMA funds will no longer be available to the FSM and RMI for this purpose once the agreements end; however, USAID will be able to provide foreign disaster assistance funding to the two countries under the same terms as it provides this assistance to other countries. After the programs and services agreements end, FEMA will be able to support disaster relief efforts only if USAID or the countries request such support on a reimbursable basis. In addition, according to State and Interior officials, telecommunications- related services that the two agencies provide to the FSM and RMI under the programs and services agreements will continue as long as the FSM and RMI provide appropriate authorization for such services. Table 5 shows the status after the programs and services agreements end of programs and services currently provided to the FSM and the RMI under the agreements. Programs Identified in Amended Compacts’ Implementing Legislation Generally Continue after Fiscal Year 2023 Although additional grants provided to the FSM and the RMI under the amended compacts’ implementing legislation will end in fiscal year 2023, the countries’ eligibility for programs now provided under that legislation will generally continue under current U.S. law. Grants provided under the amended compacts’ implementing legislation for (1) judicial training in the FSM and the RMI, and (2) agricultural and planting programs on the RMI’s nuclear-affected Enewetak Atoll are scheduled to end. However, under current U.S. law, legal authorities permitting the operation of other programs will remain available to the FSM and RMI after fiscal year 2023. Eligibility under these legal authorities continues either because the amended compacts’ implementing legislation does not specify an ending date or because other provisions in current U.S. law make the FSM and RMI eligible for the program. Programs provided in the amended compacts’ implementing legislation include U.S. Department of Agriculture Rural Utilities Service grant and loan programs; U.S. Department of Education Pell grants for higher education and grants under Part B of the Individuals with Disabilities Education Act for children with disabilities; programs for nuclear-affected areas in the RMI; and additional programs provided by the Departments of Commerce and Labor as well as law enforcement assistance provided by the U.S. Postal Service. See table 6 for a summary of the programs identified in the amended compacts’ implementing legislation and their status as of the end of fiscal year 2023. Programs Identified in Other Legislation Generally Continue after Fiscal Year 2023 In addition to being eligible for the programs provided through the compact, its associated agreements, and the amended compacts’ implementing legislation, the FSM and RMI are also eligible for a number of programs under other provisions of current U.S. law. The FSM and RMI have each received funds from the U.S. Department of Agriculture for forestry and rural housing programs, multiple U.S. Department of Health and Human Services public health program grants, U.S. Department of the Interior technical assistance and historic preservation programs, and the DOT FAA airport improvement program, among others. Under current U.S. law, the legal authorities permitting the provision of these programs in the FSM and RMI would not necessarily change after 2023. Table 7 shows the FSM’s and RMI’s eligibility for these additional grants and programs under current law after fiscal year 2023.

Why GAO Did This Study In 2003, the United States approved amended compacts of free association with the FSM and RMI, providing a total of $3.6 billion in economic assistance in fiscal years 2004 through 2023 and access to several U.S. programs and services. Compact grant funding, overseen by the Department of the Interior (Interior), generally decreases annually. However, the amount of the annual decrease in grants is added to the annual U.S. contributions to the compact trust funds, managed by joint U.S.-FSM and U.S.-RMI trust fund committees and chaired by Interior. Trust fund earnings are intended to provide a source of income after compact grants end in 2023. This testimony summarizes GAO's May 2018 report on compact grants and trust funds ( GAO-18-415 ). In that report, GAO examined (1) the use and role of U.S. funds and programs in the FSM and RMI budgets, (2) projected compact trust fund disbursements, and (3) trust fund committee actions needed to address the 2023 transition to trust fund income. For this testimony, GAO also reviewed key variables for its trust fund model as of June 2019 to determine whether these variables had substantially changed. In addition, GAO reviewed the status of Interior's response to GAO's May 2018 recommendations. What GAO Found The Federated States of Micronesia (FSM) and the Republic of the Marshall Islands (RMI) rely on U.S. grants and programs, including several that are scheduled to end in 2023. In fiscal year 2016, U.S. compact sector grants and supplemental education grants, both scheduled to end in 2023, supported a third of the FSM's expenditures and a quarter of the RMI's. Agreements providing U.S. aviation, disaster relief, postal, weather, and other programs and services are scheduled to end in 2024, but some U.S. agencies may provide programs and services similar to those in the agreements under other authorities. GAO's 2018 report noted that the FSM and RMI compact trust funds face risks and may not provide disbursements in some future years. GAO projected a 41 percent likelihood that the FSM compact trust fund would be unable to provide any disbursement in 1 or more years in fiscal years 2024 through 2033, with the likelihood increasing to 92 percent in 2054 through 2063. GAO projected a 15 percent likelihood that the RMI compact trust fund would be unable to provide any disbursement in 1 or more years in fiscal years 2024 through 2033, with the likelihood increasing to 56 percent in 2054 through 2063. Potential strategies such as reduced trust fund disbursements would reduce or eliminate the risk of years with no disbursement. However, some of these strategies would require changing the trust fund agreements, and all of the strategies would require the countries to exchange a near-term reduction in resources for more-predictable and more-sustainable disbursements in the longer term. Interior has not yet implemented the actions GAO recommended to prepare for the 2023 transition to trust fund income. The trust fund committees have not developed distribution policies, required by the agreements, which could assist the countries in planning for the transition to trust fund income. The committees have not developed the required fiscal procedures for oversight of disbursements or addressed differences between the timing of their annual determinations of the disbursement amounts and the FSM's and RMI's annual budget cycles. What GAO Recommends In its May 2018 report, GAO made three recommendations to Interior regarding each country's trust fund to address trust fund disbursement risks. Interior concurred with GAO's recommendations and discussed actions in response at subsequent trust fund committee meetings, with implementation targeted for 2023.

gao_GAO-19-720T

gao_GAO-19-720T_0

Background At VA, and indeed at all federal agencies, strategic human capital management plays a critical role in maximizing the government’s performance and assuring its accountability to Congress and the nation as a whole. As we have long reported, there is a direct link between the effectiveness of an agency’s personnel management efforts and its ability to carry out its mission. Addressing challenges in areas such as disaster response, homeland security, economic stability, and numerous other complex and evolving issues requires a skilled federal workforce able to work seamlessly with other agencies, levels of government, and nongovernmental entities. In our March 2019 report, we identified key trends in agency operations and attitudes toward work that are affecting how federal work is done and, consequently the skills and competencies that workers will need to accomplish agency missions. Agencies will need to apply appropriate talent management strategies that are adapted to these trends to recruit, develop, and retain a high-performing workforce and better meet their missions (see fig. 1). Staffing Challenges at VA Are Systemic, Long-Standing, and Undermining Its Mission Over the past two decades, we and others have expressed concern about certain human capital practices at VA and its components. For example, in November 2018, VA’s Office of Inspector General identified leadership and workforce investment as a major management challenge. The Inspector General noted that the root cause for many of the issues it identified at VA was poor and unstable leadership as well as staffing shortages. Similarly, in May 2019, we reported that leadership turnover impeded VA’s ability to address a number of management challenges we identified such as managing acquisitions, managing risk, and improving veterans’ health care. At VHA, we found that serious human capital shortfalls are undermining its ability to meet the health care needs of veterans. Key examples from our prior work include the following: In March 2019, we reported that VHA’s 172 medical centers have large staffing shortages, including physicians, registered nurses, physician assistants, psychologists, and physical therapists, as well as HR specialists and assistants. As of December 2018, VA reported an overall vacancy rate of 11 percent at VHA medical facilities, including vacancies of over 24,000 medical and dental positions and around 900 HR positions. In July 2016, we found that losses in VHA’s five clinical occupations with the largest staffing shortages, including physicians, registered nurses, and psychologists, increased from about 5,900 employees in fiscal year 2011 to about 7,700 in fiscal year 2015. Voluntary resignations and retirements were the primary drivers. VHA’s exit survey indicated that advancement issues or dissatisfaction with certain aspects of the work, such as concerns about management and obstacles to getting the work done, were commonly cited as the primary reasons people left. In December 2016, we found that several problems combined to impede VHA’s ability to improve delivery of health care services to veterans. These problems include high attrition (often involving transfers to other federal agencies), increased workload, and burnout among VHA’s HR staff. Another issue is a lack of effective internal control practices to support HR operations such as information systems that meet operational needs (see fig. 2). In our preliminary findings in a forthcoming report on the extent to which succession planning policies and procedures at VA and its components are consistent with key leading practices, we have identified several concerns. For example, according to VA officials, the agency has not produced a department-wide succession plan since 2009 due to leadership turnover. Department-wide, around 30 percent of VA employees on board as of September 30, 2017, will become eligible to retire in the next 5 years. Effective succession planning can help VA ensure it has a pipeline of talent to meet current and future mission requirements. In our prior work, we noted that effective succession planning is more than filling existing vacancies with people that have the same occupational skills and competencies. Rather, succession planning focuses on current and future needs and develops pools of high-potential staff to meet the organization’s mission over the long term. Continued Leadership Attention to Implementing Our Prior Recommendations and Other Talent Management Strategies Could Help VA Better Serve Veterans We have designated 40 of our prior recommendations to VA as priority recommendations because, upon implementation, they may have an especially significant impact on VA’s operations. Twelve of these priority recommendations are aimed at strengthening VA’s human capital management efforts and will help address VA’s challenges in such areas as recruiting and retaining doctors and nurses, performance management, and employee misconduct. To date, VA has implemented six of these priority recommendations, but needs to take additional action on the other six. While VA agreed or partially agreed with and is taking steps to implement five of these remaining priority recommendations, it disagreed with one related to developing a process to accurately count all physicians at each VA medical center because it does not believe this affects its ability to assess workload. Nevertheless, we continue to believe that VHA needs a systematic process to identify all physicians working at VA medical centers as part of the agency’s efforts to monitor and assess workload. The six unimplemented priority recommendations are for VA to 1. develop a process to accurately count all physicians providing care at each VA medical center (recommended in 2017), 2. develop a modern and effective performance management system in which VA managers make meaningful distinctions in employees’ performance ratings (recommended in 2016), 3. ensure that ratings-based performance awards are administered in a manner that is consistent with leading practices (recommended in 2016), 4. develop a plan to implement a modern information technology system to support employee performance management processes (recommended in 2016), 5. collect complete and reliable misconduct and associated disciplinary action data (recommended in 2018), and 6. ensure that employees who report wrongdoing are treated fairly and protected against retaliation (recommended in 2018). We will continue to monitor VA’s progress in implementing these and our other open recommendations. Beyond these specific recommendations, VA and other agencies can use talent management strategies to better compete for critical positions in a tight labor market and to help meet agency missions. In our prior work we noted that while these strategies are not an exhaustive list, collectively they suggest basic steps that agencies can take within existing authorities to address the demographic and technological trends affecting work that are discussed earlier in this statement. These strategies include: Align human capital strategy with current and future mission requirements. With shifting attitudes toward work, technological advances, and increased reliance on nonfederal partners, agencies need to identify the knowledge and skills necessary to respond to current and future demands. Key practices include identifying and assessing existing skills, competencies, and skills gaps. Acquire and assign talent. To ensure agencies have the talent capacity to address evolving mission requirements and negative perceptions of federal work (e.g., that it is too bureaucratic), agencies can cultivate a diverse talent pipeline, highlight their respective missions, recruit early in the school year, support rotations, and assign talent where needed. Incentivize and compensate employees. While federal agencies may struggle to offer competitive pay in certain labor markets, they can leverage existing incentives that appeal to workers’ desire for schedules and locations that provide work-life balance. Engage employees. Engaged employees are more productive and less likely to leave. Agencies can better ensure their workforces are engaged by managing employee performance, involving employees in decisions, and developing employees. Strategic Human Capital Management Is at Risk Government-wide and Is Impacting Agencies’ Missions A number of the staffing challenges facing VA are actually part of a broader set of human capital issues affecting government as a whole. As we noted in our March 2019 update of government high-risk areas, the federal government faces long-standing challenges in strategically managing its workforce. We first added strategic human capital management to our list of high-risk government programs and operations in 2001. Although Congress, OPM, and individual agencies have made improvements since then, strategic human capital management remains a high-risk area because mission-critical skills gaps within the federal workforce pose a high risk to the nation. Of the 34 other high-risk areas on our 2019 High-Risk List, skills gaps played a significant role in 16 of the areas, including information technology management and acquisitions, strengthening management functions at the Department of Homeland Security, and, as noted above, veterans’ health care at VA. While causes for these skills gaps related to high-risk areas vary, they often occur because of a shortfall in talent management activities such as robust workforce planning or training. Additionally, the changing nature of federal work and the high percentage of employees eligible for retirement have the potential to produce gaps in leadership and institutional knowledge and could threaten to aggravate the problems created from existing skills gaps. For example, 31.6 percent of permanent federal employees who were on board as of September 30, 2017 will be eligible to retire in the next 5 years, with some agencies having particularly high levels of employees eligible to retire. High-performing organizations have found that the full life cycle of human capital management activities needs to be fully aligned and focused on the cost-effective achievement of an organization’s mission. These activities include workforce planning, recruitment, on-boarding, compensation, engagement, succession planning, and retirement programs. Further, adding to agencies’ staffing challenges is the fact that much has changed since the Civil Service Reform Act of 1978 and the Classification Act of 1949 laid the foundation of much of today’s federal personnel system. Agencies’ missions have evolved and employees’ expectations of work and the workplace are changing. As a result, the extent to which the current and future workforce finds the government’s employment policies and practices relevant is an open question. We and others have identified several structural challenges within the federal human capital system that impede the ability of agencies to recruit, retain, and develop workers, both today and in the future. For example: Classification system. The General Schedule classification system—which defines and organizes federal positions primarily to assign rates of pay—has not kept pace with the government’s evolving requirements. Recruiting and hiring. Federal agencies need a hiring process that is applicant friendly and flexible while also meeting policy requirements. Pay system. Employees are compensated through an outmoded system that (1) rewards length of service rather than individual performance and contributions, and (2) automatically provides across- the-board annual pay increases, even for poor performance. Performance management. Federal agencies have faced long- standing challenges developing modern, credible, and effective employee performance management systems and dealing with poor performers. Going forward, to help agencies effectively carry out their missions, OPM and federal agencies must take some important steps to address ongoing human capital problems. These actions include continuing to develop the capacity to measure and address existing mission-critical skills gaps and using workforce analytics to predict and mitigate future gaps. Chairman Takano, Ranking Member Roe, and Members of the Committee, this completes my prepared statement. I would be pleased to respond to any questions you may have at this time. If you or your staff have any questions about this testimony, please contact Robert Goldenkoff, Director, Strategic Issues, at (202) 512-2757 or GoldenkoffR@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this testimony are Shirley Hwang (Assistant Director), Alexander Ray (Analyst-In-Charge), Sarah Green, Allison Gunn, and Shelby Kain. 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.

Why GAO Did This Study VA operates one of the largest health care delivery systems in the nation and provides billions of dollars in benefits and services to veterans and their families. However, VA faces serious and long-standing problems with management challenges and veterans' access to health care and disability benefits. These issues contributed to GAO's decision to list several areas involving VA on GAO's High-Risk List, including managing acquisitions, managing risk and improving veterans' health care, and improving and modernizing VA's disability programs. This testimony discusses (1) human capital challenges facing VA and its components, (2) GAO recommendations addressing some of those challenges, and (3) how those challenges are related to a broader set of government-wide human capital problems. This testimony is based on GAO's work on VA issued since 2017, as well as GAO's work on government-wide strategic human capital management issued since July 2014. To conduct these studies, GAO reviewed key agency documents and government-wide employment data and interviewed knowledgeable agency officials and managers, as well as subject matter specialists. What GAO Found Serious human capital shortfalls are undermining the Department of Veterans Affairs' (VA) ability to provide veterans with quality and timely services. Over the past two decades, GAO has identified major challenges with VA human capital practices. For example, in March 2019, GAO found large staffing shortages, including physicians and registered nurses, at the Veterans Health Administration's (VHA) 172 medical centers. In December 2016, GAO found that high attrition, increased workload, and burnout among VHA's human resources (HR) staff, along with ineffective internal controls to support its HR operations, have impeded VHA's ability to serve the nation's veterans (see figure). Continued leadership attention to addressing GAO's recommendations could help VA better execute its mission. GAO has made numerous recommendations to VA, 40 of which were designated as priorities because they could significantly improve VA's operations. Twelve of the 40 were aimed at strengthening VA's human capital management efforts. Of these, six have been addressed. However, VA still needs to take additional actions on the other six, such as developing a modern and effective performance management system. Beyond these priority recommendations, VA can use key talent management strategies that GAO has identified for acquiring, incentivizing, and engaging employees and thus be more competitive for a high-performing workforce in a tight labor market. Some of the challenges facing VA are part of a larger set of human capital issues affecting government as a whole. Although Congress, the Office of Personnel Management, and individual agencies have made improvements in recent years, human capital management in general remains a high-risk area because of mission-critical skills gaps within the federal workforce. Structural issues impede the ability of agencies to recruit, retain, and develop workers, including outmoded position classification and pay systems, ineffective recruiting and hiring processes, and challenges in dealing with poor performers. What GAO Recommends GAO has designated 40 of its prior recommendations to VA as priorities for implementation. Twelve of these priority recommendations are aimed at strengthening VA's human capital management efforts. To date, VA has implemented six of these priority recommendations, but needs to take additional action on the other six. GAO will continue to monitor VA's progress in implementing these recommendations.

gao_GAO-20-404

gao_GAO-20-404_0

Background TSA and Industry Roles in Securing Passenger Rail The Aviation and Transportation Security Act designated TSA as the primary federal agency responsible for security in all modes of transportation, which includes physical security and cybersecurity. Passenger rail operators, however, have the day to day responsibility for carrying out safety and security measures for their systems. Unlike the aviation environment, where TSA has operational responsibility for screening passengers and baggage for prohibited items prior to boarding a commercial aircraft, the agency has a limited operational role for securing mass transit (including rail). To secure passenger rail, TSA primarily partners with public and private transportation operators to address their security needs by conducting vulnerability assessments and sharing intelligence information and key practices, among other measures. The agency also engages with the passenger rail industry through associations, such as APTA and Association of American Railroads. Additionally, TSA is responsible for assessing the risk from terrorism and cyber threats to passenger rail, as well as other transportation modes. In addition to engaging with domestic passenger rail stakeholders, TSA’s Office of Policy, Plans, and Engagement is responsible for coordinating domestic and international multimodal transportation security polices, programs, directives, strategies, and initiatives, including conducting outreach to foreign stakeholders. TSA also engages with foreign stakeholders through TSARs. TSARs are primarily located in posts overseas and communicate with foreign government officials to address transportation security matters involving all modes of transportation, including aviation, rail, mass transit, highways, and pipelines. The TSAR role was originally created in response to the 1988 bombing of Pan Am Flight 103 over Lockerbie, Scotland, when the Aviation Security Improvement Act of 1990 was enacted, which provided that foreign security liaison officers were to serve as liaisons to foreign governments in carrying out U.S. government security requirements at specific airports. TSARs are responsible for ensuring the implementation of TSA’s requirements primarily as they relate to passenger and cargo air transportation departing the specific country en route to the United States. The primary focus of the role remains on aviation; however it has evolved over time to include maritime and land transportation. Physical and Cybersecurity Threats to Passenger Rail According to TSA, recent attacks overseas and online terrorist messaging point to public transportation systems, which include passenger rail systems, as continued high-value targets for terrorists. In general, passenger rail systems are open and designed to expedite the free flowing movement of large numbers of passengers through multiple stations. As such, these systems are inherently vulnerable to physical attacks (such as improvised explosive devices, active shooters, and chemical or biological attacks) due in part to factors such as high ridership, open access points, limited exit lanes, and fixed, publically available schedules. In addition, TSA has reported that risks increase in urban areas where multiple transportation systems and high volumes of travelers merge at intermodal stations. Transportation systems, including passenger rail systems, rely on technology and internet-connected devices to manage and secure certain business/enterprise functions, such as the operation’s website, communications, and reservations and ticketing mechanisms. They also increasingly rely on computer-networked systems for tracking, signals, and operational controls of transportation equipment and services. As dependence on these systems increases, so does risk to the system. Cyberattacks have the potential to significantly affect both business/enterprise systems and operational control systems. Business/Enterprise systems. Cybersecurity threats include ransomware, malware, phishing, and website attacks that may compromise sensitive information and affect an operator’s ability to communicate with passengers or engage in day-to-day business functions. Operational control systems. Cybersecurity threats, which may include malware or physical manipulation of a system, such as jamming signals or damaging equipment, include threats to the systems that control signaling and train speed. For example, attackers could attempt to access positive train control systems, a computer-based system designed to automatically slow or stop a train that is not being operated safely, to disrupt services. Unintentional cybersecurity threat sources may include failures in equipment or software due to aging or user errors, such as unintentionally inserting a flash drive infected with malware or clicking on a phishing email. Intentional cybersecurity threats may include corrupt employees, criminal groups, terrorists, and nations and may be used, for example, to achieve monetary gain, or for political or military purposes. Figure 1 shows examples of the types of physical and cyber threats passenger rail systems face. DHS Risk Management Framework The NIPP outlines a risk management framework for critical infrastructure protection. In accordance with the Homeland Security Act of 2002, as amended, DHS created the NIPP in 2006 to guide the national effort to manage security risk to the nation’s critical infrastructure, including through coordination of agencies and 16 various critical infrastructure sectors, including transportation systems. Most recently updated in 2013, the NIPP uses a risk management framework as a planning methodology intended to inform how decision makers take actions to manage risk. The framework calls for public and private partners to conduct risk assessments. The NIPP defines risk as a function of three elements: threat, vulnerability, and consequence, as shown in Figure 2. Threat is an indication of the likelihood that a specific type of attack will be initiated against a specific target or class of targets. Vulnerability is the probability that a particular attempted attack will succeed against a particular target or class of targets. Consequence is the effect of a successful attack. In 2010, DHS, through TSA and the U.S. Coast Guard, developed the Transportation Systems Sector-Specific Plan to conform to requirements in the NIPP. Most recently updated in 2015, this plan describes shared goals, priorities, and activities to mitigate critical infrastructure risks, and acknowledges the increasing dependence of transportation companies on cyber systems for business, security, and operational functions. Regarding cybersecurity risks, DHS produced the Cybersecurity Strategy in 2018 to help execute its cybersecurity responsibilities during the next 5 years. In order to meet one of its objectives, DHS is to encourage the adoption of applicable cybersecurity best practices, including the NIST Framework for Improving Critical Infrastructure Cybersecurity (referred to as the NIST Cybersecurity Framework). The framework is a set of voluntary industry standards and best practices to help organizations manage security risks specific to cybersecurity. The framework consists of five functions: Identify, Protect, Detect, Respond, and Recover. When considered together, these functions provide a high-level view of an organization’s management of cybersecurity risk. NIST issued the framework in 2014 and updated it in April 2018. CISA, formerly DHS’s National Protection and Programs Directorate, manages the national effort to secure and protect against critical infrastructure risks, including cybersecurity risk, for all 16 critical infrastructure sectors, including transportation. CISA’s responsibilities include coordinating with sector-specific agencies to carry out its cybersecurity and critical infrastructure activities. TSA Conducts Passenger Rail Risk Assessments and Coordinates with CISA on Cybersecurity Risk TSA Uses Three Mechanisms to Assess Passenger Rail Risk According to TSA officials, TSA uses the TSSRA, the BASE, and threat assessments to assess risk elements for physical and cyber security in passenger rail. Such assessments may address different elements of risk—threat, vulnerability, or consequence—or the total risk for specific assets, such as airport perimeters and pipeline critical facilities. Table 1 below shows the type of risk element each assessment addresses, and whether the assessment addresses risks to intermodal stations or cybersecurity risk. TSSRA. TSA uses the TSSRA, a periodic risk assessment, to assess threat, vulnerability, and consequence for various attack scenarios across the five transportation modes for which TSA is responsible. The scenarios define a type of threat actor—including homegrown violent extremists and transnational extremists, such as Al Qaeda and its affiliates—a target, and an attack mode. For example, a scenario might assess the risk of attacks using varying types of weapons on passenger rail system assets. As part of the assessment process, TSA engages with subject matter experts from TSA and industry stakeholder representatives to compile vulnerabilities for each mode, and TSA analyzes both direct and indirect consequences of the various attack scenarios. According to TSA, the agency uses the TSSRA to provide strategic insights to inform the administration’s risk mitigation strategies, policy considerations, security countermeasures and programs, and resource allocation decisions. Our analysis of the TSSRAs issued during calendar years 2015 through 2017 indicates that TSA included intermodal station attack scenarios, but did not include cybersecurity scenarios. Specifically, the assessments featured various scenarios that targeted intermodal stations, which could include rail systems. For example, a scenario might describe attacks using various numbers of improvised explosive devices on an intermodal station. TSA did not include cybersecurity attack scenarios in the calendar year 2015, 2016, or 2017 assessments. According to the 2016 assessment and TSA officials we interviewed, threat experts have indicated that cyber threats, due to their unique nature and other factors, do not lend themselves to traditional TSSRA attack scenarios. However, as discussed below, the agency does conduct cyber threat assessments. Further, TSA’s Cybersecurity Roadmap 2018, states that, as one objective, the agency will include cybersecurity in its risk assessments for all modes. According to TSA officials, the implementation plan for the Roadmap, which was approved in September 2019, provides guidance and direction for meeting this objective. TSA officials confirmed that they plan to include basic cybersecurity scenarios for all modes in the 2020 TSSRA, and that they plan to engage with TSA mass transit experts and consult with industry experts as needed to inform future cyberattack scenarios. BASE. The BASE is a voluntary security assessment of national mass transit, passenger rail, and highway systems conducted by TSA surface transportation inspectors that addresses potential vulnerabilities, among other things. It consists of an assessment template with 17 security action items developed by TSA and the Federal Transit Administration that address, among other best practices, security training programs, risk information sharing, and cybersecurity. TSA developed this assessment in 2006 to increase domain awareness, enhance prevention and protection capabilities, and further response preparedness of passenger transit systems nationwide. The agency uses the BASE assessments to track progress in implementing specific security measures over time, offer technical assistance and share best practices to help improve the overall security posture of agencies, and inform transportation security grant funding by, among other things, identifying actions agencies have taken to reduce vulnerability. TSA officials stated that the most recent formal update to the assessment template began in 2014 and was fully implemented in 2015. The update included, among other changes, revised guidance for TSA surface inspectors and the addition of questions concerning active shooter events. In fiscal year 2016, the agency also developed a more targeted BASE assessment that focuses the assessment on an entity’s areas of concern as identified by surface inspectors in a previous BASE review of that operator. As of 2017, TSA had completed initial and follow-up assessments for the top 100 mass transit agencies in the country, which comprise approximately 80 percent of the ridership in the United States. TSA officials told us that their goal is to conduct follow-up assessments every one to three years. As previously shown in table 1, our analysis of the BASE template for mass transit and passenger rail indicates that it includes questions that address selected rail agency concerns about intermodal station security, and questions related to cybersecurity issues. Specifically, we found that while the template does not contain security action items or questions that directly refer to intermodal stations, questions in the template do correspond to topics that domestic rail agencies we interviewed identified as significant to intermodal station security, such as coordination among security forces, visible security measures, and establishing roles and responsibilities. For example, one BASE question asks if the agency’s system security plan has procedures or protocols for responding to security events with external agencies such as law enforcement or fire departments. This question corresponds to the challenge of coordination among security forces in intermodal stations identified by six of the seven agencies we interviewed. Cybersecurity is the focus of one of the security action items, which includes a series of general questions related to whether the transit agency has developed a comprehensive cybersecurity strategy. According to TSA officials, the agency added cybersecurity questions to the BASE in 2013 and the questions are intended to be a high level review. For example, the BASE addresses whether the transit agency has conducted a cybersecurity risk assessment, ensured employee training covers cybersecurity roles and threats, and established a protocol for reporting cyber incidents. It also provides a list of available cybersecurity resources for agencies to consult. Threat Assessments. TSA’s Intelligence and Analysis Office identifies security threats to mass transit and passenger rail systems through various threat assessments, including annual and semiannual Mass Transit and Passenger Rail Terrorism Threat Assessments and annual Cyber Modal Threat Assessments. TSA’s Mass Transit and Passenger Rail Terrorism Threat Assessment is produced annually and establishes the current mass transit passenger rail threat level and reviews terrorist threats against mass transit passenger rail for the past year. Threat information includes terrorist attacks on passenger rail trains, train tracks, buses, bus stops, and various stations. Additionally, the threat assessment analyzes intelligence gaps for the mass transit mode. TSA supplements the annual assessment with a semiannual threat assessment that reviews terrorist threats against mass transit and passenger rail for a 6-month period. Our analysis of threat assessments TSA issued for calendar years 2015 through 2019 indicates that they addressed stations, in general, and intermodal stations specifically, when they are the subject of an attack. For example, an attack on Manchester, England’s Victoria station, an intermodal station, was included in the 2018 Mass Transit and Passenger Rail Terrorism Threat Assessment. TSA’s Cyber Modal Threat Assessment reviews cyber threats to transportation over the course of the previous year, establishes cyber threat levels for the transportation modes for which TSA is responsible, and evaluates the threat through the next year or two. This annual assessment examines cyber threats to business and industrial control systems from state and non-state actors, including terrorist groups, pro-terrorist hacker groups, and hacktivists. Moreover, it analyzes incidents of cyberattacks and cyber espionage against U.S. and foreign transportation. Both assessments analyze threat actors and their capabilities, intent, and activities—including attacks occurring internationally—as well as tactics, techniques, and procedures that could be employed in future attacks. TSA calculates threat levels for transportation and cyber modes based on assessments of threat actor intent and capability. It may also issue additional situation-based products on emerging threats. TSA routinely shares these threat assessments with rail agencies and other stakeholders, such as industry security professionals. TSA Coordinates with CISA to Facilitate Voluntary Cybersecurity Assessments and Industry Outreach In addition to TSA’s risk assessment efforts, the agency coordinates with CISA, which conducts voluntary cybersecurity assessments as needed and requested by TSA and industry stakeholders. Specifically, CISA offers eight different voluntary cyber assessment options for public and private sector stakeholders, including mass transit and passenger rail agencies. Because CISA provides services to all 16 critical infrastructure sectors, including the transportation systems sector, officials noted that it must balance the resources it devotes to each sector. For example, CISA officials stated that they have conducted six Validated Architecture Design Review assessments on rail agencies since 2015, and currently have four pending requests from transportation agencies. The Validated Architecture Design Review evaluates systems, networks, and security services to determine if they are designed, built, and operated in a reliable and resilient manner. CISA officials also stated that they have conducted weekly vulnerability scans for one rail agency since 2015. While CISA coordinates with federal and private sector stakeholders to identify and address significant risks to critical infrastructure through its assessments, agency officials stated that they defer to TSA (as the co- sector specific agency for transportation) to take the lead in broader cyber initiatives and outreach to the transportation sector. For example, TSA officials stated that the agency included CISA in planning its cybersecurity workshops, a series of half-day workshops for surface transportation agencies to learn about cybersecurity resources from DHS and discuss nontechnical cybersecurity actions to improve their cybersecurity posture. According to TSA’s Cybersecurity Roadmap 2018, the agency plans to assess the resilience of the transportation modes to malicious cyber activity in conjunction with CISA, among other things. According to officials, TSA and CISA are collaborating or planning to collaborate on several cybersecurity assessments for passenger rail systems, including a cyber risk assessment for passenger rail cars and a cyber assessment of the mass transit and passenger rail mode. CISA officials told us that TSA, DHS’s Science and Technology Directorate, and CISA’s National Risk Management Center are in early phases of developing a cyber risk assessment for select passenger rail cars that they plan to produce in fiscal year 2020. CISA officials stated that they intend to address cyber vulnerability in the rail car assessments and plan to reach out to operators to discuss results. TSA officials told us that TSA and CISA also are considering a mass transit and passenger rail cyber assessment similar to one being developed for the pipeline mode. CISA officials stated that the planned pipeline assessment effort will include a total of 10 Validated Architecture Design Review assessments, in which TSA will help make arrangements with industry and will observe the process. TSA officials explained that expanding this effort to include passenger rail would depend on CISA’s availability to conduct assessments and balance demands in other sectors. CISA officials noted that they currently do not have the resources to support a similar plan for rail. TSA Actively Works with Domestic Stakeholders to Identify Standards and Key Practices but Provides Limited Guidance on Foreign Stakeholder Engagement TSA Works with Stakeholder Groups to Develop Domestic Standards and Recommended Practices TSA participates in APTA working groups that review and develop standards and recommended practices for passenger rail security, including those that apply to intermodal station security and cybersecurity. Specifically, from 2009 through 2019, APTA produced 45 documents related to security and emergency management standards and recommended practices, among other things. TSA is listed as a participant in 37 of the 45 documents. TSA officials noted that APTA working groups regularly review documents and issues related to security topics, including through monthly phone calls in some cases, and update them as needed. According to APTA’s Manual for the Standards Development Program, standards address safety-critical subjects and establish requirements that must be met by industry; recommended practices describe an established or generally recommended approach that does not rise to the level of a standard; and white papers are intended to provide information about complex issues that present the industry’s prevailing philosophy on the subject matter. For example: APTA offers a series of general standards, recommended practices, and white papers targeted at physical infrastructure protection at passenger facilities. These documents are not specifically directed at intermodal stations, but, according to our analysis and APTA officials, apply to such facilities as well as others. The documents address factors such as exterior door and window security, as well as securing mailrooms and utility openings, among other issues. Another APTA standard addresses security and emergency management considerations during planned special events, such as identifying transit hubs that are likely to be inundated with passengers going to and from the event. APTA offers cybersecurity recommended practices that are targeted at transit agencies in the early stages of starting a cybersecurity program, including how to obtain executive-level awareness and support and how to develop a cybersecurity awareness and training program. APTA also offers recommended practices for securing control and communications systems in transit environments, such as train control systems and fare collection systems. Table 2 provides additional examples of industry standards and key practice documents, as they relate to threats identified by domestic passenger rail stakeholders we interviewed. In addition to working with industry through APTA to develop standards and practices, TSA officials stated that the Surface Transportation Security Advisory Committee, which was established in 2019 to provide advice and recommendations to the TSA Administrator on transportation security matters, may serve as a mechanism for discussing or recommending key practices as the Committee develops. Officials noted that the Committee, which includes industry and community groups, could serve as a source for identifying forward looking best practices for rail security. The Committee held initial meetings in July 2019, October 2019, and January 2020, and proposed establishing subcommittees on topics such as cybersecurity and insider threats. None of the seven domestic rail agencies we contacted identified any security areas in which they felt recommended practices were missing. Officials from five agencies specifically commented on the usefulness of APTA publications. Officials from three agencies however, noted that many transit and rail agencies are still in the early stages of starting a cybersecurity program and that cybersecurity recommended practices are generally targeted at those agencies, as compared to agencies that already have a more sophisticated approach to cybersecurity. Officials from one agency further noted that publications related to the more technical aspects of cybersecurity (such as industrial control systems) can become outdated quickly as industry outpaces the development of security standards. TSA, CISA, and passenger rail agency officials we interviewed identified the NIST Cybersecurity Framework as the primary key practices document they reference for cybersecurity. Domestic and foreign rail agency, and industry association officials, as well as academic experts we interviewed noted that the possibility or likelihood of a cyberattack causing physical damage or harm to rail passengers or infrastructure is unlikely and largely hypothetical at this time. Academic experts we interviewed pointed to an incident in Poland in 2008 as one of the few, if only, known incidents in which a cyber-related attack on rail resulted in physical harm. In this incident, according to news reports, a Polish teenager modified a television remote control so that it could be used to control signals and switch points in a local tram system. Four vehicles derailed and 12 people were injured in the incident. Several rail agency officials and stakeholders we spoke with noted that successfully hacking into train control systems would require a highly sophisticated knowledge of the system. Officials further noted that train systems are designed to fail to safe mode and stop a train in the event of an abnormal signal, and that train operators have the ability to take over controls and manually stop trains if necessary. Officials from three rail agencies, however, stated that as agencies continue to adopt new technologies and systems become more interconnected, the potential for a cyberattack increases. Additionally, CISA officials and officials from one rail agency stated that, despite the lack of many incidents to date, protecting control systems is critical given the potential catastrophic impact of a successful attack. TSA Identifies Foreign Standards and Key Practices through Multilateral Working Groups and Bilateral Relationships, but Provides Limited Guidance to TSARs on Engaging with Foreign Rail Stakeholders According to TSA officials, TSA identifies foreign passenger rail security standards and key practices through engagement in multilateral groups and by leveraging bilateral relationships. Examples of multilateral groups include the International Working Group on Land Transport Security and the European Association of Railway Police Forces (RAILPOL). The working group, established in 2006, consists of 19 member states, including the United States. It is intended as a framework for members to openly share best practices, exchange information, and contribute to the development of surface transportation security initiatives. For example, TSA and members of the working group developed a searchable database of international surface transportation security measures (known as the SMARTbox) as a resource for surface transportation professionals to gain insights into security practices used by their peers. RAILPOL, founded in 2004, is an international association of government railway police organizations. It has 22 members, including TSA and the Amtrak Police Department. Information about intermodal stations and cybersecurity can be identified and exchanged through both of these mechanisms. For example, representatives from the United Kingdom delivered a presentation on securing intermodal stations at a 2016 working group meeting, and both working group and RAILPOL meetings have included cybersecurity discussions. Figure 3 provides an image of St. Pancras International Station in London, an intermodal station where international, local, and long distance trains converge with the London Underground. Regarding bilateral engagement, TSA identifies foreign rail security standards and practices through one-on-one relationships with other countries. TSA officials noted that their level of engagement with other countries can depend on a variety of factors, including how much the countries have in common regarding transportation systems and threats, and whether or not there are formal agreements in place that allow for regular, detailed information sharing. While some relationships are ongoing, officials stated that TSA interactions with other countries are often situational or transactional—countries may reach out either directly to TSA or through the TSAR for information about a specific issue, such as perimeter protection for surface transportation. For example: TSA holds biannual meetings with Transport Canada, the Canadian government department responsible for transportation policies and programs. Discussion topics from the meetings in 2017 and 2018 included Canadian efforts to develop passenger rail regulations, results from TSA derailment device testing, and opportunities for collaboration. According to TSA officials, TSARs in several countries have facilitated engagement with foreign surface transportation officials, including passenger rail officials. For example, officials stated that one TSAR facilitated the use of TSA’s Exercise Information System for an exercise on the metro system in a foreign city, as well as joint rail security training at TSA facilities in the United States. Officials further noted that another TSAR has taken initiative to facilitate quarterly meetings between foreign government and TSA surface transportation officials, including research and development and passenger rail officials. In addition to quarterly meetings facilitated by the TSAR, TSA officials stated that they are in regular contact with research and development officials in one country to share testing information, such as the results of derailment device testing and explosives testing on railcars, and to discuss security issues related to unmanned aircraft systems. TSA officials also reported that representatives attended an APTA- sponsored study trip to Brussels and London after the 2016 and 2017 rail attacks in those cities, in part, to observe lessons learned from the attacks. Foreign governments and international rail associations also produce a variety of passenger rail security standards and key practice documents. Table 3 below provides examples of these documents and the types of threats they address. TSA officials noted that while multilateral forums provide valuable opportunities to communicate with other countries about evolving threats, emerging security technologies, and potential key practices, interest in forums such as the International Working Group on Land Transport Security has been in decline. For example, while the working group charter calls for annual meetings and quarterly conference calls, the full group has not met since 2016. TSA and foreign government officials we spoke with stated that interest in the working group may be in decline due to factors such as retirements of key officials and lack of engagement from certain countries. These officials also noted that, as leaders in rail security, they typically provide more information about key practices to other countries in large forums than they receive. Additionally, TSA officials noted that other countries frequently used the working group- developed SMARTbox initially, but that use declined in recent years in part due to its location on the Homeland Security Information Network because users may find it difficult to navigate. Further, eight of the 10 domestic and foreign rail agencies we interviewed said they were either unfamiliar with the application or did not use it. For example, officials from one domestic agency said that there was little incentive to contribute and that they found informal networks to be more useful for sharing information. In contrast, TSA and other officials we spoke to stated that bilateral relationships with trusted partners with similar sophisticated rail operations may allow for more detailed exchanges of current and emerging key security practices. TSA Provides Limited Guidance to TSARs on Engaging with Foreign Rail Stakeholders TSA has provided limited guidance to TSARs on engagement with foreign passenger rail stakeholders through the TSAR Toolkit (or handbook), which states that TSARs should engage with officials involved in multiple modes of transit, including rail; however, the primary focus of the document is engagement with aviation stakeholders. TSA further provides comprehensive and specific guidance for TSAR aviation engagement as part of its foreign airport assessments and air carrier inspections, but does not do so for surface transportation. As discussed above, according to TSA officials, some TSARs have taken the initiative to facilitate meetings and share testing and training information related to surface transportation, including passenger rail. Passenger rail officials we talked to in one country stated that these TSAR-led initiatives served as a valuable source of information and communication with TSA. In addition, one TSA official cited the value of discussing preliminary testing findings, as well as new guidelines on topics such as security in station designs, which address concerns about security in public spaces. These efforts, however, are dependent on the individual initiative of each TSAR and are not universal. For example, one TSAR we interviewed stated that TSA’s expectations and priorities for surface transportation engagement were unclear and, as a result, he focused almost exclusively on aviation. TSA officials stated that they have focused TSAR guidance on aviation engagement because of the agency’s regulatory role in this area, which, as discussed above, includes foreign airport assessments and air carrier inspections. In lieu of detailed guidance on surface transportation, officials noted they defer to the individual TSARs on how or whether to engage foreign surface transportation stakeholders. Officials emphasized this individual approach and stated that in some countries, TSAR engagement on passenger rail security issues may be limited by legal or cultural barriers. Because rail (unlike aviation) does not directly connect to the United States in most cases, officials noted that there may be less incentive for some host countries to engage. Further, some countries may not have a rail system, or may not be as advanced in rail security policies and procedures, and therefore may be less able to offer key practices. In November 2019, TSA officials noted that they were considering adding guidance for engaging with surface transportation officials and addressing intermodal concerns to TSAR Regional Operational Implementation Plans. According to officials, these plans provide targeted guidance to TSARs for engagement within their specific regions. As of February 2020, officials stated that draft plans for two regions (Western Hemisphere and Africa/Middle East) were under review at TSA. Officials further stated that these drafts, and drafts for the remaining regions currently in development, would include surface transportation-related guidance. TSA officials stated that they hoped to complete all regions’ plans by the end of calendar year 2020, but they did not provide documentation for us to verify that the final plans would contain surface transportation guidance for TSARs. The 2018 TSA Administrator’s Intent document includes a goal to promote security partnerships across surface transportation systems by, in part, identifying and communicating best practices and lessons learned to stakeholders and international partners. In addition, the NIPP states that officials should share actionable and relevant information across the critical infrastructure community to build awareness and enable risk informed decision making. The TSAR Toolkit further states that, even in locations without modal connections to the United States, there is still great value in establishing key points of contact who can share best practices or facilitate the exchange of information in the event of an emergency in modes of transit outside of aviation. As the primary overseas point of contact for security matters involving all modes of transportation, TSARs are responsible for developing bilateral relationships and facilitating information sharing with foreign stakeholders, among other things. Further leveraging formal or informal bilateral relationships could allow TSA to obtain additional passenger rail security information. While several TSARs have individually taken initiative with regard to rail, without additional guidance from TSA, there is no assurance that they will engage in these exchanges with modes outside of aviation. As a result, TSA is less likely to be fully aware of key passenger rail security practices in other countries, such as those listed in table 3 above, among others. Moreover, specific guidance will also provide TSARs with clear expectations for engaging with stakeholders, and provide TSA with greater assurance that they are engaging in a consistent manner. TSA’s new Regional Operational Implementation Plans provide an opportunity for TSA to more clearly incorporate targeted guidance to encourage TSAR outreach and information sharing in specific areas. Recent efforts by TSARs in several countries demonstrate practices, such as opening lines of regular communication on surface transportation, including passenger rail, which could be replicated in other countries. TSA Uses Various Mechanisms to Share Security Standards and Key Practices but Does Not Fully Incorporate NIST Cybersecurity Standards in the BASE TSA Shares Information about Standards and Key Practices through Its Participation in Working Groups, and through Assessments and Exercises According to TSA officials and domestic rail stakeholders we interviewed, TSA uses various mechanisms such as the Transit Policing and Security Peer Advisory Group, monthly conference calls with rail stakeholders, and the annual APTA roundtable meeting to share and discuss a range of security information with stakeholders, including information about standards and key practices. These mechanisms provide opportunities to discuss issues related to intermodal stations and cybersecurity key practices. TSA also shares information about key practices with domestic stakeholders through voluntary TSA programs such as BASE, the Intermodal Security Training and Exercise Program, and the Visible Intermodal Prevention and Response program. TSA officials provided information about how they incorporate information from foreign threats and attacks into these programs. Specifically: TSA officials noted that TSA initially developed the BASE program around standards that were produced by APTA and other industry partners following the 2004 terrorist attacks on commuter trains in Madrid and the 2005 terrorist attacks on the London subway system. According to TSA officials, the APTA standards and recommended practices, which evolve based on threats and lessons learned, form the basis for the BASE assessment template. One way in which TSA helps communicates these standards and practices to agencies is through the questions in the template. Officials noted that lessons learned from foreign rail security incidents have been used to further support certain security concepts in the BASE, such as assessment questions related to whether agencies engage in public outreach for security awareness (e.g. “If You See Something, Say Something”) and report suspicious activity. TSA officials reported that they consider overseas and domestic attack methods and tactics when planning Intermodal Security Training and Exercise Program exercises to raise awareness of emerging tactics and threats. These exercises are intended to share best practices and lessons learned, among other things. Officials noted that they recently incorporated cyber, chemical, and vehicle- ramming attacks into the program’s objectives based on recent domestic and overseas incidents, and that they shared resources, information, and best practices for security solutions. For example, officials reported conducting two regional exercises that focused on chemical threat elements as the result of a 2017 plot in Australia. TSA further reported hosting a series of vehicle ramming program workshops in the wake of attacks in New York City and Europe. According to TSA officials, TSA has not made any recent changes to the Visible Intermodal Prevention and Response program directly as a result of lessons learned or key practices resulting from a foreign rail security incident; however, officials said they regularly integrate information about foreign incidents and threats when planning program deployments. Officials also noted that the majority of current deployments are for surface transportation, which includes rail. Regarding cybersecurity, TSA has shared information about cybersecurity key practices, including the NIST Cybersecurity Framework, through a series of regional cybersecurity Intermodal Security Training and Exercise Program workshops since 2017. These “5N5” workshops listed five nontechnical cybersecurity actions an agency could take in 5 days, including: (1) develop familiarity with the NIST Cybersecurity Framework; (2) implement a unique password change policy; (3) understand the latest phishing and spam trends and how to message awareness; (4) differentiate access control among staff; and (5) report cybersecurity incidents. Six of the seven domestic rail agencies we spoke with were generally satisfied with TSA’s efforts to share security and key practice information; however officials from two of these six agencies also expressed concerns about timeliness and quality of cybersecurity information provided by TSA. For example, officials from one agency stated that they received limited cybersecurity information from TSA and that the information they did receive was of limited use because it was targeted at agencies without a sophisticated cybersecurity program. An official from another agency noted that while there were opportunities to discuss cybersecurity, the information provided was often general in nature and there was limited time for discussion in certain mechanisms because of the large number of people involved. This official also noted that while the information TSA provides is valuable and there are mechanisms available to share information about a range topics, discussions are typically related to security incidents and threats, as opposed to key practices. TSA officials acknowledged that the agency’s cybersecurity efforts were still in the early stages. They further noted that the implementation plan for the 2018 Cybersecurity Roadmap, which, among things, calls for improving information sharing and partnering with stakeholders to promote the adoption of best practices and industry and/or international standards, was only recently signed in September 2019. In addition to TSA’s information sharing mechanisms, domestic rail agency officials we spoke to reported learning about foreign key practices through personal experience and direct engagement with foreign rail counterparts. For example, officials from two agencies we spoke to hosted visits from foreign rail officials to study security measures, among other things. Officials from one agency noted they provided information to Hong Kong through APTA on key practices for managing large protest crowds in an urban transit environment. Officials from another agency noted that they participate in international information sharing surveys and research to learn about cybersecurity practices by foreign rail operators, and sent representatives to an international mass transit training forum on the development of threat, vulnerability, and risk assessments. Domestic rail agencies also identified several changes they have made to their physical security systems as a result of key practices or lessons learned from foreign rail incidents. For example: increasing random patrols and high visibility deployments of security officers, changing security camera placement to better capture station exits, and increasing security awareness messaging to employees and passengers. Additionally, officials from one agency noted that they revised subway evacuation plans to direct people towards areas less vulnerable to an attack after reviewing lessons learned from recent vehicle-based attacks in Europe. With regard to cybersecurity, one domestic agency we spoke to noted that recent wide-spread global cyberattacks reinforced challenges they have securing legacy Information Technology systems against threats such as ransomware threats. As a result, the agency is focused on identifying expiring technologies and replacing those that can no longer be patched or updated. Officials from another agency noted that they have increased the number of firewalls they use to further segment and protect systems. Table 4 below provides information on mechanisms that can be used to identify and share rail security key practice information, as identified by TSA and domestic stakeholders. TSA Does Not Fully Incorporate NIST Cybersecurity Standards into Its BASE Assessments While TSA has taken initial steps to share cybersecurity key practices and other information with passenger rail stakeholders, the BASE assessment, does not fully reflect the updated cyber key practices presented in the NIST Cybersecurity Framework, nor does it include the framework in a list of available cyber resources. As discussed above, TSA uses the BASE assessment to share security best practices with transit agencies, among other things. Our review of the BASE cybersecurity questions in the template found that they cover selected activities associated with three of the five functions outlined in the framework– Identify, Protect, and Respond. For example, the BASE asks agencies if they ensure training reinforces cybersecurity roles and responsibilities, which corresponds to the awareness and training category of the NIST Protect function. However, the remaining two functions—Detect and Recover—are not represented in the BASE. According to the framework, when considered together, these functions provide a high-level, strategic view of the life cycle of an organization’s management of cybersecurity risk. TSA officials stated that they regularly review the BASE and noted that the questions are intended to reflect both industry key practices and agency policy; however, they also stated that the agency has not updated the BASE cybersecurity questions since NIST released its Cybersecurity Framework in 2014. In January 2020, officials responsible for the BASE acknowledged that the cybersecurity questions should be updated to reflect the framework. TSA officials also noted that they would want to align changes to the BASE cybersecurity questions with any new guidance or direction provided by the newly established Surface Transportation Security Advisory Committee. As of January 2020, the Committee is in its initial start-up phase, and has not yet provided any reports or recommendations or published a timeline or project plan. Further, because the framework functions organize basic cybersecurity activities at their highest level, incorporating elements of all five functions into the BASE template should not require additional guidance from the Committee. The 2015 TSA Transportation Systems Sector-Specific Plan states that encouraging the adoption of the NIST Cybersecurity Framework across all transportation modes supports the plan’s goal to manage the security risks to the physical, human, and cyber elements of critical transportation infrastructure. The plan also states that encouraging the adoption of the framework contributes to several of the NIPP’s calls to action related to sharing actionable and relevant information. TSA considers the framework a best practice document. By updating the BASE cybersecurity questions to align more closely with the core functions in the NIST Cybersecurity Framework, TSA could better assist passenger rail and other operators in identifying current key practices and improving their cybersecurity posture. As a result, transit operators would be more aware of cybersecurity vulnerabilities and better prepared to reduce the impact from a cybersecurity incident. In addition, this would create a more consistent cybersecurity approach from TSA, since the agency promotes the framework through other mechanisms, such as the series of cybersecurity workshops, as noted above. Conclusions Recent physical and cyberattacks in U.S. cities and Europe demonstrate the evolving nature of the threats to passenger rail and highlight the importance of working with both domestic stakeholders and foreign rail security partners. As such, TSA actively engages with domestic passenger rail stakeholders, but could do more to engage with foreign stakeholders. TSARs stationed abroad are well positioned to further leverage bilateral relationships with foreign passenger rail stakeholders, and several TSARs have taken initiative to do so. However, TSA provides only limited guidance to TSARs on surface transportation engagement. Without specific guidance, there is no assurance that TSARs will engage in these exchanges with modes outside of aviation. TSA’s new Regional Operational Implementation Plans provide an opportunity to more clearly incorporate targeted guidance to encourage TSAR outreach and information sharing in specific areas. Additionally, such guidance will provide TSA with greater assurance that TSARs are engaging with foreign stakeholders in a consistent manner. TSA uses various mechanisms to share security standards and key practice information with rail stakeholders, including through BASE assessments. The cybersecurity questions in the BASE template, however, do not fully reflect two of the five core areas identified in the NIST Cybersecurity Framework. By updating the BASE cybersecurity questions to align more closely with current key practices such as the framework, TSA could better assist passenger rail and other operators in improving their cybersecurity posture. As a result, transit operators would be more aware of cybersecurity vulnerabilities and better prepared to reduce the impact from a cybersecurity incident. Recommendation for Executive Action We are making two recommendations to TSA. The TSA Administrator should ensure that the TSAR Regional Operational Implementation Plans include guidance on how TSARs are to engage with foreign surface transportation stakeholders, including passenger rail stakeholders. (Recommendation 1) The TSA Administrator should update the BASE cybersecurity template to ensure it reflects cybersecurity key practices, including the Detect and Recover functions outlined in the NIST Cybersecurity Framework. (Recommendation 2) Agency Comments and Our Evaluation We provided a draft of this report to DHS for review and comment. DHS provided written comments, which are reprinted in appendix II, and also provided technical comments, which we incorporated as appropriate. DHS concurred with both recommendations and described actions TSA plans to take to address them. Specifically, to address recommendation 1, TSA plans to draft an Operational Implementation Plan, which will provide guidance to TSARs for engaging with foreign surface transportation stakeholders, including in passenger rail security. According to TSA, this plan will also serve as the outline for the development of Regional Operational Implementation Plans, which will help align resources worldwide. To address recommendation 2, TSA plans to update the BASE Cybersecurity Security Action Item section to ensure it reflects the NIST Cybersecurity Framework Detect and Recover functions. These actions, if fully implemented by TSA, should address the intent of both recommendations. We are sending copies of this report to the appropriate congressional committees and the acting Secretary of Homeland Security. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact Triana McNeil at (202) 512-8777 or McNeilT@gao.gov. Contact points for our Office of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributors to this report are listed in appendix III. Appendix I: Physical Security and Cybersecurity Key Practices Cited by Domestic and Foreign Stakeholders We asked domestic and foreign passenger rail agencies and foreign passenger rail stakeholders we interviewed to identify some security related key practices or lessons learned that they employ, including, but not limited to, intermodal stations and cybersecurity. Table 5 below provides examples of common security practices both domestic and foreign officials identified; table 6 shows several additional key practices foreign rail stakeholders cited. These tables are not intended to be a comprehensive list, but provide examples of key security practices utilized by selected domestic and foreign rail agencies. Figure 4 below shows an example of a Project Servator poster displayed during an exercise at St. Pancras International Station in London. As noted in table 6 above, foreign passenger rail stakeholders cited Project Servator as a key rail security practice. Appendix II: Comments from the U.S. Department of Homeland Security Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In additional to the contact named above, Christopher Ferencik (Assistant Director), Sarah Turpin (Analyst in Charge), Chuck Bausell, Benjamin Crossley, Suzanne Kaasa, Tracey King, Ronald La Due Lake, William Reed, and Adam Vogt made key contributions to this report.

Why GAO Did This Study Recent physical and cyberattacks on rail systems in U.S. and foreign cities highlight the importance of strengthening and securing passenger rail systems around the world. TSA is the primary federal agency responsible for securing transportation in the United States. GAO was asked to review TSA's efforts to assess passenger rail risk, as well as its role in identifying and sharing security standards and key practices. This report addresses (1) TSA's efforts to assess risk; (2) the extent to which TSA works with U.S. and foreign passenger rail stakeholders to identify security standards and key practices; and (3) the extent to which TSA shares passenger rail security standards and key practices with stakeholders. GAO analyzed TSA risk assessments from fiscal years 2015 through 2019 and reviewed TSA program documents and guidance. GAO interviewed officials from TSA, and from seven domestic rail agencies, three foreign rail agencies, and two foreign government agencies. The results from these interviews are not generalizable but provide perspectives on topics in this review. What GAO Found The Transportation Security Administration (TSA) assesses passenger rail risks through the Transportation Sector Security Risk Assessment, the Baseline Assessment for Security Enhancement (BASE), and threat assessments. TSA uses the risk assessment to evaluate threat, vulnerability, and consequence for attack scenarios across various transportation modes. TSA surface inspectors use the baseline assessment, a voluntary security review for mass transit, passenger rail, and highway systems, to address potential vulnerabilities and share best practices, among other things. TSA works with U.S. stakeholders to identify security standards and key practices and identifies foreign standards and practices through multilateral and bilateral exchanges. However, TSA Representatives (TSARs), the primary overseas point of contact for transportation security matters, lack specific guidance on foreign rail stakeholder engagement. As a result, TSA is less likely to be fully aware of key practices in other countries, such as station security guidance. Specific guidance would provide TSARs with clear expectations and encourage more consistent engagement with foreign rail stakeholders. Public Awareness Campaign Canine Units Emphasize security awareness Detection of vapor from explosives TSA shares standards and key practices with stakeholders, including those related to cybersecurity, through various mechanisms including BASE reviews; however, this assessment does not fully reflect current industry cybersecurity standards and key practices. For example, it does not include any questions related to two of the five functions outlined in the National Institute of Standards and Technology's Cybersecurity Framework—specifically the Detect and Recover functions. Updating the BASE questions to align more closely with this framework would better assist passenger rail operators in identifying current key practices for detecting intrusion and recovering from incidents. What GAO Recommends GAO is making two recommendations: (1) that TSA update TSAR guidance to include engaging with foreign passenger rail stakeholders; and (2) that TSA update the BASE cybersecurity questions to ensure they reflect key practices. DHS concurred with both recommendations.

gao_GAO-20-496

gao_GAO-20-496_0

Background The Partnership Program is part of the Bureau’s Integrated Partnership and Communications (IPC) operation, which contains the Bureau’s varied partnership and outreach activities aimed at spreading the word about the importance of participating in the census. Specifically, the IPC operation is intended to: engage and motivate people to self-respond; raise and keep awareness high throughout the entire 2020 Census to encourage response; support field recruitment efforts for a diverse, qualified census workforce; and support dissemination of census data to stakeholders and the public. Several components of the IPC operation, including the Partnership Program, are described in more detail below. Figure 1 also demonstrates these components in example situations. Partnership Program The Partnership Program has two main components that deliver outreach at the national and local levels, respectively: the National Partnership Program (NPP) and the Community Partnership and Engagement Program (CPEP), as shown in table 1. NPP and CPEP are intended to complement and leverage their respective expertise to help maximize participation by partners. According to Bureau officials, the Bureau’s Census Open Innovation Labs is another important component that aims to integrate these partnership activities and provide guidance at the national and local levels on various strategies for messaging with partners. The roughly 1,500 Partnership specialists hired for 2020 are temporary Bureau employees who implement CPEP and are responsible for, among others things, establishing local partnerships and engaging those partners to host activities and events (known as “commitments”) within the communities they serve. These commitments could include activities and events such as conducting knowledge-sharing seminars, issuing press releases, providing questionnaire assistance, producing pamphlets, and hosting field recruiting events, among others. For the 2020 Census, the Bureau had a goal of establishing 300,000 community partners nationwide. In 2010, the Bureau secured approximately 256,000 local partners. The Partnership program also supports inter-governmental census outreach activities through state complete count commissions (SCCCs) at the state level and complete count committees (CCCs) at the tribal and local levels. These groups foster a collaborative partnership between political, business, and community leaders to deliver messages on the importance of participation in the census. The Bureau had a goal of encouraging every state to create an SCCC, formed at the highest level of state government, such as the governor’s office. SCCCs seek to leverage the state’s vested interest in a timely and complete population count. According to Bureau planning documentation, CCCs are usually formed by the highest elected official in that jurisdiction, such as a tribal leader, a mayor, county commissioner, or regional chairman. The Bureau notes that partnership specialists are the primary contact between the Bureau and the CCC. Partnership specialists may conduct workshops, train CCC members, and provide or direct CCCs to census promotional materials. Bureau officials noted that initial partnership specialists were hired earlier in the decennial cycle for 2020 in part to help form CCCs. Partnership specialists are also to play a role in assisting CCCs with identifying hard- to-count populations within their respective communities and assist in developing strategies to reach those communities. Paid Advertising Similar to previous censuses, for 2020, the Bureau plans to use multiple paid media outlets such as radio, television, newspapers, magazine, and billboards as one of its means to promote awareness and encourage participation in the census. Along with the traditional advertising, the Bureau also plans to use digital advertising such as web banner ads, video ads, digital extensions of traditional outlets, and social media channels. According to Bureau officials, they plan to be able to use paid advertising modes to target specific audiences including hard-to-count populations that may be more apt to trust such communication. For example, the Bureau’s 2020 communication plan cites research that states that the majority of Asian-Americans use traditional media— whether in print, online, or from a mobile device—as their primary source of local news. The Bureau plans to implement its 2020 paid media campaign in four phases (see table 2). As of early May 2020, the Bureau has not yet indicated the revised timeframes for the phases of paid advertising to reflect the COVID-19 disruptions. Bureau officials told us that they are planning to spend an additional $160 million on advertising as part of its public education and outreach campaign, originally estimated to cost over $500 million. According to the Bureau, its paid advertising campaign is designed to reach more than 99 percent of the nation’s 140 million households. Media Relations For the 2020 Census, to distribute these messages, the Bureau will conduct outreach to the media including providing talking points, media lists, news releases, fact sheets, frequently asked questions, and other scripts and messaging for multiple media platforms, such as a variety of radio, podcasts, and special events. Media-focused materials such as press kits and press releases will be available in the media section of 2020Census.gov. As part of its media relations effort, the Bureau has also developed a crisis communications plan for handling major events and potential messaging disturbances. According to the Bureau, its crisis communication plan is intended to allow it to respond quickly and effectively to any events or actions that jeopardize the public’s confidentiality or reduce its willingness to respond to the 2020 Census. For example, the Bureau stated that in response to the COVID-19 outbreak, it adapted part of its advertisement messaging to re-emphasize the importance of responding to the census online, avoiding the need for in-person follow-up interviews. Social Media Outreach The Bureau’s social media strategy for 2020 reflects the increased number of social media platforms that were not available during the last census, such as Snapchat, Instagram Live Stories, and Facebook Messenger. According to the Bureau, social media will play a critical role in raising awareness of the census—particularly among hard‐to‐count audiences—as well as enhancing customer service efforts, promoting recruiting efforts, driving online completion of the census, assisting with data dissemination, and mitigating disinformation. The Bureau’s social media outreach efforts are to leverage source material on the Bureau’s website. Such material includes prepared language for posts and graphics, lists of influential partners, messages for those partners, customer service-themed frequently asked questions that are tailored for social media, methods for sharing content across the Bureau’s regional offices, and social media events. Statistics in Schools In March 2020, the Bureau kicked off its Statistics in Schools (SIS) program. During the week of March 2-6, 2020, prior to the widespread closure of schools due to the COVID-19 outbreak, the Bureau provided daily modules to educators for dissemination to students. These modules included such topics as an introduction to the census, a virtual tour of the Bureau headquarters, and a take-home assignment in which students and their families were asked to summarize a discussion of what they want to see in their communities. According to the Bureau planning documentation, the SIS materials were to encourage students to pass along the importance of counting everyone to an adult in their home who will complete the census. SIS was also intended to raise awareness among students themselves, which can be important in instances where the presiding adult(s) have limited English proficiency and have to rely on children to translate or interpret information from English into their native language. The Bureau Has Taken Actions to Address Prior Challenges, but Issues for Continued Monitoring Remain We, the Bureau, and others, such as the Bureau’s National Advisory Committee, have previously identified several challenges the Bureau has faced related to its partnership and outreach efforts in prior censuses. While the Bureau has taken important actions to address these challenges, current events such as the COVID-19 outbreak provide a salient basis for which to continue to monitor these challenges and any effects they may have on the census. Challenge 1: Enumerating Hard-to-Count Groups Nature of Challenge. The Bureau strives to conduct an accurate count of the nation’s population. However, some degree of inaccuracy is inevitable. While the Bureau reported that the 2010 Census did not have a significant net undercount or net overcount nationally, the Bureau also reported that errors in coverage were unevenly distributed through the population, as figure 2 shows. In addition to the undercounted groups shown below, prior censuses, such as the 1990 Census, also showed statistically significant undercounts of Non-Hispanic Asians, American Indians off (as well as on) Reservations, and Native Hawaiian or Pacific Islanders. These errors are problematic because certain groups such as minorities, young children, and renters are more likely to be missed in the census, while other groups such as those who may own a second, seasonal home are more likely to be counted more than once. The Bureau has noted these historical trends and has classified these and other subpopulations as hard to count. According to the Bureau, hard-to-count groups can share some or all of the following characteristics: Hard to Locate: Some groups are hard to locate because where they live is unknown, or they move frequently. Hard to Persuade: Other groups are hard to persuade to participate in the census. Hard to Interview: Some groups may have communication barriers, such as limited English proficiency. Hard to Contact: Other groups may live in places with access barriers, such as residents of gated communities or renters with doormen and locked buildings. Adding to the challenge in 2020 is the possibility that, as the Bureau adapts its timing and procedures for outreach and promotion, as well as data collection, there is a possibility of uneven effects resulting on the participation rates of different groups. For example, messaging and operations that emphasize the importance of filling out the census online, in response to the COVID-19 outbreak, may not be applicable to communities or groups with limited internet access. If social distancing measures result in fewer successful interviews during Non-Response Follow-Up, for instance, then these groups with less internet access will be at relatively greater risk of being missed by the census. Adapting field enumeration procedures to implement social distancing might also be less effective in addressing respondent concerns about interacting with strangers in apartment buildings or other densely populated areas if a census worker cannot practically distance themselves from the door. Actions Taken. As part of its integrated communications plan, the Bureau’s lead communications contractor segmented the market and developed a series of strategic frameworks targeting advertising and messaging to hard-to-count subpopulations through demographic profiles. These groups include: persons experiencing homelessness; households with young children; lesbian, gay, bisexual, transgender, queer, and questioning persons; persons with disabilities; the young and mobile. In addition to demographic profiles for each of these groups, the contractor compiled lists of relevant community partners and consulted with stakeholders to construct a “day-in-the-life” analysis and develop advertising strategies to align tailored census advertising with the experiences of each group. According to Bureau documentation, these strategic frameworks provide the NPP data to decide how to best target related outreach resources. The Bureau also developed for 2020 a publicly-available tool that displays the areas of the country that are considered hard to count, according to an index of demographic indicators known as the Low Response Score. Using this index, the Bureau is able to monitor incoming census data for those areas, such as response rates and hiring and recruitment numbers, to see if efforts to reach hard-to-count areas are succeeding amid challenging current events, such as the COVID-19 outbreak, which could affect willingness of workers to participate in work taking them door-to- door to nonrespondents. Using funds appropriated under the Further Continuing Appropriations Act, 2020, the Bureau also developed a mobile questionnaire assistance (MQA) initiative to deploy census workers to specific locations, such as grocery stores, houses of worship, and community centers, or at specific events, such as festivals, to assist residents of low-response areas in filling out the census. For the beginning of the self-response period, partnership specialists identified the initial locations and times for the initiative within hardest-to-count census tracts. The Bureau plans to then monitor actual self-response data later in the operation to target those areas reporting the lowest response rates. In April 2020, the Bureau announced an indefinite delay of this latter stage of MQA in response to the outbreak of COVID-19. Basis for Continued Monitoring. Participation in the census and availability of nonrespondents for follow-up will help indicate whether the Bureau is successfully reaching hard-to-count groups. With self-response and follow-up for nonrespondents still ongoing, it is too early to know the effectiveness of the Bureau’s outreach efforts. However, multiple streams of data will provide indications of Bureau success in enumerating areas and demographic groups considered hard to count: As during the 2018 Census Test, the Bureau is monitoring active data on self-response rates at the local level, which it can compare across areas it deems hard to count. The Non-Response Follow-Up operation will yield data on interview rates—namely, the rate at which census workers are able to complete interviews with residents who had not yet responded to the census. These rates can also be compared across areas deemed hard to count. Demographic evaluation data and the Bureau’s post-enumeration survey will further provide insight into whether racial, ethnic, and other demographic groups were counted accurately. Challenge 2: Mobilizing Partnership and Outreach Resources Nature of Challenge. For its partnership and outreach strategies to be effective, the Bureau must have the necessary people and resources in the right places to execute those strategies. In prior censuses, we have reported on issues related to staffing and promotional materials faced by the Partnership Program. Specifically, during the 2000 Census we noted that partnership resources were stretched thin and in some cases took effect too late in the decennial cycle. We recommended that the Bureau review its staffing levels to ensure adequate support to partners, a recommendation the Bureau subsequently implemented. During the 2010 Census, partnership specialists expressed concern about the timely availability of promotional materials, and the impact on their ability to build relationships with potential partners. We recommended in December 2010 that the Bureau ensure that promotional materials are provided to partnership staff when they are hired, a recommendation that the Bureau agreed with but that remained open at the beginning of our audit work in December 2019. Similarly, in a 2012 Bureau evaluation of the 2010 NPP, the Bureau found that some national partners felt that promotional materials needed to be better tailored to target their audiences’ needs and that these partners had difficulty accessing relevant census data for their audiences. The Bureau also observed that improvements were needed in the distribution of promotional materials by region. Actions Taken. The Bureau increased its partnership specialist hiring from roughly 800 in 2010 to a little more than 1,500 in 2020. As figure 3 shows, collectively these partnership specialists were able to secure more than 307,000 community partners by the end of February 2020—higher than the Bureau’s goal of 300,000 and the roughly 256,000 local partners the Bureau had by the end of the 2010 Census. Partnership staff were also able to encourage the creation of state complete count commissions (SCCCs) in every state and territory except Nebraska and South Dakota. These SCCCs are complemented across the country by more than 10,000 complete count committees (CCCs) at the local level. The Bureau also set a priority of getting partners in the right places. In addition to the Bureau’s overall goal for community partners, the Bureau also set out to get at least one community partner in 100 percent of census tracts that, according to their predicted indicator of low response, are hardest to count. As of late March 2020, the Bureau had made progress towards this goal, establishing partnerships in 85 percent of such areas. The Bureau also provided partnership specialists a centralized website where they can access promotional materials for distribution to partners. These materials included fact sheets, brochures, and marketing messages translated into 13 languages. Bureau officials indicated that these materials were developed and published on the website prior to the completion of partnership specialist training. According to officials, the website was made available to partnership specialists when they started work. In addition to national and community partners, tribal, state, and local CCCs also have access to these materials. The Bureau also provided guidance on how to order hard copies of such materials through the U.S. Government Publishing Office. We found during the course of our audit work that these actions implemented our December 2010 recommendation. Basis for Continued Monitoring. The Bureau encountered some issues in onboarding partnership specialists and tracking the establishment of partnerships in hard-to-count areas. We will continue to monitor any effects of these issues as part of our ongoing work: The Bureau experienced delays in onboarding partnership specialists, which resulted in less time to form partnerships and meant less time for community engagement and education activities leading up to the census. While the Bureau was able to increase the number of partnership specialists from 2010, the Bureau did not get all of its more than 1,500 partnership specialists on board until November 2019, more than 4 months later than its initial goal. The Bureau successfully surpassed its nationwide goal for registering more than 300,000 community partners. However, we reported in February 2020 that the Bureau missed interim goals for getting at least 200,000 partners in place by January 1, 2020, and at least 250,000 partners in place by February 1, 2020. We reported during the 2000 Census on the benefits of having Partnership Program resources on the ground earlier in the decennial cycle. The Bureau varied by region in terms of getting community partners located in hard-to-count areas. As noted, the Bureau had a goal of getting at least one partner in each of the hardest-to-count census tracts by March 2020. Nationwide, the Bureau was able to achieve this in 85 percent of such areas as of late March 2020. According to Bureau data, five of the six regional offices reached at least 82 percent of this goal, with the Los Angeles region the farthest along at 90 percent as of late March 2020. On the other hand, only 70 percent of the hardest-to-count tracts in the Philadelphia region, which covers the Mid-Atlantic States, had at least one community partner. Bureau officials noted that partners in adjacent tracts can provide relevant services to the hardest-to-count tracts. Officials indicated that they would continue monitoring progress using this measure. Given the effect of the COVID-19 outbreak on partner activities, having partnership specialists on boarded and partnerships formed later than anticipated meant several months less time for in-person community engagement and education activities leading up to the census in some areas. In April 2020, Bureau officials told us that partnership specialists had been instructed to continue interactions with partners via conference calls, text, email, and other virtual means. Officials also cited the ability of partnership specialists to support virtual partner commitments such as radio interviews and virtual town halls, and officials noted that the Government Publishing Office could continue to directly supply hard copies of promotional materials to partners. Going forward, response rates the Bureau achieves in areas where the Bureau lagged in registering partners may shed light on whether or not having full partnership coverage in certain hard-to-count areas is associated with lower response rates. Additionally, the Bureau is conducting a survey of public awareness of and sentiment toward the 2020 Census, with a goal of evaluating the effectiveness of public communication efforts. This survey will also provide data that could help answer the question of whether having less time on the ground than anticipated for the full complement of partnership specialists and community partners affected community awareness of the census. Challenge 3: Coordinating Outreach across the Bureau’s Organization and Operations Nature of Challenge. Partnership and outreach activities, along with local enumeration activities, span numerous decennial operations and phases of data collection, and the Bureau has at times struggled to fully integrate these efforts. For instance, we reported in July 2018 that, during the planning stages for the 2020 Census, the Bureau’s management of initiatives aimed at enumerating hard-to-count groups was decentralized and not fully integrated across operations. We recommended that the Bureau take steps to ensure greater integration of efforts going forward. Coordination across regions of the country can also be a challenge. Bureau officials overseeing CPEP told us that during the initial staffing of partnership specialists, they observed variations in how job expectations were communicated across regional offices through training. The COVID-19 outbreak further complicates coordination across the country where state, tribal, and local public health rules, guidance, and enforcement may affect how partners in different locations need to interact with the census, and messaging to Bureau partnership staff and partners may not be amenable to a one-size-fits-all approach nationwide. The challenge of coordination can also have effects on how local managers of census activities perceive the efficiency of census efforts. Partnership specialists are expected in 2020 to work on initiatives, such as MQA, to which area census office staff also contribute. However, partnership specialists do not report to managers within the area census office, but directly report to the Bureau’s regional offices. As part of our December 2010 report on hard-to-count populations, we found that about half of all local census office managers we surveyed were dissatisfied with the level of coordination between local census office staff and partnership staff, noting duplication of effort in some cases. We recommended that the Bureau develop mechanisms to increase coordination between partnership and local census office staff, a recommendation that remains open. Actions Taken. In January 2020, in response to our July 2018 recommendation cited above, the Bureau provided us evidence of ongoing, multi-team discussions focused on integrating perspectives on decisions and planning for hard-to-count enumeration activities. In closing the July 2018 recommendation, we noted that the Bureau’s use of such an integrated approach will help ensure that the Bureau’s otherwise decentralized efforts to address hard-to-count challenges will be more effective. For field operations, the Bureau is holding a series of weekly teleconferences, to which area census office managers and all partnership specialists are invited to hear about updates from Bureau management on field operations and partnership and outreach activities. Bureau officials told us that having these calls, along with providing emailed summaries of the calls, helps to standardize any updates to guidance on field procedures or messaging across the area census office and partnership staff. For example, the Bureau sent an email in mid- March 2020 to partnership staff that included guidance on emphasizing the importance of internet self-response when responding to inquiries on the COVID-19 outbreak. The Bureau indicated taking steps to standardize partnership staff training and integrate partnership staff into field enumeration operations. Specifically, officials told us they convened a nationwide, standardized training curriculum after the Bureau had hired all of its partnership specialists. This curriculum replaced what had been a regional office- specific approach. The Bureau developed guidance for partnership specialists on how to provide direct feedback to area census offices when they become aware of a potential facility or location that should be added to the Bureau’s database for enumeration. Specifically, if partnership staff become aware of soup kitchens, homeless shelters, and other locations potentially being missed, they are instructed to submit information including the address, type of living quarters, and any applicable contact information to the relevant area census office. Moreover, the Bureau indicated it had assigned at least one partnership specialist to every area census office to improve coordination between the Partnership Program and local census operations. In addition, the Bureau created a rumor control page on its website to quickly disseminate factual content in response to misinformation. Examples of misinformation that the Bureau has detected include the false suggestion of a citizenship question being on the census and the posting of a potentially fraudulent census job website. The Bureau’s rumor control page includes the email address rumors@census.gov for the public to report possible misinformation and disinformation. The web page also has a set of frequently asked questions on topics such as data confidentiality and ways to participate in the census, thus helping to ensure that partnership staff, partners, and the public alike with access to the internet have access to accurate online information. Basis for Continued Monitoring. As the Bureau experiences inevitable turnover in partnership specialists and needs to update its messaging and outreach to reflect changing conditions on the ground, such as the COVID-19 outbreak, ensuring consistent training and guidance across its partnership staff and keeping all staff abreast of changes will be important. As noted above, our December 2010 recommendation for the Bureau to increase coordination between the Partnership Program and its area census offices remains open. While partnership specialists were assigned to area census offices, the Bureau has not put in place formal expectations for how partnership staff should support area census office staff and their activities. To fully implement this recommendation, the Bureau would need to document how partnership specialists and area census office staff are expected to work together. We will also continue to monitor how area census office managers express their perspectives on the effectiveness of the Partnership Program in supporting implementation of census field operations. For example, we surveyed area census office managers in March 2020 and found that nationally a plurality (42 percent) who responded said they were dissatisfied with the level of communication and coordination between the Partnership Program and their offices, while 41 percent said they were satisfied. A plurality of respondents (44 percent) were also dissatisfied with the level of clarity of roles and responsibilities between the Partnership Program and their offices, while 38 percent of respondents were satisfied. Open-ended comments we received from survey respondents contained ideas for improving the level of communication between partnership staff and area census office staff, such as having partnership staff be more integrated within the area census office structure. We are sharing such suggestions with the Bureau in near- real time as we receive them for the Bureau to consider moving forward. We will continue to periodically survey the office managers throughout the summer of 2020. Challenge 4: Measuring Outcomes Nature of Challenge. We have previously reported on challenges the Bureau has faced across operations in measuring outcomes. Namely, during the 2018 Census Test we reported that the Bureau’s primary measure of operational progress during Non-Response Follow-Up overstated the amount of completed workload because it emphasized the number of follow-up attempts made instead of the number of cases in which the Bureau got the data it needed—the preferred outcome of the operation. Similarly, our 2017 review of in-office address canvassing—when the Bureau attempted to use aerial imagery to identify areas of the country called blocks that did not require fieldwork to validate the address list— found that the Bureau’s performance measures looked at how many of the addresses in the address list were reviewed in-office, rather than how many addresses did not require additional fieldwork. Separately measuring the contributions of partnership and outreach activities on the desired outcome of maximizing the number of residents who respond to the census is inherently difficult. In one of its evaluations of the 2010 Census, though, the Bureau noted that the Partnership Program needed to do a better job of measuring and rewarding the intensity of partner effort throughout census operations, rather than simply looking at the number of contacts made or the number of partners. As Bureau officials noted to us, the quality of partners ultimately matters more than the quantity. Actions Taken. In its performance measures of the Partnership Program, the Bureau has taken steps beyond measuring the number of partners nationwide. For example, as noted previously, the Bureau prioritizes and tracks the extent to which it secures partnerships in hard-to-count areas, acknowledging that partnerships are needed everywhere but are particularly critical in areas that may otherwise be less likely to have high response rates. Additionally, the Bureau tracks the number of events partners complete and the number of commitments partners signed up for but did not fulfill by their planned dates (known as overdue commitments). As of early March 2020, community partners had fulfilled more than 270,000 committed events nationwide, while nearly 14,000 commitments were overdue. Basis for Continued Monitoring. Going forward, Bureau evaluations of the 2020 Census will demonstrate what, if any, lessons learned the Bureau is able to draw regarding the quality of its 2020 partnership and outreach efforts. The Bureau has multiple sources of data that it is, and will be, collecting to better understand how success can be defined: While the Bureau monitors and reports the number of overdue partnership commitments (cited above), it is not clear what the current level of nearly 14,000 overdue commitments means in the context of the universe of more than 270,000 total committed events. When we asked about this issue, Bureau officials indicated that they monitor the status of overdue commitments mostly to see if partners are staying active in their communities. Officials also said they use the metric to verify that partnership specialists are recording any follow-up they do with partners that have not fulfilled their commitments. Similarly, the Bureau’s plan for the MQA initiative cites the number of census questionnaire responses received through MQA as an indicator of how successful the Bureau is in connecting partners and other census questionnaire assistance resources to areas that need them. Yet, it is unclear from Bureau documentation what levels of MQA uptake, if any, would constitute success. While low MQA uptake, for instance, could mean that partnership specialists and census response representatives were not successful in identifying the right times and locations for MQA, it could also mean that residents were successfully able to respond to the census via other means, or that they did not attend MQA events due to the COVID-19 outbreak. When we followed up with Bureau officials, they indicated that they have planned a series of required debriefings and exit interviews with partnership specialists, which will include MQA effectiveness as a topic. Such debriefings and exit interviews will likely be particularly important, given the MQA change necessitated by the COVID-19 outbreak. Lastly, the COVID-19 outbreak could further complicate the Bureau’s ability to determine the ultimate effectiveness of its partnership and outreach—how well the Bureau achieves the goal of counting everybody once, only once, and in the right place. The Bureau plans to estimate census quality by relying in part on interviews conducted door-to-door in a nationwide sample of households, scheduled for the summer and fall of 2020. The COVID-19 outbreak could prompt the Bureau to delay the related field operation to collect the data or affect household responsiveness to in-person visits. As the timing of this report coincides with ongoing implementation of the self-response period and field enumeration operations, planned Bureau evaluations and assessments will be best positioned to identify any lessons learned and determine whether, in light of the challenges cited above, the Bureau’s partnership and outreach efforts were successful in maximizing participation in the census. Agency Comments and Our Evaluation We provided a draft of this report to the Department of Commerce. In its written comments, reproduced in appendix I, the U.S. Census Bureau said it agreed with the findings of our report and would continue to work to implement related open recommendations from our prior reports. We are sending copies of this report to the Secretary of Commerce, the Undersecretary of Economic Affairs, the Director of the U.S. Census Bureau, and the appropriate congressional committees. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report please contact me at (202) 512-6806 or mihmj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix II. Appendix I: Comments from the Department of Commerce Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments J. Christopher Mihm, (202) 512-6806 or mihmj@gao.gov In addition to the contact named above, Ty Mitchell (Assistant Director), Devin Braun (Analyst-in-Charge), Mark Abraham, Carole J. Cimitile, Alexandra Edwards, Amalia Konstas, Kerstin Meyer, Cynthia Saunders, Kate Sharkey, Farrah Stone, Jon Ticehurst, Peter Verchinski, and Alicia White made key contributions to this report.

Why GAO Did This Study The decennial census is used to apportion seats in Congress, redraw congressional districts, and allocate hundreds of billions of dollars in federal assistance annually and helps to guide public policy decisions based on social, economic, and demographic data. While recent censuses appear to have been increasingly accurate, measurement errors are not evenly distributed across the population. Given the uses of census data, ensuring an accurate count is important. As part of its partnership and outreach efforts, the U.S. Census Bureau's (Bureau) Partnership Program works with local and national organizations, businesses, and governments to promote awareness of and participation in the census, as well as to help recruit census workers. GAO was asked to review the Bureau's partnership and outreach efforts, including paid advertising and targeted communications. This report examines the Bureau's progress in addressing selected prior census challenges in these areas. GAO reviewed relevant Bureau planning documentation, collected regular Bureau reports on progress, and interviewed Bureau officials responsible for partnership and outreach efforts. GAO provided a draft of this report to the Bureau. The Bureau agreed with the report's findings. What GAO Found The Partnership Program, a core component of the Bureau's partnership and outreach activities, delivers outreach to partnering organizations at the national and local levels in order to ensure a more complete and accurate count. These partners include retail associations, tribal, state, and local governments, local businesses, and non-profit organizations, among others. Roughly 1,500 partnership specialists, who are temporary Bureau employees responsible for building relationships with and obtaining commitments from these partners, help to implement the Partnership Program, which exists alongside several other components of the Integrated Partnership and Communications operation, as shown below. The Bureau experienced delays, however, in getting these employees onboarded. The Bureau has taken important actions to address challenges that GAO, the Bureau, and others have previously identified. These challenges include: (1) Enumerating hard-to-count groups; (2) Mobilizing partnership and outreach resources; (3) Coordinating outreach across the Bureau's organization and operations; and (4) Measuring outcomes. Events taking place during implementation of partnership and outreach activities, such as the COVID-19 outbreak, provide a salient basis for which to continue to monitor these challenges and any effects they may have on the census. Moreover, continued monitoring of the Bureau's survey of public awareness of and sentiment toward the census, for example, will provide information on whether difficulties experienced in getting partnership specialists onboarded had an effect on the success of the Bureau's outreach.

gao_GAO-19-571T

gao_GAO-19-571T_0

Background VA is responsible for providing benefits to veterans, including health care, disability compensation, and various types of financial assistance. In fiscal year 2019, VA received a total budget of $201.1 billion, and the largest discretionary budget in its history—$86.6 billion, about $20 billion higher than in 2015. The department operates one of the largest health care delivery systems in the nation through its Veterans Health Administration (VHA), with 172 medical centers and more than 1,000 outpatient facilities organized into regional networks. VA has faced growing demand by veterans for its health care services, with the total number of veterans enrolled in VA’s health care system rising from 7.9 million to more than 9 million from fiscal year 2006 through fiscal year 2017. In fiscal year 2019, VHA received $73.1 billion of VA’s $86.6 billion discretionary budget. In addition to providing health care services, VA provides cash benefits to veterans for disabling conditions incurred in or aggravated by military service. To carry out its mission, VA spends tens of billions of dollars to procure a wide range of goods and services, including medical supplies; to construct hospitals, clinics, and other facilities; and to provide the information technology (IT) to support its operations. We have made hundreds of recommendations to improve VA’s management and oversight of the services it provides to veterans. Specifically, since 2000, we have made 1,225 recommendations to VA. While VA has implemented most of the recommendations, a number remain open, as of April 2019. Specifically, more than 125 recommendations related to VA health care remain open, including 17 recommendations that have remained open for 3 years or more; 15 recommendations related to improving VA acquisition management remain open, including 1 recommendation that has remained open for 3 years or more; and 12 recommendations related to management of disability claims workloads. In 2017, we began sending letters to VA and appropriate congressional committees identifying priority recommendations for VA to implement in order to significantly improve its operations. We categorized these recommendations into nine areas: (1) veterans’ access to timely health care; (2) veterans’ community care program; (3) human capital management; (4) information technology; (5) appeals reform for disability benefits; (6) quality of care and patient safety; (7) national policy documents; (8) contracting policies and practices; and (9) veterans’ access to burial options. Overall Rating for the Managing Risks and Improving VA Health Care High-Risk Area Remained Unchanged in 2019 Since we designated VA health care as a high-risk area in 2015, VA has begun to address each of the identified five areas of concern related to managing risks and improving VA health care: (1) ambiguous policies and inconsistent processes; (2) inadequate oversight and accountability; (3) IT challenges; (4) inadequate training for VA staff; and (5) unclear resource needs and allocation priorities. Since our 2017 High-Risk Report, ratings for all five criteria remain unchanged as of March 2019. Specifically, the leadership commitment and action plan criteria remain partially met. Although VA has experienced leadership instability over the past 2 years in several senior positions, a new Secretary was confirmed in July 2018. Secretary Wilkie has demonstrated his commitment to addressing the department’s high- risk designation by, among other things, creating an office to direct an integrated, focused high-risk approach and communicating to VA leaders the importance of addressing our recommendations and working with GAO. The Secretary’s actions, to date, have allowed the department to maintain its leadership commitment rating as of March 2019. The action plan criterion also remains partially met as of March 2019. In March 2018, VA submitted an action plan to address the underlying causes of its high-risk designation, but the plan did not clearly link actions to stated outcomes and goals or establish a framework to assess VA’s progress. VA officials told us that instead of revising the March 2018 action plan, it will incorporate its plans to address the high-risk designation into the department’s current initiatives. Specifically, VA is currently implementing the VHA Plan for Modernization, through which the department intends to modernize VA’s structure, culture, governance, and systems through organizational improvements. VA officials have indicated that the VHA Plan for Modernization is intended, among other things, to address the high-risk areas for VA health care. VA officials also told us they are currently developing operational plans for the VHA Plan for Modernization, and these plans will include goals, time frames, and metrics, among other things. VA estimates that the operational plans will be complete by September 2019. The monitoring, demonstrated progress, and capacity criteria remain unmet since our 2017 High-Risk Report. In order to address the monitoring and demonstrated progress criteria, VA’s ongoing revisions to its action plan need to include the addition of certain essential components, including metrics, milestones, and mechanisms for monitoring and demonstrating progress in addressing the high-risk areas of concern. VA’s capacity rating also remains not met. Though the department took steps to establish offices, workgroups, and initiatives to address its high-risk designation, many of these efforts are either in the initial stages of development or resources have not been allocated. For each of the five identified areas of concern related to managing risks and improving VA health care, ratings reflect the level of progress VA has made to address them. Ambiguous policies and inconsistent processes. Since our 2017 High-Risk Report, ratings for all five criteria remain unchanged for this area of concern as of March 2019. Leadership commitment: partially met. In September 2017, we reported that VHA had approximately 800 national policies, the majority of which were outdated. VHA reported reducing the number of national policies by 26 percent, and work continues in this area. In addition, VHA established an inventory of approximately 55,000 local policies as of October 2017. In October 2018, VHA noted its plans to determine who is responsible for monitoring implementation of national and local policy, as well as the alignment between these levels of policy. At that time, VHA also discussed its future plans to monitor the implementation and alignment of national and local policy and update its national policy directive by the end of June 2019. Additionally, VA has implemented a structure for leadership input into the policy process, such as at the VHA Chief of Staff level. However, senior leadership has lacked the stability needed to ensure issued policy meets agency goals. Capacity: not met. Since 2017, VA has issued an updated directive on policy management, and put in place procedures to train staff and obtain input from all levels on policy development. However, VA continues to face challenges in this area because it is reliant on contracts and information technology resources, which if delayed, can impede progress toward meeting goals. Action plan: partially met. Since 2017, VA has further refined its root cause analysis for this area of concern. In June 2017, VA also identified the following as enterprise-wide root causes of its high-risk designation: disjointed strategic planning; poorly defined roles, responsibilities, and decision authorities; poor horizontal and vertical integration; lack of reliable data and analysis; ineffective human capital management; and inadequate change management. VA relied on these root cause analyses as the foundational drivers for the VHA Plan for Modernization. However, VA has not used these analyses to develop and prioritize appropriate milestones and metrics in the action plan. Monitoring: not met. Since the March 2018 action plan lacked specific metrics and mechanisms for assessing and reporting progress, it is not clear how VA is monitoring its progress. Demonstrated progress: not met. Our work continues to indicate VA is not yet able to show progress in this area. Since its 2015 high-risk designation, we have made 50 new recommendations in this area of concern, 32 of which were made since our 2017 report was issued. For example, In November 2017, we reported that, due in part to misinterpretation or lack of awareness of VHA policy, VA medical center officials did not always conduct or document timely required reviews of providers when allegations were made against them. We also found that VHA was unable to reasonably ensure appropriate reporting of providers to oversight entities such as state licensing authorities. As a result, VHA’s ability to provide safe, high quality care to veterans is hindered because other VA medical centers, as well as non-VA health care entities, may be unaware of serious concerns raised about a provider’s care. We recommended that VHA direct medical centers to document and oversee reviews of providers’ clinical care after concerns are raised, among other recommendations. All of our recommendations remain open. As of January 2019, VA estimated completing the recommended revisions to its policy and audit processes in August 2019 and August 2020, respectively. In July 2018, we reported that VA collected data related to employee misconduct and disciplinary actions, but data fragmentation, reliability issues, and inadequate guidance impeded department-wide analysis of those data. Thus, VA management is hindered in making knowledgeable decisions regarding the extent of misconduct and how it was addressed. We recommended that VA develop and implement guidance to collect complete and reliable misconduct and associated disciplinary-action data department-wide, whether through a single information system, or multiple interoperable systems. VA concurred with this priority recommendation, which remains open. VA reported that it expects to implement one or more information systems that will collect misconduct and associated disciplinary action data in January 2020. Inadequate oversight and accountability. Since our 2017 High-Risk Report, ratings for one criterion improved and four remain unchanged for this area of concern as of March 2019. Leadership commitment: partially met. VA has made organizational changes, including establishing the Office of Integrity, to standardize and streamline the agency’s oversight of its programs and personnel. However, since 2017, the lack of stability in the Under Secretary for Health position has hindered its ability to demonstrate sustained commitment to improving this area of concern. Capacity: not met. VA has begun to implement capacity-building initiatives directed at improving oversight and accountability. For example, VHA’s Office of Internal Audit and Risk Assessment, a key component of the department’s oversight and accountability model, began conducting audits in 2018. However, according to VA’s action plan, the department has yet to allocate resources for this office, such as sufficient staff to carry out its activities. Action plan: partially met. In March 2019, the rating for this criterion improved to partially met. In 2018, VA conducted an analysis of the root causes contributing to findings of inadequate oversight and accountability, an important step in identifying the underlying factors contributing to this area of concern. addresses gaps in physician staffing, including those for mental health providers, which may affect veterans’ access to care, among other issues. We recommended that VHA should develop and implement a process to accurately count all physicians providing care at each medical center, including physicians who are not employed by VHA. VHA did not concur with this recommendation, which we reiterated in our priority recommendation letter. In a series of reports from 2012 through 2018, GAO found VA’s wait time data unreliable for primary and specialty care, as well as for care in the community. GAO also found that VA did not measure the full wait times that veterans experience in obtaining care across these settings. Specifically, in December 2012, we made two recommendations to VA to improve the reliability and oversight of wait time measures, both of which are designated as priority, and remain open. Similarly, in June 2018, we reported that VHA could not systematically monitor the timeliness of veterans’ access to Veterans Choice Program care because it lacked complete, reliable data to do so. Specifically, we found (1) a lack of data on the timeliness of accepting referrals and opting veterans in to the program, (2) inaccuracy of clinically indicated dates, which are used to measure the timeliness of care, and (3) unreliable data on the timeliness of urgent care. We recommended that VA take steps to improve the timeliness and accuracy of data on veterans’ wait times for care and its oversight of the future community care program that will consolidate other community care programs with the Veterans Choice Program, whose authority sunsets on June 6, 2019. VA concurred with eight of the 10 recommendations related to these findings, all of which remain open. VA reported that, in order to improve wait times data accuracy under the Veterans Community Care Program, it intends to implement several initiatives through September 2019. In September 2018, we reported on the timeliness of third-party administrators’ payments to community providers under VA’s largest community care program, the Veterans Choice Program. Although VA has taken steps to improve the timeliness of claim payments to these providers, VA is not collecting data or monitoring compliance with third-party administrators’ customer service requirements for provider calls. This could adversely affect the timeliness with which community providers are paid, possibly making them less willing to participate and affecting veterans’ access to care. We recommended that VA collect data on and monitor compliance with its requirements pertaining to customer service for community providers. VA agreed with the recommendations, but has not yet implemented them. In November 2018, we reported that VHA’s suicide prevention media outreach activities declined in recent years due to leadership turnover and reorganization. Additionally, we found that VHA did not assign key leadership responsibilities or establish clear lines of reporting for its suicide prevention media outreach campaign, which hindered its ability to oversee the campaign. In April 2019, VA implemented one of the recommendations by providing a new oversight plan for its suicide prevention media outreach campaign. It plans to implement the remaining recommendation by working with communications experts to develop metrics, targets, and an evaluation strategy to improve its outreach efforts. In April 2019, we reported that VHA’s appraisal process for assessing medical center director performance relies heavily on medical center performance information. VHA designed the Strategic Analytics for Improvement and Learning (SAIL) system to provide internal benchmarking of medical center performance and to promote high quality health care delivery across its system of regional networks and medical centers. SAIL was evaluated in 2014 and 2015 by VHA and an external contractor, respectively, but VHA has not assessed the recommendations from those evaluations, or taken action on them. The evaluations, which found issues related to the validity and reliability of SAIL and its ratings for measuring performance and fostering accountability, together included more than 40 recommendations for improvement. Without ensuring that the recommendations resulting from these previous evaluations are assessed and implemented as appropriate, the identified deficiencies may not be adequately resolved, and VHA’s ability to hold officials accountable for taking the necessary actions may be diminished. VA concurred with the two recommendations we made to address these findings, both of which remain open. Information technology challenges. Since our 2017 High-Risk Report, ratings for one criterion regressed, one improved, and three remain unchanged this area of concern as of March 2019. Leadership commitment: not met. In March 2019, the rating for this criterion declined to not met. In January 2019, the Senate confirmed a new VA Chief Information Officer (CIO). This is the fourth official to lead VA’s IT organization since our 2017 High-Risk Report, and the frequent turnover in this position raises concerns about VA’s ability to address the department’s IT challenges. Capacity: not met. In May 2018, VA awarded a contract to acquire the same commercial electronic health record system as the Department of Defense (DOD). However, VA is early in the transition and its actions are ongoing. Additionally, VA has developed a strategy for decommissioning its legacy IT systems, which are tying up funds that could be reallocated for new technology to enable improved veteran care, but has made limited progress in implementing this effort. Action plan: partially met. In March 2019, the rating for this criterion improved to partially met. In 2018, VA conducted an analysis to identify the root causes of IT challenges, which informed the goals in its action plan. However, VA’s action plan contained significant information gaps, including missing interim milestone dates. These information gaps raise questions about VA’s commitment to addressing IT-related root causes and need to be addressed before we can consider this criterion met. Monitoring: not met. The March 2018 action plan lacked specific metrics and mechanisms for assessing and reporting progress. Demonstrating progress: not met. Our work continues to indicate VA is not yet able to show progress in this area. Since its 2015 high-risk designation, we have made 14 new recommendations in this area, 12 of which were made since our 2017 report was issued. For example: In June 2017, to address deficiencies we found related to VA’s pharmacy system, we recommended that VA take six actions to provide clinicians and pharmacists with improved tools to support pharmacy services to veterans and reduce risks to patient safety. This included assessing the extent to which the interoperability of VA and DOD’s pharmacy systems impacts transitioning service members. VA generally concurred with these recommendations, all of which remain open. In April 2019, we testified that from 2001 through 2018, VA pursued three efforts to modernize its health information system— the Veterans Health Information Systems and Technology Architecture (VistA). (See Fig. 2.) However, these efforts resulted in high costs, created challenges ensuring the interoperability of health data, and ultimately did not result in a modernized VistA. Specifically, in December 2017, we reported that VA obligated over $1.1 billion for contracts with 138 contractors during fiscal years 2011 through 2016 for two modernization initiatives, an Integrated Electronic Health Record program with the DOD and VistA Evolution. We have ongoing work that examines the cost to VA of VistA and the department’s actions to transition from VistA to a new electronic health record system. Regarding the department’s most recent effort, the Electronic Health Record Modernization, we testified in April 2019 that the governance plan for this program was not fully defined, which could jeopardize its fourth attempt to modernize its electronic health record system. VA plans to implement the same electronic health record system the DOD is currently deploying. The new system is intended to be the authoritative source of clinical data to support improved health, patient safety, and quality of care provided by VA. VA has not fully implemented our priority recommendation calling for the department to define the role of the Interagency Program Office in the governance plans for acquisition of the department’s new electronic health record system. VA concurred with this recommendation and reported that the Joint Executive Committee, a joint governance body, approved a role for the Interagency Program Office, but as of April 2019 VA has yet to provide us with documentation of this development. We also testified in April 2019 that VA has not yet fully addressed the recommendation we made in September 2014 to expedite the process for identifying and implementing an IT system for the Family Caregiver Program. We reported in September 2014 that the Family Caregiver Program, which was established to support family caregivers of seriously injured post-9/11 veterans, has not been supported by an effective IT system. Specifically, we reported that, due to limitations with the system, the program office did not have ready access to the types of workload data that would allow it to routinely monitor workload problems created by the program. Without such information, the program’s workload issues could persist and impact the quality and scope of caregiver services, and ultimately the services that veterans receive. VA concurred with our recommendation and subsequently began taking steps to implement a replacement system. However, the department has encountered delays and reported recently initiating an effort to implement a new IT system to support the program based on existing commercially available software. We have ongoing work to evaluate VA’s effort to acquire a new IT system to support the Family Caregiver Program. Inadequate training for VA staff. Since our 2017 High-Risk Report, ratings for one criterion improved and four remain unchanged for this area of concern as of March 2019. Leadership commitment: not met. VA officials have reported progress in establishing a process to develop an enterprise-wide annual training plan to better ensure that VA staff are adequately trained to provide high-quality care to veterans. However, the actions necessary to complete and implement this training plan are not reflected in VA’s March 2018 action plan for the training area of concern, raising questions about the process through which it will be developed. The lack of progress in setting clear goals for improving training demonstrates that VA lacks leadership commitment to address our concerns in this area. Capacity: not met. VA has created working groups and task forces— such as the Learning Organization Transformation Subcommittee in the National Leadership Council—with specific responsibilities. However, VA’s ability to demonstrate capacity is limited because, according to VA’s March 2018 action plan, the department relies on external contractor support services to meet training goals. Action plan: partially met. In March 2019, the rating for this criterion improved to partially met. VA completed a root cause analysis for training deficiencies, which informed the goals underlying its action plan. However, the action plan continues to have deficiencies identified in 2017. For example, not all goal descriptions correspond to planned actions and the action plan lacks detail about how and which data will be collected to assess progress. Monitoring: not met. The March 2018 action plan lacked specific metrics and mechanisms for assessing and reporting progress. Demonstrated progress: not met. Our work continues to indicate that VA is not yet able to show progress in this area. Since its 2015 designation, we have made 11 new recommendations in this area of concern, 3 of which were made since our 2017 report was issued. For example, in April 2018 we reported that, while the department has recommended training for patient advocates—staff members who receive and document feedback from veterans or their representatives—it has not developed an approach to routinely assess their training needs or monitored training completion. The failure to conduct these activities increases VA’s risk that staff may not be adequately trained to advocate on behalf of veterans. As a result, we recommended VHA develop an approach to routinely assess training needs and monitor training completion. VA concurred with our recommendations, which remain open. Unclear resource needs and allocation priorities. Since our 2017 High-Risk Report, ratings for one criterion improved and four remain unchanged for this area of concern as of March 2019. Leadership commitment: partially met. In December 2017, a VA Chief Financial Officer (CFO) was confirmed after the department spent over 2.5 years under an interim CFO. In addition, VA is in the process of establishing a new office to estimate workforce resource requirements. Capacity: not met. VA has established functions intended to inform cost analyses of major VA initiatives, including a new financial management process to replace its outdated financial systems. However, it is unclear in its action plan the extent to which VA has identified the resources needed to establish and maintain these functions. Action plan: partially met. In March 2019, the rating for this criterion improved to partially met. Since our 2017 High-Risk Report, VA conducted a root cause analysis of this area of concern. However, VA’s action plan lacks metrics for monitoring progress and does not include all of VA’s ongoing actions, such as efforts to assess current and future regional demand for veterans’ health care services. Monitoring: not met. Since VA’s action plan lacks specific metrics and mechanisms for assessing and reporting progress, it is not clear how VA is monitoring its progress. Demonstrating progress: not met. Our work continues to indicate VA is not yet able to show progress in this area. Since its 2015 designation, we have made 16 new recommendations in this area of concern, 10 of which were made since our 2017 report. For example: In May 2017, we reported identifying several limitations with VA’s clinical productivity metrics and statistical models for tracking clinical efficiency; this limits VA’s ability to assess whether resources are being used effectively to serve veterans. Specifically, we found that productivity metrics may not account for all providers or clinical services, reflect the intensity of clinical workload, and reflect providers’ clinical staffing levels. Additionally, we found that efficiency models may also be adversely affected by inaccurate workload and staffing data. As a result, VA cannot systematically identify best practices to address low productivity and inefficiency as well as determine the factors VA medical centers commonly identify as contributing to low productivity and inefficiency. We made four recommendations to address these findings; three of which VA implemented in the spring of 2018 by improving productivity metrics and staffing and workload data. To implement the remaining recommendation, VA should establish a process to oversee medical centers’ plans for addressing low clinical productivity and inefficiency. In August 2018 we reported that VA medical centers face challenges operating their Sterile Processing Services programs— notably, addressing workforce needs, such as lengthy hiring time frames and limited pay and professional growth potential. VHA’s Sterile Processing Services workforce challenges pose a potential risk to VA medical centers’ ability to ensure access to sterilized medical equipment. Until VHA examines these workforce needs, VHA won’t know whether or to what extent the reported challenges adversely affect VA medical centers’ ability to effectively operate their Sterile Processing Services programs and ensure access to safe care for veterans. We recommended that VA examine workforce needs and take action based on this assessment, as appropriate. VA concurred with this recommendation, which remains open. VA Acquisition Management Was Added to GAO’s High-Risk List in 2019 In light of numerous contracting challenges that we have identified, and given the significant investment in resources to fulfill its critical mission of serving veterans, we added VA acquisition management as a new high- risk area in 2019. VA has one of the most significant acquisition functions in the federal government, both in dollar amount of obligations and number of contract actions. Specifically, about a third of VA’s discretionary budget in fiscal year 2018, or about $27 billion, has been used to contract for goods and services. We have identified challenges in the following areas of concern related to VA’s acquisition management: (1) outdated acquisition regulations and policies; (2) lack of an effective medical supplies procurement strategy; (3) inadequate acquisition training; (4) contracting officer workload challenges; (5) lack of reliable data systems; (6) limited contract oversight and incomplete contract file documentation; and (7) leadership instability. Outdated acquisition regulations and policies. VA’s procurement policies have historically been outdated, disjointed, and difficult for contracting officers to use. In September 2016, we reported that (1) the acquisition regulations contracting officers currently follow have not been fully updated since 2008 and (2) VA had been working on completing a comprehensive revision of its acquisition regulations since 2011. VA’s delay in updating this fundamental source of policy has impeded the ability of contracting officers to effectively carry out their duties. We recommended in September 2016 that VA identify measures to expedite the revision of its acquisition regulations and clarify what policies are currently in effect. VA concurred with this priority recommendation and, as of January 2019, had rescinded or re-issued updated policy memoranda for all information letters, which VA previously used to provide guidance that was temporary in nature. VA has also made some progress in updating its acquisition regulations, but more work remains to be done over the next several years. As of April 2019, VA reports that 15 of the 41 parts in its acquisition regulations update were published as final rules, 10 were issued as proposed rules for public comment, and the remainder are at an earlier stage of the rulemaking process. All parts are scheduled to be out for public comment by March 2020, but the final rules are not expected to be published until April 2021. Lack of an effective medical supplies procurement strategy. VA’s program for purchasing medical supplies has not been effectively executed, nor is it in line with practices at leading hospitals. To support more efficient purchasing of medical supplies for its 172 medical centers that serve the needs of about 9 million veterans, VA launched the Medical Surgical Prime Vendor-Next Generation (MSPV-NG) program in December 2016. MSPV-NG was part of VA’s overall effort to transform its supply chain and achieve $150 million in cost avoidance. In November 2017, we reported that VA’s approach to developing its catalog of supplies was rushed and lacked key stakeholder involvement and buy-in. It also relied on establishing non-competitive blanket purchase agreements for the overwhelming majority of products, resulting in low utilization by medical centers. VA had set a target that medical centers would order 40 percent of their supplies from the MSPV-NG catalog, but utilization rates were below this target with a nationwide average utilization rate across medical centers of about 24 percent as of May 2017. This low utilization adversely affected VA’s ability to achieve its cost avoidance goal. We recommended in November 2017 that VA develop, document, and communicate to stakeholders an overarching strategy for the program. VA concurred with this priority recommendation and is developing strategies to address it. First, in February 2019, VA developed and documented a new, overarching acquisition strategy for its Medical Surgical Prime Vendor (MSPV) program, and has begun the process of communicating it to key stakeholders, including clinical and logistics staff. Further, VA is developing a separate strategy to involve clinicians in developing requirements with plans to complete a pre-pilot of this strategy by September 2019. In response to a congressional request to assess these and other program changes, we recently began a review of VA’s MSPV program. Inadequate acquisition training. VA acquisition training, at times, has not been comprehensive nor provided to staff that could benefit from it. A 2006 statute required, and a 2016 Supreme Court decision (Kingdomware Technologies, Inc. v. United States) reaffirmed, that VA is to give preference to veteran-owned small businesses when competitively awarding contracts—a program known as Veterans First. In September 2018, we reported that training on VA’s Veterans First policy did not address some of its more challenging aspects. For example, many of the contracting officers we interviewed were uncertain about how to balance the preference for veteran-owned small businesses with fair and reasonable price determinations when lower prices might be found on the open market. In addition, VA provided several installments of online training sessions on the Veterans First policy to contracting officers but did not make them mandatory. As a result, only 52 percent of VA contacting officers completed the follow-up training by the spring of 2018. We recommended in September 2018 that VA provide more targeted training to contracting officers on how to implement the Veterans First policy, particularly in the area of making fair and reasonable price determinations, and assess whether this training should be designated as mandatory. VA concurred, and in April 2019, VA’s Chief Acquisition Officer (CAO) stated that VA is taking steps to make this training mandatory. VA also reported that its Acquisition Academy will provide Veterans First training to all contracting staff on May 30, 2019. Contracting officer workload challenges. The majority of our reviews since 2015 have highlighted workload as a contributing factor to the challenges that contracting officers face. Most recently, in September 2018, we reported that about 54 percent of surveyed VA contracting officers said their workload was not reasonable and found that workload stresses have exacerbated the struggles that they face implementing the department’s Veterans First policy. In addition, in September 2016, we reported that VHA contracting officers processed a large number of small dollar-value actions to support medical center operations, many of which involve emergency procurements of routine items to support immediate patient care. Contracting officers and the department’s Acting CAO told us that these frequent and urgent small-dollar transactions reduce contracting officers’ efficiency and ability to take a strategic view of VHA’s overarching procurement needs. We reported in November 2017 that emergency procurements accounted for approximately 20 percent—$1.9 billion—of VHA’s overall contract actions in fiscal year 2016. Figure 3 shows the percent of VHA contract actions designated as emergencies in fiscal year 2016 by each network contracting office. We recommended in November 2017 that VHA network contracting offices work with medical centers to identify opportunities to more strategically purchase goods and services frequently purchased on an emergency basis. VA concurred with this recommendation and recently offered to provide us with a demonstration of the supply chain dashboard that VA uses to track items purchased on an emergency basis, which we plan to attend by the end of May 2019. VA also agreed to conduct an analysis of its purchase card spending to identify items that should be purchased through its MSPV program. VA expects to complete this analysis by July 2019. If implemented, this would allow for both greater contracting officer efficiency and cost savings. For example, based on a similar recommendation we made in 2012, VA began more systematically employing strategic sourcing in FY 2013, and in subsequent fiscal years reported about $10 billion in savings over a 5-year period. Lack of reliable data systems. The lack of accurate data has been a long-standing problem at VA. In September 2016, we reported that VA had not integrated its contract management and accounting systems, resulting in duplicative efforts on the part of contracting officers and increased risk of errors. We and VA’s Inspector General each recommended that VA perform data checks between the two systems. VA concurred with this recommendation and some VA contracting organizations have made efforts to address this risk. Further, VA reported in March 2019, that it plans to adopt a new integrated financial and contract management system, which it plans to install VA-wide over a 9- year period, with the final site receiving the system in 2027. Limited contract oversight and incomplete contract file documentation. VA has had difficulty ensuring that its contracts are properly monitored and documented. In September 2018, we reported that, although VA obligated $3.9 billion to veteran-owned small businesses in fiscal year 2017, its contracting officers were not effectively monitoring compliance with key aspects of the department’s Veterans First policy, such as limits on subcontracting (which ensure that the goal of the program—to promote opportunities for veteran-owned businesses—is not undermined). In many cases, we found that clauses requiring compliance were not included in the VA’s contracts and orders with veteran businesses because the contracting officers either forgot to include them or were unaware of the requirement. The contracting officers we spoke with also said that they do not have sufficient time or knowledge to conduct oversight. Through limited reviews, VA has identified a number of violations that would warrant a broader assessment of the fraud risks to the program. We recommended in September 2018 that VA establish a mechanism to ensure that mandatory subcontracting-related clauses be consistently incorporated into set-aside contracts with veteran-owned businesses and that VA conduct a fraud risk assessment for the Veterans First program. VA concurred with these recommendations and is taking steps to implement them. For example, VA reported in April 2019 that it had made modifications to its electronic contract management system to ensure the clauses would be included in set-aside contracts and anticipated completing testing of the modifications in May 2019. We also reported in September 2016 that a number of VA contract files we reviewed were missing key documents, increasing the risk that key processes and regulations were not followed. We recommended that VA focus its internal compliance reviews to ensure that required contract documents are properly prepared and documented. VA concurred with this recommendation. Since then, VA has made policy changes that revised its processes for compliance reviews of contract documentation. We are currently following up with VA to obtain the results of its compliance reviews to determine if VA has fully implemented this recommendation. Leadership instability. We have previously reported, most recently in September 2018, that procurement leadership instability has made it difficult for the VA to execute and monitor the implementation of key acquisition programs and policies. For example, changes in senior procurement leadership, including the CAO and VHA’s Chief Procurement and Logistics Officer, occurred during the implementation of MSPV-NG and similar instability in leadership affected the MSPV-NG program office itself. Overall, the MSPV-NG program office has had four directors, two of whom served in an acting capacity, since its inception in 2014. To address this instability, we recommended in November 2017 that VA appoint a non-career employee as the CAO and prioritize the hiring of the MSPV-NG program office’s director position on a permanent basis. VA concurred with these recommendations and implemented them in 2018. Stable leadership should help bring consistent and much needed direction to the MSPV-NG program, but we recently identified other areas within the VA where sustained leadership is also needed. For instance, in September 2018, we reported there have been six Acting Directors within the past 2 and a half years within an oversight office that helps assess whether VA is in compliance with aspects of its Veterans First policy. Ratings for the VA Disability High-Risk Areas Either Remained Unchanged or Regressed in 2019 We designated improving and modernizing federal disability programs as high risk in 2003. An estimated one in six working-age Americans reported a disability in 2010. Many of these Americans need help finding or retaining employment, or rely on cash benefits if they cannot work. Three of the largest federal disability programs—one run by VA— disbursed about $270 billion in cash benefits to 21 million people with disabilities in fiscal year 2017. However, federal disability programs, including VA’s, struggle to meet their needs. In particular, VA struggles to manage its disability claims workloads, and, when determining whether individuals qualify for disability benefits, VA relies on outdated eligibility criteria. Managing disability claims workloads. Since our 2017 High-Risk Report, our assessment of ratings for all five criteria remains unchanged for this area of concern for VA as of March 2019. Leadership commitment: met. VA has maintained leadership focus on managing initial disability claims and appeals workloads through various initiatives to improve benefits processing and reduce backlogs. Enhancing and modernizing VA’s disability claims and appeals processes are goals in its 2018–2024 strategic plan. Capacity: partially met. VA has continued building the capacity to process initial disability claims, such as using an electronic system to distribute claims ready for decisions to available staff. On appeals, VA is reforming its process, onboarding hundreds of new staff, and implementing new technology. However, as we reported in March 2018, VA’s appeals plan does not provide reasonable assurance that it will have the capacity to implement the new process and manage risks. VA agreed with our recommendation to better assess risks associated with appeals reform and took some steps to address risks, such as limited testing of the new process. However, as of April 2019 VA has not fully addressed this recommendation. For example, VA has not developed plans to fully address risks, such as veterans choosing more resource-intensive options at higher rates than expected. Action plan: partially met. VA continues to implement plans to reduce the initial disability claims backlog. For appeals reform, VA submitted its appeals plan in November 2017 and provided several progress reports throughout 2018. In March 2018, we reported that VA’s plan for implementing a new disability appeals process did not explain how VA would assess the new process compared to the legacy process, and did not fully address risks associated with implementing a new process. We made two recommendations to improve VA’s disability benefit appeals process, including that VA (1) clearly articulate in its appeals plan how it will monitor and assess the new appeals process compared to the legacy process, and (2) ensure that its appeals plan more fully addresses related risks, given the uncertainties associated with implementing a new process. As of April 2019, VA has taken actions to address our recommendations, although key steps remain. For example, VA has not fully articulated detailed steps and time frames for assessing the relative performance of the new and legacy appeals processes. Without this assessment, VA cannot determine the extent to which the new process will achieve final resolution of veterans’ appeals sooner than the legacy process. Monitoring: partially met. VA monitors the timeliness of initial disability claims and legacy appeals, and has set timeliness goals for some, but not all, of the appeal options under the new process. VA’s plans also signal how it intends to monitor the allocation of staff for concurrent workloads in its legacy and new appeals processes. However, as of April 2019, VA has yet to specify a complete set of balanced goals for monitoring the new and legacy appeals processes (including timely and accurate processing of appeals while ensuring veteran satisfaction). Demonstrated progress: partially met. VA reported it reduced the backlog of initial disability claims from 611,000 in March 2013 to about 81,000 at the end of fiscal year 2018. However, VA’s Inspector General reported in September 2018 that VA overstated its performance by only reporting about 79 percent of the backlog. For appeals, VA addressed some gaps in its plan for implementing appeals reform, in accordance with our 2017 and 2018 recommendations, and has prioritized processing of legacy appeals. However, as of September 2018, VA still had a backlog of about 396,000 legacy appeals. Updating disability benefit eligibility criteria. Since our 2017 High-Risk Report, VA’s ratings for the action plan and monitoring criteria regressed while the other three remain unchanged as of March 2019. Leadership commitment: met. VA has sustained leadership focus on updating its Veterans Affairs Schedule for Rating Disabilities (VASRD)—used to assign degree of disability and compensation levels for veterans with military service-connected injuries or conditions—to reflect advances in medicine and labor market changes. Capacity: partially met. In August 2017, VA officials told us that it had taken actions to hire more staff for the regulations updates and leverage outside researchers to evaluate veterans’ loss of earnings in the current economy. However, as of September 2018, the agency was still working to hire these staff. Moreover, VA’s current earnings loss study covers only 8 of over 900 diagnostic codes and 2 of 15 body systems. VA needs to continue its current hiring and earnings loss planning efforts to ensure it has the capacity to comprehensively update the VASRD. Action plan: partially met. In March 2019, the rating for this criterion declined to partially met. As of April 2019, VA’s efforts to update the VASRD included new plans to conduct earnings loss studies. Veterans Benefits Administration officials stated they completed a study for eight diagnostic codes under two body systems, and the agency is determining whether its current approach for evaluating earnings loss is applicable to updating other diagnostic codes. However, we lowered VA’s prior rating of met to partially met because its latest August 2018 updated plan, issued since our 2017 High-Risk Report, provided limited detail on key planned activities, potentially jeopardizing its third attempt at modernization over the past decade. For example, VA’s plans do not indicate how and when VA will assess the applicability of its current approach, and does not include plans for updating earnings loss information for the remaining diagnostic codes and body systems. Monitoring: partially met. In March 2019, the rating for this criterion declined to partially met. According to VA officials, VA continues to track its progress toward finishing the medical updates by fiscal year 2020 and has updated its project plan to reflect delayed time frames. However, we lowered VA’s prior rating for this criterion from met to partially met because VA’s plans have changed since our last update, and although it is conducting a study to update earnings loss information for some diagnostic codes and body systems, its plan does not include timetables for monitoring these or future updates to earnings loss information. Demonstrated progress: partially met. VA reported that as of December 2018, it promulgated final regulations for 6 of 15 body systems, proposed regulations for 2, and is reviewing draft regulations for the remaining 7. However, VA has fallen about 4 years behind in its efforts to fully update the VASRD and has not completed earnings loss updates. Other Government- Wide High-Risk Areas Have Implications for VA Operations Several other government-wide high-risk areas include VA and its operations. These areas include (1) improving the management of IT acquisitions and operations, (2) strategic human capital management, (3) managing federal real property, and (3) ensuring the cybersecurity of the nation. Improving the management of IT acquisitions and operations. The executive branch has undertaken numerous initiatives to better manage the more than $90 billion that is annually invested in IT across the government. However, our work shows that federal IT investments, including those made by VA, too frequently fail or incur cost overruns and schedule slippages while contributing little to mission-related outcomes. Thus, in 2015, we added improving the management of IT acquisitions and operations to the High-Risk List. To address the portion of the high-risk area for which it is responsible, VA should, among other things, implement our past recommendations on improving IT workforce planning practices and establishing action plans to modernize or replace obsolete IT investments. In August 2018, for example, we found that VA’s policies did not fully address the role of its CIO consistent with federal laws and guidance in the areas of IT workforce, IT strategic plan, IT budgeting, and IT investment management. Until VA fully addresses the role of the CIO in all of its policies, it will be limited in addressing longstanding IT management challenges. We recommended that VA’s IT management policies address the role of the CIO for key responsibilities in the four areas we identified. VA concurred with this recommendation, which remains open. Strategic human capital management. This area was added to our High-Risk List in 2001 and continues to be at risk today because mission-critical skills gaps both within federal agencies and across the federal workforce are impeding the government from cost-effectively serving the public and achieving results. As of December 2018, VA reported an overall vacancy rate of 11 percent at VHA medical facilities, including vacancies of over 24,000 medical and dental positions and around 900 human resource positions. Also, with 32 percent of the VA workforce eligible to retire in the next 5 fiscal years, VA must address these mission-critical skill gaps and vacancies that we continue to identify in our work. In December 2016, for example, we found that VHA’s limited human resources capacity combined with weak internal control practices has undermined VHA’s human resources operations and its ability to improve delivery of health care services to veterans. Further, VHA is challenged by inefficiencies in its performance management processes, including the lack of a performance appraisal IT system, which prevents it from identifying trends and opportunities for improvement. VHA can better support medical centers by establishing clear lines of accountability for engagement efforts, collecting and leveraging leading practices, and addressing barriers to improving engagement. We made three recommendations to VA to improve its performance management system. VA partially concurred with these recommendations, which remain open. Managing federal real property. Since federal real property management was placed on the High-Risk List in 2003, the federal government has given high-level attention to this issue. However, federal agencies, including VA, continue to face long-standing challenges, including (1) effectively disposing of excess and underutilized property, (2) relying too heavily on leasing, (3) collecting reliable real property data for decision making, and (4) protecting federal facilities. In January 2019, for example, we reported that VA has enhanced its data collection on vacant properties, but the agency does not collect information needed to track and monitor disposal projects at the headquarters level. Without information on the status of disposal projects, VA cannot readily track and monitor its progress and identify areas where facilities’ managers may need additional assistance. As a result, we recommended that VA improve its procedures related to disposal of excess and underutilized property to help local facility managers plan, implement, and execute projects to dispose of those properties. In addition, VA should collect key information on the status of these disposal projects to help manage the process and identify areas where management attention is needed. VA concurred with the three recommendations we made related to these findings, all of which remain open. Ensuring the cybersecurity of the nation. We have designated information security as a government-wide high-risk area since 1997. We expanded this high-risk area in 2003 to include protection of critical cyber infrastructure and, in 2015, to include protecting the privacy of personally identifiable information. Federal agencies and our nation’s critical infrastructures are dependent on IT systems and electronic data to carry out operations and to process, maintain, and report essential information. The security of these systems and data is vital to public confidence and national security, prosperity, and well- being. Because many of these systems contain vast amounts of personally identifiable information, agencies must protect the confidentiality, integrity, and availability of this information. In addition, they must effectively respond to data breaches and security incidents when they occur. In May 2016, for example, we found that VA had developed a risk assessment for their selected high-risk systems, but had not always effectively implemented access controls. These control weaknesses included those protecting system boundaries, identifying and authenticating users, authorizing access needed to perform job duties, and auditing and monitoring system activities. Weaknesses also existed in patching known software vulnerabilities and planning for contingencies. An underlying reason for these weaknesses is that the key elements of information security programs had not been fully implemented. VA concurred with all of our five recommendations related to improving its cybersecurity controls. However, two recommendations—which specifically call for the department to conduct security control assessments and develop a continuous monitoring strategy—remain open. In November 2018, the department’s inspector general reported that VA had made progress in developing, documenting, and distributing policies and procedures to support its security program, but identified IT security as a major management challenge due to the persistence of deficiencies. For example, the inspector general identified significant deficiencies related to access, configuration management, change management, and service continuity. In addition, VA’s financial statement auditor reported deficiencies in the department’s IT security controls as a material weakness for financial reporting purposes. The auditor has reported IT security controls as a material weakness for more than 10 years. VA’s Transformational Efforts Are Ongoing Since his confirmation in July 2018, Secretary Wilkie has demonstrated his commitment to addressing the department’s high-risk designations by, among other things, creating an office to direct an integrated approach for high-risk concerns and communicating to VA leaders the importance of addressing our recommendations. Additionally, VA leadership has also encouraged senior leaders to meet with GAO subject matter experts from acquisition, performance, human capital, and financial management, among other areas, to discuss leading practices and VA’s modernization efforts. In addition, senior leaders from GAO and VA meet regularly to identify and address the root causes of high-risk issues, and discuss the status of our recommendations and VA’s efforts to address them. Fully addressing these issues will require sustained leadership attention on these issues as well as leadership stability—something that VA has not had in recent years. In particular, in the 2 years prior to Secretary Wilkie’s confirmation, VA experienced leadership instability with senior- level vacancies in key positions, including the Under Secretary for Health, CIO, and Deputy Under Secretary for Health for Community Care. In addition to sustained leadership, VA must develop action plans for addressing the high-risk issues. As noted earlier, VA officials have stated that they are currently working to address our high-risk concerns through the implementation of the VHA Plan for Modernization. The plan, which identifies high-level implementation targets through 2020, provides a framework to address the Secretary’s four priorities: (1) improving training and customer service; (2) implementing the VA MISSION Act and improving veterans’ access to care; (3) connecting the VA’s electronic health records system to the DOD’s to ensure a continuum of care for transitioning service members; and (4) transforming VA’s business systems. As part of this effort, VA is focused on “10 lanes of effort,” including transitioning to the same electronic health record system the DOD is currently deploying, and transforming its business systems— including its human resource management, finance and acquisition management, and supply chain functions—to improve the quality and availability of services at VA medical centers. In closing, VA has launched several significant efforts to address many of the underlying management challenges it faces, including transforming its electronic health record and financial management systems, updating its medical surgical prime vendor program, and implementing the VA MISSION Act. Any one of these efforts would be a significant undertaking for an agency given their scope, time frames, and costs, and VA is attempting to concurrently implement them. If successful, these efforts could be transformative for VA. Sustained congressional oversight of VA’s efforts will also be needed. We stand ready to support this oversight through continued monitoring of VA’s efforts as it ensures that the modernization efforts integrate and address many of the concerns that led to the designation of various VA areas as high risk. Chairman Pappas, Ranking Member Bergman, and Members of the Subcommittee, this concludes my statement. I would be pleased to respond to any questions you may have. GAO Contacts and Staff Acknowledgments For further information about this statement, please contact Debra A. Draper at (202) 512-7114 or draperd@gao.gov or Sharon M. M. Silas at (202) 512-7114 or silass@gao.gov for VHA health care issues; Shelby S. Oakley at (202) 512-4841 or oakleys@gao.gov for VA acquisition management issues; or Elizabeth H. Curda at (202) 512-7215 or curdae@gao.gov for VA disability claims issues. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Key contributors to this statement were Ann Tynan, Mark Bird, David Bruno, Keith Cunningham, Cathleen Hamann, Lisa Gardner, Steven Lozano, William Reinsberg, Maria Storts, Jamie Whitcomb, Amanda Cherrin (Analyst-in-Charge), Kate Tussey, Jeff Hartnett, and Teague Lyons. Vikki Porter and Jacquelyn Hamilton also contributed to this statement. This is a work of the U.S. government and is not subject to copyright protection in the United States. The published product may be reproduced and distributed in its entirety without further permission from GAO. However, because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately.

Why GAO Did This Study VA is responsible for providing benefits and services to veterans, including health care, disability compensation, and various types of financial assistance. In fiscal year 2019, VA received a total budget of $201.1 billion and a discretionary budget of $86.6 billion—the largest in VA's history—to carry out its mission. GAO, along with the VA Inspector General and other entities, continues to identify significant deficiencies in VA's governance structures and operations—all of which can affect the care provided to our nation's veterans. This testimony focuses on the status of VA's efforts to address GAO's high-risk designations and open GAO recommendations in the following areas: VA health care, acquisition management, and disability claims workloads and benefit eligibility criteria, among other areas. It is primarily based on GAO's March 2019 high-risk update and a body of work that spans more than a decade. What GAO Found The Department of Veterans Affairs (VA) has longstanding management challenges. As a result, GAO added several VA programs to its High-Risk List. This list focuses attention on government operations that are most vulnerable to fraud, waste, abuse, or mismanagement, or in need of transformation. These include managing risks and improving VA health care, VA acquisition management, and improving and modernizing VA disability programs, including managing claims and updating eligibility criteria. VA health care was designated high risk in 2015 due to concerns about VA's ability to ensure the cost-effective and efficient use of resources to improve the timeliness, quality, and safety of health care for veterans. GAO identified five areas of concern: (1) ambiguous policies and inconsistent processes; (2) inadequate oversight and accountability; (3) information technology challenges; (4) inadequate training for VA staff; and (5) unclear resource needs and allocation priorities. VA's efforts to address each of these areas have been impeded by leadership instability. However, since his July 2018 confirmation, Secretary Wilkie has demonstrated his commitment to address the department's high-risk designations. His actions to date have allowed the department to maintain its leadership commitment rating of partially met in GAO's 2019 High-Risk update. VA also partially met the action plan criteria. As of March 2019, it did not meet the other three criteria for removal from the High-Risk List (agency capacity, monitoring, and demonstrated progress). This is, in part, because GAO continues to have audit findings that illustrate that the five areas of concern have not been fully addressed. For example: In a series of reports from 2012 through 2018, GAO found VA's wait time data unreliable for primary and specialty care as well as for care in the community. GAO also found that VA did not measure the full wait times that veterans experience in obtaining care across these settings. In November 2017, GAO reported that VA medical center officials did not always conduct or document timely required reviews of providers when allegations of wrongdoing were made against them. In April 2019, GAO found that VA's governance plan for modernizing its electronic health record system was not fully defined, potentially jeopardizing its fourth attempt at modernization. In April 2019, GAO reported that VA's appraisal process for assessing medical center director performance relies heavily on a system with long-identified deficiencies that remain unaddressed, thus diminishing VA's ability to hold officials accountable. In its 2019 High-Risk Report, GAO added VA acquisition management as a high-risk area in light of the department's numerous contracting challenges and the significant federal investment in serving veterans. To date, GAO has identified challenges in the following areas: (1) outdated acquisition regulations and policies; (2) lack of an effective medical supplies procurement strategy; (3) inadequate acquisition training; (4) contracting officer workload challenges; (5) lack of reliable data systems; (6) limited contract oversight and incomplete contract documentation; and (7) leadership instability. For example, as of May 2019, VA does not have updated acquisition regulations and officials expect to have a full update by 2021; a process which has been in place since 2011. GAO designated improving and modernizing federal disability programs, including VA's program, as high risk in 2003. GAO identified two areas of concern related to VA: (1) managing disability claims workload and (2) updating disability benefit eligibility criteria. As a result of these concerns, veterans may not have their disability claims and appeals processed in a timely manner. GAO reported in March 2018 that VA is making a major effort to reform its appeals process by onboarding new staff and implementing new technology. However, its appeals planning process does not provide reasonable assurance that it will have the capacity to successfully implement the new process and manage risks. VA agreed with GAO's recommendation to better assess risks associated with appeals reform. VA leadership has committed to addressing GAO's high-risk concerns and has launched several transformational efforts. For example, VA is currently implementing the Veterans Health Administration Plan for Modernization, a framework that aims to modernize the department, as well as the VA MISSION Act of 2018. This Act requires VA to consolidate programs that allow veterans to receive care outside VA. If successful, these efforts could be transformative for VA. However, such success will only be achieved through sustained leadership attention and detailed action plans that include metrics and milestones to monitor and demonstrate VA's progress. Sustained congressional oversight will also be essential. What GAO Recommends Since 2000, GAO has made more than 1,200 recommendations to reduce VA's high-risk challenges, and VA has implemented approximately 70 percent. GAO will continue to monitor VA's progress in implementing the remaining open recommendations.

gao_GAO-19-426

gao_GAO-19-426_0

Background Overview of U.S. Pipeline System The national pipeline system consists of more than 2.7 million miles of networked pipelines transporting natural gas, oil, and other hazardous liquids. Natural gas and hazardous liquid pipelines—primarily buried underground in the continental United States—run under remote and open terrain, as well as densely-populated areas. There are three main types of pipelines based on the types of materials transported: Hazardous liquid: About 216,000 miles of hazardous liquid pipeline transport crude oil, diesel fuel, gasoline, jet fuel, anhydrous ammonia, and carbon dioxide. Natural gas transmission and storage: About 319,000 miles of pipeline—mostly interstate—transport natural gas from sources to communities. Natural gas distribution: About 2.2 million miles of pipeline—mostly intrastate—transport natural gas from transmission sites to consumers. Figure 1 depicts the network of hazardous liquid and natural gas transmission pipelines in the United States. More than 3,000 pipeline companies operate the nation’s pipeline systems, which can traverse multiple states and the U.S. borders with Canada and Mexico. Many pipeline systems are comprised of the pipelines themselves, as well as a variety of facilities, such as storage tanks, compressor stations, and control centers. Most pipeline systems are monitored through automated industrial control systems or Supervisory Control and Data Acquisition (SCADA) systems using remote sensors, signals, and preprogramed parameters to activate and deactivate valves and pumps to maintain flows within established tolerance levels. Threats to Pipeline Safety and Security Pipeline accidents can occur from a variety of causes, including third- party excavation, corrosion, mechanical failure, control system failure, and operator error. Natural forces, such as floods and earthquakes, can also damage pipelines. Although pipeline releases have caused relatively few fatalities, a single pipeline accident can be catastrophic in terms of public safety and environmental damage. Figure 2 shows notable pipeline accidents since September 2010. According to TSA, pipelines are also vulnerable to physical attacks by crude or unsophisticated tactics, such as rudimentary explosives, arson, or equipment sabotage—largely due to their stationary nature, the volatility of transported products, and the dispersed nature of pipeline networks spanning urban and outlying areas. Threats to the nation’s pipeline systems include sabotage by activists, physical attack by terrorists, and cyber attack or intrusion by nations. In October 2016, environmental activists forced the shutdown of five crude oil pipelines in four states: Minnesota, North Dakota, Montana, and Washington State. Further, in January 2019, the Director of National Intelligence stated that China has the ability to launch cyber attacks that have caused localized, temporary disruptive effects on critical infrastructure—such as disruption of a natural gas pipeline for days to weeks—in the United States. Key Critical Infrastructure Protection Guidance and Presidential Directives Federal policy and public-private plans establish the roles and responsibilities for the protection of critical infrastructure, including pipelines. These policies and public private plans include Presidential Policy Directive /PPD-21 (PPD-21) and the National Infrastructure Protection Plan (NIPP). PPD-21, issued in February 2013, was developed to advance a national unity of effort to strengthen and maintain secure, functioning, and resilient critical infrastructure, which includes pipelines. PPD-21 reflects an all-hazards approach to protecting critical infrastructure, by accounting for the protection of critical infrastructure from natural or manmade threats or incidents. Examples of threats or incidents include natural disasters, cyber incidents, industrial accidents, pandemics, acts of terrorism, sabotage, and destructive criminal activity targeting critical infrastructure. PPD-21 also identifies the 16 critical infrastructure sectors and assigns roles and responsibilities for each sector among nine designated federal sector-specific agencies as shown in Figure 3. While PPD-21 identifies the critical infrastructure sectors and assigns responsibility for each sector’s sector-specific agency, the NIPP outlines critical infrastructure stakeholder roles and responsibilities regarding critical security and resilience. The NIPP describes a voluntary partnership model as the primary means of coordinating government and private sector efforts to protect critical infrastructure. As part of the partnership structure, the designated sector-specific agencies serve as the lead coordinators for security programs of their respective sector. For example, DHS and DOT are designated as co-sector-specific agencies for the transportation systems sector, which includes pipelines. Each sector also has a government coordinating council, consisting of representatives from various levels of government, and many have a sector coordinating council (SCC) consisting of owner-operators of these critical assets or members of their respective trade associations. For example, the Transportation Government Coordinating Council has been established, and the Pipeline Modal SCC has also been established to represent pipeline operators. Pipeline Stakeholder Roles and Responsibilities Protecting the nation’s pipeline systems is a responsibility shared by both the federal government and private industry. As a result, several federal departments, agencies, and the private sector have significant roles in pipeline safety and security. The entities primarily responsible for pipeline safety and security are included below. Transportation Security Administration (TSA). TSA has primary oversight responsibility for the physical security and cybersecurity of transmission and distribution pipeline systems. Within TSA, the Policy, Plans, and Engagement’s Pipeline Security Branch is charged with overseeing its pipeline security program. Pursuant to the Implementing Recommendations of the 9/11 Commission Act of 2007 (9/11 Commission Act), TSA’s Pipeline Security Branch issued voluntary Pipeline Security Guidelines in 2011, and released revised guidelines in March 2018. Further, in accordance with the 9/11 Commission Act, TSA’s Pipeline Security Branch also identifies the top 100 critical pipeline systems in the nation. TSA also ranks the relative risk among these top 100 systems. Additionally, the Pipeline Security Branch is responsible for conducting voluntary security reviews, which assess the extent to which these 100 pipeline systems are following the intent of TSA’s Pipeline Security Guidelines. Pipeline and Hazardous Materials Safety Administration (PHMSA). PHMSA, within DOT, is responsible for regulating the safety of hazardous materials transportation and the safety of pipeline systems, some aspects of which may relate to pipeline security. PHMSA develops regulations for domestic interstate and intrastate natural gas and hazardous liquid pipelines. Its regulatory programs are focused on ensuring safety in the design, construction, operation, and maintenance of pipelines. Under PHMSA’s pipeline safety program, pipeline operators have primary responsibility for ensuring the integrity of their pipelines. PHMSA and some state pipeline safety offices are responsible for conducting inspections to oversee operators’ compliance with federal pipeline safety regulations and other federal requirements. Inspectors from PHMSA’s five regional offices and states are responsible for inspecting nearly 3,000 companies that operate 2.7 million miles of pipelines. Private sector. Although TSA has primary federal responsibility for overseeing interstate pipeline security, private sector and publicly-owned pipeline operators are responsible for implementing asset-specific protective security measures. As we previously reported, since the September 11th terrorists attacks, operators have increased their attention on security by incorporating security practices and programs into their overall business operations. Pipeline operators’ interests and concerns are primarily represented by five major trade associations—the Interstate Natural Gas Association of America, American Gas Association, American Public Gas Association, American Petroleum Institute (API), and Association of Oil Pipe Lines. According to TSA officials, pipeline operators, and association representatives, these associations have worked closely with the federal government on a variety of pipeline security-related issues, including collaborating on TSA’s voluntary standards and information sharing. MOU Annex Delineates Pipeline Security Roles and Responsibilities But Has Not Been Reviewed to Consider Pipeline Security Developments Since 2006 MOU Annex Delineates Pipeline Security and Safety Roles and Responsibilities The MOU Annex delineates TSA and PHMSA mutually agreed-upon pipeline security roles and responsibilities, consistent with their respective missions, and acknowledges that both agencies benefit by sharing each other’s expertise, among other things. Specifically, the MOU Annex identifies 11 program areas, where TSA and PHMSA agreed to coordinate their respective roles and responsibilities. The first program area for example, calls for both agencies to coordinate efforts to identify critical infrastructure, and to share relevant data and observations found during respective safety inspections and security assessments. Another program area addresses coordination in developing transportation security standards, regulations, guidelines, or directives. The MOU Annex further provides that TSA and PHMSA are to seek early and frequent coordination in developing such standards, regulations, guidelines, or directives. They are also to review the adequacy of existing standards in the private and public sector, and identify any gaps that should be addressed through rulemaking, guidelines, or directives, among other items. For a complete listing of the MOU Annex’s 11 program areas, including TSA and PHMSA roles and responsibilities and agreed-upon actions, see appendix I. TSA and PHMSA Do Not Have Timeframes for Reviewing the MOU Annex to Assess Pipeline Security Roles and Responsibilities TSA and PHMSA have both noted various developments that have occurred since 2006 that may affect their roles and responsibilities related to pipeline security. However, the MOU Annex has not been updated since its inception in 2006 to consider incorporating these changes which includes subsequently issued presidential directives, the establishment of the Cybersecurity Infrastructure and Security Agency (CISA), and distinctions between current TSA and PHMSA current inspection operations. As a result, the Annex is not current and may not fully reflect the agencies’ pipeline safety and security-related activities. For example, Homeland Security Presidential Directive/HSPD-7 (HSPD-7), which is cited as an underlying authority in both the 2004 MOU and 2006 MOU Annex was revoked and replaced by PPD-21 in 2013. According to PPD-21, the directive advances a national unity of effort to strengthen and maintain secure, functioning, and resilient critical infrastructure by, among other things, refining and clarifying critical infrastructure-related functions, roles, and responsibilities across the federal government. PPD- 21 further provides, however, that plans developed pursuant to HSPD-7 shall remain in effect until specifically revoked or superseded. According to TSA and PHMSA officials, statements of Executive Branch policy including presidential directives such as PPD-21 include changes that could impact their pipeline security and safety roles and should be considered in any future revisions to the MOU Annex. Further, PHMSA officials also told us that TSA and PHMSA’s roles and responsibilities in identifying critical infrastructure should be reviewed given the establishment of the CISA in November 2018. CISA, formerly the DHS National Protection and Programs Directorate, is responsible for, among other things, coordinating a national effort to secure and protect against critical infrastructure risks. These responsibilities include coordinating with sector-specific agencies to carry out its cybersecurity and critical infrastructure activities. TSA and PHMSA officials stated that they have closely coordinated in identifying critical infrastructure when responding to past national emergencies. For example, TSA identified and provided PHMSA with information on the pipelines that supplied fuel to specific airports during the hurricane seasons in 2017 and 2018. However, PHMSA officials stated that both TSA and PHMSA should consider reviewing how these types of efforts may need to be coordinated with CISA in the future and whether any adjustments to respective roles and responsibilities in the MOU Annex are needed. In addition, representatives from all of the industry associations that we interviewed stated that the agreement should be revised to consider how the establishment of CISA may impact current TSA and PHMSA pipeline security roles and responsibilities. TSA officials stated that they do not believe that the establishment of CISA impacts TSA’s roles and responsibilities for identifying pipeline critical infrastructure. While CISA may or may not have impacts on TSA and PHMSA’s pipeline security roles, reviewing the MOU Annex in light of new developments, such as the CISA, would allow the TSA and PHMSA to determine whether updates are necessary. TSA and PHMSA officials stated that distinctions in current inspections and enforcement operations necessitate a revision to the MOU Annex. The MOU Annex states that agencies are to explore opportunities for collaboration in inspection and enforcement activities. According to TSA and PHMSA officials, they have since explored the possibility for conducting joint activities and found that distinctions in their respective operating environments and roles and responsibilities do not allow for joint inspection and enforcement activities. For example, PHMSA conducts physical inspections of facilities to assess pipeline operators’ compliance with pipeline safety regulatory requirements and relies on a range of enforcement activities, such as civil penalties to ensure that pipeline operators correct safety violations and prevent safety problems. TSA, however, conducts voluntary security assessments of pipeline’s corporate security programs and critical facilities and relies on pipeline operators’ willingness to participate and implement recommended changes to improve pipeline security. As a result, TSA and PHMSA officials stated that pipeline operators are reluctant to participate in a voluntary assessment that might include PHMSA inspectors because they represent a regulatory agency. TSA, PHMSA and industry association representatives we interviewed agreed that the annex should be updated to accurately reflect current distinctions in the agencies’ roles and responsibilities and their respective operating environments. PHMSA officials stated that they had planned to review the MOU Annex in 2018 to assess current roles and responsibilities and determine whether any updates to the MOU Annex were needed, but efforts were delayed because of competing priorities such as addressing the aftermath of major hurricanes in 2017 and 2018. Specifically, TSA and PHMSA had agreed to an initial list of timeframes for reviewing the MOU Annex and these timeframes called for the agencies to complete the MOU Annex revision in 2018. However, as of March 2019, TSA and PHMSA have yet to complete the review and although both agencies stated that the review is ongoing, neither agency could provide updated timeframes for completion. Furthermore, while the Annex recognizes that TSA and PHMSA may propose agreed-upon amendments or modifications to the agreement, it does not call for regular or periodic reviews to identify whether any updates or revisions are needed and, as appropriate, implemented. TSA and PHMSA officials, as well as the industry association representatives we interviewed all reported that the MOU Annex helped to coordinate pipeline security and safety efforts because: (1) it is a signed written agreement that can be readily consulted; (2) it memorialized respective TSA and PHMSA roles and responsibilities for government leaders and staff at the time; and (3) it can be modified or amended as needed. Standards for Internal Control in the Federal Government states that periodic review of policies, procedures, and related control activities should occur to determine their continued relevance and effectiveness in achieving identified objectives or addressing related risks. In addition, documentation of any changes made as a result of such reviews, such as changes to an entity’s roles and responsibilities or in technology, should occur to ensure that such controls are clear over time as staff change within an organization. Standards for project management state that managing a project involves, among other things, developing a timeline with milestone dates to identify points throughout the project to reassess efforts under way to determine whether project changes are necessary. By developing and implementing mutually agreed upon time frames for reviewing the annex and updating it, as appropriate, TSA and PHMSA could better ensure that the roles and responsibilities for TSA and PHMSA remain current. Additionally, including a provision in the annex for periodically reviewing for needed updates would help ensure the agreement consistently reflects relevant and updated information on TSA and PHMSA’s roles and responsibilities. TSA and PHMSA Communicate Their Roles through Guidelines and Other Methods, and Selected Industry Stakeholders Reported the Agencies’ Roles Are Clear TSA and PHMSA have communicated their respective pipeline safety and security roles and responsibilities by issuing pipeline security guidance and safety regulations, issuing a joint advisory bulletin, and maintaining informal contacts with pipeline stakeholders when conducting outreach activities, pipeline security assessments, or safety inspections. TSA security guidelines. TSA’s Pipeline Security Branch first issued its voluntary Pipeline Security Guidelines in 2011, and revised them in March 2018. The guidelines include TSA’s recommendations for pipeline industry security practices, such as establishing a corporate security program, conducting security vulnerability assessments, and identifying critical facilities. The guidelines also recommend facility security and cybersecurity measures, which serve as the basis for the pipeline security assessments conducted by TSA’s Pipeline Security Branch. PHMSA regulations. PHMSA’s Office of Pipeline Safety issues and enforces intrastate and interstate regulations covering aspects of pipeline safety, including the design, construction, operation and maintenance, and spill response for hazardous liquid and gas pipeline facilities, including liquefied natural gas facilities. Advisory bulletins. PHMSA also issues advisory bulletins to communicate safety-related conditions to pipeline operators, and can issue advisory bulletins in coordination with TSA to notify pipeline operators of a security incident. Such bulletins may include identifying the affected operators, describing the threat, and providing information on federal resources for assistance. For example, in response to physical intrusions on pipelines and a coordinated campaign by domestic saboteurs, and to remind pipeline operators of the importance of safeguarding and securing their pipelines from physical and cyber intrusion or attack, PHMSA, in coordination with TSA, issued an advisory bulletin in 2016. The bulletin also included a brief discussion of TSA’s and PHMSA’s roles on pipeline safety and security. Forums and routine interactions with operators. TSA and PHMSA officials also reported that they communicate their agencies’ respective roles and responsibilities for pipeline safety and security to stakeholders when conducting general outreach, information sharing efforts, or inspections or assessments. TSA and PHMSA officials noted that these activities provide opportunities for agency officials and pipeline stakeholders to clarify their roles and responsibilities should pipeline operators have questions. Examples of such community outreach activities include attending meetings of the Oil and Natural Gas subsector SCC or the Pipeline Modal SCC, and TSA’s annual International Pipeline Security Forum. TSA officials also said that TSA’s monthly and quarterly unclassified threat briefings provided TSA officials and pipeline stakeholders the opportunity to discuss and clarify their roles and responsibilities. Additionally, TSA produces classified and unclassified threat assessments on physical and cyber threats to pipelines, which according to agency officials can help to clarify TSA’s security role. Finally, TSA and PHMSA officials said that pipeline security assessments and safety inspections and other enforcement activities that the agencies regularly conduct are also opportunities to communicate their roles and responsibilities. For example, TSA officials reported that should an operator ask for assistance regarding a safety issue while TSA staff was conducting a security review, TSA staff would be able to refer the operator to PHMSA to address the issue. Similarly, PHMSA officials stated that inspectors would refer an operator to TSA or its pipeline security guidelines should the operator have questions regarding, for example, what security measures to implement. The representatives of the four pipeline associations we interviewed reported that TSA and PHMSA had clearly communicated their respective roles and responsibilities to pipeline stakeholders. Specifically, all of the association representatives said that their membership understood that TSA is responsible for pipeline security matters and PHMSA is responsible for pipeline safety matters. For example, one industry association representative stated that they had contacted their members to determine whether they were unclear regarding TSA’s and PHMSA’s respective roles and responsibilities and that members reported the roles were clear to them. Further, another association representative reported that the initial security reviews and outreach efforts that TSA conducted after the pipeline security program was created helped pipeline operators to understand that its role was to oversee pipeline security. In addition, all of the association representatives we interviewed stated that the MOU Annex helped ensure that TSA and PHMSA understood and respected each other’s roles and responsibilities. As a result, according to the association representatives, their pipeline operator membership had not experienced challenges associated with overlapping or duplicative efforts on the part of TSA and PHMSA pipeline safety or security programs. TSA Communicated Pipeline Incident Response Protocols in Its 2010 Plan, but Has Not Updated the Plan to Address Changes in Key Areas TSA Has Established a Pipeline Incident Response Protocol Plan That Communicates Agencies’ Roles and Responsibilities During Pipeline Incidents In accordance with the 9/11 Commission Act, TSA issued its Pipeline Security and Incident Recovery Protocol Plan in March 2010. The plan’s stated intent is to establish a comprehensive interagency approach to counter risks, coordinate federal agencies’ actions, and minimize the consequences of incidents involving pipeline infrastructure as well as recovery time from them. The plan also defines the roles and responsibilities of federal agencies; tribal, state, and local governments; and the private sector during a pipeline incident. It also defines the measures they may take related to pipeline infrastructure security incidents. According to the plan TSA, PHMSA, the Department of Energy (DOE), and the Federal Bureau of Investigation (FBI) have principal roles in pipeline incident response, while other agencies such as the U.S. Coast Guard, the Federal Emergency Management Agency (FEMA), and the National Transportation Safety Board (NTSB) have supporting roles. The following are examples of agencies’ roles and responsibilities in each of the plan’s three response phases. Prevention/protection. TSA is responsible for monitoring pipeline owner and operators’ implementation of its pipeline security guidelines, and PHMSA is responsible for enforcing its pipeline safety regulations. TSA, in addition to the FBI, is responsible for assessing the credibility of any physical or cyber threat information it receives and sharing any intelligence related to pipeline security with pipeline owners and operators. Response. TSA is responsible for coordinating information sharing between federal agencies and pipeline stakeholders, and PHMSA is responsible for coordinating federal agency activities with the affected pipeline operator and state pipeline safety agency. The plan also states that the FBI is responsible for investigating attempted or successful attacks on pipeline infrastructure including those that are believed to have a nexus to terrorism. Recovery. PHMSA is primarily responsible for working with the pipeline operator, along with other supporting federal agencies, to facilitate service restoration. DOE is responsible for monitoring flows of throughput in the affected pipeline system or systems, assessing regional, national, and global impacts of an incident on energy infrastructure throughout all three phases. Appendix I provides more details on key federal agencies’ and pipeline operators’ roles and responsibilities, as well as the actions they may take in response to an incident as detailed in the plan. TSA Has Not Updated Its Incident Response Plan to Address Changes in Pipeline Security Threats, Technology, and Federal Laws and Policies TSA’s plan states that it will be updated periodically to address changes in pipeline security threats, technology, and federal laws and policies. Further, Standards for Internal Control in the Federal Government states that periodic review of policies, procedures, and related control activities should occur to determine their continued relevance and effectiveness in achieving identified objectives or addressing related risks. In addition, internal control standards also states that changes in an entity’s programs or activities, organizational structure, personnel, or technology can affect the operating environment and management can respond by revising internal controls on a timely basis to ensure effectiveness. However, TSA has not reviewed or revised its 2010 plan to ensure it addresses changes in at least three key areas: cybersecurity-related laws and policies, incident management policies, and DHS’s terrorism alert system as described below. TSA’s 2010 plan includes some discussion of cyber threats and refers operators to guidance they may use to better secure their SCADA and control systems. However, the plan does not identify the cybersecurity roles and responsibilities of federal agencies that are identified in the plan, such as DOE, Federal Energy Regulatory Commission (FERC), or the FBI, or discuss the measures these agencies should take to prevent, respond to, or support pipeline operators following a cyber incident involving pipelines. TSA’s 2010 plan also has not been updated to reflect current cybersecurity incident response guidance. In December 2016, DHS issued its National Cyber Incident Response Plan (NCIRP). The NCIRP is to be the primary framework for stakeholders, including pipeline operators, to understand how federal departments and agencies provide resources to support response operations for a significant cyber incident. NCIRP identifies the FBI and the National Cyber Investigative Joint Task Force as responsible for investigating reported cyber incidents. NCIRP also identifies the National Cybersecurity and Communications Integration Center (NCCIC), an agency within DHS, as responsible for providing technical assistance to affected entities, such as pipelines, to mitigate vulnerabilities and reduce impacts of cyber incidents. NCCIC is also to share information across the public and private sectors to protect against similar incidents in the future. In addition, NCIRP provides guidance detailing when and to which federal agencies or entities the public should report a cyber incident. These include the FBI, the National Cyber Investigative Joint Task Force, U.S. Secret Service, and NCCIC. For example, NCIRP states that any cybercrime—including computer intrusions or attacks, theft of trade secrets, criminal hacking, terrorist activity, espionage, sabotage, or other foreign intelligence activity—is to be reported to FBI field offices’ cyber task forces. However, TSA’s plan does not include this information or describe what measures, if any, the agencies with pipeline-related roles and responsibilities listed in NCIRP are to take in response to a pipeline cyber incident. Moreover, the 2010 plan does not account for other agencies whose roles and responsibilities are related to critical infrastructure, such as pipelines and cybersecurity. Specifically, the plan does not account for the role of NCCIC, which was established in 2009. In addition, TSA’s 2010 plan does not account for CISA’s role in cyber threat response activities or how it may affect other agencies’ roles and responsibilities for pipeline incident response. TSA officials acknowledged that reviewing and, as appropriate, revising the plan would be beneficial to ensuring the plan addresses current pipeline security threats, technology, and federal laws and policies. They stated TSA had not updated the plan to include cybersecurity response protocols because an overarching cybersecurity response protocol for all critical infrastructure sectors—not just pipelines—should first be developed. According to TSA officials, developing a pipeline cybersecurity response protocol would require a whole-of-government approach, as well as coordination with private sector and input from many sectors because of the challenges and complexity of critical infrastructure cybersecurity in general. However, through NCIRP, DHS provided a cybersecurity response protocol across all critical infrastructure sectors in December 2016. Further, NCIRP states that public and private sector entities should consider creating an operational cyber incident response plan to further organize and coordinate their efforts in response to cyber incidents. Therefore, TSA could potentially provide such an operational cyber incident response plan for the pipeline sector in its plan. TSA could also better ensure that pipeline operators understand how federal agencies may provide support in response to a cyber incident by periodically reviewing and, as appropriate, revising the plan to include its cyber incident response plan. Representatives of the four pipeline associations we interviewed told us that their membership more clearly understood federal agencies’ roles and responsibilities related to physical incidents than to cybersecurity. For example, for physical incidents the representatives stated that their members clearly understood that they are to first notify local first responders (often through the emergency 911 system) and appropriate state or federal regulators, and are to contact either the National Response Center or TSA’s Transportation Security Operations Center (TSOC), depending on the nature of the incident. However, they stated that they did not believe all of their members clearly understand that they are to report any actual or suspected cyber incidents that could impact pipeline industrial control systems or other information technology-based systems to the NCCIC. All of the association representatives told us that the process for reporting a cyber incident is less clear because, in part, of the large number of federal agencies with a cybersecurity-related role. One of the representatives also attributed the lack of clarity to the reorganization of NCCIC, and the establishment of CISA. Further, all of the representatives we interviewed indicated that clarifying the cybersecurity roles and responsibilities of DOE, Federal Energy Regulatory Commission (FERC), and TSA would, among other things, improve operators’ ability to appropriately report and respond to a cyber incident. Federal Incident Management Policies TSA also has not updated the plan to address changes in federal incident management and response policies that have occurred since the plan was developed in 2010. The plan states that it is to be consistent with the National Response Framework (NRF) and the National Incident Management System (NIMS) incident command system procedures. The NRF was first issued in 2008 and described the roles, responsibilities and coordinating structures for delivering core capabilities during incident response. According to FEMA, it revised the NRF in 2013 and 2016 to reflect lessons learned from real world events and other experiences since the framework was first developed. Likewise, NIMS was developed in 2004 as a comprehensive, national approach to incident management that was to be applicable at all jurisdictional levels and across functional disciplines, such as law enforcement, public health, or public works. According to FEMA, it revised NIMS in 2017 to reflect and incorporate policy updates and lessons learned from exercises and real-world incidents. The revision was also intended to clarify that NIMS applies to all stakeholders with incident management roles, and to enhance guidance on information management processes, data collection plans, social media integration, and the use of geographic information systems. TSA officials acknowledged the benefit of periodically reviewing, and if necessary, revising the plan to reflect FEMA’s revisions to NIMS or the NRF, but had not done so because of competing priorities. DHS’s Terrorism Alert System TSA has also not updated the plan to address changes DHS made to its terrorist alert system in 2011. Consistent with the 9/11 Commission Act, the plan describes actions that federal agencies can take at each color- coded level of the Homeland Security Advisory System to ensure the increased security of pipeline infrastructure. For example, under the protect/prevent phase, the plan states that when there is a high risk of a terrorist attack (i.e., red: severe condition) and threat is general and not specific to pipelines, TSA and PHMSA are to coordinate to identify the potential for any related or cascading events that may impact the pipeline sector. However, if there is a specific threat to pipelines, TSA, in collaboration with pipeline operators, is to identify any immediate protective measures that pipeline operators are to implement. TSA is also to ensure pipeline operators have the information necessary to implement these measures, and, if necessary, to issue security directives. In 2011, DHS replaced the four color-coded alert system of the Homeland Security Advisory System with the National Terrorism Advisory System, which has only two alert levels (elevated threat and imminent threat). TSA issued revised protective measures that pipeline operators are to take under either threat condition in April 2011 and March 2018. However, TSA has not updated the plan to communicate the actions federal agencies can take at either level of the National Terrorism Advisory System to ensure the increased security of pipeline infrastructure. TSA officials acknowledged that periodically reviewing and, as appropriate, revising the plan would help to clarify federal agencies’ roles and responsibilities for addressing pipeline security. TSA officials reported that they have not updated the plan since 2010 because they faced competing priorities. However, as described earlier, TSA’s incident response plan was developed to provide a comprehensive interagency approach to important activities such as countering risks, coordinating federal agencies’ actions and minimizing the consequences of incidents involving pipeline infrastructure. Further, the plan itself states that it will be updated periodically to address changes in pipeline security threats, technology, and federal laws and policies. By periodically reviewing and, as appropriate, revising its Pipeline Security and Incident Recovery Protocol Plan, TSA could better ensure that the plan addresses all possible and relevant threats to pipeline systems, such as cybersecurity, and fully incorporates relevant changes, such as those related to incident management and DHS’s terrorism alert system. By doing so, TSA could also provide greater assurance that federal agencies understand the actions they are to take to prevent, respond to, or recover from a physical or cyber incident. Conclusions TSA and PHMSA share responsibility for safeguarding the nation’s pipeline systems from catastrophic events. While the 2006 MOU Annex delineates TSA’s and PHMSA’s mutually agreed-upon pipeline security roles and responsibilities, it has not been reviewed since its inception to consider pipeline security developments. By developing and implementing a mutually agreed upon timeline with timeframes for reviewing the annex and as appropriate, updating it, TSA and PHMSA could better ensure that their roles and responsibilities are properly documented and updated in a timely manner to remain current. Furthermore, by revising the MOU Annex to include a provision for periodically reviewing the annex for needed updates, TSA and PHMSA could better ensure the agreement consistently reflects relevant and updated information on their roles and responsibilities. Similarly, TSA’s Pipeline Security and Incident Recovery Protocol Plan— which defines the roles and responsibilities of federal agencies; tribal, state, and local governments; and the private sector for responding to a pipeline incident—also has not been updated to reflect changes in federal laws or policies since the plan was issued in 2010. By periodically reviewing and, when appropriate, updating its Pipeline Security and Incident Recovery Protocol Plan, TSA could better ensure that the plan addresses and fully incorporates changes relevant to cybersecurity, incident management and DHS’s terrorism alert system, among others. By doing so, TSA could also better ensure that federal agencies’ actions are well coordinated in response to a pipeline-related physical or cyber incident, and that pipeline stakeholders understand federal agencies’ roles and responsibilities in preparing for, responding to, or supporting pipeline operators to restore service after a pipeline-related physical or cyber incident. Recommendations for Executive Action We are making a total of five recommendations including three to TSA and two to PHMSA: The TSA Administrator should work with the PHMSA Administrator to develop and implement a timeline with milestone dates for reviewing and, as appropriate, updating the 2006 MOU Annex. (Recommendation 1) The PHMSA Administrator should work with the TSA Administrator to develop and implement a timeline with milestone dates for reviewing and, as appropriate, updating, the 2006 MOU Annex. (Recommendation 2) The TSA Administrator, in consultation with the PHMSA Administrator should revise the 2006 MOU Annex to include a provision requiring periodic reviews of, and as appropriate, corresponding updates to the Annex.(Recommendation 3) The PHMSA Administrator, in consultation with the TSA Administrator should revise the 2006 MOU Annex to include a provision requiring periodic reviews of, and as appropriate, corresponding updates to the Annex.(Recommendation 4) The TSA Administrator should periodically review, and as appropriate, update the 2010 Pipeline Security and Incident Recovery Protocol Plan to ensure the plan reflects relevant changes in pipeline security threats, technology, federal law and policy, and any other factors relevant to the security of the nation’s pipeline systems. (Recommendation 5) Agency Comments and Our Evaluation We provided a draft of this report to DHS and DOT. DHS and DOT provided written comments which are reproduced in appendices III and IV respectively. We also provided draft excerpts of this product to the American Petroleum Institute (API), the American Gas Association, the Interstate Natural Gas Association of America, and the American Public Gas Association. For those who provided technical comments, we incorporated them as appropriate. With regard to our first recommendation, that TSA work with the PHMSA to develop and implement a timeline with milestone dates for reviewing and, as appropriate, updating the 2006 MOU Annex, DHS stated that TSA will work with PHMSA to develop and implement a timeline with milestone dates for reviewing and updating, as appropriate, the 2006 MOU Annex. DHS estimated that this effort would be completed by August 31, 2019. This action, if fully implemented, should address the intent of this recommendation. With regard to our second recommendation, that PHMSA work with TSA to develop and implement a timeline with milestone dates for reviewing and, as appropriate, updating the 2006 MOU Annex, DOT concurred and stated it would provide a detailed response within 180 days of the issuance of this report. With regard to our third recommendation, that TSA, in consultation with PHMSA, revise the 2006 MOU Annex to include a provision requiring periodic reviews of, and as appropriate, corresponding updates to the Annex, DHS stated that TSA will, in consultation with PHMSA, revise the 2006 MOU Annex to include a provision requiring periodic reviews of, and as appropriate, corresponding updates to the Annex. DHS estimated that this effort would be completed by March 31, 2020. This action, if fully implemented, should address the intent of this recommendation. With regard to our fourth recommendation, that PHMSA, in consultation with TSA, revise the 2006 MOU Annex to include a provision requiring periodic reviews of, and as appropriate, corresponding updates to the Annex, DOT concurred and stated it would provide a detailed response within 180 days of the issuance of this report. With regard to our fifth recommendation, that TSA periodically review, and as appropriate, update the 2010 Pipeline Security and Incident Recovery Protocol Plan to ensure the plan reflects relevant changes to pipeline security threats, technology, federal law and policy, and any other factors relevant to the security of the nation’s pipeline systems, DHS concurred and estimated that TSA will complete its first review by December 31, 2019. DHS further stated that it will establish a timeline for updating the plan should the review determine that an update is necessary. This action, if fully implemented, should address the intent of this recommendation. We are sending copies of this report to the appropriate congressional committees; the Secretary of Homeland Security, Secretary of Transportation; and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact William Russell at (202) 512-8777 or russellw@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix V. Appendix I: 2006 Memorandum of Understanding (MOU) Program Areas and Accompanying Text The Transportation Security Administration (TSA) and Pipeline and Hazardous Materials Safety Administration (PHMSA),”the parties”, recognize that the following program areas are important to the development and deployment of an enhanced security strategy for the transportation of hazardous materials by all modes, including pipeline. Appendix II: Summary of Key Federal Agencies’ and Pipeline Operator’s Roles and Responsibilities This appendix summarizes the roles and responsibilities of key federal agencies as well as the actions that they may take in response to an incident as detailed in Transportation Security Administration’s (TSA) 2010 Pipeline Security and Incident Recovery Protocol Plan. A summary of pipeline stakeholder’s roles, responsibilities, and examples of actions that may be taken during each incident response phase is presented below. Prevention/Protection. During the prevention/protection phase, pipeline operators are to use TSA’s pipeline security guidance and the Pipeline and Hazardous Materials Safety Administration’s (PHMSA) safety regulations as the framework to prepare and prevent against an incident. TSA is responsible for monitoring pipeline owners and operators’ implementation its security guidelines, and PHMSA is responsible for enforcing its safety regulations. The plan also states that during this phase TSA is to assume a primary role for ensuring federal agencies’ actions are coordinated through protective security advisors (PSAs). In addition, the Federal Bureau of Investigation (FBI) is responsible for assessing the credibility of a known threat, preparing and implementing a preliminary investigative plan, and, if necessary, disseminating public safety notifications. The Department of Energy (DOE) is responsible for assessing and monitoring pipeline systems for supply shortages. The prevention/protection section of the plan also describes how agencies are to share and assess threat information. For example, the plan states that TSA, PHMSA, or any federal agency that receives threat information regardless of the source, must immediately notify the FBI. It also states that if the FBI receives intelligence about a pipeline threat, it is to share this information with TSA. TSA is then to notify the pipeline operator and, if necessary, provide recommendations for additional protective measures. Finally, the prevention/protection section of the plan defines actions various agencies can implement during a heightened security threat level to increase protection from a potential attack. For example, when there is a high risk of a terrorist attack (i.e., red: severe condition) and threat is general and not specific to pipelines, TSA and PHMSA are to coordinate to identify the potential for any related or cascading events that may impact the pipeline sector. If there is a specific threat to pipelines, TSA is, in collaboration with pipeline operators, to identify any immediate protective measures that ought to be taken by pipeline operators, and ensure pipeline operators have the information necessary to implement them, and, if necessary issue security directives. Response. According to the plan, pipeline owners or operators are to notify local first responders and state regulators through the emergency 911 system. After the pipeline operator has notified local government, they are to contact the National Response Center (NRC) if the incident results in an unintentional release or causes significant damage. As we previously reported, pipeline operators are also requested to report any physical security incident that is indicative of a deliberate attempt to disrupt pipeline operations or activities that could be considered precursors to such an attempt to TSA’s Transportation Security Operations Center (TSOC). Once TSA has been notified of an incident by a pipeline operator, its Pipeline Security Branch is to monitor the incident, notify relevant federal agencies, and, if deemed appropriate, activate the Interagency Threat Coordination Committee (ITCC). PHMSA may also deploy on-scene pipeline inspectors and investigators which are to among other things, coordinate federal agencies’ activities with the affected pipeline operator and state pipeline safety agency, provide subject matter expertise to the incident command, and direct safe restoration of pipeline facilities and services. The plan also states that, during the response phase, responsibility for investigating the incident falls to NTSB or the FBI depending on whether the incident is determined to be the result of criminal activity. The FBI is solely responsible for investigating any pipeline security incident that appears to be an intentional criminal act. For example, if the incident were suspected to be the result of terrorist attack, the National Joint-Terrorism Task Force would conduct an investigation of the attack, and if appropriate, with assistance from other FBI assets. If, however, the incident resulted in fatalities, substantial property damage, or significant injury to the environment, NTSB would have responsibility for investigating the incident, and may issue safety recommendations to help prevent future accidents. Recovery. When response activities are complete, PHMSA is to have primary responsibility for overseeing pipeline operators’ safe restoration of service with TSA and other federal agencies serving primarily in support roles. PHMSA, for example, is to work with the owner/operator to facilitate restoration of service by, among other things, providing technical oversight, advice, and guidance to owner/operators; coordinating recovery activities with state pipeline safety agency, and evaluate whether to a special permit is necessary to facilitate an expedited restoration of services. Meanwhile, DOE is to continue to assess the impacts of an incident on energy infrastructure, and advise federal, state, tribal, and local authorities on priorities for energy restoration, assistance, and supply. Appendix III: Comments from the Department of Homeland Security Appendix IV: Comments from the Department of Transportation Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Ben Atwater, Assistant Director and Michael C. Lenington, Analyst-in-Charge, managed this assignment. Nanette Barton, Eric Hauswirth, Susan Hsu, and Thomas Lombardi also made significant contributions to this report.

Why GAO Did This Study More than 2.7 million miles of pipeline transport natural gas, oil, and other hazardous liquids needed to operate vehicles and heat homes, among other things, in the United States. Responsibility for safeguarding these pipelines is shared by TSA, within the Department of Homeland Security (DHS); PHMSA, within the Department of Transportation (DOT); and pipeline operators. TSA oversees the security of all transportation modes, including pipelines. PHMSA oversees pipeline safety. DHS and DOT signed a MOU on their roles across all transportation modes in 2004. In 2006, TSA and PHMSA signed an annex to the MOU (MOU Annex) to further delineate their pipeline security-related responsibilities. The TSA Modernization Act includes a provision for GAO to review DHS and DOT roles and responsibilities for pipeline security. This report addresses, among other things: (1) the extent the MOU Annex delineates TSA's and PHMSA's pipeline security roles and responsibilities; and (2) the extent TSA has communicated federal incident response procedures for pipeline breaches to stakeholders. GAO reviewed the MOU annex and related documents and TSA's Pipeline Security and Incident Recovery Protocol Plan, and interviewed officials from PHMSA, TSA, and four pipeline associations. What GAO Found The memorandum of understanding (MOU) Annex signed by the Transportation Security Administration (TSA) and Pipeline and Hazardous Materials Safety Administration (PHMSA) in 2006 delineates their mutually agreed-upon roles and responsibilities for pipeline security, but has not been reviewed to consider pipeline security developments since its inception. As a result, the annex may not fully reflect the agencies' pipeline security and safety-related activities. Efforts to update the annex were delayed by other priorities. As of June 2019, there are no timeframes for completion. By developing and implementing timeframes for reviewing the MOU Annex and updating it, as appropriate, TSA and PHMSA could better ensure any future changes to their respective roles and responsibilities are clearly delineated and updated on a regular basis. TSA's Pipeline Security and Incident Recovery Protocol Plan, issued in March 2010, defines the roles and responsibilities of federal agencies and the private sector, among others, related to pipeline security incidents. For example, in response to a pipeline incident, TSA coordinates information sharing between federal and pipeline stakeholders and PHMSA coordinates federal activities with an affected pipeline operator to restore service. However, TSA has not revised the plan to reflect changes in at least three key areas: pipeline security threats, such as those related to cybersecurity, incident management policies, and DHS's terrorism alert system. By periodically reviewing and, as appropriate, updating its plan, TSA could better ensure it addresses changes in pipeline security threats and federal law and policy related to cybersecurity, incident management and DHS's terrorism alert system, among other things. TSA could also provide greater assurance that pipeline stakeholders understand federal roles and responsibilities related to pipeline incidents, including cyber incidents, and that response efforts to such incidents are well-coordinated. What GAO Recommends GAO is making five recommendations, including that: (1) TSA and PHMSA develop and implement a timeline for reviewing and, as appropriate, updating the 2006 MOU Annex; and (2) TSA periodically review, and as appropriate, update its 2010 pipeline incident recovery plan. DHS and DOT concurred with these recommendations.

gao_GAO-20-345

gao_GAO-20-345_0

Background Schools generally report incidents of restraint and seclusion to their district, and districts are responsible for reporting incidents of restraint and seclusion to Education’s CRDC. Districts are expected to report the number of incidents and the number of students affected for all schools in their district and to use Education’s definitions of restraint and seclusion to determine whether an incident occurred. Education defines two types of restraint: physical and mechanical (see table 1). Education’s definition of a physical restraint makes a distinction between a restraint and a physical escort. Similarly, Education’s definition of seclusion makes a distinction between seclusion and a timeout (see figures 1 and 2). Every two years, OCR administers the CRDC to nearly every public school district in the country. In turn, districts self-report information on a wide variety of topics, including course offerings, discipline, and restraint and seclusion. Education collects these data through an online submission tool. CRDC activities, such as data collection and quality, are managed by a company under contract with Education. The data submission period for the 2017-18 school year ended June 21, 2019. School districts have one year from the end of the submission period to make a request to amend submitted data. As of March 2020, Education had not announced when it will publicly release these data. Education’s primary data quality checks for the CRDC data, including the restraint and seclusion data, are built into the CRDC submission tool. The online CRDC submission tool automatically performs checks that flag data errors or potential errors. These “business rules” occur in real time as districts enter data or after they upload files. The 2015-16 submission tool used three business rules related to restraint and seclusion; for the purpose of this report, we focused on the two rules most relevant to our work. The first business rule pertains to the reporting of zeros for very large districts only—that is, those districts with 100,000 or more enrolled students (see fig. 3). If a very large school district enters zero incidents of restraint and seclusion, it receives a message prompting it to review its enrollment counts and reported incidents, or provide an explanation using a reason code and comment. Importantly, if districts have not collected the data required for the CRDC—or if the data are unavailable for some other reason—districts are to leave relevant data cells blank. A zero in a data cell should represent an actual count— that is, the district restrained or secluded zero students. The second rule applies to schools that both report more than 100 incidents of restraint or seclusion and enter a greater number of students affected by restraint or seclusion than the number of incidents reported (see fig.4). Districts are asked to resolve this error by adjusting their counts so the number of students subjected to restraint or seclusion is less than the number of incidents of restraint or seclusion, or to provide an explanation using a reason code and comment. The last step in the data submission process is the district certification. To complete this step, the district superintendent or an authorized designee must indicate agreement with a statement that acknowledges that they are responsible for verifying the data, the information provided is “true and correct,” and a willfully false statement is punishable by law. The CRDC submission tool is designed so that it should not allow a district to certify its submission unless all required data pass the system validation checks, or all errors are explained. Education also reviews CRDC data quality during other phases of data collection to identify potential data quality issues to improve future collections. For school year 2015-16, Education’s contractor reviewed data quality during and after the collection phase. Education’s contractor contacted school districts about potential errors that Education determined were easily adjustable and asked them to review and correct data or provide an explanation if no corrections were determined to be necessary. Further, some states and school districts have laws, regulations, and/or policies regarding restraint and seclusion. These laws vary from state to state, and sometimes require schools or districts to annually report incidents of restraint and seclusion to either the state or local education agency. In January 2019, Education announced an initiative to address inappropriate use of restraint and seclusion on children with disabilities. As part of this initiative, OCR announced plans to conduct 50 data quality reviews of the 2015-16 restraint and seclusion CRDC data submitted across each of OCR’s 12 enforcement regions. This review had not been completed at the time of our review. In August 2019, Education announced plans for OCR to work collaboratively with Education’s National Center for Education Statistics (NCES). Education stated that the collaboration would help ensure that CRDC data are reliable and authenticated in a manner that provides a more accurate picture of key civil rights issues in education. According to the announcement, under a new agreement, NCES and OCR will work together to improve the quality of the CRDC data by providing school districts with technical assistance, and by reviewing and revising data quality procedures. NCES is the primary federal entity responsible for collecting and analyzing statistical data related to education. Key Quality Control Processes for Restraint and Seclusion Are Ineffective or Do Not Exist We identified four key issues for which Education’s quality control processes for its CRDC restraint and seclusion data are largely ineffective or do not exist (see table 2). CRDC Business Rule Targeting Zero Reporting Is Largely Ineffective Education’s business rule to detect potentially erroneous reporting of zero incidents applied to only 30 of the more than 17,000 school districts nationwide, rendering it largely ineffective for checking the 70 percent of districts that reported zero incidents of restraint or seclusion. This is because the rule only applied to districts with over 100,000 enrolled students. However, in its January 2019 data quality review of 50 districts’ restraint and seclusion data, OCR found erroneous reporting of zeros in districts of all sizes. Of the 50 districts OCR contacted, OCR determined that 40 districts should amend their original 2015-16 submissions. When we analyzed the 2015-16 CRDC restraint and seclusion data, we found that almost three-quarters of small districts reported zeros, while about one-third of large and one-fourth of very large districts reported zero incidents (see table 3). The findings from Education’s data quality review, along with those from our analysis, suggest that misreporting is a problem among districts of all sizes. For the 2017-18 CRDC data collection, Education lowered the threshold for the rule to detect potentially erroneous reporting of zeros to include districts with 50,000 enrolled students, rather than only districts with more than 100,000. However, the new rule counts students with disabilities and students without disabilities separately. To trigger the rule, a district would have to have at least 50,000 students with disabilities and report zero incidents for them, or have at least 50,000 students without disabilities and report zero incidents for them. Of the nation’s more than 17,000 school districts, only 3 (Chicago, Los Angeles, and New York) have at least 50,000 students with disabilities; only 95 have at least 50,000 students without disabilities. Education’s CRDC data show that restraint and seclusion disproportionately affect students with disabilities and its data quality review showed that substantial portions of districts of all sizes inaccurately reported zeros. However, Education could not provide a data-driven basis for the 100,000 or 50,000 student enrollment thresholds or for creating separate thresholds for students with and without disabilities. Rather, Education stated that the thresholds were a management policy decision inherited from previous administrations. Because Education’s business rule targeting districts that report zeros was inadequate, in June 2019, GAO recommended, among other things, that for the 2017-18 data collection Education contact districts that reported all zeros for restraint and seclusion to ask them to ensure that the zeros actually represented zero incidents, and Education did so after the data collection closed. Absent a business rule targeting all districts reporting zeros during data submission, inaccuracies in future data collections will likely be missed, and Education risks expending more time and resources with repeated manual follow up after the fact. Federal standards for internal control state that management should evaluate both internal and external sources of data for reliability. Absent reliable and accurate data, the public’s confidence in the CRDC restraint and seclusion data may be further undermined, and the utility of a dataset intended to assist with federal civil rights monitoring, enforcement, and oversight will remain limited. While it is clear that some school districts have reported inaccurate restraint and seclusion data, Education officials do not fully understand why this is occurring. In technical comments on a draft of this report, Education stated that it is committed to learning more about why this is occurring. While we do not know all of the reasons districts fail to report accurate data, our interviews with over 50 school and district officials provide some insight. School officials in the nine districts we visited cited a variety of reasons districts might not report, including that they were not collecting the data because their state did not require reporting, and that their school district only required them to collect data for students with Individualized Education Programs (IEPs). More fully understanding why districts report inaccurate data is key to correcting the issue. Federal standards for internal control also state that managers should use quality information to achieve the entity’s objectives, assess the risks facing the entity as it seeks to achieve its objectives, and use this assessment to develop appropriate risk responses. By not identifying school districts’ reasons for reporting zero incidents of restraint and seclusion, Education will not know how to best support districts in improving the accuracy of their reported data. Future CRDC data will remain inaccurate, significantly limiting the utility of a key tool on which OCR relies to help it enforce federal civil rights laws. CRDC Lacks Business Rules to Detect Very Low or Very High Rates of Restraint or Seclusion Education has no business rules that flag school districts reporting very low or very high rates of restraint or seclusion, nor has it completed initial efforts to determine a range of rates that might warrant further exploration. Given widespread concerns about misreporting, we devised two possible ways to test for these types of outliers. First, we looked beneath the district level to examine school-level reporting patterns within districts. When we tested the nation’s 30 largest school districts (those with more than 100,000 students), we found patterns that may suggest underreporting in at least 13 of them, in addition to the 10 that reported zeros for the 2015-16 school year. In these 13 districts, we found that all of the incidents of restraint reported occurred in no more than 15 percent of a district’s schools; the rest of the schools in those districts reported zero incidents. (See fig. 5.) For example, the Chicago Public School District—the third largest school district in the country, with nearly 400,000 students enrolled—reported a total of 47 incidents of restraint for school year 2015-16. All of these incidents were reported by just two of its 579 schools. The district’s six incidents of seclusion were clustered in one school. In the Los Angeles Unified school district, the second largest school district in the country, 82 of its 785 schools reported a total of 108 incidents of restraint, with 65 schools reporting exactly one incident each. The district reported no incidents of seclusion. Education has a business rule that targets large districts (for 2017-18, those with over 50,000 students), but only when all schools in a district report zeros. Thus, as long as a large district reports at least one incident of either restraint or seclusion, the business rule would not be triggered. Education’s post-collection data quality reviews for school year 2015-16 did not test below the district level to look for potential underreporting within a district. Second, we tested for outliers by comparing per capita rates of restraint or seclusion in the 30 very largest districts (over 100,000 students enrolled) to average rates in all school districts. In the 30 districts, we found that in addition to the 10 districts that reported zeros, nine districts reported fewer than three incidents of physical restraint per 10,000 students, which is lower than 95 percent of all districts reporting incidents. (See fig. 6.) For example, DeKalb County school district in Georgia reported 0.3 incidents per 10,000 students, and Charlotte Mecklenburg school district in North Carolina reported 0.5 incidents per 10,000 students. We also tested for districts with very high rates of physical restraint. For the 2015-16 school year, we found 52 districts that were outliers, most of which served comparatively large populations of students with disabilities. Forty-nine of these 52 districts had rates of physical restraint per enrolled student higher than 99 percent of all districts that reported incidents of physical restraint. Almost half reported an average of 10 or more incidents per student affected, and almost two-thirds of the districts reported restraining from 25 to100 percent of their students. The Learning Tree preschools in Alabama, which enrolled a combined 135 preschool students ages 3 to 5, reported that it restrained nearly two-thirds of its students in school year 2015-16. Further, Learning Tree reported 5,963 incidents of physical restraint affecting 84 students, or an average of 71 incidents of physical restraint per preschooler. The Morris-Union Jointure Commission School District in New Jersey, where almost all of its 281 students were identified as having a disability, reported restraining over one-third of its students. These students were restrained an average of 20 times in school year 2015-16. (See table 4.) We found a similar pattern in the 2015-16 seclusion data, identifying 36 outlier school districts. For 22 of the 36 districts, the rates of seclusion were higher than 99 percent of districts reporting. (See appendix IV for more information.) For example, CRDC data for the Sangamon Area Special Education District in Illinois, which enrolled 74 students in grades 1 through 12, showed the district secluded one-third of its students an average of 27 times each in school year 2015-16. Similarly, data for the Bi-County Special Education Cooperative, also in Illinois, showed the district secluded over two-thirds of its 48 enrolled students an average of 13 times each in school year 2015-16. Federal standards for internal control state that management is to determine if controls individually and in combination with other controls are capable of achieving an objective and addressing related risks. An internal control design may be deficient when a control necessary to meet an objective is missing. Without business rules or similar analytical processes to flag these outliers, they may continue to go undetected by Education and other stakeholders. Education may be missing an opportunity to identify districts with disproportionately low or high incidents of restraint and seclusion to determine where technical assistance or other intervention may be warranted. Such information is particularly critical given widespread concerns about underreporting and misreporting, and its stated interest in protecting students’ civil rights. CRDC Business Rule Targeting Illogical Data Is Largely Ineffective Education has a business rule that identifies illogical data; that is, when schools report more students affected than incidents of either restraint or seclusion. However, the rule is largely ineffective because it was not designed in a way that would detect logical inconsistencies in the majority of cases, as the rule would have only applied to schools with at least 100 incidents. When we tested Education’s rule on the 462 schools that reported at least 100 incidents in 2015-16, we found no logical inconsistencies in the data. However, when we tested the rule on all schools, we found logical inconsistencies in the data reported by 592 schools with fewer than 100 incidents. For example, a school in Indiana reported that it restrained 156 students, but only reported 80 incidents. (See table 5.) Education could not provide any data-driven basis for its threshold of 100 incidents for this business rule. Education officials said that the threshold was inherited from previous administrations’ business rules. Collecting accurate data is key to the Office for Civil Rights’ (OCR) mission to ensure equal access to education and to promote educational excellence throughout the nation. In addition, federal standards for internal control state that when evaluating the design of internal controls, such as business rules, management should determine if controls are capable of achieving an objective and addressing related risks. An internal control design is deficient if, even though it operates as designed, it does not meet the control objectives. Our analysis shows that the business rule is not effective in its current form, because 592 schools were able to report illogical, and therefore incorrect, data. School and School District Officials Lacked a Common Understanding of Federal Restraint and Seclusion Definitions We talked to more than 50 officials in nine school districts in Kentucky, Washington, and Wisconsin about their interpretations of the CRDC’s definitions of restraint and seclusion. These school districts all reported incidents of restraint and seclusion in 2015-16, but officials we interviewed differed in their interpretations of terms used in the CRDC definitions, such as alone and escort. As a result, districts varied in how they counted incidents of restraint and seclusion. Further, officials we spoke with in the three state educational agencies and all seven stakeholder groups with expertise on the use of restraint and seclusion in public schools also said there was ambiguity regarding terms used in the definitions. For example, an official from one stakeholder group that represents some of the nation’s largest school districts said that its constituents provided feedback that restraint and seclusion terms were ambiguous, open to interpretation, and do not provide enough clarity. Civil Rights Data Collection Definition of Seclusion: Seclusion is the involuntary confinement of a student alone in a room or area from which the student is physically prevented from leaving. It does not include a timeout, which is a behavior management technique that is part of an approved program, involves the monitored separation of the student in a non- locked setting, and is implemented for the purpose of calming. With respect to the definition of seclusion, district and school officials varied in their interpretations of the word alone, and consequently, whether the incident should be counted as seclusion. Officials in three districts said that an incident was not seclusion as long as a teacher was in the room with the student, while officials in several other districts said that even if a teacher was present, it could still be seclusion if the student was prevented from leaving. (See sidebar.) Officials in the nine districts we visited also varied in their interpretation of the word area. Because the CRDC’s definition of seclusion states that seclusion can occur in an area, officials from one stakeholder group representing thousands of school administrators wondered whether it should be considered seclusion if a child is in a classroom with 20 other students and is required to stay alone in the corner of the room. Officials from a district in Wisconsin said that if a student is taken away from peers and placed in one area of the same room, but cannot leave that area, it still might be seclusion, even if the student and peers are in the same room. Officials in another district in Wisconsin said that sending a student to a corner does not count as seclusion. However, they said that the use of mobile partitions to close off an area of a room could constitute seclusion. The phrase physically prevented from leaving also elicited differing interpretations. Officials from the stakeholder group representing administrators said the definition is not clear about what counts as “prevented from leaving”: a barrier, such as a door; the presence of another adult watching the child; or both. School officials we spoke with had differing interpretations of this phrase, which affected how they counted and reported incidents of seclusion. School officials in a district in Kentucky said that the phrase means closing the door and keeping it closed. However, officials in another school in the same district did not specify the use of a door, stating instead that “physically prevented from leaving” means the student cannot walk out of the room. A school official in Washington said that it would count as seclusion if staff put a student in a motorized wheelchair in a room and deactivated the wheelchair’s power. Civil Rights Data Collection Definition of Physical Restraint: Physical restraint is a personal restriction that immobilizes or reduces the ability of a student to freely move their torso, arms, legs, or head. The term does not include a physical escort. Physical escort means a temporary touching or holding of the hand, wrist, arm, shoulder or back for the purpose of inducing a student who is acting out to walk to a safe location. With respect to the definition of physical restraint, school district officials we interviewed varied in their interpretations of the term escort, which the CRDC definition specifies is not a physical restraint. (See sidebar.) While officials in three districts said that an escort meant providing a physical prompt to a student who was not resisting relocation, officials in four districts said that moving a student who was resisting staff still counted as an escort. For example, officials in a school in Wisconsin said that if the student who is resisting is “carried away” from a location, that action would not meet the definition of escort and would count as restraint. Yet a district in Kentucky counted moving students against their will—including by carrying them—as escorts and did not report them as restraints. Further demonstrating differing interpretations of these terms, officials in four districts said they reported all escorts as restraints in the CRDC. Education does not provide schools or school districts with any information that could help clarify its definitions or provide examples on how schools and school districts should apply the definitions of restraint and seclusion to common classroom situations. For example, while Education’s “Restraint and Seclusion: Resource Document” includes the CRDC definitions of restraint and seclusion, it does not include clarifying information or examples about how to apply the definitions. Officials in the schools and districts we visited inconsistently interpreted the definitions for restraint and seclusion; moreover, officials from the seven stakeholder groups we interviewed said the definitions were unclear. These findings raise concern about whether restraint and seclusion data reported by school districts to the CRDC are being reported in a way that is consistent with the CRDC definitions. Federal guidance on data reliability states that data should be well defined enough to yield similar results in similar analyses. In addition, federal standards for internal control state that agency management should use quality information to achieve the entity’s objectives, noting that such data should be reasonably free from error and bias and faithfully represent what they purport to represent. Absent data on restraint and seclusion that is what it purports to be, Education will continue to lack quality information key to fulfilling its mission of ensuring equal access to education nationwide. All Nine School Districts We Visited Used Data to Reduce Incidence of Restraint and Seclusion and Developed Strategies for Improved Reporting All Nine Districts Used their Data on Restraint and Seclusion to Reduce Use of Restraint and Seclusion Officials in all nine school districts we visited said they used their data on restraint and seclusion to help reduce its use. In addition to collecting data for CRDC reporting purposes, these districts also collected and used more current and more detailed data to help reduce the use of restraint and seclusion. Officials in seven of the nine districts said they began collecting the data when their state passed a law requiring reporting. District officials identified several benefits to collecting data and using it to develop strategies to reduce use of restraint and seclusion. Specifically, officials said that the data helped them identify the following: Behavior patterns. Officials in several districts told us that collecting and reviewing data on restraint and seclusion helps them identify patterns in staff and student behavior that may contribute to use of these practices. Specifically, by identifying the circumstances under which a student’s behavior tends to escalate, staff can strategize how to more effectively respond so as to prevent the need to use restraint or seclusion. For example, one official in an elementary school in Wisconsin said that if staff notice more incidents occur on particular weekdays, they can examine those days to understand what may be affecting students’ behavior. Similarly, a teacher of students with autism in a middle school in Washington said that reviewing data helps staff, such as teachers, paraprofessionals, and administrators, determine what triggered a student’s behavior and then determine what to do differently to avoid triggering the student. In all three states, we visited districts that required staff to participate in a debriefing after each incident in an effort to understand what might have triggered the event and to discuss strategies to deescalate future incidents. For example, officials in Washington said that the building administrator and all staff involved discuss every incident. District officials consider this an important step for reducing use of restraint and seclusion, and said holding the discussions was a “game- changer.” Need for training. Officials in several districts said they examine data on restraint and seclusion at the classroom and school level to determine if staff need additional training, including on how to manage student behavior, or appropriately use restraint or seclusion. For example, a behavior coach for a Kentucky school district said that the data on restraint and seclusion helps her determine if certain teachers could benefit from more training on de-escalation techniques. A director of student services in Washington said that he was concerned about the rates of restraint and seclusion in the district, and after implementing more training for teachers, the rates declined. Officials in another Washington district said that after the district began collecting data in response to state law, they discovered that staff were using restraint and seclusion as punishment. As a result, district officials said that they coached teachers on how to manage behavior differently and emphasized that restraint and seclusion should not be an everyday occurrence. Officials at an elementary school in Wisconsin said that de- escalation training helps staff understand that students are trying to communicate with their behaviors. They said that when staff adopt the perspective that students are trying to communicate, staff also see the value of collecting data to improve how they respond to the students’ behaviors. Need for student supports. Officials in five of the nine districts we visited spoke about using restraint and seclusion data to assess when a student required additional support services to be successful in the classroom. For example, officials in one district in Kentucky said their data provides evidence for obtaining additional staff or social emotional learning resources for students. Similarly, officials in one Wisconsin district said the data can be used to allocate funding for school-based services to help address underlying causes of behavior. Officials in another Wisconsin district said that a jump in restraints or seclusions of a particular student could indicate that the student’s individualized education program needs to be adjusted. All Selected Districts Developed Strategies to Encourage Reporting of Incidents Officials in all the districts we visited also shared strategies on how they improved their CRDC data reporting, including communicating with staff about how data are used, training on how to report, and developing processes that encourage reporting. Specifically: Communication and culture. Officials in the majority of school districts said they routinely reviewed their data with school staff and emphasized the value of collecting data on restraint and seclusion. Officials in a district in Wisconsin said that they monitor data on restraint and seclusion on a monthly basis for students with and without disabilities, which increases interest among school staff about what causes the incidents. Officials in three districts we visited said that they explain to staff that documenting incidents of restraint or seclusion ensures that students obtain the support services they need. In a Wisconsin district, officials said they emphasize that reporting helps the students and keeps the school safe by making the district aware that more supports are needed. Officials at a school in Wisconsin said that some staff might worry that the data reflect poorly on them or might fear repercussions, but district officials have worked to shift the culture of reporting to focus on continuous improvement and problem solving. Similarly, officials in another Wisconsin district said that schools might be concerned about the data being used against them; therefore, district officials try to create a culture of curiosity around the data, rather than a culture of punishment. Accountability. To encourage staff to report incidents, officials in some districts developed processes that increased accountability for reporting. For example, school officials at an elementary school in Wisconsin said an administrative assistant in the main office immediately logs calls from classroom teachers requesting help managing a student’s behavior. Officials said this process provides accountability. Two districts said that they used a team approach for restraint or seclusion, which included someone to observe and someone to record details of the intervention, such as the time it began or the events that preceded it. Having multiple people involved increased the likelihood that relevant facts were recorded. Officials in a district in Washington said that keeping teachers and staff honest about reporting requires reiterating the process and procedures, reviewing the forms with staff, and following up with schools that fail to submit reports. Officials in a district in Wisconsin said they have advised staff to write the incident down on paper until staff are able to enter it in the district’s electronic reporting system. Training. Officials from all nine districts said they encourage reporting by provided training on how to report incidents. Generally, this information was incorporated into trainings on when to use restraint and seclusion and how to deescalate a student’s behavior. Officials from five stakeholder groups we interviewed, all of whom have expertise related to the use of restraint and seclusion in public schools, stated that training was necessary to both raise awareness of the requirement to report incidents and to ensure that incidents were reported accurately. For example, in de-escalation training for teachers in a Washington district, the trainer provides examples of restraint and seclusion; presents a variety of scenarios, including ambiguous ones, for discussion; and reviews the appropriate staff response. Conclusions The Civil Rights Data Collection (CRDC) is a longstanding and critical aspect of Education’s Office for Civil Rights’ overall enforcement and monitoring strategy. Collecting accurate data through the CRDC can help Education in its mission to ensure equal access to education, promote educational excellence for all, and enforce various federal civil rights laws prohibiting discrimination on the basis of race, color, national origin, sex, and disability. However, the significant data quality problems that both Education and we identified with the CRDC data on restraint and seclusion, combined with the significant weaknesses we found in Education’s data quality control processes, cast serious doubt on the accuracy of these data. As a result, it is impossible to accurately determine the frequency and prevalence of restraint and seclusion among K-12 public school students. The four recommendations in our June 2019 report urged Education to take immediate steps to address the widespread potential misreporting of zeros for its 2017-18 CRDC. Education took some steps to address the issues we raised, but has not yet fully addressed them. Moreover, those recommendations were intended as stop-gap measures to improve the quality of the 2017-18 data being collected in real time precisely because the CRDC’s business rules related to restraint and seclusion are inadequate. Therefore, addressing our recommendations would not solve the issues that are the subject of this report. Our work makes it clear that an overhaul of the quality control processes is needed to correct fundamental problems with federal restraint and seclusion data collected through the CRDC. Two of the CRDC’s key business rules meant to check data quality and flag potential errors in restraint and seclusion data are poorly designed and the thresholds that trigger these rules have no data-driven basis. Further, Education does not have business rules designed to flag outlier schools and school districts that report relatively low or high rates of restraint and seclusion, nor has it determined a range of rates that might warrant further exploration. Until Education more fully understands why so many school districts are underreporting and misreporting federal restraint and seclusion data, it will likely not be able to help districts improve their reporting, thereby improving the accuracy and utility of the data. There were widely varied interpretations of federal restraint and seclusion definitions among the 50 school and district officials with whom we spoke and officials from the seven key stakeholder groups we interviewed echoed these concerns. As a result, we have concerns that school districts may be inconsistently counting and reporting instances of restraint and seclusion for federal reporting purposes. Clarifying the definitions, for example by explaining to districts how they can be applied to common classroom scenarios, could help produce more consistency in reporting. Ultimately, the issues we found with Education’s restraint and seclusion data have consequences for the students who are restrained or secluded in school and whose restraint or seclusion goes un-reported. When federal data are misreported to the public, it undermines confidence in that data and fails to provide decision makers with reliable information on which to make informed policy decisions to protect students. In addition, Education lacks information that could help it determine whether schools’ use of these practices may be excessive, discriminatory, or both. Recommendations for Executive Action GAO is making six recommendations on restraint and seclusion to the Department of Education’s Office for Civil Rights. The Assistant Secretary for the Office for Civil Rights should revise its CRDC business rule to require that every school district reporting zeros, regardless of district size or numbers of students with disabilities, affirm the zeros are correct during the CRDC data submission process. (Recommendation 1) The Assistant Secretary for the Office for Civil Rights should develop and implement a CRDC business rule that targets schools and school districts that report very low numbers of incidents and set data-driven thresholds to detect such incidents. (Recommendation 2) The Assistant Secretary for the Office for Civil Rights should develop and implement a CRDC business rule that targets schools and schools districts that report very high number of incidents and set data-driven thresholds to detect such incidents. (Recommendation 3) The Assistant Secretary for the Office for Civil Rights should apply the CRDC business rule targeting illogical data at the school level to all schools, regardless of the number of incidents reported. (Recommendation 4) The Assistant Secretary for the Office for Civil Rights should identify the factors that cause underreporting and misreporting of restraint and seclusion and take steps to help school districts overcome these issues. (Recommendation 5) The Assistant Secretary for the Office for Civil Rights should further refine and clarify federal restraint and seclusion definitions and take steps to ensure that this information is conveyed to school districts. This could include providing common classroom scenarios that highlight the differences between a restraint and an escort, and a time out and a seclusion. (Recommendation 6) Agency Comments and Our Evaluation We provided a draft of this report to the Department of Education for review and comment. In its formal comments, which are reproduced in appendix II, Education agreed with all six recommendations. Education also provided technical comments, which we incorporated, as appropriate. In agreeing with GAO’s six recommendations, Education stated that it would determine the best means to implement them. Education also stated that it is fully committed to working with public schools, state educational agencies, and school districts to help ensure accurate reporting of federal restraint and seclusion data, and to improve the quality of the information for all users of CRDC data. We appreciate Education’s willingness to address the serious data quality issues affecting the CRDC restraint and seclusion data. In its response, Education stated that the agency has already made significant improvements to the CRDC in general and has made specific improvements with respect to restraint and seclusion data, especially in response to the four recommendations we made in our June 2019 correspondence. Education asked that we acknowledge the progress it feels it has made in this regard, and we have done so. Importantly, however, our June 2019 recommendations were intended as stop-gap measures to improve the quality of the 2017-18 data that was already being collected in real time precisely because the CRDC’s business rules related to restraint and seclusion were inadequate. Therefore, steps Education has taken toward addressing them do not address the underlying data quality issues that are the subject of this report. In other words, the recommendations in this report urge Education to address data quality problems at the front-end by applying adequate business rules at the time districts submit their data. This could reduce the need for follow-up with districts to correct potentially inaccurate data. More information about our assessment of the steps Education has taken to address the four recommendations from the June 2019 report are available on our website. Education also stated that because our draft report did not mention the methodological improvements OCR made to address the quality of restraint and seclusion data for the 2017-18 CRDC data collection, our draft report overstates the relevance of the data issues from the 2015-16 collection. Education also stated that it provided us with information about the methodological improvements in December 2019, and, in its formal response, requested that we reflect the information in this report. We disagree with this perspective. After we completed our audit work for this engagement, Education provided us an excerpt from its post-collection data quality report for school year 2017-18. At that time and again in its technical comments on this report, Education stated that “information shared with GAO about the results of the 2017-18 data quality review process and what might be addressed is still confidential.” As of March 23, 2020 Education described the 2017-18 data quality control process as “incomplete.” Lastly, the 2017-18 CRDC data, which are the topic of the excerpt Education provided to us, are not yet available. Under our auditing standards, we cannot opine on the quality of data we could not independently assess or on the efficacy of process improvements associated with those data. Education also raised concerns about how we weighted our interviews with school officials, and it questioned the relevance of our discussions about selected school districts’ use of restraint and seclusion data not reported for CRDC purposes. We disagree. Education was concerned about the weight GAO placed on information obtained from 50 officials in 11 school districts across 3 states whom we interviewed during the course of our audit work. As stated in the report, this information cannot be generalized to all districts. However, we believe it provides useful insights into how some districts use their restraint and seclusion data to reduce the incidence of these practices and improve the accuracy of their data. The widespread disagreement among the 50 school officials with whom we spoke also highlights confusion about how to accurately and consistently apply CRDC definitions of restraint and seclusion. This finding is supported by the views of seven nonfederal advocacy organizations that represent parents and families; individuals with disabilities; and other stakeholders, such as representatives of relevant school and special education professional associations. Education questioned the relevance of discussing the benefits that selected school districts said they derive from using restraint and seclusion data not reported for CRDC purposes. Education stated that “attempting to generalize comments about how these nine school districts use restraint and seclusion data” seems inconsistent with Governmental Accounting Standards Board (GASB) statistical principles. We believe that describing selected school districts’ use of their restraint and seclusion data is within the scope of our stated audit objectives. In addition, the explanatory statement from the House Committee on Appropriations accompanying the Consolidated Appropriations Act of 2018 includes a provision for us to provide examples of how schools are adopting effective alternatives to these practices and reducing the incidence of seclusion and restraint, among other things. Further, Education mistakenly asserts that none of the data and analyses that the school districts collected, performed, or used are part of the CRDC and none could be feasibly collected by the CRDC. We have further clarified in the final report that portions of the data these school districts collect are used for CRDC reporting purposes. For example, some of the data elements are the same ones that districts use to calculate aggregate incident counts, which are required by the CRDC. We do not recommend that Education collect such detailed data or perform such analyses. Regarding Education’s concern about “GASB statistical principles” and case selection, all GAO performance audits are subject to Generally Accepted Government Auditing Standards (GAGAS); in contrast, GASB’s Generally Accepted Accounting Principles apply to financial audits of public entities. The applicable methodological guidance we followed -- Selecting a Sample of Nongeneralizable Cases for Review in GAO Engagements -- is designed to ensure that GAO policies on evidence and GAGAS are met, and conforms to the generally accepted principles and practices of the appropriate disciplines. When providing illustrative examples, it is neither necessary nor appropriate to use statistical methods to analyze and interpret evidence. Finally, in its comments, Education stated that it is critical that we emphasize that the CRDC is an aggregate of self-collected and self- reported data from school districts and that the district superintendent or an authorized designee certifies that the data they submit are “true and correct.” We agree, and acknowledged this in several places in both the draft and final reports. At the same time, we believe that self-certified data does not absolve Education of its responsibility to ensure the quality of the data it collects and publicly reports – especially given the CRDC’s longstanding role in Education’s overall enforcement of various federal civil rights laws prohibiting discrimination on the basis of race, color, national origin, sex, and disability. Self-reported data by nature are subject to error, making the need for effective quality control measures before, during, and after collection a necessity. We are sending copies to the appropriate congressional committees, the Secretary of Education, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (617) 788-0580 or nowickij@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix V. Appendix I: Objectives, Scope, and Methodology This report examines (1) the effectiveness of CRDC data quality control procedures for its restraint and seclusion data, (2) how selected districts interpret the CRDC definitions of restraint and seclusion and (3) how selected districts use data on restraint and seclusion and encourage staff to report incidents. Below are the details of our analysis to determine the extent to which Education ensures the quality of restraint and seclusion data reported by school districts, and of our interviews with officials in selected districts about how they apply Civil Rights Data Collection (CRDC) definitions of restraint and seclusion and use restraint and seclusion data. To inform all of our objectives, we interviewed federal agency officials, representatives from several nonfederal advocacy organizations that represent parents and families, individuals with disabilities, and other stakeholders, such as representatives of professional associations. We also reviewed agency documentation, relevant federal laws, regulations and policies, and selected state laws. Analysis of National Restraint and Seclusion Data To determine the extent to which Education ensures the quality of restraint and seclusion data reported by school districts, we analyzed Education’s Civil Rights Data Collection (CRDC) for school year 2015-16. Specifically, we analyzed the CRDC to determine the extent to which districts reported zero incidents of restraint and seclusion, to identify outliers (districts that reported a high or low incidence of restraint and seclusion), and to identify illogical data. CRDC is a biennial survey that is mandatory for nearly every public school and school district in the United States and is conducted by Education’s Office for Civil Rights. The CRDC collects data on the nation’s public schools (pre-K through 12th grade) that includes the use of restraint and seclusion, student demographics and enrollment numbers, educational and course offerings, and disciplinary actions. In school years 2013-14 and 2015-16, the CRDC collected data from nearly every public school in the nation (approximately 96,000 schools in 17,000 school districts in school year 2015-16). CRDC data are self-reported by districts and schools, and consequently there is potential for misreporting of information. After reviewing their CRDC data, school districts can submit revised data to Education. The public-use data file of the CRDC for school year 2015-16 was the primary source of data for our analyses and the most recent data available at the time. We also used restraint and seclusion data from school year 2013- 14 primarily to analyze how use of restraint and seclusion may have changed between the two time periods. The CRDC collected data on (1) mechanical restraint, (2) physical restraint, and (3) seclusion. Using these data, we performed the following analyses to determine potential inaccuracies or underreporting in the CRDC. Analysis of Extent of Districts Reporting Zeros To examine the extent to which school districts reported zeros, we calculated the percentage of districts and schools reporting zeros for restraint (both mechanical and physical) and for seclusion. We performed this calculation for both districts and schools nationally and by state, district size, and school type (e.g., charter, traditional, and special education schools). Although Education has a business rule that targets very large districts that report zero incidents of restraint or seclusion, we calculated the number of all districts and schools that reported zeros to understand the prevalence of zeros in the reported data. Analysis of Relatively Low Rates of Restraint and Seclusion Incidents To test for potential underreporting, we first limited our analysis to the restraint and seclusion data reported by the 30 largest school districts in the nation (districts with over 100,000 students enrolled). Because of these districts’ size, we reasoned that they would be more likely to have incidents of restraint and seclusion to report. Our analysis found that 20 of the 30 largest school districts reported incidents, and thus we focused our analysis of underreporting on the 20 largest districts that reported incidents. For each of the 20 districts, we calculated the percentage of schools that reported incidents. To compare the 20 largest districts that reported incidents with all 5,252 districts that reported incidents, we calculated the rates of restraint and seclusion per enrolled student and calculated percentile ranges. (See table 6.) We determined that nine of the 20 districts had incidents of physical restraint per enrolled student that were below the 5th percentile of all districts reporting incidents of physical restraint. Analysis of Relatively High Rates of Restraint and Seclusion Incidents To identify school districts with relatively high rates of restraint and seclusion, we examined districts that reported having more incidents than students enrolled. This analysis potentially indicates that some students may have been restrained or secluded multiple times. To illustrate, if a school district reported that it had 24 students enrolled, and also reported that it had 100 incidents of restraint, these reported data would indicate that the reporting was erroneous or that some students were restrained multiple times. Based on this analysis, we then calculated the average number of incidents (of restraint and seclusion) per student affected. Analysis of Extent of Illogical Data To test for illogical data, we analyzed the restricted-use restraint and seclusion data file for schools that reported more students affected than incidents. To illustrate, if a district reported that a school had restrained 80 students, and also reported that the school had 40 incidents of restraint, these reported data are illogical. Education has a business rule to detect illogical data at the school level, but the rule applies only to schools with more than 100 incidents. For our analysis, we looked for all schools with illogical data to determine the prevalence. School District Interviews on Interpreting CRDC Definitions of Restraint and Seclusion To determine how selected school districts interpret the CRDC definitions of restraint and seclusion, we selected 11 schools and nine school districts in three states to serve as illustrative examples. In total, we interviewed about 50 school officials. Information we collected from our 11 selected schools and nine districts cannot be generalized to all districts and schools nationwide. We selected states, districts, and schools to obtain a range of perspectives on federal reporting of restraint and seclusion data. Our selection also accounted for other criteria, such as selecting states that had laws requiring reporting; high or low rates of reporting zeros among districts; relatively high or low rates of restraint or seclusion per capita; grade levels served (e.g., K-6 or 9-12); school type (e.g., traditional or charter); and significant changes—increase or decrease—in incidents across reporting periods. We also selected districts that had reported incidents. As a result, we selected nine school districts to visit: two in Kentucky, three in Washington, three in Wisconsin, and a charter district in Wisconsin (see table 7). To determine how district and school officials, such as assistant superintendents, program managers, department directors, principals, and teachers, were interpreting the CRDC definitions of restraint and seclusion, we made the following statements and asked the following questions in our interviews. 1. We are going to talk to you about the definitions of restraint and seclusion that appear in the CRDC. We have heard that these definitions are not always clear to educators, so we want to get your feedback. Mechanical Restraint: the use of any device or equipment to restrict a student’s freedom of movement. Do you think this definition is clear or does it leave room for ambiguity? Physical Restraint: a personal restriction that immobilizes or reduces the ability of a student to move his or her torso, arms, legs, or head freely. The term does not include a physical escort. Physical escort means a temporary touching or holding of the hand, wrist, arm, shoulder or back for the purpose of inducing a student who is acting out to walk to a safe location Do you think this definition is clear or does it leave room for ambiguity? How do you differentiate between physical escort and physical restraint? Does breaking up a fight constitute a restraint? Seclusion: the involuntary confinement of a student alone in a room or area from which the student is physically prevented from leaving. It does not include a timeout, which is a behavior management technique that is part of an approved program, involves the monitored separation of the student in a non-locked setting, and is implemented for the purpose of calming. Do you think this definition is clear or does it leave room for ambiguity? How do you differentiate between timeout and seclusion? What does physically prevented (from leaving) mean? In what types of physical spaces can seclusion occur? Does your district have dedicated spaces for seclusion rooms? Can you describe where they are generally located, e.g., which types of schools or classrooms? 2. How do staff determine when an incident needs to be recorded as a restraint? 3. How do staff determine when an incident needs to be recorded as a seclusion? We conducted this performance audit from November 2018 to April 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the Department of Education Appendix III: Restraint and Seclusion Data for 20 Largest School Districts Reporting Incidents Appendix IV: School Districts with Relatively High Rates of Reported Seclusion Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Sherri Doughty (Assistant Director), Lara Laufer (Analyst-in-Charge), Morgan Jones, and Kristin Petroff, made key contributions to this report. Also contributing were James Bennett, Deborah Bland, Tonnye Conner-White, Holly Dye, Gretta Goodwin, Sheila R. McCoy, Jean McSween, John Mingus, James Rebbe, and Manuel Valverde.

Why GAO Did This Study Every 2 years, Education requires nearly all school districts to report incidents of restraint and seclusion. Generally, restraint is restricting a student's ability to move, and seclusion is confining them alone in a space they cannot leave. The House Committee on Appropriations' explanatory statement accompanying the Consolidated Appropriations Act of 2018 included a provision for GAO to evaluate the CRDC's restraint and seclusion data. This report examines (1) the effectiveness of CRDC data quality control procedures, (2) selected districts' interpretation of CRDC's restraint and seclusion definitions, and (3) selected districts' use of data. GAO analyzed CRDC's quality control processes for school year 2015-16, and interviewed officials from seven stakeholder groups and over 50 school and district officials in three states. GAO selected states, districts, and schools to obtain a range of perspectives on using restraint and seclusion data and interpreting CRDC definitions of restraint and seclusion. Selection criteria included changes in reported incidents year to year and laws requiring districts to report incidents to states. What GAO Found The Department of Education's (Education) quality control processes for data it collects from public school districts on incidents of restraint and seclusion are largely ineffective or do not exist, according to GAO's analysis of school year 2015-16 federal restraint and seclusion data—the most recent available. Specifically, Education's data quality control processes were insufficient to detect problematic data in its Civil Rights Data Collection (CRDC)—data Education uses in its efforts to enforce federal civil rights laws (see figure). For example, one rule Education used to check the quality of data submitted only applied to very large school districts, although GAO and Education's own analyses found erroneous reporting in districts of all sizes. Education also had no rules that flagged outliers that might warrant further exploration, such as districts reporting relatively low or high rates of restraint or seclusion. GAO tested for these outliers and found patterns in some school districts of relatively low and high rates of restraint or seclusion. Absent more effective rules to improve data quality, determining the frequency and prevalence of restraint and seclusion will remain difficult. Further, Education will continue to lack information that could help it enforce various federal civil rights laws prohibiting discrimination. Officials in the nine school districts GAO visited lacked a common understanding of the CRDC's restraint and seclusion definitions. Similarly, officials GAO interviewed in all three state educational agencies (Kentucky, Washington, and Wisconsin) and all seven stakeholder groups expressed similar concerns about the clarity of these definitions. For example, officials inconsistently interpreted the word alone in the definition of seclusion and, therefore, on whether to count an incident if a teacher was in the room. Absent clearer definitions, Education will continue to lack quality information on restraint and seclusion in public schools. Officials in school districts GAO visited identified several benefits to collecting these data, including identifying patterns in student behavior and developing interventions that can reduce the need for restraint and seclusion. Officials also said that analyzing their data helped them identify needs for additional staff training and student support services. What GAO Recommends GAO made six recommendations, including that Education expand its CRDC business rules to cover all districts, develop additional quality controls to address misreporting, address factors underlying misreporting, and refine and clarify its definitions. Education agreed with these recommendations.

gao_GAO-19-636

gao_GAO-19-636_0

Background Securing the Mass Transit Environment Mass transit rail operators have primary responsibility for securing their own systems. Unlike the aviation environment, where TSA has operational responsibility for screening passengers and baggage for prohibited items prior to boarding a commercial aircraft, TSA has no operational role for securing mass transit, such as employing screeners or purchasing and acquiring security equipment. Rather, TSA regularly partners with mass transit operators to address their security needs by conducting vulnerability assessments, sharing intelligence information and best practices, and working to mitigate security risks to their systems by assessing commercially available security technologies, among other measures. Mass transit operators which can be public or private entities, administer and manage all transit activities and services, including acquiring and operating any technologies designed to augment their existing security infrastructure. Securing transit systems presents inherent challenges for mass transit operators for numerous reasons. In general, mass transit systems are designed to expedite the movement of large numbers of people through multiple stations, situated along extended routes, and technologies used in the mass transit environment should not disrupt the efficiency of these operations. In addition, individual stations within these systems frequently include multiple points of entrance and exit that vary in the extent to which they may be accessed by passengers. For example, open systems include walk-up platforms with little to no barrier to entry, while other, more closed systems, typically include dedicated points of entry and exit that allow or prohibit entry access through various mechanisms. Given the size and complexity of these systems, it can be difficult for operator personnel to comprehensively monitor them for security threats. Finally, the large number of riders that pass through these systems during peak hours generally makes the sustained use of some security measures and technologies (e.g., metal detectors) difficult because such measures could result in long lines that would disrupt scheduled service. Though mass transit systems are difficult to secure, mass transit operators commonly employ a number of standard security measures for the transit environment, including: public awareness announcements and signage (e.g., reminders to report unattended baggage or suspicious activities to operator personnel immediately); use of canine teams; access controls, such as the use of lasers for intrusion detection; station design (e.g., designing transit stations to limit recess areas where bombs could be hidden, such as under a bench or a ticket machine); and video surveillance, which includes video cameras that transmit a signal to a set of television monitors to display real-time footage of transit system platforms, entrances, exits, etc. (see figure 1). Requirements, Roles, and Responsibilities for Mass Transit R&D Federal requirements for mass transit R&D efforts guide department and agency efforts. Specifically, the Implementing Recommendations of the 9/11 Commission Act of 2007 requires that S&T, in consultation with TSA and the Federal Transit Administration, carry out an R&D program to improve the security of public transportation systems. Additionally, Executive Order 13416, Strengthening Surface Transportation Security, requires the Secretary of Homeland Security to coordinate research, development, testing, and evaluation of technologies related to the protection of surface transportation, including commercial off-the-shelf products. To implement these requirements and other related activities to help secure mass transit systems, DHS, S&T, and TSA, in coordination with mass transit operators, have assumed the following roles and responsibilities: DHS. DHS carries out its requirement to coordinate department-wide R&D (including that for mass transit and other surface transportation modes) through its Integrated Product Team (IPT) process. The IPTs themselves are comprised of officials across DHS components, and are tasked with identifying DHS technology capability gaps, which are defined as differences between a department or agency’s current capabilities and those capabilities needed to perform its mission. Each IPT is responsible for identifying capability gaps related to a broad security area (such as preventing terrorism) and potential R&D efforts to address those gaps. Capability gaps relevant to surface transportation, including mass transit, fall under the Prevent Terrorism IPT and its Explosive Screening sub-IPT. S&T’s Research Council receives IPT information on security gaps and is responsible for prioritizing them across all IPTs. The results are used to inform which R&D research projects S&T and DHS components will undertake (see figure 2). S&T. Once capability gaps are agreed upon and prioritized by DHS components through the IPT process, S&T undertakes R&D projects intended to address the highest prioritized capability gaps. S&T can undertake R&D projects on behalf of any of the department’s components, including TSA, and will either initiate development of a new technology solution or coordinate or adapt existing technologies to meet the project’s needs. Once S&T’s R&D efforts result in a preliminary technology, or prototype, S&T will begin the testing process. S&T’s technology development process includes developmental and operational testing phases that are carried out by staff at laboratories S&T contracts with. Developmental testing is typically conducted in simulated environments, such as laboratories, test facilities, or engineering centers, which can sometimes be representative of the complex operational environment (i.e., an actual subway station). Operational testing includes field tests performed under realistic conditions by actual users (i.e., the transit operators testing in a subway or rail station) and overseen by S&T in order to determine the effectiveness and suitability of a prototype technology. Once testing is complete, S&T contracts with staff at partner laboratories and works with private sector industry partners on further developing the product for the commercial market, where it can be purchased by mass transit operators to help secure their systems. TSA. While TSA does not conduct R&D for mass transit security, it does sponsor testing of commercially available security technologies for the mass transit environment. This testing can take place in both laboratory environments and in operational, real-world environments. Regarding the latter, TSA uses its Surface Transportation Operational Test Bed Program (Operational Test Bed Program) to place commercially available security technologies in surface transportation environments, such as mass transit systems, for performance testing and evaluation. TSA established the program to assess the effectiveness of emerging and existing security technologies in real-world environments; verify prior laboratory testing performance results versus performance in a TSA mass transit system (or other surface transportation mode serving as a test bed); and develop recommendations for use of certain technologies in surface transportation. As part of the program, TSA establishes memorandums of agreement with surface transportation entities that participate in the program and also provides them with logistical support, such as installing technology and providing personnel to help operators with technology training and operating needs. As of 2019, there are nine mass transit operators participating as test beds in the program (test beds), and TSA officials reported working to add two more mass transit agencies as test beds. In addition, there are five other surface transportation test beds, including two pipeline and two freight rail test beds. Mass Transit Operators. Mass transit operators generally do not have dedicated portions of their budgets for, and therefore do not conduct, R&D. However, selected operators work with TSA to test commercially developed technology solutions intended to enhance their system security. In general, mass transit operators must assume security-related expenses for their systems, including the purchase or acquisition of surface security technologies. S&T Has One Surface Transportation R&D Program Under Way, but Is Not Following DHS Guidance to Track Its Progress S&T Is Developing Technologies to Address Explosive Threats within Mass Transit Stations In 2010, S&T’s Explosives Division began work on the Surface Transportation Explosive Threat Detection (STETD) program to address the threat of improvised explosive devices (IED) on persons or in objects within a mass transit station. S&T officials stated that, as of fiscal year 2019, the STETD program remains S&T’s sole R&D program related to surface transportation. In addition, it is the only DHS technology development program focused on developing products to address the threat of IEDs in a mass transit security environment. The STETD program consists of four separate technologies designed to address explosive threats within a mass transit station, each of which is in a different stage of development and maturation. Specifically, the four technologies are known as Forensic Video Exploitation and Analysis (FOVEA), Standoff Detection, Real-Time Threat Detection Agent, and Layered Architecture. The technologies are designed to address unique aspects of the mass transit environment (i.e., multiple access points, lack of access barriers, etc.) while also working together to provide IED detection coverage from the point at which the passenger enters a mass transit station, boards a train, and then finally exits the system. These technologies would allow mass transit operators to scan large unstructured crowds to detect concealed explosives worn or carried on a person (person-borne IED), or placed in stationary objects such as baggage, or intentionally deposited in an unnoticed location for detonation by a timer or remote control (leave-behind IED). S&T intends for STETD technologies to perform without requiring checkpoint baggage screening or other measures that could impede the traveling public moving through mass transit systems during periods of high passenger volume (e.g., rush hour). FOVEA is a software suite designed to interface with video management systems already installed in mass transit systems, and is intended to help operators use recently recorded camera footage to quickly determine a person’s movement through the system. S&T began developing the technology in fiscal year 2013, and it is generally directed at helping operators identify responsible parties when objects are left behind in a mass transit system. Specifically, FOVEA includes a number of tools to enable its video analysis (see figure 3). As of December 2018, the FOVEA video suite has been installed in the Washington Metropolitan Area Transit Authority’s Security Operations Control Center. During our site visit to the control center to view the use of the FOVEA suite, officials told us that FOVEA has enhanced the ability of personnel to analyze video footage for active law enforcement investigations. Officials also told us that because of this functionality, FOVEA is a valuable tool. S&T anticipates it will transition the technology to commercial development in fiscal year 2020. Standoff Detection S&T is developing a set of imaging sensors designed to scan unstructured crowds to detect hidden potential threat items (e.g., a person-borne IED) on travelers without requiring passengers to open bags or remove outerwear. According to S&T, the sensor technology would be placed in walls, near platforms, or other structures. These screening devices are designed to unobtrusively scan for detailed information on possible person-borne threat objects, such as wires connected to a pressure cooker, and provide alerts to operators via the Real-Time Threat Detection Agent (described below). Development of these sensors began in fiscal year 2014, and since 2017, S&T officials have been using the Massachusetts Bay Transportation Authority training facility to test prototypes of the technology. S&T expects to transition the technology to commercial development in fiscal year 2023. Real-Time Threat Detection Agent The Real-Time Threat Detection Agent technology is intended to automatically detect abandoned objects that could be potential threats, and notify transit operators of their existence. Specifically, the system is to analyze live video footage to identify, tag (i.e., mark on the video footage), and track left-behind objects (i.e., baggage possibly containing IEDs), as well as individuals associated with the object, without the need for continuous human monitoring of video footage. To track potential IEDs without human monitoring, the system is intended to have the capability to identify abandoned objects that could be potential threats and compare them against defined criteria for person-borne, as well as leave-behind, IED threats. Based on its analysis, the system would then create alerts and send them to operators, as well as to video review software, such as FOVEA. Development of this system began in fiscal year 2013, and S&T began developmental testing of the technology at Washington Metropolitan Area Transit Authority facilities in October 2018. S&T expects to transition the equipment to commercial development in fiscal year 2021. Layered Architecture Layered architecture, the final component of the STETD program, is intended to have the capability to integrate information from the various existing security technologies utilized by a mass transit system to enable more accurate threat identification. The goal of this component is to gather input from multiple pieces of technology, such as distributed sensors and tools used across a mass transit environment (to include existing sensors and other STETD technologies), to present a consolidated threat profile to operators in a command center. This technological component is the least developed of the STETD program technologies, with ongoing work focused on experimentation with different prototypes. S&T expects to transition the technology to commercial development in fiscal year 2023. STETD Program Funding Since 2013, S&T’s funding for mass transit R&D has decreased, delaying the development of associated technologies. Specifically, during fiscal years 2013 through 2017 funding for the STETD program—the only DHS R&D program focused solely on mass transit security—decreased by 78 percent, but then increased again in fiscal year 2018 (see figure 4). According to S&T officials, one reason funding was reduced during this period was to direct additional funds to R&D for a newly-identified threat. Specifically, in fiscal years 2015 and 2016, following the landing of a gyrocopter on the U.S. Capitol grounds and other incidents, unmanned aerial systems became a significant and emerging threat, and a top DHS priority. At the time, S&T had no funding allocated for a related R&D effort, so S&T leadership subsequently redirected funding from the STETD program toward R&D on unmanned aerial systems. S&T officials told us that, due to fluctuations in funding, in addition to other factors, the program’s completion date has shifted from 2017 to 2023, which has delayed efforts to make the technologies commercially available to mass transit operators. Although S&T began increasing funds for the STETD effort in 2018, according to program officials, decreases in program funding have delayed program deliverables and pushed timelines out. For example, according to S&T officials, lower levels of program funding have made it difficult to employ highly-skilled contract staff at the laboratories S&T partners with to carry out STETD R&D, which has slowed the pace of development. S&T officials also stated that a lack of funding and changes in funding slow down what is already a technically challenging development effort. As S&T officials explained, the STETD program is pushing the performance boundaries and capabilities of existing technologies, and in some cases, inventing entirely new technologies for screening highly trafficked environments. S&T Has Not Developed Milestones to Effectively Track the Program’s Progress S&T is not using milestones that fully adhere to DHS guidance for milestone descriptions to track its progress on developing STETD technologies. Specifically, DHS budget development guidance directs DHS components to develop program milestones that are specific, measurable, results-oriented and relevant, and time-bound. To be results-oriented and relevant, milestones must clearly link to activities in program strategy, budget, or other planning documents. Linking milestones to such activities allows parties reviewing the milestones, such as DHS leadership and Congress, to understand how achieving the milestones move the development process forward overall. We assessed all 22 STETD milestones that S&T has used to report progress on the program from fiscal years 2013 through fiscal year 2018. We found that 17 of the 22 milestones were not results-oriented as required by DHS guidance because they did not clearly link to any key activities described by STETD program documents. Specifically, one STETD program document identified several key activities for completing work on the technologies, including a requirements development phase, developmental testing, and operational testing, but STETD milestones did not clearly link to these activities. For example, One fiscal year 2018 STETD milestone for the layered architecture technology was to conduct a simulation and analysis of layered sensing configurations to optimize sensor placement and system performance, to be completed by the end of fiscal year 2018. While this milestone was specific, measurable, and time-bound, it did not clearly link to a key activity (e.g., developmental or operational testing) identified in the STETD program’s plan, and thus was not results- oriented. Another milestone from 2013 was to demonstrate advanced leave- behind detection software in a mass transit system. The milestone was to be met by the second quarter of fiscal year 2014, and remained unmet as of March 2019. Because the milestone was not results-oriented (i.e., it did not link back to activities in program documents), it was unclear how failing to achieve the milestone impacted, or potentially delayed, the overall technology development process. S&T officials explained because they are dealing with technology innovation and invention, they plan to develop milestones that closely align to program plans after they develop a potential technology solution that is ready for developmental or operational testing. However, according to STETD program plans, it can take several years for STETD technologies to begin developmental testing. For example, one STETD technology (layered architecture) is not expected to begin developmental testing until the fiscal year 2021-2022 time frame. Therefore, under S&T’s current practice, the program would not begin using results-oriented milestones, clearly linked to program plans, for this STETD technology until more than ten years after work on the program was initiated. Furthermore, we found that for one STETD technology (FOVEA), the program did not consistently use results-oriented milestones after the technology began developmental testing. Specifically, in fiscal year 2015 S&T began developmental testing for FOVEA, but, two of three FOVEA milestones reported in fiscal years 2017 and 2018 were not results- oriented. Without milestones for the STETD program that reflect DHS guidance to clearly link the milestone to key events in program planning documents, Congress and DHS decision makers cannot fully assess whether the STETD program is meeting its goals within identified time frames. Additionally, DHS decision makers are not positioned to identify adjustments that may be needed to facilitate the achievement of program goals. We previously recommended in March 2019 that S&T take steps to more fully incorporate DHS’s budget development guidance, to include more results-oriented milestones, for its R&D programs. DHS concurred with this broader recommendation, and as of June 2019, is taking initial steps to ensure its implementation. TSA Prioritizes Tests of Technology Products, but Projected Funding Shortfalls May Reduce the Scope of Future Testing TSA Prioritizes the Testing of Technology Products to Secure Mass Transit Systems through Its Operational Test Bed Program TSA sponsors tests of commercial products at contracted partner laboratories, as well as at mass transit stations and other surface transportation venues through its Operational Test Bed Program. The purpose of these tests is to inform surface transportation operators about different technological products that could address their security needs and to confirm whether the products will operate effectively. As part of the testing process, TSA officials assist in product installation, hold technical demonstrations, and provide training for mass transit officials. Once product testing concludes, TSA officials document test results in written assessments, which they make available to mass transit and other surface transportation stakeholders to review upon request. TSA officials told us they use a number of methods to identify products that are currently available in the commercial marketplace and could be tested. Officials stated they conduct market research on vendors currently making technology products that could potentially meet the needs of mass transit operators, such as portable screening devices used to detect potential person-borne IEDs. To do so, officials maintain and utilize an existing list of vendors, conduct market research on their products, attend relevant symposiums and university conferences on technological advancements, and solicit information on these products and their capabilities from both vendors and operators who have used them. TSA officials also work with national laboratories to assist with relevant research on specific technological products and their capabilities that could be good candidates for testing. To determine which technology products to test, TSA prioritizes technologies that can be used in the mass transit environment. TSA officials told us that because of the level of risk facing mass transit systems, they generally try to ensure that the commercial products that are tested address capability gaps relevant to the mass transit environment. In addition, TSA officials stated that while they try to address all identified surface transportation capability gaps, due to limited resources, they tend to select, on an annual basis, products for testing related to the following gaps: anomaly and explosive detection, high throughput threat detection, intrusion detection, infrastructure protection, and chemical and biological threat security, all of which have applicability to the mass transit environment. Moreover, TSA officials told us that anomaly and explosive detection and high throughput threat detection are the technology gaps that are critical to securing mass transit systems. TSA officials also identify other criteria used for testing, including performance requirements and vendor claims about their products’ ability to address specific capability gaps. Regarding performance requirements, TSA officials told us any products selected for testing must meet a set of minimum performance requirements in order to be considered appropriate for addressing the unique security needs for different surface transportation modalities. For example, for an explosive screening product used in mass transit, TSA officials established requirements for probability of detection and probability of false alarm. TSA officials stated they work with mass transit operators and others to identify these requirements to ensure that IED detection technologies can effectively identify threats without disrupting mass transit operations. Lastly, TSA officials said they also try to select technologies that can be used to secure multiple surface transportation modes. For example, officials said they may select an intrusion detection sensor for testing that could be adapted for securing pipelines or freight rail yards. TSA Tested Approximately 110 Existing Technology Products to Secure Surface Transportation Systems from Fiscal Years 2013 through 2018, but May Have to Reduce the Scope of Future Testing From fiscal years 2013 through 2018, TSA sponsored laboratory and field tests of approximately 110 commercial products that are designed to address identified surface transportation capability gaps (such as intrusion detection and explosive detection). These tests take place in either of two environments—laboratory or field (i.e., within a mass transit venue)—and are designed to address surface transportation security capability gaps. Since 2013, 67 percent of the technology products (72 products) assessed by TSA focused on detection-related gaps, most of which were related to intrusion detection. The remainder of products tested addressed interoperable information systems or a combination of capability gaps, such as anomaly and explosive detection and chemical and biological threat security (see figure 5). Of the products that addressed more than one capability gap, 78 percent (14 of 18) of these products could be used for anomaly and explosive detection, as well as high throughput threat detection. TSA officials told us that because anomaly and explosive detection and high throughput threat detection technologies can be easily transported to different locations within a station, they are of particular interest to mass transit operators. On our site visits to two mass transit operators that TSA utilizes to test technology products, we observed the testing and use of commercial products. These products were designed to detect anomalies on the underside of railcars as well as among persons traversing transit platforms, terminals, and stations. The products were being used to support both normal operations and a national security special event (see figure 6). Mass transit officials told us during our site visits that they considered these commercial technologies to be useful additions to their existing security measures. They also said the test results, which TSA made available to them, were helpful in determining whether to invest in purchasing the products for long-term use. In addition to allowing TSA to perform technology assessments, the program also gives transit operators hands-on experience with technologies they are unfamiliar with. For example, an official from one mass transit test bed told us that equipment often performs differently when the manufacturer’s employees are operating it due to their prior experience with, and dedicated training on, the equipment. This official noted that transit system employees, who often do not have similar experience and training with a particular technology, can sometimes have different performance results when operating this equipment. The official told us that TSA’s Operational Test Bed Program gives transit employees an opportunity to develop structured, hands-on experience with certain products with TSA’s assistance, allowing them to understand the full potential and capabilities of the technology. This official said that, in her case, observing fellow transit employees using a particular technology for several hours convinced her of the product’s application in a real-world environment, and subsequently was an important factor in her decision to recommend its purchase for use by her transit agency. Although mass transit operators we spoke with valued the Operational Test Bed Program, TSA has decreased funding for the program since fiscal year 2013. Specifically, our analysis of program funding showed that the program experienced an approximately 70 percent decrease in funding from fiscal years 2013 through 2018 (see figure 7). TSA officials stated that recent decreases in program funding, coupled with projected funding shortfalls for the Operational Test Bed Program for 2019 through 2024, will limit the program’s capacity to conduct testing and assessments of technologies. Specifically, the TSA program manager for the Operational Test Bed Program told us that the recent decreases in funding for the program to its current level will materially impact the operation of the program moving forward. Furthermore, a TSA May 2019 budget planning document shows that, to fully meet project requirements, the program will require approximately $20 million in additional funding for fiscal years 2019 through 2024. Should the program not receive this funding, TSA officials stated they would not be able to test as many products or address as many surface transportation capability gaps through the program. They also stated that the funding shortfalls would limit TSA’s analysis of technology performance. Program managers are in the process of identifying additional funding requirements for the program through TSA’s internal budget review process. Mass Transit Stakeholders Effectively Collaborate to Identify Capability Gaps and Test Security Technologies, but TSA Does Not Comprehensively Share Technology Assessments S&T, TSA, and Mass Transit Operators Have Effectively Collaborated to Identify Mass Transit Capability Gaps and Test Technology Solutions through Four Key Mechanisms S&T, TSA, and mass transit operators regularly collaborate on issues related to identifying mass transit capability gaps and testing security technologies to address those gaps. During the course of our review, we identified four key mechanisms that S&T, TSA, and mass transit operators use to collaborate on mass transit security issues—the DHS IPT process’s sub-IPT focusing on surface transportation capability gaps (including those pertaining to mass transit); the Intermodal Transportation Systems Research and Development Working Group (RDWG); TSA’s Operational Test Bed Program; and the Transit Policing and Security Peer Advisory Group (PAG) (see table 1). We assessed the effectiveness of collaboration in these four mechanisms using our leading collaboration practices (see side bar) and found that each of them generally followed these practices. We reported in March 2019 that DHS-wide R&D collaboration has improved through the IPT process but some challenges remain, such as ensuring all components participate in the process, among other things. However, as discussed below, we found that S&T and TSA collaborate effectively through the IPT process for identifying mass transit security capability gaps and security technologies. sustained over the long-term? If leadership is shared, have roles and responsibilities been clearly identified and agreed upon? Have participating agencies clarified roles and responsibilities? participants been included? Do they have the ability to commit resources for their agency? mechanism be funded and staffed? Have online collaboration tools been developed? Written Guidance and Agreements: If appropriate, have participating agencies documented their agreement regarding how they will be collaborating? Have they developed ways to continually update and monitor these agreements? Prevent Terrorism sub-IPT on Explosive Screening. S&T and TSA collaborate on identifying surface transportation (including mass transit) capability gaps through the Prevent Terrorism’s sub-IPT on Explosive Screening. At the federal level, we found that TSA and S&T’s ongoing use of this mechanism met several leading collaboration practices, including bridging organizational cultures, leadership, clarity of roles and responsibilities, participants, and written guidance and agreements. Specifically, the Explosive Screening sub-IPT has a formal structure that is outlined in the Prevent Terrorism IPT’s charter. This written agreement establishes leadership and clarifies the roles and responsibilities of each of the participants. As key participants and voting members, S&T and TSA possess the necessary expertise to identify capability gaps for mass transit (i.e., S&T has expertise in technology research and development, and TSA has in-depth knowledge of the mass transit and other surface transportation sectors). Moreover, by requiring S&T and TSA to work together to prioritize capability gaps, the process allows them to operate across agency boundaries (i.e., to bridge their respective organizational structures). In 2012, S&T officials stated that they collaborated with TSA as members of the Explosive Screening sub-IPT to identify anomaly and explosive detection and high-throughput threat detection as the highest priority capability gaps for surface transportation. These gaps were the basis for S&T’s STETD program. RDWG. We also found that the RDWG facilitates effective collaboration between S&T, TSA, and surface transportation stakeholders, including 30 mass transit operators. The RDWG generally follows leading collaboration practices related to bridging organizational cultures, leadership, clarity of roles and responsibilities, participants, resources, and written guidance and agreements. Specifically, the RDWG is a working group with an established charter that brings together federal and surface transportation representatives to operate across agency and sector boundaries to identify surface transportation-related capability gaps, including those for mass transit. The RDWG charter clarifies the participant roles and responsibilities and establishes a framework for nominating new members to ensure all relevant participants are included. TSA, as the designated chair, funds the working group and ensures its continuation. In addition to identifying surface transportation security gaps, an S&T official told us that the members of the RDWG review the prior year’s security capability gaps to determine whether they are still relevant and if there are commercially available technologies to address them. The S&T official explained that they then use the results of these reviews to inform their work on the Explosive Screening sub-IPT, specifically using them as basis for R&D project requirements. For example, S&T officials said that anomaly and explosive and high- throughput threat detection were the gaps identified by the participants of the RDWG and then reported through the DHS IPT process, which ultimately helped inform the scope of S&T’s STETD program. In addition, S&T officials told us that they use the RDWG annual meetings to communicate to surface transportation stakeholders the progress they have made to address these gaps through the STETD program. Operational Test Bed Program. TSA’s Operational Test Bed Program facilitates collaboration between S&T, TSA, and mass transit operators on the testing and evaluation of security technologies. The program, through its memorandums of agreement with mass transit operators, generally follows leading collaboration practices related to collaboration criteria for bridging organizational cultures, clarity of roles and responsibilities, participants, resources, and written guidance and agreements. Specifically, memorandums of agreement serve as a mechanism for TSA and S&T to operate outside of their agency boundaries to test technologies in real-world environments (i.e., mass transit systems). The agreements also clarify the roles and responsibilities and serve as guidance for TSA, S&T, and mass transit operators on how testing is to be carried out. For example, the agreements clarify responsibilities for installing and operating test equipment. The program is funded and managed by TSA, and S&T and mass transit operators leverage TSA’s resources through their participation. Collaboration through the program has led to numerous benefits for TSA, S&T, and mass transit operators. According to TSA officials, the program allows TSA to fulfill federal requirements to test and evaluate mass transit security technologies and to expand the market for these products. Additionally, S&T has used the program’s established agreements with transit systems to facilitate the testing of STETD program prototypes. Lastly, transit operators use the program to obtain first-hand information on the performance of security technologies within their system. For example, an official from one mass transit operator participating in the test bed program told us they purchased two different types of passive millimeter wave scanners that they tested through the program and found to be effective. Transit Policing and Security Peer Advisory Group (PAG). The PAG facilitates collaboration among mass transit stakeholders to share information and meets the bridging organizational cultures, leadership, participants, resources, and written guidance and agreements collaboration key features criteria. Specifically, the PAG follows several practices through its charter, which designates a transit police chief as the chair of the group. The charter also outlines the participating mass transit stakeholders and allows them to share information across agency boundaries on security-related issues, including information on incident response, emerging threats, and other best practices on mitigating security issues. Officials from all nine of the mass transit operators we spoke with are members of the PAG, and seven of them stated that the PAG fosters collaboration between mass transit operators by providing a forum for transit officials to connect and share information. Particularly, one mass transit operator stated that the PAG was beneficial to the safety and security of her system because it has allowed mass transit officials to share experiences and disseminate best practices for responding to threats. Another official said that the PAG helped him stay informed about the current risks facing all of the mass transit operators, and without the group, collaboration probably would not happen. TSA Engages in Collaborative Activities to Share Information on Security Technologies, but Does Not Comprehensively Share Technology Assessment Information TSA engages in a number of collaborative activities to share information on security technologies with mass transit stakeholders to help improve their technology investments. Specifically, TSA disseminates a quarterly newsletter to surface transportation stakeholders (including mass transit operators) which summarizes TSA’s efforts to address various surface transportation security issues. Among other things, the newsletter shares information on the technologies for which TSA has sponsored testing in different mass transit operator systems. Additionally, TSA maintains a collective email account that was created for all mass transit operators to send TSA suggestions for technologies or products to test, among other things. Further, TSA officials stated that they notify and communicate to regional mass transit stakeholders any information on upcoming test bed demonstrations, so these stakeholders can attend in person if they prefer. Finally, according to the TSA program manager, TSA officials utilize the American Public Transportation Association’s annual conference, industry symposiums, and security roundtables to engage with mass transit stakeholders to share information on security threats, capability gaps, and technology. In addition, to assist transit operators, TSA produces a number of assessments and reports (products) that include performance information on a range of technologies. These products include: TSA’s Market Survey. TSA officials regularly update a market survey, which contains a list of commercial vendors who develop technological products capable of addressing surface transportation security issues. This list, which does not contain sensitive information and may be readily shared with operators, includes vendors and their associated products that TSA believes may be applicable to mass transit security; it does not catalog the list of products TSA has sponsored testing for. TSA officials populate this list by attending relevant symposiums and university conferences, as well as soliciting input from partner laboratories. Surface Transportation Sensor Catalog. The Surface Transportation Sensor Catalog documents the technology assessments performed by TSA and contains summaries of various security technologies evaluated since 2007. In addition to evaluated products, it also contains summaries of the vendor product demonstrations received by TSA since fiscal year 2016. TSA officials stated that the catalog is updated each year with new entries for recently evaluated products, and is intended as a resource for both TSA and its stakeholders to have greater awareness of technologies and help them make more informed technology investment decisions. State of Technology Reports. TSA publishes a State of Technology report that provides a detailed overview of a specific challenge to surface transportation stakeholders (such as person-borne IEDs) and gives a high-level summary of available products that could address it, and a technology maturation roadmap (with objectives) that needs to be implemented in order to meet surface transportation stakeholders’ operational and security needs to address those specific threats. Although these TSA products contain technology assessment and other information that would benefit mass transit operators seeking to purchase and implement security technologies, mass transit operators may not be receiving them. Specifically, TSA shares these products with transit operators upon request. However, officials from seven of the nine mass transit operators that we spoke with said they wanted more technical assessment information on commercially available security technologies, indicating that they may not be routinely requesting, and therefore not receiving, the TSA products that would provide this information. In addition, four of the nine said they would like TSA’s assessment information on technologies to be more accessible. Finally, an official from one mass transit system who previously worked for TSA on mass transit issues, and thus has knowledge of the broader mass transit community, stated that many operators would benefit from the Surface Transportation Sensor Catalog, but smaller operators are not aware of this resource and therefore do not know how to request it. TSA officials stated they do not routinely share information on security technologies with mass operators for two reasons. First, TSA officials explained that many of the in-depth reports that result from its testing of security technologies contain sensitive information and cannot be distributed without first assessing whether the requester is eligible to receive it. However, officials from most of the mass transit operators we spoke with said they would like more technical assessment information. Therefore it could be useful for mass transit operators to know when TSA publishes these reports so they can request the full report for review. This notification could consist of non-sensitive, high-level information on technologies assessed so that mass transit operators could request the information in full. Second, the TSA program manager responsible for these assessments told us that his office does not have sufficient resources to develop and maintain a centralized, web-based repository that would allow mass transit operators to search and retrieve sensitive information independently, such as the sensor catalog and technology assessment reports. Despite these limitations, in the past, TSA has shared information related to technology assessments routinely with mass transit operators. For example, TSA used to post a verified technology list on a Federal Emergency Management Agency web page. A TSA official stated that the information posted on the website included summary information on technology evaluations and other technology information. Further, TSA received feedback from surface transportation (including mass transit) stakeholders that the information posted was useful. TSA no longer posts information on the web page because the Federal Emergency Management Agency no longer maintains the website. In addition, TSA officials stated that they had plans to include more comprehensive information about TSA’s technology assessments within an online information resource known as the DHS Responder Knowledge Base—a department-wide database previously developed to house information for first responders. In April 2019 officials from DHS’s Countering Weapons of Mass Destruction Directorate said they had plans to reach out to TSA and other DHS components on how to utilize the Responder Knowledge Base as a repository for their reports and other sensitive information. Furthermore, officials from that directorate told us that the database is about 2 years from being launched, and they do not have a specific completion date. Standards for Internal Control in the Federal Government states that management should communicate externally to their stakeholders through the appropriate means. Further, the 2013 National Infrastructure Protection Plan states that in order to ensure that situational awareness capabilities keep pace with the evolving risk environment, officials should improve practices for sharing information that will improve security and resilience. Until the Responder Knowledge Base is operational, mass transit operators could benefit from TSA routinely sharing appropriate information on the technology assessments and other performance information at its disposal. For example, TSA could leverage the resources of existing coordination mechanisms, like the PAG, or develop a listserv to automatically notify a more comprehensive group of mass transit operators of the existence of a new technology assessment or sensor catalog. Notifying more mass transit operators on an ongoing basis that this information is available would help ensure they have the benefit of all relevant TSA information when making strategic security technology investments. Further, doing so would help mass transit operators to better use their limited resources to acquire proven technologies that could enhance the overall security posture of their systems. Conclusions Monitoring and securing surface transportation systems continues to present unique challenges. With respect to mass transit systems, for example, operators must balance the need to efficiently move passengers through the system with the need to screen for explosives and other threats. Since 2010, S&T’s STETD program has been the only DHS R&D program that has developed technologies to address these challenges. Although S&T has made progress, as of fiscal year 2019, none of the technologies associated with the STETD program have matured enough to undergo commercial development, and the program’s completion date has been extended from fiscal year 2017 to fiscal year 2023. While fluctuations in the program’s funding have contributed to delays, S&T has not followed DHS guidance for developing milestones that would help officials understand whether the program is achieving key activities identified in planning documents when faced with funding and other challenges. Without milestones that clearly convey an understanding of the program’s progress, DHS decision makers are not positioned to identify any adjustments that may be needed to facilitate the achievement of program goals. S&T, TSA, and mass transit operators effectively collaborate through a number of stakeholder groups to identify mass transit security gaps and to test possible technology solutions that could address them. TSA also supports greater awareness of available technologies by publishing key information on commercially available products (such as technology assessment results) and making it available to mass transit operators upon request. However, TSA does not comprehensively or routinely share this information, and seven of the nine mass transit operators we spoke with stated they wanted more technology assessment information. Without a mechanism to share technology assessments and related information with more mass transit operators and on a routine basis, TSA cannot ensure that mass transit operators will be fully informed about available technologies they could use to secure their systems. Moreover, without this information, mass transit operators may not be positioned to make the best possible use of the limited funding available for purchasing these technologies. Recommendation for Executive Action We are making two recommendations, one to DHS and one to TSA. The Deputy Secretary of Homeland Security should ensure that S&T take steps to more fully incorporate practices for developing milestones within DHS’s budget preparation guidance, into the Surface Transportation Explosive Threat Detection program. (Recommendation 1) The Administrator of TSA should develop a mechanism to more routinely and comprehensively share appropriate information on the performance of mass transit security technologies (such as the annual sensor catalog and security technology assessments) with mass transit operators and stakeholders until DHS completes work on a more permanent information sharing resource. (Recommendation 2) Agency Comments and Our Evaluation We provided a draft of this report to S&T and TSA for review and comment. DHS provided written comments which are reprinted in appendix I. In its comments, DHS concurred with both recommendations and described actions planned to address them. S&T and TSA also provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees and the acting Secretary of Homeland Security. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact William Russell at (202) 512-8777 or russellw@gao.gov. Contact points for our Office of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix II. Appendix I: Comments from the U.S. Department of Homeland Security Appendix II: GAO Contact and Staff Acknowledgments William Russell (202) 512-8777 or RussellW@gao.gov. Staff Acknowledgments In addition to the contact named above, Christopher Ferencik (Assistant Director), Mona Nichols Blake (Analyst in Charge), Jason Blake, Frederick K. Childers, Michele Fejfar, Jonathan G. Felbinger, Eric Hauswirth, Tracey King, Kristiana D. Moore, Claire Peachey, Jack Sheehan, Sarah Veale, and Robert Ward made key contributions to this report.

Why GAO Did This Study Since 2016, bombings of subways and bus systems in foreign cities and attempted attacks in U.S. cities demonstrate continued security threats to mass transit and other surface transportation systems. S&T and TSA are the primary federal entities responsible for researching, developing, and testing technologies designed to address threats to these systems. GAO has previously identified challenges with S&T's oversight of R&D projects. GAO was asked to review S&T and TSA's roles in developing and testing surface transportation security technologies. This report, among other objectives, (1) assesses the extent to which S&T is developing technologies to secure surface transportation systems and progress made, and (2) identifies the key mechanisms that S&T, TSA, and stakeholders use to collaborate and share information on identifying capability gaps and security technologies, and analyzes the extent to which they are effective. GAO assessed S&T's mass transit program because it was the only active R&D effort for surface transportation security. GAO interviewed officials from S&T, TSA, and nine mass transit operators; observed technologies; reviewed documentation; and analyzed budget information from fiscal years 2013 to 2018. GAO also used GAO's leading collaboration practices to assess collaboration on security technologies. What GAO Found The Department of Homeland Security's (DHS) Science and Technology Directorate (S&T) has one research and development (R&D) effort focused on surface transportation, the Surface Transportation Explosive Threat Detection (STETD) program, which is developing technologies to secure mass transit systems (see figure). DHS guidance requires S&T to develop results-oriented milestones to track progress. GAO found, however, that S&T has not used milestones that fully adhered to DHS guidance. For example, most STETD program milestones did not clearly link to key activities described in program plans. As a result, DHS may not have the information needed to determine whether the STETD program is meeting its goals. S&T, TSA, and stakeholders effectively collaborate, but TSA could better share test results with mass transit stakeholders. For example, S&T, TSA, and mass transit operators regularly collaborate on issues related to identifying mass transit capability gaps and testing security technologies to address those gaps. Nevertheless, GAO found TSA's efforts to share information on existing technologies to secure mass transit could be improved. Specifically, TSA regularly assesses commercially available technologies, but does not routinely or comprehensively share its results with mass transit operators. For example, TSA's reports on its testing of commercially available products would provide mass transit operators with technical assessment information. However, seven of the nine mass transit operators GAO spoke with asked for more technical assessment information on existing commercial technologies, indicating that they may not be receiving the TSA products that would provide this information. Sharing this information more routinely and comprehensively with mass transit operators would allow TSA to better inform them about the capabilities of technologies that could be acquired to secure thteir systems. What GAO Recommends GAO is making two recommendations: that S&T incorporate DHS milestone guidance for its STETD program, and that TSA develop a mechanism to routinely and comprehensively share security technology information with mass transit operators. DHS concurred with both recommendations.

gao_GAO-20-262T

gao_GAO-20-262T_0

Background Globalization of Drug Manufacturing Drugs sold in the United States—including active pharmaceutical ingredients and finished dosage forms—are manufactured throughout the world. According to a May 2019 FDA report, in fiscal year 2018 about 40 percent of establishments manufacturing drugs for the U.S. market were located domestically and more than 60 percent of establishments manufacturing for the U.S. market were located overseas. As of March 2019, FDA data show that India and China had the most manufacturing establishments shipping drugs to the United States, with about 40 percent of all foreign establishments in these two countries. (See fig. 1.) Types of Inspections Drugs manufactured overseas must meet the same statutory and regulatory requirements as those manufactured in the United States. FDA’s Center for Drug Evaluation and Research (CDER) establishes standards for the safety, quality, and effectiveness of, and manufacturing processes for, over-the-counter and prescription drugs. CDER requests that FDA’s Office of Regulatory Affairs (ORA) inspect both domestic and foreign establishments to ensure that drugs are produced in conformance with applicable laws of the United States, including current good manufacturing practice (CGMP) regulations. FDA investigators generally conduct three main types of drug manufacturing establishment inspections: preapproval inspections, surveillance inspections, and for-cause inspections, as described in table 1. At times, FDA may conduct an inspection that combines both preapproval and surveillance inspection components in a single visit to an establishment. FDA uses multiple databases to select foreign and domestic establishments for surveillance inspections, including its registration database and inspection database. Because the establishments are continuously changing as they begin, stop, or resume marketing products in the United States, CDER creates an establishment catalog monthly. The catalog is prioritized for inspection twice each year. In our 2008 report we found that, because of inaccurate information in FDA’s databases, the agency did not know how many foreign drug establishments were subject to inspection. For example, some establishments included in FDA’s registration database may have gone out of business and did not inform FDA that they had done so or did not actually manufacture drugs for the U.S. market. In our report, we noted that some foreign establishments may register because, in foreign markets, registration may erroneously convey an “approval” or endorsement by FDA, when in fact the establishment may never have been inspected by FDA. We recommended that FDA take steps to improve the accuracy of this registration information. In our 2010 and 2016 reports we found that FDA had taken steps to improve the accuracy and completeness of information in its catalog of drug establishments subject to inspection, such as using contractors to conduct site visits to verify the existence of registered foreign establishments and confirm that they manufacture the products that are recorded in U.S. import records. To prioritize establishments for surveillance inspections, CDER applies a risk-based site selection model to its catalog of establishments to identify those establishments (both domestic and foreign) that, based on the characteristics of the drugs being manufactured, pose the greatest potential public health risk should they experience a manufacturing defect. This model analyzes several factors, including inherent product risk, establishment type, inspection history, and time since last inspection, to develop a list of establishments that FDA considers to be a priority for inspection. Through this process, CDER develops a ranked list of foreign and domestic establishments selected for inspection that is submitted to ORA. To be efficient with its resources, according to FDA officials, ORA staff may shift the order of establishments to be inspected on CDER’s prioritized list based on geographic proximity to other planned inspection trips. FDA Inspection Workforce Investigators from ORA and, as needed, ORA laboratory analysts with certain expertise are responsible for inspecting drug manufacturing establishments. FDA primarily relies on three groups of investigators to conduct foreign inspections: ORA investigators based in the United States, who primarily conduct domestic drug establishment inspections but may sometimes conduct foreign inspections. Members of ORA’s dedicated foreign drug cadre, a group of domestically based investigators, who exclusively conduct foreign inspections. Investigators assigned to and living in the countries where FDA has foreign offices, including staff based in the foreign offices full time and those on temporary duty assignment to the foreign offices. FDA began opening offices around the world in 2008 to obtain better information on the increasing number of products coming into the United States from overseas, to build relationships with foreign stakeholders, and to perform inspections. FDA full-time foreign office staff are posted overseas for 2-year assignments. FDA staff can also be assigned to the foreign offices on temporary duty assignments for up to 120 days. In fiscal year 2019, there were full-time and temporary duty drug investigators assigned to FDA foreign offices in China and India. Post-Inspection Activities FDA’s process for determining whether a foreign establishment complies with CGMPs involves both CDER and ORA. During an inspection, ORA investigators are responsible for identifying any significant objectionable conditions and practices and reporting these to the establishment’s management. Investigators suggest that the establishment respond to FDA in writing concerning all actions taken to address the issues identified during the inspection. Once ORA investigators complete an inspection, they are responsible for preparing an establishment inspection report to document their inspection findings. Inspection reports describe the manufacturing operations observed during the inspection and any conditions that may violate U.S. statutes and regulations. Based on their inspection findings, ORA investigators make an initial recommendation regarding whether regulatory actions are needed to address identified deficiencies using one of three classifications: no action indicated (NAI); voluntary action indicated (VAI); or official action indicated (OAI). Inspection reports and initial classification recommendations for regulatory action are to be reviewed within ORA. For inspections classified as OAI—where ORA identified serious deficiencies—such inspection reports and classification recommendations are to be reviewed within CDER. CDER is to review the ORA recommendations and determine whether regulatory action is necessary. CDER also is to review inspection reports and initial classification recommendations for all for-cause inspections, regardless of whether regulatory action is recommended by ORA. According to FDA policy, inspections classified as OAI may result in regulatory action, such as the issuance of a warning letter. FDA issues warning letters to those establishments manufacturing drugs for the U.S. market that are in violation of applicable U.S. laws and regulations and may be subject to enforcement action if the violations are not promptly and adequately corrected. In addition, warning letters may notify foreign establishments that FDA may refuse entry of their drugs at the border or recommend disapproval of any new drug applications listing the establishment until sufficient corrections are made. FDA may take other regulatory actions if it identifies serious deficiencies during the inspection of a foreign establishment. For example, FDA may issue an import alert, which instructs FDA staff that they may detain drugs manufactured by the violative establishment that have been offered for entry into the United States. In addition, FDA may conduct regulatory meetings with the violative establishment. Regulatory meetings may be held in a variety of situations, such as a follow-up to the issuance of a warning letter to emphasize the significance of the deficiencies or to communicate documented deficiencies that do not warrant the issuance of a warning letter. The Number Of Foreign Inspections Declined In Recent Years, And The Majority Of Such Inspections Identified Deficiencies Total Number of FDA Foreign Drug Inspections Has Decreased Since Fiscal Year 2016 after Several Years of Increases Our preliminary analysis of FDA data shows that from fiscal year 2012 through fiscal year 2016, the number of FDA foreign drug manufacturing establishment inspections increased but then began to decline after fiscal year 2016 (see fig. 2). In fiscal year 2015, the total number of foreign inspections surpassed the number of domestic inspections. From fiscal year 2016 to 2018, both foreign and domestic inspections decreased—by about 10 percent and 13 percent, respectively. FDA officials attributed this decrease to vacancies in the number of investigators available to conduct inspections (which we discuss later in this testimony statement) and to inaccurate data used to select establishments for inspection in fiscal years 2017 and 2018. Despite steps taken to improve the accuracy and completeness of FDA data on foreign establishments, data challenges we identified in our 2008 report continue to make it difficult for FDA to accurately identify establishments subject to inspection. Specifically, since 2017, FDA has pursued an initiative to inspect approximately 1,000 foreign establishments that lacked an inspection history and, as of November 2019, officials said all of these establishments had either been inspected or were determined to not be subject to inspection. However, officials told us that this effort contributed to the decline in the number of foreign inspections conducted because of how data inaccuracies affected the process for selecting establishments for inspection. Specifically, after selecting uninspected foreign establishments for inspection, FDA determined that a sizeable percentage of these establishments were not actually subject to inspection (e.g., about 40 percent of those assigned to the China Office in fiscal years 2017 and 2018). These foreign establishments were thus removed from the list for inspection for the given year. FDA officials told us that the next highest priority establishments identified through the risk- based model to replace those establishments were domestic. As a result, the number of foreign establishments actually inspected decreased. As part of our ongoing work, we plan to examine the accuracy and completeness of information FDA maintains about foreign establishments and the application of its risk-based site selection process. FDA continues to conduct the largest number of foreign inspections in India and China, with inspections in these two countries representing about 40 percent of all foreign drug inspections from fiscal year 2016 (when we last reported on this issue) through 2018. (See table 2.) In addition to India and China, the rest of the countries in which FDA most frequently conducted inspections has generally been the same since our 2008 report. Most Foreign Inspections Are Surveillance Inspections Our preliminary analysis of FDA data shows that each year from fiscal year 2012 through 2018, at least 50 percent of FDA’s foreign inspections were surveillance inspections. In contrast to preapproval inspections, surveillance inspections are used to ensure drugs already on the market are manufactured in compliance with FDA regulations. In recent years, the proportion of foreign surveillance inspections has increased. As figure 3 shows, in fiscal year 2012, 56 percent of foreign inspections were surveillance-only inspections; in contrast, from fiscal year 2016 through 2018, about 70 percent of foreign inspections were surveillance-only, which was comparable to the percentage for domestic inspections during that period. This is a significant increase from the 13 percent of foreign inspections that were surveillance-only when we made our 2008 recommendation that FDA inspect foreign establishments at a comparable frequency to their domestic counterparts (85 percent of which were surveillance-only at that time). FDA has implemented changes to its foreign drug inspection program since our 2008 report that may have contributed to the increase in surveillance inspections. Prior to 2012, FDA was required to inspect domestic establishments that manufacture drugs marketed in the United States every 2 years, but there was no similar requirement for foreign establishments. As a result, and as we reported in 2008, foreign inspections were often preapproval inspections driven by pending applications for new drugs. FDA thus conducted relatively few surveillance-only inspections to monitor the ongoing compliance of establishments manufacturing drugs that were already on the market, with just 13 percent of foreign inspections conducted for surveillance purposes at the time of our 2008 report. However, in 2012, the Food and Drug Administration Safety and Innovation Act eliminated the 2-year requirement for domestic inspections, directing FDA to inspect both domestic and foreign establishments on a risk-based schedule determined by an establishment’s known safety risks, which was consistent with our 2008 recommendation. FDA Identified Deficiencies during the Majority of Foreign Inspections Our preliminary analysis of FDA data shows that from fiscal year 2012 through 2018, FDA identified deficiencies in approximately 64 percent of foreign drug manufacturing establishment inspections (3,742 of 5,844 inspections). This includes deficiencies necessitating a classification of VAI or the more serious OAI, as described in the text box. Inspection Classifications Based on their inspection findings, FDA investigators make an initial recommendation regarding the classification of each inspection: No action indicated (NAI) means that insignificant or no deficiencies were identified during the inspection. Voluntary action indicated (VAI) means that deficiencies were identified during the inspection, but the agency is not prepared to take regulatory action, so any corrective actions are left to the establishment to take voluntarily. Official action indicated (OAI) means that serious deficiencies were found that warrant regulatory action. About 59 percent of domestic inspections (3,702 out of 6,291) identified deficiencies during this time period. (See fig. 4.) This proportion is similar to what we found when we last looked at this issue in 2008, when FDA identified deficiencies in about 62 percent of foreign inspections and 51 percent of domestic inspections from fiscal years 2002 through 2006. Our preliminary analysis showed that serious deficiencies identified during foreign drug inspections classified as OAI—which represented 8 percent of inspections from fiscal year 2012 through 2018—include CGMP violations such as those related to production and process controls, equipment, records and reports, and buildings and facilities. For example: Failure to maintain the sanitation of the buildings used in the manufacturing processing, packing, or holding of a drug product (21 C.F.R. § 211.56(a) (2019)). At an establishment in India producing finished drug products, the investigator reported observing a live moth floating in raw material used in the drug production, and that the facility staff continued to manufacture the drug products using the raw material contaminated by the moth, despite the investigator pointing out its presence. Failure to perform operations relating to the manufacture, processing, and packing of penicillin in facilities separate from those used for other drug products (21 C.F.R. § 211.42 (d) (2019)). At an establishment in Turkey that manufactured penicillin and other drugs, the investigator reported that the manufacturer had detected penicillin outside the penicillin manufacturing area of the establishment multiple times. According to FDA, penicillin contamination of other drugs presents great risk to patient safety, including potential anaphylaxis (even at extremely low levels of exposure) and death. The identification of serious deficiencies is not unique to foreign inspections. For example, at a domestic establishment producing finished drug products, the investigator observed brown stains, white residues, and brown stagnant water in manufacturing equipment. Some investigators who conduct foreign inspections expressed concern with instances in which ORA or CDER reviewers reclassify the investigator’s initial inspection classification recommendations of OAI to the less serious classification of VAI. We plan to examine this issue as part of our ongoing work. FDA Continues To Face Challenges Filling Vacancies Among Staff Conducting Foreign Inspections Our ongoing work showed FDA’s foreign inspection workforce has staff vacancies, which FDA officials said contributed to the recent decline in inspections. As previously mentioned, FDA uses multiple types of staff resources to conduct foreign drug inspections—including ORA investigators based in the United States, members of ORA’s dedicated foreign drug cadre based in the United States, and investigators assigned to FDA’s foreign offices. However, each of these groups has current vacancies. According to FDA officials, the agency is trying to fill vacancies in each of these groups, but the lower staff numbers may limit FDA’s ability to conduct more foreign inspections. ORA investigators based in the United States. This group of investigators conducts the majority of foreign inspections; about 76 percent of foreign inspections in fiscal year 2018 involved an ORA investigator based in the United States who conducts both foreign and domestic inspections. FDA officials said that the more experienced investigators from this group are expected to conduct three to six foreign inspections per year, and investigators hired using generic drug user fees are expected to inspect nine to 12 foreign establishments per year. As of June 2019, there were 190 investigators eligible to conduct foreign drug inspections, but officials said that as of November 2019, the agency had an additional 58 vacancies in this group. Officials said that the agency was in the process of hiring 26 ORA investigators based in the United States to fill these vacancies, with 32 vacancies remaining. FDA officials attributed the vacancies to multiple factors: investigator retirements, investigator movement to other parts of FDA, and the need to hire to additional investigator positions using generic drug user fees. Officials also said that a reorganization within ORA led to a reduced number of investigators who conduct drug manufacturing establishment inspections. While FDA recently filled several of the vacancies, officials told us that new investigators are not typically used for foreign inspections until they have been with the agency for 2 to 3 years. ORA dedicated foreign drug cadre. About 15 percent of foreign inspections in fiscal year 2018 involved an investigator from ORA’s dedicated foreign drug cadre—a group of ORA investigators based in the United States who exclusively conduct foreign inspections. FDA officials said that members of the cadre are expected to conduct 16 to 18 foreign inspections each year. According to FDA, the cadre had 20 investigators in 2012 and 15 investigators in 2016. However, the cadre had only 12 investigators as of November 2019, out of 20 available slots. According to FDA officials, the agency is attempting to fill these positions from the current ORA investigator pool, but officials are not confident that all 20 slots will be filled. Investigators assigned to FDA’s foreign offices. Approximately 7 percent of foreign inspections in fiscal year 2018 involved investigators from FDA’s foreign offices. The investigators conducting these inspections are those based in the China and India foreign offices—the countries where most drug inspections occur— and also include those on temporary duty assignment to these offices. According to FDA officials, these investigators are expected to conduct 15 foreign inspections each year. We have noted high vacancy rates for these foreign offices in past reports. While these vacancy rates have decreased over time, vacancies persist. As of November 2019, FDA’s China office had three of 10 drug investigator positions vacant (a 30 percent vacancy rate), while FDA’s India office had two of six drug investigator positions vacant (a 33 percent vacancy rate). FDA has taken steps to address vacancies in the foreign offices, but continues to face challenges. In our 2010 report, we recommended that FDA develop a strategic workforce plan to help recruit and retain foreign office staff. FDA released such a plan in March 2016, but the long- standing vacancies in the foreign offices raise questions about its implementation. FDA officials told us that one challenge in recruiting investigators for the foreign offices is that well-qualified investigators for those positions need foreign inspection experience. For example, an official in FDA’s India office told us that foreign inspections can be challenging and the India office does not have the resources to develop or train new investigators. Therefore, it is important to recruit investigators who have experience conducting foreign inspections, and such investigators are recruited from ORA. Thus, vacancies in the other two groups of investigators can influence the number of staff available to apply for positions in the foreign offices. Further, according to FDA officials, after employees have accepted an in-country position, the agency can experience significant delays before they are staffed in the office due to delays in processing assignments. For example, an official in FDA’s India office said that investigators need to complete a week-long security training program and must obtain the security clearance needed to work at the U.S. Embassy, which is where FDA’s foreign office is located. However, the official told us that there are limited availabilities for that training and background checks for security clearances can take time. According to this official, FDA investigators do not usually receive first priority for the training. FDA estimates that it can take as little as 1 year to over 2 years to clear background and medical checks and arrive at a foreign office. For example, an investigator in FDA’s China office told us that as a result of these requirements and other issues, it took nearly 2 years for the investigator to arrive at the office after FDA had accepted the investigator’s application. According to FDA’s own strategic workforce plan for the foreign offices, these types of delays have resulted in staff changing their decision after accepting a position in the foreign offices. Persistent Challenges Unique To Foreign Inspections Raise Questions About Their Equivalence To Domestic Inspections Our preliminary analysis indicates that FDA continues to face unique challenges when inspecting foreign drug establishments—as compared to domestic establishments—that raise questions about the equivalence of these inspections. Specifically, based on our interviews with drug investigators in the dedicated foreign drug cadre and FDA’s foreign offices in China and India, we identified four challenge areas related to conducting foreign inspections, which are described below. Of the four challenge areas identified, three areas—preannouncing inspections, language barriers, and lack of flexibility—were also raised in our 2008 report. Preannouncing Inspections. As we reported in 2008, the amount of notice FDA generally gives to foreign drug establishments in advance of an inspection is different than for domestic establishments. Domestic drug establishment inspections are almost always unannounced, whereas foreign establishments generally receive advance notice of an FDA inspection. According to FDA officials, FDA is not required to preannounce foreign inspections. However, they said the agency generally does so to avoid wasting agency resources, obtain the establishment’s assistance to make travel arrangements, and ensure the safety of investigators when traveling in country. FDA does conduct some unannounced foreign inspections, particularly if the investigators conducting the inspection are based in FDA’s foreign offices. However, FDA officials told us that FDA does not have data on the frequency with which foreign drug inspections are unannounced, nor the extent to which the amount of notice provided to foreign establishments varies. According to FDA officials, this is because FDA does not have a data field in its database to systematically track this information. However, the officials estimated that the agency generally gives 12 weeks of notice to establishments that investigators are coming when investigators are traveling from the United States. While investigators in FDA’s China and India offices do conduct unannounced or short-notice inspections, these staff do not perform most of the inspections in these countries. (See table 3). Our preliminary work indicates that preannouncing foreign inspections can create challenges and raises questions about the equivalence to domestic inspections. Of the 18 investigators we interviewed, 14 said that there are downsides to preannouncing foreign inspections, particularly that providing advance notice gives foreign establishments the opportunity to fix problems before the investigator arrives. For example, when an inspection is preannounced, it gives establishments time to clean up their facility and update or generate new operating procedures. However, establishments are expected to be in a constant state of compliance and always ready for an FDA inspection, and several investigators told us seeing the true day-to-day operating environment for an establishment is more likely during an unannounced inspection. Of the 18 investigators we interviewed, 12 said that unannounced inspections are generally preferable to preannounced inspections. One investigator told us that, although they believe the best way to ensure industry compliance to CGMPs is for establishments to not know when FDA is coming for an inspection, there is no data that would allow the agency to evaluate whether unannounced inspections are better than preannounced inspections. In addition, some investigators told us that it is still possible to identify serious deficiencies during preannounced inspections. For example, investigators can still identify issues by looking at the firm’s electronic records, including time-stamped data relating to the creation, modification, or deletion of a record. Three investigators also told us that in some cases there can be benefits to announcing inspections in advance. For example, for preapproval inspections, announcing the inspection in advance gives the establishment time to organize the documentation and staff needed to conduct the inspection. Language Barriers. Our preliminary work indicates that language barriers—which we first reported as a challenge to conducting foreign inspections in our 2008 report—can add time to inspections and raise questions about the accuracy of information FDA investigators collect and thus about the equivalence to domestic inspections. FDA generally does not send translators on inspections in foreign countries. Rather, investigators rely on the drug establishment to provide translation services, which can be an English-speaking employee of the establishment being inspected, an external translator hired by the establishment, or an English-speaking consultant hired by the establishment. Of the 18 investigators that we interviewed, 14 said that language barriers can be a challenge to conducting foreign inspections and were especially challenging in parts of Asia, including China and Japan. Seven investigators told us this is less of a challenge for inspections conducted in other foreign countries, including India and countries in Europe, because workers at establishments in these countries are more likely to speak English, and documentation is also more likely to be in English. Investigators told us that compared to domestic inspections, it can be more challenging and take longer to complete typical inspection-related activities, such as reviewing documentation or interviewing employees, if the investigator needs to rely on translation. Fourteen of the 18 investigators we interviewed said that there can be concerns related to relying on establishment staff and independent translators. Specifically, 11 investigators told us there can be uncertainties regarding the accuracy of the information being translated, particularly when investigators rely on the translation provided by an employee of the establishment being inspected. For instance, one investigator said that there is more risk of conflict of interest if the establishment uses its own employees to translate. Another investigator said that they went to a drug establishment in China that told FDA it had English-speaking employees to translate the inspection, but that this was not the case, and the investigator had to use an application on their phone to translate the interviews. In addition, the firm representative providing the translation may be someone that does not have the technical language needed, which can make it harder to communicate with firm staff and facilitate the inspection. One investigator told us that the independent translators hired by firms are sometimes consultants and, in those instances, it can seem like the consultants are coaching the firm during the inspection. FDA officials told us that when they conduct unannounced for-cause inspections in China, investigators bring locally employed staff who work in FDA’s China office to act as translators. The investigators we interviewed said that in such instances, they valued knowing that the translation they were getting was accurate. However, FDA does not have the resources to provide locally employed staff on every inspection, according to an FDA official. We will continue to examine this issue with FDA as part of our ongoing work. Lack of Flexibility. Our preliminary work indicates that, as we first reported in 2008, the overseas travel schedule can present unique challenges for FDA’s domestically based investigators—including both ORA investigators and members of the dedicated foreign dug cadre— who conduct the majority of foreign inspections. Eight of the 12 dedicated foreign drug cadre investigators that we interviewed told us that there is little flexibility to extend foreign inspections conducted by domestically based investigators because the inspections they conduct on an overseas trip are scheduled back-to-back in 3-week trips that may involve three different countries. This raises questions about their equivalence to domestic inspections. For instance, extending one inspection would limit the amount of time the investigator has to complete their other scheduled inspections, some investigators told us. In addition, eight investigators told us that domestically based staff are generally unable to extend the total amount of time spent on an overseas trip—one investigator told us that an investigator would have to find something really bad to justify an extension. In contrast, FDA officials told us that inspections conducted by in-country investigators in China or India, and domestic inspections in the United States, are generally scheduled one at a time and can thus more easily be extended if the investigator needs additional time to pursue potential deficiencies. However, in-country investigators are not involved in the majority of inspections conducted in China or India. Three investigators from the dedicated foreign drug cadre told us that when they travel overseas, they adjust their inspection approach to help ensure they finish foreign inspections on time. For example, one investigator told us an investigator may start the inspection in an area of the establishment that was noted as having issues during the last inspection. However, one investigator said that sometimes it is not possible to cover everything in depth during a foreign inspection. Another investigator told us that they focus on identifying the most serious issues during a foreign inspection, and that less serious issues can be identified in the establishment inspection report for reference in the next inspection. Five investigators also noted that they work long hours during their inspection to ensure they can complete the needed work. While FDA may assign more than one investigator to an inspection to complete needed work, one investigator said that FDA does not usually assign more than one person to an inspection because investigators are expected to have the experience to conduct inspections by themselves. From fiscal years 2012 to 2018, the majority of both foreign and domestic inspections were conducted by one person—77 percent and 66 percent, respectively. Post-Inspection Classification Process. According to FDA officials, starting in fiscal year 2018, FDA implemented a new post-inspection classification process: when an ORA investigator recommends an OAI classification following an inspection, ORA compliance is required to send that inspection report to CDER for review within 45 calendar days from the inspection closeout. Among other things, the process was intended to help ensure FDA can communicate inspection results to domestic and foreign establishments within 90 days of the inspection closeout, as committed to under the Generic Drug User Fee Amendments of 2017(GDUFA II). FDA officials told us that the changes also required an additional ORA review for foreign inspection reports to align that process with the process for domestic inspection reports. Although the 45-day reporting time frame for potential OAI classifications is a requirement for both domestic and foreign inspections, adding the additional level of review within ORA effectively shortened the amount of time investigators have to document findings for foreign inspections. Our preliminary work indicates that the post-inspection reporting time frames can create challenges for domestic investigators that conduct foreign inspections and raise questions about the equivalence to domestic inspections. Eight of the 18 investigators that we interviewed said shortening the time for completing reports and adding a level of review has made it more challenging to meet reporting requirements, especially if serious deficiencies are identified during the inspection. Investigators told us that for a potential OAI inspection, they now need to send the inspection report to their supervisor for endorsement within 10 days of the closeout of a foreign inspection, regardless of when the investigator’s next inspection is scheduled for, or whether the investigator has to travel from overseas back to the United States after the inspection. For example, if a domestic investigator finds serious deficiencies on the first inspection, thus indicating an initial OAI classification, the investigator needs to write and send the related inspection report to the ORA supervisor for endorsement before returning home from the 3-week overseas trip to meet the required time frame. One investigator told us that, as a result of the time pressures, post-inspection reports may be less thorough, and that some inspection observations could be better supported if investigators had more time to write the reports. In conclusion, foreign manufacturing establishments continue to be a critical source of drugs for millions of Americans, and FDA inspections are a key tool to ensure the quality of these drugs. Over the years since we first examined this issue, FDA has made significant changes to adapt to the globalization of the drug supply chain and has greatly increased the number of inspections that it conducts of foreign establishments. Notably, it has markedly increased the percentage of foreign inspections conducted to monitor drugs already on the market, which we previously noted were vital to FDA oversight of foreign establishments. However, the agency continues to be faced with many of the same challenges in the oversight of foreign establishments that we identified in our 2008 report. Our preliminary work has identified inspection decreases, related in part to FDA challenges filling investigator vacancies. We have also identified a variety of unique challenges that investigators face in foreign inspections. As we continue to conduct our work, we will further examine the cumulative effect of these challenges that raise questions about FDA’s ability to conduct equivalent inspections in foreign establishments. We will examine the extent to which FDA has assessed its oversight of drugs manufactured overseas and the steps it is taking to mitigate any risks, and make recommendations as appropriate. Chair DeGette, Ranking Member Guthrie, and Members of the Subcommittee, this completes my prepared statement. I would be pleased to respond to any questions that you may have at this time. GAO Contact and Staff Acknowledgments If you or your staff have any questions about this testimony, please contact Mary Denigan-Macauley, Director, Health Care at (202) 512-7114 or DeniganMacauleyM@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. GAO staff who made key contributions to this testimony are William Hadley (Assistant Director); John Lalomio (Analyst-in- Charge); Katherine L. Amoroso; George Bogart; Zhi Boon; Derry Henrick; Laurie Pachter; and Vikki Porter. 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.

Why GAO Did This Study More than 60 percent of establishments manufacturing drugs for the U.S. market were located overseas in fiscal year 2018. FDA has estimated that about 40 percent of finished drugs and 80 percent of active drug ingredients are manufactured overseas. FDA is responsible for overseeing the safety and effectiveness of all drugs marketed in the United States, regardless of where they are produced and conducts inspections of both foreign and domestic drug manufacturing establishments. GAO has had long-standing concerns about FDA's ability to oversee the increasingly global supply chain, an issue highlighted in GAO's High Risk Series for the last 10 years. GAO recommended in 2008 (GAO-08-970) that FDA increase the number of inspections of foreign drug establishments. GAO found in 2010 (GAO-10-961) and 2016 (GAO-17-143) that FDA was conducting more of these foreign drug inspections, but GAO also reported that FDA may have never inspected many establishments manufacturing drugs for the U.S. market. This statement is based on ongoing work and provides preliminary GAO observations on 1) the number of foreign inspections FDA has conducted, 2) inspection staffing levels, and 3) challenges unique to foreign inspections. For this work, GAO examined FDA data, visited FDA foreign offices in China and India, and interviewed drug investigators based in these offices and in the United States. What GAO Found GAO's preliminary analysis of Food and Drug Administration (FDA) data shows that from fiscal year 2012 through 2016, the number of foreign drug manufacturing establishment inspections increased. From fiscal year 2016 through 2018, both foreign and domestic inspections decreased—by about 10 percent and 13 percent, respectively. Howevever, the total number of foreign inspections surpassed the number of domestic inspections in 2015, and a growing percentage of FDA's foreign inspections (43 percent in 2018) were conducted in China and India, where most establishments that ship drugs to the United States were located. FDA officials attributed the decline, in part, to vacancies among investigators available to conduct inspections. GAO previously noted the vital role that inspections play in FDA's oversight of foreign establishments. FDA has vacancies among each of the groups of investigators who conduct foreign inspections. FDA had 190 investigators in the United States who conduct the majority of foreign inspections, but an additional 58 positions were vacant. FDA was in the process of filling 26 of these vacancies, with 32 remaining. However, according to FDA officials, it could be 2 to 3 years before new staff are experienced enough to conduct foreign inspections. FDA also faces persistent vacancies among investigators in its foreign offices. FDA investigators identified persistent challenges conducting foreign inspections, raising questions about the equivalence of foreign to domestic inspections. For example, while domestic inspections are almost always unannounced, FDA's practice of preannouncing foreign inspections up to 12 weeks in advance may give manufacturers the opportunity to fix problems. Investigators from FDA's China and India offices do conduct some unannounced inspections, but they are involved in a small percentage of inspections in these countries (27 percent and 10 percent, respectively). Further, FDA continues to rely on translators provided by the foreign establishments being inspected, which investigators said can raise questions about the accuracy of information FDA investigators collect. What GAO Recommends GAO will continue to assess these issues as part of ongoing work, and make recommendations as appropriate.

gao_GAO-19-495

gao_GAO-19-495_0

Background IRA owners are not permitted to engage in certain prohibited transactions involving IRA assets. Prohibited transactions generally fall into two categories: Transaction involving disqualified persons. An IRA is prohibited from engaging in a transaction with disqualified persons, such as members of the IRA owner’s family or an IRA fiduciary. Transaction involving self-dealing. An IRA owner who is a fiduciary is prohibited from engaging in a transaction with the IRA where the IRA owner personally benefits (other than through the receipt of a distribution). We previously reported that prohibited transactions are more likely to arise when IRA owners make unconventional IRA investments. Unlike conventional IRA investments in publicly traded stocks, bonds, and mutual funds, unconventional investments in real estate, virtual currency, or private equity are more likely to involve the IRA owner, disqualified family members, or other disqualified persons. For example, an IRA invested in rental real estate can leave IRA owners susceptible to a number of prohibited transactions, such as renting to family or paying for repairs with personal funds. IRA owners may face adverse and potentially severe tax consequences if they are found to have engaged in a prohibited transaction. Specifically, the IRA could lose its tax-favored status. The account would then be treated as distributing all its assets to the IRA owner at the fair market value on the first day of the year in which the prohibited transaction occurred. The IRA owner may be subject to additional income taxes because of any early distribution from an IRA. The prohibited transaction may also be subject to excise taxes. The Employee Retirement Income Security Act of 1974 (ERISA), which established IRAs and rules prohibiting certain IRA transactions, assigned IRA oversight roles to both DOL and IRS. To avoid confusion over dual jurisdiction, a 1978 Executive Order further clarified the agencies’ roles and responsibilities regarding prohibited transactions. As a result, the authority to interpret the prohibited transaction rules and grant exemptions to those rules was transferred to DOL. The transfer did not affect IRS’ ability to enforce the excise tax provisions or the tax consequences for IRA owners who are found to have engaged in a prohibited transaction. However, in enforcing such tax consequences, IRS is bound by the regulations, rulings, opinions, and exemptions issued by DOL. DOL has the authority to grant administrative exemptions to the prohibited transaction rules on either an individual or a class basis. DOL can grant prospective exemptions for a transaction that an IRA is considering, as well as retroactive exemptions for transactions that have already occurred. DOL Has Not Sufficiently Documented Internal Policies and Procedures for Reviewing Prohibited IRA Transaction Exemption Applications To grant an exemption from prohibited IRA transaction rules, DOL evaluates applications using statutory criteria, and follows administrative procedures codified in regulations. Generally, DOL may not grant an exemption unless it finds the exemption to be: in the interest of the plan and its participants and beneficiaries, and protective of the rights of plan participants and beneficiaries. Before granting an exemption, DOL generally must publish a notice of proposed exemption in the Federal Register inviting interested parties to comment on the proposed exemption. DOL Has a Process to Grant Administrative Exemptions for Otherwise Prohibited IRA Transactions DOL regulations lay out the process for filing and processing prohibited transaction exemptions applications. Among other things, the regulations explain: who may apply, what information must be included with an application, when a conference with DOL can be requested, when a request for reconsideration of a DOL decision can be made, how DOL and the applicant will notify interested persons if DOL decides a tentative approval is warranted. DOL also publishes a booklet that provides an explanation of the regulations and applicable laws, and includes additional information for applicants like examples of common types of exemption requests. IRA owners or their fiduciaries file applications for exemptions with DOL’s Office of Exemption Determinations which is part of EBSA. Applicants can research information about past exemptions granted by the agency on EBSA’s website. As explained in the DOL booklet describing the application requirements, applicants have the burden of demonstrating that they should be granted an exemption. If DOL tentatively denies an application, applicants have options for requesting that the denial be reconsidered. Within 20 days of the tentative denial, applicants can request a conference with DOL, or notify DOL of their intent to submit additional information. If, after a conference has been convened, DOL issues a final denial of the application, DOL will entertain one request for reconsideration if the applicant presents significant new facts or arguments, which, for good reason, could not have been submitted earlier. After DOL publishes a notice of proposed exemption in the Federal Register that describes the pending application, the applicant must notify interested persons of the pending exemption. Often, the contents of the information sent to all interested persons, the manner in which it is sent, and any associated deadlines will have previously been agreed to by DOL and the applicant. DOL may also hold public hearings during the comment period. For example, if the transaction involves potential fiduciary self-dealing or conflicts of interest, any individual potentially adversely affected by the exemption may submit a request for a public hearing to DOL. If granted, DOL publishes information about the exemption in the Federal Register and on its website. Figure 1 provides an overview of the exemption application process. The regulations describe circumstances in which DOL will ordinarily not consider an application. For example, DOL generally will not consider an individual application if DOL already has under consideration a class exemption relating to the same type of transaction. DOL will also not consider an application for transactions subject to DOL or IRS investigations. DOL requires applicants to disclose in their applications whether exemption transactions are, or have been, subject to an investigation or enforcement action by DOL or IRS. In addition, if the applicant or any other party in interest becomes the subject of an investigation or enforcement action, the applicant is required to promptly notify DOL. If applicants find that their prospective transaction is substantially similar to other transactions for which the agency has previously granted exemptions, they can follow an expedited process by submitting an “EXPRO” application. EXPRO applications are required to cite prior exemptions granted by DOL to demonstrate that the proposed IRA transaction is substantially similar to other IRA transactions for which DOL has previously provided an exemption. Specifically, EXPRO applicants must cite as substantially similar, either (1) two individual exemptions granted by DOL within the previous 5 years, or (2) one individual exemption granted within the past 10 years, and one transaction authorized pursuant to the EXPRO class exemption within the past 5 years. The applicant must give notice to all interested persons, and the applicant must resolve all substantive adverse comments provided by interested persons before DOL will grant final approval. The time to complete the exemption process can range from a few months to more than a year. DOL officials told us that the process generally takes about 1 year for an individual IRA application that is relatively simple or routine. EXPRO applications have been processed in as few as 78 days. According to DOL officials, the process can start before an applicant submits a formal application because applicants can, and do, request informal consultations and conferences with DOL. DOL officials explained that sometimes potential applicants decide not to file an application after an informal conference because applicants realize that their application would likely be denied. DOL officials explained that during the review process, they first confirm their understanding and characterization of the proposed exemption through correspondence with the applicants. Then, in response, DOL often sets conditions under which relief from the prohibited transaction rules is contingent, such as on the applicant taking additional actions and remaining in compliance with those conditions. For example, if an applicant wants to sell or purchase an asset in what would be an otherwise prohibited IRA transaction, DOL may stipulate that the applicant first obtain an independent appraisal or valuation assessment to determine a fair-market value of that asset. After applications are formally submitted, many IRA applicants withdraw during DOL’s review process. Over an 11-year period, we found that of the 124 IRA applications, applicants withdrew roughly half (56) before the review process was completed (see table 1). Of the remaining 68 applications that continued with the review process, DOL granted 48, denied 16, and closed four application cases for administrative or other reasons. DOL officials did not dispute the results of our analysis, but they said that it would be misleading to conclude that DOL is more likely to grant than deny applications. Rather, they said that their practice of encouraging applicants to consult with DOL in advance leads some potential applicants to decide not to pursue an exemption. In our review of processed applications, we found that most of the applications involved the sale of IRA assets. We found that 88 of the 124 applications were for transactions involving the sale of IRA assets. Most of these were sales of securities or real property (see appendix I for additional information). The next most common type of transaction was for the purchase of assets (21 applications), and most of those also involved securities or real property. The remaining applications involved other transactions, including leases, loans, and extensions of credit. DOL Lacks Documented Policies and Procedures to Manage Its Reviews and Data DOL has not sufficiently documented internal policies and procedures to manage and help ensure effective internal controls of its prohibited transactions exemption process. While DOL regulations and guidance detail the requirements for applicants, DOL generally lacks internal documentation of the steps and actions DOL officials are to follow when processing applications, and the roles and responsibilities of agency officials. DOL officials told us that they use a case tracking system to record and track applications. When an application is received by DOL, the division chief of EBSA’s Office of Exemption Determinations (OED) reviews the application and assigns it to an OED supervisor. Either the division chief or the supervisor enters preliminary information from the application into the system, and classifies the transaction by applying one or multiple subject matter codes. The supervisor then reviews the information in the applicant’s case file and assigns the case to an OED analyst. DOL officials told us that any interim data, such as the publication date for a proposed exemption, is entered by the supervisor in the system. If an application is withdrawn by an applicant, denied, or granted, the supervisor records this information in the system, including the dates of these actions. When a case is closed, the analyst completes a close-out index form and submits it to the supervisor for review, and the supervisor enters a closing code in the system. DOL officials told us that they can use the system to generate management reports, such as on the number of applications filed and the amount of time to process cases. Neither the process described above, nor the different roles and responsibilities of the OED division chief, supervisors, and analysts in that process, were documented in the internal documents that DOL provided. A system reference guide included instructions to system users for how to input and modify case records, generate reports, and add or modify users. The reference guide also included screen prints indicating which fields are required by the system to process a case. However, the reference guide did not contain information about responsibilities and duties for these data entry activities, and how those duties are assigned. The documentation provided is unclear regarding who within OED is ultimately responsible for making final decisions on applications. According to Standards for Internal Control in the Federal Government, documentation of an agency’s policies and procedures is a necessary part of an effective internal control system. Such documentation can appear, for example, in management directives or operating manuals, and it should be readily available for examination. Policies and procedures can also help document internal control responsibilities within the agency. DOL officials told us that OED is a small and compact organization, and as such, its policies and procedures can easily be communicated “person to person” and through onsite training. DOL officials also said that the process for entering data is not difficult, and there are few opportunities for error because nearly all data on applications is prepopulated. The principles of internal control, however, apply to both large and small organizations. The level and nature of documentation may vary based on the size of the organization and the complexity of the processes the organization performs, but documentation is still necessary. By documenting policies and procedures, management will be better positioned to monitor whether the organization’s activities are aligned with those policies and procedures, and assess whether the organization is achieving its objectives. Documenting procedures also would provide greater transparency about how applications are handled, and can reduce the risk of employees carrying out their duties inconsistently. For a small organization like OED, documentation of policies and procedures provides a means to retain organizational knowledge, and can help ensure continuity of and consistency in operations if key personnel leave the organization unexpectedly. DOL and IRS Currently Share Some Information on Exemption Applications, but More Formalized Collaboration Could Improve Their Oversight Efforts Some information sharing takes place between DOL and IRS on applications for IRA prohibited transaction exemptions, but no formal mechanism exists to help guide collaboration between the two agencies. As previously discussed, DOL and IRS share oversight responsibility for prohibited IRA transactions. Based on our review of applications and DOL data as well as interviews with agency officials, we found that interactions between DOL and IRS regarding applications for prohibited transaction exemptions are infrequent and limited in scope. Of the 124 applications we reviewed, only eight were coded in OED’s Case Tracking System as having “external contact with IRS,” and DOL officials confirmed that this accurately reflects the level of interagency coordination. DOL officials stated that they sometimes contact IRS about exemption applications, and IRS officials confirmed to us that they periodically receive communications from DOL. IRS officials also told us that they occasionally contact DOL. Both agencies described to us how their current interaction occurs. DOL officials told us that they coordinate with IRS in the following ways: If, during the application review process, OED staff identify applications that may warrant further review or investigation for tax violations, they refer the case to EBSA’s Office of Enforcement, which may then coordinate or refer the case to IRS. DOL officials said that OED staff review the IRS “Dirty Dozen” list of potentially abusive tax scams and schemes. IRS officials said that when possible prohibited transactions arise during an examination that might require DOL input, IRS examiners reach out to DOL to ensure that IRS understands DOL decisions on those transactions. DOL officials said that, in their view, most requested prohibited IRA transaction exemptions do not require extensive interaction with IRS. They questioned the potential usefulness of information about denied or withdrawn applications that might be shared with IRS, but said that IRS could certainly obtain this information if IRS requested it. IRS officials, however, told us that more information from DOL about prohibited IRA transactions and requested exemptions could be useful in carrying out IRS’s oversight responsibilities. For example, DOL does not share information on denied or withdrawn applications with IRS, information that IRS officials told us would be helpful to them. We found that denial information could be useful to IRS as illustrative examples of prohibited transactions for examiner training and educational outreach to IRA owners. Information about the types of transactions in withdrawn applications could also help IRS identify emerging issues or trends in potential prohibited transactions marketed to IRA owners. Although some limited collaboration between DOL and IRS exists, the agencies have not applied to their oversight of prohibited transactions some key practices we have identified in prior reviews of interagency collaboration. Specifically, developing a mechanism to formalize the sharing of information between DOL and IRS could help support current collaboration activities, and could be useful in helping the agencies identify opportunities for greater collaboration going forward. Furthermore, documentation is a necessary part of an effective internal control system. Documenting the procedures for interagency collaboration would improve internal control over the agencies’ activities. A formal agreement, such as a memorandum of understanding (MOU) or other mechanism, can further help agencies monitor, evaluate, and update interagency collaboration. For example, DOL and IRS have previously formalized their collaboration regarding oversight of a different type of retirement savings vehicle— employer-sponsored retirement plans. DOL and IRS have oversight responsibilities for employer-sponsored retirement plans, such as pensions, and in 2003, DOL and IRS completed an MOU to implement collaboration between the two agencies with regards to investigations of and litigation involving employer-sponsored retirement plans. The employer retirement plan MOU and the implementing guidance contain some features of interagency collaboration mechanisms that we have identified in prior work. For example: The responsibilities of each agency are documented, and responsible agency components and officials are identified. The agencies use collaboration tools (checklists) for determining whether issues presented in an examination or investigation by one agency should be referred to the other. A system and process exists to track referrals, and the agencies reconcile their data about referrals (including pending referrals) quarterly. The employer retirement plan MOU also established a process to periodically monitor its effectiveness, and the MOU was last updated in 2013. Developing a similar mechanism to formalize the sharing of information between DOL and IRS regarding IRA prohibited transaction exemptions could help the agencies better support their current coordination efforts and identify additional opportunities for greater collaboration. Conclusions IRAs are a key vehicle for individuals to save for retirement. IRA owners’ decisions to invest in unconventional assets can expand their role and responsibilities substantially. The consequences for account owners who make a mistake can be severe. When IRA owners request an exemption from rules on prohibited transactions, DOL evaluates applications using statutory criteria, and follows administrative procedures codified in regulations. However, DOL has not sufficiently documented internal policies and procedures for how to manage its process for granting exemptions. Such documentation is a necessary part of an agency’s effective internal control system. DOL and IRS share oversight responsibility of prohibited IRA transactions. While the two agencies do share some information, they do not have a formal mechanism to guide and monitor their collaboration. By formalizing interagency collaboration, such as through an MOU or other mechanism, DOL and IRS could help reinforce their current information sharing and potentially identify new opportunities to improve their oversight efforts through greater collaboration. Documenting procedures for DOL and IRS collaboration on prohibited IRA transactions would also help introduce better internal control over these activities. Recommendations for Executive Action We are making a total of three recommendations, including two to DOL and one to IRS. The Secretary of Labor should document internal policies and procedures for managing the IRA prohibited transaction exemption process. (Recommendation 1) The Secretary of Labor, in consultation with the Commissioner of Internal Revenue, should establish a formal means, such as a memorandum of understanding or other mechanism, to support and guide DOL’s and IRS’s collaborative efforts to oversee IRA prohibited transaction exemptions. (Recommendation 2) The Commissioner of Internal Revenue, in consultation with the Secretary of Labor, should establish a formal means, such as a memorandum of understanding or other mechanism, to support and guide DOL’s and IRS’s collaborative efforts to oversee IRA prohibited transaction exemptions. (Recommendation 3) Agency Comments We provided a draft of this report to the Secretary of Labor, the Commissioner of Internal Revenue, and the Secretary of the Treasury for review and comment. In its comments, reproduced in appendix II, DOL generally agreed with our two recommendations directed to it. For recommendation 1, DOL plans to create an internal procedure manual formalizing OED’s administrative case processing procedures to help in passing along institutional knowledge. For recommendation 2, DOL agreed to periodically discuss all IRA exemption cases with IRS and did not elaborate on the formal means for this information sharing. DOL also provided technical comments which we incorporated as appropriate. In its comments, reproduced in appendix III, IRS generally agreed with our recommendation directed to it. For recommendation 3, IRS said it is committed to discussing an appropriate mechanism, including periodic meetings, to formalize collaboration on IRA prohibited transaction exemptions. IRS plans to consider expanding its formal collaboration with DOL as part of the next periodic update of the existing employer plan MOU. IRS also provided technical comments which we incorporated as appropriate. The Department of the Treasury provided technical comments which we incorporated as appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees, the Secretary of Labor, the Secretary of the Treasury, and the Commissioner of Internal Revenue. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact James R. McTigue, Jr. at (202) 512-9110 or Charles A. Jeszeck at (202) 512-7215. You may also reach us by email at mctiguej@gao.gov or jeszeckc@gao.gov. GAO staff making key contributions to this report are listed in appendix IV. Appendix I: Applications for Individual Retirement Account Exempted Transactions by Type Appendix II: Comments from the Department of Labor Appendix III: Comments from the Internal Revenue Service Appendix IV: GAO Contact and Staff Acknowledgments GAO Contacts Staff Acknowledgments In addition to the contacts named above, MaryLynn Sergent and David Lehrer (Assistant Directors), Ted Burik, Susan Chin, Steven Flint, Emily Gruenwald, Mark Kehoe, Jungjin Park, and David Reed made key contributions to this report. James Bennett, Amy Bowser, Jacqueline Chapin, Edward J. Nannenhorn, Andrew J. Stephens, Walter Vance, and Adam Wendel also provided support.

Why GAO Did This Study IRA owners are able to invest in a wide variety of assets, but they are prohibited from engaging in certain transactions involving IRA assets. IRA owners who engage in prohibited transactions may incur increased income tax liability, additional taxes, and the loss of the tax-advantaged status of their accounts. DOL can grant exemptions from the prohibited transaction rules. IRS enforces tax laws relating to IRAs and can assess additional taxes. GAO was asked to examine (1) DOL's process for granting exemptions for prohibited IRA transactions and outcomes of that process, and (2) the extent to which DOL and IRS collaborate on oversight of prohibited transaction rules for IRAs. GAO reviewed relevant federal laws and regulations; examined agency guidance, exemption process documentation, and application case files; assessed interagency coordination using internal control standards and prior work on interagency collaboration; and interviewed DOL and IRS officials. What GAO Found The Department of Labor (DOL) has a process to grant administrative exemptions for individual retirement account (IRA) transactions that would otherwise be prohibited by law, such as an IRA buying investment property from the IRA owner. DOL evaluates applications using statutory criteria and follows administrative procedures codified in regulations. Applications for proposed transactions that are substantially similar to certain other transactions previously granted exemptions may follow an expedited process. As shown in the figure, GAO found that roughly half (56) of the IRA prohibited transaction exemption applications it reviewed were withdrawn by the applicant before the review process was completed. In reviewing processed applications, GAO found that most of the prohibited transactions for which an exemption was sought involved the sale of IRA assets. With regard to DOL's application review process, GAO found that DOL has not sufficiently documented internal policies and procedures to help ensure effective internal control of its process. Documenting procedures could increase transparency about how applications are handled, reduce the risk of DOL employees carrying out their duties inconsistently, and provide a means to retain organizational knowledge should key personnel leave unexpectedly. Although DOL and the Internal Revenue Service (IRS) share some information as part of their oversight responsibility for prohibited IRA transactions, no formal mechanism exists to help guide collaboration between the agencies. Of the 124 IRA applications GAO reviewed, only eight reflected DOL contact with IRS. GAO found that DOL has information about requested exemptions to prohibited IRA transaction rules that could be useful to IRS in carrying out its oversight responsibilities. For example, DOL does not share information on denials—information that could be useful as prohibited transaction examples for IRS examiner training and educational outreach to IRA owners. In prior work on interagency collaboration, GAO has found that formal agreements, such as a memorandum of understanding, can help agencies monitor, evaluate, and update interagency collaboration. Formalizing the sharing of information between DOL and IRS regarding IRA prohibited transaction exemptions could help the agencies better support their current coordination efforts and identify additional opportunities for greater collaboration. What GAO Recommends GAO is recommending that DOL and IRS establish a formal means—such as a memorandum of understanding or other mechanism—to collaborate on oversight of prohibited IRA transaction exemptions. GAO is also recommending that DOL document policies and procedures for managing the exemptions process. DOL and IRS generally agreed with GAO's recommendations.

gao_GAO-20-176

gao_GAO-20-176_0

Background Lebanon is a small, religiously diverse country bordering the Mediterranean Sea (see fig. 1). Religious tensions among Lebanon’s Maronite Christians, Sunni Muslims, Shiite Muslims, and others, have for many years contributed to conflicts within Lebanon as well as with neighboring countries. According to State, Lebanon’s political system is characterized by sectarian divisions and pressures from external and internal forces that limit its ability to function. Upon gaining independence from France in 1943, Lebanese leaders adopted a power-sharing agreement, in which each of the country’s officially recognized religious groups were to be represented in the government according to their share of the population based on the 1932 census. This unwritten agreement established a status quo in which the president must be a Maronite Christian (the largest single denomination in 1932), the prime minister a Sunni Muslim, and the speaker of parliament a Shia Muslim. Tensions over the balance of power among these groups have provoked conflict. During the Lebanese Civil War from 1975 to 1990, both Syrian and Israeli forces occupied the country. In the midst of the civil war and Israel’s occupation of southern Lebanon, Hizballah emerged in Lebanon as a powerful Islamic militant group. In 2000, Israeli forces withdrew from southern Lebanon. In 2005, owing to pressure from the international community, Syrian forces withdrew from Lebanon following the assassination of Lebanon’s prime minister. Parliamentary elections in that year led to a member of Hizballah holding a cabinet position for the first time, and at least one member of Hizballah has held a cabinet position ever since. Instability arising from the Syrian civil war that began in 2011 has also exacerbated sectarian conflict and created new challenges within Lebanon. In particular, that war has caused an influx of over 1.3 million Syrian refugees into Lebanon, a country with a population of only 4.5 million. The Syrian civil war has also increased the risk of terrorist incidents in Lebanon, as foreign terrorist fighters have crossed Lebanese borders going to and from the conflict. U.S. Security-Related Objectives and Associated Funding Since 2013, the United States’ primary goal in providing security assistance to Lebanon has been to strengthen its state institutions to allow them to exert sovereign authority and enhance security. Since at least 2015, the primary objectives supporting this goal have focused on 1) building the capacity of Lebanese security forces to exert sovereign authority over Lebanese territory, including at the border and by maintaining internal security; and 2) enhancing the capacity of Lebanese security forces to respond to terrorist and criminal threats. Through both these objectives, the U.S. also seeks to delegitimize and marginalize Hizballah by helping to support legitimate state institutions. To achieve these objectives, a number of agencies and offices within State and DOD provide support to the LAF, which is generally responsible for providing border security, counterterrorism, and national defense, and to the ISF, or national police force, which is generally responsible for maintaining law and order within Lebanon. See table 1. U.S. support for Lebanese forces has included a variety of assistance, including training, equipment, and sustainment, as shown in figure 2. State and DOD reported that they obligated nearly $1.5 billion in security assistance funding for Lebanon in fiscal years 2013 through 2018. The largest security assistance programs were State’s Foreign Military Financing program, which provides grants and loans to foreign governments for the acquisition of U.S. defense equipment, services, and training, and DOD’s Global Train and Equip program, which funds training and equipment for foreign military forces to conduct counterterrorism operations and enhance maritime and border security. These two programs collectively accounted for nearly 80 percent of assistance. State provided about 56 percent of the overall funding and DOD contributed 44 percent, as shown in figure 3. End-Use Monitoring for Security Assistance DOD and State are required to conduct end-use monitoring (EUM) for some of the equipment provided to Lebanon. In 1996, Congress amended the Arms Export Control Act to require the President to establish a program for monitoring the end-use of defense articles and defense services sold, leased, or exported under the act, including through Foreign Military Sales, Foreign Military Financing, or the Foreign Assistance Act of 1961. The law requires that the program be designed to provide reasonable assurances that recipients are complying with restrictions imposed by the U.S. government on the use, transfer, and security of defense articles and defense services, and that such articles and services are being used for the purposes for which they are provided. DOD’s Defense Security Cooperation Agency (DSCA) is responsible for EUM for Foreign Military Sales. The President is also required to take all reasonable steps to ensure that equipment made available to foreign countries for international narcotics control under the Foreign Assistance Act are used only in ways consistent with the purposes for which such equipment was made available. State’s INL implements this requirement through its End-Use Monitoring Program. DSCA administers the Golden Sentry program, which DOD uses to comply with requirements related to the end-use of defense articles and services transferred to foreign governments. DOD officials at the Office of Defense Cooperation-Beirut conduct the EUM activities established and overseen by DSCA. DSCA’s policy manual for EUM, the Security Assistance Management Manual, and the associated standard operating procedures for Beirut require DOD officials to, among other things, conduct two levels of monitoring: routine EUM and enhanced EUM. Routine EUM: DOD conducts routine EUM for defense articles and services that do not have any unique conditions associated with their transfer. In conducting routine EUM, DOD personnel are required to observe and report any potential misuse or unapproved transfer of U.S.-origin defense articles. Routine EUM is to be conducted in conjunction with other required security-related duties, using any readily available information. For example, U.S. officials might observe how a host country’s military uses U.S. equipment when they visit a military installation on other business. DOD policy states that routine EUM must be documented at least quarterly. DOD policy does not require inventories and physical security checks as part of routine EUM. Enhanced EUM: DOD conducts enhanced EUM for defense services, technologies, or articles specifically identified as sensitive. Lebanon has five types of sensitive defense articles that require enhanced EUM—night vision devices, sniper rifles, light attack aircraft, unmanned aerial vehicles, and Hellfire missiles. DOD policy requires serial number inventories for defense articles needing enhanced EUM within 90 days of delivery of the articles and thereafter within one year of the last inventory performed. In addition, the purchase agreements authorizing the sale of an item may contain specialized notes directing the purchaser to adhere to certain physical security and accountability requirements. In addition to enhanced and routine EUM, DSCA is required to conduct periodic Compliance Assessment Visits to evaluate the Office of Defense Cooperation in Beirut’s compliance with DOD’s EUM policy and the Lebanese government’s compliance with physical security and accountability requirements. U.S. Agencies Have Reported Progress and Challenges in Meeting Security Objectives in Lebanon, but Performance Information Gaps Limit Monitoring of Activities U.S. Agencies Reported Both Progress and Challenges in Meeting Security Objectives in Lebanon According to State and DOD assessments, reports, and interviews with State and DOD officials, the LAF’s border security and counterterrorism capabilities have demonstrated some notable improvements from 2013 to 2018. For example, a 2013 DOD assessment noted that the Lebanese government lacked effective control over its sovereign territory and indicated the LAF leadership was reluctant to engage aggressively in counterterrorism operations. By 2018, however, U.S. agencies reported that, following the expansion of LAF Land Border Regiments, Lebanon had established control of a large part of its borders for the first time in its history. In addition, U.S. agencies reported that the LAF had enhanced its capacity in counterterrorism and counter-narcoterrorism, resulting in more operations. In 2017, for example, the LAF undertook a successful operation to expel ISIS elements along the border with Syria, making Lebanon, DOD officials noted, the only country in the region to successfully expel ISIS from its territory without the involvement of U.S. ground forces. Similarly, State reported improvements in the ISF’s capabilities. For example, INL reported that its ISF training program has become increasingly specialized because of the force’s improved capabilities. According to State reporting, from 2008 through 2012, INL focused its training for the ISF on basic skills. As the ISF became more capable, however, INL reported that the ISF assumed responsibility for all basic training, allowing INL to focus its resources on providing specialized courses. Some examples of these specialized courses include advanced technical radio training and advanced interview and interrogation training. INL also reported that providing equipment and facilities to the ISF helped further to enhance ISF capabilities. For instance, INL reported that the ISF uses the academy INL constructed for it in 2015 in Aramoun for advanced forensics training. In addition, the ISF improved its overall investigative capacity and counterterrorism capabilities since 2013, as shown in a 2017 assessment of State’s Antiterrorism Assistance program. U.S. officials stated that the quality of the working relationship between the U.S. and Lebanon is an important component of success, and Lebanese officials said that U.S. assistance is critical to achieving their mission. U.S. officials noted that the LAF and ISF have been some of the most committed U.S. partners in the region. The LAF and ISF officials we met with also said that U.S. assistance enhances their capabilities and allows them to do their jobs more effectively. One ISF unit, for example, stated that buses purchased with U.S. assistance allow it to transport large numbers of personnel to mission locations. In addition, one LAF unit noted that U.S.-provided armored personnel carriers form the backbone of the LAF’s armored brigades. Despite reported progress, U.S. agencies indicated that some challenges remain for the ISF and the LAF. While the ISF’s capabilities have improved since 2013, U.S. officials said it continues to be more capable in and around Beirut than in other parts of the country. As a result, the LAF often provides internal security to supplement the ISF outside of the capital. Additionally, the ISF needs to improve its internal coordination of cybercrime cases and analyses of digital evidence, according to a 2017 assessment of State’s Antiterrorism Assistance program. For example, the ISF units handling digital investigations and processing, the assessment noted, were fractured and divided, resulting in overwhelming workloads for some units and underutilization of others. DOD assessments also noted that the LAF continues to have some capability gaps, including an ongoing need for equipment and challenges with operating and maintaining U.S.-provided equipment. For example, LAF personnel have expressed concerns about the complexity and sustainability of some U.S. systems, such as the M2 Bradley Fighting Vehicles and A-29 light attack aircraft. Additionally, while U.S. officials stressed they have no desire for direct confrontation between Lebanese security forces and Hizballah, U.S. agencies report that Hizballah’s presence within Lebanon remains a challenge for both the ISF and LAF. In 2018, for example, State reported that Hizballah was the most capable terrorist organization in Lebanon and that it continued to exert control over some areas of the country. Embassy Beirut Has Taken Steps to Review Performance, but Information Gaps Limit its Ability to Monitor Security- Related Activities In addition to periodically assessing long-term performance, State’s Foreign Affairs Manual and internal guidance outline a number of good practices for ICS management. First, the Foreign Affairs Manual says all missions, such as Embassy Beirut, should have an ICS with a hierarchy of goals, objectives, sub-objectives, and, as needed, key activities. Second, missions must assess progress against ICS strategic objectives at least annually. Third, State internal guidance says it is a good practice for missions to establish ICS performance indicators with targets to show the expected change over the course of each period of performance. Fourth, it is also a good practice for missions to practice regular, ongoing data collection against key performance indicators to gauge the direct and near-term effects of activities. The 2018 ICS for Lebanon includes a hierarchy of goals, objectives, and sub-objectives, in line with the guidance in State’s Foreign Affairs Manual. For instance, the Lebanon ICS has objectives with sub- objectives that include activities outlining how to accomplish those objectives. The 2018 ICS contains 19 security-related activities with corresponding performance indicators for State and DOD activities, such as training Lebanese security forces in counterterrorism or border security operations. The hierarchy included in the 2018 ICS represents an improvement from the previous ICS, developed in 2015, which included information on goals, objectives, and sub-objectives, but did not outline specific activities or performance indicators. State guidance notes the benefit of such a hierarchy is that it shows the individuals who work on such activities how their actions contribute to achieving mission objectives. According to State officials, Embassy Beirut conducted an annual review of the ICS in October 2019. The goals of the annual review, according to State officials, were to assess progress against the ICS objectives and to remove or add goals, objectives, and key activities as needed. In July 2019, Embassy Beirut officials told us that they planned to conduct a review of the ICS approximately one year after its approval, which was in August 2018. However, State officials told us that leadership turnover in the summer of 2019 resulted in Embassy Beirut delaying the review until October 2019. Embassy Beirut, however, has not established targets for all of the 19 security-related performance indicators in its 2018 ICS. The Foreign Affairs Manual emphasizes that having targets to indicate the expected change over the course of each period of performance is a good practice. Several of Embassy Beirut’s security-related ICS indicators lack such targets, making it difficult for State to use the indicators to assess progress because it cannot compare the actual results of activities to the expected results. For example, several of the embassy’s security-related performance indicators deal with the number of people trained or improvements in specific capabilities of the security forces. Because the embassy has not established targets for these particular indicators, State cannot quantify the results it expects to achieve or determine how the actual results compare to those expectations. State officials noted that some bureaus have established performance indicators that are the same as, or similar to, security-related performance indicators in the ICS and some of those indicators have targets. For example, INL officials noted that INL has a Country Plan for Lebanon that has performance indicators and targets similar to some of the security-related performance indicators found in the ICS. However, many of the security-related activities included in the ICS are implemented by more than one agency or bureau. Therefore, the performance indicators for these activities would require targets that account for all the implementers. Additionally, Embassy Beirut did not have complete performance data for its security-related ICS performance indicators. State’s Foreign Affairs Manual emphasizes that regular, ongoing data collection against performance indicators to gauge the direct and near-term effects of activities is a good practice. Federal standards for internal control also state that agencies should use quality information that is, among other things, complete. Information is complete if it includes relevant data needed by decision makers to assess performance or to allocate resources. When we requested information on progress made toward the security-related indicators in the 2018 ICS, Embassy Beirut provided incomplete data for 11 of the 15 security-related indicators we analyzed. Data for six of these 11 were incomplete because the indicator called for quantitative data that were not included. For example, three of the six quantitative indicators called for data on the number or percentage of people trained. Embassy Beirut provided information that stated training had occurred, but did not quantify the number or percentage of people trained, as called for by the indicators. Data for the other five of these 11 indicators were incomplete because the indicators called for qualitative data that were not included. For example, three of the five qualitative indicators dealt with improving the capacity or capabilities of Lebanese units, but the information Embassy Beirut provided did not include a description of whether or how Lebanese units improved in those areas. Embassy Beirut provided complete data for four of the 15 indicators we analyzed, as shown in table 2. For three of the four indicators, Embassy Beirut provided the quantitative data called for by the indicator. For the remaining indicator, which dealt with number of personnel trained and the completion of facility upgrades, the embassy provided data on the number of personnel trained and a description of the status of the upgrades. According to Embassy Beirut officials, individual programs have targets and collect performance data associated with the security-related ICS performance indicators, but the Embassy did not have such information consolidated in a centralized document covering the time period we reviewed. Officials further noted that the ICS contains performance indicators, but not specific targets, as the ICS was not a vehicle for establishing specific targets when it was drafted in 2018. Additionally, State officials at headquarters stated that they do collect performance data related to some of the Lebanon ICS security-related indicators, but they did not provide evidence that this data is available to or used by Embassy Beirut as part of its ICS review. To review targets and indicators as part of the annual ICS review, Embassy Beirut officials said they planned to use evaluations and assessments of programs conducted by State and DOD headquarters entities or third parties. However, these assessments and evaluations cannot provide complete data on Embassy Beirut’s security-related performance indicators because not all of the security assistance programs in Lebanon have conducted them. In addition, these assessments and evaluations do not take place annually, which limits Embassy Beirut’s ability to use them on an ongoing basis to monitor strategic activities. Without setting targets and collecting complete data on performance indicators, Embassy Beirut will be limited in its ability to monitor its progress toward achieving the expected results of its security-related activities. State documents indicate that sound program design and performance management serve as the basis for efficient and effective use of department resources to achieve strategic objectives. If Embassy Beirut does not address the gaps in its performance information, it will be limited in its ability to ensure the intended alignment of policy, planning, resources, and programs through its annual reviews of the ICS. State and DOD Use Two Primary Safeguards to Limit the Risk That U.S. Assistance for Lebanon Will Benefit Terrorist Organizations State and DOD’s two primary safeguards to limit the risk that U.S. security assistance to Lebanon will benefit terrorist organizations are: 1) reviewing Lebanese security organizations for ties to terrorist organizations and 2) vetting individual recipients of assistance. For the first safeguard, State examines Lebanese security organizations for associations with foreign terrorist organizations (FTO) prior to providing support. Annual State, Foreign Operations, and Related Appropriations acts for fiscal years 2013 through 2018 included provisions to restrict funding for the ISF or the LAF if they are controlled by a U.S.-designated FTO. According to State officials, under these provisions, State regularly evaluates the LAF and ISF to determine if they have strong individual or organizational connections or alignment of purpose with Hizballah or any other FTO. State officials said they have determined that both the LAF and ISF are independent institutions that Hizballah does not control. State officials added that some longstanding divisions exist between Hizballah and the ISF, in particular. For example, one State official noted that Hizballah has assassinated ISF leaders in the past. Furthermore, members of the ISF are not allowed to be members of any political party, according to State officials. Second, State and DOD vet members of the Lebanese security forces who will receive U.S. assistance, such as training, for ties to terrorism. State and DOD vet by checking the names and other biographic or biometric information of potential recipients of assistance against information about known or suspected terrorists and their supporters. State and DOD officials conduct name-check vetting using one or more of three methods: In-country screening: State officials said they review a variety of sources in Lebanon to screen all potential recipients of State and DOD-funded training. Consular Affairs officials use State’s Independent Namecheck application to vet all potential trainees in country. This application allows overseas posts to screen names of individuals through State’s Consular Lookout and Support System (CLASS) database. CLASS contains records from numerous U.S. agencies on persons with immigration violations and terrorism connections, among other potential visa ineligibilities. In addition, officials said they may examine other sources, including local law enforcement or U.S. intelligence community sources. Terrorist Screening Center: State INL sends the names of potential ISF trainees to the Terrorist Screening Center, a multi-agency center administered by the Federal Bureau of Investigation, for further vetting. INL officials noted that this step does not result in many more exclusions beyond the initial in-country screening, but it serves as an additional check to ensure INL funding does not benefit FTOs. Nonimmigrant visa vetting: Any potential trainees who apply to come to the U.S. for training undergo vetting for a nonimmigrant visa, which includes interagency counterterrorism checks. According to DOD officials, a majority of their LAF trainees receive training in the United States. Some trainees under State programs also receive training in the United States. Officials said they believe these vetting procedures provide sufficient assurances that LAF and ISF trainees are not members of an FTO. They also stated they receive a relatively small number of “hits,” or indications that screening uncovered derogatory information. In these cases, officials said they remove the individual from the training roster and screen a substitute applicant instead. According to State officials, INL is in the process of moving its namecheck vetting from the Terrorist Screening Center to an internal State office. From 2012 to 2017, State piloted a counterterrorism vetting program for five countries, including Lebanon, through the Risk Analysis and Management (RAM) team in State’s Bureau of Administration. Vetting for Lebanon conducted through the pilot focused primarily on vetting contractors and grantees that would potentially implement U.S. assistance programs, including a security assistance program in 2015. RAM officials said that they resumed vetting in February 2019 for some programs in Lebanon, as determined by programming offices based on program-specific risk assessments that identify risks that can be mitigated through namecheck vetting. These officials said all the screening they conducted for Lebanon during the initial pilot phase was for programs determined to be of low or medium risk and, as of November 2019, they had not found derogatory information for any of the screened individuals. State and DOD Conducted All Required Checks of Equipment in Lebanon, but DOD Did Not Meet Its Timeliness Standards on Nearly One-Third of Observations State Inspected All Equipment Provided to the ISF Annually, in Accordance with Its Standards INL conducted annual inspections of equipment it provided to the ISF, as required by State policy. According to INL’s annual EUM reports, from 2013 through 2018, INL annually inspected 100 percent of the equipment valued at over $2,500 and defense articles regardless of value provided to the ISF, either by on-site inspection or host government verification. During our visit to an ISF site in Beirut, Lebanon, we found that all 16 items included in our random, non-generalizable sample were either physically present or accounted for through documentation. We observed 12 of the 16 items, such as police motorcycles and buses. The ISF provided documentation showing that the remaining four items, all trucks, were unavailable for inspection because the ISF had deployed them on missions. Figure 4 shows police motorcycles provided to the ISF that were inventoried by serial number. DOD Conducted Enhanced EUM Checks for All Items Requiring Them, but Did Not Meet Its Timeliness Standards for Nearly One-Third of Observations To provide reasonable assurance that recipients comply with U.S. government restrictions on the use and security of defense articles, DOD’s EUM standards require the Office of Defense Cooperation in Beirut to conduct enhanced EUM for designated sensitive defense articles, such as night vision devices provided to the LAF. U.S. officials must conduct an initial inventory of equipment requiring enhanced EUM within 90 days of delivery and must visually inventory 100 percent of enhanced EUM-designated defense articles within one year of the last inventory, or within 90 days of an acceptable reason for missing an inspection (such as the item was deployed), and enter inventory information into DOD’s SCIP database. DOD officials accounted for all of the 2,991 items subject to enhanced EUM from 2013 through 2018 at least once, according to our analysis of SCIP data. DOD officials in Beirut said they conducted serial number inventories of all items requiring enhanced EUM from 2013 through 2018, as required by DOD’s EUM program. During our visit to Lebanon in April 2019, DOD officials in Beirut physically observed nearly 100 percent (270 of 271) of the defense articles requiring enhanced EUM at the three LAF locations we visited. Only one of the 271 items was unavailable for inspection and the LAF provided documentation showing it was out for repair. Figure 5 shows night vision devices provided to the LAF that were inventoried by serial number. DOD reporting, including a 2017 DSCA Compliance Assessment Visit and U.S. Central Command Inspector General reports, indicates that the LAF has generally complied with DOD requirements to account for and secure equipment and conduct compliance checks of all required equipment. DOD officials said the LAF is transparent about the location of the equipment and goes out of its way to ensure DOD officials are able to account for it. The officials also said the LAF is rigorous about safeguarding all required equipment and consistently meets standards equivalent or similar to U.S. standards for equipment accountability. While our analysis showed that DOD generally accounted for items requiring enhanced EUM, we also found that DOD did not always conduct inspections consistent with its timeliness standards. If DOD does not inspect an item within the timeframes required by its standards, DOD considers the inspection delinquent. Our analysis of the duration between inspections from 2014 through 2018 showed delinquencies in each year and, in total, 32 percent (4,533) of the 14,287 recorded observations we analyzed for timeliness were delinquent. We found that 86 percent of the 2,874 items we analyzed for timeliness had at least one delinquent inspection during the 6 years we reviewed, and 61 percent had two or three delinquent inspections. While inspections were often delinquent, we found that the length of time items remained delinquent was not extensive, with the average length of each delinquency lasting 2.6 months. Only 1 percent of recorded observations showed a delinquency of 6 months or longer. Figure 6 shows the duration of delinquencies for those inspections that were delinquent. Officials from both the Office of Defense Cooperation in Beirut and DSCA stated that the method that DSCA uses to determine inspection due dates for annual inspection plans impedes the Office of Defense Cooperation’s ability to meet DOD’s timeliness standards. DSCA assigns due dates for items based on a general category code instead of an individual item’s serial number, which according to DOD officials, does not allow the Office of Defense Cooperation-Beirut to plan inspections in a way that meets DOD’s timeliness standards. For example, one type of night vision device represents 61 percent of the 2,991 items requiring enhanced EUM. Because these items all have the same general category code, DSCA designates all of them as due for inspection on the same day, regardless of when DOD officials last inspected each individual item. As a result, the inspection due dates DSCA establishes may be inconsistent with DOD’s guidance, which complicates planning and could result in some items having nearly 2 years between inspections before DCSA flags them as delinquent. Addressing how DSCA determines inspection due dates for items requiring enhanced EUM is important for ensuring the Office of Defense Cooperation has the information it needs to meet DOD’s timeliness standards for equipment accountability. According to DOD officials, as of April 2019, equipment on order for the LAF would double the number of items subject to enhanced EUM inspections. This increase underscores the importance of providing the Office of Defense Cooperation-Beirut accurate inspection due dates for the equipment provided to Lebanon. By not assigning inspection due dates consistent with DOD standards, DSCA hinders the Office of Defense Cooperation’s ability to plan effectively. It also increases the likelihood DOD will experience continued delays in conducting the required checks that ensure the proper safeguarding and usage of sensitive defense articles. Conclusions Recognizing Lebanon’s importance to the security and stability of the Middle East, U.S. agencies invested nearly $1.5 billion in security assistance to the country from fiscal years 2013 through 2018. However, the prominent role of Hizballah in the Lebanese government complicates the U.S. relationship with Lebanon and heightens the importance of ensuring strong management controls over U.S. assistance. U.S. agencies report that the LAF and ISF have improved in their capabilities to secure Lebanon’s border and to combat terrorist activity. Embassy Beirut has also taken a number of steps to track progress toward meeting U.S. security-related objectives. Gaps in the embassy’s performance information, however, limit its ability to monitor the ongoing progress of specific activities and to make informed decisions about where to allocate resources and attention. State and DOD did conduct end-use checks of all required items and their reporting indicates the LAF and ISF have generally taken appropriate steps to safeguard equipment. DOD, however, did not meet its timeliness standards for nearly one-third of all observations of sensitive equipment from 2013 through 2018. DSCA does not assign inspection due dates in a way that is consistent with DOD standards, which may limit DOD’s ability to fully ensure items requiring enhanced end-use monitoring are safeguarded and used as intended in a timely manner. Recommendations for Executive Action We are making a total of three recommendations, including two to State and one to DOD: The Secretary of State should direct the Department’s relevant bureaus to work with Embassy Beirut to establish, as appropriate, and consolidate targets for each of the security-related performance indicators. (Recommendation 1) The Secretary of State should direct the Department’s relevant bureaus to work with Embassy Beirut to collect and review performance data for key security-related performance indicators. (Recommendation 2) The Secretary of Defense should direct DSCA to revise the inspection due dates it establishes for items requiring enhanced EUM for the Office of Defense Cooperation in Beirut to align with DOD’s standards for EUM by considering the date of last inspection. (Recommendation 3) Agency Comments We provided a draft of this report to the Departments of State and Defense for comment. In its comments, reproduced in appendix II, State concurred with the recommendations that Embassy Beirut 1) establish, as appropriate, and consolidate targets for; and 2) collect and review performance data for its security-related performance indicators. State also provided technical comments, which we incorporated as appropriate. We requested comments on a draft of this product from DOD. The Director for Egypt, Israel, and the Levant in the Office of the Secretary of Defense for Policy provided us with the Department’s comments in an email stating that DOD concurs with the recommendation that DOD direct DSCA to revise the inspection due dates it establishes for items requiring enhanced EUM for the Office of Defense Cooperation in Beirut to align with DOD’s standards for EUM by considering the date of last inspection. The Director noted that the current SCIP EUM software complicates annual inventory planning and reporting and that DSCA’s EUM personnel have documented a software modification requirement that would allow them to implement the recommendation. Additionally, she stated that final design and budget decisions for fiscal year 2021 are not yet complete and the magnitude of this software modification is a major task that is core to EUM programming. We are sending copies of this report to the appropriate congressional committees and the Secretaries of State and Defense. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff has any questions about this report please contact me at (202) 512-2775 or fielde1@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. Appendix I: Objectives, Scope, and Methodology For fiscal years 2013 through 2018, we (1) examined to what extent the Department of State (State) and the Department of Defense (DOD) assessed the progress of their efforts to meet strategic objectives related to security for Lebanon; (2) described what safeguards State and DOD have put in place to limit the risk of U.S. security assistance provided to Lebanon benefitting terrorist organizations; and (3) analyzed to what extent State and DOD conducted end-use monitoring (EUM) checks of equipment provided to Lebanese security forces. To determine to what extent State and DOD assessed the progress of their efforts, we reviewed agency documentation and interviewed State, DOD, and Lebanese government officials. We reviewed Embassy Beirut’s Integrated Country Strategies (ICS), for fiscal years 2015 through 2018 to determine agencies’ strategic objectives for security assistance. We compiled information from State and DOD assessments and performance reporting on security assistance programs operating in Lebanon from fiscal years 2013 through 2018, including assessments, evaluations, and surveys. While we did not independently evaluate the quality of these documents, we did review their methodologies and determined that the approaches taken generally appeared reasonable. We also reviewed State and DOD performance reporting, such as program annual reports and Embassy Beirut’s Performance Plans and Reports for fiscal years 2013 through 2018. We then reviewed the compiled evidence to determine what this reporting showed about to what extent agencies had made progress toward their strategic objectives from fiscal years 2013 through 2018. We reviewed Embassy Beirut’s process for monitoring progress on its 2018 ICS—including what information the embassy compiles and how it determines whether programs are achieving their intended results. We reviewed State’s Foreign Affairs Manual and federal standards for internal control to identify key practices for ICS management. We compared the 2018 ICS to these key practices and requested information on Embassy Beirut’s assessment of progress on 19 security-related activities and indicators included in its 2018 Lebanon ICS. Embassy Beirut provided information for each of the activities and indicators as of May 2019. We reviewed the information provided by Embassy Beirut to determine if it was complete. Four of the 19 performance indicators covered activities for which performance data was not yet available. We did not include these four indicators in our analysis. For the remaining 15 indicators, we determined that Embassy Beirut provided complete data if it included relevant data needed by decision makers to assess performance. According to federal internal control standards, relevant data have a logical connection with identified information requirements. For example, if the information required for an indicator was quantitative in nature (such as the number or percentage of people trained), then we considered the information provided to be relevant if it included quantitative data that directly addressed the indicator. To describe what safeguards U.S. agencies have put in place to prevent U.S. security assistance from benefitting terrorist organizations, we reviewed legislative requirements, State policy guidance, and agency documentation. We reviewed annual appropriations acts from fiscal years 2013 through 2018 to determine what, if any, restrictions were placed on funding for Lebanon to ensure assistance did not benefit terrorist organizations. We reviewed State policy guidance on counterterrorism vetting and interviewed State and DOD officials in Washington, D.C. and Beirut, Lebanon about the steps they take to prevent assistance from benefitting U.S.-designated foreign terrorist organizations, including Hizballah. We reviewed agency documentation, including a risk assessment, an interagency memo, and State memos requesting the release of funding. We reviewed what safeguards State and DOD use to limit the risk of U.S. security assistance benefitting terrorist organizations, but did not analyze how the agencies made determinations when applying these safeguards. To evaluate to what extent State and DOD conducted EUM checks of equipment provided to Lebanese security forces, we reviewed agency documentation and data and interviewed State and DOD officials in Washington, DC and Beirut, Lebanon. We also reviewed State and DOD EUM standards to determine what requirements the agencies established for their respective programs and conducted site visits in Lebanon. To evaluate to what extent State conducted EUM checks in accordance with its standards, we reviewed State’s Bureau of International Narcotics and Law Enforcement Affairs (INL) annual end-use monitoring reports for 2013 through 2018 and analyzed INL EUM data. We also interviewed State officials in Washington, D.C. and Beirut, Lebanon about their processes for conducting and recording EUM. Because INL officials told us its EUM annual reports are the agency’s official documents for tracking adherence to EUM requirements, we used the reports in our analysis of State’s compliance with its standards. We interviewed INL officials about any identified discrepancies within the annual reports or between the annual reports and INL’s EUM data and determined that the reports were sufficiently reliable for our purposes. In Beirut, Lebanon, we visited two Lebanese Internal Security Forces (ISF) sites to observe the ISF’s processes for safeguarding and inventorying equipment. At the ISF’s Mobile Forces site, we reviewed a random, nongeneralizable sample of 16 items requiring EUM—all of which were vehicles, including buses, motorcycles, and trucks. We reviewed the serial numbers of items that were available on-site and reviewed ISF documentation accounting for those items that were not immediately available. To evaluate to what extent DOD had conducted EUM checks in accordance with its standards, we analyzed data from DOD’s Security Cooperation Information Portal (SCIP) database for 2013 through 2018 and interviewed officials from DOD’s Defense Security Cooperation Agency (DSCA) in Washington, D.C. and the Office of Defense Cooperation in Beirut, Lebanon. To analyze SCIP data, we compared observations recorded in the database against DOD’s standards. DOD’s Security Assistance Management Manual standards for EUM state that Security Cooperation Offices, like the Office of Defense Cooperation in Beirut, must visually inventory 100 percent of in-country enhanced EUM- designated defense articles within one year from the last inventory performed, except for those enhanced EUM-designated defense articles not available for observation (such as deployed or returned to the United States for repair), or as stipulated otherwise in the SCIP-EUM database or by separate policy memo. According to DOD’s standards, enhanced EUM-designated items not available for inventory during their annual inventory cycle due to deployment, returned to the United States for repair, or other legitimate reason, must be inventoried within 90 days after returning from deployment or repair. Each observation in the SCIP database represented a single inspection or attempted inspection of an item and includes, among other things, the item’s serial number, equipment category type, location, status, and date observed. Because we analyzed multiple years of data, DOD recorded more than one observation for almost all items. To evaluate to what extent DOD met its standards, we used the following parameters in our analysis: We determined that an observation met the requirement for being inventoried within one year from the last inventory performed if it occurred within 12 months of the last observation of the same item. Using a standard of 12 months between visits provides a small amount of leeway to account for the fact that security conditions, Lebanese Armed Forces (LAF) scheduling, or other factors (such as the 365th day falling on a holiday or weekend) outside of the Office of Defense Cooperation’s control could impact the exact date on which inventories were scheduled. If, for example, an item was inspected in February 2017 and again in February 2018, our analysis would consider it timely regardless of the actual date of inspection. We considered items that were unavailable for inspection due to deployment, repair, and security conditions to be accounted for because they were unavailable for legitimate reasons. If the Office of Defense Cooperation recorded an observation showing that an item was unavailable for legitimate reasons within 12 months of the last observation, we considered that observation to be timely. Because the SCIP dataset we analyzed does not include the date an item was returned from deployment or repair, we determined that a reinspection was timely if it was conducted within 4 months of the observation indicating the item was unavailable for inspection. We used a 4-month standard by examining the average and median length of time for a reinspection, which were 3.7 months and 3 months, respectively. The 4-month standard provides some time for an item to be returned before triggering the 90-day reinspection requirement. After we applied the 4-month standard, 367 observations, or about 3 percent of all observations we analyzed for timeliness, were still considered delinquent because they had not been reinspected within 4 months. On average, the items that were considered delinquent under this standard were reinspected about 8 months after they were considered delinquent, or about 12 months after the last attempted visit, indicating, on average, that these items were not inspected again until the next annual cycle. Because we do not know the date on which an item was returned, however, our analysis may slightly over-count delinquencies resulting from an item being unavailable for inspection. Our analysis only examined the time between recorded observations. Therefore, it did not count any items that were delinquent as of the end of 2018 if no observation had been recorded. The SCIP dataset includes no observation for 609 items in 2018. Of these, 117 were disposed of, lost, or expended in combat prior to 2018, 476 were observed in 2017 but delinquent as of the end of 2018, and 16 items were last observed before 2017. Additionally, due to data limitations, we did not analyze whether the first observation for each item was timely. Because our data set started in 2013, we did not have data on the date of the last observation for items delivered prior to 2013. DOD’s standards also state that DOD officials must first inspect items requiring enhanced end-use monitoring within 90 days of the item’s delivery. However, the SCIP data we analyzed only included the date the item was entered into SCIP, rather than the item’s delivery date, so we could not analyze whether the first inspection for items delivered after 2013 was timely. We did not include these observations in our analysis of timeliness. We conducted logical tests of the SCIP data, interviewed knowledgeable DOD officials about the database, and discussed our analysis with DSCA and ODC officials. We determined the data were sufficiently reliable for our purposes. We traveled to Tripoli, Lebanon and visited three LAF facilities to observe DOD procedures for conducting end-use monitoring and to see how the LAF safeguarded the equipment provided to them. We observed DOD’s enhanced end-use monitoring process for the 271 items in these three locations. We conducted this performance audit from October 2018 to December 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the Department of State Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Elizabeth Field, (202) 512-2775 or fielde1@gao.gov. Staff Acknowledgments In addition to the contact named above, Biza Repko (Assistant Director), Kara Marshall (Analyst-in-Charge), Adam Brooks, Lisa G. Shibata, Aldo Salerno, Neil Doherty, Martin de Alteriis, and Ashley Alley made key contributions to this report.

Why GAO Did This Study Since 2013, State and DOD have obligated nearly $1.5 billion in assistance to support Lebanese security forces. U.S. support for Lebanon is complicated due to the prominent role in the country of Hizballah, an Iranian-backed terrorist organization, which retains considerable influence as a major political party and a militia. The U.S. support includes equipment and training to build the capacity of Lebanese security forces. The equipment provided is subject to end-use monitoring requirements, which seek to ensure items are properly accounted for in Lebanon's inventory. GAO was asked to review U.S. security assistance provided to Lebanon since 2013. For fiscal years 2013 through 2018, this report (1) examines to what extent State and DOD assessed progress toward achieving strategic objectives related to security; (2) describes safeguards to limit the risk of U.S. assistance benefitting terrorist organizations; and (3) evaluates State and DOD end-use monitoring checks of equipment provided to Lebanese security forces. GAO analyzed State and DOD reports, documents, and data; and interviewed officials in Washington, D.C. and Beirut, Lebanon. What GAO Found The Departments of State (State) and Defense (DOD) reported progress in meeting security objectives in Lebanon, but gaps in performance information limit their ability to fully assess the results of security-related activities. State and DOD report improvements in Lebanese security forces' capabilities in key areas, such as border security. As part of monitoring such improvements and assessing the performance of security activities in Lebanon, State created related indicators but has not established targets for all of these indicators. Furthermore, State's data were incomplete for 11 of the 15 indicators GAO analyzed. For example, performance data for three indicators did not identify the number or percentage of people who received security training, as called for by the indicator. Without addressing these gaps, State has limited ability to determine to what extent it is achieving the intended results of its security-related activities in Lebanon. State and DOD use two primary safeguards to limit the risk of terrorist organizations benefitting from U.S. assistance to Lebanon. First, State routinely reviews the leadership of the Lebanese military and police forces and has determined they are not controlled by a foreign terrorist organization. Second, State and DOD vet potential trainees to ensure they do not have known or suspected ties to terrorism. Consistent with end-use monitoring requirements, State and DOD conducted required inventory checks of equipment provided to Lebanese security forces, but DOD did not meet its timeliness standards for nearly one-third of its observations. According to DOD officials, the method DOD uses to determine when it should complete annual inspections does not consider the date of the equipment's last inspection, which results in some inspections taking longer than prescribed by DOD's timeliness standards. Without conducting checks in a timely manner, DOD cannot fully ensure the equipment is properly accounted for and safeguarded. What GAO Recommends GAO is making three recommendations: 1) State should establish, as appropriate, and consolidate targets for its security-related performance indicators; 2) State should collect complete performance data for security-related indicators; and 3) DOD should revise its approach for determining when end-use monitoring inspections are to be completed to consider the date of last inspection. State and DOD concurred with these recommendations.

gao_GAO-19-537

gao_GAO-19-537_0

Background The DATA Act was enacted May 9, 2014, for purposes that include expanding on previous federal transparency legislation by requiring the disclosure of federal agency expenditures and linking agency spending data to federal program activities, so that both policymakers and the public can more effectively track federal spending. The act also holds agencies accountable for submitting complete and accurate data to Treasury and requires that agency-reported award and financial data comply with OMB and Treasury data standards. The DATA Act requires OMB and Treasury to establish government-wide data standards that to the extent reasonable and practicable provide consistent, reliable, and searchable spending data for any federal funds made available to or expended by federal agencies. These standards specify the data elements to be reported under the DATA Act and define and describe what is to be included in each data element, with the aim of ensuring that data will be consistent and comparable. The DATA Act requires OMB and Treasury to ensure that the standards are applied to the data made available on USAspending.gov and also requires agencies’ OIGs and GAO to review these data and report on their completeness, timeliness, accuracy, and quality. Sources of Data on USAspending.gov USAspending.gov has many sources of data, including data that agencies submitted to Treasury through their financial management systems and other data extracted from government-wide award systems that collect data from federal agencies and external award recipients. Treasury’s DATA Act Broker (Broker) is a key component of the data collection and reporting framework. The Broker enables agencies to upload, validate, and certify financial data and create linkages between the financial and award data for publication on the USAspending.gov website. Agencies are expected to submit three data files with specific details and data elements to the Broker from their financial management systems in accordance with Treasury guidance documents. File A: Appropriations account includes summary data such as the fiscal year cumulative federal appropriations account balances and includes data elements such as the agency identifier, main account code, budget authority appropriated amount, gross outlay amount, and unobligated balance. File B: Object class and program activity includes summary data such as the names of specific activities or projects as listed in the program and financing schedules of the annual budget of the U.S. government. File C: Award financial includes award transaction data such as the obligation amounts for each federal financial award made or modified during the reporting quarter (e.g., January 1, 2017, through March 31, 2017). The Broker also extracts data from four government-wide award reporting systems: the Federal Procurement Data System–Next Generation (FPDS-NG), System for Award Management (SAM), Financial Assistance Broker Submission (FABS), and the FFATA Subaward Reporting System (FSRS). These systems supply award and sub-award data (e.g., federal grants, loans, and contracts) to USAspending.gov. The systems compile data that agencies and external federal award recipients submit to report procurement and financial assistance award data required under the Federal Funding Accountability and Transparency Act of 2006 (FFATA). The four files produced with data that the Broker extracts from the four systems are as follows: File D1: Procurement includes data on the attributes of the award and the awardee and the recipient of the award (extracted from FPDS-NG on a daily basis) for procurement awards (contracts), if any, and contains elements such as the total dollars obligated, current total value of award, potential total value of award, period of performance start date, and other data to identify the procurement award. File D2: Financial assistance includes award and awardee attribute data (extracted from FABS nightly) on financial assistance awards (grants and loans) and contains data elements such as the federal award identification number, the total funding amount, the amount of principal to be repaid for the direct loan or loan guarantee, the funding agency name, and other data to identify the financial assistance award. File E: Additional awardee attributes includes additional data (extracted from SAM) on the award recipients and contains elements such as the awardee or recipient unique identifier; the awardee or recipient legal entity name; and data on the award recipient’s five most highly compensated officers, managing partners, or other employees in management positions. File F: Subaward attributes includes data (extracted from FSRS) on awards made to subrecipients under a prime award, if any, and contains elements such as the subaward number, the subcontract award amount, total funding amount, the award description, and other data to facilitate the tracking of subawards. According to Treasury guidance, after agencies submit Files A, B, and C, the Broker runs a series of validations and produces warnings and error reports for agencies to review. After passing validations for these three files, the Broker generates Files D1 and D2 containing details on procurement and assistance awards and performs a cross-file validation of linkages between File C and Files D1 and D2, which generates error and warning reports, as appropriate. The Broker also generates Files E and F containing data on highly compensated officers and subawards associated with the prime awards. There are no field-level or cross-file validations for Files E and F. With their quarterly submission, agency senior accountable officials (SAO) are required to certify the data submissions and to provide assurance over the alignment of Files A through F and that the data are valid and reliable in accordance with OMB guidance. According to Treasury officials, once the certification is submitted, a sequence of computer program instructions or scripts is issued to transfer the data from the Broker to tables set up in a database used as a source for the data on the website. Data are then displayed on USAspending.gov along with certain historical data from other sources, including Monthly Treasury Statements. Federal Financial Management SSPs OMB and Treasury implementation guidance called for customer agencies to consider how best to leverage their SSPs to capture data for their submissions, engage with their SSPs throughout the implementation process, and document the SSP role in agency DATA Act submissions. According to the SSPs, 60 non–Chief Financial Officers Act of 1990 (CFO Act) agencies use a federal SSP for all or part of the data submissions out of the 82 reporting data under the DATA Act as of the fourth quarter in fiscal year 2018. In 2014, Treasury designated four federal financial management SSPs to provide financial management services to other federal agencies. Although the four Treasury-designated federal financial management SSPs have changed over the years, the four federal financial management SSPs, which performed DATA Act services for external customers as of December 2018, are as follows: The Administrative Resource Center (ARC) is a Treasury SSP that provided financial management services to 42 customer agencies external to Treasury. According to ARC, 21 of those agencies received DATA Act services from ARC. The Enterprise Services Center (ESC) is a Department of Transportation financial management SSP that provided services to seven external customer agencies. According to ESC, six of those customer agencies received DATA Act services from ESC. Pegasys Financial Services (PFS) is a Department of Agriculture financial management SSP that provided services to 37 external customer agencies. According to PFS, 24 of those customer agencies received DATA Act services from PFS. The Interior Business Center (IBC) is a Department of the Interior financial management SSP that provided services to 18 external customer agencies. According to IBC, nine of those customer agencies received DATA Act services from IBC. The DATA Act requires agencies’ OIGs to issue reports assessing the quality of the agencies’ spending data submissions and compliance with the DATA Act. In the OIGs’ reports covering their agencies’ second quarter fiscal year 2017 submissions, nine OIGs reported issues with their agencies’ use of an SSP for DATA Act submissions. Five of the nine OIGs issued recommendations related to these issues, and four agencies concurred with the recommendations. For example, one OIG recommended that its agency work closely with its SSP to address timing and coding errors that the SSP caused for future DATA Act submissions. Another OIG recommended that its agency work with its SSP to identify OMB requirements that the SSP is to perform and insert them into the service-level agreement, in order to address errors caused by confusion as to whether the SSP or the agency should submit certain types of data. Although our prior reports on the DATA Act included recommendations, our recommendations were not related to SSPs’ implementation of the DATA Act. The DATA Act requires OIGs and GAO to issue their second reports on data quality in November 2019. Federal SSPs Provide a Variety of DATA Act Services for Their Customer Agencies Customer Agencies Use Federal Financial Management SSPs for a Variety of Services The 27 customer agencies that responded to our survey reported that the four federal financial management SSPs provide a variety of services (see fig. 1). All 27 agencies reported using their federal financial management SSPs for DATA Act services, and almost all of the agencies used their SSPs for several other financial management services, such as general ledger accounting, financial reporting, and hosting the customer agencies’ financial systems. As such, the SSPs play a key role in helping to ensure that these customer agencies successfully carry out the requirements of the DATA Act and submit Files A, B, and C from their financial management systems. In addition, 17 agencies reported using their SSPs for payroll or budget execution services, while fewer reported using their SSPs for other financial management services, such as grant or loan processing. DATA Act services. All 27 customer agencies responding to our survey reported using an SSP for DATA Act services. As discussed in more detail below, these DATA Act services may include activities such as preparing DATA Act files from financial systems, consolidating DATA Act files from multiple agency component entities, reconciling DATA Act files to other source data, and uploading DATA Act files to the Broker for validation. General ledger accounting. Twenty-six agencies reported using SSPs for general ledger accounting, which may include activities such as general ledger setup and maintenance, posting transactions to the general ledger, accrual and liability processing, and period-end general ledger closing. Financial reporting. Twenty-six agencies also reported using SSPs for financial reporting, which may include activities such as Treasury reporting, financial statement preparation, cash forecasting and reporting, and financial performance and operational reporting. Financial system hosting. Twenty-five agencies reported using SSPs for financial system hosting, which may include services such as systems management and monitoring, disaster recovery, help desk administration, network security compliance and controls, and continuity of operations plans and testing. Invoice processing. Twenty-four agencies reported using SSPs for invoice processing, which may include services such as recording receiving and acceptance reports, recording invoices, matching invoices to receiving and acceptance reports, and routing invoices to obtain approval for payment. Budget execution. Seventeen agencies reported using their SSPs for financial management services related to budget execution, which may include activities such as budget setup and maintenance, fund allocation and control, and budgetary reporting. Payroll. Seventeen agencies reported using SSPs for payroll. SSP payroll services may include recording payroll and benefit payments; reconciling payroll service data with financial management data; and recording credits, payment adjustments, and employee receivable offsets. Procurements/contracts. Thirteen agencies reported using SSPs for procurement and contract services. SSP procurement and contract services may include recording credits and payment adjustments; auditing payments; processing payments for incurred expenses and payments in advance; and capturing award identifier data, such as the Procurement Instrument Identifier (PIID) and agency Unique Record Identifier (URI) to support DATA Act reporting. Grants processing. Seven agencies reported using SSPs for grants processing, which may include recording requests for grant payments, matching grant payment requests to obligating documents, routing grant payment requests for approval, and generating payment transactions. These processes also include payments for expenses and payments in advance as well as capturing award identifier data, such as Federal Award Identification Numbers (FAIN) and Catalog of Federal Domestic Assistance (CFDA) codes to support DATA Act reporting. Loans processing. Three agencies reported using SSPs for loan processing. Loan processing services may include recording requests for loan payments, matching loan payment requests to obligating documents, generating payment transactions, resolving payment issues, and recording credits and payment adjustments. Customer Agencies Rely on Their SSPs to Perform Various DATA Act Services All 27 agencies responding to our survey reported that their federal SSPs perform a variety of DATA Act services or activities, as shown in figure 2. Preparing data files A, B, or C and uploading them to the Broker are the most prevalent DATA Act services or activities that the federal SSPs perform, whereas fewer than half of the SSPs certify and publish the files for the agency after receiving agency approval. All 27 agencies reported that their SSPs prepare at least one of the Files A, B, or C using data from either SSP or customer agency financial systems. In addition, 15 of the 27 agencies reported that their SSPs consolidate DATA Act files from multiple agency components. Seventeen agencies reported that their SSPs reconcile Files A, B, or C to other source data. For example, a reconciliation of general ledger and subledgers may include verifying that (1) general ledger account balances can be traced to aggregated or discrete agency transactions and (2) aggregated or discrete agency transactions can be traced to the point of origination and source documents. Twenty-five of the 27 agencies reported that their SSPs upload Files A, B, or C to the Broker for validation. In turn, the Broker runs a series of data validations and produces warnings and error reports for agencies to review after the files are submitted. Twenty-one agencies reported that their SSPs address these warnings and errors on their behalf. After warnings have been reviewed and all errors have been addressed, Files A, B, and C have been uploaded and Files D1, D2, E, and F have been generated, the agency’s SAO is required to certify the validity and reliability of the data submissions in accordance with OMB guidance. Twenty-four agencies reported that their SSPs provide final Files A, B, or C for the customer agency to review and certify in the Broker, and 11 agencies reported that their SSPs finalize the files in the Broker and click the Certify and Publish button after receiving agency approval to certify. Most Customer Agencies and the Four Federal SSPs Reported DATA Act Challenges and Have Taken Steps to Address Them Customer Agencies and SSPs Reported Various Challenges Affecting Timeliness, Completeness, and Accuracy of Their DATA Act Submissions We asked customer agencies in our survey to specify the challenges associated with using an SSP that they experienced since their initial DATA Act submission; the SSP’s role in these challenges; and whether the challenges affected the timeliness, completeness, or accuracy of their submissions. Sixteen of the 27 customer agencies that responded to our survey identified one or more challenges associated with using an SSP (see fig. 3), many of which affected the timeliness, completeness, and accuracy of agency submissions. The survey questions and summarized results are shown in appendix II. In addition, officials from all four federal SSPs described various challenges they experienced in helping their agency customers with DATA Act submissions. The challenges reported by these 16 customer agencies and the federal SSPs are summarized below. Depending on the effectiveness of agencies’ and SSPs’ actions to address them (as discussed further below), these challenges may increase the risk that agencies will be unable to submit quality data in accordance with the DATA Act. Dependencies. Ten agencies reported that they have experienced challenges related to agency submission activities that depend on relationships with, or actions being taken by, the SSP before the agency can proceed. One agency reported that it must rely on its SSP to prepare, validate, and finalize all DATA Act files prior to agency certification and the files are often submitted close to the due date. Another agency reported that its SSP provided DATA Act submission files to the agency the day before or near the certification deadline. Relying on SSPs to prepare DATA Act files in a timely manner increases the risk that agencies may be unable to certify and publish their DATA Act submissions on time. Resources. Seven agencies said that they have experienced resource challenges related to a lack of funding or human resources at the customer agency or its SSP. One agency noted that its SSP has only a small group of people that assist with all of its SSP’s services, making it challenging in particularly busy seasons (such as the close of the fiscal year) for the SSP to meet internal deadlines and resolve data discrepancies affecting the timeliness, completeness, and accuracy of data submissions. Another agency reported challenges with funding resources, noting that the agency has been unable to use its SSP’s integrated financial and procurement system because of the costs associated with implementation, operation, and maintenance. The shortage of resources to bring on more staff or improve systems increases the risk that agencies may be unable to submit quality data and fully carry out DATA Act requirements. Competing priorities. Five agencies said that they have experienced challenges related to statutory, regulatory, policy, or other matters that have competing priorities or conflicting requirements that may affect an agency or its SSP’s DATA Act submission process. One agency reported that the fourth quarter DATA Act reporting deadline falls within fiscal year- end reporting time frames, requiring the agency to prioritize fiscal year- end reporting over some of the DATA Act reporting tasks. Similarly, officials from an SSP told us that they also experienced challenges with the short turnaround times required to incorporate system updates for DATA Act submissions. Data quality. Four agencies reported that they have experienced challenges related to meeting DATA Act requirements for data quality because they use an SSP, including completeness and accuracy of agency data to be reported as well as SAO certification and reporting of nonfinancial data elements. One agency reported that its SSP included extraneous transactions in its File C that were not required for DATA Act reporting. This created a high volume of warning messages when the Broker compared the data with file D2 during the validation process. Another agency reported that its SSP provided incomplete files for agency certification and that the data in the files did not reflect the data in the agency’s financial reports. These challenges not only increase the risk of lower-quality agency data submissions but may also require SSPs and customer agencies to expend additional resources to address warning messages that the Broker generated. Guidance. Four agencies reported experiencing challenges involving incomplete, unclear, missing, and evolving OMB and Treasury guidance related to SSP implementing requirements and Broker changes, including data elements, the technical schema, and other key policies. One agency noted that guidance on performing quarterly certifications is not readily available. Two agencies reported challenges involving a lack of guidance on how to communicate error corrections and desired changes to their SSPs. Lack of guidance could result in misunderstandings or miscommunications between the SSP and its customer agency, increasing the risk of delays or errors in the agency’s data submissions. Technology. Four agencies reported that they have experienced technology challenges with developing and submitting required files. These challenges include SSP infrastructure issues, such as integrating multiple existing and disparate management systems or their SSPs needing to modify existing systems to implement the DATA Act. Agencies reported that some systems are unable to include the required data elements for all reported transactions. Similarly, an SSP official told us prior to our survey of customer agencies that the SSP also experienced technical challenges with systems and data that have since been resolved. Another SSP experienced challenges with its financial systems not capturing award identification data elements, such as the PIID. Such limitations in customer agency and SSP technology may require the use of limited resources for error corrections and manual work-arounds, increasing the risk of reporting errors, and may hamper customer agencies’ and SSPs’ ability to submit quality data in accordance with the DATA Act. Project management. Two agencies reported that they have experienced challenges related to their SSPs’ project management, such as the lack of a designated project manager and inadequate documentation of progress made or key decisions. Specifically, both customer agencies said that their SSPs did not provide data in a timely manner for their review prior to submission. One agency reported that although this did not affect the timeliness of its submission, it affected data quality because the agency did not have sufficient time to test and implement sufficient internal controls and validation procedures prior to data being published on USAspending.gov. Additionally, the same agency reported that there is no senior project manager at the SSP who oversees the processes used to provide financial management services to the agency. This challenge may affect agencies’ ability to resolve errors, increasing the risk that they submit incorrect data. However, none of the four federal SSPs described any project management challenges when we asked them prior to our survey what challenges they faced in carrying out their roles and responsibilities for assisting their customer agencies with implementing the DATA Act. Agencies Reported Steps They Have Taken to Address Identified DATA Act Submission Challenges and Ensure Data Quality We asked customer agencies in our survey to describe the steps they and their SSP have taken to address reported challenges and help mitigate risks associated with them. Of the 16 agencies that identified challenges, 12 agencies reported that they had already taken steps to address them and five agencies said they were aware of steps their SSPs had taken. As discussed in more detail below, communication and coordination between the SSP and Treasury, as well as customer agency technological improvements and manual work-arounds were the steps most often reported by agencies to address identified challenges. Communication and coordination. Eight of the 12 agencies described communication efforts with their SSPs or Treasury to facilitate coordination and seek information needed to address their challenges associated with using an SSP. These efforts include requesting information from the SSP, Treasury, and other government resources to obtain additional knowledge regarding the DATA Act and to prepare internal guidance and procedures. One agency reported implementing a weekly meeting with its SSP on DATA Act reporting. Another agency reported that its SSP has been very proactive in sharing information (bulletins, updates, etc.) and assisting with submitting DATA Act information. According to the eight customer agencies, increased communication and coordination has helped to address several technology, dependency, and resource challenges. Technology improvements. To address technology challenges, four agencies discussed making improvements in technology at both the agency and the SSP. The improvements include implementing an integrated financial and procurement system platform and working with the software vendor to obtain access to FPDS-NG for anticipated 2019 procurement activity reporting. Another agency is currently implementing a technological solution to aid in consolidating and reconciling files. In addition to technological solutions, two agencies reported using manual work-arounds, such as developing and implementing internal manual processes to reconcile and correct data files. Some agencies also discussed actions their SSPs had taken to address technology issues, including two SSPs that are working with the developers to address software issues. One agency reported that in addition to addressing technology challenges, these improvements also provided substantial cost savings. We also asked the 27 customer agency survey respondents to specify the internal control processes and activities they use to assure the quality of data submitted to the Broker. Twenty-four of 27 agencies reported that they use various processes and activities to provide such assurance. Specifically, these 24 agencies reported that they reconcile data files to other agency data and sources (e.g., the Governmentwide Treasury Account Symbol Adjusted Trial Balance System). These reconciliations can help identify errors in data files and ensure that they are consistent with other agency data. In addition, 18 agencies reported that they review their SSPs’ Service Organization Control (SOC) reports to identify any internal control deficiencies, and nine agencies reported that they implemented controls to address control deficiencies identified in their SSPs’ SOC reports. Twenty agencies reported that they review or verify agency data displayed on USAspending.gov. By reviewing these data, agencies can confirm that the data that they uploaded to the Broker are presented accurately on the website. Nineteen agencies reported that they incorporate the results of OMB Circular No. A-123 reviews that affect their DATA Act submissions. OMB Circular No. A-123 provides a methodology for agency management’s reporting on internal controls over reporting, and it also establishes an assessment process based on our Standards for Internal Control in the Federal Government that management must implement in order to properly assess and improve internal controls over operations, reporting, and compliance. Agencies Described Useful Practices for Working with SSPs on DATA Act Submissions Twenty of the 27 agencies that responded to our survey described useful practices in working with an SSP on DATA Act submissions. These agencies reported most often that discussing issues with the SSP and performing data reconciliations or comparisons were helpful. For example, 12 agencies reported that working and communicating with the SSP was useful. They also reported that having a readily available point of contact for better communication and communicating early about the need to complete the DATA Act submission helps to resolve any concerns prior to the due date. One agency reported that it conducts weekly meetings with its SSP to discuss key topics, including implementation, data quality, and reporting processes and procedures. Eleven agencies reported that conducting data reconciliations or comparisons, and creating a standard operating procedure to ensure that their data are consistently reviewed, reconciled, corrected, and certified, was also useful. One agency noted that in addition to the quarterly files that require certification, the SSP also provides monthly files that the agency can review to provide additional time to correct any identified errors. These 20 agencies also suggested other practices for successfully working with an SSP on DATA Act submissions, such as automating reconciliations and other internal control processes to increase efficiency, implementing continuous training and monitoring, assigning an accountant as an agency contact, and conducting an analysis of agency risk as recommended in the Data Quality Playbook. One agency reported that it performs extensive comparisons of agency-generated data reports to SSP-prepared data files, and that it partially automated this process to help increase efficiency. Agencies Described the Need for Additional Steps, Tools, and Guidance Related to Using an SSP for DATA Act Submissions Nine of the 16 agencies that identified challenges associated with using an SSP reported that their agencies need to take additional steps to address their identified challenges. Some of these agencies reported that they still need to address issues such as correcting data and improving communication with their SSPs. These agencies also reported the need to negotiate annual service-level agreements with their SSPs to address resource and competing priority concerns. One of these agencies reported that it continues to work with its SSP to understand what is lacking in the process of correcting PIID information for obligations. A few agencies reported the need to develop internal guidance on topics such as data quality plans per OMB guidance and a reconciliation process to address their data quality challenges. Finally, one agency reported that it is in the process of hiring additional personnel to address its challenges with competing priorities. While customer agencies are primarily responsible for the quality of their DATA Act submissions, six of the 16 agencies that reported challenges reported that their SSPs also need to take certain steps to address identified challenges, such as communicating with the customer agency and making technology improvements. For example, one agency reported that its SSP does not provide the customer agency with updated data submission files after the agency has requested changes. The customer agency suggested that the SSP provide updated files more often to help the agency ensure that the changes are included in the final file it submits to the Broker. Another agency reported that its SSP has been experimenting with different methods to eliminate cross-file warnings and errors in File C that need to be addressed. Five of the 27 customer agencies we surveyed reported that additional tools or guidance from OMB, Treasury, or other entities (such as the SSP) could assist agencies with using an SSP for DATA Act submission. Specifically, three agencies reported that they would like to have guidance, including standard operating procedures, for communicating and working with their SSPs. One agency suggested additional OMB or Treasury training on compliance with the DATA Act, and one agency suggested improvements by Treasury to prevent Broker errors that result from normal business scenarios and require manual work-arounds to the agencies’ system-generated files. OMB staff told us that they are involved with the DATA Act Executive Steering Committee, working closely with Treasury to oversee all aspects of both policies and implementation related to federal spending transparency efforts. According to OMB staff, neither the SSPs nor their customer agencies have reported any current challenges with DATA Act submissions to OMB. OMB staff stated that effective implementation of OMB Memorandum M-18-16 guidance to agencies and SSPs, which discusses establishing entity-level controls related to using SSPs, would help to ensure that the SSPs provide quality services. In April 2019, OMB issued Memorandum M-19-16 on shared services, which among other things described the process and desired outcomes for shared services and established a governance and accountability model for achieving them. For example, as it relates to the DATA Act, the memorandum calls for the Shared Services Governance Board to leverage the DATA Act Executive Steering Committee’s work on DATA Act standards. Treasury officials told us prior to our survey that they held two workshops for SSPs in the early stages of implementation to address specific concerns and questions on DATA Act implementation. Treasury officials stated that the only challenge reported by SSPs to the department related to linking and certifying award data using the Broker when the awarding agency and the file C reporting agency were different. To address this concern, Treasury added clarification in guidance on files A, B, and C submissions and a new Broker feature allowing agencies to specify whether the award data submitted in Files D1 and D2 comes from the funding agency or the awarding agency. Agency Comments We provided a draft of this report to OMB; Treasury, including ARC; ESC; PFS; and IBC for comment. OMB, ESC, and IBC told us that they had no comments on the draft report. Treasury and PFS provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the Director of the Office of Management and Budget, the Secretary of the Treasury, the four federal financial management shared service providers, and interested congressional committees and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-9816 or rasconap@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix III. Appendix I: Objectives, Scope, and Methodology The Digital Accountability and Transparency Act of 2014 (DATA Act) requires us to review Office of Inspector General (OIG) reports and issue reports of our own assessing and comparing the completeness, timeliness, accuracy, and quality of the data that federal agencies submit under the act and the implementation and use of data standards. We issued our first report on data quality in November 2017, as required. In July 2018, we issued a report on our review of OIG reports on agencies’ first DATA Act submissions and in the course of our review found that some OIGs reported challenges involving the use of federal shared service providers (SSP) that helped agencies implement the DATA Act. For this report, our objectives were to describe (1) the types and variations of services that the federal SSPs provide to their financial management customer agencies to assist them in implementing the DATA Act and meeting the act’s requirements and (2) any challenges that federal SSPs and their financial management customer agencies have encountered in their efforts to ensure the quality of data submissions consistent with the standards established under the DATA Act and the steps they have taken to address those challenges. To address our first objective, we interviewed four federal SSPs and surveyed their financial management customer agencies to identify the types and variations of services the SSPs provide related to DATA Act implementation and meeting the act’s requirements. We also obtained and reviewed selected service-level agreements executed between the four SSPs and their financial management customer agencies to determine the types and variations of DATA Act services that the SSPs provided to them. In December 2018, we emailed a survey questionnaire to 67 customer agencies that the four federal SSPs told us were external customers for DATA Act or other financial management services and that also submitted data under the DATA Act as of December 2018. During the survey, we determined that 60 of those 67 agencies actually received DATA Act services from a federal SSP and were eligible members of our study population. We received survey responses from 31 agencies by our January 2019 deadline and, based on our review, determined that 27 were eligible and sufficiently complete for our purposes. After excluding ineligible agencies from our population, the response rate was 45 percent. In developing, administering, and analyzing the survey, we took steps to minimize the five types of potential errors, described below, that may affect survey results. Because we surveyed all agencies in our population, there was no sampling error. To minimize the effects of coverage error—the exclusion of some eligible members of the population, or inclusion of ineligible members—we identified as ineligible and removed seven initially identified agencies because we determined that they did not use a federal SSP to provide DATA Act services. Measurement error may result from differences in how a question is interpreted and the sources of information available to respondents. To help prevent measurement error, we conducted pretests of the draft questionnaire with four customer agencies, each using a different SSP, and made revisions to improve the validity and minimize the burden of responding to our questions. Nonresponse error may result when a survey fails to capture information from all agencies selected in the survey, and it may introduce bias if those agencies that did not respond would have given materially different answers than those that did. To maximize survey response, we sent multiple email reminders to the surveyed agencies and extended the submission deadline. While we do not have evidence of material bias from those not responding, we limit our survey results in this report as representing only those 27 agencies responding. Finally, to limit the possibility of processing error, survey responses were checked for invalid or illogical answer patterns, and data edits were made as necessary to facilitate processing and analysis of the results. This analysis was verified by a separate data analyst. Table 1 lists the 27 customer agencies (by shared service provider) for which we obtained, reviewed, and included customer agency survey responses. The survey questions and summarized results are shown in appendix II. To meet our second objective, we interviewed the officials of the four federal SSPs prior to our survey to obtain information on the challenges the SSPs and their customer agencies encountered and steps taken to address them. We reviewed and analyzed the 27 customer agency survey responses to identify challenges responding agencies reported since their initial DATA Act submissions because they are working with an SSP and any steps SSPs and their financial management customers took to address challenges and to help ensure the quality of data submissions. We did not corroborate the customer agencies’ survey responses with the four federal SSPs, Office of Management and Budget (OMB), or Department of the Treasury (Treasury). To identify steps taken to address challenges, we also obtained and reviewed any reports related to DATA Act implementation that the SSPs or their respective OIGs produced. In addition, we interviewed OMB staff and Treasury officials about any guidance they have provided or actions they have taken to assist the four SSPs and their financial management customers with any challenges related to DATA Act compliance. We conducted this performance audit from July 2018 to July 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Results of Survey of Federal Financial Management Shared Service Provider Customer Agencies United States Government Accountability Office Survey of Customer Agencies’ Use of Shared Service Providers for Digital Accountability and Transparency Act (DATA Act) Submissions This questionnaire asks about your agency’s use of and relationship with your federal shared service provider (SSP). It should be completed by officials knowledgeable about the DATA Act services your federal SSP provides. Please submit only one survey response per agency, but consult with other officials as needed; when answering, please consider any agency component activity and experiences together, and answer at the agency level to the best of your ability. If your agency uses more than one federal SSP, please be sure to include information about both SSPs in your answers. This is a fillable PDF form. You can click buttons and type into highlighted boxes throughout the form; the boxes will accommodate more text than is immediately visible. Save this file to a drive now, and save your answers periodically as you go. When completed, save this file and email it to DATAActImplementation@gao.gov. If a “Submit” button appears in the upper right corner of your screen, you may also use that to automatically email your completed questionnaire (some viewers will not see this button depending on your system’s Javascript settings). If you have any questions, or feel this questionnaire was sent to your office in error, please contact 1. What is the name of the person completing this questionnaire, title, agency name, and contact information? (Please submit only one survey response per agency) Phone: 2. What is the name and title of the individual who reviews and certifies your agency’s DATA Act submission in the Treasury broker as ready for publishing? Title: a. Is this person also your agency’s DATA Act Senior Accountable Official (SAO)? 3. Which federal SSP(s) (if any) does your agency currently use for the financial management services listed below? (Select all that apply) 4. Which specific DATA Act services/activities does the federal SSP(s) identified in question 3 perform for your agency (in whole or in part)? (Select all that apply) Finalize the files in the Treasury broker by clicking the “Certify and Publish” button after receiving agency certification Other SSP services or activities (specify in the box below) 5. Which, if any, of the following activities does the federal SSP(s) you identified above initiate for your agency (in whole or in part) for the broker to perform? (Select one answer in each row) 6. What are the steps taken by your agency to certify the final DATA Act files before they are published (e.g., by whom and how are the data validated, reviewed, and comments (if any) provided on the files)? Summary included in report 7. Since your agency’s initial DATA Act submission, has your agency experienced any challenges in the following areas because it is working with an SSP, and did the challenge(s) in working with the SSP have an impact on the timeliness, completeness, and accuracy of any of your agency’s submissions? (Select all the area(s) with challenges and any impacts that apply) Technology issues Including challenges with developing and submitting required files, and SSP infrastructure issues such as integrating multiple existing and disparate financial and management systems, or the SSP needing to install new systems or modify existing systems to implement the DATA Act. Dependencies Agency submission activities depend on relationships with or actions being taken by the SSP before the agency can proceed. Guidance Incomplete, unclear, missing, and evolving guidance related to the SSP implementing requirements and broker changes, including data elements, the technical schema, and other key policies issued by OMB and Treasury. Resources Lack of funding or human resources by your agency or SSP. Project management Challenges related to the SSP’s project or program management, such as lack of a designated project manager and inadequate documentation of progress made or key decisions. Data quality Issues related to meeting DATA Act requirements for data quality because of the use of an SSP, including completeness and accuracy of agency data to be reported, as well as Senior Accountable Official certification and reporting of nonfinancial data elements. Competing priorities Statutory, regulatory, policy or other matters that have competing priorities or conflicting requirements that may affect an agency or their SSP’s DATA Act submission process. Challenge? Contract Management Challenges related to the management of the Service Level Agreement (SLA) and/or tasks and services that the SSP provides for the customer agency or the SSP should be providing but are not in the SLA. Other challenges (specify in box below) 8. What were the specific challenge(s) with your agency’s DATA Act submissions identified in question 7, and how did your SSP play a role? If your agency used multiple SSPs, please specify to which provider the challenge(s) was related. a. What steps, if any, has your agency taken to address these challenge(s)? b. What steps, if any, remain to be taken by your agency to address these challenge(s)? c. What steps, if any, are you aware of that your SSP has taken to address these challenge(s)? d. What steps, if any, remain to be taken by your SSP to address these challenge(s)? Summary included in report 9. What management or oversight practices has your agency found to be useful in working with your SSP on DATA Act submissions? 10. What internal control processes and activities does your agency use to provide assurance over the quality of data submitted to the Treasury broker and displayed on USAspending.gov? Internal control processes and activities: Reconcile data files to other agency data and sources (e.g., SF 133, GTAS) Review SSP’s Statement on Standards for Attestation Engagements No. 18/Service and Organization Controls (SOC) reports to identify any internal control deficiencies. (Describe internal control deficiencies related to DATA Act submissions, if any, in box below) Implement complementary controls to address SSP control deficiencies identified in the SOC report. (Describe controls implemented, if any, in box below) Review/verify agency data displayed on USAspending.gov Incorporate the results of A-123 reviews that have an impact on DATA Act Other internal control processes or activities (specify in box below) 11. What additional tools or guidance, if any, are needed from OMB, Treasury, or others to assist with your agency’s use of the SSP or DATA Act submission? 12. Please provide any additional comments or explanations not already discussed above. Please save and e-mail your responses to DATAActImplementation@gao.gov. Thank you for completing our questionnaire! Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Michael LaForge (Assistant Director), Laura Pacheco (Auditor in Charge), Umesh Basnet, Thomas Hackney, Roy Kilgore, and Diane Morris made major contributions to this report. Other key contributors include Dave Ballard, Jenny Chanley, Peter Del Toro, Patrick Frey, Ricky Harrison, Maxine Hattery, Jason Kelly, James Kernen, Christina Quattrociocchi, Carl Ramirez, Michelle Sager, and James Sweetman. Related GAO Products DATA Act: OMB Needs to Formalize Data Governance for Reporting Federal Spending. GAO-19-284. Washington, D.C.: March 22, 2019. Streamlining Government: OMB and GSA Could Strengthen Their Approach to Implementing a New Shared Services Plan. GAO-19-94. Washington, D.C.: March 7, 2019. Open Data: Treasury Could Better Align USAspending.gov with Key Practices and Search Requirements. GAO-19-72. Washington, D.C.: December 13, 2018. DATA Act: Reported Quality of Agencies’ Spending Data Reviewed by OIGs Varied Because of Government-wide and Agency Issues. GAO-18-546. Washington, D.C.: July 23, 2018. DATA Act: OMB, Treasury, and Agencies Need to Improve Completeness and Accuracy of Spending Data and Disclose Limitations. GAO-18-138. Washington, D.C.: November 8, 2017. DATA Act: As Reporting Deadline Nears, Challenges Remain That Will Affect Data Quality. GAO-17-496. Washington, D.C.: April 28, 2017. DATA Act: Office of Inspector General Reports Help Identify Agencies’ Implementation Challenges. GAO-17-460. Washington, D.C.: April 26, 2017. DATA Act: Implementation Progresses but Challenges Remain. GAO-17-282T. Washington, D.C.: December 8, 2016. DATA Act: OMB and Treasury Have Issued Additional Guidance and Have Improved Pilot Design but Implementation Challenges Remain. GAO-17-156. Washington, D.C.: December 8, 2016. DATA Act: Initial Observations on Technical Implementation. GAO-16-824R. Washington, D.C.: August 3, 2016. DATA Act: Improvements Needed in Reviewing Agency Implementation Plans and Monitoring Progress. GAO-16-698. Washington, D.C.: July 29, 2016. DATA Act: Progress Made but Significant Challenges Must Be Addressed to Ensure Full and Effective Implementation. GAO-16-556T. Washington, D.C.: April 19, 2016. DATA Act: Data Standards Established, but More Complete and Timely Guidance Is Needed to Ensure Effective Implementation. GAO-16-261. Washington, D.C.: January 29, 2016. DATA Act: Progress Made in Initial Implementation but Challenges Must be Addressed as Efforts Proceed. GAO-15-752T. Washington, D.C.: July 29, 2015.

Why GAO Did This Study Over the past 2 decades, the federal government has undertaken efforts to save money and increase efficiencies by encouraging agencies to use administrative and operational services and processes that other federal and external parties provide, commonly referred to as shared services. The DATA Act was enacted to increase accountability and transparency and, among other things, establish government-wide data standards. Certain agencies have used shared services of federal SSPs to implement the act. The act also requires a series of oversight reports by agencies' Offices of Inspector General (OIG) and GAO. OIGs for five agencies have made recommendations related to agencies' use of SSPs for DATA Act services, and four agencies concurred with the recommendations. The objectives of this report are to describe (1) the types and variations of services that federal SSPs provide to their financial management customer agencies to assist them with implementing the DATA Act and meeting the act's requirements and (2) the challenges federal SSPs and their financial management customer agencies have encountered in their efforts to ensure the quality of data submissions consistent with DATA Act standards and steps they have taken to address those challenges. To address these objectives, GAO interviewed staff at four federal SSPs, OMB, and Treasury; reviewed selected agreements between the SSPs and their customer agencies; conducted a survey of customer agencies from December 2018 to January 2019; and analyzed the survey responses. What GAO Found GAO found that the 27 agencies that responded to its survey use federal shared service providers (SSP) for a variety of services, including financial system hosting, general ledger accounting, financial reporting, and various Digital Accountability and Transparency Act of 2014 (DATA Act) services. Sixteen of the 27 SSP customer agencies reported that they experienced challenges associated with using an SSP, many of which affected the timeliness, completeness, or accuracy of agency DATA Act submissions. Ten of these agencies experienced challenges with depending on an SSP to take actions before the agency could proceed. Agencies responding to GAO's survey also reported other challenges, such as a lack of guidance from the Office of Management and Budget (OMB) and the Department of the Treasury (Treasury), limited customer agency and SSP resources, SSP errors affecting data quality, and inadequate SSP project management activities. Twelve of these 16 agencies stated that they are taking steps to address these challenges—such as increasing communication with their SSPs, making technology improvements, and performing manual work-arounds to reconcile and correct data files. Nine agencies reported remaining additional steps, for example, correcting data errors and developing a reconciliation process and internal guidance on topics such as data quality plans. While agencies are primarily responsible for the quality of DATA Act submissions, five agencies also reported that their SSPs had taken similar steps to address identified challenges. Twenty of the 27 agencies described useful practices for working with SSPs on DATA Act submissions, including the agency discussing issues with the SSP and obtaining data files from the SSP each month to provide additional time to correct any identified errors. Treasury officials stated prior to GAO's survey that they held workshops for SSPs in the early stages of DATA Act implementation and clarified guidance issued in June 2018 to specifically address their concerns and questions. After GAO's survey, in April 2019, OMB issued a memorandum on shared services that among other things described the process and desired outcomes for shared services and established a governance and accountability model for achieving them.

gao_GAO-19-498

gao_GAO-19-498_0

Background Prescription Drug Supply Chain Several entities are involved with, and pay different prices for, prescription drugs as they move from the manufacturer to the beneficiary (a system referred to as the prescription drug supply chain). In general, manufacturers develop and sell their drugs to wholesalers, and wholesalers then sell the drugs to pharmacies. In the Part D program, CMS pays Part D plan sponsors to provide drug coverage, and plan sponsors may charge beneficiaries monthly premiums in exchange for coverage. Plan sponsors and PBMs negotiate reimbursement rates for the drugs provided to beneficiaries. When the beneficiary purchases a drug, the pharmacy is paid by the Part D plan sponsor, or through the PBM on the sponsor’s behalf, and by the beneficiary through any applicable cost-sharing. (See fig. 1 for a flow chart showing the relationship between certain entities in the prescription drug supply chain when a Part D plan sponsor uses a PBM.) Prescription Drug Plan Services Services associated with developing and managing a prescription drug plan performed by PBMs, Part D plan sponsors, or both, include: Formulary development. Determining the list of drugs covered under the plan (the formulary), including assignment of covered drugs to tiers that correspond to different levels of beneficiary cost sharing and placing restrictions on drugs included in the formulary. Part D plan sponsors submit formularies for their plans to CMS for review and approval annually. Pharmacy network development. Creating a network of pharmacies where beneficiaries may fill their prescriptions and negotiating drug prices and reimbursement rates with those pharmacies. This can also include developing “preferred networks,” whereby beneficiaries pay lower cost-sharing and pharmacies agree to receive lower prices for drugs in exchange for increased volume of prescriptions purchased. Utilization management services. Utilization management services include processes such as: Prior authorization. A requirement that beneficiaries obtain approval for a drug by the PBM or plan sponsor before obtaining the drug if it is to be covered by the plan. Step therapy. A requirement where more expensive drugs are covered only if beneficiaries try less expensive alternatives first and find them not to be effective. Medication therapy management. A program required by CMS designed to improve medication adherence and reduce the risk of adverse drug events through discussion with targeted beneficiaries and prescriber intervention. Drug utilization review. A concurrent examination by the PBM or plan sponsor of prescriptions at the time of purchase by the beneficiary to assess safety considerations, such as potential adverse interactions, and compliance with clinical guidelines (including quantity and dose). These reviews can also occur retrospectively to analyze beneficiaries’ drug utilization and physicians’ prescribing patterns. Negotiation of rebates from manufacturers. Negotiating rebates for Part D plan sponsors with manufacturers in exchange for driving more utilization of a manufacturer’s drug. This can include more favorable placement on the sponsor’s formulary. The rebate terms do not have to be disclosed to the public, but plan sponsors must report rebate amounts to CMS. PBM Revenue PBMs may earn revenue from providing drug benefit management services to Part D plan sponsors in a number of ways, including: (1) payments from plan sponsors for administering services, such as drug benefits claim processing; (2) retention of a portion of drug rebates that PBMs negotiate on behalf of the plan sponsor and fees for managing and distributing those rebates; (3) spread pricing; and (4) payments from manufacturers for various services. PBMs may provide drug benefit management services to Part D plan sponsors and commercial plans, such as employer-sponsored health plans. Commercial plans may pay PBMs in ways similar to Part D plans (e.g., rebate retention and claims processing fees). Part D Coverage and Payments Part D plan sponsors are also required to provide access to all or substantially all drugs covered under certain therapeutic classes of drugs, known as Medicare protected classes: (1) anticonvulsants, (2) antidepressants, (3) antineoplastics, (4) antipsychotics, (5) antiretrovirals, and (6) immunosuppressants for the treatment of transplant. Plans are limited in the formulary restrictions they can apply to these drugs. Additionally, CMS generally requires Part D plan sponsors to provide coverage for at least two drugs in each class. CMS makes payments prospectively to Part D plan sponsors for beneficiary drug coverage. CMS pays plan sponsors monthly, and these payments are determined through annual bids submitted in June of the preceding program year, which runs from January 1 through December 31. Those bids reflect the plan sponsors’ estimates of program costs and rebates and other price concessions that the sponsor expects to receive during the ensuing program year. At the end of the program year, CMS reviews cost data submitted by plan sponsors through PDE records and their submission of rebate and other price concession data and compares estimated payments with actual costs incurred, with CMS either reclaiming some funds or making additional payments. Thus, the final plan payments by CMS are based on the costs actually incurred by Part D plan sponsors minus rebates and other price concessions that are either passed along to the plan sponsors or retained by the PBMs. Implications of Rebates and Other Price Concessions Rebates and other price concessions reduce the cost of the Part D program to beneficiaries and the federal government. In developing their bids, Part D plan sponsors may subtract rebates and other price concessions that are passed along to them from their estimated drug costs. When they do, rebates and other price concessions reduce a plan sponsor’s estimate of liability that is reflected in bid amounts, which, in turn, reduce beneficiary premiums because they are based, in part, on the bid amount. This downward pressure on premiums is one reason that premiums remained relatively unchanged between 2010 and 2015, according to CMS, even though total gross Part D drug costs grew about 12 percent per year in that period. Rebates have additional implications for Part D beneficiaries and the Part D program more generally. Since beneficiary cost sharing is calculated based on the price of the drug at the time of purchase (i.e., before rebates are paid), beneficiaries pay higher cost sharing than they would if rebates were paid at the point of sale. In addition, higher pre-rebate drug prices may result in beneficiaries more quickly reaching the catastrophic coverage phase, where the federal government’s share of drug costs increases, and the plan sponsors’ share decreases. Part D Plan Sponsors Used a PBM to Provide Most Drug Benefit Management Services in 2016, and Use Was Concentrated among Five PBMs Seventy-four percent of the drug benefit management services provided under 624 Part D plan sponsor contracts were performed by a PBM alone or in conjunction with a Part D plan sponsor in 2016. We found that plan sponsors performed the remaining 26 percent of services themselves. In addition, a PBM was used to provide one or more of the 10 key drug benefit management services under nearly all of the 624 Part D plan sponsor contracts (99.7 percent), and the manner in which they used them varied, as summarized below: Number of drug benefit management services provided. Part D plan sponsor contracts varied by the number of services provided by PBMs. Eighty-nine percent of Part D plan sponsor contracts used a PBM alone or in conjunction with a plan sponsor for at least half of the 10 drug benefit management services; 15 percent of contracts used a PBM alone or with a plan sponsor for all 10 services. Number of PBMs used. Part D plan sponsor contracts varied in the number of PBMs used to provide one or more of the 10 drug benefit management services. Fifty-four percent of contracts used one PBM, 35 percent used two or three PBMs, and 11 percent used four or more PBMs. Types of drug benefit management services provided. Part D plan sponsor contracts varied by the drug benefit management services they used a PBM to provide. PBMs alone or with the plan sponsor more frequently provided claims adjudication (99 percent of Part D plan sponsor contracts), pharmacy network development (92 percent), and rebate and other price concession negotiations (83 percent). In contrast, PBMs alone or with the plan sponsor less frequently provided a pharmacy and therapeutics committee (45 percent), enrollee appeals and grievance process-management (30 percent), and enrollment processing (34 percent). Part D plan sponsors mainly used five PBMs in 2016. Of the 103 PBMs that provided at least one drug benefit management service to the 624 Part D plan sponsor contracts in 2016, the following five provided at least one service to 528 (85 percent) plan sponsor contracts in 2016: CVS Caremark, OptumRx, Express Scripts, Medimpact, and Argus. These five PBMs also provided the largest number of services to Part D plan contracts in 2016. For example, CVS Caremark, by itself or with another PBM or plan sponsor, provided 17 percent of services that PBMs provided to Part D plan sponsors’ contracts in 2016, the most of any PBM. See appendix II for more information on variation in Part D plan sponsor contracts’ use of PBMs, factors that influence sponsors’ decision to use a PBM, and additional information on the PBMs used by Part D plan sponsors. PBMs Primarily Earned Part D Revenue through Fees Paid by Plan Sponsors, Not Rebate Retention, and Reported That This Differed from PBMs’ Commercial Plan Revenue Our review of 20 service agreements between Part D plan sponsors and PBMs found that the primary revenue source for PBMs from services they provided to Part D plans was (1) a volume-based fee paid by plan sponsors based on the number of paid claims that the PBM processed; (2) a flat monthly per-member, per-month fee paid by plan sponsors; or (3) a combination of the two. Nineteen of the 20 service agreements that we reviewed stated that PBMs were to be paid in one of these ways. None of the service agreements tied these fees to the price of a drug paid to the pharmacy. Representatives we interviewed from all seven of the PBMs confirmed that a Part D plan sponsor-paid fee for the PBM’s services was the primary way they earned revenue from their Part D clients. We also examined PBM revenue reported to CMS by Part D plan sponsors in their rebates and other price concession data—also referred to as direct and indirect remuneration (DIR)—in 2016, the most recent data available at the time of our analysis. These data show that PBMs passed nearly all rebates received from manufacturers through to Part D plan sponsors in 2016. Part D plan sponsors reported to CMS that, of the approximately $18 billion in rebates that PBMs negotiated with pharmaceutical manufacturers that year, PBMs retained $74.3 million, or about 0.4 percent, and passed through the remaining 99.6 percent to plan sponsors. The small amount of PBM rebate retention in the Part D program was also reflected in the service agreements we examined and in our interviews with PBM representatives. Sixteen of the 20 service agreements that we reviewed included provisions that required the PBM to pass through all rebates to the Part D plan sponsor; one other agreement required at least 95 percent to be passed through to the plan sponsor. The other three service agreements that we reviewed either did not include provisions related to rebate retention or redacted such information. Officials we interviewed from four of the seven PBMs told us their PBMs passed through to Part D plan sponsors all rebates obtained from manufacturers. Representatives of one PBM noted that plan sponsors, in turn, may use rebates to help offset the growth in drug costs, helping lower premiums for beneficiaries. Representatives from the other three PBMs noted that the amount of retained rebates was relatively small, consistent with the data reported to CMS. PBMs and Part D plan sponsors may earn non-rebate revenue from manufacturers for providing certain services. The service agreements we examined included examples of this revenue, including fees for rebate program administration, prescriber education programs, and programs designed to ensure patients adhere to, and comply with, recommendations regarding a particular prescription. The full amount that PBMs and Part D plan sponsors earned from manufacturers for non- rebate services in 2016 was $516.5 million. Although CMS requires these fees to be reported to the agency by plan sponsors, CMS does not break out how much of the money was received by PBMs and how much was received by plan sponsors. PBMs earned little Part D revenue from spread pricing—keeping the difference between the amount the PBM paid the pharmacy for a drug and the amount the PBM charged the plan for the drug, from 2014 through 2016. PBMs earned about $300,000 from spread pricing in 2016, according to CMS rebate and other price concession data. CMS data also show that PBMs earned no revenue from spread pricing in either 2014 or 2015. PBMs generally earn more from spread pricing and rebate retention from commercial plans than they do from Part D, according to officials from three PBMs. Officials from two of these PBMs said CMS reporting requirements have removed much of the incentive in Part D for PBMs to earn revenue from spread pricing because of the complexity of the requirements and the criticism from health care providers when reports to CMS containing these amounts are publicized. See appendix III for more information on Part D plan sponsor reporting to CMS of the amounts of revenue—other than rebates and discounts—that manufacturers provide to their PBMs; and on PBM and Part D plan sponsor perspectives on PBM revenue earned from spread pricing, the effect of CMS requirements on spread pricing revenue, and differences between PBMs’ Part D and commercial business lines. Rebates and Other Price Concessions Grew Faster Than Part D Expenditures from 2014 through 2016 Growth in the amount of rebates and other price concessions provided by manufacturers and others to Part D plan sponsors and PBMs outpaced growth in gross and net Part D expenditures for all brand-name and generic drugs from 2014 through 2016. Gross expenditures reflect what was paid to the pharmacy by the Part D plan sponsor—or the PBM on the sponsor’s behalf—and by the beneficiary for a given drug. Net expenditures reflect any rebates and discounts obtained by plan sponsors and PBMs after a beneficiary receives a drug. During this time, gross Part D expenditures increased 20 percent, from $120.7 billion in 2014 to $145.1 billion in 2016. The amount of rebates and other price concessions obtained for these drugs increased 66 percent during the same period, from $17.5 billion to $29 billion. As a result, rebates and other price concessions as a proportion of gross expenditures increased from 14 percent of gross expenditures in 2014 to 20 percent in 2016. This resulted in an increase in net Part D expenditures of 13 percent, from $103.2 billion in 2014 to $116.1 billion in 2016 (see fig. 2). Rebates accounted for most of the total of rebates and other price concessions obtained for Part D drugs from 2014 through 2016. Rebates are generally paid by manufacturers to Part D plan sponsors, or PBMs on sponsors’ behalf, after a drug is purchased from a pharmacy. In 2016, rebates accounted for 92 percent ($27 billion) of the $29.1 billion in rebates and other price concessions. The proportion was generally consistent in 2014 and 2015, with rebates accounting for 93 and 91 percent of total rebates and other price concessions, respectively. Pharmacy-related price concessions, which include any monies obtained by plan sponsors and PBMs from a pharmacy after a beneficiary purchases a drug, accounted for nearly all the rest of rebates and other price concessions—7 percent—in 2016. The amount of pharmacy- related price concessions increased 295 percent from 2014 through 2016 ($538 million to $2.1 billion). The 444 highest expenditure, highest utilization brand-name drugs accounted for the majority of expenditures and received the vast majority of rebates and other price concessions in 2016. These drugs accounted for 65 percent of the $145 billion in Part D expenditures and received 90 percent of the $29.1 billion in rebates and other price concessions obtained for Part D drugs. Of the 444 highest expenditure, highest utilization brand-name drugs in 2016, the 200 highest utilization and the 200 highest expenditure drugs received a greater amount of rebates and other price concessions than the 200 highest expenditure per utilization drugs. (See table 1.) Furthermore we found that brand-name drugs received greater amounts of rebates and other price concessions than generic drugs. Specifically, among the 444 highest expenditure, highest utilization brand-name drugs and the 476 highest expenditure, highest utilization generic drugs, brand-name drugs received 98 percent of rebates and other price concessions in 2016. Consistent with the results for all Part D drugs, from 2014 through 2016 rebates and other price concessions outpaced growth in gross and net expenditures for the three groups of highest expenditure, highest utilization brand-name drugs in our analysis (see table 2 for information on these brand-name drugs). The three groups of brand-name drugs generally had higher percent changes in rebates and other prices concessions and in gross and net expenditures than did all Part D drugs, which includes generics. For example, from 2014 through 2016, net expenditures for the 200 highest expenditure brand-name drugs increased 27 percent compared to a 13 percent increase for all Part D drugs. Of the three groups, the 200 drugs with the highest expenditure per utilization had the largest percentage increases in expenditures and rebates and other price concessions. However, these drugs had relatively low gross expenditures, rebates and other price concessions, and utilization compared with the other two groups. Increases in expenditures for the three groups of drugs in our analysis were primarily accounted for by increases in the price per drug rather than changes in utilization, as indicated by the growth in expenditures exceeded growth in their utilization. Net expenditures per beneficiary were similar if a Part D plan sponsor used a PBM for rebate negotiations or if it conducted its own negotiations. Specifically, in 2016, median net expenditures per enrollee were similar for plan sponsors using a PBM and those that did not at $2,557 and $2,570, respectively. Rebates and other price concessions accounted for a median of 12 percent of gross Part D expenditures for plan sponsors using a PBM for their negotiations and a median of 10 percent for plan sponsors that did not. The majority—82 percent—of plan sponsors used a PBM to obtain rebates and other price concessions on their behalf. The plan sponsors that performed their own negotiations generally had higher enrollment than those that used a PBM—a median of approximately 47,000 beneficiaries, compared to approximately 13,000 beneficiaries (see table 3). See appendix IV for additional information on expenditures and rebates and other price concessions obtained for the 444 highest expenditure, highest utilization brand-name Part D drugs in 2016. The appendix also contains information on expenditures and rebates and other price concessions obtained by the Part D plan sponsors whose representatives we interviewed. Part D Drug Prices Were Significantly Lower Than List Prices for Brand- Name Drugs in Retail Pharmacies; Drugs Sold in Specialty Pharmacies Received Fewer Discounts In 2016, the highest expenditure, highest utilization brand-name drugs sold in retail pharmacies received discounts off of manufacturer list prices that were significantly higher than those sold in specialty pharmacies. Of the 444 highest expenditure, highest utilization brand-drugs in our analysis, 244 were sold in retail pharmacies. For this group, gross Part D prices—those paid to the pharmacy by the Part D plan sponsor, PBMs on the sponsor’s behalf, and the beneficiary—were 17 percent lower than manufacturer list prices for these drugs. When rebates and other price concessions were accounted for, net Part D prices were 41 percent lower than manufacturer list prices. In contrast, the 200 drugs sold in specialty pharmacies received fewer discounts off of manufacturer list prices. For these drugs, median gross and net prices were 15 percent and 16 percent, respectively, lower than manufacturer list prices (see fig. 3). As a result, drugs sold in retail pharmacies received median discounts (41 percent) that were 2.5 times larger than those sold in specialty pharmacies (16 percent). See appendix V for more information on prices for the highest expenditure, highest utilization brand-name drugs and for information on prices for selected generic drugs. Utilization Management Was Generally Associated with Financial Savings and Improved Health Indicators, but Its Effect on Medication Adherence and Access Was Less Clear Our review of 52 peer-reviewed studies indicates that utilization management services were associated with financial savings or improved beneficiary health indicators. However, the effects on ensuring that beneficiaries take their medication as prescribed (adherence) and access to clinically appropriate prescriptions were less clear. The studies examined the effects of 10 different types of utilization management services in three areas: (1) financial savings; (2) beneficiary health indicators; and (3) beneficiary medication adherence and access: Financial savings. Twenty-seven of the 36 studies we reviewed that examined financial savings found that utilization management services were associated with savings for the Medicare program, Part D plans, or beneficiaries. For example, all eight studies that examined the relationship between generic substitution and financial savings found savings. Of the 10 studies that did not find financial savings, five found no statistically significant impact of the utilization management service on savings, three found the utilization management service was associated with a decrease in savings, and two found both an increase and decrease in savings for different types of utilization management services. Beneficiary health indicators. Twelve of the 20 studies that examined beneficiary health indicators found that utilization management services were associated with improvement, such as a reduction in adverse drug events. Ten of the 12 studies that found improvement examined either medication therapy management programs or comprehensive medication reviews. The other two studies that found improvement looked at drug utilization reviews, which examine a beneficiary’s prescriptions to identify safety considerations, such as potential adverse interactions with other drugs and compliance with clinical guidelines. Of the eight studies that found no improvement, one found that a health indicator worsened, and four found improvement in at least one health indicator and a decline in at least one other indicator. Beneficiary medication adherence and access. Of the 15 studies that examined the effect of utilization management services on beneficiaries’ medication adherence or access to clinically appropriate drugs, 10 examined medication therapy management programs or comprehensive medication reviews. Seven of these 10 found improvement in medication adherence. In contrast, the other five studies that examined adherence and access found negative, mixed, or no effects associated with prior authorization and step therapy. For example, two studies examined the effect of prior authorization and step therapy and found that these utilization management services resulted in increased access problems. Two other studies examined the relationship of prior authorization and step therapy adherence and found a mixed impact. The remaining study examined the relationship of only prior authorization with the time needed to access medications and found no clinically significant impact. Stakeholders we interviewed generally agreed that utilization management services resulted in financial savings but differed in their views regarding the effect of utilization management services on beneficiaries’ medication adherence and access to clinically appropriate drugs. In interviews with representatives from PBMs, Part D plan sponsors, and a manufacturer trade association, these stakeholders generally agreed that utilization management services resulted in financial savings. While representatives from most Part D plan sponsors and PBMs told us that utilization management services have resulted in no adverse impact on medication adherence and access to prescriptions, representatives of the three drug manufacturers we interviewed told us that utilization management services limit medication adherence and access to medications by, for example, delaying therapy to needed drugs. See appendix VI for more information about the effects of utilization management services from the peer-reviewed studies we examined and the stakeholders we interviewed. See appendix VII for the articles included in our literature review. Agency Comments The Department of Health and Human Services provided technical comments on a draft copy of this report, which GAO incorporated as appropriate. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the appropriate congressional committees and the Secretary of Health and Human Services. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or dickenj@dickenj@gao.gov.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix VIII. Appendix I: Objectives, Scope, and Methodology This appendix provides details on our scope and methodology in addressing each of our five reporting objectives: (1) the extent to which Part D plan sponsors contract with pharmacy benefit managers (PBM) to deliver drug benefit management services to Medicare beneficiaries; (2) how PBMs earn revenue from the services they provide to Part D plan sponsors; (3) trends in rebates and other price concessions obtained by Part D plan sponsors and PBMs from manufacturers and others for Part D drugs; (4) the extent to which prices for Part D drugs are discounted off of manufacturer list prices; and (5) what is known about savings and other effects from utilization management services commonly used in Part D. In addition, the appendix describes the steps we took to assure the reliability of the data we analyzed. Interviews For all our objectives, we obtained the perspectives of stakeholders on Part D plan sponsors’ use of PBMs as well as information on sponsors’ efforts to control Part D expenditures and drug utilization. We spoke to representatives from 17 small, mid-sized, and large Part D plan sponsors: Aetna, Anthem, Banner Health, Cambia Health, Cigna, CVS, Express Scripts, Kaiser, Health Care Service Corp, Health Plan of San Mateo, Henry Ford Health System, Humana, Missouri Highways and Transportation Commission, Rite Aid, United Health Care, University of Pittsburgh Medical Center, and WellCare. We spoke with seven PBMs: Argus, CVS Caremark, EnvisionRx, Express Scripts, MedImpact, Prime Therapeutics, and OptumRx. To obtain other drug industry perspectives, we spoke with representatives from three drug manufacturers: Eli Lilly, Gilead, and Amgen. We also spoke with one entity that is both a wholesaler and pharmacy services administrative organization: AmerisourceBergen. Additionally, we spoke with other industry and advocacy organizations, including groups representing drug manufacturers, Part D plan sponsors, pharmacies, and PBMs: America’s Health Insurance Plans, Biotechnology Innovation Organization, Community Oncology Alliance, National Association of Chain Drug Stores, National Association of Specialty Pharmacies, National Community Pharmacists Association, Patients for Affordable Drugs, Pharmacy Benefit Management Institute, Pharmaceutical Care Management Association, and Pharmaceutical Research and Manufacturers of America. The Extent to Which Part D Plan Sponsors Contract with PBMs to Deliver Drug Benefit Management Services to Beneficiaries To determine the extent to which PBMs provided services to Part D plan sponsors, we analyzed the Centers for Medicare & Medicaid Services’ (CMS) Health Plan Management System (HPMS) data that identified the entity or entities responsible for performing each of 10 key drug benefit management services for plan sponsors’ Part D contracts in 2016, the most recent available expenditure and rebate and other price concession data at the time of our analysis. CMS provided HPMS data for the 624 Part D plan sponsor contracts that were effective in 2016. The data contained the entity or entities reported by each plan sponsor as performing each service. Using this information, we identified for each contract whether the plan sponsor performed a service itself; contracted with a PBM to perform the service; or performed the service in coordination with a PBM. For a given contract, we counted as being a PBM any entity that was not the plan sponsor that performed one or more drug benefit management services. We manually reviewed those PBMs against a list of PBM members from a PBM trade organization. We used internet searches to confirm the entity was not the plan sponsor in instances when it was not listed in the trade organization’s member directory. In doing so, we also identified whether the plan sponsor shared common ownership with the PBM responsible for providing the drug benefit management service. For example, there were instances where the plan sponsor and PBM were sister organizations owned by the same parent company. In this situation, we counted the PBM as a separate entity from the plan. In addition, we analyzed PBM use by plan sponsor contract enrollment size using CMS contract enrollment information from June 2016. Additionally, we used HPMS data to examine plan sponsor contracts’ variation in the number of PBMs used, the types of services that PBMs provided, and the use of PBMs by contract enrollment size. We also identified the PBMs that provided the most services and described the services they provided. Last, we interviewed Part D plan sponsor representatives to understand the considerations that influenced their decision about how and whether to use a PBM. How PBMs Earn Revenue from the Services They Provide to Part D Plan Sponsors To determine how PBMs earned revenue from services they provide to Part D plan sponsors, we relied on four information sources. First, we reviewed selected service agreements between PBMs and Part D plan sponsors. The service agreements generally contain detailed information on the services that the PBM will provide, how the plan sponsor will pay the PBM for those services, and the rates that pharmacies will be paid for Part D drugs. We asked CMS for a list of all service agreements it approved between January 2016 and May 2018 that were in effect as of June 2018. CMS provided us with a list of 119 service agreements. Using June 2018 Part D publicly available enrollment data from CMS, we obtained from CMS the 20 service agreements for Part D plans sponsors with the largest enrollment in June 2018. While most of the service agreements included sufficient information to determine how the PBMs were paid, some did not, and, where appropriate, we noted these instances in our findings. Second, we examined PBM revenue reported to CMS by Part D plan sponsors in their rebates and other price concession data—also referred to as direct and indirect remuneration (DIR)—for 2014, 2015, and 2016. These rebate and other price concession submissions contain information on the various sources of revenue and expenses incurred by PBMs and plan sponsors. Third, we reviewed applicable CMS regulations and guidance on the reporting of PBM and Part D plan sponsor revenue and expenses. Fourth, we interviewed PBM representatives about the extent to which PBMs retained rebates or passed them through to plan sponsors and, in some cases, the reasons for this decision. We also asked certain PBM representatives whether their revenue sources for Part D, specifically rebate retention and spread pricing, differed from PBMs’ and plan sponsors’ commercial business and, if so, the reasons for any differences. Rebates and Other Price Concessions Obtained by Part D Plan Sponsors and PBMs from Manufacturers and Others for Part D Drugs To examine rebates and other price concessions obtained by Part D plan sponsors and PBMs from manufacturers and others for Part D drugs, relative to overall Part D expenditures, we analyzed plan sponsors’ gross and net expenditures for Part D drugs for 2014 through 2016, the most recent data available at the time of our analysis. Gross expenditures reflect what was paid to the pharmacy by the plan sponsor, PBMs on the sponsor’s behalf, and the beneficiary for a given drug. Net expenditures reflect any rebates and other price concessions obtained by Part D plan sponsors and PBMs after a beneficiary receives a drug. To calculate gross expenditures, we used Medicare prescription drug event (PDE) data to calculate gross brand-name and generic drug expenditure and utilization information for all Part D plan sponsors’ contracts. We used Red Book, a compendium published by Truven Health Analytics, to determine whether drugs were brand-name or generic. We then identified individual brand-name and generic drugs by grouping expenditure claims with the same active ingredient, strength, dosage form, and route of administration (known as ISDR). We calculated brand- name and generic drug expenditures based on a drug’s ingredient cost, dispensing fees, sales tax, and applicable vaccine administration fees. We used PDE data to calculate gross expenditures for all Part D plan sponsors at both the contract and plan sponsor level. We used DIR data to determine the amount of rebates and other price concessions and subtracted this amount from this data to calculate net expenditures. We also obtained plan sponsor enrollment data using publicly available CMS data for June 2016, which allowed us to calculate gross per beneficiary expenditures. We also examined differences in the amount of rebate and other price concessions obtained relative to expenditures for Part D plan sponsors that used a PBM relative to those that did not. We determined PBM involvement in rebate and other price concession negotiations for individual plan sponsors using 2016 HPMS data. We specifically looked at each entity listed in HMPS as negotiating rebates and other price concessions with drug manufacturers and others. We were able to determine whether a PBM or plan sponsor performed this service for 197 plans sponsors. However, there were 20 Part D plan sponsors where a PBM or plan was not solely listed as performing the rebate and other price concession service. In these instances, we could not identify which entity negotiated rebates and other price concessions and therefore excluded them from this analysis. To obtain more information on drugs that have the greatest fiscal impact on the Part D program and beneficiaries, we calculated gross and net expenditures for the brand-name and generic drugs with the highest expenditures, highest utilization, and highest expenditure per utilization in 2016. For both brand-name and generic drugs, we identified the following: the 200 brand-name and 200 generic drugs with the highest expenditures in 2016; the 200 brand-name and generic drugs with the highest utilization in 2016 (based on number of 30-day prescriptions); and the 200 brand-name and generic drugs with the highest expenditures per utilization (i.e., highest expenditure per number of 30-day prescriptions). As a result of overlap in the groups of drugs, these criteria yielded two groups: the 444 highest expenditure, highest utilization brand-name drugs and the 476 unique highest expenditure, highest utilization generic drugs. These 920 drugs accounted for 81 percent of total Part D expenditures in 2016. We used drug-level rebate and other price concessions data to calculate net drug prices for these drugs by subtracting rebate and other price concessions for each drug from gross expenditures. The Extent to Which Prices for Part D Drugs Are Discounted Off of Manufacturer List Prices To determine the extent to which Part D drug prices are discounted off of manufacturer list prices, we compared the median gross and net prices for the 444 brand-name and 476 generic highest expenditure, highest utilization drugs to (1) list prices established by manufacturers, and (2) the cost to pharmacies of acquiring these drugs. For list prices, we used 2016 average wholesale price (AWP) data from Truven Health Analytics’ Red Book. AWP is a common benchmark drug price used in the negotiation of payment rates between Part D plan sponsors and pharmacies. Because AWP is updated on an ongoing basis, we calculated a day-weighted per unit price that takes into account the number of days that the reported price was in effect in 2016. We then determined the median AWP price for each drug product based on the ISDR. We refer to the median price as the manufacturer list price. For pharmacy acquisition costs, which reflect the price pharmacies paid to obtain the drug, we used retail community pharmacy acquisition cost data from National Average Drug Acquisition Cost (NADAC) data. NADAC does not contain data from non-retail pharmacies, such as mail- order or specialty pharmacies. For our groups of 444 brand-name and 476 generic drugs, we separated drugs sold in retail community pharmacies from those sold in specialty pharmacies. If a drug did not have pharmacy acquisition cost data from NADAC, we considered that drug to be sold in specialty pharmacies and, thus, a specialty drug. We used 2016 PDE data to determine the gross per unit Part D price for a drug by dividing the gross expenditures for the drug by the total quantity dispensed of it. For example, a drug that had 1,000 units prescribed to Medicare beneficiaries and $5,000 in gross expenditures would have a gross per unit price of $5. We determined net per unit Part D prices for the drugs in our two study groups by dividing the amount of rebates and other price concessions for each drug by the quantity dispensed of it and then subtracting the amount of rebates and other price concessions per quantity from the gross Part D price for each drug. For each drug, we then determined the median pharmacy acquisition cost (if available), median gross Part D price, and median net Part D price as a proportion of median manufacturer list price by dividing each price by the median manufacturer list price. We then reported the median value for these pricing points for the highest expenditure, highest utilization drugs in our analysis. Analysis of Literature on Effect of Utilization Management Services To determine what is known about the impact of utilization management services that PBMs commonly provide to Part D plan sponsors, or that plan sponsors may perform themselves, we conducted a literature search for studies that examined the effect of utilization management services in Part D (regardless of whether they were provided by a PBM or another entity) on the following outcomes: (1) financial costs or savings, (2) beneficiaries’ health indicators, and (3) beneficiaries’ access to clinically appropriate medications or taking their medications as prescribed (adherence). The literature search was performed from April 2018 to July 2018 using keyword searches in bibliographic databases, including ProQuest, EBSCO, and Scopus. We limited our search to studies published beginning in 2006—the year the Part D program began. For our searches, we developed a list of search terms for our literature review by reviewing relevant background documentation and several database searches. The search terms included: “utilization management,” “prior authorization,” “quantity limits,” “step therapy,” “generic substitution,” “drug utilization review,” “quantity edit,” “medication therapy management,” and “comprehensive medication review,” combined with “access,” “adherence,” “health benefit,” “clinical outcome,” “generic use,” “cost effectiveness,” “savings,” “costs,” and “Medicare.” The literature search generated 700 studies. We reviewed this list by examining the abstracts for those studies that addressed the effects of utilization management services in Part D and were published in peer- reviewed journals. We identified 48 studies that met our criteria then added four more that met the criteria from several literature reviews we examined, resulting in a final group of 52 peer-reviewed studies that we analyzed. We analyzed these studies to group them by type of utilization management service evaluated and type of outcome measured. We documented any methodological limitations of these studies but did not exclude any of them on this basis. See the bibliography in Appendix VII for a list of the 52 studies in our review. We also interviewed PBMs, plan sponsors, and drug manufacturers to obtain their views regarding the impact of utilization management services in Part D plans and asked them to recommend additional studies on utilization management services. We did not assess the methodology or data reliability of the studies provided to us by these drug supply chain stakeholders; none of them met our criterion of being published in peer- reviewed journals. We used these studies to better understand stakeholder perspectives. Data Reliability To ensure the data used to produce this report were sufficiently reliable, we took several steps. We performed data reliability checks on the HPMS data by reviewing the data for missing values and errors, checking the information against other publicly available sources, and interviewing knowledgeable agency officials. We performed data reliability checks on the PDE and DIR data by reviewing relevant documentation, checking the data for outliers and errors, and interviewing knowledgeable agency officials. We performed data reliability checks of the AWP and NADAC data sets by testing the data for missing data and outliers and reviewing relevant documentation. After taking these steps, we determined the data were sufficiently reliable for the purposes of our reporting objectives. We conducted this performance audit from May 2017 to July 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings based on our audit objectives. Appendix II: Medicare Part D Plan Sponsors’ Use of Pharmacy Benefit Managers (PBM) Appendix II: Medicare Part D Plan Sponsors’ Use of Pharmacy Benefit Managers (PBM) This appendix provides additional detail on the use of PBMs by Part D plan sponsors to provide prescription drug benefit management services to Medicare beneficiaries. Part D Plan Sponsors’ Use of PBMs We examined Centers for Medicare & Medicaid Services’ (CMS) data to identify the 10 key drug benefit management services provided by PBMs under 624 Part D plan sponsor contracts in 2016, the most recent available expenditure and rebate and other price concession data at the time of our analysis, and found the following variation in plan sponsor use of PBMs: Services provided by PBMs. Part D plan sponsors’ contracts varied by the services provided by PBMs in 2016. Plan sponsors’ use of a PBM for drug benefit services—either alone or with the plan sponsor—for their 624 contracts varied from 30 percent for enrollee appeals and grievance process-management to 99 percent for claims adjudication. For seven of the 10 drug benefit management services, PBMs—either alone or in conjunction with the plan sponsor—provided services to more than half the sponsor contracts (see fig. 4). Number of PBMs used. Part D plan sponsor contracts varied in the number of PBMs used to provide one or more of the 10 drug benefit management services. For example, 54 percent of plan sponsors’ contracts used a single PBM, while 11 percent used four or more PBMs (see fig. 5). Use of PBMs by enrollment. Smaller Part D contracts—those with contract enrollment below the median enrollment of all Part D contracts—used a PBM more often than larger contracts—those with enrollment at or above the median. For instance, 87 percent of smaller Part D plan sponsor contracts used a PBM alone or with the plan sponsor for rebate and price concession negotiations, compared to 77 percent of larger contracts. Similarly, 54 percent of smaller Part D contracts used a PBM alone or with a plan for a pharmacy and therapeutics committee, compared to 35 percent of larger contracts. Use of financially related PBMs. Part D plan sponsors’ contracts varied by their use of PBMs with which they were related by common ownership—either as a subsidiary or a sister company. In 2016, plan sponsors used a PBM with which they were related by common ownership for 17 percent of the 624 Part D plan sponsors contracts. Larger contracts—those with enrollment at or above the median— were more likely to use a PBM related by common ownership than smaller contracts. Larger contracts used a financially related PBM for 24 percent of drug benefit management services, compared to 10 percent of drug benefit management services provided to smaller contracts. Factors Influencing Plans’ Decisions to Use a PBM The Part D plan sponsor representatives with whom we spoke noted several considerations that influenced their decision about how and whether to use a PBM. One plan sponsor noted that small plans may lack the resources to conduct their own rebate negotiations and, therefore, may use a PBM instead. Three other plan sponsors noted they switched from conducting their own rebate negotiation with manufacturers to using a PBM. Two plan sponsors said this switch was due to PBMs’ ability to obtain larger rebates than the plan sponsor could, and the third determined a PBM would help it achieve the best value and quality, while meeting Part D’s regulatory requirements. In contrast, representatives of three other Part D plan sponsors noted advantages of performing drug benefit management services themselves. For example, one plan sponsor noted that it performs almost all drug benefit management services internally, as it believes doing so improves quality through better communication and care coordination with pharmacies. Another plan sponsor noted the decision not to contract out certain services to a PBM may be influenced by a desire for more customization over formulary management and greater control over prior authorization. Representatives of one plan sponsor noted that their plan does not use a PBM because they believe they are more effective in developing formularies with better utilization management and greater use of generic drugs than are PBMs. Variation in the Number of Services Provided by PBMs to Part D Plan Sponsors Our analysis of CMS data for the 624 Part D plan sponsor contracts found that the five PBMs that provided the largest number of services to Part D plan sponsors’ contracts in 2016 also generally provided a full range of PBM services to them. Four of the top five PBMs provided all 10 drug benefit management services to plan sponsors’ contracts while the fifth PBM conducted claims adjudication but used an intermediary to conduct rebate negotiations. (See table 4). Furthermore, the top five PBMs provided a high proportion of the services that Part D plan sponsors most commonly used a PBM to provide. For example, CVS Caremark provided claims adjudication to 144 (23 percent) of Part D plan sponsor contracts, and OptumRx provided this service to 138 (22 percent). In contrast, we found that Part D plan sponsors used a large number of PBMs to provide a limited range of drug benefit management services. For example, 48 percent of PBMs provided only one type of drug benefit management service to plan sponsors’ contracts, and 22 percent of PBMs provided only one service to only one plan sponsor contract. For instance, there were 10 unique entities counted as PBMs in our analysis that provided only customer service support to one plan sponsor contract. One PBM representative noted in an interview that it is relatively common for plan sponsors and PBMs to contract with other vendors to provide additional assistance with drug benefit management services. One plan sponsor told us, for example, that its PBM uses a vendor to manage customer service calls. Appendix III: Information on Pharmacy Benefit Manager (PBM) Revenue Earned from Manufacturers and from Spread Pricing This appendix provides additional detail on (1) non-rebate revenue that PBMs may earn for services provided to manufacturers and Medicare Part D plan sponsors, and (2) PBM perspectives on Centers for Medicare & Medicaid Services (CMS) policies relating to spread pricing in Part D. Non-rebate Revenue That PBMs May Earn for Services Provided to Manufacturers and Part D Plan Sponsors PBMs and Part D plan sponsors may earn non-rebate revenue from manufacturers for providing certain services. Even though this money is reported to CMS as part of the rebate and other price concession submission, not all of it is considered rebates or other price concessions, which will lower plan liability in determining bids and thereby lower premiums. Of the $516.5 million in non-rebate revenue paid by manufacturers in 2016, $440 million, or about 85 percent, represented the amount paid for the services that exceeded the fair market value of the service and is considered rebates and other price concessions. These may be used to reduce the drug costs incurred by the plan sponsor. Therefore, this revenue factors into bid determinations and may be used to reduce premiums. The remaining $78.6 million in payments from manufacturers were considered “bona fide service fees”—fees paid by manufacturers to Part D plan sponsors and PBMs for services that the manufacturer would otherwise perform, or contract for, and that represented the fair market value of those services. Such fees do not reduce the plan sponsor’s drug costs and, therefore, could not factor into reducing premiums. The determination of a bona fide service fee as reported to CMS is made by the drug manufacturer and the Part D plan sponsor and is not routinely evaluated by CMS, agency officials told us. However, CMS requires that the PBM and manufacturer have information documenting the fair market value of the service. Stakeholder Perspectives on PBM Revenue Earned from Spread Pricing, Rebate Retention, and Differences from the Commercial Sector CMS requires Part D plan sponsors to report revenue earned from rebates retained by the PBM. This revenue increases the plan’s liability, which increases the amount of plan bids and, therefore, result in higher premiums. In contrast, rebate revenue passed through by PBMs to Part D plan sponsors lowers the plan’s liability, reduces plans bids, and, therefore, lowers beneficiary premiums. Some PBMs earn more revenue from spread pricing in their commercial business than in Part D, officials from three PBMs told us. Officials from two of these PBMs noted that CMS requirements create a disincentive to engage in spread pricing that is not present in the commercial sector. Beginning in 2010, CMS required that plan sponsors base the amount of beneficiary cost-sharing on the amount received by the pharmacy for a drug—known as the “pass-through price.” CMS also required that an estimate of rebates or other price concessions be included in the administrative costs submitted by the plan sponsor for bid determinations. Part D plan sponsors can still agree to pay the PBM based on the higher price of the drug without accounting for rebates, known as the lock-in price. However, the difference between that amount and the pass-through price would increase the bid determination and ultimately increase the premiums that plans charge beneficiaries. Because there are no similar requirements pertaining to the commercial prescription drug benefit market, spread pricing is more common there, CMS officials told us. Appendix IV: Expenditures and Rebate and Other Price Concession Information for Medicare Part D Drugs This appendix provides information on (1) pharmacy-related price concessions for all Medicare Part D drugs and (2) expenditure and rebate and other price concession information for the 444 highest expenditure, highest utilization brand-name Part D drugs in 2016. The appendix also contains additional information on expenditures and rebates and other price concessions obtained by the 16 Part D plan sponsors whose representatives we interviewed. Pharmacy-Related Price Concessions for All Part D Drugs The amount of pharmacy-related price concessions obtained by Part D plan sponsors, or pharmacy benefit managers (PBM) on plan sponsors’ behalf, increased 295 percent from 2014 through 2016, from $538 million to $2.1 billion (see fig. 6). These monies account for any adjustments to the price of the drug paid to the pharmacy after the point sale, such as a pharmacy returning money that was overpaid by the plan sponsor or vice versa. It can also include monies paid based on pharmacies’ performance in meeting agreed-upon performance metrics—for example, fees a pharmacy pay plan sponsors, or bonuses pharmacies receive from plan sponsors, based on their performance. In 2016, Part D plan sponsors received $2.3 billion from pharmacies and paid out $211 million, for a net of $2.1 billion in pharmacy-related price concessions. Five of the seven PBMs and seven of the 12 Part D plan sponsors whose representatives we interviewed said they have performance-based arrangements with pharmacies. One plan sponsor noted that its performance agreement involves paying bonuses to pharmacies that exceed performance measures, while charging fees to pharmacies that did not meet the measures. The sponsor said this is part of an attempt to move from paying for volume to paying for value. Another plan sponsor told us there has been an improvement in pharmacy performance as a result of the program. Representatives from pharmacy industry groups said these pharmacy- related fees have put increasing pressure on pharmacies. For example, one group noted there is no standardization across measures with each plan sponsor using its own measures, and it is difficult for pharmacies to tie a fee to a specific pharmacy location or claim. Another group noted that fees may be imposed on pharmacies for performance measures not directly applicable to the pharmacy. For example, the group said specialty pharmacies have been assessed fees for beneficiary lack of adherence to maintenance medications, such as blood pressure medications, that these pharmacies do not commonly provide. Expenditure and Rebate and Other Price Concession Information for 444 Highest Expenditure, Highest Utilization Brand-Name Drugs PBMs and Part D plan sponsors obtained rebates and other price concessions for 441 (99 percent) of the 444 highest-expenditure, highest- utilization brand-name drugs in 2016. The amount of rebates and other price concessions for each drug ranged from $1,300 to $1.8 billion in 2016, with a median of $3.3 million. Rebates accounted for $24.5 billion of the $26 billion in rebates and other price concessions (94 percent) obtained by plan sponsors and PBMs for these 444 drugs. As a proportion of gross Part D expenditures—the amount paid by plan sponsors, or the PBM on the sponsors’ behalf, and by beneficiaries—for the 444 drugs ranged from -0.5 percent to 70.5 percent. (See fig. 7.) Expenditures and Rebates and Other Price Concessions for the 444 Highest Expenditure, Highest Utilization Part D Brand-Name Drugs, by Therapeutic Drug Class, 2016 Expenditures and rebates and other price concessions varied by therapeutic class for the 444 highest expenditure, highest utilization drugs in 2016. Among those with 10 or more drugs in their class, gross expenditures ranged from $2.9 billion to $21.2 billion, and rebates and other price concessions ranged from $170 million to $8.7 billion (see table 5). Four classes—endocrine metabolic agents, anti-infective agents, respiratory agents, and central nervous system agents—accounted for 54 percent of the gross Part D expenditures, and 62 percent of rebates and other price concessions for the 444 highest expenditure, highest utilization drugs. When accounting for rebates and other price concessions, these drugs accounted for 51 percent of net Part D expenditures. Gross and Net Part D Expenditures Varied among Selected Part D Plan Sponsors in 2016 Rebates and other price concessions as a proportion of gross expenditures varied from 4 percent to 27 percent in 2016 for the 17 Part D plan sponsors whose representatives we interviewed. Gross Part D expenditures per beneficiary ranged from $1,772 to $5,583, and net Part D expenditures per beneficiary ranged from $1,687 to $4,837 (see table 6). Appendix V: Information on Discounts Off Manufacturer List Prices for Brand-Name and Generic Medicare Part D Drugs This appendix contains additional information on the gross and net discounts for the highest expenditure, highest utilization brand-name and generic Medicare Part D drugs in 2016. Information on the Extent to Which Brand-Name Part D Drugs Were Discounted Off Manufacturer List Prices The amount of discounts in 2016 for the 444 highest expenditure, highest utilization brand-name drugs varied by whether they were sold in retail or specialty pharmacies. Discounts also varied by whether the brand-name drugs were highest expenditure, highest utilization or highest expenditure per utilization drugs. Of the 444 highest expenditure, highest utilization brand-name drugs, 244 were sold in retail pharmacies and 200 were sold in specialty pharmacies. Brand-name retail drugs. The three groups of drugs all had pharmacy acquisition costs that were 81 percent of manufacturer list prices and gross Part D prices that were between 83 and 84 percent of manufacturer list prices in 2016. However, the net prices varied, ranging from 55 percent of manufacturer list price for the highest utilization drugs to 77 percent for the highest expenditure per utilization drugs (see table 7). Brand-name specialty drugs. The 38 highest expenditure drugs and 187 highest expenditure per utilization drugs sold in specialty pharmacies had median gross prices that were between 84 and 85 percent of manufacturer list price and net prices that were 84 percent of manufacturer list price in 2016. We also found variation in brand-name prices across therapeutic classes for the 244 highest expenditure, highest utilization Part D drugs sold in retail pharmacies. In 2016, median gross Part D prices for the brand- name drugs sold in retail pharmacies were similar across the nine therapeutic classes we analyzed, ranging from 81 percent to 84 percent of the manufacturer list price. However, there was a much wider range among median net prices, from 43 percent to 83 percent of manufacturer list price. Anti-infective agents had the lowest percentage point changes in their prices from gross to net (1 percentage point), while endocrine metabolic agents, cardiovascular agents, respiratory agents, ophthalmologic agents, and genitourinary agents had the largest changes, with declines from gross to net of greater than 30 or more percentage points (see table 8). In contrast, there was little variation in both median gross and net prices across all therapeutic classes for brand-name drugs sold in specialty pharmacies. The range in median gross prices as a proportion of manufacturer list prices across the six therapeutic classes was 83 percent to 86 percent, and the range in median net prices as a proportion of manufacturer list prices was 80 percent to 84 percent. Information on the Extent to Which Generic Part D Drugs Were Discounted Off Manufacturer List Prices In 2016, discounts off of the manufacturer list price varied by whether the generic drug was sold in retail pharmacies or in specialty pharmacies. Of the 476 highest expenditure, highest utilization generic drugs in our analysis, the 367 sold in retail pharmacies had a median gross and net Part D price that were 66 percentage points lower than the manufacturer list price, and 13 percentage points higher than the pharmacy’s cost of acquiring the drugs. The 109 generic drugs sold in specialty pharmacies received far fewer discounts off of manufacturer list price than drugs sold in retail pharmacies. Median gross and net prices for those drugs sold in specialty pharmacies were both 26 percentage points lower than manufacturer list prices (see fig. 8). Therefore, generic drugs sold in retail pharmacies received median discounts (66 percent below manufacturer list prices) that were 2.5 times larger than those generic drugs sold in specialty pharmacies (26 percent below manufacturer list prices). We also found pricing variation by whether the generic drugs were in the 200 highest expenditure, 200 highest utilization group, or the 200 highest expenditures per utilization group. Generic retail drugs. Of the 367 generic drugs sold in retail pharmacies, 200 were in the group of the 200 highest utilization generic drugs, 198 were in the group of the 200 highest expenditure generic drugs, and 91 were in the group of the 200 generic drugs with the highest expenditure per utilization. We found that the gross Part D price for the highest utilization drugs was 14 percent of the manufacturer list price, while the gross price for the highest expenditure drugs was 34 percent of the manufacturer list price. However, the Part D gross price for the highest expenditure per utilization drugs was 63 percent of the manufacturer list price. The difference in gross and net Part D price as a percentage of manufacturer list price was one percentage point or less for all three groups of drugs (see table 9). Generic specialty drugs. Of the 109 generic drugs sold in specialty pharmacies, none was in the group of the 200 highest utilization generic drugs, two were in the group of the 200 highest expenditure generic drugs, and all 109 were in the group of the 200 highest expenditure per utilization generic drugs. The gross Part D price for the highest expenditure per utilization drugs sold in specialty pharmacies was 74 percent of the manufacturer list price, and these drugs received no additional rebates and other price concessions. There was variation in generic drug pricing across the eight therapeutic classes for generic drugs sold in retail pharmacies. Median gross Part D prices for generic retail drugs ranged from 14 percent of manufacturer list prices for cardiovascular agents to 56 percent of manufacturer list prices for dermatological agents (see table 10). However, there was little difference between in median gross and net Part D prices as a percentage of manufacturer list price for generic retail drugs in any therapeutic class, with the percentage difference ranging from 0 percent to 2 percent. There was little variation in median gross and net prices across the therapeutic classes for generic drugs sold in specialty pharmacies. The range in median gross prices as a percentage of manufacturer list prices was 73 to 75 percent (see table 11). There was little difference between median gross and Part D net prices as a percentage of manufacturer list price, with the percentage difference between median gross and net prices 1 percent or less for all classes. Appendix VI: Studies and Stakeholders’ Views on Effects of Utilization Management Services This appendix contains additional details on our review of 52 peer- reviewed studies on the effects of utilization management services on (1) financial savings, (2) beneficiary health indicators, and (3) beneficiary medication adherence and access, as well as stakeholders’ views on these effects. Effect of Utilization Management Services on Financial Savings Of the 36 studies that examined the effect of utilization management services on financial savings, 18 examined medication therapy management programs and eight examined generic substitution. The two groups of studies found the following: Medication therapy management programs or comprehensive medical reviews. Thirteen of the 18 studies that examined the relationship between a medication therapy management program or comprehensive medical review and financial savings found an increase in savings. For example, one study found that a medication therapy management program conducted by telephone decreased beneficiary drug costs by $682 per beneficiary for participants, compared to an increase of $119 for those not in the program. Generic and therapeutic substitution and generic dispensing rate. Of the 8 studies that examined the relationship between generic and therapeutic substitution and financial savings, all found an increase in savings. For example, a 2013 study examined the potential financial savings to beneficiaries and Part D plan sponsors of generic and therapeutic substitution of commonly prescribed drugs. The study estimated that in 2007, generic and therapeutic substitutions could have resulted in an average annual savings of $127 and $389 per person, respectively. Additionally, eight of these 36 studies examined the generic dispensing rate, and all eight found that utilization management led to an increase in the rate. The generic dispensing rate—the percent of prescriptions dispensed with a generic drug instead of a brand-name drug—represents a source of financial savings through a reduction in the use of brand- name drugs, which are generally more expensive than generics. For example, a 2017 study analyzed 2012 Part D data to examine the impact of prior authorization and step therapy on generic use among low-income subsidy beneficiaries. This study found that those randomly assigned to a plan using both prior authorization and step therapy had an increased generic dispensing rate of 3 to 15 percentage points for all three classes of drugs examined. Effect of Utilization Management Services on Beneficiary Health Indicators Twelve of the 20 studies that examined beneficiary health indicators found that utilization management services were associated with improved indicators, while the other eight found a mixed impact, no impact, or a decline. Examples of studies that looked at the association of utilization management services with beneficiary health indicators include: A study analyzing data from three Part D plan sponsors, which found there was a nearly 50 percent reduction in the use of potentially harmful drugs by beneficiaries 6 months after the implementation of a retrospective drug utilization review program. A randomized trial of medication therapy management for Part D beneficiaries found a nearly 60 percent reduction in beneficiaries’ drug therapy problems over time among two groups after the medication therapy management intervention. Effect of Utilization Management Services on Beneficiary Medication Adherence and Access Fifteen studies examined the effect of utilization management services on beneficiary medication adherence and access. Seven of the 10 studies that examined the effect of either medication therapy management programs or comprehensive medication reviews on beneficiaries’ medication adherence (taking medication as prescribed) found improvement. For example, a 2016 study used data from Part D and the U.S. Renal Data System to examine the relationship of medication therapy management eligibility with immunosuppressant drug adherence 12 months after beneficiaries received a kidney transplant. The study found that medication therapy management-eligible transplant recipients were 14 percent more likely to have improved adherence than transplant recipients who were not eligible. The other three studies that examined medication therapy management programs or comprehensive medication reviews found no statistically significant impact on adherence. The effect of two other utilization management services—prior authorization and step therapy—on beneficiary medication adherence and access (the ability to obtain clinically indicated prescriptions) is unclear, according to the studies we reviewed. The two studies that examined the relationship of prior authorization and step therapy with adherence both found a mixed impact. For example, one study examined the impact of a health plan requiring either prior authorization or step therapy on medication use among dual-eligible nursing home residents. The study found that some residents whose new plan required prior authorization or step therapy for their current medication were more likely to have gaps in medication use than those without for two of six classes of drugs in 2006, but no gaps for any of the classes for in 2007 and 2008. The two studies that examined the relationship of prior authorization and step therapy with access found an increase in medication access problems, but they did not focus exclusively on the Medicare population. For example, one study used 2006 data from a random sample of psychiatrists surveyed about their patients to examine the relationship of prior authorization and step therapy with medication access problems among dual-eligible psychiatric patients. The study found that patients in plans with prior authorization and step therapy requirements were 2.8 and 1.8 times more likely, respectively, to have experienced medication access problems than patients in plans without these requirements. This study examined the transition of dual-eligible beneficiaries from Medicaid drug coverage to Medicare Part D when the program began in 2006, so the results may not be generalizable to the entire Medicare population at present. Stakeholder Perspectives on the Effect of Utilization Management on Financial Savings, Beneficiary Health, Medication Adherence and Access to Clinically Appropriate Medications Most representatives of pharmacy benefit managers (PBM), Part D plan sponsors, and a manufacturer trade association we interviewed generally agreed that utilization management services resulted in financial savings by requiring the use of generic drugs. Representatives of 10 of 14 plan sponsors and six of eight PBMs we interviewed stated that utilization management services generally resulted in financial savings. Representatives of one Part D plan sponsor stated that its utilization management services resulted in annual savings of approximately 3 percent. However, representatives of one Part D plan sponsor and one PBM noted that not all utilization management services result in savings. For example, they noted that improving care with medication therapy management programs may increase drug costs through increased utilization. Additionally, representatives of one Part D plan sponsor noted the savings from utilization management services in commercial plans may be greater than in Part D because the use of manufacturers’ copay coupons are prohibited in federal health care programs, including Part D. While the coupons reduce or eliminate beneficiaries’ out-of-pocket co-payments for certain brand-name drugs, thereby encouraging their use, the coupons do not affect the amount that the plans pay for drugs. Therefore, to the extent that beneficiaries in their commercial plans use coupons, Part D plan sponsors have a greater incentive to employ utilization management services in these plans to reduce the use of more expensive brand-name drugs. Representatives of Part D plan sponsors and PBMs we interviewed differed with manufacturers and, in some cases, with each other on the effects of utilization management services on various non-financial aspects of drug utilization: Beneficiary health. Representatives from all three manufacturers we interviewed stated that utilization management services negatively affected beneficiary health by reducing their access to necessary medications. In contrast, seven of the 11 Part D plan sponsors and four of the five PBMs that discussed the effect of utilization management services on beneficiary health stated that utilization management services generally resulted in improved beneficiary health. Representatives of certain PBMs and one Part D plan sponsor provided us examples of the ways utilization management services have improved their beneficiaries’ health, such as through opioid quantity limits. One Part D plan sponsor noted that point-of-sale utilization management services warn pharmacies of therapeutic duplications, toxicities across multiple prescriptions, or interactions of certain drugs with health conditions. Medication access. Representatives from all three drug manufacturers noted that utilization management services impose limits on beneficiaries’ access to drugs, while seven of nine Part D plan sponsors and three of the four PBMs who discussed this stated utilization management services had no significant restrictions on beneficiaries’ access to necessary medications. Representatives from one plan sponsor noted there are appeals processes to ensure beneficiaries’ access is not adversely impacted by utilization management services. Medication adherence. Representatives from all three manufacturers told us that utilization management services limit beneficiaries’ adherence to their medications, such as by causing delays in therapy, while seven of eight Part D plan sponsors and all four PBMs who discussed this stated utilization management services had no adverse impact on beneficiaries’ adherence to their medications. Representatives from one plan sponsor and two PBMs stated that utilization management services may have a positive impact on adherence, such as by lowering copays through generic substitution. Medicare protected classes and utilization management. Representatives from Part D plan sponsors, PBMs, and manufacturers differed in their views on the effect of Part D utilization management services restrictions on protected class drugs on beneficiary health. Representatives of two PBMs told us the effect was positive, as beneficiaries who use these drugs do not experience disruptions in therapy. Representatives of two other PBMs said there was no effect, and one said there was a negative effect—as plan sponsors were required to cover certain less effective drugs. Representatives of one PBM said that, for example, patients in commercial health plans do not have any problems accessing protected class drugs that are subject to utilization management. These representatives noted that the Centers for Medicare & Medicaid Services provides for adequate access. However, one manufacturer told us that utilization management services for HIV drugs are rightly restricted in Part D, as these services may cause disruptions in therapy, which can lead to drug resistance and poorer health outcomes. Representatives of five Part D plan sponsors said Medicare’s restrictions on the use of utilization management services for protected class drugs have had a negative impact on beneficiary health because, for example, they limit plans’ ability to ensure that a prescribed drug is appropriate, such as ensuring that a cancer drug is appropriate for a beneficiary’s weight. Another plan sponsor representative told us the restrictions may have a positive impact by reducing increases in medical costs, while another plan sponsor said the restrictions have had no impact. Appendix VII: Bibliography of Peer Reviewed Studies Used in GAO’s Literature Review Abbass, I. M., E. O. Caplan, D. B. Ng, R. Kristy, C. R. Schermer, P. Bradt, J. M. Collins, W. Man, M. Chan, and B. T. Suehs. “Impact of Overactive Bladder Step Therapy Policies on Medication Utilization and Expenditures among Treated Medicare Members.” Journal of Managed Care & Specialty Pharmacy, vol. 23, no. 1 (2017): 27-37. Agarwal, A., R. A. Freedman, F. Goicuria, C. Rhinehart, K. Murphy, E. Kelly, E. Mullaney, M. St. Amand, P. Nguyen, and N. U. Lin. “Prior Authorization for Medications in a Breast Oncology Practice: Navigation of a Complex Process.” Journal of Oncology Practice, vol. 13, no. 4 (2017): e273-e282. Almodovar, A. S., D. R. Axon, A. M. Coleman, T. Warholak, and M. C. Nahata. “The Effect of Plan Type and Comprehensive Medication Reviews on High-Risk Medication Use.” Journal of Managed Care & Specialty Pharmacy, vol. 24, no. 5 (2018): 416-422. Alston, G., and C. Hanrahan. “Can a Pharmacist Reduce Annual Costs for Medicare Part D Enrollees?” The Consultant Pharmacist, vol. 26, no. 3 (2011): 182-189. Ben-Joseph, R., C.C. Chen, A. P. De, R.L. Wade, and D. Shah. “Consequences of Patient Access Restrictions to Branded Oxycodone Hydrochloride Extended-Release Tablets on Healthcare Utilization and Costs in US Health Plans.” Journal of Medical Economics, vol. 17, no. 10 (2014): 708-718. Bergeson, J. G., K. Worley, A. Louder, M. Ward, and J. Graham. “Retrospective Database Analysis of the Impact of Prior Authorization for Type 2 Diabetes Medications on Health Care Costs in a Medicare Advantage Prescription Drug Plan Population.” Journal of Managed Care Pharmacy, vol. 19, no. 5 (2013): 374-384. Bloudek, L. M., D. Makenbaeva, and M. Eaddy. “Anticipated Impact of Generic Imatinib Market Entry on the Costs of Tyrosine Kinase Inhibitors.” American Health & Drug Benefits, vol. 8, no. 9 (2015): 472-480. Branham, A., J. Moose, and S. Ferrari. “Retrospective Analysis of Medication Adherence and Cost Following Medication Therapy Management.” Innovation in Pharmacy, vol. 1, no. 1 (2010): 1-8. Branham, A. R., A. J. Katz, J. S. Moose, S. P. Ferreri, J. F. Farley, and M. W. Marciniak. “Retrospective Analysis of Estimated Cost Avoidance Following Pharmacist-Provided Medication Therapy Management Services.” Journal of Pharmacy Practice, vol. 26, no. 4 (2013): 420-427. Buhl, A., J. Augustine, A. M. Taylor, R. Martin, and T. L. Warholak. “Positive Medication Changes Resulting from Comprehensive and Noncomprehensive Medication Reviews in a Medicare Part D Population.” Journal of Managed Care & Specialty Pharmacy, vol. 23, no. 3 (2017): 388-394. Caffiero, N., T. Delate, M. D. Ehizuelen, and K. Vogel. “Effectiveness of a Clinical Pharmacist Medication Therapy Management Program in Discontinuation of Drugs to Avoid in the Elderly.” Journal of Managed Care & Specialty Pharmacy, vol. 23, no. 5 (2017): 525-531. Caplan, E. O., M. C. Guy, J. Chang, and K. Boesen. “Telephone-Based Cardiovascular Medication Therapy Management in Medicare Part D Enrollees with Diabetes.” The American Journal of Pharmacy Benefits, vol. 9, no. 2 (2017): 47-54. Chinthammit, C., E. P. Armstrong, K. Boesen, R. Martin, A. M. Taylor, and T. Warholak. “Cost-Effectiveness of Comprehensive Medication Reviews Versus Noncomprehensive Medication Review Interventions and Subsequent Successful Medication Changes in a Medicare Part D Population.” Journal of Managed Care & Specialty Pharmacy, vol. 21, no. 5 (2015): 381-389. Chisholm-Burns, M. A., C. A. Spivey, E. A. Tolley, and E. K. Kaplan. “Medication Therapy Management and Adherence among US Renal Transplant Recipients.” Patient Preference and Adherence, vol.10 (2016): 703-709. Corsi, K., V. Lemay, K. K. Orr, and L. Cohen. “Pharmacist Medication Therapy Management in Home Health Care: Investigation of a Sustainable Practice Model.” Journal of the American Pharmacists Association, vol. 58, no. 4S (2018): S64-S68. Dai, R. and J. Robst. “The Relationship between Plan Characteristics and Medicare Prescription Drug Plan Bids.” Applied Economics Letters, vol. 19, no. 1 (2012): 99-104. De Lott, L. B., J. F. Burke, K. A. Kerber, L. E. Skolarus, and B. C. Callaghan. “Medicare Part D Payments for Neurologist-Prescribed Drugs.” Neurology, vol. 86, no. 16 (2016): 1491-1498. Dodson, S. E., J. F. Ruisinger, P. A. Howard, S. E. Hare, and B. J. Barnes. “Community Pharmacy-Based Medication Therapy Management Services: Financial Impact for Patients.” Pharmacy Practice, vol. 10, no. 3 (2012): 119-124. Duru, O. K., S. L. Ettner, N. Turk, C. M. Mangione, A. F. Brown, J. Fu, L. Simien, and C.W. Tseng. “Potential Savings Associated with Drug Substitution in Medicare Part D: The Translating Research Into Action for Diabetes (TRIAD) Study.” Journal of General Internal Medicine, vol. 29, no. 1 (2014): 230-236. Egilman, A. C., J. D. Wallach, J. S. Ross, and S. S. Dhruva. “Medicare Spending and Potential Savings on Brand-Name Drugs with Available Generic Substitutes Excluded by 2 Large Pharmacy Benefit Managers, 2012 through 2015.” JAMA Internal Medicine, vol. 178, no. 4 (2018): 567- 569. Fox, D., L. D. Ried, G. E. Klein, W. Myers, and K. Foli. “A Medication Therapy Management Program’s Impact on Low-Density Lipoprotein Cholesterol Goal Attainment in Medicare Part D Patients with Diabetes.” Journal of the American Pharmacists Association, vol. 49, no. 2 (2009): 192-199. Gellad, W. F., J. M. Donohue, X. Zhao, M. K. Mor, C. T. Thorpe, J. Smith, C. B. Good, M. J. Fine, and N. E. Morden. “Brand-Name Prescription Drug Use among Veterans Affairs and Medicare Part D Patients with Diabetes: A National Cohort Comparison.” Annals of Internal Medicine, vol. 159, no. 2 (2013): 105-114. Gernant, S. A., M. E. Snyder, H. Jaynes, J. M. Sutherland, and A. J. Zillich. “The Effectiveness of Pharmacist-Provided Telephonic Medication Therapy Management on Emergency Department Utilization in Home Health Patients.” Journal of Pharmacy Technology, vol. 32, no. 5 (2016): 179-184. Gold, J. A., B. French, and L. C. Vermeulen. “Reduction of Use of Potentially Inappropriate Medications in the Elderly.” Wisconsin Medical Journal, vol. 107, no. 4 (2008): 213-214. Hoadley, J. F., K. Merrell, E. Hargrave, and L. Summer. “In Medicare Part D Plans, Low or Zero Copays and Other Features to Encourage the Use of Generic Statins Work, Could Save Billions.” Health Affairs, vol. 31, no. 10 (2012): 2266-75. Hui, R. L., B. D. Yamada, M. M. Spence, E. W. Jeong, and J. Chan. “Impact of a Medicare MTM Program: Evaluating Clinical and Economic Outcomes.” The American Journal of Managed Care, vol. 20, no. 2 (2014): e43-e51. Huskamp, H. A., D. G. Stevenson, A. J. O’Malley, S. B. Dusetzina, S. L. Mitchell, B. J. Zarowitz, M. E. Chernew, and J. P. Newhouse. “Medicare Part D Plan Generosity and Medication Use among Dual-Eligible Nursing Home Residents.” Medical Care, vol. 51, no. 10 (2013): 894-900. Moczygemba, L. R., J. C. Barner, J. C. Brannier, and E. R. Gabrillo. “Outcomes of a Medicare Part D Telephone Medication Therapy Management Program.” Journal of the American Pharmacists Association, vol. 52, no. 6 (2012): e144-e152. Moczygemba, L. R., J. C. Barner, K. A. Lawson, C. M. Brown, E. R. Gabrillo, P. Godley, and M. Johnsrud. “Impact of Telephone Medication Therapy Management on Medication and Health-Related Problems, Medication Adherence, and Medicare Part D Drug Costs: A 6-Month Follow Up.” The American Journal of Geriatric Pharmacotherapy, vol. 9, no. 5 (2011): 328-338. Moore, J. M., D. Shartle, L. Faudskar, O.S. Matlin, and T.A. Brennan. Impact of a Patient-Centered Pharmacy Program and Intervention in a High-Risk Group. Journal of Managed Care Pharmacy, vol. 19, no. 3 (2013): 228-236. Newman-Casey, P. A., M. A. Woodward, L. M. Niziol, P. P. Lee, and L. B. De Lott. “Brand Medications and Medicare Part D: How Eye Care Providers’ Prescribing Patterns Influence Costs.” Ophthalmology, vol. 125, no. 3 (2018): 332-339. Null, K. D., K. Moll, A. Sadosky, M. K. Pasquale, J. C. Cappelleri, and B. Parsons. “Trends Associated with Implementing and Lifting a Pregabalin Step Therapy Policy.” American Journal of Pharmacy Benefits, vol. 8, no. 2 (2016): e17-e24. Olvey, E. L., M. C. Guy, J. Chang, and G. H. Skrepnek. “Cost- Effectiveness of Medication Therapy Management in Part D Diabetic Enrollees.” American Journal of Pharmacy Benefits, vol. 6, no. 5 (2014): e147-e156. Pai, A.B., A. Boyd, J. Depczynski, I.M. Chavez, N. Khan, and H. Manley. Reduced Drug Use and Hospitalization Rates in Patients Undergoing Hemodialysis Who Received Pharmaceutical Care: A 2-Year, Randomized, Controlled Study. Pharmacotherapy, vol. 29, no. 12 (2009): 1433-1440. Patel, R. A., M. P. Walberg, E. Tong, F. Tan, A. E. Rummel, J. A. Woelfel, S. M. Carr-Lopez, and S. M. Galal. “Cost Variability of Suggested Generic Treatment Alternatives under the Medicare Part D Benefit.” Journal of Managed Care Pharmacy, vol. 20, no. 3 (2014): 283-290. Pindolia, V. K., L. Stebelsky, T. M. Romain, L. Luoma, S. N. Nowak, and F. Gillanders. “Mitigation of Medication Mishaps via Medication Therapy Management.” The Annals of Pharmacotherapy, vol. 43, no. 4 (2009): 611-620. Shen, X., B. C. Stuart, C. A. Powers, S. E. Tom, L. S. Magder, and E. M. Perfetto. “Impact of Formulary Restrictions on Medication Use and Costs.” The American Journal of Managed Care, vol. 23, no. 8 (2017): e265-e274. Starner, C. I., S. A. Norman, R. G. Reynolds, and P. P. Gleason. “Effect of a Retrospective Drug Utilization Review on Potentially Inappropriate Prescribing in the Elderly.” The American Journal of Geriatric Pharmacotherapy, vol. 7, no. 1 (2009): 11-19. Stebbins, M.R., D.J. Kaufman, and H.L. Lipton. “The PRICE Clinic for Low-Income Elderly: A Managed Care Model for Implementing Pharmacist-Directed Services.” Journal of Managed Care Pharmacy, vol. 11, no. 4 (2005): 333–341. Steele, K. M., J. F. Ruisinger, J. Bates, E. S. Prohaska, B. L. Melton, and S. Hipp. “Home-Based Comprehensive Medication Reviews: Pharmacist’s Impact on Drug Therapy Problems in Geriatric Patients.” The Consultant Pharmacist, vol. 31, no. 10 (2016): 598-605. Stockl, K. M., D. Tjioe, S. Gong, J. Stroup, A. S. M. Harada, and H. C. Lew. “Effect of an Intervention to Increase Statin Use in Medicare Members Who Qualified for a Medication Therapy Management Program.” Journal of Managed Care Pharmacy, vol. 14, no. 6 (2008): 532-540. Suehs, B. T., A. Louder, M. Udall, J. C. Cappelleri, A. V. Joshi, and N. C. Patel. 2014. “Impact of a Pregabalin Step Therapy Policy among Medicare Advantage Beneficiaries.” Pain Practice, vol. 14 (5): 419-426. Tang, Y., W. F. Gellad, A. Men, and J. M. Donohue. “Impact of Medicare Part D Plan Features on Use of Generic Drugs.” Medical Care, vol. 52, no. 6 (2014): 541-548. Thatcher, E. E., E. M. Vanwert, and S. R. Erickson. “Potential Impact of Pharmacist Interventions to Reduce Cost for Medicare Part D Beneficiaries.” Journal of Pharmacy Practice, vol. 26, no. 3 (2013): 248- 252. Touchette, D. R., A. L. Masica, R. J. Dolor, G. T. Schumock, Y. K. Choi, Y. Kim, and S. R. Smith. “Safety-Focused Medication Therapy Management: A Randomized Controlled Trial.” Journal of the American Pharmacists Association, vol. 52, no. 5 (2012): 603-612. Ward, M. A., and Y. Xu. “Pharmacist-Provided Telephonic Medication Therapy Management in an MAPD Plan.” The American Journal of Managed Care, vol. 17, no. 10 (2011): e399-e409. Welch, E. K., T. Delate, E. A. Chester, and T. Stubbings. “Assessment of the Impact of Medication Therapy Management Delivered to Home-Based Medicare Beneficiaries.” The Annals of Pharmacotherapy, vol. 43, no. 4 (2009): 603-610. West, J. C., J. E. Wilk, I. L. Muszynski, D. S. Rae, M. Rubio-Stipec, C. L. Alter, W. E. Narrow, and D. A. Regier. “Medication Access and Continuity: The Experiences of Dual-Eligible Psychiatric Patients during the First 4 Months of the Medicare Prescription Drug Benefit.” The American Journal of Psychiatry, vol. 164, no. 5 (2007): 789-796. West, J. C., J. E. Wilk, D. S. Rae, I. L. Muszynski, M. Rubio-Stipec, C. L. Alter, K. E. Sanders, S. Crystal, and D. A. Regier. “First-Year Medicare Part D Prescription Drug Benefits: Medication Access and Continuity among Dual Eligible Psychiatric Patients.” The Journal of Clinical Psychiatry, vol. 71, no. 4 (2010): 400-410. Winston, S., and Y.S. Lin. “Impact on Drug Cost and Use of Medicare Part D of Medication Therapy Management Services Delivered in 2007.” Journal of the American Pharmacists Association, vol. 49, no. 6 (2009): 813-820. Woelfel, J. A., S. M. Carr-Lopez, M. D. Santos, A. Bui, R. A. Patel, M. P. Walberg, and S. M. Galal. “Assessing Medicare Beneficiaries’ Willingness-to-Pay for Medication Therapy Management Services.” The Consultant Pharmacist, vol. 29, no. 2 (2014): 104-109. Zillich, A. J., M. E. Snyder, C. K. Frail, J. L. Lewis, D. Deshotels, P. Dunham, H. A. Jaynes, and J. M. Sutherland. “A Randomized, Controlled Pragmatic Trial of Telephonic Medication Therapy Management to Reduce Hospitalization in Home Health Patients.” Health Services Research, vol. 49, no. 5 (2014): 1537-1554. Appendix VIII: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgements In addition to the contact named above, Robert Copeland, Assistant Director; William A. Crafton, Analyst-in-Charge; Britt Carlson, Kaitlin Dunn, Andrew Emmons, Michael Rose, and Dan Ries made key contributions to this report. Also contributing were George Bogart, Yesook Merrill, Laurie Pachter, and Vikki Porter.

Why GAO Did This Study Total expenditures for the Medicare Part D drug program exceeded $100 billion in 2016. Part D plan sponsors may use a PBM to provide drug benefit management services for Part D coverage, such as negotiating drug rebates and other price concessions and paying pharmacy claims. Policymakers have sought a better understanding of PBMs' roles in the drug supply chain and plans' and PBMs' efforts to manage Part D drug spending and use. GAO was asked to examine the role of PBMs in the Part D program. This report examines, among other objectives, (1) the extent to which Part D plan sponsors use PBMs, (2) trends in rebates and other price concessions obtained by both PBMs and plan sponsors for Part D drugs, and (3) how PBMs earn revenue for services provided to Part D plans. GAO analyzed Centers for Medicare & Medicaid Services (CMS) data on Part D plan sponsors' use of PBMs in 2016 as well as CMS drug expenditure, pricing, and rebate and other price concession data for all Part D drugs from 2014 through 2016 (the most recent available data at the time of our analysis). GAO reviewed service agreements between Part D plan sponsors and PBMs that were approved by CMS from January 2016 through May 2018 and had the highest enrollment as of June 2018. GAO spoke with CMS officials and 38 stakeholder groups including PBMs, Part D plan sponsors, pharmacy representatives and drug manufacturers. What GAO Found Medicare Part D plan sponsors used pharmacy benefit managers (PBM) to provide 74 percent of drug benefit management services and performed the remaining 26 percent of services themselves in 2016—the most recent year of data at the time of our analysis. Plan sponsors are private entities that operate drug plans; PBMs are organizations that help manage drug benefits. Rebates and other price concessions—discounts generally paid by manufacturers to Part D plan sponsors and PBMs after the sale of a drug at the pharmacy—grew faster than Part D expenditures from 2014 through 2016. Specifically, gross expenditures (the amount paid to pharmacies by plan sponsors, or by the PBM on the sponsor's behalf, and by the beneficiary) increased 20 percent, to $145.1 billion. During this period, rebates and other price concessions increased 66 percent, to $29 billion—20 percent of 2016 gross expenditures. Consequently, net expenditures (gross expenditures less rebates and other price concessions) increased only 13 percent, to $116.1 billion. PBMs primarily earned Part D revenue through a volume-based fee paid by plan sponsors based on PBM-processed claims; a per-member, per-month fee paid by plan sponsors; or a combination of the two. PBMs also earned revenue from the rebates they negotiated with manufacturers for Part D drugs, which accounted for $18 billion of the $26.7 billion in rebates in 2016. PBMs retained less than 1 percent of these rebates, passing the rest to plan sponsors. Plan sponsors in turn may use rebates to help offset the growth in drug costs, helping control premiums for beneficiaries. The Department of Health and Human Services provided technical comments on a draft of this report, which GAO incorporated as appropriate.

gao_GAO-19-423

gao_GAO-19-423_0

Background Insurance allows individuals and businesses to manage risk by providing compensation for certain losses or expenses, such as those from car accidents, fires, medical services, or inability to work. According to NAIC, as of December 31, 2017, there were 2,509 property/casualty companies and 852 life insurance companies in the United States and its territories. In 2017, premiums written for the property/casualty sector totaled $602.2 billion in 2017 and premiums written for the life and health sector totaled $683.2 billion. As we have noted in recent reports, advances in technology and widespread use of the internet have brought about significant changes in the financial industry. For example, in recent years technology has changed consumer expectations and preferences, with younger consumers especially being well-versed in new technologies and looking to take a more hands-on approach to managing their finances. Similarly, over the last 5 years, established insurers and insurtech companies have used technology to offer simpler insurance products and streamlined customer experiences. Insurtech companies have been playing a variety of roles in the U.S. insurance market. Key players in insurtech include the following: Insurtech companies (typically startups) that are licensed insurance companies. Insurtech startups offer innovative products and services and are active in all major insurance products and all lines of business, with concentrations in the property/casualty business. For example, according to its website, Lemonade Insurance Company is a property/casualty insurer that sells products exclusively through mobile applications (apps) and its website. It offers renters, condominium, and homeowners insurance in several states. Another example is Root, which describes itself as an automobile insurance company that uses a smartphone app to understand individual driving behavior. Customers can download the Root app to their smartphones, obtain a personalized quote after a 2–3 week test drive, and purchase and manage their policy entirely within the mobile app. Insurtech companies that do not provide insurance themselves, but offer technology solutions for insurers. For example, according to the website for Groundspeed Analytics, they use AI and data science methods to provide information for the commercial property/casualty insurance industry to help identify potential areas of profit and enhance the customer experience. According to the website for Habit Analytics, they use real-time consumer data, sourced from smartphones and connected devices in homes, to create behavioral profiles that enable insurance companies to provide input for their risk models. Many established insurers have been acquiring such companies. Established insurers that use technologies or partner with insurtech companies. For example, the insurer Nationwide notes on its website that it created Nationwide Ventures to invest in startups, pilot new technologies, and test new solutions and business models by exploring topics that range from analytics and automation technology to new insurance and financial services platforms. According to analysis by the Deloitte Center for Financial Services and data collected by research firm Venture Scanner, as of mid-2018 there were more than 1,000 insurtech firms established in more than 60 countries, with more than half of those launched in the United States since 2008. State Licensing Regulation for Admitted and Nonadmitted Insurance Markets Insurance companies are regulated principally by the states and are licensed under the laws of a single state, known as the state of domicile. Companies may conduct business in multiple states, but the state of domicile serves as an important regulator. State regulators license insurance agents, generally review and approve insurance products and premium rates, and examine insurers’ financial solvency and market conduct. As we have previously reported, state regulators typically conduct financial solvency examinations every 3–5 years, while market conduct examinations are generally done in response to specific consumer complaints or regulatory concerns. To help ensure that policyholders continue to receive coverage if their insurer becomes insolvent or unable to meet its liabilities, states also have guaranty funds (separate for life and property/casualty insurance), which are funded by assessments on insurers doing business in those states. Individuals who wish to sell, solicit, or negotiate insurance in the United States must generally be licensed as producers, a term including insurance agents and insurance brokers. Insurance agents typically represent only one insurance company. Insurance brokers represent multiple insurance companies and are free to offer a wider range of products to their clients. Brokers can search the market and obtain multiple price quotes to fit their clients’ needs. Producers must comply with state laws and regulations governing their activities. NAIC notes that as of September 2018, more than 2 million individuals and more than 200,000 business entities were licensed to provide insurance services across all lines of insurance in the United States. Traditional insurers, sometimes referred to as admitted insurers, can be licensed to sell several lines or types of coverage to individuals or families, including personal lines—such as homeowners, renters, and automobile insurance—and commercial lines—such as general liability, commercial property, and product liability insurance. Admitted insurers can sell insurance in one or more states but, according to NAIC, must be licensed to operate in every state in which they sell coverage. To help ensure adequacy and fairness in pricing and coverage, state regulators oversee the insurance rates and forms of admitted insurers. State regulators also may require admitted insurance companies to maintain specific levels of capital to continue to conduct business. The surplus lines insurance market, also known as the nonadmitted market, can provide insurance coverage for risks that traditional insurers are unwilling or unable to cover. The risks covered can include potentially catastrophic property damage and liability associated with high-hazard products, special events, environmental impairment, and employment practices. In the absence of the surplus lines market, NAIC notes that some insureds in those markets would be unable to secure coverage. In most states, surplus lines insurers cannot write insurance coverage that is available from admitted insurers and only may write coverage rejected by a number of admitted insurers, according to NAIC. Furthermore, in those states, the surplus lines insurance broker must conduct a “diligent search” of the admitted insurance market to determine if comparable coverage is available. The broker can write coverage only if a specified number of admitted insurers have declined to offer such coverage. According to NAIC, new and innovative insurance products for which there is no loss history may be difficult to appropriately price. According to stakeholders we interviewed, the nonadmitted market is therefore a common entry point into the insurance market for insurtech firms that want to sell insurance products. NAIC notes that, after a new coverage has generated sufficient data, the coverage often eventually moves to, and is sold by, insurers in the admitted market. For example, private flood insurance was developed and first offered in the nonadmitted market but now also is offered in the admitted market. The nonadmitted market is generally regulated somewhat differently than the admitted market. According to NAIC, surplus lines insurers are subject to regulatory requirements and are overseen for solvency by their domiciliary state or country, but surplus lines transactions are regulated through the licensing of surplus lines brokers. NAIC states these brokers are responsible for ensuring that the surplus lines insurer meets eligibility criteria to write policies in the state and is financially sound. Furthermore, NAIC notes surplus lines brokers and producers must be licensed to sell surplus lines insurance in each state in which they operate. State insurance departments may have authority to suspend, revoke, or not renew the license of a surplus lines broker or producer. Unlike admitted insurers, surplus lines insurers may not have access to state guaranty funds that are available to help pay claims in the event of an insurer insolvency. In addition, according to NAIC, surplus lines insurers generally have more freedom to change policy coverages and premium rates than admitted insurers. NAIC stated that state regulators require both nonadmitted and admitted insurance companies to maintain specific levels of capital to continue to conduct business. According to NAIC, most state insurance regulators also can use their authorities under state statues such as an unfair trade practices act to ensure consumers are protected (for example, to ensure that claims are paid and insurers or brokers do not misrepresent policy terms) and to remedy other bad conduct. Other Participants in the Regulatory Framework for Insurance NAIC assists state regulators with various oversight functions. While NAIC does not regulate insurers, it provides services designed to make certain interactions between insurers and regulators more efficient. These services include providing detailed insurance data to help regulators understand insurance sales and practices; maintaining a range of databases useful to regulators; and coordinating regulatory efforts by providing guidance, model laws and regulations, and information-sharing tools. The Federal Insurance Office was established in the Department of the Treasury (Treasury) by the Dodd-Frank Wall Street Reform and Consumer Protection Act. The office is headed by a director appointed by the Secretary of the Treasury. The Federal Insurance Office monitors all aspects of the insurance industry (including by identifying issues or gaps in insurance regulation that could contribute to systemic risk in the insurance industry), and helps develop federal policy on international insurance matters, but is not a regulatory agency itself. The office also serves as an information resource for the federal government and coordinates with federal regulators, state insurance regulators, and NAIC. The Federal Insurance Office also represents the United States in the International Association of Insurance Supervisors and coordinates federal efforts in international insurance matters. Emerging Use of Technologies Can Reduce Insurance Costs and Expand Product Choices but Creates Privacy and Other Challenges In recent years, the insurance industry has begun to adopt several types of technology that are designed to provide a range of benefits to insurers and consumers (policyholders), including improved risk monitoring, reduced costs, and improved underwriting. However, the use of these technologies also can create challenges for insurers and potential risks for consumers, including changed business models, pricing fairness, and privacy issues. Insurance Industry Increasingly Using Mobile Apps, Big Data, and Other Technologies Based on our literature review and interviews with stakeholders, we identified six key technologies that have seen increased use in the insurance industry in recent years and one technology (blockchain) that has seen limited adoption and which the industry has been exploring for wider use. Mobile apps. A mobile app is software designed to run on a mobile device, such as a smartphone or tablet computer. Insurance industry stakeholders told us that several insurers have adopted mobile apps to make their products and services available on mobile devices. For example, insurers have adopted mobile apps that allow consumers to purchase products online. An increased number of insurers in recent years also have adopted mobile apps that allow customers to complete tasks online such as submitting insurance claims and turning on-demand insurance coverage on or off. Insurers also have been using mobile apps to capture consumer data and usage patterns (behaviors). AI, algorithms, and machine learning. AI is the development of computer systems to perform tasks and make decisions that historically have required human intelligence to perform. Machine learning is a subset of AI and focuses on the ability of machines to receive a set of data and learn for themselves, changing algorithms as they learn more about the information they process. (Algorithms are sets of rules that a computer or computer program follows to compute an outcome.) In the insurance industry, AI includes applications that provide specific expertise or allow for task completion. For example, AI provides on-line “chatbots” (sometimes called robo-advisory services) that answer questions specific to an insurance product or service. When a consumer communicates with a chatbot, the chatbot takes the information the consumer provided and enters it into an algorithm. Based on protocols outlined in the algorithm, the chatbot provides a response to the consumer’s question. As the conversation moves forward, the chatbot will adapt to answer more questions using machine learning in real-time. According to insurance industry stakeholders, insurers have been using algorithms to analyze information obtained from other technology sources to determine what a consumer’s risk profile is and then determine the consumer’s premium rate based on their risk profile. Big data. Big data are large volumes of data (often aggregated from multiple sources to develop data sets). As we have noted in other work, big data are frequently analyzed using predictive analytics, machine learning, and data mining to identify trends, patterns and characteristics. The insurance industry uses big data in several ways, including analyzing consumer information, identifying risk patterns and pricing risk, and analyzing information related to risk pooling. Insurers also use big data to streamline and more accurately underwrite products. For instance, an insurer may use big data to determine whether consumers are high- or low-risk based on factors identified from extensive datasets such as what they purchase online or how they shop for insurance online. This is similar to lenders’ usage of big data. In a previous report, we noted that lenders were using big data to evaluate risk and make lending decisions using real- time nontraditional information gathered from social media sites. Internet of things. The internet of things refers to semi-autonomous and internet-capable devices (such as machinery, home appliances, thermostats, and smartphones) that have sensors that interact with the physical environment and typically contain elements for processing and communicating information. Some insurers stated that the internet of things could be used in the insurance industry to track and reduce risk, detect problems, and mitigate potential claims. For example, a homeowner could have a smart home thermostat that sends alerts when the power goes off and indoor temperature decreases. With the homeowner able to address the issue in real time, the homeowner could mitigate the risk of frozen pipes bursting and potentially prevent a loss and an insurance claim. According to CBInsights, insurers have partnered with insurtech firms that provide this technology to offer real-time monitoring. Drones. Drones are remotely piloted aircraft systems. Insurers have been using drones for a variety of purposes in the insurance industry. For example, insurers use drones to obtain aerial footage over a disaster area to determine the amount of damage to a house or crop field. Insurance companies also use drones to verify information submitted by a policyholder in a claim or help determine the risk presented by difficult-to-reach areas of a property, such as a roof. Telematics. Telematics combines telecommunications and information processing to send, receive, and store information related to specific items such as automobiles or water heaters. Telematics often uses sensors to relay information such as global positioning system location, speed, and water levels. For example, sensors in an automobile can provide data on a driver’s behavior (such as speed, hard braking, and turning radius). The insurer may use that information to determine the driver’s risk profile and help determine the premium rate for that driver. These technologies can be used together. For example, a telematics device can be used to provide data to a mobile app, which can then send the information to an AI algorithm to determine whether a claim should be paid. See figure 1 for examples of the types of technologies that insurers may use to automate the claims process. Blockchain/ distributed ledger technology and smart contracts. The insurance industry has been studying whether blockchain technology could be used to improve insurance processes. Blockchain refers to a type of distributed ledger technology—in which multiple entities and locations share and synchronize datasets—that facilitates and permanently records virtual transactions. Information is uploaded and recorded in a series of secured blocks; the information uploaded cannot be modified or erased once uploaded into the blockchain (thus providing an accurate history of specific transactions and information). According to insurers, blockchain could be used by the industry to track insurance coverage history, expedite the claims process, provide an audit trail of insurance transactions, and address cybersecurity issues. For instance, a blockchain could expedite the claims process by allowing agents, policyholders, and repair companies immediate, secure access to certain data that are part of the claim only as the data are needed. “Smart contracts” include provisions for contract performance that can be executed by a computer algorithm (for instance, on a blockchain). For example, an insurer stated that a smart contract for homeowners insurance might stipulate that if an earthquake of a specific size occurred in a policyholder’s residential area, a claim payment for damage in a specified dollar amount automatically would be made from the insurer to the policyholder. According to NAIC, adoption of blockchain technology in insurance is limited at this time. Technologies Can Create Benefits but Also Present Risks to Insurers and Consumers According to stakeholders with whom we spoke and literature we reviewed, the use of technology in the insurance industry creates potential benefits but also can create risks for both insurers and consumers. We present stakeholder views on the benefits and challenges technology presents in the primary areas they identified as being affected by technology, which include (1) pricing and risk evaluation, (2) consumer protection, (3) business operations and risk monitoring, and (4) product offerings. See figure 2 for a summary of the potential benefits and challenges we discuss. According to stakeholders we interviewed and literature we reviewed, the use of technology for determining insurance pricing and coverages creates several benefits and risks for insurers and consumers: Increased underwriting accuracy. Insurers and others told us that insurers have been using technologies that provide enhanced analytic capabilities or data from previously unavailable sources to increase the accuracy of underwriting. These technologies allow insurers to make new connections between policyholder characteristics and risk. That is, insurers are using big data, AI, and algorithms to obtain and analyze more information about consumers than they previously had been able to obtain. For instance, a property/casualty insurer could collect data on when consumers set their home alarms and use this and other risk information to refine risk determinations for those individuals. Another example is when insurers use data collected from telematics devices in automobiles to inform the insurer about the policyholder’s risk of being involved in an accident. A better understanding of the risk presented by policyholders can help insurers more accurately and effectively price and manage risks. More individualized pricing. Insurers also have been using technologies to underwrite policies in a way that results in more individualized pricing, which benefits insurers and could benefit some consumers. That is, big data can allow an insurer to use factors for which traditional underwriting typically has not accounted. According to stakeholders we interviewed, doing so allows an insurer to place an individual in a smaller risk pool than if traditional underwriting factors were used and to price coverage for that individual more in line with the risk that individual presents. This can help an insurer better manage its level of risk by offering lower prices to lower-risk customers, charging more for higher-risk customers, or even declining to offer coverage to consumers it considers high-risk. Some stakeholders told us that technologies allow consumers to receive more individualized premium rates, based on their risk characteristics, than had been possible. For example, some insurers have been using telematics devices to obtain information on policyholder driving habits and the risk level they present and adjust premium rates based on this information. As a result, consumers who engage in safer driving practices receive the benefit of lower premiums. Policyholders also could use such information to take actions that will lower their risk level and therefore their premiums. For instance, consumers could seek to reduce specific driving behaviors, such as fast stops or starts, which negatively affect their premium rate. However, consumers with higher-than-average risks could end up paying more or perhaps be declined coverage. Stakeholders including an industry representative and a law firm in the field indicated that insurers also might use data to exclude high-risk consumers from marketing. For example, an insurer might not choose to market to high-risk consumers to discourage them from buying their insurance. This approach, in theory, helps insurers decrease the number of high-risk policyholders they insure but could create difficulties for some seeking coverage. Two industry representatives and an academic in the field indicated that the potential for decreased risk pooling creates a difficult question about the minimum extent of pooling that is socially desirable. For example, these stakeholders stated that when insurance underwriting becomes too individualized, it might no longer serve an insurance function; that is, there is very little pooling of risk. They stated it may be a desirable social benefit to have a certain level of risk pooling to allow more people to effectively manage their risk. In a November 2018 issue paper, the International Association of Insurance Supervisors noted the potential effect of more individualized underwriting on the fairness of consumer outcomes. Among other findings, the paper noted the collection of more data on policyholders may enable a more specific risk categorization that could affect risk pooling principles and lead to issues around affordability of certain insurance products or even availability (the potential for exclusion). The association noted that insurance supervisors should monitor whether such negative consumer impacts become a trend and, if so, raise awareness at the appropriate policy and political level(s). Validating consumer data and models. Insurers and insurtech firms increasingly have been using AI and data collection algorithms to gather data through mobile, wearable, and other internet-connected devices and from online sites. According to two academics in the field, collecting consumer data in large quantities and from multiple disparate sources, including social media, poses challenges for insurers in relation to validating those data. Insurers and insurtech firms also face challenges associated with validating models that use the data. Although AI and machine learning can help insurers and agents underwrite risk more accurately, these stakeholders said that these tools and processes can increase risk because the collected information may be inaccurate or inappropriately used in determining premium rates. For example, while models may indicate that certain factors developed by AI from social media and other sources are associated with increased policyholder risk, it may be difficult or impossible for insurers to validate the accuracy of such data. In addition, it can be a challenge for insurers to ensure that the use of such data and models does not result in the use of prohibited factors in determining premium rates, such as race or sex. For example, several stakeholders told us that certain factors, while not specifically disallowed by insurance regulations, could end up serving as a proxy for a disallowed factor. One example cited by a stakeholder was the use of information on consumer magazine subscriptions, which are not prohibited on their own, but could serve as proxies for factors that are prohibited. Finally, it can be a challenge for insurers to document and explain to regulators how rating models that use AI and machine learning work and provide assurance that the rates produced by the models are not unfairly discriminatory toward policyholders. For example, some industry stakeholders we interviewed said that these models are often developed by data scientists and not actuaries, as had been the case in the past. Unlike actuaries, they said data scientists who develop rating models may not fully understand insurance-specific requirements, such as setting premium rates that are not unfairly discriminatory, and may struggle to measure the impact of new variables used in the models. Furthermore, data scientists may be unfamiliar with insurance rules and regulations and may not understand how to communicate their work to state insurance regulators. One regulator described to us how one insurance company was unable to explain how one of the factors that it entered into its advanced risk model—proximity of a home to a day care center—related to the risk that a consumer posed. An actuarial group suggested a greater collaboration between actuaries and data scientists could provide greater assurance that such rating models meet regulatory requirements. Quality of data used in pricing. According to some stakeholders, insurers’ use of nontraditional data and AI to develop insurance pricing models creates two potential risks for consumers that parallel some of the risks for insurers. First, as previously mentioned, insurer’s use of nontraditional data and AI can create a risk that factors unrelated to the risk presented by a consumer could be used to set his or her premium rate. Stakeholders including a regulator said that algorithms or big data may allow insurers to correlate certain factors with higher claim rates, although the factors do not actually relate to risk and may even act as a proxy for a prohibited factor such as race or sex. As a result, some stakeholders noted that using such information to determine a premium rate could be unfairly discriminatory. Some stakeholders also said that such factors unintentionally could become proxies for prohibited rating factors— such as race. For example, using information on a consumer’s purchase history could serve as a proxy for race. Second, some stakeholders indicated that when insurers use AI to generate information on consumers, it is difficult to ensure these data are accurate. Because the data were not explicitly provided by the consumer, the consumer does not have a chance to correct or dispute the data. For example, if an insurer uses AI to pull data from a consumer’s social media accounts, those data could be incorrect or outdated, but the consumer would not know the data were being used as a factor in determining his or her premium rate. This would prevent the consumer from correcting the information if it was wrong. Some stakeholders indicated that if an insurer has difficulty understanding the factors and algorithms being used to price the insurance product, the consumer most likely will not be able to understand them. Consumer Protection According to stakeholders with whom we spoke and literature we reviewed, some uses of technology can pose risks in terms of the protection of consumer data. In addition, the use of the nonadmitted market by insurtech companies and insurers may result in more limited financial protections for consumers. Cost of protecting consumer data. As noted earlier, insurers collect and use consumer data in large quantities and from multiple disparate sources, including social media, posing challenges for protecting those data. For example, according to representatives of one property/casualty industry association we interviewed, it can be expensive to maintain the appropriate level of cybersecurity (including technical and organizational measures) to prevent any unauthorized access or use of the additional volumes and types of customer information used in recent years. Consumer privacy concerns. Stakeholders noted that insurers’ expanded use of consumer data raises concerns about the privacy of such data. For example, an automobile insurer may collect data on a consumer using a telematics device installed in the consumer’s vehicle. While an insurer may use data on the consumer’s driving habits for the purpose of adjusting premium rates, the device also may collect information on where and when a consumer drives. This is information consumers may not wish others to possess. One academic also said there is concern about the ownership of the data collected through telematics and other technologies, such as AI, for the purposes of insurance. For instance, if an insurer obtained data from a policyholder’s automobile with a telematics device, a question exists about whether policyholders would have the right to take those data to another insurer if they switched insurers or whether the data belong to the first insurer. As we have described in other work, this presents a larger privacy issue as it may not be possible for a consumer to know exactly what is collected, or when and how the data are used. This lack of knowledge reduces the consumer’s control over their personal information and limits their ability to track what data belong to them. Some stakeholders mentioned concerns about insurers collecting information from social media and other sources that consumers did not explicitly consent to provide to insurers. The European Union (EU) General Data Protection Regulation, which includes regulations governing consumer consent, had an entry into force and application date of May 25, 2018. According to an industry analyst, the General Data Protection Regulation applies to insurance companies around the world, including those in the United States, that process the personal data of EU residents, regardless of the nationality of the person in question or the location of the company. Furthermore, the analyst notes that the regulation strictly defines legal uses of individuals’ data and requires companies to ensure individuals can explicitly and individually consent to other uses of their data. In prior reports, we also noted data privacy concerns in relation to lender use of financial technology. Consumer protection concerns due to use of the nonadmitted market. The nonadmitted market is a common entry point for insurtech firms because of that market’s usefulness for innovative insurance products with little loss history. However, the sale of consumer insurance through nonadmitted insurers raised concerns among several stakeholders. As we noted in a prior report, nonadmitted insurers may face fewer regulatory constraints than traditional insurers in the prices they can charge and their ability to create and offer new products. While data do not exist on the number of insurtechs using the nonadmitted market, industry representatives told us that because of this greater regulatory freedom, a number of insurtechs choose to operate as nonadmitted insurers or as brokers selling policies through nonadmitted insurers. As described in the Background, when consumers purchase insurance from nonadmitted insurers, they do not have some of the same consumer protections they would have if they purchased coverage from an admitted insurer. For example, regulators conduct limited reviews of the prices charged and the products sold by nonadmitted insurers. And as noted earlier, if nonadmitted insurers became insolvent, state guaranty funds may not be available to help pay policyholder claims. As we previously reported, some regulations serve to push potential policyholders toward the admitted market because of the better financial protections it provides (such as rate approvals and access to state guaranty funds). For example, as noted earlier, a broker placing coverage with a nonadmitted insurer generally must conduct a diligent search for available coverage in the admitted market every time a potential policyholder requests coverage in the nonadmitted market. This helps ensure coverage is purchased from an admitted insurer as often as possible. Stakeholders offered differing assessments on the extent of any related risks to consumers resulting from insurtech use of the nonadmitted market. For example, an industry representative said the nonadmitted market is not appropriate for most consumer products because of the lower consumer protections as compared with the admitted market. Two insurtech firms also have raised questions about the ability of insurtech companies and other market participants to properly comply with diligent search requirements. For example, an industry representative told us it does not seem possible to satisfy the diligent search requirement when products are sold on-demand through a mobile app. Furthermore, the representative raised the question of how a broker could legitimately search the admitted market for coverage in cases in which an insurer offers immediate coverage as soon as consumers complete applications on their smartphones. Conversely, some insurers, regulators, and NAIC said that nonadmitted insurers are appropriately regulated and consumers are not necessarily at any greater risk than when purchasing coverage from admitted insurers. Also, several states have eliminated the diligent search requirements. However, a consumer advocate noted that such deregulation raises further consumer protection issues in a market where less regulation is already a concern for consumers. Business Operations and Risk Monitoring According to the literature we reviewed and stakeholders we interviewed, insurers have been using various technologies to reduce their operating costs but may face risks that affect their operations and business models. Reduced costs. Stakeholders described how adopting various technologies has led to reduced costs in four operational areas for insurers: Communicating with customers. Insurers have been using mobile apps and chatbots to reduce the cost of providing information to potential customers. For example, a consumer might be shopping online for an insurance policy late in the evening. The insurer can use a chatbot to interact with that consumer and answer questions about insurance products. In the past, this might not have been possible if an agent was not available to work nonstandard business hours or insurers might have needed to hire and retain more agents to work evenings and weekends. Underwriting. Insurers have been using technology to reduce the cost of underwriting insurance. For example, according to two insurtech firms and one industry representative we interviewed, some insurers review multiple sources of data with AI to automatically review the information in a consumer’s insurance application, rather than incurring the costs of hiring staff to do so. Through the industry article review and stakeholder interviews, we found that insurers also use the internet of things to obtain data from smart home alarms to monitor consumer usage of alarm systems and thereby assess consumer risk levels. This reduces the costs associated with determining and analyzing risk factors. Claims processing. According to some stakeholders we interviewed, insurers now have the capability to digitally collect and automatically analyze claim evidence, thereby reducing staffing needs and realizing cost savings. For example, consumers can use their smartphones to take photographs of their vehicles after an accident and send the photographs and other information to their insurers through mobile apps. On receipt of the photographs, insurers can use AI algorithms to verify the damage shown—decisions that historically required human intelligence to perform—and automatically start the claims process for the consumer. Fraud. Insurers are able to detect fraud, or decide which claims need to be investigated further by employees, with information verified using big data, the internet of things, and telematics. For instance, an insurer may verify information provided in a claim against information obtained from a smart device to determine if the information provided by the policyholder was accurate. An insurer also might identify a false burglary claim by verifying whether an alarm was set during the time frame identified in the claim and reviewing video from home security cameras. Connecting to legacy computer systems. Some industry stakeholders and association representatives we interviewed stated that established insurers face significant challenges using new technologies because they first have to replace legacy computer systems or customize their systems to interface with new technologies properly. According to industry stakeholders, legacy computer systems were, in some ways, built around satisfying regulatory requirements rather than enhancing the consumer experience or providing more desirable products. They noted it can be costly and difficult to replace such systems or to modify them to interface with more consumer-centered systems, such as those being developed by insurtech companies. Changing roles for insurers and agents. According to some insurance industry stakeholders, emerging uses of key technologies and innovative business models could lead to changes in insurers’ roles and products. For example, with the advent of self-driving vehicles, the liability for accidents could shift from the driver to the vehicle maker or the company that produced the self-driving system. In such cases, they said insurance coverage primarily would be sold to those entities rather than the consumer, and the demand for and amount of consumer automobile coverage sold could decrease substantially. This could cause a shift in demand for products from consumers to commercial lines, resulting in the potential loss of business for some agents and insurers. Some industry stakeholders we interviewed also told us that as more technologies (such as telematics or other smart devices) were adopted to help consumers mitigate risk, insurers likely would have to shift their business model. That is, they would have to move from a model focused on sales of policies, in which agents play a central role, to a model focused on providing consulting services to consumers to help them prevent and mitigate risk and loss. Risk monitoring. Insurers have been using big data with data aggregation and mining to improve monitoring of insured risks. More specifically, several stakeholders told us that these tools and analytical methods can help insurers quickly analyze volumes of data from many sources in or near real time. For example, several stakeholders gave the example of an insurance company using sensors or other devices to continuously collect verified data on movements of insured ships and their cargo. Such data can be useful to insurers for understanding the risks associated with providing insurance coverage and even can be used to provide the ship carrying the cargo the appropriate insurance documentation required for the port of entry. Several stakeholders also told us that some insurtech companies have been using telematics to collect real-time data on driver behavior, which they combine with other information such as credit scores, to develop a fuller and more accurate picture of the risk presented by a given policyholder. Insurers then can use these risk profiles to determine whether to change a policyholder’s rates or continue to insure them. Several stakeholders indicated that such real-time information is likely more accurate than previous risk- assessment methods. Product Offerings According to stakeholders we interviewed and literature we reviewed, the use of various technologies to create new product offerings has created several benefits for insurers and consumers. Ability to offer on-demand products. Technologies have been helping insurers tailor products to specific consumer needs and expand offerings to niche markets. Some insurtech companies have started offering on-demand insurance (insurance that policyholders can turn on and off as needed). For example, one regulator and an academic said that market research data demonstrated that consumers want to be able to turn on insurance for their drones when the drones are in use and turn it off when the drones are idle. Insurers also have been developing similar on-demand products for drivers working for rideshare companies such as Lyft and Uber and for Airbnb and VRBO rentals (to cover the gaps that traditional homeowners insurance, which generally provides coverage on a long-term basis, might have in relation to short-term rentals of homeowners’ properties). On-demand products allow insurers to diversify their product lines and attract more consumers, which is discussed later in this report. Increased convenience. With some insurers providing mobile apps and chatbots, consumers are able to access insurance products and information 24 hours a day. For example, consumers can use mobile apps to get immediate quotes and underwriting decisions from some insurers. In the past, consumers likely would have had to visit an insurance agent or fill out a lengthy application and wait much longer for an underwriting decision. And as previously discussed, some insurers allow their policyholders to submit claim information and photographs of damage through a mobile app without speaking with an agent. Increased consumer choice. According to NAIC and an insurtech firm, consumers can benefit from the increased choice that comes from insurers using technology to offer additional products and services. For example, consumers obtain the ability to purchase insurance for certain time periods for certain items such as drones and action cameras, home sharing, or mile-based automobile insurance. NAIC and the insurtech firm said that some insurers that offer insurance to rideshare operators allow the policyholders to turn the coverage on when they are working and off when they are not. This can reduce premium rates for policyholders who only occasionally work as rideshare drivers. According to the industry articles we reviewed and the stakeholders with whom we spoke, insurers’ use of technology also has benefitted consumers by leading to the development of aggregator websites that bring together quotes from multiple insurers and allow consumers to comparison shop for insurance products. Some insurers said technology may soon give consumers the added ability to further customize their insurance policies by allowing them to select among various available coverages and terms and essentially create a policy that best suits their needs. NAIC and State Regulators Initiated Actions to Address Challenges That Stakeholders Said Could Affect Adoption of Technologies NAIC, state regulators, and others have initiated a number of actions intended to monitor and address industry and regulator concerns associated with insurtech, including any insurance rules and regulations that could affect insurers’ adoption of technologies. These actions address challenges in areas including (1) evaluation of underwriting methodologies, (2) approvals for new insurance products, (3) customer notification methods and time frames, (4) anti-rebating laws, (5) cybersecurity, and (6) regulator skillsets and resources. NAIC and State Regulators Have Taken Actions Designed to Monitor Insurtech Concerns and Maintain Insurer Oversight and Consumer Protection NAIC and state regulators have initiated a number of actions intended to monitor concerns that regulations could affect insurers’ adoption of innovative technologies while maintaining oversight of consumer protection issues. First, to monitor technology developments that may affect the state insurance regulatory framework and to develop regulatory guidance, as appropriate, NAIC created an Innovation and Technology Task Force. According to NAIC, this task force provides a forum for regulator education and discussion of innovation and technology in the insurance sector. For example, the task force has held discussions on the collection and use of data by insurers and state insurance regulators—as well as new products, services, and distribution platforms—to educate the regulators on how these developments affect consumer protection, privacy, insurer and producer oversight, marketplace dynamics, and the state-based insurance regulatory framework. In addition, the task force has held forums on emerging issues related to companies or licensees leveraging new technologies. Areas discussed included developing products for on-demand insurance purposes, reviewing new products and technologies affecting the insurance space, and potential implications for the state-based insurance regulatory structure. In addition, in 2012 the EU-U.S. Insurance Dialogue Project was formed, in which EU and U.S. insurance regulators discuss emerging technology issues in the international insurance industry. During the project’s sixth forum in November 2018, the regulators and representatives from industry and consumer organizations discussed challenges and opportunities relating to issues including cyber risks, the use of big data, and AI. According to a project publication, the dialogue project enhanced mutual understanding of respective regulatory frameworks and initiatives between the United States and European Union, which will help ensure effective coordinated supervision of cross-border insurance groups for the benefit of policyholders. In 2018, the project published an issues paper on big data. The paper discusses data collection, portability, quality, and availability and how insurers and third parties use data in marketing, rating, underwriting, and claims handling. Future work by the project may include discussion of insurers’ use of third-party vendors, disclosures to applicants, and insurers’ use of AI models. NAIC and State Regulators Initiated Actions to Address Specific Insurtech Challenges NAIC and state regulators have initiated a number of actions intended to address industry and regulator concerns about certain insurance rules and regulations that a number of them said could affect insurers’ adoption of technologies. Evaluating Underwriting Methodologies That Use Technology Stakeholders, including regulators, told us that regulators can face challenges in assessing new underwriting methodologies, such as those that use predictive analytics or AI. Reviewing predictive analytics can be a challenge for regulators because of the amount of data used to develop a model, the complexity of techniques, and limited staff resources (discussed in more detail later in this section). In addition, insurers employ different technological approaches, and their documentation and explanation of the methods and approaches differ. Finally, the data and models insurers use dynamically change and may have to be re- submitted for review even before regulators have an opportunity to review the original submission. One state regulator and an industry stakeholder also told us that while an insurer may know the universe of factors from which an AI system pulls, the insurer may not know, or be able to describe for regulators, how the system uses those factors to determine a premium rate. In turn, this may prevent regulators from understanding the system or validating the insurer’s assertions about the system. For example, one state regulator told us that after presenting a rate scheme based on nontraditional factors, an insurer was unable to provide assurances or explanation to the regulator that the resulting premium rates were not unfairly discriminatory. In 2018, NAIC’s Casualty Actuarial and Statistical Task Force began developing a white paper on best practices state regulators can use when reviewing predictive models and analytics filed by insurers to justify rates and guidance they can use for their review of rate filings based on predictive models. NAIC officials told us the Casualty Actuarial and Statistical Task Force will receive comments on the white paper and then evaluate how to incorporate best practices into the Product Filing Review Handbook and recommend such changes to other NAIC working groups. Approvals for New Products Insurtech firms and other stakeholders told us that working through other regulatory processes, such as the insurance product filing and approval process, often can be inefficient and time consuming because insurers must file in every state in which they wish to sell a product and state requirements can vary. We have noted such difficulties in the insurance market in general. These challenges can be exacerbated by rapid technological evolution in insurer products and risk models. In addition, some stakeholders noted that a lengthy product approval process can be challenging for technology-oriented products. For instance, an insurtech firm may develop a new product quickly to meet consumer demand but might not be able to get the product to market quickly. Some also said that products might become obsolete before the filing approval process was completed. Some stakeholders told us that such challenges can motivate insurtechs to sell insurance through nonadmitted insurers because such insurers have more freedom in altering and selling new products. As we have noted, doing so can bring risks for consumers. In December 2017, the American Insurance Association proposed the Insurance Innovation Regulatory Variance or Waiver Act (Proposed Model Law) to NAIC. The proposed model law would urge allow regulators to create regulatory “sandboxes,” wherein certain regulatory requirements would be waived for insurers seeking to pilot innovative products. Specifically, the proposed model law would authorize insurance regulators to grant variances, waivers, or no-action letters with respect to statutory or regulatory requirements that make it more difficult to introduce new insurance technologies, products, or services. Under the proposed model law, regulators also would be authorized to attach terms and conditions meant to protect consumers to such variances or waivers. Some stakeholders with whom we spoke believed that regulatory sandboxes would not work in the U.S. state-based regulatory framework. For example, some stakeholders told us it would be inappropriate for a state to change legal or regulatory requirements for some but not all insurers or grant exceptions to laws passed by a state legislature to some insurers and not others, as it would no longer be a level playing field. State regulators generally told us they believe that the current regulatory framework provides state regulators with enough flexibility to allow for technology-based innovation. Accordingly, some states have been promoting the use of innovation in the insurance industry by hosting technology sandboxes, where technology companies meet regularly with state regulators to improve companies’ knowledge of insurance regulations and also educate regulators about how the technologies work. According to stakeholders, these technology sandboxes are not the same as regulatory sandboxes that have been established in other nations, as they do not allow waivers of laws and regulations for insurtech companies to test their products. Paper Notification Requirements Insurtech firms we interviewed told us that regulations that require paper notifications and U.S. mail delivery for certain processes can make it difficult or more costly for them to offer products with features such as immediate underwriting or on-demand policies. For example, according to insurers and other industry stakeholders, some state laws require that insurance policy cancellation notices be sent by U.S. mail rather than by email. One insurtech firm told us that it would be very costly to meet requirements for mail delivery of insurance policies and cancellation notices because they would have to set up another delivery mechanism (in addition to their electronic notification system). Industry stakeholders also told us that certain laws and regulations that require a minimum period of time before a consumer-initiated policy cancellation takes effect can present challenges for products designed to allow consumers to immediately turn certain coverage on or off. For instance, if consumers used a mobile app to indicate they wanted to turn their automobile insurance coverage off temporarily, it could be unclear if this constituted an actual policy cancellation. Some stakeholders are concerned that states may require an insurance company to give the policyholder a written notice of cancellation at least 30 days before the end of the policy term. Similarly, industry stakeholders told us that some current state regulations could impede on-demand coverage because policies usually must indicate that coverage begins at 12:01 a.m. on the day after a policy is signed and approved. For instance, for on-demand policies that allow on/off subscription at the consumer’s request, it can be unclear whether they are covered the minute that they initiate the coverage, or if they must wait until the following day for coverage to be effective. According to NAIC, many states have taken steps to work within or modify existing laws and regulations to adapt to the increased use of technology in the insurance industry. For example, to address concerns that insurers are required to provide customers with a written, 30-day notice of a policy cancellation, NAIC conducted an analysis in 2018 that found that many states instead require “adequate” notice and that approximately 44 states allow notices to be provided electronically. However, some stakeholders in the insurance industry told us that state cancellation notice requirements are still a barrier to innovation. Anti-Rebating Laws According to industry stakeholders, many states have anti-rebating laws that generally prohibit insurers from providing consumers with anything of value as an inducement to purchase insurance. NAIC Model Law 880 states that unless expressly provided by law, no insurer may knowingly pay any rebate or incentive to an insured to induce them to purchase a specific product. Insurers, industry stakeholders, and regulators (including NAIC’s Innovation and Technology Task Force) told us that anti-rebating laws can be a barrier to innovation because they could preclude insurers from offering devices that could be used to help insurers and consumers monitor risk. For example, if an insurer offered a policyholder free use of a telematic device (to help insurers collect real- time data and potentially help the policyholder make driving habits safer), it could be considered an inducement and violate anti-rebating laws. The same possibility exists if an insurer were to provide a policyholder with a device to monitor the operating conditions of a boiler to prevent potential water damage should a problem arise. As a result, anti-rebating laws may make it difficult for insurers to make use of certain technologies that could benefit both insurers and policyholders. In contrast to the consensus on the legitimacy of electronic communications, there is little consensus among states on addressing insurers’ concern that anti-rebating laws are a barrier to innovation. According to NAIC, states vary widely on the types of items insurers are allowed to provide for free to customers, with some states having dollar limits on allowable items or allowing items that are specifically linked in a policy. In other cases, it is unclear what is allowable. At NAIC’s fall 2018 meeting, participants noted that some of the NAIC bulletins related to the anti-rebating model law have not addressed whether technologies such as telematics that provide benefits to consumers are considered rebates. According to NAIC, others noted that states typically have taken the position that if a rebate or incentive reduces risk that is the most important issue for all parties involved. NAIC officials noted during the fall 2018 meeting that they will continue to monitor the issues involved. Cybersecurity NAIC adopted a model law and states have passed new laws governing cybersecurity and data protection to safeguard the increasing amount of personal data used by insurers. In 2017, NAIC approved the Insurance Data Security Model Law, which creates a legal framework for requiring insurance companies to operate cybersecurity programs. The law outlines planned cybersecurity testing, creation of an information security program, and incident response plans for breach notification procedures. The NAIC model law is only a guideline until adopted by individual states, but NAIC noted that in 2018 and 2019, Michigan, Ohio, Mississippi, and Alabama adopted laws based on the NAIC model and additional states have pending legislation. In an October 2017 report, Treasury endorsed the model law and recommended that Congress consider preempting the states if the law were not adopted over the next 5 years. At the state level, New York’s Department of Financial Services noted it was the first state agency to establish cybersecurity regulations, which became effective March 1, 2017. In May 2018, South Carolina enacted the South Carolina Department of Insurance Data Security Act, which NAIC has characterized as an adoption of the model law. In December 2018, Michigan adopted a similar law. Separately, in June 2018 California passed a law giving consumers more control over their personal information. California’s law generally requires companies to report to customers, upon their request, the categories of personal information they collected about the customer, the business or commercial purpose for collecting and selling such personal information, and what categories of third parties received it. Hiring and Retaining Staff with Technical Expertise According to industry and regulatory stakeholders, the complexity and evolving nature of the models and approaches used by insurers may outpace the rate at which regulators can educate themselves on those models and approaches. For example, regulators trained in the current rating models may need to acquire new skills to understand and validate advanced and evolving models. In addition, stakeholders told us that new technologies used by insurers can pose significant challenges to regulators partly because of the resource requirements. For instance, regulators and other stakeholders told us that regulators often do not have enough staff with technical expertise, such as data analytics skills, and find it challenging to hire and retain such staff due to limited resources. NAIC has initiated actions to address concerns that state insurance regulators may not have staff with the knowledge or skill sets to address more complex predictive models. For example, in 2018 NAIC management conducted a survey of states regarding the appropriate skills and potential resources NAIC membership may need to deal with big data. Subsequently, in April 2019, NAIC management made recommendations to its Big Data Working Group to hire a technical staff resource to provide technical support for state insurance regulators in the review of actuarial models; develop a tool for state insurance departments to share information on model reviews; and develop a training and education program. NAIC officials told us they also plan to develop a white paper to provide state regulators with guidance on the use of chatbots and AI in the distribution of insurance and the regulatory supervision of these technologies. As many of the regulatory initiatives that NAIC and states have undertaken to address challenges associated with the implementation of new technologies are under development (or recently developed), the impact of these actions on innovation and consumer protection is unknown. It will be important for NAIC and state insurance regulators, as well as the Federal Insurance Office, to continue monitoring developments in these areas. Agency and Third Party Comments We provided a draft of this report to Treasury and NAIC for review and comment. Treasury and NAIC provided technical comments that we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of the Treasury, the Chief Executive Officer of the National Association of Insurance Commissioners, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions concerning this report, please contact me at (202) 512-8678 or ortiza@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix II. Appendix I: Objectives, Scope, and Methodology This report (1) identifies uses of technologies and the benefits and challenges they might present for insurers and their customers, and (2) discusses what stakeholders identified as key challenges that could affect the adoption of new technologies, and actions that have been taken to address those challenges. While insurance technology (insurtech) does not have a standard definition, for the purposes of this report we defined it as the use of emerging technologies by insurance companies. We focused on insurtech activities in the property/casualty and life sectors of the U.S. insurance market, including information on personal and commercial insurance where available. We did not include the health insurance sector in our scope because of significant differences between that sector and the property/casualty and life insurance sectors in terms of the types of products offered and the methods by which they are sold and regulated. To identify technologies being used in the insurance industry and gain insights about their (potential) benefits and challenges for insurers and customers, we conducted a literature review of scholarly and peer- reviewed material, trade and industry articles, government reports, conference papers, general news, association, nonprofit, and think tank publications, hearings and transcripts, and working papers that described these technologies and their uses. We conducted searches of the ProQuest and HeinOnline databases to identify studies published from January 2015 through June 2018 that were relevant to our research objectives. Because insurtech is a fairly new field, we found few academic publications related to our objectives. We also conducted background research for examples of technologies being used in the insurance industry and their associated benefits and challenges. We also conducted semi-structured interviews with cognizant stakeholders and reviewed documents provided by them to obtain information on and descriptions of current, in-development, and potential future uses of existing or new technology in the insurance industry. We also obtained their views on the benefits and challenges experienced or expected by insurance companies as well as the (potential) benefits and challenges for consumers. We conducted more than 35 interviews with representatives of regulatory organizations, including the Federal Insurance Office; National Association of Insurance Commissioners (NAIC); state insurance regulators in Arizona, California, Connecticut, and Michigan; and the National Council of Insurance Legislators. We also interviewed three academics, representatives of one consumer group, 13 traditional insurance and reinsurance providers and industry associations, two actuarial professional associations, four consulting groups, two law firms in the field, and seven insurtech firms. We identified potential interviewees by conducting internet research, reviewing literature search results, and reviewing recommended interviewees from our initial interviews. We selected interviewees based on their relevance to the scope of our review. Based on our literature review and interviews with stakeholders, we identified seven recently used and emerging technologies in the insurance industry: (1) mobile applications; (2) artificial intelligence (AI), algorithms, and machine learning; (3) big data; (4) internet of things; (5) blockchain/ distributed ledger technology and smart contracts; (6) drones; and (7) telematics. To obtain information about challenges that could affect the adoption of innovative technologies, we identified relevant laws and regulations pertaining to insurance technology innovation by reviewing prior GAO reports on financial regulation, interviewing regulators and industry participants, and analyzing relevant documents, including relevant NAIC model laws and state laws and regulations. We also conducted semi- structured interviews with and reviewed documents provided by the key stakeholders identified in the first objective to identify (1) any actions NAIC and selected state insurance regulators were taking on new insurance technologies, and what challenges, if any, insurers’ use of new technologies creates for regulators; (2) what is known about the impact of any actions taken by NAIC and state insurance regulators on innovation among insurance companies and on consumer protection; and (3) stakeholders’ views on the applicability of foreign regulatory actions for U.S. insurtech markets. We conducted this performance audit from April 2018 to June 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Patrick Ward (Assistant Director), Deena Richart (Analyst in Charge), Gina Hoover, Hadley Nobles, Akiko Ohnuma, and Tyler Spunaugle made key contributions to this report. Also contributing were Emei W. Li, Barbara Roesmann, Jena Y. Sinkfield, Frank Todisco, and Helen Tulloch.

Why GAO Did This Study The innovative use of technology by insurance companies (insurtech) is growing and offers the potential to improve customer experiences while also lowering insurer costs. Some stakeholders have raised questions about how certain uses of insurtech could create both risks for consumers and challenges for regulators, and whether some challenges might slow technological innovation in the insurance sector. GAO was asked to provide information on insurtech activities in the property/casualty and life insurance sectors. This report (1) identifies new uses of technologies and potential benefits and challenges for insurers and their customers; and (2) discusses what stakeholders identified as key challenges that could affect the adoption of new technologies, and actions taken to address those challenges. GAO reviewed available literature; analyzed relevant laws and regulations; and conducted interviews with more than 35 stakeholders, including federal and state regulators, technology companies, insurers, and consumer groups (selected based on literature reviews and recommendations, and for relevance to the scope of GAO's review). GAO is not making any recommendations in this report. What GAO Found Insurtech companies (recently established companies bringing technology-enabled innovations to the insurance industry) as well as established insurers have begun to use technologies, including artificial intelligence (AI) and mobile applications, in an attempt to improve risk assessment and enhance customer experiences. For example: Consumers can purchase insurance products specifically tailored to their situation and needs, such as renters or auto insurance that can be turned on and off as needed using a mobile app. Some insurers have begun to use nontraditional data (such as from social media) to analyze policyholder risk, and use AI and complex algorithms to reduce costs by automating information gathering and risk assessment. However, implementing these technologies can create potential challenges for insurers and risks for consumers, including the following: The use of AI to create underwriting models for determining premium rates can make it challenging for insurers to ensure that factors prohibited by regulation (such as race) are not used in models. Such models are often developed by data scientists who, unlike actuaries, may not fully understand insurance-specific requirements. Insurer collection and use of consumer data not provided by the consumer raise questions about data accuracy, privacy, and ownership. Some insurtechs sell coverage through nonadmitted insurers. As we have previously reported, nonadmitted insurers—unlike traditional insurers—are not required to be licensed in each state in which they sell insurance, and receive less regulatory oversight of their policies and rates. Also, if nonadmitted insurers became insolvent, state guaranty funds would not be available to help pay policyholder claims. Stakeholders with whom GAO spoke identified challenges they said might affect adoption of innovative technologies. These include paper-based documentation requirements that do not accommodate online insurance transactions, and challenges for regulators in the evaluation of complex rating models. The National Association of Insurance Commissioners (NAIC) and state regulators have initiated a number of actions designed to address such concerns. For example: State insurance regulators, through an NAIC task force, have been examining regulatory areas that may pose obstacles for innovation, such as requirements for paper documentation or signatures. NAIC issued draft best practices for states to use when reviewing complex rating models. NAIC adopted a model law that creates a legal framework for states to use to require insurance companies to operate cybersecurity programs and protect consumer data. Because many of these regulatory initiatives are still in development (or recently developed), the effect on innovation and consumer protection is unknown.

gao_GAO-19-523

gao_GAO-19-523_0

Background SBA oversees a number of programs designed to provide small businesses with resources and tools, including access to capital, help with federal contracting opportunities, and entrepreneurial counseling and training. SBA’s Size Standards and Industry Classification Codes Federal procurement regulations generally define a small business as one that is independently owned and operated and not dominant in its field and that meets the size and criteria or standards established by SBA. The Small Business Act of 1953 authorized SBA to establish size standards for determining eligibility for all procurement programs in which small business status is required or advantageous. SBA uses the North American Industry Classification System (NAICS) as the basis for its size standards. The standards vary by industry and are generally expressed as the average number of employees over a 12-month period or average annual receipts in the previous 3 years. As of August 2019, employee- based size standards for federal procurement purposes ranged from 100 to 1,500, and revenue-based size standards ranged from $1.0 million to $41.5 million. The number of employees or average annual receipts indicates the maximum size allowed for a business and its affiliates to be considered small. The Small Business Jobs Act of 2010 required SBA to review at least one-third of all size standards during every 18-month period from the date of its enactment and all size standards at least once every 5 years thereafter. SBA completed the first 5-year review in 2016. According to SBA’s size standard methodology, the agency assesses industry structure and the overall degree of competitiveness of an industry and of firms in the industry when establishing size standards. To assess industry structure, SBA analyzes four primary factors: average firm size, degree of competition in an industry, start-up costs and entry barriers, and distribution of firms by size. SBA also considers the ability of small businesses to compete for contracting opportunities under the current size standards. According to SBA officials, for industries with $20 million or more in federal contracting annually, SBA also examines the small business share of federal contract dollars relative to the small business share of total industry receipts. Small Business Goals and Set-Aside Contracts for Small Businesses Each year, SBA negotiates small business prime contracting goals with federal agencies that have procurement authority so that, in the aggregate, the federal government meets its goal of awarding 23 percent of prime contract dollars to small businesses. In September 2018, we reported that SBA considers prior-year achievement and other factors in setting annual agency goals. SBA’s procurement center representatives and OSDBUs assist agencies in meeting small business goals. Agency contracting officers have the authority to enter into, administer, or terminate contracts and are responsible for helping agencies meet small businesses goals, including by setting aside contracts for small businesses. One of the first steps in the federal acquisition process is assignment of the NAICS code that best describes the principal purpose of the acquisition and corresponding size standard (see fig. 1). Generally, the FAR states that if the contract is valued under the simplified acquisition threshold, the contracting officer must set it aside for small businesses. If valued above the simplified acquisition threshold, the contracting officer conducts market research to determine whether a contract should be set aside for small businesses. For contracts not set aside for small businesses, contracting officers generally must include specific small business subcontracting goals for the prime contractor to meet. Past Performance Requirements The FAR requires agencies to evaluate price or cost to the government in every source selection and evaluate the quality of the product or service in the acquisition by considering one or more noncost factors, such as past performance. Contracting officers generally have discretion under the FAR to choose evaluation factors and their relative weights, rating systems, and the past performance they will consider. For example, a contracting officer could consider technical excellence, past performance (including relevance), and price. To select a firm for contract award, agency officials evaluate offers against the criteria specified in the solicitation. According to the FAR, if an offeror has no record of relevant past performance, the offeror cannot be evaluated favorably or unfavorably on past performance. Contracting officers use the Contractor Performance Assessment Reporting System to enter and review evaluations of past performance. The FAR generally requires agencies to document contractor performance on contracts or orders that exceed certain dollar thresholds. Once a project is complete, the assessing official rates the contractor on elements such as quality of the product or service, schedule, cost control, management, and small business utilization. This information then becomes available to other agencies for making source selection decisions. Indefinite Delivery/Indefinite Quantity Contracts Indefinite delivery/indefinite quantity (ID/IQ) contracts provide flexibility when an agency cannot specify the quantities or timing of a product or service. Contracting officers may issue ID/IQ contracts as single-award or multiple-award contracts. Single-award refers to a situation in which one contract is awarded under a solicitation. The FAR establishes a preference for awarding multiple-award ID/IQ contracts—instances in which more than one prime contractor is awarded a task-order contract (for services) or delivery-order contract (for supplies) under a single solicitation. Once agencies determine their specific needs, such contracts allow agencies to establish a pool of qualified contractors to compete for future orders under streamlined procedures. Contractors compete to be in the pool and generally compete again for task or delivery orders. Multiple-award ID/IQ contracts can be unrestricted (open competition for businesses of all sizes) or restricted to small businesses. They also can have one pool of contractors with separate “tracks” for small and nonsmall businesses to ensure contract opportunities for small businesses (that is, some orders are set aside for small businesses). After a multiple-award ID/IQ contract has been awarded, an agency places delivery or task orders, generally using the fair opportunities process. An order, which is placed when a specific need arises, obligates funds and authorizes work. Orders must be within the scope, issued within the period of performance, and be within maximum value or quantities agreed to in the contract. For multiple-award ID/IQ contracts, the FAR requires that each awardee be given a fair opportunity to compete for subsequent orders. In April 2017, we found that in fiscal years 2011–2015, federal agencies obligated more than $130 billion annually through ID/IQ contracts. We also found that contracting officers said it was easier and faster to place an order under an existing ID/IQ contract than to award a separate contract when a specific need arose. SBA’s Mentor-Protégé Programs A mentor-protégé program is an arrangement in which mentors—typically experienced prime contractors—provide technical, managerial, and other business development assistance to eligible small businesses, or protégés. SBA established the 8(a) Mentor-Protégé Program in 1998 for mentors to partner with 8(a) socially and economically disadvantaged businesses to improve the ability of 8(a) businesses to compete for prime contracts and subcontracts. The Small Business Jobs Act of 2010 and National Defense Authorization Act for fiscal year 2013 authorized SBA to establish a government-wide mentor-protégé program for all small businesses, which SBA named the All Small Mentor-Protégé Program. Small businesses that have a mentor-protégé relationship through either program can form a joint venture with a mentor (which can be a mid-sized or large business) and compete for set-aside contracts as long as the protégé is a small business with at least a 51 percent interest in the joint venture. Team Arrangements Contractor team arrangements take two forms: two or more companies form a partnership or joint venture to act as a prime contractor or a prime contractor agrees with one or more companies to have them act as its subcontractors under a specified federal contract or acquisition program. Companies generally form a contractor team arrangement before submitting an offer. Businesses of all sizes can form joint ventures to compete for contracts. Joint ventures generally have to consist only of small businesses to compete for small business set-aside contracts—the exception being small and nonsmall businesses entered in a mentor- protégé agreement under one of SBA’s programs. A Very Small Percentage of Small Businesses Grew to Be Mid-Sized and Continued to Receive Federal Contracts during Fiscal Years 2008–2017 Most Small Businesses Awarded Set-Aside Contracts in 2017 Were Well Below SBA Size Standards Most small businesses awarded set-aside contracts in fiscal year 2017 did not appear poised to outgrow their size standard. According to FPDS-NG data, about 86 percent of the 121,604 set-aside contracts awarded in that year were to small businesses with revenue or employees at or below 25 percent of the size standard for their industry (see fig. 2). These businesses received about 64 percent of the dollar obligations for set-aside contracts in fiscal year 2017. The small businesses closest to their SBA standards (above 75 percent of the size standard) were awarded about 2 percent of the set-aside contracts and about 7 percent of the contract dollar obligations in fiscal year 2017. We performed the same analysis for fiscal years 2013–2016, and the results across the four quartiles were generally the same throughout the time period. A Very Small Percentage of Small Businesses Awarded Set-Aside Contracts in Fiscal Year 2008 Grew to Mid-Sized by 2013 and Continued to Obtain Contracts Based on our review, a very small percentage of the small businesses that were awarded set-aside contracts in fiscal year 2008 grew to mid- sized in subsequent years and continued to receive any type of contract. As shown in table 1, more than 93 percent of the businesses that were awarded only set-aside contracts in fiscal year 2008 and received any federal contract (including a set-aside or competed contract) in fiscal year 2017 remained small. About 2.5 percent of such businesses had become mid-sized by fiscal year 2017. In addition, we analyzed the extent to which small businesses that grew to be mid-sized in 2013 continued to receive any type of contract in fiscal years 2014–2017. Of the 5,339 small businesses awarded only set- aside contracts in fiscal year 2008 and awarded any sort of federal contract in fiscal year 2013, 104 grew to mid-sized by fiscal year 2013. Of those 104 mid-sized businesses, 23 remained mid-sized in subsequent years and were awarded 75 contracts, and three grew to large and were awarded six contracts (see table 2). Seventeen of the 104 mid-sized businesses became small again. Thirty-seven of the 104 mid- sized businesses were awarded 306 contracts and were categorized as small, mid-sized, or large depending on the NAICS code listed in the contract. That is, businesses can be awarded contracts under several NAICS codes, each with a different size standard. Of the 24 mid-sized businesses not awarded any contracts in 2014–2017, nine were no longer registered in the System for Award Management, a central registration system for federal contractors. Mid-Sized Businesses Can Provide Services and Goods on Contracts Set Aside for Small Businesses under Certain Circumstances Businesses Can Keep Contracts If They Exceed Size Standards during the Life of a Set-Aside Contract Federal regulations generally allow a small business with a contract to continue performing under its contract if it outgrows the size standard that it met in its initial offer. If a business qualified as small and was awarded a single-award contract under a small business set-aside, it generally would be considered small for contracting purposes for the life of that contract. The business can continue providing the service or product. Additionally, the agency can continue counting the contract towards its small business goals unless the business is required to recertify, whether through a regulatory or contractual requirement, and in doing so is deemed other than small. Once the contract ends, the follow-on or renewal contract is a new contract; size is determined as of the date the business bids on the new contract. The regulations are applied similarly to multiple-award contracts. Some multiple-award contracts are set aside for small businesses only. If a business qualified as small at the time of its initial offer, it is generally small for each order issued against the contract for the life of the contract even if it outgrows the size standard. Multiple-award contract orders awarded to businesses that have grown to be other than small during the course of the set-aside contract generally still may be counted toward agency small business goals. They would not be counted if the contractor were required to recertify, whether by a regulatory or contractual requirement, and in doing so was deemed other than small. There are a few instances in which a business must recertify its size status after its initial offer. In the case of an awarded multiple-award, set- aside contract, this would make the concern ineligible for the placement of orders or exercise of options. For example: SBA has stated that mergers and acquisitions create an exception to the general rule that a firm’s size and status is determined at the time of the initial award. Generally, if a business becomes other than small pursuant to a merger or acquisition after its initial offer, the business must recertify its size. Certain requirements for recertification become effective just before the end of the fifth year on a multiple-award contract. A multiple-award contract that runs for more than 5 years, including options, requires each business to recertify size within 120 days before the end of the fifth year and 120 days before exercising options thereafter. The determination of small or other than small is based on the size standard at the time of the recertification. Size determinations are not permanent; a business can recertify later as small if it meets the size standard. A contracting officer may require a business to recertify its size status in response to a solicitation for an order. An SBA recertification determination is based on the size as of the date the business submits its response to the order. The rules are different for agreements, including blanket purchase agreements. A blanket purchase agreement is a simplified method of filling anticipated repetitive needs for supplies or services that functions as a “charge account” with qualified sources of supply. Where the agreement is a set-aside or a reserve award to any type of small business, a business must qualify as small both at the time of the offer and at the time of the order to be considered for the order. The agency may count the business toward its small business goal if the business is small at the time of the order. Businesses That Form Joint Ventures under SBA Mentor-Protégé Programs Generally Can Access Set-Aside Contracts Both SBA’s 8(a) and All Small Mentor-Protégé programs allow the mentor (including those that are mid-sized businesses) and the protégé to form a joint venture and bid on set-aside contracts based on the protégé’s status as a small business. Once a protégé no longer qualifies as small, the mentor-protégé joint venture will no longer be eligible to bid for new small business set-asides. But, a change in protégé size generally does not affect contracts previously awarded to a joint venture between the protégé and the mentor. The mentor-protégé joint venture may seek any small business contract for which the protégé would qualify. Therefore, the size of the mentor generally does not affect whether a mentor-protégé joint venture can bid for a small business contract. According to SBA officials, the agency does not track the size of mentors. As of September 2018, there were 106 joint ventures formed under the All Small Mentor-Protégé program and 171 joint ventures under the 8(a) Mentor-Protégé program (see table 3). We analyzed FPDS-NG and SBA data to determine the size of the mentors participating in joint ventures under SBA’s All Small Program that were awarded set-aside contracts in fiscal years 2016–2018. Of the 29 joint ventures awarded set-aside contracts during these years, 13 of the 26 mentors were mid-sized businesses. Options Proposed by Stakeholders for Assisting Mid-Sized Businesses Vary in Terms of Their Potential Benefits and Involve Tradeoffs We reviewed options proposed in literature to enhance contracting opportunities for mid-sized businesses and asked stakeholders for their perspectives on potential benefits and drawbacks. Some options for increasing federal contracting opportunities for mid- sized businesses identified in our literature review would help mid-sized businesses more than others, according to stakeholders. They noted that establishing a set-aside for mid-sized businesses—the option designed to help mid-sized businesses most directly—also would pose challenges for small businesses and agencies. In contrast, some options primarily would help small businesses that were growing (revenue or employees approaching the size standards). This, in turn, could offset any of the advantages that mid-sized businesses would derive. For instance, benefiting small businesses could increase competition and result in fewer awards to mid-sized businesses. As shown in table 4, we grouped the options into four categories: (1) establishing a set-aside for mid-sized businesses, (2) modifying the rules for multiple-award contracts, (3) changing how past performance is considered when evaluating bid proposals, and (4) modifying SBA’s size standards. A Mid-Sized Set-Aside Could Increase Contracting Opportunities but Affect Other Businesses and Agencies Several stakeholders told us that establishing a separate set-aside category for mid-sized businesses would increase contracting opportunities for mid-sized businesses, but others expressed concerns that the potential threat to small businesses and administrative burden on agencies might outweigh this benefit. Some literature suggests that when businesses outgrow their size standards they struggle to compete against much larger, established businesses for contracts. Also, literature we reviewed suggested small business goals motivate agencies to set aside more and larger contracts for small businesses, resulting in a scarcity of smaller contract solicitations for mid-sized businesses. Members of Congress have proposed establishing pilot programs that would help mid- sized businesses, either defined by business or contract size. Several stakeholders commented directly on separate set-asides and contracting opportunities for mid-sized businesses. An OSDBU director noted former small businesses (those that outgrew their size standard) would benefit from an opportunity to compete with firms of similar size for prime contracts. However, some stakeholders believed the set-aside would not increase opportunities for mid-sized businesses. Specifically, one trade association executive noted that this option continues to shelter small businesses that become mid-sized businesses from competition with larger businesses. The stakeholder added that a set-aside would not address the ability of mid-sized businesses to compete against large businesses on an unrestricted basis. Another trade association executive said there still would be a need to help mid-sized firms develop expertise and encourage competition. Some stakeholders believed the option would have a limited impact or was not necessary. Specifically, one trade association executive said that its members want a good path to growth for small businesses, not a set- aside. Another trade association executive similarly believed mid-sized businesses want to open up contract opportunities, not restrict them by creating more set-asides. One stakeholder also argued that the option could create incentives for large businesses to split their companies to fit new set-aside size standards. Most stakeholders believed the set-aside for mid-sized businesses would take away opportunities from small businesses, with several noting that contracts that normally would be set aside for small businesses might be set aside for mid-sized businesses instead. SBA officials stated that a set- aside would have a negative effect on all small business programs and support the use of larger contracts, resulting in fewer contract awards to small businesses. However, one trade association representative said the set-aside could be structured so that small businesses still could compete for the mid-sized set-aside contracts. An OSDBU director told us contracting officers could limit the effect on small businesses by considering small businesses first, mid-sized businesses second, and large businesses last. In this scenario, it would be large businesses that would be most affected by a set-aside for mid-sized businesses. Stakeholders cited more limitations than benefits for agencies if this option were implemented. Most stakeholders told us a mid-sized business set-aside would increase agency burden, including additional time and cost to define and implement the new set-aside and additional tracking and reporting costs. SBA officials noted that it would create an additional burden for contracting officers and that further study would be needed before implementing a mid-sized set-aside. Some stakeholders also noted the potential burden on agencies of complying with additional contracting goals, with one OSDBU director saying that agencies do not have the resources to meet current small business contracting goals, let alone meet mid-sized contracting goals. Another OSDBU director believed that mid-sized business set-asides likely would violate the World Trade Organization’s Government Procurement Agreement because the United States negotiated exclusions for small businesses in the agreement, but not for mid-sized businesses. Some stakeholders believed it would be very difficult for agencies to define a mid-sized business. One OSDBU director told us that contracting officers would have to perform new market research for mid-sized set-asides and abide by a new layer of requirements. Several stakeholders questioned which agency actually would administer the new set-aside program. Stakeholders identified a few benefits for agencies. Two stakeholders told us agencies could benefit from having a larger supplier base and more choices for services. One OSDBU director said agencies might benefit from retaining former small business contractors for a longer time, and a researcher said agencies might gain access to talent and value they might not get from large businesses. Modifying Rules for Multiple-Award Contracts Could Help Mid-Sized Businesses Stakeholders told us that allowing small businesses that grow beyond their size standards to move to the unrestricted version of multiple-award contracts could help mid-sized businesses. As discussed previously, multiple-award contracts can be unrestricted or restricted to small businesses or have separate tracks for small and nonsmall businesses (such as by using set-aside orders). This option proposes that small businesses on the restricted track of a multiple-award contract that outgrow the contract’s small business size standard be moved to the unrestricted track. This practice can be referred to as “on-ramping.” According to the stakeholder proposing this option, if a small business contractor grew to mid-sized, but could not transition to the unrestricted track, all the effort the business put into winning the contract would be wasted simply because it grew. Some multiple-award contracts allow small businesses that outgrow the size standard to move to the unrestricted track of the multiple-award contract, but this is not always the case. Agencies have discretion when making this determination. If a business is allowed to move to the unrestricted track of such a contract, it would be able to place bids on additional orders resulting from the contract. In cases in which a business cannot move to the unrestricted track, it has to leave the contract after completing any ongoing orders. For example, the General Services Administration’s One Acquisition Solution for Integrated Services allows businesses that have outgrown their size standards to move to the unrestricted track, while EAGLE II does not. Most stakeholders we interviewed said this option could increase contracting opportunities for growing small or mid-sized businesses. An OSDBU director said the option would let businesses that grew to be mid- sized move to the unrestricted pool of the multiple-award contract so they could keep their existing contract. A researcher said the option gives more time for small and mid-sized businesses to prepare for full and open competition. But two stakeholders noted that mid-sized businesses already in the unrestricted pool may be negatively affected by increased competition from additional contractors placed in the pool for task orders. Stakeholders offered differing opinions on how this option would affect agencies. Several stakeholders said that agencies would benefit from being able to retain contractors even if the contractors outgrew their size standard. For example, a trade association executive said it would be less disruptive for the agency if the business could continue its contract. Two stakeholders thought that moving a business to the unrestricted pool would reduce agency time and paperwork (compared to re-competing the contract and performing additional evaluations). However, several stakeholders told us that allowing small businesses that grew beyond the size standards to “on-ramp” might increase administrative burden on agencies. For example, it might take longer for an agency to evaluate proposals for unrestricted task order competitions if the pool of competitors grew. SBA officials expressed concern that if task order competitions grew too large, businesses in the unrestricted pool that objected to the increased competition from new contractors might pursue litigation. Changing Past Performance Requirements Could Increase Contracting Opportunities for Mid- Sized and Small Businesses but May Increase Risk for Agencies Stakeholders told us that changing past performance requirements could increase prime contracting opportunities for mid-sized and small businesses, but might increase risk for agencies. Some of the literature we reviewed considered requirements based on the size or number of past contracts (“quantitative past performance requirements”) as a barrier to entry for mid-sized businesses. Options have been proposed that would regulate what types of past performance contracting officers consider and how they establish solicitation requirements. Lower or Eliminate Quantitative Requirements for Past Performance This option proposes lowering (for example, limiting their use or making their terms more flexible) or eliminating quantitative requirements for past performance. Nearly all the stakeholders we interviewed said that lowering quantitative requirements would increase contracting opportunities for mid-sized businesses, small businesses, or both. For example, one researcher said that mid-sized and small businesses would benefit because the barriers to entry on some large contracts would be lowered. Similarly, an OSDBU director said that smaller mid-sized businesses and small businesses get shut out of contract competitions because they cannot meet the past performance requirements, and lower past performance requirements would give them a chance to compete. One trade association executive also pointed out that this option could help mid-sized and small businesses develop a performance record for future solicitations. More than half of the stakeholders told us that eliminating quantitative past performance requirements entirely also would increase contracting opportunities for mid-sized businesses, small businesses, or both. Specifically, two stakeholders said eliminating these quantitative requirements would enable mid-sized and small businesses without records of past performance to substantiate their qualifications in other non-quantitative ways. Stakeholders noted trade-offs for agencies. Some stakeholders believed lowering past performance requirements would benefit agencies because more contractors would be eligible to bid. For example, one OSDBU director said agencies might receive proposals from businesses that could not have met quantitative requirements but have enough expertise to submit a high-quality proposal. However, some stakeholders said lowering quantitative requirements may increase agency burden, citing a longer evaluation period due to a larger pool of bidders. More than half of the stakeholders said eliminating requirements entirely would increase the burden on agencies, for various reasons. One researcher said agencies might not have staff with the technical expertise to assess bids based on a strictly qualitative evaluation. Additionally, several stakeholders noted challenges for agencies in obtaining qualitative performance information using the Contractor Performance Assessment Rating System, citing rating subjectivity and verification difficulties. Several stakeholders said lowering or eliminating quantitative requirements for past performance would increase the risk to the agency of awarding contracts to firms that cannot successfully complete the project. For example, one OSDBU director pointed out that “conventional wisdom” for contracting officers is that a project’s success rate is higher when a company can meet higher past performance requirements. Another OSDBU director said that qualitative forms of evaluation, such as testimony from another agency, are not sufficient and could put the agency’s project at risk for lack of an objective measure of a contractor’s capabilities. In contrast, the third OSDBU director said that quantitative requirements do not lower the risk to the agency because completing a certain number of contracts is not a guarantee of satisfactory future performance. Require Agencies to Consider Past Performance of Each Company in Team Arrangements This option proposes that contracting agencies be required to consider the past performance of individual companies in team arrangements as opposed to evaluating only the aggregate past performance of contractors in team arrangements. The literature suggested that requiring agencies to consider each team member’s past performance would provide incentives to mid-sized businesses to work together to compete for contracts with past performance requirements that each would not be able to meet individually. According to SBA officials, the Small Business Act already requires agencies to consider the past performance of each participant in a joint venture or team for bundled contracts and multiple-award contracts above a certain dollar threshold. Also, agencies generally consider the relevant past performance information of individual members of a team arrangement in certain situations if they will perform major or critical aspects of the requirement. However, there are other situations—such as when the contract is not specifically for a small business but instead is bid on by a joint venture that includes a small business—in which agencies are not required to consider each team member’s past performance. Stakeholders identified some benefits to a more flexible consideration of past performance for mid-sized businesses. More than half of the stakeholders believed this option would increase contracting opportunities for mid-sized businesses. For example, a trade association executive said mid-sized businesses currently struggle to fulfill past performance requirements, and this would allow them to combine their past performance with another business to qualify for new and larger contract opportunities. However, some stakeholders noted that mid-sized companies probably have won prime contracts. Therefore, they already might have the requisite past performance to bid on a contract. Nearly all the stakeholders we interviewed thought this option would increase contracting opportunities for growing small businesses because they would be able to team with a small or nonsmall business to bid on contracts for which they otherwise would not have the past performance to qualify. One researcher described a dilemma for small businesses: they cannot compete for contracts without past performance, but they cannot get past performance without winning contracts. SBA officials said that businesses prefer that the past performance of each member be considered instead of the past performance of the joint venture, which could be minimal, especially if it was a new joint venture. Stakeholders identified trade-offs for agencies. More than half of stakeholders said considering past performance of both members in a team arrangement would benefit agencies because more contractors could meet requirements to bid. One OSDBU director said that this option also might allow agencies to benefit from the enhanced capacity and innovative solutions offered by mid-sized businesses. However, several stakeholders cautioned that this option could increase risk for an agency. For example, one trade association representative believed that because team arrangements are the companies’ creation and the government has no involvement in administering them, there is more risk to the agency that the contracting team might not be able to complete the contract. Consider Subcontracting Past Performance in Contract Evaluation This option proposes that agencies should consider subcontracting past performance when evaluating bid proposals. It has been suggested that this could be done in two ways. First, agencies could be required to consider a business’s past performance as a subcontractor—a route for many small businesses to gain access to federal contracts—when competing for prime contracts. Second, agencies could be required to count the past experience of both the prime contractor and its significant subcontractors towards a solicitation’s past performance requirements. Stakeholders had differing opinions on whether this option would help mid-sized businesses. Several stakeholders believed that allowing mid- sized businesses to leverage their subcontracting experience to meet requirements would increase contracting opportunities for these firms. Furthermore, one researcher thought mid-sized businesses could secure more subcontracting opportunities because large firms might be more willing to team with them. However, some stakeholders believed this option would not increase contracting opportunities for mid-sized businesses, with two stating that this option is less important for them because they likely outgrew their size standard by winning set-aside prime contracts. An OSDBU director thought it also might increase competition from smaller firms. Nearly all of the stakeholders we interviewed said this option could increase prime contracting opportunities for growing small businesses. For example, one OSDBU director said that considering subcontracting as past performance would help small businesses compete for prime contracts, grow, and move forward. A researcher noted this could help small businesses transition to mid-sized. SBA officials similarly stated that small businesses want agencies to consider their subcontracting past performance so they can access contracts for which they would not normally qualify. In May 2019, SBA officials said they were working on implementing legislation that requires SBA to create a pilot program to provide past performance ratings for small business subcontractors. Similar to previous options, stakeholders contrasted the benefits and drawbacks of increased competition for agencies. Several stakeholders thought this option would expand the pool of bidders, making contracts more competitive and bringing more value to agencies. However, four stakeholders noted that verifying prime and subcontracting experience could create more work for agencies. Specifically, two of the four noted that it could be difficult for agencies to determine subcontracting past performance because the ratings in the Contractor Performance Assessment Rating System are tied to the prime contractors. Some stakeholders also noted this option could increase the risk to agencies that projects would not be completed successfully. Specifically, because subcontracting agreements are between the subcontractor and prime contractor, and therefore are not enforceable by agencies, a prime contractor might not use a subcontractor whose past performance was considered during the evaluation process. For example, a researcher pointed out that a prime contractor could use the subcontractor’s experience to win a contract, but then not use the subcontractor for any of the work. Modifying SBA’s Size Standards Could Help Some Small Businesses Transition to Mid-Sized Several stakeholders noted that modifying SBA’s size standards would not help mid-sized businesses as such; rather, the modifications could allow a few mid-sized businesses to become eligible for small business set-aside contracts again and help growing small businesses prepare for the transition to mid-sized. It has been noted that agencies increasingly use large, multiple-award contracts that can cause small businesses to outgrow their size standard before they build the capacity (financial resources, business infrastructure, or past performance records) they need to successfully compete for contracts. To address such issues, options have been proposed to modify SBA’s size standards. Change the Calculation for Revenue-Based Size Standards This option would change the number of years of revenue considered when applying revenue-based size standards. SBA would allow businesses to consider their past 5 years of revenue, pick the lowest 3 years in that period, and average them to determine if they met revenue- based size standards. In a December 2018 amendment to the Small Business Act, Congress extended the number of years of revenue that service businesses use to calculate their size from 3 to 5 years but included no provision related to selecting lowest-revenue years. Stakeholders expressed reservations regarding this option for mid-sized businesses. Some stakeholders said that this option may not increase opportunities for mid-sized businesses because they already had outgrown their size standard. Several stakeholders also said the option offered only a temporary solution. One noted that this option would delay “graduation” from the size standard, but would not address the issue that mid-sized businesses need to continue to grow to secure additional federal contracts. An OSDBU director told us it is critical that small businesses develop and execute a marketing and business plan to transition from small to successful mid-size. Nearly all the stakeholders we interviewed noted that allowing businesses to choose their lowest 3 years of revenue in a 5-year period could prevent an outlier revenue year from causing a small business to prematurely outgrow its size standard. For example, an OSDBU director said that a large, 1-year award is not indicative of a business’s revenue over the long term. Nearly all of stakeholders also said that enabling businesses to choose the lowest 3 years of revenue would help ease the transition to mid-sized. For instance, an OSDBU director said businesses could stay below the size standards for longer and establish a performance record to help secure future contracts. However, several stakeholders expressed concern that very small businesses might lose contracting opportunities due to increased competition (that is, more and larger-sized firms would remain under the size standards). Finally, SBA officials pointed out that this option could be perceived as unfair because it would not benefit businesses in industries with employee-based size standards. Subtract Research and Development Expenses This option proposes that businesses be able to subtract research and development expenses from their total revenue when calculating their eligibility for small business status. The stakeholder proposing this option said that businesses close to the size standard have to focus their revenue on pursuing contracts that will support their company as they transition to full and open competition, and so cannot spare money to further invest in researching and developing new products or processes that might improve their business. In addition to encouraging more investment in research and development, subtracting these expenses would lower revenue and allow some mid-sized businesses to be classified as small again. One researcher told us this was a strong option for mid-sized businesses, particularly information technology businesses, because research and development investment is such a large part of their expenses. However, some stakeholders said this option would not increase contracting opportunities for mid-sized businesses. For example, one OSBDU director believed this option would not benefit mid-sized businesses because it did not help these businesses to compete with larger businesses. More than half of the stakeholders we interviewed said that this option might encourage small businesses to invest in research and development. However, several stakeholders noted that this option only would help the small percentage of small businesses that perform research and development. SBA officials pointed out that this option could be perceived as unfair because not all businesses have research and development expenses. They also pointed out that modifying revenue calculations would not benefit manufacturing businesses, which invest more in research and development than other sectors but primarily use employee-based size standards. Stakeholders also noted potential trade-offs for agencies. Several stakeholders told us this option would benefit the government by encouraging investment in research and development with one stakeholder stating that it might result in higher-quality bids. However, several stakeholders and SBA officials also told us that allowing small businesses to subtract research and development expenses would increase the administrative burden on agencies or add too much complexity. For example, two OSDBU directors said it would be difficult for an agency to verify that research and development expenses were correctly claimed and subtracted from revenue. SBA officials noted that there were no industry-by-industry data on research and development expenses. Some stakeholders observed this option could lead to an increase in fraud or manipulation, with one trade association executive saying the option would not increase innovation, just claimed expenses. Raise Revenue-Based Size Standards This option proposes increasing SBA’s revenue-based size standards. The trade association representative who proposed the option believed that small business size standards should be raised so that high revenue- generating small businesses that still are not dominant in their field would not be shut out of set-asides. Increasing revenue-based size standards would benefit some mid-sized businesses by making them eligible again for small business set-asides. However, more than half of the stakeholders told us the option would have a limited impact—it would apply only to the mid-sized businesses small enough to fall under the newly raised standard—or no impact at all (for most other mid-sized businesses). To illustrate the limited impact, one OSDBU director used the example of management consulting services (NAICS 541611), which has a size standard of $15 million (revenue). If the standard were increased to $17 million, it might not affect many businesses. Rather, it would help only the $16 million company to compete for set-asides again, the director said. Furthermore, one researcher said that the increase would not address the systemic disadvantage that mid-sized businesses face in competing with large businesses. Stakeholders identified tradeoffs for small businesses related to this option. Nearly all stakeholders said that raising revenue-based size standards could help growing small businesses better prepare to transition to mid-sized while remaining eligible for set-asides. One OSDBU director said small businesses could add to their performance record and have more time to become competitive with larger businesses. One researcher said that small businesses could get additional time to diversify contract portfolios and fund professional certifications. However, some stakeholders cautioned the option could harm very small businesses because as one stakeholder explained, there would be more competitors for small business set-asides. As noted previously, we found that most small businesses awarded set-aside contracts in 2017 were well below the size standards. Stakeholders also identified trade-offs for federal agencies. Several stakeholders said agencies would benefit from the increased competition. For example, one OSDBU director said agencies might have more bidders, which could lower pricing. Some stakeholders said the option could help agencies reach small business goals more easily because more businesses would be considered small. However, a stakeholder advised that agencies also might need more time and resources to evaluate an increased number of bids. SBA officials explained to us that they comprehensively review all the size standards every 5 years, looking at factors, such as industry trends and small business market share. They contended that if contracts became larger to the detriment of small businesses, small businesses then would have a decreased market share. If small businesses were losing market share, that would be captured by the SBA size standard methodology and the size standards would be adjusted accordingly. They also noted that revenue-based size standards were getting higher and higher as a result of adjustments during SBA’s reviews and adjustments for inflation and that further increases might allow firms that were dominant in their industry to be small, which is contrary to statute. Finally, they stated that just raising size standards without taking into account industry structure and market conditions would enable more experienced businesses to qualify as small and hurt small businesses that need federal assistance the most, especially in competing for set-aside contracts. Agency Comments and Our Evaluation We provided a draft of this report to the Department of Homeland Security, Department of Defense, General Services Administration, and SBA for their review and comment. The Department of Homeland Security provided technical comments, which we incorporated where appropriate. In emails, the OSBDU director at the Department of Defense and an audit liaison at the General Services Administration stated that the agencies did not have any comments. SBA provided technical comments in an email from the GAO Liaison, which we incorporated as appropriate. We considered a number of these comments to be more than technical in nature and therefore, summarize them here: SBA offered new views on three specific options for increasing federal contracting opportunities for mid-sized businesses presented in the report, which we incorporated where appropriate. SBA also made the larger point that they believe any option to help mid-sized businesses would hurt small businesses. In discussing the various options in the report, we present the views of various stakeholders and SBA on how the options would affect small businesses. SBA stated that we created our own methodology for determining a mid-sized business by multiplying the current size standards and that a formal study should be performed to establish a baseline definition of a mid-sized business. Our goal was not to establish a baseline definition of a mid-sized business. As we note in the report, there is no statutory or regulatory definition of a mid-sized or large business. We applied multipliers to SBA’s size standards only for the purposes of our analysis—specifically, to identify businesses that had outgrown small business size standards and continued to receive federal contracts. SBA stated that we multiplied size standards by a factor of five to define mid-sized businesses in all industries and cited analysis that it had done that indicated that more than 95 percent of businesses are at or below SBA’s size standards. The agency concluded that this means that in some industries, almost all firms would be considered mid-sized under our definition of mid-sized. However, only those firms with revenue or employees up to five times above the SBA small size standard would be considered mid-sized in our analysis. We counted any businesses with revenue or employees at or below the small size standard as small. SBA stated that considering a factor of two or three times the SBA size standards to identify mid-sized businesses would improve our results. We considered a number of different factors when developing our methodology. As noted in the report, we used five times the small size standard to distinguish between mid-sized and large businesses based on the distribution of contracts and obligations among businesses in these two groups. SBA stated that (1) the report should explain the basis and method for selecting the sample of 5,339 businesses awarded set-aside contracts in 2008 and (2) a sample of 104 out of 5,339 firms over that period of time was too small to be generalizeable. The 5,339 businesses awarded set-aside contracts in 2008 and awarded any sort of federal contract in 2013 were not a sample; rather, they were all the businesses that met these criteria. Therefore, we did not generalize to the population based on a sample. Our analysis showed that only 104 of these 5,339 businesses grew to mid-sized by 2013. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the Secretary of Defense, the Acting Secretary of Homeland Security, the Administrator of the General Services Administration, and the Acting Administrator of SBA. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact William B. Shear at (202) 512-8678 or shearw@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Major contributors to this report are listed in appendix III. Appendix I: Objectives, Scope, and Methodology This report analyzes (1) the extent to which small businesses grew to be mid-sized and continued to receive federal contracts; (2) instances in which mid-sized businesses can perform work on contracts set aside for small businesses; and (3) options for increasing federal contracting opportunities for mid-sized businesses and views on the strengths and limitations of the options. We present information on contracts awarded to small, mid-sized, and large businesses in fiscal year 2017 in appendix II. For the first objective, we analyzed data from the Federal Procurement Data System-Next Generation (FPDS-NG) for fiscal years 2008 through 2017 (the most recent complete data available when we began our review). For a consistent set of data across the 10-year period, we selected all contracts awarded above the micro-purchase threshold. To determine the relevant contract size standard, we matched the size standards data from the Small Business Administration (SBA) to the contract awards data from FPDS-NG based on the year of award and the contract’s North American Industry Classification System (NAICS) code. We then compared the businesses’ annual revenue or number of employees to SBA size standards. We assessed the reliability of the FPDS-NG data we used by performing electronic testing of selected data elements and reviewing existing information about FPDS-NG and the data the system produces. We determined that these data were sufficiently reliable for the purposes of determining the extent to which small businesses that grew into mid-sized businesses continued to receive federal contracts and the size of businesses awarded contracts during a specific time period. To determine the extent to which contracts were set aside for small businesses, we calculated the percentage of new contracts awarded in fiscal year 2017 that were small business set-asides. To determine the size of the small businesses awarded these set-aside contracts, we divided SBA’s size standards into four segments for each NAICS code— below or at one-fourth of the size standard, above one-fourth to one-half of the size standard, above one-half to three-fourths of the size standard, and above three fourths of the size standard—and determined the number and obligations of set-aside contracts awarded to small businesses in each quartile for fiscal year 2017. We completed this same analysis for fiscal years 2013 through 2016 to see if the results were similar. To determine the extent to which small businesses grew to be mid-sized and continued to receive federal contracts, we used FPDS-NG data from fiscal years 2008 through 2017. Because there is no statutory or regulatory definition of a mid-sized or large business, we applied a number of multipliers to determine size. Businesses with revenues or employees at or below the SBA small size standards were small. We considered businesses with revenue or employees up to five times above the SBA size standard as mid-sized businesses. We considered businesses with revenue or employees more than five times the size standard as large businesses. We used five times the small size standard to distinguish between mid-sized and large businesses based on the distribution of contracts and obligations among businesses in these two groups. We discussed this approach and methodology with SBA officials and officials at three federal agencies that had large obligations for small business contracts in fiscal year 2017. These officials did not raise any questions about our approach, and some reiterated that there was no legal definition of mid-sized businesses. Using these definitions, we selected businesses awarded only small business set-aside contracts in fiscal year 2008 and determined whether these businesses also were awarded any type of federal contract in fiscal year 2017 and if they were in the same or different size category in fiscal year 2017. We then determined the number of businesses awarded set- aside contracts in fiscal year 2008 and awarded any sort of federal contract in 2013 that had become mid-sized in fiscal year 2013 and the extent to which those businesses were awarded any contracts in subsequent years. We also determined the percentage of competed contracts awarded to small, mid-sized, and large businesses in fiscal year 2017. For purposes of this report, competed contracts are those competed using (1) full and open competition, (2) full and open competition after exclusion of sources, and (3) simplified acquisition procedures. To determine the industry sectors with the largest number of set-aside and competed contracts in fiscal year 2017, we collected and analyzed FPDS data for each of the two-digit NAICS industry sectors. See appendix II for more information. For our second objective, we reviewed the Federal Acquisition Regulation and small business laws and regulations to identify provisions that allow small businesses that grow into mid-sized businesses to continue providing services and goods on contracts set aside for small businesses. We reviewed SBA documentation related to its 8(a) and All Small Mentor- Protégé programs because forming joint ventures with small businesses under these programs is one way that mid-sized businesses can provide services and goods under set-aside contracts. We analyzed lists from SBA of the businesses that entered into mentor-protégé agreements as of July 2018 and the mentor-protégé agreements that had formed joint ventures as of September 2018. Using FPDS-NG data, we determined the number of joint ventures formed under the All Small Mentor-Protégé program that had been awarded set-aside contracts from fiscal years 2016 through 2018. We began with 2016 to allow time after the program was created in 2013 for businesses to enter into agreements and form joint ventures. We ended with 2018 because it was the most recent complete year of data available when we conducted this analysis. Using the same multiplier methodology designed for our first objective, we determined the number of mentors awarded set-aside contracts as part of a joint venture that were mid-sized businesses. We assessed the reliability of the SBA and FPDS-NG data we used by interviewing SBA officials about their data and performing electronic testing. We determined that these data were sufficiently reliable for determining the number of mentors awarded set-aside contracts as part of a joint venture that also were mid-sized businesses. We were not able to perform a similar analysis for joint ventures formed under SBA’s 8(a) Mentor-Protégé program because SBA does not maintain a Data Universal Numbering System number for mentors participating in that program. This number is needed to determine the size of the mentor. To identify stakeholder views on options for increasing federal contracting opportunities for mid-sized businesses, we identified a number of proposed options by reviewing literature, including sources identified during our background research and initial interviews. We also conducted a literature search. We used ProQuest to search 13 databases—including Business Premium Collection, EconLit, Global Newsstream, Policy File Index, and ProQuest Dissertations and Theses Global Research Library. We also conducted searches using Lexis Advanced, EBSCO Business Source Corporate Plus, Dialog, DTIC, Scopus, and HeinOnline. The search was limited to 11 years (2008–2018) and to scholarly, trade, think- tank, and government publications. For the searches, we used keywords such as “advanced small businesses,” “federal contracting,” “mid-tier/mid-sized small businesses,” “middle market,” “IDIQ,” “challenges,” and “opportunities.” Our searches yielded 199 sources. To select relevant sources, an analyst reviewed the titles and abstracts and selected 21 as likely to propose options for increasing federal contracting opportunities for mid-sized businesses. A second analyst reviewed the first analyst’s selection for concurrence. The 21 sources we selected included trade association reports, congressional testimonies, and research reports. One analyst read the 21 sources and identified any specific options discussed. We eliminated suggestions or recommendations that were unclear, duplicative, or unconnected to mid-sized businesses. A second analyst read the same sources and verified that the first analyst had correctly identified all the options pertaining to increased contracting opportunities for mid-sized businesses. From this analysis, we compiled a final list of 14 options for which we would obtain stakeholder views. The options selected were grouped into four categories. The list of options included in the report is not exhaustive; the options are intended only to be illustrative of potential approaches to enhancing contracting opportunities for mid-sized businesses. To obtain stakeholders’ views regarding the strengths and limitations of these options, we selected three categories—trade associations, researchers, and federal agencies—from which to develop a nongeneralizable sample of individuals to interview. To identify trade associations, we compiled a list of 20 trade associations that represented small and mid-sized businesses from the literature search and previous GAO work on small business contracting. We searched each organization’s website for any publications the organization may have published on small and mid-sized businesses and federal contracting. We used search terms such as “federal contracting,” “mid-size,” and “size standards.” We identified six trade associations using this process. Because one did not respond to our request, our sample included representatives of the remaining five associations. We also selected three researchers who published on mid-sized businesses and federal contracting. In addition, we selected the directors of the Offices of Small and Disadvantaged Business Utilization (OSDBU) at three federal agencies—Department of Defense, Department of Homeland Security, and General Services Administration—that were among the top five agencies in terms of total dollar obligations for small business contracts in fiscal year 2017. We then interviewed the 11 stakeholders. For each interview, we asked them to provide their views on the strengths and limitations of each option in relation to small, mid-sized, and large businesses and for federal agencies. We performed a content analysis to analyze the responses. First, we created preliminary codes that represented key themes across the interviews of the strengths and limitations of the 14 options, such as “would increase administrative burden on agencies” or “would increase contracting opportunities for mid-sized businesses.” A methodologist reviewed the coding system to ensure it was logical. We pre-tested the coding of responses from three interviews to ensure the appropriateness of the codes. One analyst coded each response to a particular strength or limitation of an option and a second analyst reviewed the coding. If a response did not align with a strength and limitation, the response was coded as “unclassified.” The team discussed the results of the initial coding analysis and made some adjustments to the codes. Once the coding scheme was finalized and the responses from the remaining eight interviews were coded by an analyst, a second analyst reviewed the coding. If the second analyst disagreed with the coding of a particular response, the two analysts spoke and achieved concurrence. After response coding was completed, we tabulated the responses based on the codes. A second person verified the calculation of the stakeholders’ response totals. Because we selected a nongeneralizable sample of stakeholders to interview, their views are not generalizable to other stakeholders who have knowledge about options for increasing contracting opportunities for mid-sized businesses, but their views offered important perspectives. To characterize the number of stakeholders who offered the same opinion, we used “nearly all” for nine or 10 stakeholders, “most” for seven or eight stakeholders, “more than half” for six stakeholders, “several” for four or five stakeholders, and “some” for three stakeholders. We also interviewed SBA officials to obtain their views on how the options might affect small businesses, as well as to gather information related to our other two objectives. We conducted this performance audit from April 2018 to August 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Information on Contracts Awarded to Small, Mid-Sized, and Large Businesses in Fiscal Year 2017 In this appendix, we provide information on contracts awarded to small, mid-sized, and large businesses in fiscal year 2017. Because there is no statutory or regulatory definition of a mid-sized or large business, we applied multipliers to the Small Business Administration’s (SBA) size standards. Small businesses were those with revenue or employees at or below the size standard for their industry. We considered mid-sized businesses as those with employees or revenue up to five times above the size standard and large businesses as those with employees or revenue more than five times the size standard. Percentage of Competed Contracts Awarded in Fiscal Year 2017 by Size Our analysis of Federal Procurement Data System-Next Generation (FPDS-NG) data showed that mid-sized businesses received the smallest share—9 percent—of competed contracts (compared with small and large businesses) in fiscal year 2017 (see fig. 3). For purposes of this report, competed contracts are those competed using (1) full and open competition, (2) full and open competition after exclusion of sources, and (3) simplified acquisition procedures. Dominant Industry Sectors in Federal Contracting We analyzed FPDS-NG data to determine the number of set-aside and competed contracts awarded in fiscal year 2017 by industry sector. In fiscal year 2017, the largest number of set-aside contracts were awarded in the following sectors: manufacturing; professional, scientific, and technical services; and construction (see fig. 4). In that same year, the largest number of competed contracts were awarded in the following sectors: manufacturing; professional, scientific, and technical services; and wholesale trade. Similarly, in fiscal year 2017 the largest contract obligations (set-aside and competed) were awarded in the sectors of construction; manufacturing; and professional, scientific, and technical services (see fig. 5). For competed contracts, we analyzed FPDS-NG data to determine if the size of businesses awarded contracts varied by industry sector. In fiscal year 2017, small and large businesses were generally awarded more competed contracts than mid-sized businesses, regardless of sector (see table 5). The industry sectors in which small and large businesses were awarded the most competed contracts in fiscal year 2017 were manufacturing; professional, scientific, and technical services; and wholesale trade. Similarly, in fiscal year 2017, mid-sized businesses were awarded the most competed contracts in the manufacturing and professional, scientific, and technical services sectors. The third dominant sector for mid-sized businesses was information. Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact name above, Paige Smith (Assistant Director), Nancy Eibeck (Analyst in Charge), Edward Chiu, Sarah Garcia, Julia Kennon, Jill Lacey, Barbara Roesmann, Jessica Sandler, and Jena Sinkfield made significant contributions to this report.

Why GAO Did This Study Small businesses that receive federal contracts set aside for them may outgrow the size standards the Small Business Administration (SBA) uses to define small businesses. (Size standards vary by industry and generally are based on employees or revenue.) Questions have been raised about the extent to which mid-sized businesses can compete with large businesses for federal contracts. GAO was asked to provide information on federal contracting opportunities for mid-sized businesses. This report analyzes, among other objectives, (1) the extent to which small businesses grew to mid-sized and continued to receive federal contracts and (2) options for increasing contracting opportunities for mid-sized businesses. GAO analyzed federal contracting data for fiscal years 2008–2017 (most recent and complete). In the absence of legal definitions of “mid-sized” and “large,” GAO multiplied relevant size standards for small businesses to arrive at parameters for mid-sized and large businesses for its analysis. GAO reviewed literature to identify options for increasing contracting opportunities and interviewed SBA officials and a nongeneralizable selection of 11 stakeholders—trade association representatives, researchers, and small business directors at three agencies with large obligations for small business contracts in fiscal year 2017—to obtain views on the options. SBA provided comments, which we addressed as appropriate. What GAO Found From fiscal year 2008 through 2017, very few small businesses that were awarded limited competition (set-aside) contracts grew to be mid-sized and continued to receive contracts. (GAO defined mid-sized businesses as having revenue or employees up to five times above the small business size standard.) Of the 5,339 small businesses awarded set-aside contracts in fiscal year 2008 and awarded any sort of federal contract (including set-aside or competed) in 2013, 104 became mid-sized by fiscal year 2013. Of those 104 businesses, 23 remained mid-sized through 2017 and won 75 contracts. Another three businesses became large and won six contracts. Options for increasing federal contracting opportunities for mid-sized businesses that GAO identified in its review include establishing a separate set-aside category, changing consideration of past contracting performance, and modifying size standards. Stakeholders told GAO some options would help mid-sized businesses more than others. While a set-aside category for mid-sized businesses would increase opportunities for mid-sized businesses, stakeholders generally believed it could decrease opportunities for small businesses and increase agency burden (time and costs to implement the set-aside). Requiring agencies to consider businesses' past performance as subcontractors or as part of a team would help both mid-sized and growing small businesses by making them more competitive for contracts. Stakeholders said raising size standards based on revenue would allow a limited number of mid-sized businesses to be eligible for set-asides again, but not help the vast majority of mid-sized businesses.

gao_GAO-19-305

gao_GAO-19-305_0

Background The NDAA requires DHS to develop and implement 43 border security metrics in four domains—between POEs, at POEs, the maritime border, and air and marine security in the land environment. Within DHS, CBP and the Coast Guard have primary responsibility for border security within these four domains. CBP and its subcomponents are to secure U.S. borders at and between POEs by preventing inadmissible people and illicit goods from entering the United States, among other responsibilities. Within CBP, the primary offices and components involved in border security are the Office of Field Operations at POEs, Border Patrol between POEs, and Air and Marine Operations for air and marine security in the land and maritime domains. The Coast Guard and CBP’s Air and Marine Operations share responsibility for security of the nation’s maritime borders. Table 1 shows examples of border security metrics by domain and responsible DHS component. According to DHS officials, within DHS, two subcomponents within the Office of Strategy, Policy, and Plans were responsible for coordinating the department’s effort to develop the fiscal year 2017 Border Security Metrics Report. A senior DHS official explained that the report was initially tasked to the Unity of Effort Integration Office, which was part of the Unity of Effort initiative started in 2014 to better understand border security efforts along the southwest border including exploring the development of border security metrics. OIS assumed responsibility for the report in 2017. According to OIS officials, to prepare the report, they obtained data and information related to each NDAA metric from the administrative records of the DHS components with primary responsibilities for border security in the four domains. For example, OIS requested data and information on “turn backs” and “got aways” from Border Patrol—the lead component for the between POE domain—which records sector estimates of turn backs and got aways based on direct and indirect observations. Of the 43 metrics the NDAA listed for inclusion in the Border Security Metrics Report, the majority were counts and rates of border security activities. The remaining metrics were estimates or were not specifically described. For example, the number of apprehensions in each Border Patrol sector is a count metric. In contrast, a rate metric compares one value or number against another. For example, the wait time ratio compares the average wait times to total commercial and private vehicular traffic volumes being processed at a land POE. An estimate is used for metrics of flows or activities that are largely undetected and therefore cannot be measured directly and must be estimated, such as the number of undetected unlawful entries. A few metrics are a combination of counts or rates with an estimate. For example, the metric for total inadmissible travelers at POEs counts known inadmissible travelers that are intercepted at POEs, and also requires an estimate of how many inadmissible travelers may have successfully entered at a POE without being detected, which cannot be directly measured. The NDAA did not specifically describe some metrics. For example, while the NDAA asked for an examination of each of the eight consequences under the Consequence Delivery System, it did not specify how this examination was to be carried out or what it was to include. While many of the metrics required by the NDAA can be addressed with data from DHS’s administrative records, certain metrics that rely on estimates necessitate the use of alternative methodologies and in some cases, specialized technical expertise. For example, DHS contracted with the Institute for Defense Analyses to assist with the development of a statistical model for estimating undetected unlawful entries. In its fiscal year 2017 Border Security Metrics Report, DHS provided information on its methodological approaches, such as how it estimated undetected unlawful entries. DHS Reported Information on Most Required Metrics and Generally Used Quality Information but Did Not Identify Some Limitations DHS Reported Information on 35 of the 43 Required Metrics and Generally Included Elements Listed in the NDAA In its first Border Security Metrics Report, DHS reported information on 35 of the 43 metrics called for by the NDAA. The metrics DHS provided spanned the four domains outlined in the NDAA and included a mix of counts, rates, estimates, or a combination thereof as shown in figure 1. For 18 of the 35 border security metrics DHS included in its report, we found DHS generally included elements listed in the NDAA. For example, the NDAA asked for the number of detected unlawful entries between POEs, and in its report DHS provided information on the number of detected unlawful entries over a 10-year period. As another example, the NDAA asked for the number of cargo containers at sea ports that were identified to be potentially high-risk. In response, DHS provided information on the number of potentially high-risk containers from fiscal years 2013 through 2016 and also provided contextual information about trends in the volume of such containers over time. See table 2 for more information on these examples as well as other examples of the types of information included in DHS’s fiscal year 2017 Border Security Metrics Report. For some metrics, DHS also provided information in addition to the elements listed in the NDAA. For example, the NDAA described the “AMO apprehensions assisted” metric as a count of the number of apprehensions that were assisted by CBP’s AMO through the use of unmanned aerial systems and manned aircraft. In addition to the counts for such assists, DHS also provided data on the flight hours expended to assist with these apprehensions. For 17 of the 35 reported metrics, we identified differences between the metric as described by the NDAA and as reported by DHS. The differences we identified generally fell into two categories: Metric differed in scope or calculation. Some of the metrics DHS reported on differed in scope or in their calculation from what the NDAA described for reasons such as data availability, among other factors. For example, DHS’s fiscal year 2017 Border Security Metrics Report scoped three metrics on unlawful border crossings between POEs (the “attempted unlawful border crosser apprehension rate,” the “estimated undetected unlawful entries,” and the “probability of detection rate”) to only include data for the southwest border. In these instances, the report noted that a methodology for estimating data on unlawful crossings for the northern border had not yet been completed but that research was underway to do so. As an example of a difference in calculation, DHS presented the interdiction effectiveness rate for each southwest border sector as an alternative to the metric “unlawful border crossing effectiveness rate in each Border Patrol sector.” According to DHS’s report, the department used the interdiction effectiveness rate because it had not yet produced and validated sector-level estimates of unlawful entries required to calculate the unlawful border crossing effectiveness rate. In its report, DHS stated it expects these estimates to be available for the 2019 report. Alternative metric provided. For the situational awareness in the maritime environment metric, DHS stated that it is in a multi-year process to develop a metric that meets the intent of the NDAA. As an alternative, DHS instead provided data on the number of aircraft and vessel operational hours that contributed to maritime domain situational awareness. See appendix I for additional information about any differences we identified for each metric. The eight metrics on which DHS did not provide information spanned all four domains. In its report, DHS explained that the eight omitted metrics were either still in development, under review within the department, or officials were in the process of collecting data for them. Table 3 lists the eight metrics on which DHS did not provide information and the date DHS estimated it will report on each metric. DHS Components Generally Have Processes to Help Ensure Reliable Data and Quality Information, but DHS Does Not Have a Systematic Process for Reviewing the Reliability of Data to Identify Limitations In general, DHS components responsible for collecting the data used in the metrics DHS reported have processes to help ensure the reliability of the data and the quality of the information provided. DHS also identified and disclosed limitations with some of the data elements or methodologies used for the metrics in its report. However, DHS does not have a systematic process for reviewing the reliability of data to identify limitations related to the metrics, and we identified at least one additional limitation for 21 of the 35 metrics on which DHS reported where DHS did not disclose such limitations or could have been more transparent about the limitations or assumptions in its report. Data are considered reliable when they are reasonably free from error and bias. Quality information is derived from relevant and reliable data and is considered to be, among other things, complete, accurate, and timely. The specific processes DHS components use to ensure data reliability vary from metric to metric. Examples of processes DHS or its components have implemented to help ensure the reliability of the data and the quality of information provided include: Issuing guidance and monitoring implementation. In September 2012, Border Patrol headquarters officials issued guidance to help provide a more consistent, standardized approach for the collection and reporting of turn back and got away data by Border Patrol sectors. Each sector is individually responsible for monitoring adherence to the guidance. According to DHS’s report, command staff at Border Patrol stations ensure agents are aware of and utilize proper definitions for apprehensions, got aways, and turn backs at their respective stations and also ensure that the necessary communication takes place between and among sectors and stations to minimize double-counting when subjects cross over multiple areas of responsibility. Supervisory reviews of data entries. With regard to data on AMO vessel and aircraft missions, AMO guidance mandates that supervisors perform a review of all pre- and post-mission data entries to help ensure accurate entry of mission information. AMO officials confirmed that supervisors review the data being entered into the database. Additionally, officials said AMO data teams run monthly validation checks of data entered to check for completeness and accuracy, such as out-of-range values. Using built-in electronic safeguards. CBP’s databases for entering and maintaining data elements—including travelers or passengers seeking admission, known inadmissible aliens at POEs, referrals for secondary examinations, major infractions, and private vehicles processed at a POE—have built-in processes to detect and prevent potential data entry errors. More specifically, as an officer enters a record, the systems check for valid entries into relevant fields and provide an error message to the officer for entries that appear to be invalid (e.g., if an officer leaves a mandatory field blank or enters contradictory information such as charging an individual with a crime while also entering a request for expedited removal). In some cases, the systems will prevent a record from being saved if any required fields are blank. Comparing data against other sources. As part of the Coast Guard’s data reliability processes for data on maritime migrant interdictions used in the “known maritime migrant flow rate” metric, Coast Guard officials said that analysts cross-check the data entered into their database with other Coast Guard reporting documents, such as internal spreadsheets, to ensure accuracy. Independent assessment of performance measure data. Some border security metrics are similar to, or use the same data elements as, performance measures DHS reports annually in response to the Government Performance and Results Act Modernization Act (GPRAMA) of 2010. For those performance measures, DHS annually assesses a subset of measures and their data for completeness and reliability using independent review teams. For example, in May 2017 an independent review team assessed the “migrant interdiction effectiveness in the maritime environment” performance measure, which uses the same data as the border security metric, “known maritime migrant flow rate.” The review team found the measure to be complete and reliable and the data to be of good quality overall, but also recommended that the Coast Guard and DHS continue work on an improved database to enhance the consistency of data collection, among other things. In addition to the components having processes to help ensure the reliability of the data and the quality of the information used in the report, DHS took steps to be transparent in its presentation of the metrics by identifying and disclosing known limitations with some of the data elements or methodologies used for the metrics in its report. Communicating the extent to which such limitations exist and their potential impact is important to help facilitate the appropriate use and understanding of the data and the metrics. DHS identified and disclosed limitations related to the potential for misclassification of observations, the potential for cases not being entered or recorded correctly, and methodological limitations, among other things. For example, one of the key limitations DHS’s report identified for the data on turn backs and got aways is that they are based on potentially subjective observations of agents who have to make a determination on how to classify them based on what they observed or the available evidence (e.g., tracks, sensor activations, interviews with apprehended subjects, camera views, etc.). Further, DHS’s report explained that agents may face challenges in making that determination because some unlawful border crossers may enter the United States to drop off drug loads or to act as decoys to lure agents away from a certain area and then return to Mexico, and therefore may be misidentified as turn backs, for example. As another example, DHS identified limitations due to cases not being entered or recorded correctly. For the “known maritime migrant flow rate” metric, DHS used data on the total number of maritime migrants interdicted. In its report, DHS explained that a potential limitation of this data element is that the Coast Guard relies on international and domestic partners to report their interdictions for compilation in its database. Consequently, the accuracy and completeness of the data depend on whether those reports are made by those partners and the accuracy of their reports. See appendix I for additional information about the limitations identified for each metric. Even as DHS identified and disclosed limitations related to some of its metrics, we identified at least one additional limitation for 21 of the 35 metrics on which DHS reported where DHS did not disclose such limitations or could have been more transparent about the limitations in its report. Examples of such instances include: Potential for cases not being entered or recorded correctly. In our previous work we found that mission data for unmanned aerial systems were inconsistently collected across operation locations. Specifically, in February 2017 we reported that there were instances where no assist information was recorded in AMO’s data system even though such assets participated in investigations and operations. Because AMO’s data may not reflect all asset assists, we recommended that AMO update and maintain guidance for recording mission information in its data collection system and provide training to users of the system. For its fiscal year 2017 Border Security Metrics Report, DHS used asset assists data in metrics such as the “AMO individuals detected,” “AMO apprehensions assisted,” and “illicit drug seizures assisted by AMO,” but did not disclose this limitation in its report. Potential for data to be changed over time. Border Patrol officials told us that data on the apprehension of unaccompanied alien children may change over time because original apprehension records from a shared database have, in some instances, been updated by staff from U.S. Immigration and Customs Enforcement (ICE) Enforcement and Removal Operations (ERO). Officials said that in January 2015 they noticed that ERO staff were inadvertently overwriting Border Patrol’s original data entries about the status of apprehended children when they made updates to those children’s records. For example, if a child was unaccompanied at the time of his or her apprehension and was recorded as such by Border Patrol in the initial record entry, ERO may have changed the “unaccompanied” status in the system after they matched the child with a family member or sponsor. As a result, data may not be reconcilable with initial apprehension counts over time. DHS did not fully disclose limitations for some metrics. We identified instances where DHS could improve transparency about the assumptions or limitations of the data presented in its report. For example, in 2014 Border Patrol implemented a standard, southwest border-wide methodology to improve reporting of turn backs and got aways. While DHS made mention of this change in the text of the report, the data for these metrics are presented in tables without any table notes or disclosures within the table about this change. Further, DHS’s report does not discuss how the change may affect comparability of the data. Consequently, a reader may not be aware that data for before 2014 in a table are not necessarily comparable to the data for 2014 and after in the same table. Without a comprehensive identification of the limitations of the metrics and their associated data, and without an adequate disclosure of those limitations, the value of DHS’s report as a source of information to Congress, policymakers, and the public may be diminished. The metrics in the report were specifically identified and requested by Congress in the NDAA, and provide Congress with important information about the outputs and outcomes of DHS’s border security policies and investments that could be used to inform decision-making. However, those reading the report may not be aware of important contextual information because DHS did not identify and disclose some limitations, thereby creating the potential for the data to be misinterpreted. According to DHS officials who prepared the report, while they took steps to identify methodological limitations of the metrics, no process currently exists to systematically review the reliability of operational data used for public reporting purposes, such as in the metrics report. Specifically, DHS officials within OIS told us that while they were responsible for leading and managing the preparation of the report, they largely relied on the DHS components from which they collected the data to assess the data’s reliability and communicate identified limitations. OIS officials explained that many of the data elements used, such as those from AMO or the Coast Guard, were ones with which they were not familiar or had not worked with previously in their area of immigration statistics. OIS officials also noted that in some cases, the data had previously been used in performance measures or had been collected and tracked for several years, so they trusted the components’ processes for ensuring their reliability and identifying limitations, but reviewed the data provided where possible and consulted with the components as needed. However, OIS officials said that while they included as much information in the report as was known about identified limitations with the existing operational data, no additional effort was made to systematically review the underlying reliability of the data to comprehensively identify limitations that should be acknowledged when publicly reported because no department-wide process exists to do so. Standards for Internal Control in the Federal Government state that management officials should evaluate data sources for reliability and communicate quality information, including relevant data from reliable sources, to achieve an agency’s objectives. The quality information can then be used by agency management and external stakeholders such as policymakers, to make informed decisions and evaluate performance, among other things. Further, DHS’s Management Directive on Information Quality states that data and information disseminated by the department should, among other things, have full, accurate, transparent documentation, and error sources affecting data quality should be identified and disclosed to users. Additionally, our previous work on approaches for verifying and validating performance information found that communicating significant data limitations and their implications allows stakeholders to judge the data’s credibility for their intended use and to use the data in appropriate ways. By developing and implementing a process to systematically review the reliability of the data or consider the results of assessments components have completed, comprehensively identify any limitations, and communicate the data or methodological limitations with the metrics, DHS would improve the quality of the information available to Congress, DHS leadership, and the public. Doing so would also facilitate a better understanding and appropriate interpretation and use of the data in the context of the Border Security Metrics Report, thereby enhancing the report’s value as a source of information for future decision-making. DHS’s Model to Estimate Unlawful Border Entries Uses Assumptions that Have Not Been Validated and Does Not Convey Uncertainty of Estimates DHS Used a Statistical Model to Estimate Unlawful Border Entries for Three Metrics Based upon statistical modelling, DHS developed a Model-based Apprehension Rate to calculate the total number of unlawful border entries between land POEs, including entries both detected by Border Patrol and “estimated undetected unlawful entries.” DHS reported that in fiscal year 2016 there were about 624,000 detected entries (which include apprehensions, turn-backs, and got aways) and estimated that there were about 62,000 undetected unlawful entries. DHS also used the Model- based Apprehension Rate to develop two other metrics in the fiscal year 2017 Border Security Metrics Report: (1) A “probability of detection rate,” which is the estimated proportion of the number of detected unlawful border entries to the total number of unlawful entries between land POEs. DHS estimated that in fiscal year 2016, 91 percent of unlawful border crossers were detected and 9 percent were not detected. (2) The “attempted unlawful border crosser apprehension rate,” which is the estimated proportion of unlawful border entrants apprehended by Border Patrol to the total number of unlawful entrants between land POEs. DHS estimated that in fiscal year 2016, 65 percent of individuals were apprehended by Border Patrol and 35 percent of individuals attempting an unlawful border entry either got away or entered the United States undetected. DHS based its statistical model upon research conducted by the Institute for Defense Analyses that leveraged long-standing research using capture-recapture models. Originally developed and utilized in biological and ecological sciences, capture-recapture models have been applied to other disciplines, including social science. According to the Institute for Defense Analyses, capture-recapture models have been the core approach for academic efforts to model the process of unlawful entry into the United States across land borders for several decades. To develop its statistical model, DHS used a capture-recapture methodology to calculate a probability of apprehension by counting the number of unlawful border crossers that were apprehended multiple times. At a high-level, capture-recapture involves taking an initial sample of the population of interest, in this case individuals attempting to cross the border unlawfully. Then, separately, a second, independent sample of the same population is taken. The samples are then compared to determine the number of individuals who appear in both samples. When the number of individuals who appear in both samples (e.g., individuals who have been apprehended twice) is low, it can be inferred that the overall population of interest (e.g., total unlawful border crossers) is much larger than the total number of individuals in the two samples. On the other hand, if the recapture rate is high, then it can be inferred that the overall population of interest is not much larger than the total number of individuals in the two samples. In the context of unlawful border crossing, when an individual’s first attempt at unlawfully crossing the border is successful, the individual enters the United States and no apprehension is made. However, if an individual is apprehended, Border Patrol records an apprehension of this individual in a DHS data system and the individual is potentially subject to consequences for entering unlawfully, such as administrative enforcement and removal, criminal prosecution, or being barred from legally entering the United States in the future. The individual is then returned to his or her home country, where the individual can then choose whether or not to make another attempt to unlawfully cross the border. During a second attempt to unlawfully cross the border, the individual faces the same possible outcomes (enter the United States unlawfully or apprehension by Border Patrol). Figure 2 provides the framework for DHS’s Model-based Apprehension Rate. DHS modified the traditional capture-recapture methodology by calculating a deterrence rate of 60 percent in fiscal year 2016 to account for individuals who choose not to make another unlawful border crossing attempt. The deterrence rate accounts for an individual being deterred from attempting to unlawfully cross the border again; that is, DHS assumed that some percentage of apprehended individuals, once returned to their country, will remain in their home country. DHS calculated the deterrence rate based upon a survey of Mexican individuals who were apprehended and returned to the border region of Mexico by U.S. immigration authorities. DHS assumed the remaining 40 percent of individuals who were apprehended and removed to their home country in fiscal year 2016 remain undeterred and will attempt to unlawfully cross the border again. Historically, DHS (and its predecessor the Immigration and Naturalization Service) did not use statistical models to calculate an apprehension rate but relied on apprehensions as a proxy measure for all unlawful entries (both observed and unobserved) between POEs. DHS also included in its report information on the apprehension rate using this method. Specifically, DHS also calculated an Observational Apprehension Rate based on direct observations (unlawful border crossers observed by Border Patrol) and indirect observations (residual evidence of a border crosser, i.e., footprints) of attempted unlawful border crossers. Using the observational apprehension rate, DHS calculated that in fiscal year 2016, it apprehended 79 percent of unlawful border crossers. DHS’s Statistical Model Uses Assumptions about Border Crossers that Have Not Been Validated and May Affect Results DHS made assumptions about border crossers to develop its statistical model and described these assumptions in its report; however, DHS did not validate some of these assumptions or determine how they potentially could affect the accuracy of the Model-based Apprehension Rate through the use of sensitivity analyses. More specifically, DHS’s model incorporates several assumptions related to border crossers. Among others, these assumptions include: the rate at which individuals will be deterred from crossing again remains the same, regardless of the number of attempts an individual has made; individuals who indicate an intent to stay near the U.S.-Mexico border will attempt re-entry; a single apprehension rate applies to diverse groups of border crossers, regardless of their nationality or the number of attempts an individual has made; and certain individuals will not evade Border Patrol. However, the validity of some of these assumptions—which affect the Model-based Apprehension Rate—is uncertain. For example, DHS’s model estimates the rate at which a diverse group of border crossers attempting to evade detection will be apprehended by Border Patrol. This group includes both Mexicans and non-Mexicans and individuals who attempt to cross again after varying amounts of time. Despite this diversity, the model assumes that all crossers have the same chance of apprehension on each attempt to cross the border. This assumption allows DHS to apply the estimated apprehension rate developed based on a sample of Mexicans re-apprehended within 90 days—the group for whom relevant data exist—to a broader population of individuals regardless of the number of attempts the border crossers have made or their nationality. However, DHS did not make efforts to determine the extent to which an apprehension rate based on Mexican citizens re- attempting entry within 90 days would reflect apprehension rates for non- Mexicans and individuals crossing again after longer periods. Additionally, DHS assumes that the apprehension rate never varies between an individual’s attempts at crossing the border. For example, DHS assumes that an individual making a first attempt at crossing the border faces the same odds of apprehension as an individual making a fourth or fifth attempt at crossing the border. However, DHS has not explored the possibility that, for example, individuals may gain experience and knowledge from border-crossing attempts that could help them better evade Border Patrol on subsequent attempts. Further, DHS’s model assumes that certain individuals unlawfully crossing the border, such as those seeking asylum, will not evade apprehension and will turn themselves in to Border Patrol. Specifically, in addition to individuals who ultimately do seek asylum, DHS also includes within this group and applies this assumption to individuals apprehended as a family unit and unaccompanied minors. Under this assumption, 100 percent of such individuals are apprehended. According to DHS’s fiscal year 2017 Border Security Metrics Report, these individuals have historically been released into the United States with a Notice to Appear in immigration court for legal proceedings on a future date, rather than being subject to immediate DHS enforcement consequences such as voluntary return. Therefore, DHS assumes that 100 percent of these individuals will self-present to Border Patrol because, in doing so, they are able to claim asylum or other protection and potentially remain in the United States. However, representatives from the Institute for Defense Analysis stated that while anecdotally self-presenting rates of these individuals are high, more rigorous analysis is needed to accurately estimate a self- presentation rate. For example, it is possible that not all families crossing the border unlawfully may seek to self-present to Border Patrol; some may attempt to evade capture and enter the United States undetected. In this case, DHS may be underestimating the number of individuals who unlawfully cross the border and enter the United States by assuming 100 percent of these individuals will self-present to Border Patrol agents. Additionally, DHS noted in its fiscal year 2017 Border Security Metrics Report that this assumption does not reflect the actual behavior of all border crossers in this group. OIS officials stated that they based this assumption on interviews with Border Patrol agents but had not done formal or quantitative analysis to support this assumption. Further, OIS officials stated that they did not have a strong alternative assumption to use instead and therefore assumed that 100 percent of individuals within this group are apprehended. DHS described these assumptions in its report but did not provide quantitative information on the extent to which these assumptions affected the Model-based Apprehension Rate through the use of sensitivity analyses. Sensitivity analyses help to convey the extent to which changing the values of variables, assumptions, data, or other input affects statistical estimates. For example, sensitivity analyses could provide information on how different assumptions about unlawful border crossers’ behavior and other inputs to the statistical model could have affected the Model-based Apprehension Rate. OIS officials stated that while they had started to run sensitivity analyses by modifying certain assumptions, they had not completed the analysis and did not include results of the sensitivity analyses in the report. The Office of Management and Budget’s (OMB) statistical standards for federal agencies include providing the results of sensitivity analyses for key methodological assumptions to ensure that these assumptions do not unduly affect the results of the model. By including the results of sensitivity analyses in its Border Security Metrics Report, DHS would allow Congress and the public to better understand the potential limitations associated with its model and make independent assessments on its accuracy. DHS Did Not Convey the Statistical Uncertainty of Its Estimated Apprehension Rate DHS used a statistical model to develop the Model-based Apprehension Rate but did not provide information on the level of uncertainty related to this estimate. Rather, the fiscal year 2017 DHS Border Security Metrics Report provided a single rate that does not fully convey the difficulty and uncertainty of estimating partially unobserved metrics, such as unlawful entries and the probability of detection. Specifically, using the Model- based Apprehension rate, DHS estimated that 65 percent of unlawful border crossers were apprehended in fiscal year 2016, and the remaining 35 percent entered the United States. However, like all statistical models, DHS’s estimate is based upon a limited sample of data and may be affected by random variation, meaning that DHS does not have complete certainty that its rate is accurate. DHS included a discussion of limitations in the report but did not quantify its degree of uncertainty. According to the OMB statistical standards for federal agencies, possible variation in estimates should be noted, such as by reporting the range of each estimate. Measures of statistical uncertainty, such as margins of error or confidence intervals, help to convey the amount by which estimates might vary due to randomness in the data and allows consumers of the estimates to evaluate their accuracy. OIS officials stated that they agree that providing measures of statistical uncertainty would help Congress and the public better understand the Model-based Apprehension Rate to evaluate border security. Officials told us that the office had begun to develop measures of statistical uncertainty but did not complete this effort because the staff member who was working on the analyses recently left the office. Further, OIS officials stated that they were unsure when they would be able to provide measures of statistical uncertainty in future reports. Including measures of statistical uncertainty in future reports would allow Congress, policy makers, and the public to more fully evaluate the extent to which the metrics that use the Model-based Apprehension Rate are valid. Further, while DHS may ultimately adopt a new, simulation-based model in the future, described later in this report, it plans to use the current Model- based Apprehension Rate for estimates in its Border Security Metrics Report for the foreseeable future. Therefore, providing this additional information about the estimates would allow DHS to more accurately convey how limitations in available data and methods could affect the results and provide more useful information about migration and border enforcement. Additionally, to the extent DHS adopts a new estimating metric, that estimate may have some level of uncertainty associated with it. DHS Is Developing Another Model to Better Reflect Unlawful Border Entries at the Border DHS is developing another model because its current statistical model may not sufficiently reflect conditions at the southwest border. Specifically, DHS’s current statistical model does not fully account for the changing population of unlawful border crossers. The capture-recapture methodology, which underlies the Model-based Apprehension Rate, was developed to sample homogenous populations that behave in set, uniform ways. However, those crossing the border have become increasingly diverse in recent years. Our analysis of DHS data used to develop the Model-based Apprehension Rate shows that the number of unlawful border crossers whose characteristics and behavior are best reflected in the statistical model has declined. For example, our analysis illustrated that the population that conforms best to the model’s assumptions—adult Mexicans travelling without dependents who do not plan to claim asylum and who are returned to Mexico in a short amount of time—has fallen from over 60 percent of apprehensions in fiscal year 2000 to less than 25 percent of apprehensions in 2016, as shown in figure 3. Conversely, the number of individuals who are excluded from the statistical model such as non-Mexicans, and individuals whose behavior may not reflect the model’s assumptions, such as asylum-seekers or those who have not departed the United States (e.g., because they are awaiting immigration court proceedings) have increased over time, as shown in figure 4. For example, the percentage of individuals apprehended at the border who are excluded from the model because they await immigration court proceedings increased from 26 percent in fiscal year 2000 to almost 70 percent in fiscal year 2016. DHS acknowledged these trends in its fiscal year 2017 Border Security Metrics Report and noted them as a limitation to the effectiveness of its model. OIS officials further noted that some of these limitations are difficult to address within the bounds of the statistical model. For example, to properly account for non-Mexicans, OIS officials stated that they would need information on the rate at which non-Mexicans are deterred from crossing the border. However, it would be difficult and costly to obtain this information through the use of a survey and real-world data does not already exist, according to OIS officials. To help address limitations of its current statistical model, DHS has invested in another research project to estimate the number of unlawful border crossers between land POEs, including unknown border entries. Border Patrol contracted with Johns Hopkins Applied Physics Laboratory to undertake a project that aims to use a combination of statistical modeling and data from sensors along the border to estimate the total number of unlawful border entries between land POEs, including entries both detected by Border Patrol and those not detected by Border Patrol. According to project documentation we reviewed, the project plans to leverage the CBP Tactical Simulation, an agent-based simulation of tactical border operations. CBP Tactical Simulation incorporates information on terrain at the border based on geographic information systems and sensors along with probability models that reflect how Border Patrol agents and unlawful border crossers behave in given circumstances. Border Patrol and OIS officials told us that this project would be more adaptable to changing border conditions and could help the agency address limitations associated with the Model-based Apprehension rate. Specifically, according to OIS officials, a simulation-based estimate would rely upon fewer assumptions about the types of individuals who unlawfully cross the border as compared to the current Model-based Apprehension rate. However, Border Patrol officials noted that estimates of unobservable phenomena, such as unobserved border entries, always face some limitations in their accuracy and that the new model may still rely upon samples of data that would have associated uncertainty as well as assumptions that would need to be validated. Ultimately, though, Border Patrol officials stated that the simulation-based model may be an improvement upon the current Model-based Apprehension rate. Border Patrol officials stated that the first iteration of the model would be presented to Border Patrol leadership for their review at the end of fiscal year 2019 and if at that time Border Patrol leadership approves the model, the earliest the simulation-based estimate could potentially be incorporated into the DHS Border Security Metrics Report would be for fiscal year 2020. Exploring alternative models is a positive step for DHS, however given that the project is in the early stages, it is too early to tell if it will be able to address the limitations we identified associated with the current model. Our Prior Work Has Identified Other Metrics DHS Could Use to Help Measure the Effectiveness of Border Security In addition to the NDAA metrics, we have identified other metrics that DHS could use to help measure the effectiveness of border security. In particular, based on the findings from our previous reviews of border security programs and efforts, we have recommended that DHS use metrics that are relevant to each of the four domains listed in the NDAA— between POEs, at POEs, the maritime border, and for air and marine security in the land domain. For example, Between POEs domain. In February 2017, we reported on the use of border fencing along the southwest border and found that CBP collects data that could help provide insight into how border fencing contributes to border security operations, including the location of illegal entries. For example, we found that CBP collects data it could potentially use to determine the extent to which border fencing diverts illegal entrants into more rural and remote environments, and border fencing’s impact, if any, on apprehension rates over time. However, CBP had not developed metrics that systematically use these data to assess the contributions of border fencing to its mission. To better position CBP to make resource allocation decisions with the best information available to inform competing mission priorities and investments, we recommended that the Chief of the Border Patrol develop metrics to assess the contributions of pedestrian and vehicle fencing to border security along the southwest border using CBP data. DHS agreed with the recommendation and stated that it planned to develop metrics for use in its operational control framework for southwest border security operations. As of October 2018, DHS stated that the department planned to test the metrics and implement them in the framework by September 2019. At POEs domain. In July 2017, we reported on the Importer Security Filing (ISF) program and found that while ISF rule data have improved the program’s ability to identify high-risk cargo shipments, CBP could collect additional performance information to better evaluate program effectiveness. While evaluating the direct impact of using ISF rule data to assess shipment risk is difficult, we identified examples of how CBP could better assess the ISF program’s effectiveness. For example, CBP could track the number of containers not listed on a manifest—which could pose a security risk—it identifies through reviewing vessel stow plans. Collecting this type of additional information would help CBP better assess whether the ISF program is improving its ability to identify high-risk shipments. Therefore, we recommended that CBP identify and collect additional performance information on the impact of the ISF rule data, such as the identification of shipments containing contraband, to better evaluate the effectiveness of the ISF program. DHS agreed with the recommendation and reported that it is working to assess additional performance metrics to evaluate the effectiveness of the ISF program and anticipates completing the assessment by end of December 2019. Maritime border domain. In October 2017, we reported on the Coast Guard’s performance goals and found that although the Coast Guard’s performance goals are generally aligned with its statutory missions, the Coast Guard does not explain why certain aspects of mission performance are measured while others are not. For example, we found that while the Coast Guard’s mission is to interdict all illegal drugs, the agency’s two performance goals related to that mission were for cocaine interdiction only, excluding many other substances. We recommended that the Coast Guard either develop new performance goals to address mission activity gaps, or explain in the Coast Guard’s Annual Performance Report why certain aspects of mission performance are measured while others are not. Developing new goals to address missions, or describing how existing goals sufficiently assess mission performance, could better convey the Coast Guard’s progress in achieving its missions. DHS agreed with the recommendation and in February 2018, the Coast Guard provided us with its updated fiscal year 2017 Annual Performance Report. We found that while the updated report explained why performance goals related to its drug interdiction mission focus solely on cocaine interdiction, for the four other performance goals we previously identified as not fully addressing all related mission activities, the updated report did not include additional goals or explain why certain aspects of mission performance are not measured. We continue to believe that in instances in which performance goals do not fully address all of the respective mission activities, the Coast Guard’s Annual Performance Report should include an explanation. Air and marine security in the land domain. In May 2017, we reviewed DHS’s efforts to address subterranean, aerial, and maritime smuggling of drugs and humans. We found that while DHS established high-level performance measures and collected data on smuggling by ultralight aircraft, it had not assessed its efforts specific to addressing this smuggling method. Additionally, we found that DHS had similarly not assessed smuggling methods such as tunnels, panga boats (a fishing vessel), and recreational vessels. We recommended that DHS direct CBP, ICE, and Coast Guard to establish and monitor performance measures and targets related to ultralight aircraft, cross-border tunnels, panga boats, and recreational vessel smuggling to help provide reasonable assurance that efforts to address these smuggling methods are effective. By establishing measures and monitoring performance against targets, managers could obtain valuable information on successful approaches and areas that could be improved to help ensure that technology investments and operational responses to address these smuggling methods are effective. DHS agreed with the recommendations for measures related to ultralight aircraft and cross-border tunnels. DHS reported that AMO and Border Patrol have drafted a performance measure for ultralight aircraft, however, reviews and approval of the measure will not be completed until November 2019. As of June 2018, DHS reported that ICE was leading the development of measures related to cross-border tunnels. DHS did not agree with the recommendation to establish measures and monitor performance against targets for smuggling by panga boats and recreational vessels because the department believed measures and targets would not provide the most useful strategic assessment of operations to prevent all illicit trafficking, regardless of area of operations or mode of transportation. We continue to believe that the recommendation is valid and recognize the value of high-level strategic performance measures. However, such high-level measures may not provide sufficiently detailed performance information to allow DHS to identify successful approaches to addressing smuggling by panga boats and recreational vessels and areas for improvement. Further, establishing performance measures and targets related to smuggling by panga boats and recreational vessels could, in turn, better position DHS to understand the overall smuggling threat. Appendix II provides additional information on these and other metrics we have previously recommended that DHS could use to help measure the effectiveness of border security in the four domains. Conclusions Securing U.S. borders is a complex undertaking that spans multiple domains and locations. It is also a key part of DHS’s mission for which DHS has made significant investments over the years. Given the complexity and breadth of border security efforts, having data and information available on the state of border security is important for DHS as well as policymakers and the public to understand the effectiveness of those investments. DHS’s fiscal year 2017 Border Security Metrics Report makes an important contribution in providing such data and information. DHS components generally have processes to help ensure the reliability of the data used in the metrics report and DHS identified and disclosed some data and methodological limitations with the metrics. However, DHS did not systematically review the reliability of data used in all metrics to identify and disclose limitations and their potential implications for the metric. Without complete information about the limitations of the data or the metric methodologies used in the report, Congress, policymakers, and the public may not be aware of important context or information needed to fully and appropriately understand the data being presented. By developing and implementing a process to systematically review the reliability of the data, as well as comprehensively identify limitations and communicate limitations of the metrics, DHS would improve the quality of the data and information provided in the report which would facilitate a better understanding and appropriate interpretation of the data and information provided. To develop three metrics in the report, DHS used a statistical model that incorporated untested assumptions about the behavior of unlawful border crossers that may not reflect real-world conditions. DHS was transparent about the limitations of its model, but providing the results of sensitivity analyses and measures of statistical uncertainty related to the model would allow Congress, policymakers, and the public to better understand its potential limitations and more fully evaluate the validity of DHS’s metrics that use estimates. Recommendations for Executive Action We are making the following four recommendations to DHS: The Secretary of Homeland Security should develop and implement a process to systematically review the reliability of the data used in its Border Security Metrics Report and comprehensively identify any limitations with the data and methodologies that underlie its metrics. (Recommendation 1) The Secretary of Homeland Security should ensure the communication of the limitations of the metrics identified through the systematic review in the department’s annual Border Security Metrics Report. (Recommendation 2) The Under Secretary for the Office of Strategy, Policy, and Plans should include the results of sensitivity analyses to key assumptions in its statistical models of unlawful entry estimates in its annual Border Security Metrics Report. (Recommendation 3) The Under Secretary for the Office of Strategy, Policy, and Plans should include measures of statistical uncertainty for all metrics based on estimates derived from statistical models in its annual Border Security Metrics Report. (Recommendation 4) Agency Comments We provided a draft of this report to DHS and the Office of National Drug Control Policy for review and comment. DHS provided written comments, which are reproduced in appendix III and discussed below. DHS also provided technical comments, which we incorporated as appropriate. The Office of National Drug Control Policy indicated via e-mail that it did not have any comments on the draft report. In its comments, DHS concurred with our recommendations and stated that it planned to implement 3 of the 4 by October 2020. With respect to our second recommendation, DHS requested that we consider it closed as implemented because the department already detailed some of the limitations in its fiscal year 2017 report, and plans to continue to identify known limitations and the progress made to mitigate previously identified limitations in future reports. As discussed in this report, we agree that DHS identified and disclosed limitations for some metrics in its fiscal year 2017 Border Security Metrics Report; however, we identified at least one additional limitation for 21 of the 35 metrics on which DHS reported that DHS did not disclose or about which it could have been more transparent. To address the intent of this recommendation, once DHS has implemented a process to systematically review the reliability of the data used in its report and comprehensively identified related limitations, it should disclose those limitations in its annual Border Security Metrics Report. We are sending copies of this report to the appropriate congressional committees and the Secretary of the Department of Homeland Security. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. Contacts points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. If you or your staff have any questions about this report, please contact me at (202) 512-8777 or gamblerr@gao.gov. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Additional Information on Metrics Contained in the Department of Homeland Security’s Fiscal Year 2017 Border Security Metrics Report This appendix provides additional information on our analysis of the suitability and validity of the metrics the Department of Homeland Security (DHS) reported in its fiscal year 2017 Border Security Metrics Report for each of the four domains listed in the National Defense Authorization Act for Fiscal Year 2017 (NDAA)—between ports of entry, at ports of entry, the maritime border, and air and marine metrics in the land domain. Specifically, this appendix provides information on the metrics including their status, descriptions, differences between what DHS reported for the metrics and how they were described or defined by the NDAA, limitations, and any additional information or planned actions by DHS, where applicable. Between Ports of Description This metric is a rate comparing apprehensions to the total number of attempted unlawful border crossers. As such, this metric requires an estimate of the number of unlawful entry attempts that are not detected, which is added to the number of detected unlawful border crossers to create the denominator. The Department of Homeland Security (DHS) provided two methods for calculating this rate in its report. The first method, called the Model-based Apprehension Rate, uses a statistical model based on the capture-recapture methodology to estimate the rate.The second method, called the Observational Apprehension Rate, calculates the ratio of apprehensions to the sum of apprehensions and got aways. unlawful entries is limited to the southwest border. According to the report, research is underway on methods to produce estimates for the northern border. Limitations The Observational Apprehension Rate incorporates data on apprehensions, and got aways, while the Model- based Apprehension Rate is based on an estimate for undetected unlawful entries. Consequently, the limitations for those metrics also apply here. For more information on the limitations for those metrics, see the respective sections below. DHS identified: The observational apprehension rate excludes unobserved got aways. Additional information and planned actions by the Department of Homeland Security In its report, DHS noted that it has taken steps to improve situational awareness along the border and mitigate limitations. These steps include investing in technology, refining observational estimates, and developing a methodology to estimate statistical reliability. According to U.S. Border Patrol officials, investments in new technology have enabled U.S. Border Patrol to better detect cross-border activities. For additional information on the data elements used for this metric and DHS’s planned actions, see the respective sections below on apprehensions, got aways, and the estimate for undetected unlawful entries. Description This metric is a count of the total number of attempted unlawful border crossers between land ports of entry who were directly or indirectly observed or detected by U.S. Border Patrol. The Department of Homeland Security (DHS) calculated this metric by adding turn backs, got aways, and apprehensions of unlawful border crossers. Patrol officials, the northern border has different immigration dynamics than the southern border and accounts for a significantly smaller number of turn backs and got aways overall, so northern border data were not included. Limitations Because this metric incorporates data on apprehensions, got aways, and turn backs, the limitations for those metrics also apply here. For more information on the limitations for those metrics, see the respective sections below. Additional information and planned actions by the Department of Homeland Security For additional information on the data elements used for this metric and DHS’s planned actions, see the respective sections below. Description This metric is an estimate of the number of attempted unlawful border crossers that are not directly or indirectly observed or detected by U.S. Border Patrol (Border Patrol). The Department of Homeland Security (DHS) used a statistical model, based on capture-recapture methodology, to estimate total successful unlawful entries, and subtracted detected got aways to calculate the total number of undetected unlawful entries. unlawful entries is limited to the southwest border. According to DHS’s report, research is under way to produce this estimate for the northern border. DHS does not currently have reliable data on the estimated share of migrants who, following an unsuccessful unlawful entry attempt, are deterred from making a subsequent reentry attempt. For its model, DHS used data from a survey of recently removed Mexicans, which asked them about their intentions to re-enter the United States. According to DHS’s report, a shortcoming of the survey is that it does not take account of shifting border enforcement efforts, potential changes in behavior by individuals who have been exposed to consequence programs, or other deterrent factors along the border. Consequently, any resulting undercount in the estimate of the deterred population results in a downward bias. The population that conforms best to the model’s assumptions represents a diminishing share of southwest border apprehensions. Specifically, in its report DHS said that Mexican adults removed to the nearest border accounted for about 95 percent of apprehensions in the 1990s. However, because of recent changes at the border, including changes in the composition of border flows (i.e., rising numbers of Central Americans and asylum seekers) and in Border Patrol’s enforcement strategy, the population best reflected in the model has declined to as few as 20 percent of apprehensions in recent years. Further, DHS noted that some alien populations, such as those seeking asylum and who do not evade detection by Border Patrol agents, are also excluded from the model. However, these populations make up an increasing share of apprehensions in recent years. The model uses restrictive assumptions about which re-apprehensions to include. For example, the model excludes apprehensions occurring at check points and other remote locations and those occurring more than 4 days after an illegal entry. According to DHS, these assumptions result in a downward bias. We identified: DHS described assumptions it made in its report but did not provide quantitative information on the extent to which they affected its estimated undetected unlawful entries through the use of sensitivity analyses. Sensitivity analyses help to convey the extent to which changing the values of variables, assumptions, data, or other input affects statistical estimates. By including the results of sensitivity analyses in its Border Security Metrics Report, DHS would allow Congress and the public to better understand the potential limitations associated with its model and make independent assessments on its accuracy. DHS did not provide information on the statistical level of uncertainty related to this rate, such as margins of error or confidence intervals. This information would help convey how the estimates might vary due to randomness in the data. Instead, DHS provided a single rate that does not fully convey the difficulty and uncertainty of the estimate. This metric incorporated data on apprehensions and got aways. For more information on the limitations associated with those metrics, see the respective sections below. Additional information and planned actions by the Department of Homeland Security According to DHS, officials are continuing to improve the accuracy of the existing statistical model for estimating unlawful border crossers but are also considering alternative methodologies. U.S. Customs and Border Protection has contracted with Johns Hopkins Applied Physics Laboratory to develop a new model for estimating the flow of unlawful border crossers. This model uses a combination of statistical modelling, data from sensors along the border, and probability models that reflect how Border Patrol agents and unlawful border crossers behave in given circumstances. Border Patrol officials estimated that the earliest the simulation-based estimate could potentially be incorporated into the DHS Border Security Metrics Report would be for fiscal year 2020. Description This metric is a count of the number of unlawful border crossers who, after making an unlawful entry into the United States, responded to law enforcement efforts by returning promptly to the country from which they entered. These data came from U.S. Border Patrol (Border Patrol) records. Border Patrol officials, the northern border has different immigration dynamics than the southern border and accounts for a significantly smaller number of turn backs overall, so northern border data were not included. Officials stated that while the current emphasis of reporting is on the southwest border, efforts are underway to identify and find ways to capture data that are important and reflective of the effectiveness in addressing threats specific to the northern border. The estimate aggregates potentially subjective observations from thousands of individual agents. Some unlawful border crossers may enter the United States to drop off drug loads or to act as decoys to lure agents away from a certain area and then return to Mexico, and therefore may be misidentified as turn backs. In our previous work we identified differences in the procedures for reporting and classifying turn backs across sectors, and noted how factors such as terrain and weather may impact agents’ abilities to accurately detect turn backs. According to DHS, since 2014, Border Patrol has implemented a standard, southwest border-wide methodology to improve reporting and mitigate the potential subjectivity of observations by agents. Therefore, data before 2014 are not necessarily comparable to data from 2014 and later. DHS presented the data in a table without explaining that the methodology used to categorize and count turns backs changed in 2014. Additional information and planned actions by the Department of Homeland Security According to DHS’s report, Border Patrol has taken steps to implement a standard, southwest border-wide methodology to improve reporting of potential turn backs. In addition, DHS’s report said that command staff ensure all agents are aware of and utilize proper definitions for apprehensions, got aways, and turn backs at their respective stations. They also ensure necessary communication takes place between and among sectors and stations to minimize double-counting when subjects cross through more than one station. DHS’s report noted that Border Patrol headquarters components validate data integrity. Description This metric is a count of the number of unlawful border crossers who are directly or indirectly observed entering unlawfully, are not apprehended, and are not turn backs. These data came from U.S. Border Patrol (Border Patrol) records. Border Patrol officials, the northern border has different immigration dynamics than the southern border, so northern border data were not included. Officials stated that while the current emphasis of reporting is on the southwest border, efforts are under way to identify and find ways to capture data that are important and reflective of the effectiveness in addressing threats specific to the northern border. Limitations DHS identified: The count aggregates potentially subjective observations from thousands of individual agents. In previous work we identified differences in procedures for reporting and classifying got aways across sectors, and noted how factors such as terrain and weather may impact agents’ abilities to accurately detect got aways. According to DHS, since 2014, Border Patrol has implemented a standard, southwest border-wide methodology to improve reporting and mitigate the potential subjectivity of observations by agents. Therefore, data before 2014 are not necessarily comparable to data from 2014 and later. DHS presented the data in a table without explaining that the methodology used to categorize and count turns backs changed in 2014. For information on limitations with the model-based estimate for undetected unlawful entries, see the section for estimated undetected unlawful entries above. Additional information and planned actions by the Department of Homeland Security According to DHS’s report, Border Patrol has taken steps to implement a standard, southwest border-wide methodology to improve reporting of potential got aways. In addition, DHS’s report said that command staff ensure all agents are aware of and utilize proper definitions for apprehensions, got aways, and turn backs at their respective stations. They also ensure necessary communication takes place between and among sectors and stations to minimize double-counting when subjects cross through more than one station. DHS’s report noted that Border Patrol headquarters components validate data integrity. As a comparison against the counts of documented got aways, DHS also provided an estimate of total successful unlawful entries along the southwest border using a statistical model based on capture-recapture methodology. For more information on the methodology for this estimate, see the section titled “Estimated Undetected Unlawful Entries” in this appendix. Description This metric is a rate comparing the number of apprehensions and turn backs to the number of apprehensions, estimated undetected unlawful entries, turn backs, and got aways in each U.S. Border Patrol sector. rate is not available because sector-level estimates of unlawful entries and attempts have not yet been produced and validated. As an alternative, DHS presented data using the interdiction effectiveness rate. With this rate, the estimated undetected unlawful entries measure is replaced with known got aways. However, DHS does not have an interdiction effectiveness rate for the northern border so it solely provided data for the southwest border. According to DHS’s report, the department has not yet developed a northern border interdiction effectiveness rate because there are only a small number of attempted and successful entries along the northern border. Limitations None identified. Additional information and planned actions by the Department of Homeland Security DHS reported that sector-level estimates of unlawful entries and attempts are projected to be available in its 2019 annual Border Security Metrics Report to Congress. Authorization Act for Fiscal Year 2017 (NDAA) defined this metric as a rate comparing the estimated total undetected unlawful border crossing attempts to the unlawful border crossing effectiveness rate. The Department of Homeland Security (DHS) calculated this metric by dividing the detected unlawful entries by the estimated total unlawful entries. The number of detected unlawful entries is calculated by adding turn backs, got aways, and apprehensions. Estimated total unlawful entries is calculated by adding turn backs, apprehensions and estimated total successful unlawful entries derived from DHS’s statistical model. unlawful entries is limited to the southwest border. Additionally, DHS used detected unlawful entries as the numerator, instead of the estimated total unlawful border crossing attempts not detected as called for in the NDAA. For the denominator DHS used the estimated total unlawful entries instead of the unlawful border crossing effectiveness rate, as called for in the NDAA. Limitations Because this metric incorporates data on apprehensions, got aways, and turn backs, as well as the estimate for undetected unlawful entries, the limitations for those metrics also apply to this metric. For more information on the limitations for those metrics, see the respective sections for those metrics. Additional information and planned actions by the Department of Homeland Security For additional information on apprehensions, got aways, turn backs, and the estimate for undetected unlawful entries, and any planned actions by DHS for those metrics, see the respective sections for those metrics. Description This metric is a count of the number of apprehensions in each U.S. Border Patrol (Border Patrol) sector. Data come from Border Patrol records, and each apprehension of the same unlawful crosser in a fiscal year is counted separately, meaning these data do not represent a count of unique crossers apprehended. Border Patrol officials, the northern border has different immigration dynamics than the southern border, so northern border data were not included. Officials stated that while the current emphasis of reporting is on the southwest border, efforts are under way to identify and find ways to capture data that are important and reflective of the effectiveness in addressing threats specific to the northern border. Limitations DHS identified: In its report, DHS said that apprehensions are not a useful indicator of successful unlawful border crossings over the long-term and across multiple locations because the relationship between apprehensions and successful unlawful entries depends on the apprehension rate, which changes over time and may differ by location. Additional information None. Description This metric is a count of the number of apprehensions of unaccompanied alien children (UAC), and the nationality of such children, in each U.S. Border Patrol (Border Patrol) sector. A UAC is a child under 18 years old with no lawful immigration status, and no parent present and available in the United States to provide care and physical custody. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report The Department of Homeland Security (DHS) only included data for the southwest border. Limitations DHS identified: Agents may not be able to reliably distinguish among older children and young adults or confirm whether children are traveling alone or in family groups. We identified: We previously reported that it can be challenging to obtain accurate information about a child’s country of origin because of absence of documentation, language barriers, and coached responses by smugglers, among other reasons. Border Patrol officials said that the data on UAC may have reliability issues because original data from a shared database had been changed. Specifically, officials said that in January 2015 they noticed that Enforcement and Removal Operations staff were inadvertently overwriting Border Patrol’s original data entries about the status of migrant children apprehended once those children were placed with relatives or a foster family. Additional information and planned actions by the Department of Homeland Security According to Border Patrol officials, agents rely on statements provided by the child to determine the nationality of UACs when verifiable documentation is not available. Verifiable documentation could include biometric checks, birth certificates, state-issued identification cards, and passports. However, officials noted that this list is not all-inclusive and the processing agent determines the validity of any presented documents. Border Patrol officials said that a data integrity team regularly examines data on apprehensions and they conduct biweekly data reliability checks. Additionally, they are working with Enforcement and Removal Operations to modify the data entry process so that updates can be made without overwriting the original apprehension data entered by Border Patrol. Description This metric is a count of the number of apprehensions of family units, and the nationality of such family units, in each U.S. Border Patrol (Border Patrol) sector. A family unit is the number of individuals apprehended with a family member. For example, a mother and child apprehended together are counted as two family units. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report The Department of Homeland Security (DHS) only included data for the southwest border. Limitations DHS identified: DHS noted that the count of apprehensions for family units is considered reliable, but that agents may not be able to reliably identify family units. We identified: According to Border Patrol officials, their data entry system did not have a dedicated field for agents to record apprehensions of persons within a family unit for all of the years presented in the report. In December 2014, Border Patrol added specific data entry fields to its data entry processes for agents to input information about family units. These fields incorporated built in safeguards and edit checks to help ensure that agents make an appropriate family unit classification. Previously, Border Patrol officials said they used proxy data to identify family units. Given the additional safeguards and checks included with the new family unit data entry fields, Border Patrol officials stated that the data after December 2014 may be more reliable overall compared to previous years. Border Patrol officials stated that they have high confidence in the proxy count for data pre-2014, but acknowledged that those data may contain misclassifications of family units. Additional information According to Border Patrol officials, agents are trained in interviewing techniques and the processing agent will consider all available evidence to determine the validity of claims to familial relationships. Border Patrol officials also noted that in order to be categorized as a family unit, at least one member of the family unit must be at least 18 years of age. Consequently, related individuals younger than 18 years of age that are apprehended together would not be categorized as a family unit. Description This metric is a rate comparing the amount and type of illicit drugs seized between ports of entry in any fiscal year to the average of the amount and type of illicit drugs seized between ports of entry in the immediately preceding 5 fiscal years. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations None identified. Additional information None. Description This metric was not specifically defined in the National Defense Authorization Act for Fiscal Year 2017 (NDAA); the NDAA called for an estimate of the impact of the Consequence Delivery System (CDS) on the recidivism rate of unlawful border crossers over multiple fiscal years. The Office of Immigration Statistics (OIS) calculated this metric by providing the average annual recidivism rate for the 3 years prior to fiscal year 2012— when the CDS was implemented—and the average annual recidivism rate for the subsequent 3 years. The annual recidivist rate is calculated by dividing the number of unique crossers apprehended multiple times in a fiscal year by the total number of unique crossers in the fiscal year. to DHS’s report, recidivism data for the northern border were not available due to the small number of attempted illegal entries along the northern border. Noting the findings from our January 2017 review, DHS stated that its current recidivism measure could be strengthened by using the date an unlawful border crosser is removed or returned instead of the date they are apprehended, as well as by counting re-apprehensions within a fixed period of time defined by the crosser’s repatriation date instead of by the fiscal year. In January 2017, we reported that using a crosser’s apprehension history beyond 1 fiscal year, and excluding crossers that have not been previously removed, among other things, produces a significantly different rate compared to how DHS currently calculates it. Consequently, we recommended that DHS calculate recidivism for a period of time longer than 1 fiscal year and that DHS exclude from the recidivism calculation aliens for whom there is no record of removal and who may remain in the United States. As of December 2018, this recommendation remained open. DHS stated that changes in the recidivism rate after 2012 cannot be attributed solely to CDS because enforcement is a complex, dynamic system. We identified: Given that DHS’s methodology is to provide the 3-year average of the recidivism rate before and after CDS was implemented in fiscal year 2012, the data presented will remain static for subsequent annual reports because the periods of comparison for analyzing recidivism are fixed around a specific point in time. According to OIS officials, to help address this issue, in the next report they plan to provide individual rates for each year instead of the 3-year average. Additional information and planned actions by the Department of Homeland Security In its report, DHS noted that future reports will include estimates of the impact of CDS on both the annual recidivism rate and a longer-term recidivism rate. For example, OIS officials said they plan to update the way they calculate recidivism for future issues of the report and are developing a multivariate impact analysis that would take into consideration factors such as crossers’ demographics and immigration history. Description This metric was not specifically defined in the National Defense Authorization Act for Fiscal Year 2017 (NDAA); the NDAA called for an examination of each consequence under the Consequence Delivery System (CDS), including (1) voluntary return, (2) warrant of arrest or notice to appear, (3) expedited removal, (4) reinstatement of removal, (5) alien transfer exit program, (6) criminal consequence program, (7) standard prosecution, and (8) Operation Against Smugglers Initiative on Safety and Security. The Department of Homeland Security (DHS) presented data on the recidivism rates for each consequence between fiscal years 2012 through 2016. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report As noted above for the metric “estimates of the impact of the Consequences Delivery System on recidivism,” DHS only included data for the southwest border because recidivism data for the northern border were not available due to the small number of attempted illegal entries along the northern border. Differences in recidivism rates among the consequences may also reflect differences in the propensity of the targeted populations to attempt to re-enter. As with the metric for estimating the impact of the CDS on recidivism discussed above, DHS noted the limitation that current recidivism data are based on apprehensions within a given fiscal year, and not the date when an individual was repatriated to their country of origin. In January 2017, we reported that some unlawful border crossers were incorrectly classified based on CDS guidance. U.S. Border Patrol (Border Patrol) agents implement CDS by classifying apprehended aliens into one of seven noncriminal or criminal categories and then applying one or more of eight different consequences; therefore, determining the correct classification of the unlawful border crosser is important for identifying and applying the appropriate consequence. Our analysis of Border Patrol apprehension data from fiscal year 2013 through 2015 showed that Border Patrol did not classify 11 percent of apprehensions in accordance with the agency’s guidance. We recommended that Border Patrol provide consistent guidance for classification and take steps to ensure the integrity of classification data. Border Patrol implemented this recommendation as of December 2017, but the issue could potentially have implications for the data DHS used in this metric, which was for fiscal years 2012 through 2016. Additional information and planned actions by the Department of Homeland Security According to its report, DHS is refining its analysis and will seek to specifically address the limitations discussed above in the fiscal year 2018 version of the Border Security Metrics Report. Ports of entry are U.S. government facilities that provide for the controlled entry into or departure from the United States. There are 328 ports of entry in the United States. Specifically, a port of entry is any officially designated location (seaport, airport, or land border location) where U.S. Customs and Border Protection (CBP) officers or employees are assigned to clear passengers, merchandise and other items, collect duties, and enforce customs laws; and where CBP officers inspect persons seeking to enter or depart, or apply for admission into, the United States pursuant to U.S. immigration law and travel controls. CBP’s Office of Field Operations (OFO) is the lead DHS component responsible for carrying out activities at POEs. The 15 metrics in this domain measure the number of travelers attempting to enter the United States at ports of entry, illicit drugs seized at ports of entry, and cargo entering the United States, among other things. DHS included 11 of the 15 metrics called for in the NDAA for this domain in its fiscal year 2017 Border Security Metrics Report, as shown in table 5. DHS reported that the four metrics for which it did not provide information did not yet have a reliable methodology or were under review, and that DHS was in the process of developing methodologies to capture the data needed for the requested metrics. DHS officials said these four metrics would not be ready for inclusion in the next annual report. Description This metric is a count of total inadmissible travelers, and requires an estimate of the number of inadmissible travelers who successfully enter at a port of entry without being detected. The metric is the sum of the number of inadmissible travelers interdicted and the estimated number of inadmissible travelers who successfully enter at a port of entry without being detected. inadmissible travelers who successfully enter at a port of entry without being detected. Therefore, DHS only presented data on known inadmissible travelers. Limitations None identified. Additional information and planned actions by the Department of Homeland Security DHS projected that the department may be able to include estimates on the number of inadmissible travelers who successfully enter at a port of entry in its fiscal year 2019 Border Security Metrics Report to Congress. According to U.S. Customs and Border Protection (CBP) officials, they are in the process of determining whether CBP’s Compliance Measurement Examination (COMPEX) program could be used as a means to reliably measure undetected inadmissible travelers. Description These metrics are rates that require data on travelers seeking admission at a port of entry, interdictions of inadmissible travelers, and an estimate of the number of inadmissible travelers who successfully enter at a port of entry without being detected. The refusal rate is calculated by dividing the number of inadmissible travelers interdicted by all people seeking admission at a port of entry. The interdiction rate is calculated by dividing the number of inadmissible travelers interdicted by the total number of inadmissible travelers who attempt to enter at a port of entry. inadmissible travelers who successfully enter at a port of entry without being detected. Therefore, DHS only presented data on the refusal rate. Limitations None identified. Additional information and planned actions by the Department of Homeland Security DHS projected that the department may be able to include estimates on the number of inadmissible travelers who successfully enter at a port of entry in its next Border Security Metrics Report to Congress. According to U.S. Customs and Border Protection (CBP) officials, they are in the process of reviewing data and program policies for CBP’s Compliance Measurement Examination program to determine if the program could be used as a means to reliably measure undetected inadmissible travelers, which would then be used in calculating the interdiction rate. Description This metric is a count of the amount in kilograms of illicit drugs seized by U.S. Customs and Border Protection officers at ports of entry. In an appendix to the report, the Department of Homeland Security listed out 34 different types of illicit drugs and the amounts seized for each for fiscal years 2007 through 2016. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations None identified. Additional information None. Description This metric is a rate that compares the amount of illicit drugs seized (in kilograms) by Office of Field Operations officials at ports of entry in 1 fiscal year to the average amount seized in the immediately preceding 5 fiscal years. the metric. The Department of Homeland Security provided rates for cocaine, methamphetamine, marijuana, and heroin for fiscal years 2012 through 2016. Limitations None identified. Additional information None. Description This metric is a count of the number of infractions related to travelers and cargo committed by major violators, and an estimate of the number of major infractions not interdicted. The Department of Homeland Security (DHS) calculated an infraction rate by dividing the number of major infractions by the total number of passengers at ports of entry for fiscal years 2007 through 2016. National Defense Authorization Act for Fiscal Year 2017 (NDAA). As an alternative, for the purpose of its report, DHS defined a major infraction as an arrest, including arrests related to terrorism, drugs, criminal aliens, and currency, among other things. DHS reported that it does not have a methodology in place to estimate the number of undetected major infractions. Therefore, only data on known infractions are included. DHS only included data for passenger infractions and not cargo-related infractions. Although not requested by the NDAA, DHS provided an infraction rate by dividing the number of known infractions by the total number of travelers at ports of entry. Limitations We identified: Given that DHS’s alternative approach to this metric involves using arrests as a proxy for major infractions, it is unclear whether there is a one-to-one correspondence between the arrest of a major violator and the number of infractions committed. Additional information According to U.S. Customs and Border Protection (CBP) officials, they plan to use data from CBP’s Compliance Measurement Examination program as a means to report estimated undetected major infractions starting with DHS’s fiscal year 2019 report. Description This metric is a rate that compares the amount of cocaine seized at land ports of entry to the total estimated flow of cocaine. the total flow of cocaine through land ports of entry. The Office of National Drug Control Policy produces annual estimates for total cocaine flow into the United States, but does not have a methodology to estimate the flow of cocaine through land ports of entry alone. Therefore, the estimates the Department of Homeland Security used included cocaine flow through all domains. According to the U.S. Drug Enforcement Administration’s National Drug Threat Assessment, the southwest border remains the key entry point for the majority of the cocaine entering the United States. Limitations None identified. Additional information None. Description This metric is a rate that compares the average wait time for vehicles to pass through a land port of entry to the total number of commercial and private vehicles at each land port of entry. data were not available for every port of entry, such as small ones with negligible wait times. Limitations We identified: We reported in July 2013 that commercial vehicle wait time data were unreliable due to inconsistent data collection processes at ports, and made two recommendations to DHS to improve the reliability of the data. While DHS implemented these recommendations in 2018, older data, including the data for the years presented in the report (fiscal years 2012 through 2016), remain unreliable. Additional information and planned actions by the Department of Homeland Security U.S. Customs and Border Protection (CBP) officials clarified that the wait times shown in the report reflect the average of all hourly recordings for wait times at ports of entry rather than the average passenger or vehicle experience because CBP did not report a volume-weighted measure of wait times. According to the report, CBP’s wait time policy is currently under review and new guidance will be issued in the future to account for improvements in automation and recording. Description This metric is a rate that measures traffic volume at land ports of entry against the physical and staffing capacity at each land port of entry. The Department of Homeland Security (DHS) calculated the average number of vehicles processed per booth, per hour at each land port of entry. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None. Limitations None identified. Additional information In addition to reporting utilization at each port of entry, DHS provided the average utilization rate for all northern border land ports of entry and all southern border land ports of entry. This metric is a rate that measures the frequency of secondary examinations at each land port of entry. The Department of Homeland Security (DHS) calculated the rate by dividing the recorded number of passengers sent for secondary inspection by the total number of recorded passengers at each land port of entry. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report DHS did not include data on secondary examinations of cargo or shipments. Limitations None identified. Additional information None. Description This metric is a count of the number of cargo containers at sea ports that DHS identified as potentially high-risk using National Targeting Center (NTC) security criteria. According to the Department of Homeland Security (DHS), all international cargo shipments coming to the United States are screened to identify potentially high-risk containers, which may then be reviewed, scanned, or physically inspected prior to lading at a port of entry. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations We identified: U.S. Customs and Border Protection (CBP) officials said that the process of defining and identifying “high-risk” shipments can exclude some shipments, such as those in free trade zones. Additional information DHS’s report said that the NTC periodically refines, improves, and revises the security criteria for high-risk shipments, which can affect the number of cargo shipments identified as high-risk. Description This metric is a rate comparing the number of potentially high-risk cargo containers scanned at each sea port of entry during a fiscal year to the total number of high-risk cargo containers that entered the United States at each sea port of entry during the previous fiscal year. separate from cargo containers that were reviewed or assessed; instead, DHS tracks these inspection methods collectively. Therefore, DHS also included data on potentially high-risk cargo containers that were reviewed or assessed as well as those that were scanned in its report. Limitations DHS identified: In its report, DHS noted that ratio data are not available for fiscal year 2014 because U.S. Customs and Border Protection did not collect comparable container-level data (as opposed to shipment-level data) in fiscal year 2013. DHS also noted that the totals across the ports or field offices may include duplicate container counts. We identified: NTC officials said that the definition of “high-risk” shipments excludes some shipments, such as those in free trade zones. NTC officials noted that assessing, reviewing, and scanning containers are different activities and reflect different levels of inspection or review. For example, NTC officials said that while all containers are “assessed” in order to determine their risk level, only higher risk containers may be scanned using radiation detection and nonintrusive inspection equipment. Consequently, when DHS included data on containers that were assessed or reviewed but not scanned, the resulting count was higher. In an appendix to its report, DHS presented a column of data called the “percentage of potentially high-risk containers scanned (same fiscal year)” for each fiscal year. Given DHS’s inability to separate data on the different inspection methods, the data in this column included containers that were reviewed by all inspection methods, not just scanning. In its appendix, DHS did not present data on the number of containers that “entered the United States,” even though it used those data to calculate the ratio and they are specified in the National Defense Authorization Act for Fiscal Year 2017. As a result, it is not possible to verify the accuracy of DHS’s ratio calculations. Additional information None. The U.S. maritime border domain encompasses ports, internal or inland waters, and coastal waters, as well as the territorial sea (waters 12 nautical miles seaward of the U.S. coast), contiguous zone (waters adjacent to and seaward of territorial sea and extending 24 nautical miles from shore), and exclusive economic zone (waters seaward of and adjacent to territorial sea and extending out to 200 nautical miles from shore). U.S. Coast Guard (Coast Guard), Air and Marine Operations, and U.S. Border Patrol share responsibility for patrolling the U.S. maritime borders, and territorial sea. The Coast Guard is a component of DHS and the lead federal maritime law enforcement agency on the high seas (waters beyond 12 nautical miles seaward of the U.S. coast) and all other waters under U.S. jurisdiction. The Coast Guard responds to a variety of maritime border security issues, including trafficking of narcotics, people, illicit goods, unlawful migration, illegal exploitation of natural resources, potential terrorist activities, and the disruption of maritime commerce. The metrics in this domain measure the number of migrants and illicit drugs removed, among other things. DHS included 4 of 6 metrics called for in the NDAA for this domain in its fiscal year 2017 Border Security Metrics Report, as shown in table 6. Description This metric was not specifically defined in the National Defense Authorization Act for Fiscal Year 2017 (NDAA). The NDAA described situational awareness as the knowledge and understanding of current unlawful cross- border activity, including (1) threats and trends concerning illicit trafficking and unlawful crossings, (2) the ability to forecast future shifts in such threats and trends, (3) the ability to evaluate such threats and trends at a level sufficient to create actionable plans, and (4) the operational capability to conduct persistent and integrated surveillance of the international borders of the United States. developing a measure for situational awareness in the maritime domain that meets the intent of the NDAA. While this effort is in process, DHS presented data on U.S. Coast Guard and U.S. Customs and Border Protection (CBP) asset (aircraft and cutter or boat) hours contributing to situational awareness or interdiction support and the number of vessel manifests screened. Limitations None identified. Additional information According to CBP Air and Marine Operations officials, they did not have confidence that the data for years prior to fiscal year 2016 were consistent enough for making comparisons across years. Consequently, only data for fiscal year 2016 were included in DHS’s report for the metrics related to CBP. Description This metric is a count of the total number of undocumented migrants interdicted, identified directly or indirectly but not interdicted, or otherwise believed to have unlawfully entered the United States through the maritime border. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations Department of Homeland Security (DHS) identified: The accuracy of migrant flow counts depends on partners to report interdictions and the ability to detect migrants. According to the DHS report, the U.S. Coast Guard relies on partners to report interdictions for compilation in the U.S. Coast Guard’s database. Interdictions may be double-counted by the U.S. Coast Guard and its partners because they cooperate during operations and some interdictions by partners may not get reported. Further, some migrants may not be apprehended and leave no evidence, and are therefore excluded from the known flow figures. We identified: According to U.S. Coast Guard officials, there is no centralized database for tracking migrant interdictions, and the decentralized nature of the data collection could lead to errors. Additional information According to the U.S. Coast Guard, about 90 percent of the data on migrant interdictions and flow originate from U.S. Coast Guard records. U.S. Coast Guard officials said that as part of a department-wide initiative to standardize illegal immigration statistics, they are in the preliminary stages of building a centralized database to enter and maintain information on migrant interdictions. Additionally, officials said they take steps to ensure the reliability of externally reported data such as communicating with partners and working together to reconcile any errors. Within the U.S. Coast Guard, meetings are held regularly to discuss and vet the accuracy of migrant flow data. Description This metric is a rate comparing the amount and type of illicit drugs removed by the Department of Homeland Security (DHS) maritime security components in any fiscal year, including drugs abandoned at sea, to the average amount removed or abandoned in the immediately preceding 5 fiscal years. by all DHS maritime security components, but DHS only provided data on removals by the U.S. Coast Guard. DHS did not explain in its report why it only included data from the U.S. Coast Guard. DHS officials said that the U.S. Coast Guard is the primary DHS component involved in this activity and was the only component that provided data for this metric, but this was not noted in the report. According to U.S. Coast Guard officials, some of the data for fiscal 2013 was misreported. Specifically, the quantity removed for methamphetamine should be 0 (report shows 17.4) while the value should be 7.9 kilograms for heroin (report shows 0). Additional information None. Description This metric is a rate comparing the amount of cocaine removed by the Department of Homeland Security (DHS) maritime security components inside and outside the maritime transit zone to the total documented cocaine flow rate. DHS used estimates of noncommercial maritime cocaine flow from the Consolidated Counter Drug Database, which are derived from intelligence reporting and case data. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations DHS identified: There is less robust intelligence on noncommercial maritime cocaine flow outside the transit zone than inside the transit zone, so data for outside the transit zone are not considered reliable. Precise cocaine flow estimates through a particular mode or domain can be difficult to obtain. In our prior work, officials with the Office of National Drug Control Policy and other departments and agencies involved in U.S. counternarcotics efforts told us that it is difficult to obtain precise estimates of cocaine flow because of the difficulty in obtaining specific information about the production of cocaine and how it gets to the United States. We have also previously reported that when confronted with threats to their activities, drug-trafficking organizations use a variety of techniques to quickly change their modes of operation, thus avoiding capture of their personnel and seizure of their illegal drugs. For example, when air interdiction efforts have proven successful, traffickers have increased their use of maritime and overland transportation routes. Additional information According to U.S. Coast Guard officials, DHS officials hold quarterly inter-agency meetings to review the reliability of performance data related to cocaine interdiction performance. Air and Marine Operations (AMO) is a federal law enforcement agency within CBP that interdicts unlawful people and cargo approaching U.S. borders, investigates criminal networks, and provides domain awareness in the air and maritime environments, among other things. The metrics in this domain measure AMO’s flight hours, individuals detected, and apprehensions, among other things. DHS included 7 of 8 metrics within this domain called for in the NDAA in its fiscal year 2017 Border Security Metrics Report, as shown in table 7. DHS reported that the “AMO actionable intelligence” metric was under review and estimated that the department would provide information on this metric in its 2019 annual report to Congress. Description This metric is a rate comparing the number of flight hour requirements to the number of flight hours flown by Air and Marine Operations (AMO) in the land domain. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations Department of Homeland Security (DHS) identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. We identified: DHS used the terms “funded flight hours,” “unfunded flight hours,” and “unconstrained flight hours” in the report without clearly defining them. AMO officials stated that a definition of these terms will be included in the next report. Additional information AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Description This metric is a rate comparing the number of funded flight hours appropriated to Air and Marine Operations (AMO) to the number of actual flight hours flown. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations Department of Homeland Security (DHS) identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. Additional information AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Description This metric is a rate comparing the number of aviation missions flown by Air and Marine Operations (AMO) to the number of aviation missions cancelled by AMO due to maintenance, operations, or other causes. the number of missions cancelled due to causes within AMO control, such as maintenance, personnel, and asset availability. However, the Department of Homeland Security (DHS) used the total number of mission requests, which also includes the number of missions flown in addition to the number of missions cancelled for reasons within AMO control. Limitations DHS identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. Additional information AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Description This metric is a rate comparing the number of missions cancelled by Air and Marine Operations (AMO) due to weather compared to the total planned missions. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations Department of Homeland Security (DHS) identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. Additional information AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Description This metric is a count of the number of individuals detected by Air and Marine Operations (AMO) through the use of unmanned aerial systems and manned aircraft. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None identified. Limitations Department of Homeland Security (DHS) identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. DHS data on detections from manned aircraft were limited to those that led to apprehensions and arrests, and data from unmanned aircraft were limited to the number of Vehicle and Dismount Exploitation Radar (VADER) detections. AMO did not track data from all sensors on unmanned and manned aircraft, and considers this metric to be a work in progress. We identified: In February 2017 we reported that some mission data (such as asset assists) for unmanned aerial systems were collected inconsistently across operation locations, which could affect the accuracy of the counts provided. We recommended that U.S. Customs and Border Protection—of which AMO is a component—update and maintain guidance for recording mission information in its data collection system, and provide training to users of the system. DHS completed implementation of these recommendations in July 2018. Although the recommendations have been implemented, this limitation is relevant because the data presented (for fiscal year 2016) were collected prior to their implementation. Additional information and planned actions by the Department of Homeland Security DHS expects to provide more comprehensive data for this metric in the next annual report. AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Description This metric is a count of the number of apprehensions assisted by Air and Marine Operations (AMO) through the use of unmanned aerial systems and manned aircraft. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None. Limitations Department of Homeland Security (DHS) identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. We identified: In February 2017 we reported that some mission data (such as asset assists) for unmanned aerial systems were collected inconsistently across operation locations, which could affect the accuracy of the counts provided. We recommended that U.S. Customs and Border Protection—of which AMO is a component—update and maintain guidance for recording mission information in its data collection system, and provide training to users of the system. DHS completed implementation of these recommendations in July 2018. Although the recommendations have been implemented, this limitation is relevant because the data presented (for fiscal year 2016) were collected prior to their implementation. Additional information In addition to the number of apprehensions assisted, DHS also provided the number of enforcement flight hours used for the assists. AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Description This metric is a count of the number and quantity of illicit drug seizures assisted by Air and Marine Operations (AMO) through the use of unmanned aerial systems and manned aircraft. Differences between the National Defense Authorization Act for Fiscal Year 2017 and the Department of Homeland Security’s report None. Limitations Department of Homeland Security (DHS) identified: Data prior to fiscal year 2016 were unavailable. According to AMO officials, this is because AMO did not collect these data prior to fiscal year 2016, or because older data were not comparable. We identified: In February 2017 we reported that some mission data (such as asset assists) for unmanned aerial systems were collected inconsistently across operation locations, which could affect the accuracy of the counts provided. We recommended that U.S. Customs and Border Protection—of which AMO is a component—update and maintain guidance for recording mission information in its data collection system, and providing training to users of the system. DHS completed implementation of these recommendations in July 2018. Although the recommendations have been implemented, this limitation is relevant because the data presented (for fiscal year 2016) were collected prior to their implementation. Additional information In addition to the drug seizures assisted (in pounds), DHS also provided the number of enforcement flight hours used for the assists. AMO officials said they have taken steps to improve how they track flight hour data, such as by adding new data fields to AMO’s system and providing training to staff. Based on findings from previous reviews of border security programs and efforts, we have recommended other metrics that the Department of Homeland Security (DHS) could use to help measure the effectiveness of border security. The tables that follow provide information about these recommended metrics in each of the four domains listed in the National Defense Authorization Act for Fiscal Year 2017—between ports of entry, at ports of entry, in the maritime border domain, and the air and marine security in the land domain. The recommendations listed in the tables below remain open; however, implementing them would provide DHS with additional indicators and metrics that could provide important insights into the state of border security. Appendix III: Comments from the Department of Homeland Security Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Taylor Matheson (Assistant Director), David Alexander, Kelsey Burdick, Lilia Chaidez, Kathleen Donovan, Michele Fejfar, Sally Gilley, Christopher Hatscher, Eric Hauswirth, Mikaela Meyer, Sasan J. “Jon” Najmi, Kevin Reeves, and Jeff Tessin made key contributions to this report.

Why GAO Did This Study According to DHS, the United States has approximately 6,000 miles of land borders, 95,000 miles of coastline, and more than 300 ports of entry where travelers and cargo are inspected and processed for entry. Securing U.S. border areas is a key part of DHS's mission, and the department's ability to measure its border security efforts is essential for it to manage its responsibilities effectively and efficiently. The NDAA for Fiscal Year 2017 requires DHS to report annually on 43 border security metrics. DHS issued its first report in May 2018. The Act also includes a provision for GAO, within 270 days of receipt and biennially for the following 10 years, to review and report on the data and methodology contained in DHS's report. This report assesses the extent to which DHS: (1) reported metrics as outlined in the NDAA using quality information; and (2) validated assumptions and conveyed statistical uncertainty for unlawful entry metrics, among other objectives. GAO assessed the methodology and data in DHS's report, analyzed DHS's use of statistical models, and interviewed officials from DHS offices and components involved in developing the metrics. What GAO Found The Department of Homeland Security (DHS) reported on 35 of 43 metrics called for by the National Defense Authorization Act (NDAA) for Fiscal Year 2017 (see figure); it generally used quality information, but did not identify some data limitations. GAO found that about half of the 35 metrics generally included elements as called for by the NDAA, while 17 metrics differed, such as in scope or calculation. For example, DHS only provided information on the southwest border for some metrics, such as the estimate of undetected unlawful border crossers for which a methodology for estimating unlawful crossings for the northern border had not yet been completed. DHS components responsible for collecting the metric data generally have processes in place to ensure the reliability of the data and the quality of the information provided. DHS also identified and disclosed limitations for some, but not all, of the data elements and metrics used. For example, GAO found that DHS did not disclose limitations on data related to apprehensions of individuals that were assisted by unmanned aerial systems. By developing and implementing a process to systematically review the reliability of the data and comprehensively identify and communicate limitations, DHS would improve the quality of the information provided. DHS used a statistical model to estimate three metrics on unlawful border entries but did not validate some assumptions the model employs through sensitivity analyses and provide measures of statistical uncertainty in accordance with standards for federal agencies. For example, DHS's model assumes that 100 percent of families unlawfully crossing the border will be apprehended, but DHS did not provide information on the extent to which the assumption affected its metrics. DHS also did not provide information on the level of statistical uncertainty for the metrics, such as margins of error. Providing such information would allow Congress and the public to better understand the potential limitations and accuracy of these metrics of unlawful entry. Additionally, DHS's statistical model, which is based on Mexican adults not seeking asylum, represents a small and declining share of those apprehended at the border and DHS is developing a new model to account for current border conditions. What GAO Recommends GAO is making four recommendations, including that DHS develop and implement a process to systematically review the reliability of metric data, identify and communicate limitations of the metrics, and include the results of sensitivity analyses and measures of statistical uncertainty for metrics derived from statistical models. DHS concurred with the recommendations.

gao_GAO-19-402T

gao_GAO-19-402T_0

DHS Has Made Progress Addressing Past Challenges, But Some Actions are Still Under Way Our past work has identified progress and challenges in a number of areas related to DHS’s management of the CFATS program, including (1) the process for identifying high-risk chemical facilities; (2) how it assesses risk and prioritizes facilities; (3) reviewing and approving facility site security plans; (4) inspecting facilities and ensuring compliance; and (5) efforts to conduct outreach with stakeholders and first responders. Identifying High-Risk Chemical Facilities In May 2014, we found that more than 1,300 facilities had reported having ammonium nitrate to DHS. However, based on our review of state data and records, there were more facilities with ammonium nitrate holdings than those that had reported to DHS under the CFATS program. Thus, we concluded that some facilities weren’t required to report to DHS and some that were required may have failed to do so. We recommended that DHS work with other agencies, including the Environmental Protection Agency (EPA), to develop and implement methods of improving data sharing among agencies and with states as members of a Chemical Facility Safety and Security Working Group. DHS agreed with our recommendation and has since addressed it. Specifically, DHS compared DHS data with data from other federal agencies, such as EPA, as well as member states from the Chemical Facility Safety and Security Working Group to identify potentially noncompliant facilities. As a result of this effort, in July 2015, DHS officials reported that they had identified about 1,000 additional facilities that should have reported information to comply with CFATS and subsequently contacted these facilities to ensure compliance. DHS officials told us that they continue to engage with states to identify potentially non-compliant facilities. For example, as of June 2018, DHS officials stated that they have received 43 lists of potentially noncompliant facilities from 34 state governments, which are in various stages of review by DHS. DHS officials also told us that they hired an individual to serve as the lead staff member responsible for overseeing this effort. DHS has also taken action to strengthen the accuracy of data it uses to identify high-risk facilities. In July 2015, we found that DHS used self- reported and unverified data to determine the risk categorization for facilities that held toxic chemicals that could threaten surrounding communities if released. At the time, DHS required that facilities self- report the Distance of Concern—an area in which exposure to a toxic chemical cloud could cause serious injury or fatalities from short-term exposure—as part of its Top-Screen. We estimated that more than 2,700 facilities with a toxic release threat had misreported the Distance of Concern and therefore recommended that DHS (1) develop a plan to implement a new Top-Screen to address errors in the Distance of Concern submitted by facilities, and (2) identify potentially miscategorized facilities that could cause the greatest harm and verify that the Distance of Concern of these facilities report is accurate. DHS has fully addressed both of these recommendations. Specifically, in response to the first recommendation, DHS implemented an updated Top-Screen survey in October 2016 and now collects data from facilities and conducts more accurate modeling to determine the actual area of impact (formerly called the Distance of Concern), rather than relying on the facilities’ calculation. In response to the second recommendation, DHS officials reported in November 2016 that they reassessed all facility Top-Screens that reported threshold quantities of chemicals posing a toxic release threat, and identified 158 facilities with the potential to cause the greatest harm. In April 2018, DHS officials reported that all of these facilities have since been reassessed using updated Top-Screen information and, where appropriate, assigned a risk tier. Assessing Risk and Prioritizing Facilities DHS has also taken actions to better assess regulated facilities’ risks in order to place the facilities into the appropriate risk tier. In April 2013, we reported that DHS’s risk assessment approach did not consider all of the elements of threat, vulnerability, and consequence associated with a terrorist attack involving certain chemicals. Our work showed that DHS’s CFATS risk assessment methodology was based primarily on consequences from human casualties, but did not consider economic consequences, as called for by the National Infrastructure Protection Plan (NIPP) and the CFATS regulation. We also found that (1) DHS’s approach was not consistent with the NIPP because it treated every facility as equally vulnerable to a terrorist attack regardless of location or on-site security and (2) DHS was not using threat data for 90 percent of the tiered facilities—those tiered for the risk of theft or diversion—and using 5-year-old threat data for the remaining 10 percent of those facilities that were tiered for the risks of toxic chemical release or sabotage. We recommended that DHS enhance its risk assessment approach to incorporate all elements of risk and conduct an independent peer review after doing so. DHS agreed with our recommendations and has implemented actions to address both of them. Specifically, with regard to our recommendation that DHS enhance its risk assessment approach to incorporate all elements of risk, DHS worked with Sandia National Laboratories to develop a model to estimate the economic consequences of a chemical attack. In addition, DHS worked with Oak Ridge National Laboratory to devise a new tiering methodology, called the Second Generation Risk Engine. In so doing, DHS revised the CFATS threat, vulnerability, and consequence scoring methods to better cover the range of CFATS security issues. Additionally, with regard to our recommendation that DHS conduct a peer review after enhancing its risk assessment approach, DHS conducted peer reviews and technical reviews with government organizations and facility owners and operators, and worked with Sandia National Laboratories to verify and validate the CFATS program’s revised risk assessment methodology. To further enhance its risk assessment approach, in the fall of 2016, DHS also revised its Chemical Security Assessment Tool (CSAT), which supports DHS efforts to gather information from facilities to assess their risk. According to DHS officials, the new tool—called CSAT 2.0—is intended to eliminate duplication and confusion associated with DHS’s original CSAT. DHS officials told us that they have improved the tool by revising some questions in the original CSAT to make them easier to understand; eliminating some questions; and pre-populating data from one part of the tool to another so that users do not have to retype the same information multiple times. DHS officials also told us that the facilities that have used the CSAT 2.0 have provided favorable feedback that the new tool is more efficient and less burdensome than the original CSAT. Finally, DHS officials told us that, as of June 2018, DHS completed all notifications and processed tiering results for all but 226 facilities. DHS officials did not provide an estimated target completion date for these pending risk assessments, noting that completing the assessments is highly dependent on the facilities providing the necessary Top-Screen information. Reviewing and Approving Facility Site Security Plans DHS has also made progress reviewing and approving facility site security plans by reducing the time it takes to review these plans and eliminating the backlog of plans awaiting review. In April 2013, we reported that DHS revised its procedures for reviewing facilities’ security plans to address DHS managers’ concerns that the original process was slow, overly complicated, and caused bottlenecks in approving plans. We estimated that it could take DHS another 7 to 9 years to review the approximately 3,120 plans in its queue at that time. We also estimated that, given the additional time needed to do compliance inspections, the CFATS program would likely be implemented in 8 to 10 years. We did not make any recommendations for DHS to improve its procedures for reviewing facilities’ security plans because DHS officials reported that they were exploring ways to expedite the process, such as reprioritizing resources and streamlining inspection requirements. In July 2015, we reported that DHS had made substantial progress in addressing the backlog—estimating that it could take between 9 and 12 months for DHS to review and approve security plans for the approximately 900 remaining facilities. DHS officials attributed the increased approval rate to efficiencies in DHS’s review process, updated guidance, and a new case management system. Subsequently, DHS reported in its December 2016 semi-annual report to Congress that it had eliminated its approval backlog. Finally, we found in our 2017 review that DHS took action to implement an Expedited Approval Program (EAP). The CFATS Act of 2014 required that DHS create the EAP as another option that tier 3 and tier 4 chemical facilities may use to develop and submit security plans to DHS. Under the program, these tier 3 and 4 facilities may develop a security plan based on specific standards published by DHS (as opposed to the more flexible performance standards using the standard, non- expedited process). DHS issued guidance intended to help facilities prepare and submit their EAP security plans to DHS, which includes an example that identifies prescriptive security measures that facilities are to have in place. According to committee report language, the EAP was expected to reduce the regulatory burden on smaller chemical companies, which may lack the compliance infrastructure and the resources of large chemical facilities, and help DHS to process security plans more quickly. If a tier 3 or 4 facility chooses to use the expedited option, DHS is to review the plan to determine if it is facially deficient, pursuant to the reporting requirements of the CFATS Act of 2014. If DHS approves the EAP site security plan, it is to subsequently conduct a compliance inspection. In 2017, we found that DHS had implemented the EAP and had reported to Congress on the program, as required by the CFATS Act of 2014. In addition, as of June 2018, according to DHS officials, only 18 of the 3,152 facilities eligible to use the EAP had opted to use it. DHS officials attributed the low participation to several possible factors including: DHS had implemented the expedited program after most eligible facilities already submitted standard (non-expedited) security plans to DHS; facilities may consider the expedited program’s security measures to be too strict and prescriptive, not providing facilities the flexibility of the standard process; and the lack of an authorization inspection may discourage some facilities from using the expedited program because this inspection provides useful information about a facility’s security. We also found in 2017 that recent changes made to the CFATS program could affect the future use of the expedited program. As discussed previously, DHS has revised its methodology for determining the level of each facility’s security risk, which could affect a facility’s eligibility to participate in the EAP. Inspecting Facilities and Ensuring Compliance In our July 2015 report, we found that DHS began conducting compliance inspections in September 2013, and by April 2015, had conducted inspections of 83 of the inspected 1,727 facilities that had approved security plans. Our analysis showed that nearly half of the facilities were not fully compliant with their approved site security plans and that DHS had not used its authority to issue penalties because DHS officials found it more productive to work with facilities to bring them into compliance. We also found that DHS did not have documented processes and procedures for managing the compliance of facilities that had not implemented planned measures by the deadlines outlined in their plans. We recommended that DHS document processes and procedures for managing compliance to provide more reasonable assurance that facilities implement planned measures and address security gaps. DHS agreed and has since taken steps toward implementing this recommendation. Specifically, DHS revised CFATS Standard Operating Procedures that, as of February 2019, we are reviewing to determine if they sufficiently document the processes and procedures currently being used to track noncompliant facilities and ensure facilities implement planned measures as outlined in their approved site security plans. In August 2018, we reported that our analysis of DHS data since our 2015 report showed that DHS has made substantial progress in conducting and completing compliance inspections. Specifically, our analysis showed that DHS increased the number of compliance inspections completed per year since DHS began conducting compliance inspections in 2013 and that, for the 2,466 high-risk facilities with an approved site security plan as of May 2018, DHS had conducted 3,553 compliance inspections. Of these, DHS issued corrective actions to two facilities that were not in compliance with their approved site security plan. In our August 2018 report, we also found that DHS developed a new methodology and performance measure for the CFATS program in order to evaluate security changes made by high-risk chemical facilities, but that the methodology does not measure the program’s impact on reducing a facility’s vulnerability to an attack. We found that DHS could take steps to evaluate vulnerability reduction resulting from the CFATS compliance inspection process. We recommended that DHS incorporate vulnerability into the new methodology to help measure the reduction in the vulnerability of high-risk facilities to a terrorist attack, and use that data in assessing the CFATS program’s performance in lowering risk and enhancing national security. DHS agreed and is taking steps to implement this recommendation. Specifically, in September 2018, DHS reported making progress towards the implementation of two new performance metrics by the end of the first quarter of fiscal year 2019. DHS officials stated that these metrics should, among other things, evaluate the progress of individual facilities in enhancing their security while part of the CFATS program and be used to demonstrate an increase in the security posture across the population of CFATS facilities. Conducting Stakeholder and First Responder Outreach In April 2013, we reported that DHS took various actions to work with facility owners and operators, including increasing the number of visits to facilities to discuss enhancing security plans, but that some trade associations had mixed views on the effectiveness of DHS’s outreach. We found that DHS solicited informal feedback from facility owners and operators in its efforts to communicate and work with them, but did not have an approach for obtaining systematic feedback on its outreach activities. We recommended that DHS take action to solicit and document feedback on facility outreach consistent with DHS efforts to develop a strategic communication plan. DHS agreed and has implemented this recommendation by developing a questionnaire to solicit feedback on outreach with industry stakeholders and began using the questionnaire in October 2016. In August 2018, we reported that DHS shares some CFATS information with first responders and emergency planners, but these stakeholders may not have all of the information they need to minimize the risk of injury or death when responding to incidents at high-risk facilities. While certain facilities are required under the Emergency Planning and Community Right-to-Know Act of 1986 to report some chemical inventory information, which local officials told us they rely on to prepare for and respond to incidents at chemical facilities, we found over 200 chemicals covered by CFATS that may not be covered by these reporting requirements. We also reported that DHS developed a secure interface called the Infrastructure Protection (IP) Gateway that provides access to CFATS facility-specific information that may be missing from required reporting. However, we found that the IP Gateway is not widely used at the local level and officials from 13 of 15 selected Local Emergency Planning Committees we contacted—consisting of first responders and covering 373 CFATS high-risk facilities—said they did not have access to CFATS data in the IP Gateway. We recommended that DHS should take actions to encourage access to and wider use of the IP Gateway and explore other opportunities to improve information-sharing with first responders and emergency planners. DHS concurred with this recommendation and reported in September 2018 that they are taking actions to implement it. Specifically, DHS has revised three fact sheets and an outreach presentation to include information on the IP Gateway and how to request access to it. In addition, DHS plans to ensure contact is made with first responders representing the top 25 percent of CFATS high-risk chemical facilities by no later than March 2019 so that they are properly prepared to respond to incidents at these facilities. Chairman Thompson, Ranking Member Rogers, and Members of the Committee, this completes my prepared statement. I would be pleased to respond to any questions that you may have at this time. GAO Contact and Staff Acknowledgements If you or your staff members have any questions about this testimony, please contact me at (202) 512-3841 or andersonn@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Other individuals making key contributions to this work include Ben Atwater, Assistant Director; Hugh Paquette, Analyst-in-Charge; Chuck Bausell, Michele Fejfar, Tracey King, and Tom Lombardi. 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.

Why GAO Did This Study Thousands of facilities across the United States contain hazardous chemicals that could be used by terrorists to inflict mass casualties or harm surrounding populations. In accordance with the DHS Appropriations Act, 2007, DHS established the CFATS program to, among other things, identify and assess the security risk posed by chemical facilities. DHS inspects high-risk facilities after it approves facility security plans to ensure that the facilities are implementing required security measures and procedures. This statement summarizes progress and challenges related to DHS's CFATS program management. This statement is based on prior products GAO issued from July 2012 through August 2018, along with updates as of September 2018 on actions DHS has taken to address GAO's prior recommendations. To conduct the prior work, GAO reviewed relevant laws, regulations, and DHS policies for administering the CFATS program; how DHS assesses risk; and data on high-risk chemical facilities. GAO also interviewed DHS officials and relevant stakeholders. What GAO Found The Department of Homeland Security (DHS) has made progress addressing challenges that GAO's past work identified to managing the Chemical Facility Anti-Terrorism Standards (CFATS) program. The following summarizes progress made and challenges remaining in key aspects of the program. Identifying high-risk chemical facilities. In July 2015, GAO reported that DHS used self-reported and unverified data to determine the risk of facilities holding toxic chemicals that could threaten surrounding communities if released. GAO recommended that DHS should better verify the accuracy of facility-reported data. DHS implemented this recommendation by revising its methodology so it now calculates the risk of toxic release, rather than relying on facilities to do so. Assessing risk and prioritizing facilities. In April 2013, GAO reported weaknesses in multiple aspects of DHS's risk assessment and prioritization approach. To improve this process, GAO recommended that DHS enhance its risk assessment approach to incorporate all elements of risk and conduct a peer review after doing so. DHS implemented both recommendations by revising the CFATS risk assessment methodology to include threat, vulnerability, and consequence to better cover the range of security issues, and conducting peer reviews and technical reviews to verify and validate the CFATS program's new risk assessment approach. Reviewing and approving facility site security plans . DHS is to review facility security plans to ensure their security measures meet DHS standards. In April 2013, GAO reported a 7- to 9-year backlog for these reviews. In July 2015, GAO reported that DHS had made substantial progress in addressing the backlog—estimating that it could take between 9 and 12 months for DHS to review and approve security plans for the approximately 900 remaining facilities. DHS has since taken additional action to expedite these activities and has eliminated this backlog. Inspecting facilities and ensuring compliance. In July 2015, GAO found that nearly half of the facilities DHS had inspected were not fully compliant with their approved security plans and that DHS did not have documented procedures for managing facilities' compliance. GAO recommended that DHS document procedures for managing compliance. DHS revised CFATS procedures that, as of February 2019, GAO is reviewing to determine if they sufficiently address the recommendation. Conducting stakeholder and first responder outreach. In August 2018, GAO reported that DHS shares some CFATS information with first responders and emergency planners but these stakeholders may not have all of the information they need to minimize the risk of injury or death when responding to incidents at high-risk facilities. GAO recommended that DHS should, among other things, take actions to explore opportunities to improve information-sharing with first responders and emergency planners. DHS concurred with this recommendation and reported in September 2018 that it is conducting additional outreach and taking other actions to implement it. What GAO Recommends GAO has made various recommendations to strengthen DHS's management of the CFATS program, with which DHS has generally agreed. DHS has implemented or described planned actions to address these recommendations.

gao_GAO-20-118

gao_GAO-20-118_0

Background The CSA and DEA Registration The Controlled Substances Act (CSA) was enacted in 1970 to regulate and facilitate the use of controlled substances, including certain prescription drugs such as opioid pain relievers, for legitimate medical, scientific, research, and industrial purposes while preventing them from being diverted for illegal uses. According to DEA, the CSA requires DEA to maintain a “closed system” of distribution, which includes limiting the amount of certain controlled substances that are available in the marketplace by setting quotas. Various CSA provisions also require persons who handle controlled substances to register with the DEA. This includes businesses that import, export, manufacture, or distribute controlled substances; certain health care practitioners, such as physicians, licensed to dispense, administer, or prescribe them; and pharmacies authorized to fill prescriptions, referred to as “registrants.” The registration mechanism creates a “closed system” of distribution in which distribution may lawfully occur among the registrants. The closed system of distribution, along with registrant compliance with the CSA’s regulatory requirements, helps to ensure that a particular controlled substance is always accounted for by a DEA-registered entity, from its creation until it is dispensed to a patient or is destroyed. The CSA places controlled substances in one of five schedules based generally on findings related to the substance, including whether the substance has a currently accepted medical use in treatment in the United States, its relative potential for abuse, and the degree of dependence the drug or other substance may cause. For further information on this and other legal requirements, please see appendix II. Legitimate Use of Prescription Drugs, Drug Supply Chain, and Opportunities for Opioid Abuse and Diversion The prescription drug supply chain is the means through which prescription drugs are ultimately delivered to patients with legitimate medical needs. Although there can be many variations in the flow of prescription drugs through the supply chain, in a common example, prescription drugs are produced by manufacturers; are purchased and stored by distributors, who take orders and deliver them to customers such as pharmacies; and ultimately are dispensed by pharmacies to patients who have a prescription from a practitioner, as shown in figure 1. Although prescription drugs are intended for legitimate medical uses, the prescription drug supply chain may present opportunities for the drugs to be diverted and abused as the drugs move through the various components of the supply chain. For example, an individual may visit multiple practitioners posing as a legitimate patient, referred to as a doctor shopper, to obtain prescriptions for drugs for themselves or others, or criminal enterprises may rob distributors and pharmacies of prescription drugs to sell to others. Roles and Responsibilities Related to Controlled Substances DEA, through its Diversion Control Division, is responsible for preventing, detecting and investigating the diversion of controlled substances from legitimate sources while ensuring an adequate and uninterrupted supply is available for legitimate medical, commercial, and scientific needs. The division is responsible for enforcing the CSA and its regulations pertaining to pharmaceutical controlled substances and listed chemicals. In doing so, it conducts domestic investigations, among other things, in DEA’s 23 field division offices. By law, generally, manufacturers, distributors, and reverse distributors are required to report to DEA every sale, delivery, or other disposal of any controlled substance. As we previously reported, manufacturers and distributors of schedules I and II drugs and schedule III narcotics must file reports with DEA through ARCOS, a drug reporting system that allows the agency to monitor the flow of DEA controlled substances from their point of manufacture through commercial distribution channels to point of sale or distribution at the dispensing/retail level. In addition, certain schedule III non-narcotics and some schedule IV narcotics are also covered by the ARCOS reporting requirements. DEA implemented the ARCOS database in 1997, and approximately 1,250 distributors, manufacturers, and reverse distributors report more than 72 million transactions into ARCOS each year, according to DEA. Generally, certain registrants must report certain data at least quarterly and they have the option to report voluntarily on a monthly basis. By law, each registrant, such as manufacturers and distributors of controlled substances, is required (1) to design and operate a system that is compliant with applicable federal and state privacy laws to identify suspicious orders of controlled substances, and (2) upon discovering a suspicious order or series of orders, notify the DEA Administrator and the special agent in charge of the appropriate DEA field division office. The SUPPORT Act, which amended the CSA in part, also includes requirements related to preventing drug diversion. The SUPPORT Act provisions require the DEA Administrator to establish a centralized database for collecting suspicious orders reports, which is discussed in more detail below. In addition, the SUPPORT Act requires the Attorney General to make certain data available to registered manufacturers and distributors through ARCOS. The SUPPORT Act also requires the Attorney General to submit to Congress a report that provides information about how the Attorney General is using ARCOS data to identify and stop suspicious activity no later than one year after the date of enactment of the SUPPORT Act. DEA Diversion-Related Data Systems DEA operates and maintains various information systems containing registrant information, transaction data, and suspicious drug orders that support its efforts to prevent, detect, and investigate the diversion of pharmaceutical controlled substances. These include Controlled Substance Ordering System (CSOS). This system is used primarily by manufacturers and distributors, as well as pharmacies and hospitals to place orders for controlled substances in a secure electronic environment, and includes information such as the number of packages, size of packages and name of items ordered, according to DEA. ARCOS. As discussed above, ARCOS monitors the flow of transactions of schedule I, II, III and select schedule IV controlled substances from their point of manufacture to their point of sale or distribution at the dispensing or retail level (such as hospitals, retail pharmacies, practitioners, and teaching institutions). The data in ARCOS are used to, among other things, track regulatory compliance in the pharmaceutical drug industry and to detect abuse of legally manufactured pharmaceuticals that are diverted to illegal markets, according to DEA Diversion Control Division officials. Suspicious Order Reporting System (SORS). DEA developed SORS to receive and store suspicious order reports. To date, DEA has developed three versions of SORS as described below. SORS Online version. In late October 2019, DEA launched the Suspicious Orders Report System (SORS) Online, a centralized database required by the SUPPORT Act, for registrants that distribute controlled substances to report suspicious orders to DEA. Reporting a suspicious order to SORS Online constitutes compliance with the reporting requirement that registrants notify the Administrator of the Drug Enforcement Administration and the Special Agent in Charge of the Division Office of the Drug Enforcement Administration for the area in which the registrant is located or conducts business. SORS online is the third version of DEA’s SORS system that was originally developed in 2008. Unlike earlier SORS versions, SORS Online requires users to provide a reason an order is suspicious. At the time of our study, the use of SORS Online was voluntary. Registrants who are under active MOAs with DEA are reporting to the new SORS Online system, according to DEA. Follow-up version. Suspicious order reports reported by registrants since March 2017 operating under an active memorandum of agreement (MOA) with the DEA that required them to submit their reports electronically to DEA headquarters as opposed to their local DEA field division office using the SORS Follow-up version, according to DEA Diversion Control Division officials. Initial version. The initial version of SORS stores suspicious order reports for registrants with an expired MOA but who elected to voluntarily continue to report suspicious orders in the same way as under the MOA, according to DEA Diversion Control Division officials. The initial version of SORS was established in 2008. Figure 2 provides an overview of the information DEA obtains and uses to support its diversion control efforts. State Prescription Drug Monitoring Programs (PDMP) A PDMP is an electronic database that tracks controlled substance prescriptions, managed within and at the state level. State PDMPs can provide health care providers and authorities timely information about prescribing and patient behaviors that may indicate drug abuse or diversion and facilitate a response. Authorized users, such as practitioners and pharmacists, may access information submitted to PDMPs by dispensers. A state’s PDMP is housed by a specified statewide regulatory, administrative or law enforcement agency. The PDMP distributes data from the database to individuals who are authorized under state law to receive the data for purposes of their profession. PDMP data can assist law enforcement and health care providers such as practitioners and pharmacists in identifying patterns of prescribing, dispensing, or receiving controlled substances that may indicate abuse or diversion. PDMPs vary in numerous ways across states, including what data they collect; what drugs they cover; who has access to, or who is required to use, the prescription drug monitoring program; and which state agency oversees and administers the program. DEA may request state PDMP data through submitting requests or subpoenas to the state official operating the PDMP database, for example, to support diversion control investigations. The requirements on requesting and accessing state PDMP vary from state to state according to DEA Diversion Control Division officials. Officials noted that the different state-by-state requirements create difficulties for federal law enforcement during a multi- state or national case as law enforcements’ requests for data have to be addressed at the state level. Data Analytics Data-analytics activities can include a variety of techniques to prevent and detect diversion, including data matching and data mining. Data matching is the large scale comparison of records and files to detect errors or incorrect information. It can be used to verify information provided by recipients or detect unreported changes. Data mining is the use of automated computer algorithms to detect patterns, including those that are otherwise not obvious, correlations, or anomalies within large data sets indicative of potential diversion. Entities may identify many types of analytics techniques that can be used to address improper transactions, such as Rules based – Identify suspicious orders with rules, such as orders that go above a threshold; Anomaly – Detect individual and aggregated abnormal patterns versus peer group, for example, the orders from one pharmacy compared to other pharmacies in the same geographic area; and Predictive – Assess against known diversion. A provider that has characteristics similar to those of known bad actors. DEA Collects Industry-Reported Data to Help Address Opioid Diversion, but Opportunities Exist to Improve its Management and Use of Data DEA Uses Self-Reported Industry Purchase Data to Help Identify and Address Opioid Diversion Activities DEA uses industry-reported ARCOS data to help generate leads, support enforcement actions, and allocate resources. The agency uses these data in a number of ways, including supporting field diversion control activities and developing analytical products. Field-Based requests for data analysis. DEA’s Diversion Control Division’s ARCOS Unit responds to requests for data analysis from its field division offices in support of diversion control enforcement activities. According to DEA officials, this unit is responsible for the collection, maintenance, and analysis of ARCOS data. For example, DEA said this unit conducts analysis on controlled substances that are bought and sold in a particular timeframe between a seller and a buyer. The ARCOS Unit also obtains information on the quantity, dosage units, grams, and ingredients of the drugs in the sale and conducts analyses in response to specific requests from field-based investigators who send their requests to the unit. For example, DEA officials said that out of the 800 field division office requests for analysis sent to the DEA ARCOS Unit in calendar year 2018, about 60 percent of those were for “enhanced” validations. This process includes a controlled substance report which the unit provides to field investigators for their use during scheduled drug investigations, and contains a summary of, among other things, an ARCOS registrant’s reported sales and purchases compared against what other registrants report was sold to them. This process uses both automatic and manual checks. According to DEA officials, they received approximately 480 requests for enhanced validations from DEA field investigators in 2018. While DEA officials noted that DEA’s enhanced validation procedures are not documented, they acknowledged that the ARCOS Unit is in the process of developing standard operating procedures for ARCOS data quality control, including the enhanced validation process. All requests for validations submitted to the ARCOS Unit are analyzed and compiled, and sent to field-based investigators to support scheduled investigations. Although validations are primarily requested for scheduled investigations, field offices can request these reports pursuant to any scheduled or non-scheduled investigation. DEA Analytic Product - Drug Profiles. Using ARCOS data, DEA creates drug profiles for suspected bad actors at the retail level (such as certain pharmacies), who have irregular transactions–also known as outliers, according to DEA officials–in a specific area or zip code and provides this information to its field division offices. The ARCOS Unit compares this suspected “bad actor” with other area competitors. DEA Analytic Product – Annual Threat Assessments. DEA’s ARCOS Unit also uses ARCOS data to develop threat assessments annually to aid field investigators. The threat assessments use ARCOS data to provide drug-related transaction trends and patterns related to a given DEA field division office area of operations to help establish priorities and allocate resources. DEA officials noted that field division office staff use these assessments to develop work plans identifying which registrants will be subject to the office’s routine regulatory investigation that year. Field Querying of ARCOS Data. – Field division offices may also use ARCOS querying tools to analyze ARCOS data to proactively identify diversion targets, such as reviewing ARCOS data to identify information on top purchasers of controlled substances. In a written response to our questions, DEA officials told us that several ARCOS drug profiles they developed have contributed to state and federal administrative, criminal and regulatory investigations. DEA officials recently informed us that as part of a reorganization, it established the Targeting and Special Projects Section whose goal is to focus on leveraging DEA’s data capabilities. Specifically, this section is composed of two units, including the Reports Analytics Unit and the Targeting and Special Projects Unit, which was established in March 2019 and is responsible for conducting data analytics on ARCOS and other data, according to DEA Diversion Control Division officials. DEA is currently working to determine the types of analysis these units will conduct. DEA Conducts Limited Analysis of Industry- Reported Data Using Automated Computer Algorithms We found that while DEA uses ARCOS data to support ongoing investigations and conducts analysis on this data to identify investigative leads for its field division offices, it could conduct more robust analysis using automated computer algorithms to help identify questionable patterns in the data. This analysis in turn could be used to identify registrants that need to be investigated. According to DEA officials, most of the analysis DEA currently conducts on ARCOS data is used by the field division offices. For example, upon receiving information on pharmacies that have a high frequency of reporting stolen or lost-in-transit drugs a field division office may contact DEA’s ARCOS Unit to request ARCOS information. DEA then analyzes the ARCOS data to produce the requested reports to support the field’s ongoing investigations. DEA also conducts routine analysis of ARCOS data to identify high volumes of drugs sold by a distributor to a single purchaser, high volumes of drugs purchased by a single purchaser, and trends in drugs sold or purchased in a given geographic area compared to similar nearby areas. DEA officials also identified one type of analysis it conducts using a computer algorithm. Specifically, DEA uses a computer algorithm when comparing large volumes of drugs purchased in a given geographic area to the area’s population data. According to DEA officials, DEA conducts this analysis quarterly. However, DEA did not report conducting active and recurring monitoring of transactions using algorithms to detect and flag transactions that indicate potential diversion, either on a real-time or near real-time basis. We identified several additional opportunities for DEA to proactively analyze ARCOS data using computer algorithms to identify unusual patterns of drug distribution on a more routine basis. Such analyses could be used to proactively support or generate leads for investigations of potential drug diversion. For example, DEA could Analyze ARCOS data to identify unusual volumes of drugs that were disposed of rather than sold. Conduct analysis of ARCOS data to identify unusual numbers of deleted transactions or deletions of transactions of high volumes of drugs. Analyze ARCOS data by comparing the amount of drugs being acquired by a registrant to the amount of drugs accounted for, through being sold or disposed of, among other things, by each registrant to determine any differences. Analyze ARCOS data to identify trends in distribution or purchases of drugs in a given geographic area. DEA could look for unusual patterns when comparing such activity in an area with that of other nearby areas; or analyze volumes of drugs purchased in a geographic area when adjusted by the area population. In addition to the analysis noted above using ARCOS data, we also identified further analysis that DEA could perform using ARCOS data and additional available data to help identify potentially suspicious purchase or distribution patterns. Specifically, in our review and analysis of ARCOS data and information about PDMP data, we identified an opportunity for DEA to analyze ARCOS and PDMP data together for a more complete picture of drug transactions from distribution to retailers through dispensation to patients. We determined this could help in assessing whether the amount of drugs being prescribed is consistent with the amount of drugs being purchased or distributed in a given geographic area. For example, in areas where the number of prescriptions increases, a subsequent increase in drug orders and distribution to that area could be considered understandable. However, where the number of prescriptions in an area remains the same, or decreases, a significant increase in drug orders and distribution to that area could be considered unusual, especially if this pattern persists over several reporting periods. DEA stated that it occasionally performs such analysis manually, noting however that its access to PDMP data is contingent upon each state’s requirements and willingness to share its PDMP data with federal law enforcement. In July 2019, DEA officials responsible for overseeing the use and analysis of ARCOS data expressed an interest in improving DEA’s ARCOS data analytic capabilities but stated that they needed more staff and resources. Specifically, they noted they would like to hire additional staff, such as data scientists, to conduct analysis on ARCOS data using, for example, additional computer algorithms. DEA also noted that it was considering automation of additional types of analyses, but did not provide a start date or estimate as to when it would move forward on that consideration. While DEA created the new Targeting and Special Projects Section in March 2019 to enhance DEA’s data analytics and set aside some positions for program analysts and subject matter experts, among other positions, as of October 2019, DEA officials did not have any details or documentation about the data analysis efforts the new division plans to undertake. We have previously reported that new approaches to combining and “making sense of” large amounts of varied data—methods referred to as advanced analytics—are helpful to uncover patterns, identify anomalies, and provide insights not suggested by assumed hypotheses. In addition, other federal entities responsible for detecting diversion and abuse of controlled substances utilize computer algorithms as part of their analysis of available data in order to flag and prioritize potential instances of diversion for further investigation. For example, the Centers for Medicare & Medicaid Services and its National Benefit Integrity Medicare Drug Integrity contractor use proactive data analysis to detect aberrant patterns and potential diversion in drug prescribing. As a result, the contractor is able to produce “prescriber risk assessments,” which provide a comparison of controlled substance prescribing patterns across peers. The Centers for Medicare & Medicaid Services also uses proactive data analysis to identify providers with potentially inappropriate prescribing patterns, especially as it concerns opioids. Similarly, some opioid drug distributors use computer algorithms to identify suspicious orders that are the basis for the suspicious order reports they are required to provide to DEA. The establishment of this new section within DEA focused on its data analytics capabilities presents an opportunity for DEA to more proactively use data analytics with regard to its ARCOS and other data. In doing so, DEA could more effectively identify possible diversion activities or unusual activity to aid its ongoing efforts to prevent, detect, and investigate diversion more quickly and assist it in reporting on how it is using ARCOS data to identify suspicious activities. DEA Recently Developed a Centralized Database to Collect Suspicious Opioid Order Data but Lacks an Overall Structure to Manage all of its Data DEA Recently Created the Required Centralized Database for Suspicious Order Reports In October 2019, DEA established the Suspicious Orders Report System (SORS) Online, a centralized database for collecting suspicious order reports, which is required by the SUPPORT Act to be established by October 24, 2019. The SORS Online data fields include a requirement for registrants to note their reasons for identifying an order as suspicious, drug quantity, and dosage strength. The successful implementation of the centralized database is important because it could address the fragmented way in which suspicious order reports are currently submitted. However, reporting to the centralized database is currently voluntary. Registrants may notify DEA of a suspicious order using other means, including email, facsimile, or telephone. The systems and reports are not currently integrated, and investigators must query each system or office separately in order to find, for example, information related to a lead they are investigating. Currently, registrants are required upon discovery of a suspicious order or series of orders, to notify the Administrator of the DEA and the Special Agent in Charge of the division office of the DEA for the area in which the registrant is located or conducts business. Prior to DEA establishing the SORS Online centralized database, registrants with an existing or a prior MOA also have reported suspicious orders into one of two SORS databases when reporting to headquarters. The new SORS Online is the only electronic mechanism for reporting suspicious orders now, according to DEA. Registrants who are under active MOAs with DEA are reporting to the new SORS Online system, according to DEA. Registrants that are not under an MOA may also use SORS Online, but are not required to do so. Registrants not under an MOA may also use a paper-based process, among others, when reporting to the field division offices and DEA headquarters. However, no integration exists across headquarters’ and field division offices’ various electronic- and paper-based systems. DEA officials we met with said that some of the suspicious order reports received at the field division office level are stored in hard copy in accordion file folders, instead of being digitized or entered into a searchable database. Reporting to SORS Online satisfies the requirement to report such orders to the Administrator of the DEA and the Special Agent in Charge of the Division Office of the DEA for the area in which the registrant is located or conducts business. Successfully managing the SORS Online database could lead to needed efficiency improvements and more effective use of the suspicious order report data. DEA Lacks a Data Governance Structure to Manage its Data on Opioid Orders Although DEA has guidance, policies and procedures regarding the use of some of its information systems, it has not established a formalized data governance structure to manage its collection and use of data used to support the Diversion Control Division’s mission. DEA specifically has not institutionalized and clearly documented policies and procedures that describe division staff’s roles and responsibilities for collecting and analyzing data nor has it provided a structure that describes the agency’s approach to establishing and maintaining such a program. We have identified a number of issues with DEA’s management of data. For example, DEA does not have any documentation on their process for ensuring the quality of data registrants submit to its ARCOS database— the main system that enables DEA to monitor the flow of controlled substances. As a result, it is difficult to understand the controls they have over this important data. A data governance structure is defined as an institutionalized set of policies and procedures for providing data governance throughout the life cycle of developing and implementing data standards. A data governance structure also helps to ensure important data assets are formally managed and fully utilized, and can also provide consistent data management. We previously reported on key practices based on several data governance models, including developing and approving data standards, managing, controlling, monitoring, and enforcing consistent application of data standards, and delineating roles and responsibilities for decision making and accountability. Additionally, in June 2019, the Office of Management and Budget established a Federal Data Strategy (Strategy) as a framework of operational principles and practices to help agencies use and manage data. We found several areas where DEA’s current practices do not reflect select leading data governance practices. Agencies should identify data needs to answer key agency questions: We found that DEA does not have a governance structure to determine and prioritize its data requirements for either suspicious order reports it receives or data reported into its ARCOS systems. For example, DEA has not established standard requirements for the information required in a suspicious order report. As a result, distributors’ suspicious order reports vary and may contain inconsistent and insufficient data for DEA to make investigative decisions. In addition, DEA does not have a governance structure to identify agency and industry stakeholder data needs to help inform its opioid diversion control efforts. Agencies should provide resources explicitly to leverage data assets: Agencies should ensure that sufficient human and financial resources are available to support data driven agency decision- making, and accountability. As mentioned earlier, while DEA created the new Targeting and Special Projects Section in March 2019 to enhance DEA’s data analytics, as of October 2019, DEA officials did not have details or documentation about the data analysis efforts the new division plans to undertake or the resources they plan to provide for those efforts. As a result, DEA is unable to conduct the analysis that would enable it to more effectively use its existing data in making decisions about diversion related efforts. Agencies should prioritize data governance: Agencies should ensure there are sufficient authorities, roles, organizational structures, policies, and resources in place to transparently support the management, maintenance, and use of strategic data assets. Similarly, leading practices for data governance includes delineating roles and responsibilities for decision-making and accountability, including roles and responsibilities for stakeholder input on key decisions. As mentioned earlier, DEA established the Targeting and Special Projects Section in March 2019 whose goal is to focus on leveraging DEA’s data capabilities and conducting data analytics on ARCOS and other data, according to Diversion Control Division officials. While the new section appears to hold promise, DEA has not clearly defined and adopted the new section’s roles and responsibilities for managing and analyzing data across the DEA or how the new section will communicate and collaborate with other Diversion Control Division headquarters and field staff. As a result, the new division may not operate in a predictable, repeatable, and accountable way. Agencies should support non-federal stakeholders: Agencies should engage with industry, academic, and other nonfederal users of data to share expert knowledge of data assets, promote wider use, improve usability and quality, and advance innovation and commercialization. Later in this report, we identify an opportunity for DEA to collaborate with industry stakeholders and seek their input for an initiative that is supposed to assist industry stakeholders in their responsibilities to report suspicious orders to DEA. Although DEA has not incorporated these data governance practices, it is in the early stages of developing a data governance structure. As of September 2019, DEA officials told us that its Office of Information Systems’ Chief Data Officer just recently started to work with DOJ and other components to develop a data strategy in response to the recently released department wide strategy, and therefore does not have any additional documentation or information related to timelines and deliverables for formally implementing a DEA data governance or other data structure for the agency. Without additional details, such as a timeframe for developing the structure or more information about what it would entail, it is unclear how or if these efforts will incorporate leading practices for data governance and if they will be effective. Data governance processes are important for DEA given it works with an extensive and complex network of stakeholders to manage opioid diversion risks and uses industry-reported data to help it identify patterns that might indicate potential diversion. An effective data governance structure could help DEA ensure its important data assets are formally managed and fully utilized, and can also help ensure consistent data management. Industry and technology councils, domestic and international standards-setting organizations, and entities within the federal government endorse the establishment and use of a governance structure to oversee the development, management and implementation of data standards, digital content and other data assets. While DEA’s Systems Do Not Provide Complete, Real-Time Data on Suspicious Orders, Most Industry Stakeholders Said Adding Such Capabilities Would Not Provide Extensive Value DEA does not have an existing mechanism or a comprehensive database of orders before they are filled that it can analyze, on a real-time basis, to identify potentially suspicious orders. However, most industry stakeholders we spoke with on the usefulness of real-time data noted that such a mechanism would not add extensive value to diversion detection. DEA’s current data systems either contain historical, not real-time, data or do not contain all drug order data that could be reported. ARCOS. The data in the ARCOS database is historical, rather than real- time, on orders that have been filled. Every registered manufacturer is required, at such time or times and in such form as required by the Attorney General, to make periodic reports to the Attorney General of every sale, delivery or other disposal of any controlled substance. Each distributor is required to make such reports with respect to narcotic controlled substances. For example, as part of the reporting to ARCOS, acquisition and distribution transaction reports are required, by regulation, to be filed every quarter, except that a registrant may be given permission to file more frequently, but not more frequently than monthly, depending on the number of transactions being reported each time by that registrant. In addition, manufacturing transaction reports are required to be filed annually, except that a registrant may be given permission to file more frequently, but not more frequently than quarterly. CSOS. DEA does not require registrants to use CSOS and thus it is not used by all registrants. As previously discussed, CSOS is an electronic ordering system which allows registrants to place orders for controlled substances. Shipments of all ARCOS-reportable controlled substances, ordered through CSOS, are included in registrant’s periodic ARCOS reporting. Suspicious Order Reports. Suspicious order reports are intended to identify orders of unusual size, orders deviating substantially from a normal pattern, and orders of unusual frequency before they are filled. By law, each registrant is required to design and operate a system to identify suspicious orders that it receives. Registrants identify, then report, to DEA using their own systems to determine suspicious orders. As discussed previously in this report, registrants can report these orders into DEA’s newly launched SORS Online database, but reporting into this database is voluntary and registrants have an option to report in other ways, so this database does not capture all suspicious orders and is therefore not comprehensive. Suspicious orders are likely identified in close to a “real-time” basis. Orders that have been identified and reported as suspicious by the registrants, are orders that have not yet been filled. While two individual drug distribution companies we interviewed said they saw some value in real-time reporting, most industry stakeholders we spoke with on the usefulness of real-time data, including a broad cross- section of associations representing pharmacies and drug distribution companies, said that real time dissemination of suspicious orders by DEA would not add extensive value to efforts to detect possible diversion. Instead, some industry stakeholders suggested that a focus on data that provide trends over time might be more useful. As discussed earlier in this report, we provide examples of data analysis DEA currently performs and could perform on its existing data that could potentially help DEA determine or identify possible patterns of aberrant behavior in drug order information. Others we spoke with raised concerns about the varying ways companies determine what is suspicious and that using real-time data reported from DEA on these orders could be like comparing apples to oranges. Most of the associations that represent pharmacies and drug distributors that we met with indicated they did not see much value in either reporting, or receiving reports, of suspicious orders in “real-time.” For example, a representative from an association representing pharmacists told us that rarely would there be a case where a single order was so egregious that stopping it would have a significant impact on public health. This representative also noted that it would be more important to focus on historical trends, given “trends don’t happen in a day.” Other stakeholders we spoke with said that while there may be utility in real-time reporting of suspicious orders, they also had concerns about its feasibility, given current available data. They noted it would be difficult to compare suspicious order data as reported by registrants because companies rely on their own methods to determine a suspicious order. For example, Officials from an association that represents a large number of drug distributors indicated that receiving more real-time data might allow their members to have an additional check on orders that a wholesale distributor receives, but this utility would largely be contingent on the distributors’ ability to compare suspicious order reports across one another. Distributors use different criteria for determining whether an order is suspicious; there is no continuity across them; and they experience varying order volumes and patterns across their customers and over time as patient needs change. Thus, such analyses would be difficult to conduct, if they could be done at all, and would not necessarily result in useful comparisons. A representative from one drug distribution company told us that having knowledge of other distribution companies’ suspicious orders is not helpful because the company would not know how the other distributor made a determination on the suspicious order. Another representative stated that distributors are operating proprietary systems that may or may not vary substantially from each other depending on a large number of varying circumstances, and may be operating “wildly different” systems for identifying suspicious orders and therefore the information would not be valuable. Representatives from two drug distribution companies identified additional challenges to real-time reporting of suspicious orders if the determination of whether an order was suspicious or not was made by DEA. First, they did not believe DEA had sufficient resources or knowledge to identify suspicious orders. One representative said DEA does not know the history and market dynamics in the pharmaceutical industry to help inform decisions it would need to make on an order. Second, identifying an order as suspicious before it is filled would add a tremendous burden on DEA. According to one of the representatives, their company typically ships orders on the same day the order is received, consistent with “just-in-time” inventory management practices. If DEA were expected to make suspicious order determinations without the risk of disrupting patient care needs, it would be imperative for DEA to act quickly to identify suspicious orders. These distribution companies did not believe DEA would be able to identify them rapidly as needed. As noted above, DEA’s current systems are not designed for real-time reporting, and it does not have an existing mechanism or a comprehensive and complete database of orders before they are filled that it can analyze, on a real- time basis, to identify potentially suspicious orders. Officials from the association that represents a large number of drug distributors were careful to point out, however, that systems sometimes differ intentionally due, for example, to varying customer bases, service requirements and patient care needs. Thus, a certain amount of variability in suspicious order systems, criteria and decisions may be warranted, and even desirable. DEA Does Not Have Outcome-Oriented Goals and Performance Measures for its Opioid Diversion Activities While DEA has developed some performance measures to track and publicly report the progress and results of its efforts in reducing diversion, DEA has not developed objectives, outcome-oriented goals, or measurable performance targets to assess the effectiveness of its opioid diversion control data analysis efforts and the link between DEA’s use of data and progress toward its diversion goals and strategies. DEA does have performance measures including the number of civil penalties and administrative actions it has undertaken, planned or scheduled investigations completed, and community outreach events completed. While these measures are useful, they do not account for outcomes of these actions, such as their potential impact on the volume of opioids being improperly sold or purchased. DEA officials noted that it adheres to goals established through the Office of National Drug Control Policy’s National Drug Control Strategy, such as reducing the prescription opioid rate by one-third within three years, reducing overdose deaths, and within five years, ensuring all health care providers have adopted best practices for opioid prescribing. However, those goals involve a multitude of federal agencies, and are not directly related to DEA’s use of industry-reported data, nor linked specifically to DEA diversion control efforts. DEA also noted that they have a number of goals across strategies such as DEA’s 360 strategy in addition to the goals in DOJ’s strategic plan; a performance measure with a measurable target for its agency-wide objective related to dismantling drug trafficking organizations—maximizing the monetary value of currency, property, and drugs seized; and a measure for curbing opioid and other illicit drug use. GPRAMA directs agencies to develop and document goals, as well as performance measures to assess progress towards their goals. Agencies can use performance measurement to make various types of management decisions to improve programs and results, such as developing strategies and allocating resources, including identifying problems and taking corrective action when appropriate. Additionally, GPRA as amended by GPRAMA states that management should define outcome-oriented objectives in specific and measurable terms. Measurable targets help decision makers conduct assessments of whether program goals were achieved, and linkages between an organization’s goals and performance measures create a line of sight so that everyone understands how program activities contribute to the organization’s goals. DEA officials view their existing performance goals as sufficient overall. However, without defining objectives in specific measurable terms, DEA is likely not able to adequately assess whether its respective investments and efforts are helping it to limit the availability of and better respond to the opioid prescription diversion threat. Until program officials can review the effectiveness of these systems based on quantifiable benefits and measurable performance targets, they are not well-positioned to determine the extent to which suspicious order reports or ARCOS data and systems are enhancing the effectiveness of the agency’s opioid related regulatory and criminal diversion investigations, prosecutions and civil actions. Documenting program goals and developing measurable performance targets and linkage to program goals could provide DEA with the information it needs to assess progress and make informed decisions about current and future operations. DEA Has Taken Some Steps to Help Industry Report Suspicious Orders, but Has Not Addressed Identified Limitations with the Data it Shares or Receives DEA Has Developed a Tool to Share Some Drug Purchase Data with Industry, but the Tool Has Limitations DEA developed an ARCOS query option for registrants to use, called the ARCOS Enhanced Lookup Buyer Statistic Tool, in February 2019 to better support registrants’ efforts to identify and report suspicious orders. This tool allows registrants to query certain ARCOS data maintained by DEA. Although this tool was supposed to be an improvement upon a prior iteration of the lookup tool DEA had developed, distributors and an industry association representing distributors identified several limitations with the tool. Specifically: Single query challenges and no bulk downloads. The distributor can only query the tool one pharmacy at a time, even though some distributors supply thousands of pharmacies on a daily basis. Thus, if certain distributors were to query all of its pharmacies for possible suspicious order patterns, the process could be time-consuming or not feasible. DEA noted it was working on this limitation. Limited login credentials. DEA only provides each distributor with one set of login credentials, so only one employee can log in at a time to query the tool. DEA noted it was working on this limitation. Data provided in the tool are not detailed enough. The tool does not provide detailed enough information to be useful to facilitate the identification of suspicious orders. For example, when a distributor queries the tool, the search results will list the total dosage units for a particular opioid for the past six months at the pharmacy. Because some opioid drug dosages are more commonly abused than others, distributors told us that simply having the total number of dosage units is not as helpful as seeing the breakdown of the different dosage units. In another instance, the data provided to distributors does not include critical details about the number of suppliers. One distributor might have multiple warehouses and distribution centers that it uses to package and ship pharmaceutical products. In the ARCOS data that DEA provides to distributors, these individual warehouses are counted as distinct suppliers in the total supplier count data provided to the distributors. Therefore, the number of suppliers may appear inflated to the distributors, even though it is only a single company providing the products. According to DEA, the ARCOS lookup tool is meant to be a pointer and assist distributors in conducting due diligence so they can “know their customer.” Regardless if a distributor is shipping from multiple distribution centers and therefore showing as multiple suppliers in the lookup tool, these are all unique DEA registration numbers and are therefore unique suppliers to the customer. According to DEA, the important part here is that distributors can see quantity and gram totals per registrant (such as, a pharmacy customer) that they query. When evaluating whether an order is suspicious, a distributor uses its own internal transaction data to evaluate a buyer’s ordering patterns. However, purchasers of controlled substances, such as pharmacies and medical practices, may use multiple distributors for their purchases. Distributors have previously raised concerns that they did not have access to additional transaction data, such as whether the purchaser is also buying controlled substances from additional suppliers. They have noted that this additional data would be useful when making decisions about whether an order is suspicious, and specifically, that ARCOS data would be useful in helping them evaluate whether an order was suspicious. For example, in 2018, one distributor testified that, given DEA’s access to the controlled substance transaction data that distributors report, “nly DEA has visibility over the entire landscape and can track and analyze aggregate data on the distribution of controlled substances in particular jurisdictions.” In addition, an industry organization we met with provided comments to DOJ in 2017 that certain data could provide more context for them to identify problematic orders. Specifically, the organization noted that if DEA could provide ARCOS data in aggregate form without identifying individual distributors’ competitors, the distributor could consider a pharmacy’s orders in the context of the pharmacy’s overall ordering from all distributors. An industry association representing distributors, and two of the distributors that we interviewed stated that the Enhanced Lookup Buyer Statistic Tool is a step in the right direction. However, the industry association and four distributors that we interviewed stated that the tool remains limited in helping distributors improve how they identify suspicious orders, as noted above. DEA officials told us that distributors have brought some of these concerns about the ARCOS Enhanced Lookup Buyer Statistic Tool’s usability to their attention. For example, in May 2019, an industry association representing distributors sent a letter to DEA, outlining a consolidated list of industry’s concerns about the tool. In recent discussions in June 2019, DEA officials acknowledged some of these limitations and stated that some industry concerns would be easier to fix than others, but that they had not established a timeframe for when the changes would be implemented. For example, DEA officials noted it might be easier to provide additional login credentials to distributors and make the data available to be downloaded in a more functional way for distributors. For some of the other limitations industry stakeholders identified, such as providing more detailed ARCOS data to the distributors, DEA officials raised concerns. For example, DEA officials noted that distributors could use the additional detailed data as a market research tool in order for distributors to gain unfair market advantages or to learn more about their competitor’s business contracts with pharmacies. In September 2019, DEA officials told us that it was not currently addressing changes to the ARCOS Enhanced Lookup Buyer Statistic Tool, due to competing priorities within DEA. Specifically, DEA officials noted that it is focused on existing priorities related to meeting upcoming requirements mandated in the SUPPORT Act, including establishing a suspicious order centralized database, as discussed previously in this report. While we recognize that agencies need to determine and set priorities, it is important for DEA to continue to work with industry in ensuring that the tool it created to address the SUPPORT Act requirement will help industry in addressing its suspicious order reporting requirement under the CSA, as amended. The SUPPORT Act requires DEA to provide distributors with access to ARCOS data to help the distributors identify, report, and stop suspicious orders of opioids and reduce diversion rates. By identifying solutions – in consultation with industry stakeholders – to address the limitations of the ARCOS Enhanced Lookup Buyer Statistic Tool, DEA could better ensure registrants have more useful information at their disposal when evaluating whether an order is suspicious. DEA Identified Extensive Limitations with Suspicious Order Reports it Received DEA officials and DEA field division offices we interviewed identified a number of limitations with suspicious order reports they received, and, due to these limitations, they rarely use suspicious order reports to generate potential investigative leads. The issues DEA identified included: Threshold-based algorithm triggers. Several DEA headquarters and field division officials told us that some distributors used fixed thresholds to identify suspicious orders, which DEA officials stated are not helpful or useful because the information is often not actionable. Lack of documented rationale. During the course of our review, DEA officials told us that many suspicious order reports do not include the rationale for why the registrant decided the order was suspicious, making it difficult to determine which suspicious order reports might contain actionable intelligence. In September 2019, the DOJ Office of the Inspector General reported that the current regulatory language governing industry suspicious order reporting does not require manufacturers and distributors to state why they believe an order is suspicious. In October 2019, DEA launched a new centralized database of suspicious order reports, as required by the SUPPORT Act. DEA’s new reporting format of suspicious order reports includes a required field for “Reason,” for registrants to provide an explanation of why the order is suspicious. However, currently reporting to the new centralized database is voluntary. Differing methodologies. As discussed earlier in this report, the definition of a “suspicious order” may include, but is not limited to, an order of a controlled substance of unusual size, an order of a controlled substance deviating substantially from a normal pattern, and orders of controlled substances of unusual frequency. However, it is up to the individual distributor companies to decide the more specific metrics, according to DEA Diversion Control Division officials. Each distributor must design and operate a system to identify suspicious orders. Therefore, distributors utilize different methods to flag customer orders as suspicious. According to our analysis of DEA data, it has collected at least 1.5 million suspicious order reports since 2014, and these reports may contain data on attempted purchases that were denied, based on indicators of suspicious patterns. These data could help with DEA’s efforts to prevent, detect, and investigate diversion. Officials from six DEA field division offices we interviewed said they refer to suspicious order reports when conducting their routine regulatory investigations of registrants. DEA field division officials also stated that, while suspicious order reports are generally not used as the primary or sole impetus to initiate an investigation, officials will infrequently refer to related suspicious order reports when there is an ongoing criminal investigation that is initiated through other means. However, of the DEA field division offices we interviewed, officials from two offices told us that they had used a suspicious order report as the sole or primary impetus for initiating a criminal investigation in the past year – one stating that it happened once, and another estimating that it happened one to three times. Another field division told us that they had two convictions “in recent memory” that began with a suspicious order report. Three offices told us that they had not used suspicious order reports as the sole or primary impetus for a criminal investigation in the past year, and one told us they did not know if a suspicious order report had been used in that way. DEA field division offices we interviewed also identified reasons why suspicious order reports may not be as useful as they could be in helping to identify investigative leads. For example, one DEA field division office characterized the suspicious order reports they received from one particular registrant as being “spot on” and always warranting a DEA follow-up investigation, given the amount of detail and evidence of the registrant conducting its own on-site investigation into the customer. However, the same DEA office reported that other suspicious order reports were based on industry-developed thresholds that were not useful because the resultant reports did not indicate why the order was suspicious. Of the five DEA field division offices that we asked to characterize the quality of suspicious order reports, three of them reported that suspicious order reports were either “moderately” or “somewhat” useful. Officials from one field division office said that suspicious order reports are “very useful,” while officials from another DEA field division office reported that suspicious order reports are “not at all useful.” We have previously reported on these issues, including DEA communication with registrants, and in June 2015, we found that additional guidance from and additional communication with DEA was needed about registrants’ roles and responsibilities under the CSA, as amended. We recommended that DEA develop additional guidance for distributors for suspicious order monitoring and reporting. DEA did not expressly agree or disagree with our recommendation, but raised concerns about the recommendation, stating that “short of providing arbitrary thresholds to distributors, it cannot provide more specific suspicious orders guidance because the variables that indicate a suspicious order differ among distributors and their customers.” In responding to this recommendation, DEA officials told us that the agency had refocused its efforts on revising draft regulations in line with the SUPPORT Act, and that the revised draft was undergoing internal DEA and DOJ review. The agency noted that it expected the rule to codify existing legal obligations related to due diligence and suspicious order reporting and provide additional guidance regarding the nature and timing of the suspicious order reporting requirement, but also indicated that it was not possible to be certain of the precise nature of the draft rule. The 2015 recommendation remains relevant and important, and while DEA has reported taking some actions to address it, as noted above, DEA has not taken all the necessary steps to address the recommendation. We will continue to monitor DEA’s progress in addressing our recommendation. Conclusions Given the extensive and complex network of stakeholders DEA works with to manage opioid diversion risks and the agency’s use of a large amount of industry-reported data, DEA could do more to use proactive, automated computer algorithms to analyze its data sources in detecting questionable patterns in industry-reported drug transaction data. It is missing opportunities to more effectively identify questionable ordering patterns and possible diversion activities than through its current analysis methods. Using more automated analyses, similar to other federal entities that use computer algorithms as part of their analysis of available data to help flag instances of diversion, DEA could enhance its ongoing efforts to prevent, detect, and investigate diversion more quickly and assist it in reporting on how it is using ARCOS data to identify suspicious activities. Furthermore, because DEA does not have a documented data governance structure in place to manage its data, it risks challenges related to quality, availability, and integrity of the data it uses to support opioid diversion. Although DEA has started to explore developing a data governance structure, it is important for DEA to document and define its process about what the structure would entail. This would help the agency determine the effectiveness of its structure, an important consideration given the large amounts of varied data DEA receives from industry stakeholders. Also, while DEA does have some performance goals related to opioid diversion, it lacks outcome-oriented goals and measurable performance targets to assess the extent to which the industry-reported data it obtains and uses support the agency’s diversion control activities. Defining these targets could help DEA adequately assess whether its respective investments and efforts are helping it to limit the availability of and better respond to the opioid prescription diversion threat. DEA’s efforts to provide registrants with additional information to facilitate the identification of suspicious orders is promising, but has limitations. Due to these limitations, registrants, such as distributors, might not have complete information when they are identifying suspicious orders. By identifying solutions – in consultation with industry stakeholders – to address the limitations of the ARCOS Enhanced Lookup Buyer Statistic Tool, such as the need for additional login credentials or the ability to bulk download data, DEA could better ensure registrants have more useful information at their disposal when evaluating whether an order is suspicious. Finally, we continue to monitor implementation of our 2015 recommendation that DEA provide additional guidance to distributors related to suspicious orders, and we believe that it remains relevant and important. Recommendations for Executive Action We are making the following four recommendations: The DEA Administrator should develop and implement additional ways to use algorithms in analyzing ARCOS and other data to more proactively identify problematic drug transaction patterns. (Recommendation 1) The DEA Administrator, in coordination with the department-wide efforts on data strategy, should establish and document a data governance structure to ensure DEA is maximizing its management of industry-reported drug transaction data. (Recommendation 2) The DEA Administrator should establish outcome-oriented goals and associated measurable performance targets related to opioid diversion activities, using data it collects, to assess how the data it obtains and uses supports its diversion control activities. (Recommendation 3) The DEA Administrator, in consultation with industry stakeholders, should identify solutions to address the limitations of the ARCOS Enhanced Lookup Buyer Statistic Tool, to ensure registrants have the most useful information possible to assist them in identifying and reporting suspicious orders to DEA. (Recommendation 4) Agency Comments and Our Evaluation We provided a draft of this report to DOJ, including DEA, for review and comment. In its comments, reproduced in appendix III, DEA agreed with three of the four recommendations, and neither agreed or disagreed with the fourth. DEA also provided technical comments, which we incorporated as appropriate. In response to our first recommendation that DEA should develop and implement additional ways to use algorithms in analyzing ARCOS and other data to more proactively identify problematic drug transaction patterns, DEA concurred and stated it will continue to examine a variety of technologies to analyze ARCOS and other data and implement additional ways to use algorithms to more proactively identify problematic drug transaction patterns. If these and other actions to expand the agency’s analytic capabilities are effectively implemented, DEA would address the intent of our recommendation. DEA also concurred with our second recommendation that DEA, in coordination with the department-wide efforts on data strategy, should establish and document a data governance structure to ensure DEA is maximizing its management of industry-reported drug transaction data. In its response, DEA stated it is currently implementing this recommendation and will continue to mature its data governance structure. The intent of this recommendation is for DEA to establish a formalized data governance structure to manage its collection and use of data used to support the Diversion Control Division’s mission. By establishing such a structure, DEA could better ensure its important data assets are formally managed and fully utilized, and could also help ensure consistent data management across the Diversion Control Division. DEA neither agreed nor disagreed with our third recommendation that DEA should establish outcome-oriented goals and associated measurable performance targets related to opioid diversion activities, using data it collects, to assess how the data it obtains and uses supports its diversion control activities. In its response, DEA stated it recognizes that measurable performance targets related to opioid diversion activities can serve as leading practices at different organizational levels including the program, project, or activity level. However, DEA stated it needs additional clarification on the specific actions needed to fulfill this recommendation. Our recommendation is intended to ensure that DEA can demonstrate the usefulness of the data it collects and uses to support its opioid diversion control activities. We will continue to work with DEA to address the specific actions needed to assess how the data it obtains and uses support its diversion control activities to fully address the intent of this recommendation. Based on our review of DEA’s existing performance goals and targets for its opioid diversion efforts, as well as our previous work on performance measurement, we believe that further development of related performance goals and targets is warranted and could potentially improve the usefulness of the data DEA collects and uses in support of its diversion control program. DEA also stated in its comments that the limited timeframe did not allow GAO to meet with DEA officials responsible for performance metrics for opioid diversion. However, in our interviews with DEA regarding its performance metrics for opioid diversion, we submitted our questions in advance of meeting with DEA officials to allow time for the questions to be reviewed by relevant officials. DEA stated in its comments that it will ensure that GAO meets with the appropriate officials to address metrics. As stated earlier, we will continue to work with DEA to address the specific actions needed to meet the intent our recommendation. DEA concurred with our fourth recommendation that DEA, in consultation with industry stakeholders, should identify solutions to address the limitations of the ARCOS Enhanced Lookup Buyer Statistic Tool, to ensure registrants have the most useful information possible to assist them in identifying and reporting suspicious orders to DEA. DEA stated it has consulted with industry stakeholders and has identified solutions to address the limitations of the tool. We believe such consultation will be beneficial for DEA to understand its industry stakeholders’ needs and that identifying solutions for addressing these needs would help ensure registrants have the information necessary to help identify and report suspicious opioid orders. We are sending copies of this report to the appropriate congressional committees, the Attorney General, the DEA Administrator and the Secretary of Health and Human Services, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-6691 or McneilT@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Scope and Methodology To understand the extent to which DEA obtains and uses industry- reported data, and the opportunities that exist to improve how that data are obtained and used, including the feasibility of real-time reporting, we reviewed applicable laws, regulations, court cases, and DEA internal documentation. We also conducted interviews with DEA headquarters offices, including the Diversion Control Division and DEA field division offices. To determine DEA registrant legal reporting requirements related to prescription drug orders and the meaning of suspicious orders, we reviewed applicable laws and regulations, including the CSA and its subsequent amendments and related DEA regulations and guidance. In addition, we reviewed the recently enacted SUPPORT Act. To identify policies and guidelines DEA uses to obtain and review registrant-reported data, we reviewed DEA procedures for conducting drug-related investigations, information system manuals for data and information systems used by DEA, and DEA written communications to registrants and DEA forms registrants use to report prescription drug transactions to DEA. As part of our work examining the information systems used to obtain and analyze data reported by registrants, we interviewed officials who oversee the management of DEA information systems, such as Automation of Reports and Consolidated Orders System (ARCOS), Controlled Substances Ordering System (CSOS), Registrant Information Consolidated System, and the Suspicious Orders Reporting System (SORS) systems used to obtain and store suspicious order reports at DEA headquarters. We interviewed DEA officials in headquarters and field division offices to determine how information that industry members report to DEA is obtained and used to detect and identify potential diversion activities. The perspectives we gathered from field division offices cannot be generalizable to the entire population of field division offices, but did provide us with insights into the agency’s diversion efforts and use of industry-reported data. To identify what opportunities exist, if any, for DEA to improve these efforts, such as using computer algorithms or real-time reporting, we also interviewed DEA officials responsible for developing analytical products based on industry-reported data. In addition, we interviewed DEA officials at eight field division offices to learn about how diversion investigators use industry-reported data and what, if any, improvements might be needed. To identify which of the 23 DEA field division offices to interview, we prioritized our selection based on four primary criteria: 1) the controlled prescription drug availability rate in their geographic area, according to a 2017 DEA threat assessment report, indicating whether the field division office had a “high” or “moderate” rate of availability; 2) whether the office was within the location of a DOJ Opioid Fraud and Detection Unit task force location; 3) whether the office was located within top ten state with high controlled prescription drug prescribing rate, as identified by the CDC; and 4) whether the office was located within a state that the CDC identified as having a high ER visit rate for opioid overdoses. We also ensured that the DEA field division offices we interviewed represented different geographic areas within the United States. We also conducted interviews with four pharmaceutical distributors and one trade organization whose membership includes wholesale distributors. We interviewed three organizations representing pharmacies, pharmacists, and drug diversion professionals to gather their perspectives and experiences with efforts to detect and report suspicious opioid orders. We based our initial interview selection of distributors based on DEA- provided ARCOS data of opioid-related transactions, which indicated the three largest distributors for opioids. To identify smaller distributors to gather their perspectives, we contacted an industry association representing distributors to facilitate our efforts to arrange for an interview, resulting in an interview with one additional distributor. In addition, we interviewed officials from a state prescription drug monitoring program (PDMP) that collects real-time data, a Bureau of Justice Assistance grant program that supports PDMPs, and a company that operates 44 of the 53 state PDMPs to gain insights on the data they collect. The views of these organizations cannot be generalized to the entire population, but provided important insights and perspectives about suspicious order detection and reporting. We reviewed the data DEA collects to identify possible types of analyses DEA could conduct using ARCOS data to identify unusual patterns. In addition, we reviewed key data governance practices used by organizations and identified through our past work to determine the extent to which DEA has a governance structure in place to manage how it collects and uses data to support diversion control efforts. Additionally, we reviewed the June 2019 Office of Management and Budget Federal Data Strategy which provides a framework of operational principles and practices to help agencies use and manage data. The key practices we identified to compare DEA’s data governance efforts against were: identify data needs to answer key agency questions; provide resources explicitly to leverage assets; prioritize data governance; and support non-federal stakeholders. We selected these practices because they are important to early development of a data governance structure. We also reviewed the February 2019 Data Strategy, released by DOJ, that is to serve as a roadmap for DOJ components to manage their data assets. To understand the extent to which DEA assesses the results of the data it obtains and uses from its ARCOS system and through suspicious order reporting, we reviewed DEA’s performance measures and applicable laws governing performance reporting in the federal government, including the Government Performance and Results Act of 1993 (GPRA), as updated and expanded by the GPRA Modernization Act of 2010 (GPRAMA). Although GPRA and GPRAMA requirements apply to those goals reported by departments (e.g., DOJ), we have previously reported that they can serve as leading practices at other organizational levels, such as component agencies for performance management. We also reviewed related national, DOJ, and DEA strategy documents that are used to communicate diversion control goals and performance. These documents included the 2018 National Drug Threat Assessment and 2019 National Drug Control Strategy, DOJ’s department-wide strategic plan, DOJ Annual Performance Report, DEA’s 360 strategy guide, and DEA congressional budget justification documents. In addition, we evaluated DEA’s performance measures against criteria in Standards for Internal Control in the Federal Government. Furthermore, we reviewed the extent to which DEA defined objectives and outcome-oriented goals, or established measurable performance targets to evaluate the effectiveness of how it obtains and uses data and compared them to GPRAMA requirements, which may serve as leading practices for DEA. To determine what opportunities exist, if any, for DEA to improve its use and collection of industry-reported data, such as using computer algorithms or real-time reporting, we interviewed DEA officials to determine what analytics, if any, DEA is using to detect and identify potential opioid diversion activities. In our interviews with field division offices, we requested information regarding how investigators received suspicious order reports from registrants and how the investigators requested and used ARCOS and other system analysis to conduct or support their investigative work. We also interviewed officials from other entities with opioid diversion prevention responsibilities, such as state level Prescription Drug Monitoring Programs, the Department of Health and Human Services, including the Centers for Medicare and Medicaid Services, and the Department of Justice (DOJ) U.S. Attorney’s Office Opioid Fraud and Abuse Detection Unit. To obtain perspectives of industry stakeholders on how data, such as suspicious orders may be better reported to DEA, we interviewed four industry associations whose memberships include industry stakeholders. We selected these associations based on their roles in representing various DEA registrant communities, such as pharmacists, pharmacies, and distributors. We also reviewed documentation describing the data available to DEA via its ARCOS database, as well as documentation that described examples of unusual patterns of orders. Based on such information, two GAO specialists identified methods that could be implemented using computer algorithms to analyze ARCOS data to identify patterns that might indicate unusual activity. Additionally, these specialists identified related opportunities that DEA could use to analyze ARCOS data combined with data from other sources, such as prescription rate information, to identify these patterns. To address the extent to which DEA collaborates with industry stakeholders to combat opioid diversion, we examined DEA policies and procedures, and interviewed relevant DEA officials, industry associations, and private sector industry members. Specifically, we examined DEA agency-wide directives and guidance, and component management policies and procedures for providing information to industry stakeholders related to industry’s suspicious order reporting requirements, including written communication DEA sent to industry stakeholders related to suspicious order reporting. In addition, DEA officials provided us with a demonstration of SORS, ARCOS, and the ARCOS Enhanced Lookup Buyer Statistic Tool – available to distributors to help them identify and report suspicious opioid orders. We interviewed DEA officials in eight field division offices who interact with industry stakeholders on, among other things, identifying and reporting suspicious orders. These officials provided their perspectives on the usefulness of suspicious order reports to their investigations as well as other industry self-reported data collected in DEA information systems. We interviewed opioid distributors of varying sizes, as noted above, including some of the largest opioid distributors, based on DEA-provided ARCOS data of opioid-related transactions, for their perspectives on the information and tools DEA provides to them, including the Lookup Buyer Statistics Tool and the ARCOS Enhanced Lookup Buyer Statistic Tool. The views of these distributors are not generalizable to the entire population, but provide insights and information on how industry detects and reports suspicious orders through use of ARCOS data and other tools. We conducted this performance audit from January 2019 through January 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Timeline of Selected Events and Legislation Impacting or Related to Industry- Reported Data on Prescription Opioids Appendix II: Timeline of Selected Events and Legislation Impacting or Related to Industry- Reported Data on Prescription Opioids Pub. L. No. 91-513, 84 Stat. 1242 (Title II of the Comprehensive Drug Abuse Prevention and Control Act of 1970, Pub. L. No. 91-513, 84 Stat. 1236). Appendix III: Comments from the Department of Justice Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact above, Tonnye’ Conner-White (Assistant Director), Gary M. Malavenda (Analyst in Charge), David Bruno, Jill Center, Billy Commons, Peter DelToro, Kathleen Drennan, Melissa Hargy, Will Horowitz, Hayden Huang, Eric Hauswirth, Susan Hsu, Nicole Jarvis, Benjamin T. Licht, Amanda Miller, and Jan Montgomery all made key contributions to this report.

Why GAO Did This Study Since 1999, more than 700,000 people have died of a drug overdose in the United States, with about 48,000 dying of an opioid overdose in 2017 alone. The DEA administers and enforces the Controlled Substances Act as it pertains to ensuring the availability of controlled substances, including certain prescription drugs, for legitimate use while limiting their availability for abuse and diversion. The Substance Use-Disorder Prevention that Promotes Opioid Recovery and Treatment for Patients and Communities Act, enacted in 2018 included a provision for GAO to study the reporting of suspicious opioid orders on a real-time basis nationally using computer algorithms. This report examines, among other things, how DEA obtains and uses industry-reported data to identify and address suspicious opioid orders and opportunities for DEA to improve these efforts, such as using computer algorithms or real-time reporting. GAO analyzed program documentation and DEA data, and interviewed DEA and industry officials as well as officials from national associations representing distributors, investigators, state boards of pharmacy, and other federal and state agencies. What GAO Found The Drug Enforcement Administration (DEA) collects industry-reported data on the sale and purchase of controlled substances and prescription drugs, including opioids. It uses these data to support ongoing investigations into the diversion of such substances into the illegal market place and to identify investigative leads for its field division offices. GAO identified deficiencies associated with DEA's drug diversion efforts, including the following: Limited proactive and robust analysis of industry-reported data. While DEA's current data systems are not designed to conduct real-time analysis, and it conducts some analyses of industry-reported data, such as in response to requests from its field division offices, DEA could conduct more analyses using automated computer algorithms to help identify questionable patterns in the data. For example, DEA could analyze data to identify unusual volumes of deleted transactions or unusual volumes of drugs that were disposed of rather than sold. It could also analyze data to identify trends in distribution or drug purchases in a given geographic area. Other analysis DEA could perform is to look for unusual patterns when comparing drug orders in one geographic area with other nearby areas. These analyses could potentially help DEA proactively identify suspicious activities or registrants that may warrant investigation. No data governance structure to manage all drug transaction data. Although DEA has guidance, policies and procedures for the use of some information systems, it has not established a formal data governance structure to manage all data it collects and maintains, which are integral to its diversion control activities. A data governance structure is defined as an institutionalized set of policies and procedures for providing data governance throughout the life cycle of developing and implementing data standards. Industry and technology councils, domestic and international standards-setting organizations, and federal entities endorse the use of a governance structure to oversee the development, management, and implementation of data standards, digital content, and other data assets. While DEA began efforts to develop a governance structure, it is in the early stages of development and does not have additional details or documentation of its efforts. An effective data governance structure could help DEA ensure its important data assets are consistently and fully utilized. What GAO Recommends GAO is making four recommendations related to DEA's collection and use of industry-reported data. DEA agreed with three of the four recommendations, and neither agreed nor disagreed with the fourth.

gao_GAO-20-206

gao_GAO-20-206_0

Background Aviation Maintenance Workforce Different aviation industry employers have distinct workforce needs and may require workers with specific skillsets depending on the type of work performed. The aviation maintenance workforce includes FAA-certificated mechanics and repairmen, as well as non-certificated workers. FAA-certificated mechanics inspect, service, and repair aircraft bodies (airframe) and engines (powerplant), and only they can approve an aircraft for return to service. FAA-certificated mechanics can earn an airframe rating, a powerplant rating, or an airframe and powerplant (A&P) rating. It can take between 1 and 3 years to obtain the required education or training to become certificated. If an FAA- certificated mechanic changes employers, the certificate remains valid. FAA-certificated repairmen service aircraft components and must be recommended for certification by their employer to perform specific tasks such as welding or painting. It can take more than a year to obtain the required experience or training to become certificated. FAA-certificated repairmen are employed by entities such as repair stations that are authorized by FAA to perform specific tasks. A repairman certificate is only valid at the employer for which it was issued. Non-certificated aviation maintenance workers include individuals who are supervised by certificated mechanics or repairmen in performing repair work. FAA maintains data on certificated mechanics and repairmen, including data on characteristics such as age and sex. FAA also maintains some data on non-certificated workers, such as the number employed by FAA- certificated repair stations, but neither the federal government nor the aviation industry maintains data on the total number of non-certificated aviation maintenance workers. Pathways to Becoming FAA-Certificated Career pathways consist of education, training, and support services that enable individuals to obtain industry-relevant certification and employment. There are three distinct pathways to become eligible to take the FAA mechanic tests—military training and experience, AMT School, and practical, or civil work experience (see fig. 1). FAA collects data on the use of the different pathways to becoming a certificated mechanic. Individuals must pass the FAA mechanic tests to become certificated, regardless of the pathway they take to become eligible to take the tests. There are three tests—written, oral, and practical. FAA publishes testing standards for the oral and practical skills tests. Military training and experience. The Community College of the Air Force administers an FAA-approved A&P training program which consists of on-the-job training and various courses for military service members with certain experience. When service members successfully complete the program, the Joint Services Aviation Maintenance Technician Certification Council issues them a certificate of eligibility to take the FAA mechanic tests (see side bar). Aviation Maintenance Technician School. Individuals may also attend an FAA-approved AMT School to become eligible to take the FAA mechanic tests. FAA approves and oversees AMT Schools and it maintains enrollment and mechanic test pass-rate data for each school. The minimum curriculum requirements for these schools are currently prescribed in regulation. The regulation includes the subjects the curriculum must cover and the number of training hours students must complete to become eligible to take the FAA mechanic tests. Given that AMT School curriculum requirements are in regulation, FAA has in the past attempted to amend the requirements through the federal rulemaking process. Practical work experience. People can also become eligible to take the FAA mechanic tests by demonstrating practical, or civil work experience. Individuals may work under the supervision of a certificated mechanic for 18 months for either an airframe or powerplant certificate, or 30 months for an A&P certificate. Practical work experience includes apprenticeships, which combine on-the-job training with classroom instruction. For certificated repairmen, there is no prescribed test, though repairmen must demonstrate their practical experience or have completed formal training to be certificated. Avionics technicians also have no prescribed test, but may seek certain related certifications. Occupational Data The Standard Occupational Classification (SOC) system is a federal statistical standard used to classify workers into occupational categories for purposes of collecting, calculating, or disseminating data such as employment levels and pay. The SOC structure forms the basis for the occupational coding system used by BLS’ Occupational Employment Statistics survey and Current Population Survey. Aviation maintenance workers generally fall into the avionics technicians or the aircraft mechanics and service technicians occupational group (see fig. 2). Both occupational groups include certificated and non-certificated individuals. The most recent revision to the SOC was for 2018. Selected Federal Agencies and the Aviation Maintenance Workforce In addition to the Department of Transportation (DOT) and FAA, several other federal agencies play a role in developing and maintaining a qualified aviation professional workforce. For example, we previously reported on related efforts administered by DOD, DOL, VA, and Education. These agencies provide either financial assistance for education or training in aviation maintenance related fields or administer programs that support career pathways to becoming an FAA-certificated mechanic or repairman. Existing Data Provide Limited Information about the Current Workforce and Potential Future Demand Federal Data Provide Information on the Number of Certificated Mechanics and Repairmen but Likely Overestimate How Many Are Available to Fill Projected Openings As of December 2018, about 295,000 individuals held a mechanic certificate and about 35,000 held a repairman certificate. The median age of FAA-certificated mechanics and repairmen was 54 years old, according to our analysis of the FAA data. Specifically, 52 percent were between the ages of 50 and 70 years old; 19 percent were between 39 and 49; and 19 percent were between 18 and 38. The remaining 10 percent were between the ages of 71 and 89 years old (see fig. 3). In comparison, about 23 percent of the overall workforce was age 55 or over according to BLS data as of 2018. Our analysis of FAA data also found that 3 percent of all aviation maintenance certificate holders were women as of December 2018. This percentage has not changed since we last reported on this workforce in 2014. In comparison, BLS data as of 2018 show that women made-up 47 percent of the total workforce. We were not able to analyze other demographic characteristics for these certificate holders, such as race or ethnicity, because neither FAA nor BLS collects these data. FAA data for 2015 through 2018 also provide some information on the education and work experience of certificated mechanics. These data show that attending AMT School was the most common pathway certificated individuals used to qualify for the FAA tests to become mechanics (see fig. 4). In addition, FAA data provide information on the number of newly certificated individuals and indicate that FAA certificated about 8,600 mechanics and repairmen on average each year for 2014 through 2018 (see fig. 5). BLS data project an annual average of 11,800 job openings in the United States from 2018-2028 for the aircraft mechanics and service technicians occupation due to growth and replacement, which include job openings for both certificated and non-certificated workers. The supply of workers to fill any open or projected job openings in the aviation industry, however, not only depends on the number of people qualified to do the work, but also their availability and willingness to work at a certain wage and under particular working conditions. While FAA data provide information on the number of mechanic and repairman certificate holders who are qualified to perform certain work, less is known about the number of them who are available and willing to work in the aviation industry. FAA data show there were approximately 330,000 certificated mechanics and repairmen as of December 2018, but FAA officials said this number likely overestimates the number of individuals working in the aviation industry. BLS data indicate 136,900 were employed in the aircraft mechanics and service technicians occupation in 2018, but it is not clear how many of those jobs were filled by FAA- certificated workers. In addition, it is unknown how many of the approximately 330,000 certificate holders are retired, deceased, or working in other industries. Individuals who obtain a mechanic certificate from FAA may never work in the aviation industry, or may begin their career in the aviation industry and leave for a job in another industry. Several stakeholders we interviewed said FAA-certificated mechanics possess certain skills that are transferrable to other industries and leave the aviation industry to work for other employers, such as amusement parks. Furthermore, officials explained that once certificated, there is no certification renewal requirement for mechanics. Federal Data Provide Some Information on Current Pay and Demand for Aviation Maintenance Workers, but Do Not Distinguish Between Certificated and Non- Certificated Workers BLS publishes some data on pay for aircraft mechanics and service technicians, such as average hourly and annual wages. However, the occupational classification system BLS and other federal statistical agencies use for aircraft mechanics and service technicians does not distinguish between FAA-certificated and non-certificated workers. This is in part because workers are classified by the work they perform and not necessarily by certification or education, according to SOC classification system principles. As a result, it is difficult to determine employment characteristics such as pay for certificated workers, specifically. BLS data as of May 2018 show that annual wages for aircraft mechanics and service technicians ranged from about $37,000 to about $98,000. According to BLS officials, it is not uncommon for there to be a wide salary range across an occupation, as wages may vary depending on factors such as experience, education, and skills. A DOD official we interviewed also said that employers have told him that they pay certificated aviation maintenance workers more than non-certificated workers. BLS officials said they collected wage and employment data for certificated workers separate from non-certificated workers in employer surveys conducted between 2000 and 2012. However, officials said they stopped collecting these data in part because employers inconsistently reported them. Data limitations at the federal and state levels also make it difficult to determine the demand for certificated aviation maintenance workers. BLS occupational data. On the federal level, BLS occupational outlook data provide some information on potential future demand nationwide for aviation maintenance workers, but the data do not distinguish between certificated and non-certificated workers. As a result, the data provide limited detail about the demand for certificated workers, specifically. According to BLS data, total employment for the aircraft mechanics and service technicians occupation is projected to grow about 3 percent over the 2018 to 2028 time frame, which is slower than the average for all occupations. As previously mentioned, these data project an annual average of 11,800 job openings for this occupation from 2018 to 2028 due to job growth and replacement. DOL certification data. On its public website for career planning and job search, CareerOneStop, DOL provides information on certifications that are frequently mentioned in online job postings and considered to be in-demand. DOL also indicates in its online resources which certifications may draw on training and experience gained in the military. However, DOL does not track or publish data on the demand for occupational licenses, including federal licenses such as FAA’s A&P certification. DOL officials said currently there are no plans to expand the agency’s data collection to include information on the demand for occupational licenses. DOL officials added that for certain jobs that require licenses, the demand for the required licenses mirrors occupational demand for those jobs so collecting those data may not be as meaningful. Workforce Innovation and Opportunity Act plans. On the state level, Workforce Innovation and Opportunity Act plans, intended in part to outline states’ use of federal funds to help workers meet employers’ needs, provide some geographically-specific information on the demand for workers in the aviation industry. Our review of states’ most recent Workforce Innovation and Opportunity Act plans found that 19 states identified the aerospace and aviation industry as a targeted sector for development. However, only certain plans specifically mention the need for certificated mechanics; others refer to the aviation industry more broadly. Employers we interviewed had differing perspectives on potential growth in demand for aviation maintenance workers; some said they were experiencing difficulty finding enough workers to meet their needs, while others said they were not experiencing difficulty. While some stakeholders voiced concerns about the potential for a labor shortage, the selected labor market indicators we reviewed for aircraft mechanics and service technicians (unemployment, wages, and employment) from 2013 through 2018 were not all consistent with the existence of hiring difficulties. See appendix I for our analysis. Officials we interviewed from a regional airline said the majority of the airline’s certificated mechanics come to them directly after completing AMT School and that the airline was having a difficult time finding enough mechanics to fill 60 open full-time positions. On the other hand, officials we interviewed from a major airline said the airline rarely hires certificated mechanics right out of AMT School and that their employees typically come to them with a number of years of experience. The officials from the major airline said that they were not experiencing difficulty recruiting and retaining aviation maintenance workers, but noted that regional airlines may experience hiring difficulties first if there is a shortage of these workers because certificated mechanics often start their careers at regional airlines to gain practical experience before moving on to work at a major airline. Government and Industry Programs Support the Workforce, but FAA Lacks Information that Could Advance Its Workforce Development Efforts Government, Industry, and AMT Schools Administer Programs and Coordinate to Some Extent to Support the Development of Aviation Maintenance Workers’ Skills and Career Pathways Registered Apprenticeship Program Serving Underrepresented Populations The Department of Labor awarded the Connecticut Department of Labor Office of Apprenticeship Training a $1,550,000 grant to fund the Connecticut Apprenticeship Expansion Rx project, which targets the aviation industry, among others. The project aims to serve over 1,600 apprentices and provide underrepresented populations, including women, dislocated, and under- employed individuals an opportunity to acquire industry required credentials. Key partners include industry, educational institutions, and labor unions. Registered Apprenticeship Program for Airframe and Powerplant Mechanics As part of the State Apprenticeship Expansion grant, the Alaska Department of Labor and Workforce Development (DOLWD) is implementing registered apprenticeships in aviation, a relatively new industry in using the apprenticeship model. With the help of the U.S. Department of Labor Office of Apprenticeship in Alaska, two aviation occupations have been approved: Airframe & Powerplant Mechanic and Air Transport Pilot. The Alaska state Apprenticeship Coordinator is working closely with the U.S. Department of Labor Office of Apprenticeship and individual air carriers across Alaska to develop and implement registered apprenticeships for these occupations. Alaska DOLWD has approximately 12 mechanic apprentices and 2 air transport pilot apprentices with various air carrier employers. support Registered Apprenticeship Programs— employer-driven training opportunities that combine on-the-job learning with related classroom instruction. The program facilitates coordination among different stakeholders such as industry, states, and educational institutions to support apprenticeships and employment opportunities. DOL awarded almost $3.8 million in grants and contracts from 2014 through 2018 to promote these apprenticeships for aviation maintenance workers. For example, one grantee is aiming to serve underrepresented populations in the aviation industry, including women, and another is coordinating with industry to develop a registered apprenticeship program for certificated mechanics (see side bars). In addition, the United Services Military Apprenticeship Program, a DOL registered program, provides service members an opportunity to improve skills and qualify for employment in a recognized civilian trade, including as an A&P mechanic, upon completion of military service. VA’s Post-9/11 GI Bill Program. The Post-9/11 Veterans Educational Assistance Act of 2008 (Post-9/11 GI Bill) provides funding for veterans to pursue an approved program of education, including undergraduate and graduate degrees, non-college degree programs, apprenticeships, and on-the-job training. VA data show approximately $42 million in Post-9/11 GI Bill funds were awarded in fiscal year 2018 to 4,200 veterans enrolled in aviation maintenance post-secondary programs, which include programs at FAA-approved AMT Schools. Education’s financial assistance. Education provides billions of dollars in federal assistance to support students pursuing higher education, which may include training in aviation-related fields. We previously reported that in academic year 2011-2012, Education disbursed $918 million in federal grants to 142,708 recipients and $1.3 billion in federal loans to 114,564 recipients pursuing aircraft mechanic and avionics programs. DOT’s workforce development grant program. DOT is also developing a process for administering a workforce development grant program for aviation maintenance workers. Specifically, the FAA Reauthorization Act of 2018 included a provision for DOT to establish an aviation maintenance workforce development grant program. Once established, eligible entities such as aircraft repair stations, unions, educational institutions, and state or local governments may apply for grants. The program may provide grants for projects such as establishing new educational programs or scholarships for individuals seeking employment in the aviation maintenance industry and supporting service members transitioning into aviation maintenance related careers. In addition, FAA has taken steps to engage other key stakeholders on aviation workforce development initiatives. In September 2018, FAA sponsored an Aviation Workforce Symposium that included participants from industry, AMT Schools, and federal agencies such as Education and DOL. Discussion topics included building the pipeline of workers, maximizing efficiency in training, and promoting productive partnerships. Subsequent to the 2018 symposium, FAA established an Aviation Workforce Steering Committee in February 2019, in part to coordinate efforts across FAA to address various workforce related provisions included in the FAA Reauthorization Act of 2018. The steering committee finalized its charter in April 2019, and the charter states FAA’s intentions of developing productive partnerships with industry, government, and educational institutions to expand the pipeline of aviation safety professionals. As of October 2019, FAA was finalizing a working group structure to carry out the steering committee’s work that will focus on: (1) marketing/communications, (2) educational outreach, (3) training, and (4) partnerships. FAA officials also told us they plan to collaborate with other federal agencies moving forward, including Education, DOL, and DOD. For example, FAA and DOD officials said they are currently discussing options for streamlining pathways for service members with aviation maintenance backgrounds to move into civilian careers in aviation maintenance. According to a DOD official, streamlining pathways could increase the number of service members who become FAA-certificated mechanics and leverage the skills of the over 250,000 current service members with aviation maintenance backgrounds. Additional examples of states, industry employers, and AMT Schools coordinating or partnering to support the aviation maintenance workforce include: Career grants. One state we visited developed a career grant to align students’ programs of study with in-demand occupations in the state. The grant provides tuition assistance to in-state residents working toward selected certificates or degrees at eligible in-state colleges or universities, including aviation maintenance programs. Military pathway program. Officials from a regional airline we interviewed said the airline developed a military transition program to assist service members in preparing for the FAA mechanic tests. The airline funds 100 percent of the program cost, which according to officials is about $11,000 per person. As part of the program, airline officials told us they provide about $5,000 worth of tools to each participant. Training equipment and funding. Officials we interviewed from one school said they strategically opened their AMT program next to a major cargo airline so that students could benefit from employment opportunities there. The officials said the airline also benefits from the close proximity of the school in that it is able to leverage local talent, and the airline provides AMT School students with scholarships, technical support, and surplus equipment to use for training. In another example, officials from a major commercial airline told us the airline partners with over 40 AMT Schools and provides them with funding to improve operations and recruitment. Officials said the goal of the program is to ensure the airline has a pipeline of workers to fill any future job openings. FAA Does Not Use Existing Data to Strategically Target Its Resources and Workforce Development Efforts Despite FAA’s recent efforts to coordinate with other federal agencies on expanding and streamlining pathways for aviation maintenance careers, we found that FAA does not routinely analyze, collect, or coordinate with other stakeholders on certain related data. Such activities could assist FAA in measuring progress toward meeting its workforce related objectives and inform strategic decisions for promoting the development of this workforce. For example, FAA’s strategic plan includes an objective on promoting the development of a robust aviation workforce. In addition, FAA’s Aviation Workforce Steering Committee charter emphasizes providing diverse populations, including youth, women, and minorities, with clear pathways into aviation careers to expand the talent pool from which both government and industry may recruit. However, neither the strategic plan nor the steering committee charter provides specific information on how FAA plans to select and measure any efforts it undertakes related to these objectives. Prior GAO work has emphasized that strategic workforce planning requires monitoring and evaluating progress toward goals, and federal internal control standards state that management should use quality information to achieve its objectives. We identified several areas in which improved data analysis, collection, or coordination could assist FAA in measuring progress and understanding how to target its resources in support of its workforce related objectives. Demographic data. FAA collects certain demographic data on its A&P certification application, such as the age and sex of individuals; however, FAA currently uses these data only to determine eligibility and issue certificates, according to FAA officials. These data could also be used to identify patterns or relationships, such as the trend in female certificate holders by pathway, which could be useful information as FAA aims to increase opportunities for women to pursue aviation maintenance careers. In addition, FAA does not currently collect data on the race and ethnicity of certificated individuals. Such data could provide additional information on the demographics of certificated individuals and help FAA or other stakeholders monitor the progress of any efforts to diversify this workforce. FAA could also leverage BLS data on the race and ethnicity of certificated and non-certificated aircraft mechanics and service technicians more broadly as it begins to develop and implement any activities related to expanding and diversifying the talent pool for recruiting workers into aviation maintenance careers. Pathway data. FAA also maintains mechanic pathway data, but these data do not provide a complete picture of certificated individuals’ education, training, and work experience due to certain data limitations. For example, FAA does not require AMT Schools to report program completion data. As a result, it does not have information such as how many students who enter FAA-approved AMT Schools complete the program. Moreover, FAA does not analyze nationwide trends for AMT Schools using existing data on these schools (such as enrollment or mechanic test pass-rate data) or aggregate information across AMT Schools to better understand the AMT School pathway as a whole. In addition, pathway data collected by FAA do not clearly differentiate between civil and military work experience. Specifically, FAA officials said practical, or civil, work experience pathway data may include information on individuals with both prior military and civil experience. Moreover, according to an FAA official, FAA’s military experience pathway data may include individuals who completed DOD’s FAA- approved A&P training program as well as individuals who met FAA’s on-the-job training requirements through relevant military experience. Combined pathway data may limit FAA’s and DOD’s understanding of DOD’s contributions to this workforce, including the number of individuals who completed DOD’s FAA-approved A&P training program and subsequently obtained mechanic certification from FAA. Supply and demand data. Other federal agencies, such as BLS and DOD, maintain data that relate to this workforce more broadly that could be useful to FAA. For example, FAA could leverage BLS data on the projected employment of certificated and non-certificated aircraft mechanics and service technicians in conjunction with its data on newly certificated workers each year to better understand worker supply and demand. DOD also maintains information on separating service members with aviation maintenance backgrounds, who may be attractive to the commercial aviation industry. For example, according to a DOD official, in fiscal year 2018 over 22,000 service members with aviation maintenance backgrounds separated from the Air Force and Navy. Additional data analysis and coordination could potentially yield useful information on worker supply and demand and areas for promoting the development of this workforce. Without routinely analyzing its existing data on certificated workers, collecting additional data, or leveraging existing workforce data maintained by other federal agencies, FAA will not have certain information it needs to measure progress and strategically target its resources toward its objective of promoting the development of a robust aviation workforce. A robust aviation workforce, including certificated mechanics and repairmen, is necessary for maintaining a safe aviation system. FAA’s recently developed Aviation Workforce Steering Committee presents the agency with an opportunity to engage other federal agencies in discussions on how to leverage data to expand and diversify this workforce. FAA Has Proposed Changing Its Decades Old Mechanic Curriculum Requirements and Its Testing Standards, and Revisions to Them Are Ongoing Even as FAA’s strategic plan states the agency’s focus on promoting the development of a skilled aviation maintenance workforce to integrate new technologies, the agency has acknowledged that the current curriculum requirements for AMT Schools and mechanic testing standards are outdated. Efforts to revise the curriculum requirements for AMT Schools are ongoing through the rulemaking process, and FAA is also currently updating the testing standards for mechanics. The curriculum requirements for AMT Schools have remained largely unchanged for several decades despite numerous attempts to update them as aviation technology has evolved. The minimum requirements are established in regulation and list the subjects that AMT Schools must include in their training curriculum for individuals to be eligible to take FAA’s mechanic tests. FAA officials, employers, and AMT School officials we interviewed said the current curriculum requirements do not emphasize commonly used modern aircraft technologies, such as avionics and composite materials. Because the curriculum requirements are established in federal regulation, FAA has attempted several times to revise them through the rulemaking process. Table 1 provides selected changes or actions relating to these requirements. FAA officials noted several challenges to updating the AMT School curriculum requirements, including competing demands at the department level, the extent of comments FAA has received from stakeholders in response to proposed changes, and the amount of time required to coordinate with internal stakeholders during the review process. We previously reported on factors that affect the amount of time needed to issue a rule for selected agencies, which included similar challenges such as the complexity of an issue, agency management priorities, and the amount of review required at different phases of the rulemaking process. In October 2015, FAA published a notice of proposed rulemaking (NPRM) with the stated goal of updating the existing AMT School curriculum and providing an efficient means of changing specific course items by including them in each school’s operations specifications (see fig. 6). This would eliminate the need to go through the federal rulemaking process to update the curriculum. As part of its ongoing efforts to revise the curriculum requirements for AMT Schools through the rulemaking process, FAA issued a supplemental NPRM in the April 2019 Federal Register that expanded the scope of the NPRM it issued in October 2015 (see fig. 6). Comments on the supplemental NPRM were due in June 2019. As of October 2019, FAA officials said they were in the process of reviewing the comments. In a separate effort outside of the rulemaking process, FAA is currently updating the testing standards for mechanics. The standards were last revised in 2015. FAA has acknowledged that current mechanic testing standards are also outdated. As a result, aviation stakeholders have stated that the mechanic tests include outdated or irrelevant questions. For example, the practical test may include projects on wood airframes and fabric coverings, which are not common to modern commercial aircraft. FAA has stated that the revised testing standards will provide a comprehensive framework for the mechanic tests and serve as a guide for reviewing and revising the oral and written test questions and the practical projects. FAA officials said two offices within the agency are responsible for updating AMT School training curriculum requirements and mechanic testing standards and that these offices have been coordinating efforts to align the two. FAA’s efforts to modernize the curriculum requirements for AMT Schools and its efforts to update the mechanic testing standards started on slightly different paths, in part due to differences in when the working groups were formed and recommendations to address these issues were made (see fig. 7). As of October 2019, FAA had not issued a final rule for modernizing AMT School curriculum requirements as required by the FAA Reauthorization Act of 2018, and it was still in the process of updating testing standards. FAA officials have indicated that they have informed the appropriate committees in Congress that the proposed schedule for issuance of a final rule is in October 2020 and said that the revised mechanic testing standards would likely be finalized after the publication of the final rule amending the curriculum requirements for AMT Schools. An FAA official noted that any delay in finalizing the rule would likely result in a corresponding delay to finalizing the testing standards. Delaying the release of the updated mechanic testing standards could result in the prolonged use of outdated or irrelevant questions on the mechanic tests. FAA officials said that once finalized and implemented, the updated curriculum requirements for AMT Schools and the mechanic testing standards for individuals should be mostly aligned. Conclusions A sufficient supply of aviation maintenance workers is critical to maintaining a safe and robust aviation system and meeting the growing demand for air travel. Current training and skills requirements for these workers are also important because of changing flight technology. Both the federal government and other industries benefit from having a professional, trained, and qualified workforce, and addressing aviation workforce needs is a shared responsibility among these different stakeholders. As the federal agency responsible for certificating aircraft mechanics and repairmen, FAA maintains certain demographic information on these individuals that could shed light on the characteristics and employment of these personnel. However, without strategically using or analyzing the data it has along with data other stakeholders collect, FAA will not have certain information it needs to target its resources or measure and improve progress toward its aviation workforce goals. It may also miss the opportunity to provide other stakeholders with valuable information for supporting these workers. Other agencies and stakeholders may also assist FAA in understanding and promoting the development of the aviation maintenance workforce. FAA’s recently developed Aviation Workforce Steering Committee presents the agency with an opportunity to engage other federal agencies to explore potential data sources and their usefulness and discuss ways to expand, diversify, and strengthen career pathways for the aviation maintenance workforce. Recommendation for Executive Action The Administrator of FAA should direct the Aviation Workforce Steering Committee, as part of its ongoing efforts, to take steps to use existing FAA data and coordinate with other federal agencies to identify and gather the information it needs to measure progress and target resources toward its goal of promoting a robust, qualified, and diverse aviation maintenance workforce. For example, FAA could task a committee working group with developing and implementing ways to improve data sharing among federal agencies to inform decision-making on how to strengthen career pathways and better understand the supply and demand of certificated workers. (Recommendation 1) Agency Comments We provided a draft of this report to DOT, DOL, Education, DOD, and VA for review and comment. DOT provided written comments, which are reprinted in appendix II. DOT concurred with our recommendation. Specifically, DOT agreed that using existing data could potentially contribute to its efforts to develop the aviation maintenance workforce. DOT said it will ask the Aviation Workforce Steering Committee to consider using existing FAA data and to coordinate with other federal agencies regarding other potential data sources to support the FAA’s aviation maintenance workforce goals. DOL provided technical comments, which we incorporated in the report as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of the Department of Transportation, the Secretary of the Department of Labor, the Secretary of the Department of Education, the Secretary of the Department of Defense, the Secretary of the Department of Veterans Affairs, and other interested parties. In addition, the report is available at no charge on the GAO website at https://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7215 or gurkinc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. Appendix I: Labor Market Indicator Analysis While no single metric can be used to determine whether a labor shortage exists, certain indicators in conjunction with views of stakeholders can provide insight on this issue. We previously found, based on our review of economic research, that an occupation experiencing a labor shortage would exhibit the following: (1) a low unemployment rate signaling limited availability of workers in that profession, (2) increases in wages offered to draw people into that profession, and (3) increases in employment due to increases in demand for that occupation. Table 2 shows these specific indicators from 2013 through 2018 for the aircraft mechanics and service technicians occupation, measured using Bureau of Labor Statistics (BLS) Current Population Survey data. According to our analysis of BLS data from 2013 through 2018, unemployment rate and change in median wage earnings for the aircraft mechanics and service technicians occupation, which includes both Federal Aviation Administration-certificated and non-certificated workers, were consistent with the existence of hiring difficulties, while the percent change in employment was not. Data on two of the three indicators (unemployment rate and wage earnings) were consistent with difficulties in hiring aircraft mechanics and service technicians. However, because these data combine information for certificated and non-certificated workers, it is difficult to know the extent to which any hiring difficulties represent demand for certificated workers, specifically. In addition, the indicators should be viewed with appropriate caveats. For example, while median wages increased for aircraft mechanics and service technicians in 2018 compared to 2013, they did not increase in every year—and they exhibited decreases of as much as 6.7 percent. The direction of change of the employment indicator was not consistent with hiring difficulties for this occupation. As shown in table 2, from 2013 through 2018, employment for aircraft mechanics and service technicians does not appear to have changed, while employment for all other occupations increased. However, employment for this occupation did not remain constant in every year over that time period and exhibited increases of as much as 12.5 percent and decreases of as much as 21.9 percent. Appendix II: Department of Transportation Agency Comments Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Chelsa Kenney Gurkin at (202) 512-7215 or Gurkinc@gao.gov. Staff Acknowledgments In addition to the contact named above, Betty Ward-Zukerman (Assistant Director), Meredith Moore (Analyst-in-Charge), Ellie Klein, and Chris Woika made key contributions to this report. Additional assistance was provided by James Bennett, Lilia Chaidez, Holly Dye, Serena Lo, Sheila R. McCoy, John Mingus, Michael Naretta, James Rebbe, Almeta Spencer, and Andrew Von Ah.

Why GAO Did This Study FAA requires that only mechanics who are “certificated” by the FAA approve aircraft for return to service. The required training to become a certificated mechanic can take between 1 and 3 years. FAA also oversees the certification of repairmen who work on aircraft parts. Some stakeholders have expressed concern that retirements and attrition could adversely affect the capacity of this workforce to meet the growing demand for air travel, and that mechanic curriculum is outdated. The FAA Reauthorization Act of 2018 included provisions for GAO to examine the aviation workforce. This report examines (1) what available federal data reveal about the FAA-certificated aviation maintenance workforce; (2) how selected government agencies, educational institutions, and businesses provide support and coordinate to develop the aviation maintenance workforce; and (3) the progress FAA has made in updating its curriculum and testing standards for mechanics. GAO analyzed FAA data on certificate holders as of December 2018; reviewed BLS employment data for 2013 through 2018; reviewed relevant federal laws and regulations; and interviewed selected federal agency, industry, and AMT School officials. What GAO Found Federal data provide an incomplete picture of the Federal Aviation Administration (FAA)-certificated aviation maintenance workforce. A sufficient supply of certificated workers is critical for safety and to meet the growing demand for air travel. However, supply and demand data for certificated workers are limited. FAA maintains data on the number of individuals newly certificated each year (see figure), but less is known about how many certificated individuals exit the aviation industry each year and the extent of growing demand. Bureau of Labor Statistics (BLS) data provide some information on pay and demand for aviation maintenance workers more broadly, but do not differentiate between FAA-certificated and non-certificated workers due to data collection challenges. Demographic data may also be useful for workforce analysis and planning. FAA data provide some demographic information on certificated mechanics and repairmen, such as age and sex, but the agency lacks data on race and ethnicity. According to GAO analysis of FAA data, over half of the roughly 330,000 mechanics and repairmen FAA had certificated as of December 2018 were between 50 and 70 years old and 97 percent were men. Government agencies, educational institutions, and businesses coordinate to some extent in support of this workforce, but FAA lacks certain information—including information maintained by other agencies that administer related programs—that could advance its workforce development efforts. One of FAA's strategic objectives includes promoting the development of a robust, skilled aviation workforce, and the agency established a committee, in part, to explore ways to diversify this workforce; however, FAA is not currently positioned to understand whether its efforts are optimally targeted or effective. Without routinely analyzing its own data or leveraging others' data, FAA may not have certain information it needs to track or ensure progress toward its workforce development goals. FAA has acknowledged that curriculum requirements for Aviation Maintenance Technician (AMT) Schools and mechanic testing standards are outdated. Efforts to revise the curriculum requirements for AMT Schools are ongoing and FAA officials told GAO that a final rule will be published some time toward the end of 2020. FAA officials indicated that the revised mechanic testing standards would likely be finalized after. What GAO Recommends GAO recommends that FAA use its existing data and coordinate with other federal agencies to identify and gather information to measure progress and target resources toward its goal of promoting a robust, qualified, and diverse aviation maintenance workforce. FAA agreed with the recommendation.

gao_GAO-20-30

gao_GAO-20-30_0

Background The primary purpose of the CORD Project is to develop and implement strategies for reducing obesity among low-income children. According to CDC, strategies that have been used to prevent and manage obesity include screening patients using body mass index (BMI), so children and their parents understand their risks; supporting healthy behaviors—such as eating vegetables and promoting physical activity—in early care and education centers and schools; and educating parents on how to reinforce healthy living habits at home. BMI is used to determine overweight and obesity (see sidebar). Percentiles are calculated from the Centers for Disease Control and Prevention (CDC) growth charts developed from national survey data collected between 1963 and 1994. Funding for the CORD Project was first made available through the enactment of the Patient Protection and Affordable Care Act, about one year after the CORD Project was authorized. In January 2011, CDC published the funding opportunity announcement—which outlined the goals of the grant as well as the eligibility criteria and other requirements—and, in September 2011, the first demonstration projects began. Congress subsequently appropriated additional funding for the CORD Project in April 2015 and January 2018, bringing the total amount appropriated to $65 million for fiscal years 2010 through 2023. CDC officials told us that during this time period the CORD Project was the primary source of CDC funding for childhood obesity research focused on low-income children. CDC implemented the CORD Project in three separate grant phases, with different design approaches and grantees. (See fig. 1.) Across the three CORD Project phases—only the first of which is complete—CDC has awarded ten grants to entities to implement demonstration projects aimed at reducing obesity in low-income children. (See table 1.) In the first phase of the CORD Project, CDC also awarded a grant to the evaluation center to conduct a cross-site evaluation of the implementing grantees’ demonstration projects. CDC Has Made Four Key Changes to the CORD Project Design in Response to Lessons Learned and National Recommendations CDC Made Four Key Design Changes to the CORD Project between Each Grant Phase CDC made four key design changes between the three CORD phases. CDC changed the scope of the project (i.e., type of strategies implemented), the type of evaluations (i.e., how it evaluated the strategies), the purpose of the study design, and the extent of participation by state Medicaid or CHIP programs. (See fig. 2.) CDC officials designed the CORD Project based on the language and requirements in CHIPRA and the CORD Project Plan developed by HHS, according to CDC officials. For CORD phases 2 and 3, CDC officials modified elements of the design in response to lessons learned, time frames for implementation, and recommendations related to childhood obesity made by national organizations such as the U.S. Preventive Services Task Force (hereafter referred to as the Task Force). Scope of CORD Project. After CORD phase 1, CDC officials shifted the scope of the CORD Project from prevention to the treatment of children who are overweight or have obesity, according to CDC officials. Specifically, CDC designed CORD phase 1 to require grantees to implement demonstration projects that integrated public health and primary care strategies by promoting children and their families’ use of healthy behaviors and by modifying community environments. CORD phase 1 grantees implemented strategies in two types of settings: (1) community and (2) health care settings. Public health strategies are activities and programs delivered in community settings, such as schools and early care and education centers. Grantees also implemented primary care strategies, which in general are BMI screenings or other activities implemented in health care settings, such as during physician visits in federally qualified health centers. While CORD 1 grantees implemented strategies in both types of settings, the specific strategies that each CORD phase 1 grantee implemented varied. (See text box and app. I for additional information about the strategies CORD 1 grantees implemented.) Examples of Strategies Childhood Obesity Research Demonstration Phase 1 Grantees Implemented California Demonstration Project Public health strategies implemented included training staff at early care and education centers on health behavior change strategies and providing centers with large self-service water containers to promote increased water intake. Primary care strategies implemented included body mass index (BMI) screenings for children participating in early care and education centers. BMI is a measure used to determine overweight and obesity. Massachusetts Demonstration Project Public health strategies implemented included training teachers in participating elementary schools on how to implement evidence-based health education curricula that encouraged learning about nutrition and physical activity. Primary care strategies implemented included establishing a healthy weight clinic located in the participating health centers. Texas Demonstration Project Public health strategies implemented included providing classroom-based nutrition and gardening curricula in the early care and education centers. Primary care strategies included modifying electronic health records systems to increase provider awareness and action related to maintaining healthy weight, such as prompting clinicians to refer children who were overweight or had obesity to additional services. Pediatric Weight Management Interventions The Centers for Disease Control and Prevention (CDC) defines pediatric weight management interventions as intensive behavioral interventions designed to address excess weight through child and parental counseling on diet, physical activity, or behavior change management. The U.S. Preventive Services Task Force refers to these as interventions for weight management interventions. For CORD phases 2 and 3, CDC shifted the scope of the CORD Project to the treatment of children who are overweight or have obesity. Specifically, CDC modified the scope to only focus on implementing pediatric weight management interventions, one type of primary care treatment strategy (see sidebar). CDC officials told us they changed the scope of CORD phase 2 in response to the shorter, 2-year funding period authorized by law. CDC officials stated that unlike CORD phase 1, the shorter time frame for CORD phase 2 did not allow for a planning year to establish and solidify community relationships across multiple community settings while also enabling sufficient time to implement the strategies and analyze outcome and other data. CDC also modified the scope for CORD phase 2 and 3 to focus on pediatric weight management interventions in response to existing national recommendations related to childhood obesity, according to CDC officials. The Task Force recommended that primary care providers screen children 6 years and older for obesity and offer, or refer children with obesity to, pediatric weight management interventions. In making its recommendation, the Task Force found that pediatric weight management interventions should involve at least 26 hours of contact between the provider and the child, family, or both over a period of 2 to 12 months. According to CDC’s funding opportunity announcement for CORD phase 2, a 2007 expert committee convened by the American Medical Association similarly recommended that all health care providers address weight management and lifestyle issues with children at least once a year and provide behavior counseling on key obesity-related behaviors. CDC officials stated they designed CORD phase 2 to meet the guidelines and standards outlined in these recommendations. After Congress extended the CORD Project for a 6-year period beginning in fiscal year 2018 and appropriated additional funding, CDC designed the scope of CORD phase 3 as a 5-year grant to continue efforts to implement pediatric weight management interventions only. CDC officials stated they considered returning to an integration of public health and primary care strategies for CORD phase 3, similar to CORD phase 1, but decided that the best use of resources was to focus on integrating pediatric weight management interventions into communities, which includes linking families with resources already available in the community, such as low-cost physical activity offerings. Common Outcome Measures for Childhood Obesity Research Demonstration (CORD) Phase 1 Frequency of fruit and vegetable consumption Frequency of sugar-sweetened beverage consumption Physical activity Sleep time Screen time (e.g., watching television and playing video games) Body mass index Quality of life (e.g., physical, emotional, and social) Type of evaluations. In CORD phase 1, CDC awarded a grant to another entity—the evaluation center—to conduct a cross-site evaluation to aggregate results of the three implementing grantees’ demonstration projects. In designing CORD phases 2 and 3, CDC did not award grants to independent entities to conduct cross-site evaluations of the implementing grantees’ demonstration projects. In CORD phase 1, the cross-site evaluation was intended to help inform national policy decision- making, including recommendations regarding the applicability of CORD strategies in other communities. To assess the effectiveness of CORD phase 1, CDC designed the cross-site evaluation to examine the demonstration projects using a set of common outcome measures, which the evaluation center developed in collaboration with CDC officials and implementing grantees (see sidebar). CDC officials told us they removed the cross-site evaluation component for CORD phases 2 and 3 in part due to challenges executing the cross- site evaluation in CORD phase 1. Officials explained, for example, that the difficulty in developing common outcome measures that could be analyzed across the three demonstration projects that were both valid and specific enough to the strategies was a challenge given the variation in the strategies implemented by each grantee, data collection time frames, and methodologies. In addition, CDC officials stated the implementing grantees had sufficient capacity to conduct their own evaluations. For these reasons, CDC officials said they concluded that the cross-site evaluation was not an efficient use of resources. The CORD phase 1 grantees also identified the following challenges related to the cross-site evaluation: Grantees told us there was insufficient time to develop the common outcome measures prior to implementing the strategies. One grantee noted this resulted in them needing to collect some data retrospectively instead of collecting it in real time. Grantees also collected data at different time frames from each other, which resulted in limited data for measuring outcomes via the common measures. Evaluation center officials stated that the lack of a common timeline for collecting data resulted in them only being able to analyze changes in common outcomes measures at the two common time points across all three grantees—baseline and 12 months—even though some grantees collected data at later time points (e.g., 24 months after implementation began). Thus, evaluation center officials said they were unable to determine whether changes in outcomes observed were sustained 24 months after implementation. Grantees reported challenges in creating valid common outcome measures applicable across the varying age ranges, locations, and strategies implemented for the three demonstration projects that affected results of the cross-site evaluation. For example, only the Massachusetts demonstration project chose to implement strategies in the Special Supplemental Nutrition Program for Women, Infants and Children offices, making any data collected about that strategy unable to be included in the cross-site evaluation. CDC required the implementing grantees in all CORD phases to conduct their own evaluations and report on outcomes associated with the strategies implemented under their demonstration projects. Specifically, CDC expected the grantee-specific evaluations to measure health outcomes—such as changes to BMI, nutrition, and physical activity—and quality of life, and to report information on the processes, outcomes, and costs of the individual demonstration projects in the evaluations. Purpose of study design. While CDC designed CORD phases 1 and 2 to build knowledge and evidence on strategies for reducing obesity among low-income children, CDC designed CORD phase 3 to focus on translating strategies proven to reduce childhood obesity into routine use for low-income families. More specifically, for CORD phases 1 and 2, CDC required grantees to use or adapt strategies that previously had not been rigorously tested in low-income children. For example, in CORD phase 2, the Arizona demonstration project adapted a preexisting program—which was aimed at preventing child behavior issues through motivational interviewing techniques and parent education—to improve weight-related health behaviors in low-income children. By comparing low-income participants receiving the strategies with those who did not, the Arizona demonstration project aims to develop evidence about whether or not these strategies work to reduce obesity in low-income children. Arizona officials told us that while the CORD phase 2 study design is appropriate for helping to expedite the translation of knowledge into practice, it has nonetheless been challenging to implement the demonstration project in a 2-year period. The officials explained 2 years is a short period of time for this type of demonstration project. For CORD phase 3, CDC is requiring grantees to take an existing evidence-based pediatric weight management intervention and convert it into a user-friendly package of information, containing all materials clinical or community-based entities would need to easily, efficiently, and completely replicate the pediatric weight management intervention. Materials may include implementation manuals, training curricula, technical assistance, and evaluation materials. CORD phase 3 grantees are required to partner with clinical or community entities that will then use the package to implement the set of pediatric weight management interventions in their community. Additionally, CORD phase 3 grantees are required to make edits to the packaged materials based on the results of the implementation and develop sustainability and dissemination plans to implement the pediatric weight management intervention at additional locations. CDC officials and agency documentation outlined multiple reasons why they modified the study design for CORD phase 3. For example, in its funding opportunity announcement for CORD phase 3, CDC noted that there have been challenges in moving research-based, national recommendations, like Task Force recommendations, into practice. According to CDC officials, this challenge is especially great in low- income communities, where there are a limited number of available pediatric weight management interventions that are rigorous enough to meet the standards outlined by the Task Force. Additionally, officials noted that when these interventions are available, families are generally charged for the services. CDC officials told us that, according to the literature, it can take many years for evidence-based clinical interventions to make it into mainstream practice. Thus, by designing CORD phase 3 to package evidence-based pediatric weight management interventions that will be targeted to communities with low-income families, CDC officials told us they hope to reduce the number of years before adoption of such treatment strategies is prevalent. Participation by state Medicaid or CHIP program officials. In the design for CORD phase 1, implementing grantees were not required to develop relationships with officials from their state Medicaid or CHIP offices or with other payers, but these relationships were encouraged, according to CDC officials. At each subsequent CORD phase, CDC modified its expectations of grantees regarding the involvement of state Medicaid and CHIP program officials in the demonstration projects. Specifically, CDC added a requirement that the implementing grantees form a payer advisory board with representatives from state Medicaid or CHIP offices and encouraged grantees to collaborate with other relevant health care stakeholders, such as private payers, to foster discussions about how to obtain reimbursement for CORD strategies. Noting the importance of establishing these types of relationships, CDC officials told us that grant funding and in-kind donations—which CDC encouraged grantees to identify and use to supplement CORD grant funding—are not sustainable sources of funding for continued implementation of childhood obesity programs. As a result, the officials told us reimbursement from insurers, such as Medicaid or CHIP, is necessary to sustain the implemented strategies at the level of intensity required by the Task Force recommendations. For example, in CORD phase 2, officials from the Arizona demonstration project told us they included representatives from United Healthcare’s private and Medicaid health plans and also a representative from Mercy Care, a not-for-profit Medicaid plan, on their payer committee. Arizona demonstration project officials stated they were working with representatives of the state Medicaid program and private health plans to determine what kind of evidence payers would need to reimburse for obesity-related services. Reimbursement is a key focus in CORD phase 3, and CDC officials told us they plan to assist grantees in determining which services provided within the pediatric weight management interventions may be reimbursable. Specifically, CDC officials stated they will coordinate opportunities for information sharing, technical assistance, and networking between CORD phase 3 grantees, states, and CMS in order to explore broader Medicaid and CHIP coverage options for the services delivered through the grants. CMS officials noted that medical services provided under the grant could be reimbursable under states’ Medicaid and CHIP programs, including under the Early and Periodic Screening, Diagnostics and Treatment benefit. CDC Used a Similar Approach to Grantee Management in the First Two CORD Phases While CDC changed some design elements of the CORD Project between the phases, according to CDC officials, the agency used a consistent approach in managing grantees. Specifically, CDC officials told us that in CORD phases 1 and 2 they promoted collaboration between themselves and the grantees, as well as among the grantees, and monitored the grantees through regular interactions with them. CDC officials told us they used a team of personnel with different expertise to oversee the CORD phase 1 and 2 grants. For example, the team included a project officer who specialized in program management to oversee the day-to-day operations, as well as subject matter experts, including one experienced in evaluation design. CDC officials told us they interacted with CORD phase 1 and 2 grantees on regular conference calls and conducted annual site visits to each grantee. Grantees stated that CDC’s site visits aided them in implementing their demonstration projects by keeping them and their community partners accountable. In addition, grantees told us that CDC collaborated with them to provide expertise on, or troubleshoot the design of, the implementation of their demonstration projects. For example, Arizona demonstration project officials told us that CDC officials helped them to figure out how to best achieve their desired sample size for a strategy they were implementing. CDC officials told us they plan to continue a similarly collaborative management approach for CORD phase 3. CDC officials stated they also monitored CORD phase 1 and 2 grantees by requiring grantees to regularly report on their efforts and generally plan to monitor CORD phase 3 grantees the same way. For example, CDC required CORD phase 1 grantees to submit annual progress reports at least 90 days before the end of the budget period that included descriptions of progress made towards the research goals, information on expenditures, and a detailed budget justification for the new budget period. CDC also required CORD phase 1 grantees to submit both annual progress reports and a final progress report. CORD phase 1 grantees told us that CDC officials were helpful in providing administrative support that ensured grant paperwork was completed consistent with requirements. Evaluations Show Some Improvements for the Completed Demonstration Projects and CDC and Grantees Identified Factors Affecting Implementation Evaluation Center and Grantees Reported Some Improvements for Children Receiving the Strategies in the First Phase of the CORD Demonstration Projects The evaluation center’s cross-site evaluation and the implementing grantees’ evaluation findings reported some improvements in BMI and other outcomes measured among children who received CORD phase 1 strategies. Specifically, the evaluation center reported that positive changes on these outcomes were observed most often among the following groups of children, providing some evidence of the effectiveness of the strategies delivered: Children who received primary care strategies, such as individualized counseling. Children who received public health strategies, such as an evidence- based nutritional program, in addition to the primary care strategies. In evaluating the CORD 1 demonstration projects, the evaluation center did not examine which specific strategies were the most effective. The primary objective of the cross-site evaluation was to determine if there was evidence that an integrated approach had any advantage over implementing either public health only or primary care only strategies. The evaluation center examined the extent to which the three CORD 1 demonstration projects collectively were associated with positive changes over time in behavior or reductions in BMI. Because of the considerable variation in each of the three demonstration projects, the evaluation center grouped the various strategies implemented by the three grantees into two categories for the analysis: public health and primary care plus. Next, the evaluation center categorized children by the types of strategies they received (public health only, primary care plus only, or both public health and primary care plus) and by age (2 to 5 years, 6 to 8 years, and 9 to 12 years). The evaluation center tested whether each of the possible combinations of strategy and age showed improvement over a 12-month period for each common measure. Using this approach, the evaluation center found some improvements for all of the common outcomes measured; however, improvements were not observed for each strategy or age group. Specifically, of the 81 possible combinations of strategy and age, 52 demonstrated some improvement over the 12-month period; however, only 16 of them showed a statistically significant improvement. (See fig. 3.) For example, BMI improved for children over the 12-month period in three of the strategy and age combinations, but the improvement was statistically significant for just one of those combinations. Among the 52 groups that showed improvements at 12 months, most of the differences observed were very small. For example, from the start of the intervention to 12 months after the intervention, there was about a 1 percent increase in the percentage of children who reported they were physically active for 60 minutes at least one day a week. The implementing grantees—each of which conducted their own evaluations—also reported some improvements in the children who received CORD phase 1 strategies. Similar to the evaluation center’s findings, the implementing grantees did not report improvements for all participating age groups or all outcomes they examined. Among their findings, the grantees reported the following: Children at participating early care and education centers in Texas, who were exposed to strategies such as classroom-based nutrition and gardening curricula, demonstrated modest improvements in BMI over a 2-year period when compared with children who did not receive these Texas demonstration project strategies. The Texas demonstration project also reported improvements in BMI for some children who participated in a weight management program administered in YMCAs compared with a different weight management program administered in primary care clinics. Specifically, researchers found that the YMCA program was more effective in reducing BMI for low-income children at 3 months but not at 12 months after implementation of the program. Children who received both public health and primary care strategies under the California demonstration project experienced some improvement on some outcome measures when compared to children who only received one type of strategy. For example, children who are overweight or have obesity who received both public health and primary care strategies reported playing less hours of video games during the week than those who only received the primary care strategies. During CORD phase 1, the Massachusetts demonstration project observed some improvements over time in the children who received CORD strategies. For example, the percentage of seventh grade students with obesity decreased from the start of implementation compared with 24 months after implementation in the two communities where the strategies were implemented. However, these results were modest; the decrease in the percentage of students with obesity was less than 3 percent in both communities. CDC officials, implementing grantees, and the evaluation center noted that modest or no effects were likely in part due to small sample sizes because of recruitment issues. Regarding recruitment, CDC officials told us that two of the three CORD phase 1 demonstration projects had issues with recruitment that caused sample size issues and ultimately statistical power issues. Specifically, when there is a smaller sample size, a study may be underpowered, which means that statistically significant effects are less likely to be detected even when differences exist. CDC officials explained that having limited statistical power affects the ability for more specific modeling or analysis to determine for whom the strategies works best (e.g., those with obesity or severe obesity). Grantees and CDC officials told us that when faced with recruitment issues, grantees made changes to their recruitment strategies. For example, grantees reduced the minimum BMI required for children participating in the demonstration projects in an attempt to increase participation. However, grantees told us they were still not able to reach their anticipated number of participants. Additionally, an official from the evaluation center told us some common outcome measures used in the cross-site evaluation were limited. The official explained that, had the grantees had more time to reach consensus on how to collect the data for the common outcome measures, or had the common outcome measures been identified in advance of the implementing grantees developing their own evaluations using measures specific to their demonstration projects, the evaluation center might have had more precise data to demonstrate improvements among participants. In planning for CORD phase 1, CDC officials acknowledged that the demonstration projects might not result in significant changes for some outcomes. CDC’s funding opportunity announcement noted that changes in health indicators, such as BMI, are long-term objectives, and that the period of funding for the projects might be too short to demonstrate significant improvement in these outcomes. A CDC official stated that although strong results were not found across each of the demonstration projects, the results of the implemented strategies provided evidence that these strategies could be implemented in a real-world setting. Thus, they noted the lack of stronger and larger effects does not mean that the demonstration projects were not successful. CORD phase 1 grantees told us they continue to analyze the data and expect to publish additional findings, even though the grant period has concluded. For example, the evaluation center told us they had enough data from the CORD Project to continue publishing for many years and planned to publish studies examining how existing community policies— such as physical activity policies—affected the outcomes of the implemented strategies. One of the implementing grantees also told us that having more time to fully analyze, use, and publish results from the data was needed. CDC and Grantees Identified Several Factors Affecting the Implementation of Strategies to Reduce Obesity among Low- Income Children CDC officials and CORD grantees identified several factors that affected grantees’ ability to implement strategies to reduce childhood obesity among low-income children. According to CDC officials, policymakers and researchers should consider these factors when implementing similar strategies in the future. CDC officials or implementing grantees identified the following factors they observed across the CORD grantees: Staff turnover. CDC officials told us that the turnover of principals and other administrative personnel trained to provide the strategies is one factor that negatively affected the implementation of the strategies in schools or clinics. For example, CDC officials noted that in the Massachusetts demonstration project, researchers had to establish a relationship with a new principal of one of the participating schools when the other principal left, which delayed progress in implementation at that location. Similarly, the Arizona demonstration project also experienced staff turnover at the clinics, which led to a need for retraining and challenges in staff flows. CDC officials suggested that future research should consider incorporating staff retraining costs in the design of public health strategies to help mitigate this challenge. Family support. CDC officials told us that grantees had to provide more support than initially anticipated to families to better ensure their participation. CDC officials told us that grantees addressed this challenge by allowing siblings to also attend or participate in the activities or by holding activities on weekends or after school to accommodate parents’ work obligations. Strategies should be designed to be flexible for families, as there are competing demands on the families participating in the demonstration projects, CDC officials explained. Pertinent programs and policies. Implementing grantees noted that the preexistence of programs or policies that promoted healthy behaviors in the public health and primary care sectors positively affected their implementation of CORD strategies. For example, Massachusetts demonstration project officials told us that the strategies they implemented complemented an existing statewide program that promoted opportunities for healthy eating and active living in the communities, schools, childcare centers, and businesses. Grantee officials attributed the organizational commitment and motivation they observed in participating schools to these preexisting activities. Commitment from partner organizations. Implementing grantees found that the commitment of partner organizations, such as schools, was an important factor affecting implementation. According to implementing grantees, determining the willingness and ability of an organization to implement the strategies is important—by identifying, for example, leaders who support the strategies and can help ensure staff commitment to execute them. The Massachusetts demonstration project reported that 90 percent of the stakeholders they worked with noted the presence of leadership and administrative support for the project reduced feelings of conflict between program implementation and other priorities. Alternatively, the California demonstration project identified the lack of a strong supporter in a leadership position as a barrier to implementation. Parental stresses. Implementing grantees found that parental stresses related to social economic status (e.g., food insecurity or accessibility challenges, including transportation to intervention sites) was a major factor negatively affecting family participation and the implementation of the strategies. Grantees explained that understanding the effect of these stresses on a family’s ability to focus on the strategies to reduce childhood obesity is important and grantees should plan for ways to mitigate those stresses. CDC told us that the CORD phase 3 demonstration projects may be able to help mitigate some of the challenges identified from prior CORD experiences, as noted above. For example, CDC officials told us they plan to work with CORD phase 3 grantees to find ways to mitigate challenges associated with staff turnover, which could include taping trainings or allowing for virtual opportunities for retraining. Additionally, the CORD phase 3 grants are implementing pediatric weight management interventions in different settings—some in clinical settings and some in community settings—which CDC officials said may provide parents with additional flexibility to participate in the strategies. CDC and Others Have Taken Steps to Disseminate CORD Results and Continue to Promote the Use of CORD Strategies in Low-Income Communities CDC has taken steps to share CORD phases 1 and 2 design materials and available results with researchers and others. For example, CDC shared on its website information for CORD phases 1 and 2, including project summaries, background information about the grantees, and published literature describing the project designs and results. In addition, CDC shared lessons learned about the CORD Project and evidence-based childhood weight management programs during a series of webinars. According to CDC officials, the intended audience for the webinars included public health practitioners and researchers; local, state, and federal government agency officials; health care professionals; policy analysts; and community health workers. CDC officials also told us they have presented CORD results and lessons learned at conferences and at meetings organized by other HHS agencies. Specifically, a CDC official and grantees summarized results from the first phase of the CORD Project at the American Academy of Pediatrics’ Annual Conference in 2016. CORD phase 1 results were also presented at the 2018 Annual Meeting for the Association of State Public Health Nutritionists. Additionally, in November 2017, CORD phase 1 grantees met with researchers from the National Institutes of Health’s Childhood Obesity Prevention and Treatment Research program to share lessons learned from their respective research. To further disseminate CORD results, CDC officials highlighted their planned report to Congress, as required by CHIPRA, which was subsequently issued in September 2019. The report describes the findings for CORD phase 1 and provides brief descriptions of the CORD phase 2 grantees and their demonstration projects, since the results of those projects are not yet available. The report identifies CORD phase 1 findings, including information about the costs of implementing the strategies. CDC officials noted that the implemented public health strategies, such as providing classroom-based nutrition and gardening curriculum or programs that promote physical activity, cost less than primary care strategies. Specifically, CDC reported that the costs of public health strategies in early care and education centers ranged from $26 to $96 per child, the costs of some primary care strategies ranged from $164 to $181 per child, and the cost of more intensive family-based weight management programs ranged from $2,107 to $2,220 per child. CDC, in collaboration with other HHS agencies, has also taken some steps to promote the wider adoption of CORD strategies in low-income communities. For example, CDC and CMS have had preliminary discussions about how CMS could help CORD grantees understand how Medicaid and CHIP programs could reimburse for the obesity-related strategies they are implementing as part of the CORD Project, which CDC officials told us could help to sustain and expand these strategies to other low-income communities. CMS officials told us they are considering whether to issue guidance to state Medicaid and CHIP programs that explains how some states have been able to reimburse entities for the provision of overweight- and obesity-related services. CDC officials told us that after discussions with CMS officials, they provided information to CMS in October 2018 that could be used for a possible CMS information bulletin to state Medicaid and CHIP officials on childhood obesity. In addition, CDC and the Office of the Assistant Secretary for Planning and Evaluation within HHS have awarded a cooperative agreement to the National Association of Community Health Centers to increase the implementation of an evidence-based childhood weight management program—Mind, Exercise, Nutrition, Do It!—by federally qualified health centers. According to HHS officials, the National Association of Community Health Centers is assisting 14 federally qualified health centers in five states (Arizona, Florida, Illinois, Mississippi, and North Carolina) to implement this intervention and, based on lessons learned, plans to develop an implementation guide to support the expansion of this strategy to other health centers. CDC officials also told us they are coordinating with the National Cancer Institute within the National Institutes of Health to share knowledge with CORD phase 3 grantees about how to develop business models to support the expansion of successful strategies, which aligns with one of CDC’s goals for CORD phase 3 to determine how to increase the adoption of successful strategies beyond the CORD intervention sites. CDC officials and implementing grantees provided us some examples of CORD strategies and materials that continue to be used in the states where they were implemented or have been implemented in other low- income communities. Officials from the Massachusetts demonstration project told us that some of the primary care strategies they developed during CORD phase 1 are still provided in the healthy weight clinics that participated in the project. Officials from the Arizona demonstration project told us they have received funding from the U.S. Department of Agriculture to develop a new training module for health care providers interested in implementing the project’s pediatric weight management intervention. They explained that the new training module will include information on parenting strategies specific to child health behaviors (e.g., monitoring of physical activity) and examples of stories from the families who participated in the Arizona demonstration project. CDC officials also told us that materials that CORD grantees used as part of their strategies are publically available for use by researchers and other communities. These materials include a primary care resource guide developed in collaboration with the American Academy of Pediatrics, the Coordinated Approach to Child Health early childhood kit, and a healthy weight clinic implementation guide. Agency Comments We provided a draft of this report to HHS for comment. HHS provided technical comments, which we incorporated as appropriate. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the Secretary of Health and Human Services, the appropriate congressional committees, and other interested parties. In addition, this report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact me at (202) 512-7114 or dickenj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Major contributors to this report are listed in appendix II. Appendix I: Childhood Obesity Research Demonstration Grantees The Children’s Health Insurance Program Reauthorization Act of 2009 (CHIPRA) authorized the Department of Health and Human Services (HHS) to establish the Childhood Obesity Research Demonstration (CORD) Project. CHIPRA specified that HHS provide project grants to universities or other eligible entities to implement activities to reduce childhood obesity among low-income children. HHS designated the Centers for Disease Control and Prevention (CDC) as the agency responsible for designing, awarding, and managing the grants. Subsequent laws provided additional funding and extended the CORD Project for two more phases. The first phase of the CORD Project began in September 2011 and was completed in September 2016. The purpose of CORD phase 1 was to determine whether implementing strategies in public health sectors, including early care and education centers, schools and community organizations, and primary care sectors, such as health care clinics, could improve low-income children’s risk factors for obesity. CDC funded three implementing grantees: San Diego State University, the Massachusetts State Department of Public Health, and the University of Texas Health Science. CORD phase 2 started in June 2016. The purpose of this phase was to further test if strategies implemented in the primary care sector would reduce the body mass index (BMI) in children with obesity, or who were overweight with risks including medical and behavioral risks and family history. CDC funded the following grantees: the Massachusetts State Department of Public Health and Arizona State University. As of July 2019, CORD phase 2 was ongoing. Overview The California demonstration project was led by San Diego State University. The demonstration project was implemented in three rural communities—Brawley, El Centro, and Calexico—in Imperial County, California. Imperial County, California, is located on the U.S.-Mexico border, and had an estimated 174,528 residents, 77 percent of whom were of Mexican origin—including 32 percent who were foreign born—in 2010. Three-quarters of all residents reported speaking a language other than English at home. The median household income was $39,402, compared to $61,632 in the state, and income disparities are reflected further in the differential poverty rates (23 percent in Imperial County versus 14 percent in California as a whole). Early care and education centers (23 centers) Collected height and weight for children aged 2-5 years. Trained center staff on health behavior change strategies to use at their centers. Provided centers with large self-serve water containers and cooking kits with child-friendly cooking and serving items. Elementary schools (13 schools) Worked with school nurses and trainees to collect BMI measurements from kindergarteners, third graders, and fifth graders in the El Centro Elementary School District; and kindergarteners, second graders and fifth graders in the Brawley Elementary School District. Provided schools physical activity equipment. Installed water jets and other water containers to provide self-serving access by students. Developed lesson plans promoting sleep for grades kindergarten through sixth grade. Community (three community organizations and three independent restaurants) Provided a water dispenser at two community recreations centers in Brawley and El Centro and at one Boys and Girls Club in Brawley. Developed community gardens at the Boys and Girls Club in Brawley and a recreation center in El Centro. Demonstration Project Introduced healthy children’s menu items in three restaurants. The grantee conducted a non- randomized study which sought to determine whether strategies implemented in both public health sectors and primary care sectors would be more effective at preventing and controlling childhood obesity when compared with strategies implemented in public health sectors only, primary care sectors only, or when strategies were not implemented. Community health clinics (three clinic sites) Modified the clinics’ electronic health record systems to improve health care provider screening and treatment of childhood obesity including through the use of alerts and prompts. To facilitate the adoption of the system changes, a patient care coordinator was hired to work across the participating clinics. Hired community health workers and a community health worker coordinator to administer the Family Wellness Program, a 12-month program that delivered wellness and physical activity workshops, motivational interviewing, and newsletters. Overview The Massachusetts demonstration project was a led by the Massachusetts Department of Public Health. The demonstration project was implemented in the cities of Fitchburg, located in north-central Massachusetts, and New Bedford, in southeast Massachusetts. In 2010, the population of these two cities was about 40,000 and 95,000, respectively, and was predominantly non-Hispanic white (about 68 percent). Both communities had higher percentages of low-income residents than the state of Massachusetts, according to 5-year estimates from the 2008-2012 American Community Survey. Specifically, the percentage of families with children whose incomes were less than the federal poverty level was about 24 percent in Fitchburg and 27 percent in New Bedford versus 12 percent in the state. Early care and education centers (nine centers) Trained mentors to provide support to staff to implement evidence-based programs on nutrition and physical activity. Schools and after school programs (six schools and 17 after school programs) Provided evidence-based health education curricula and training to teachers to encourage student learning about nutrition and physical activity. Implemented a nutrition curriculum for after-school program staff to use with children aged 5 to12 years. Community Implemented a communications campaign, including text messaging, small billboards, transit ads, and handouts, to spread the demonstration project’s brand and to change community norms and practices in physical activity and healthy eating. Special Supplemental Nutrition Program for Women, Infants and Children (one program in each community) Collaborated with the Special Supplemental Nutrition Program for Women, Infants and Children to implement intervention activities including training nutritionists and nutrition assistants in best practices on assessment and counseling for childhood obesity prevention and developing an obesity counseling toolkit for providers. Demonstration Project The grantee used a combination of pre- post time series and quasi- experimental designs to examine the extent to which the interventions resulted in changes in BMI, individual- level lifestyle behaviors, satisfaction with health care services, and quality of life among children, as well as to health policies, programs, and environments in the two intervention cities compared to another city. Health centers (two centers) Modified existing electronic health records to deploy a computerized, point-of-care decision support alert at the time of a well-child care visit for a child who is overweight or has obesity. The alert prompted clinicians to document weight status, nutrition and physical activity counseling, and place referral to the on-site healthy weight clinic for weight management support. Implemented a healthy weight clinic in each participating health centers. Each healthy weight clinic was staffed with a physician, a nutritionist, and a community health worker who met with each patient and family. Patients participating in the healthy weight clinics engaged in dietary and physical activity assessment, goal setting, and were connected to community resources to support healthy lifestyles. Overview The Texas demonstration project was led by the University of Texas Health Science Center in Houston. The demonstration was implemented in two catchment areas in Houston and Austin, Texas. The data collected at the beginning of the project from participating early care and education centers, schools, and clinics indicated that families were low-income, with most parents reporting an annual household income of $25,000 or less. The population was predominantly Hispanic (73 to 83 percent), with approximately 44 to 55 percent predominately Spanish-speaking. Early care and education centers (28 centers) Provided classroom materials on nutrition and gardening and bilingual parent tips sheets on nutrition, activity, and screen time. Provided physical activity equipment to participating centers. Schools (40 schools) Trained school staff on a nutrition and physical education classroom curricula. Sent text messages in English or Spanish to participants once a week that emphasized program concepts and linked families to resources. Community Provided training sessions to teach community health workers, teachers, parents, physicians, and others stakeholders about advocacy and the implementation of environmental changes for healthy eating and active living. Demonstration Project Health care clinics (11 clinics) Provided BMI screening for children who are overweight or have obesity, which included decision supports to integrate guidelines for the appropriate clinical screening, evaluation and treatment into day-to-day practice. Modified electronic health records to identify children who were overweight and had obesity, provide prompts for treatment, and provide clinicians with access to referral information for weight management. The Texas demonstration project implemented and evaluated a primary and secondary obesity prevention program. In the primary prevention intervention, the grantee collected data on risk factors and the utilization of health care services and community programs. This intervention was focused on the entire community, with the goal of preventing the development of obesity. The secondary prevention program consisted of a randomized control trial, targeted to children who were already overweight or had obesity. Children and their families were randomly assigned to either a community centered or a primary care centered weight management program. The Arizona demonstration project is led by Arizona State University. The purpose of the project is to implement an adapted program in three pediatric primary care clinics located in Maricopa County, Arizona. These clinics serve a minority patient demographic of about 60 to 65 percent, of which the largest groups are Mexican American and American Indian. Program adaptation: Adaptation began by assessing the needs and capacity of a primary care organization and the families they serve. The program was then pilot-tested in a general pediatrics clinic and a clinic for children with advanced obesity to determine feasible delivery modifications as well as enhanced content for obesity management and prevention. During and at the end of the pilot trial, feedback was solicited from stakeholders and families. A draft of the adapted version of the program was then developed, additional feedback was sought from experts and stakeholders and a second pilot-testing phase was completed. Feedback was again collected from families who received the intervention and from stakeholders who participated in the pilot. The intervention protocol and content were further refined to implement in the three pediatric primary care clinics. Demonstration Project The Arizona demonstration project implemented an adapted program that was designed to target health behavior change in children ages 5 and one half to 12 years by improving family management practices and parenting skills, with the goal of preventing obesity and excess weight gain. The program is designed to tailor services based on a family assessment and to increase parent motivation. The project included a randomized control trial to evaluate the effectiveness of the adapted program within three primary care clinics in two federally qualified health centers and a children’s hospital. Effectiveness study: Participants were identified during clinic well- and sick-child visits and through queries of electronic health records. After completing a family health assessment, families were randomly assigned to the adapted program or services as usual. Participating families completed routine assessments about family health behaviors, child health behaviors, family well-being and support, and other topics. Following the assessments, feedback sessions were initiated. The first feedback session focused on understanding (a) the caregivers’ perception of their needs; (b) their child’s health, adjustment, and behavior; and (c) the caregivers’ motivation to change parenting and family management practices in support of health behavior change. Additionally, over a 6-month period, families participated in eight to 16 parenting sessions tailored to the specific needs identified in the family health routine assessment and focused on a specific behavior change goal, such as setting limits on snacking between family meals or monitoring children’s sedentary and physical activity time. In the second and third feedback sessions, the coordinator began by checking in with the family about their progress, discussing barriers they experienced, and exploring the ways that the previous feedback and parenting sessions were helpful for them in catalyzing and supporting healthy lifestyle behavior change. Additionally, coordinators provided families with referrals to existing resources in the community. In weeks where a face-to-face session was not scheduled or did not occur, the coordinator conducted a 15- to 30-minute phone-based coaching session. The purpose was to maintain contact with the family and help problem- solve challenges, reinforce positive achievements, and continually address motivation to change and barriers to engagement. Overview The Massachusetts demonstration project for CORD phase 2 is led by the Massachusetts State Department of Public Health and Massachusetts General Hospital. The demonstration project was implemented in the cities of Holyoke and New Bedford, Massachusetts. Primary care screening and assessment of child BMI: Children were referred to the demonstration project by their primary care provider during a health care visit where a height and weight was obtained and it was determined that the child was overweight or had obesity. After the referral was made, parents were mailed an introductory letter and fact sheet by the study team. A bilingual study coordinator contacted parents by phone and explained that the research study was to examine strategies to improve the care that is provided for children who require weight management. The coordinator obtained verbal informed consent from the parent and administered a 20-minute baseline survey. Demonstration Project The demonstration project implemented a randomized trial that compares the effects of a pediatric weight management program delivered in the healthy weight clinics of two federally qualified health centers with a weight management program delivered at two YMCAs. Eligible children were overweight or had obesity, ages 6 to 12 years, and received primary care at the two federally qualified health centers. Child assigned to intervention: After the parent completed the survey, the child was randomly assigned to a healthy weight clinic in one of the two federally qualified health centers or to the weight management program delivered at one of the two YMCAs. Each of the two intervention groups received an intensive 6- month intervention, followed by a 6-month maintenance period that delivered 30 or more hours of contact time over one year. In addition, children in both intervention groups were exposed to quality of care improvements in their federally qualified health centers, which included primary care provider weight management training and text messages to participating families for self-guided behavior change support. Healthy weight clinic: This intervention was clinic-based and used a multidisciplinary team, including, a pediatrician, community health worker, dietician, and access to behavioral/mental health providers, as needed. The team was trained to deliver motivational interviewing and behavioral modification techniques to engage families in setting and following through on healthy eating and activity goals. Visits alternated between group visits with other children and families in the program and individual visits for the first 6 months and individual visits in the second 6 months. During the first 6 months of the intervention, the community health worker or dietitian made bi-weekly phone calls to the family on weeks they did not have an in person visit. During the second 6 months, they provided once-monthly calls. YMCA weight management program: This intervention was a community- based intervention where staff at two local YMCAs were trained to implement the program. Two YMCA group leaders provided support, education and activities during sessions, which included goal setting and action planning, a parent discussion, and 60 minutes of physical activity for the children. The program was delivered over 12 months, which included 16 weekly sessions, followed by four sessions delivered every other week and concluded with five monthly sessions. Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Shannon Slawter Legeer (Assistant Director), Deitra H. Lee (Analyst-in-Charge), and Kristen M. Pinnock made key contributions to this report. Also contributing were Krister Friday, Richard Lipinski, Laurie Pachter, Ethiene Salgado- Rodriguez, and Emily Wilson Schwark.

Why GAO Did This Study Childhood obesity affects nearly 14 million children aged 2 to 19 years in the United States. Children in low-income families are disproportionately affected, with about 1 in 5 having obesity. Studies suggest that children with obesity are likely to become adults who are overweight or have obesity, which can contribute to poorer health and higher health care expenditures. CDC was designated as the agency to design and manage the project and has awarded grants in three separate phases. GAO was asked to examine the CORD Project, including what has been learned regarding strategies to reduce childhood obesity. In this report, GAO describes 1) the extent to which CDC changed the design of the CORD Project between grant phases, 2) the results of the CORD Project and factors that have affected implementation, and 3) efforts by CDC and others to disseminate results and lessons learned. To conduct this work, GAO reviewed planning and grant documentation for the three CORD phases, published articles about the design of CORD phase 1 and 2, and documentation describing the results of CORD phase 1. GAO also interviewed CDC officials, CORD phase 1 and 2 grantees, and officials from other HHS agencies involved in the design of the CORD Project. HHS provided technical comments on a draft of this report, which GAO incorporated as appropriate. What GAO Found The Centers for Disease Control and Prevention (CDC) has made four key changes to the design of the Childhood Obesity Research Demonstration (CORD) Project between each of the three grant phases. Established by law in 2009, the project provides research grants to develop and implement strategies to reduce obesity among low-income children. One of CDC's design changes, for example, was to modify the scope of the project (i.e., type of strategies implemented by grantees). After CORD phase 1, CDC officials shifted the scope from prevention—through the implementation of strategies in community settings, such as schools, and in health care settings—to the treatment of children who were overweight or had obesity. According to CDC officials, the agency made this change due to the shorter time frame for implementing CORD phase 2 and in response to existing national recommendations related to childhood obesity. CDC also changed the purpose of the project's study design prior to phase 3. Whereas CORD phases 1 and 2 were intended to build knowledge and evidence of effective strategies, CDC modified CORD phase 3 to focus on translating effective strategies into routine use by converting them into a package of materials that others could replicate. To evaluate the effectiveness of CORD phase 1—the only phase that is complete—CDC awarded a grant to an independent entity to aggregate results across the three grantees, and each grantee conducted their own evaluation. The evaluation center and the grantees reported some improvements in children who received CORD 1 strategies. For example, the evaluation center reported small but positive changes in outcomes measured, which included body mass index and fruit and vegetable consumption. These improvements were most often observed among children who received primary care strategies, such as individualized counseling, and children who participated in public health strategies, such as an evidence-based nutritional program, in addition to the primary care strategies. CDC and grantees identified several factors during the first two phases that affected the ability to implement strategies to reduce obesity among low-income children. For example, grantees noted that the preexistence of programs and policies that promoted healthy behaviors positively affected their implementation of CORD strategies. CDC officials identified the turnover of principals and other school or clinic staff as negatively affecting the implementation and suggested that future researchers incorporate staff retraining costs into their strategies as a way to help mitigate this challenge. CDC has taken steps to share CORD design materials and results through published literature, websites, and conferences. It has also coordinated with other Department of Health and Human Services (HHS) offices and agencies to promote the wider adoption of CORD strategies in low-income communities. For example, CDC has collaborated with an office in HHS to fund a project to increase the use of a specific weight management program used in CORD phase 1.

gao_GAO-20-228

gao_GAO-20-228_0

Background Federal law requires executive agencies to: maintain adequate inventory controls and accountability systems for property under their control; continuously survey property under their control to identify excess; and promptly report excess property to GSA and generally dispose of it in accordance with GSA regulations so that it can be made available to other federal agencies and stakeholders for reuse. In addition, the FPPMA requires agencies to assess accountable property within their control in accordance with guidance from GSA. Property can be accountable or non-accountable. Accountable property is property with a useful life of at least 2 years that an agency determines should be tracked in its property records. Each agency determines what constitutes accountable property for that agency, and for our selected agencies, consideration is given to an item’s acquisition cost and other factors, such as ease of theft and sensitivity. For our three selected agencies, these acquisition cost thresholds ranged from $5,000 to $10,000 in fiscal year 2018. Non-accountable property is property that does not meet the agency’s definition of accountable property and may include items such as office furniture and printers. Agencies typically do not track non-accountable property unless they need to for specific purposes, such as managing inventory levels. Some agency property is located in warehouses. For the purpose of this report, we used the definition of “warehouse” in FRPP guidance: “buildings used for storage, such as ammunition storage, covered sheds, and buildings primarily used for storage of vehicles or materials.” This term encompasses a broad array of property that agencies may classify differently for internal purposes but classify as warehouses for FRPP reporting in the absence of more precise categories. For example, some buildings that DOE reports as warehouses in FRPP are specialized facilities for storing nuclear and nuclear-contaminated material. In the absence of an FRPP category for nuclear storage, these buildings are classified in FRPP as warehouses. In a similar manner, FAA classified as warehouses in FRPP buildings used to house air traffic support systems, such as approach lighting systems, because no other category in FRPP was a better fit. GSA’s role in agencies’ management of property they have acquired, whether in warehouses or elsewhere, is limited until an agency declares that property as excess. Once property is declared excess, it can be transferred to another agency or certain non-federal recipients, donated, sold, abandoned, or destroyed. GSA has issued regulations that govern agencies’ actions in the property disposal process, and it administers a web-based system that facilitates property disposal. However, prior to an agencies’ identifying property as excess, GSA’s authority to issue regulations or formal guidance regarding agencies’ management of property is limited to topics that have been specifically assigned to the GSA Administrator, according to GSA officials. GSA distinguishes between formal guidance and informal guidance, and GSA officials explained this distinction as follows. Formal guidance, such as a Federal Management Regulation bulletin, must be reviewed by GSA management and general counsel officials. For example, in 2017 GSA issued a Federal Management Regulation bulletin on warehousing that summarized industry perspectives from two voluntary consensus standards, which were published by ASTM International’s asset management committee and which GSA had participated in developing. One standard addressed storage of property and the other addressed strategic warehousing. GSA officials stated they were authorized to include content on property management in the bulletin under the authority of GSA’s real-property policy program because the content supported the real property goal of reducing the federal government’s real property footprint. In contrast, informal guidance does not require review by GSA management or general counsel officials but may be published on GSA’s website or disseminated at GSA trainings. Whether formal or informal, agencies are not required to adhere to such guidance. Typically, within agencies, responsibility for managing property is generally shared between property officials and property custodians. Property officials’ primary responsibilities relate to property management. For example, they may be responsible for updating property data systems, providing property lists and instructions for property inventories, resolving issues that arise with property management, and managing the disposal process after a property custodian has determined that an item is no longer needed. Property custodians are generally program managers who are assigned responsibility for specific property items associated with the program they manage as an ancillary duty. For example, property custodians may be required to conduct physical inventories of property assigned to their program and work with a property official to resolve any issues arising during the inventory. Depending on how an agency uses its warehouses and property, property custodians may be responsible for property in a single warehouse, in multiple warehouses, or in a variety of locations. Moreover, multiple property custodians may be responsible for property in a single warehouse, as depicted in figure 1. Selected Agencies Stored Various Types of Property in Warehouses but Had Limited Information about Storage Costs Selected Agencies Stored a Variety of Property in Warehouses The three selected agencies had a total of 1,221 warehouses, with over 6.4-million square feet, that contained a broad array of property. Although comprehensive data on property in these agencies’ warehouses were unavailable, interviews, site visits, and agencies’ data on warehouses themselves provided some information on the types of property in them. We found that the agencies had some commonalities in the contents of their warehouses. For example, all three had warehouses that contained material-handling equipment, such as fork lifts, as well as excess property being processed for disposal. However, much of the property in agencies’ warehouses was specific to their missions, according to agency officials and our observations. Table 1 includes information about the agencies’ warehouses and examples of the types of agency-specific property in them. We also visited warehouses at each of the selected agencies to obtain additional information about and view the types of property stored within them, as described below. FAA. FAA had warehouses at four main sites that contained property specific to the sites’ missions, and most of the remaining warehouses were buildings that contained equipment, tools, or materials to maintain aviation support systems or housed support systems, such as approach lighting systems, according to our analysis of FAA warehouse data, FAA officials, and sites we visited. For example, we visited FAA’s warehouses at the Mike Monroney Aeronautical Center, including the Logistics Support Facility, FAA’s largest warehouse and central location for maintaining and repairing aviation support systems deployed throughout the national airspace system. Most items in the warehouse were spare parts, materials, and systems or system components that had been sent to the facility for repair. We also visited FAA’s Staging Area, which supports FAA’s manufacture and assembly of new systems to be deployed throughout the country. Accordingly, much of the property at the two warehouses that comprise this facility was equipment, parts, and material, along with the machines and tools to manufacture and assemble the material. For example, we viewed components of a wind shear alert system that were being prepared to be shipped. FAA’s Mobile Asset Deployment Center stored and maintained FAA’s mobile assets, such as air traffic control towers and housing units that FAA deploys to maintain service during disruptions such as natural disasters. Finally, we visited a 96-square-foot shack—identified in FRPP as a warehouse and pictured below—that housed an approach lighting system. (See fig. 2 for examples of FAA warehouses and property.) Office of Science. Most Office of Science warehouses were located at Office of Science national laboratories. Warehouses at the two national laboratories we visited—Argonne National Laboratory and Fermi National Accelerator Laboratory—contained a broad variety of equipment, including equipment being staged for near-term use and equipment in longer-term storage specifically designated for future projects. For example, one warehouse at Fermi National Accelerator Laboratory contained a cryogenic system acquired by CERN, the European Organization for Nuclear Research, as its contribution to a planned experiment. This cryogenic system will be used for cooling purposes. A warehouse at the same site also contained some decades-old items kept as replacements for items still in use. According to officials, many of these older items would be difficult to obtain in a reasonable time frame for a reasonable price if a replacement were needed. In addition, at Fermi National Accelerator Laboratory, we saw a large, out-of-use calorimeter— a device commonly used in physics experiments—that was being stored for eventual use in an educational display. Warehouses also contained parts, materials, and supplies for laboratory use. (See fig. 3 for examples of Office of Science warehouses and property.) BOP. Most of the BOP warehouses were located at correctional institutions throughout the country, served similar functions, and contained similar types of property for inmate use, according to BOP headquarters officials and our review of BOP real property data. The two correctional institutions we visited each had a warehouse that served as a distribution center, where items arriving at the institution were received, processed, and sent to the appropriate personnel within the institution, and a food service warehouse, where food items used to feed the inmate population were stored. At one institution, non-perishable items for inmate use, such as uniforms, mattresses, soap, and toilet paper were stored at the distribution center, while the other institution we visited stored less property at the distribution center and expedited delivery to the relevant division. Additionally, one institution used a warehouse to store dairy equipment in support of an inmate-run dairy. (See fig. 4 for examples of BOP warehouses and property.) Selected Agencies Tracked Warehouse Costs but Lacked Information on Property Storage Costs All three selected agencies tracked certain direct costs for owned and leased warehouses, including operations and maintenance costs for owned warehouses and some leased warehouses, and the rental cost for leased warehouses (see table 2). Although the agencies had this cost information, they did not use it to systematically determine how much it costs to store their property in warehouses, whether at an aggregate or per-item level. Two features of how these agencies track property and warehouse costs would make it difficult to do so. First, as mentioned above, selected agencies did not have comprehensive information on items in warehouses, information that would be needed to determine per-item storage costs. Second, selected agencies generally incurred direct costs—rent, operations costs, and maintenance costs—at a warehouse level. However, because a warehouse may have had some of its square footage dedicated to other uses, such as office or laboratory space, it would be difficult to ascertain what percentage of costs would be allocated to storage versus these other uses. Moreover, in some cases, operations costs, such as utilities, were incurred at a multi-building level, making it difficult to determine what portion of the bill is attributable to a single warehouse. Finally, selected agencies generally did not track indirect costs, such as personnel costs for conducting regular inventories and other administrative costs associated with storing property in their warehouses, according to agency officials. While none of the selected agencies systematically tracked property storage costs, we did identify one Office of Science site, one Department of Transportation site, and one Department of Justice site that analyzed the use of specific portions of warehouses for cost allocation purposes. Officials at these agencies said that this approach may create incentives for property custodians to identify excess property in a timelier manner. Argonne National Laboratory, within the DOE’s Office of Science, annually analyzes direct costs for each building, including warehouses, and charges each division within the laboratory for the space it occupies. A report assessing contractor performance at DOE’s Fermi National Accelerator Laboratory noted that implementation of such a system could be an effective way to hold divisions accountable for the number of items they have in storage. The Department of Transportation and the Department of Justice each manage a warehouse near their respective headquarters that they use to store property for various divisions within each department. The departments charge users for the portions of the warehouses they occupy. While these approaches may create incentives to identify unneeded property in a timely manner, they may not be applicable for all circumstances. For example, staff at Fermi National Accelerator Laboratory stated that they explored the cost and benefits of analyzing space use to allocate costs by user but had not found it to be cost- effective or feasible. In addition, allocating costs based on warehouse usage would be challenging if users’ space usage changes regularly. Without Guidance, Selected Agencies Did Not Systematically Assess the Ongoing Need for Property in Their Warehouses Two of the Three Agencies Specified When to Identify Some Types of Unneeded Property, but None Speciied How to Assess Most Items for Ongoing Need Two of the three agencies we reviewed had policies in place explaining the frequency in which property custodians should assess property for ongoing need. Specifically, the Office of Science and BOP had policies that called for identifying unneeded property beyond the statutory requirement to continually survey property to identify excess. For example, DOE regulations, which cover the Office of Science, require managers to perform walkthroughs at least every 2 years to identify unneeded property. According to officials, these walkthroughs are conducted by contractors that manage national laboratories. Similarly, BOP policy requires that property custodians conduct an annual site inspection to identify unneeded property prior to the annual inventory, and, according to officials, this process is overseen by the institution’s associate warden. In contrast, FAA policy does not set any timeframe for property custodians to identify unneeded property. However, according to one FAA headquarters official, assessing property for ongoing need is inherent to the inventory process, which, according to FAA policy, should occur at least every 3 years for accountable property. In addition, only DOE had specific requirements to determine if property is needed. Specifically, DOE regulation requires written justification for retention of property classified as equipment held for future projects. If equipment is retained for longer than a year, the justification is to be reviewed by a higher level of authority, and retention of such equipment for longer than 3 years requires approval by the head of the DOE field organization. The Office of Science Organizational Property Management Officer—who is responsible for reviewing contractors that manage Office of Science sites—reviewed sites’ adherence to this requirement using metrics, such as acquisition date and time in storage, according to officials. Beyond this particular requirement for DOE, none of the agencies had a systematic way to identify property that may be unneeded. Instead, they primarily relied on professional judgment to determine the ongoing need for property in warehouses in the absence of guidance on how to determine whether property is still needed. For example, FAA officials confirmed that they do not have guidance or metrics on how to identify unneeded property and typically rely on property custodians’ professional judgment. According to officials at one FAA site we visited, property custodians do not use specific criteria for identifying unneeded property because it is obvious when items are no longer needed. Similarly, at the BOP institutions we visited, officials confirmed that they rely on property custodians’ professional judgment, along with the judgment of associate wardens, to identify unneeded property during the annual site inspections, but acknowledged that this has led to different outcomes. For example, at one site we visited, site officials stated that some associate wardens are more inclined than others to require property custodians to identify property as unneeded. While officials at all of the selected agencies said they believed property custodians were able to identify unneeded property in a timely manner using their professional judgement, we identified instances, through our interviews and agency assessments, where agencies had retained unneeded property in storage. While the agencies identified and in most cases addressed these instances, these situations demonstrate the challenges associated with agencies’ existing approaches. Specifically: A 2016 report from DOT’s inspector general found that FAA property custodians allowed obsolete computers to remain on the property records, including computer systems manufactured in 2006 or earlier that were likely no longer in use because of their 3- to 4-year lifecycles. In 2018, a review found that Fermi National Accelerator Laboratory’s contractor was storing IT equipment, which had not been classified as equipment held for future projects, dating back to 1998. The report recommended that the contractor review all IT equipment for continued need and that certain items be removed from the active inventory in their asset management system. BOP headquarters officials told us that, when assisting regional office personnel in training a new property official at an institution, they noticed the institution was storing inmate clothing that exceeded the institution’s needs. According to the officials, they worked with the new property official to transfer the clothing from the institution to another BOP institution that needed it. Stakeholders Identified Systematic Methods to Assess Property for Ongoing Need Selected agencies’ limited guidance on how to identify unneeded property and reliance on professional judgment were not unique to the agencies in our review. For example, in a previous review that examined five agencies—Environmental Protection Agency, Forest Service, GSA, Department of Housing and Urban Development, and Internal Revenue Service—we found that selected agencies did not have policies and processes for identifying unneeded property on a proactive basis and relied on “triggering events,” such as an office move to make excess property decisions. Moreover, the industry and standards-setting groups we interviewed for this review indicated that these approaches were common across the federal government. However, the industry stakeholders and federal agencies that participated in ICPM that we interviewed identified more systematic ways to identify unneeded property. For example: Periodic justification for continued storage. One agency implemented a policy in 2013 requiring written justification to retain certain accountable property for certain time periods, with the time period varying for different types of property. After the initial storage time period, written justification for continued storage must be reviewed and approved by an official who is above the property custodian. According to property officials, this policy has contributed to an estimated 35 to 40 percent reduction of property held in storage. Data analytics. Officials from another agency stated that they use a logistics management application to track and analyze information, such as property age, amount, rate of usage, and warehouse space availability. As a result, the agency has identified and disposed of excess property at various warehouses that otherwise would likely have been retained. For example, according to officials, analysis conducted using this application on idle property in one warehouse informed the decision to identify as unneeded a significant amount of furniture. A previous manager had acquired the furniture for use in staff housing, but the items were not well-suited to available housing in the area. Utilization reviews. Industry groups we interviewed advocated for increased use of data to assess utilization to inform decisions on whether to retain stored property, such as utilization reviews that systematically assess property utilization and continued need. For example, when conducting a utilization review, one stakeholder recommended a process that begins with pinpointing where the inactive population of property items reside. Upon locating anything that has been inactive for a certain period of time, within a certain storage area, those items are identified as candidates for disposal. After the results come in, the property custodians can recommend that a certain amount of items on the overall list are marked for disposal. Limited Guidance Exists for Property Management While some stakeholders identified systematic ways to identify unneeded property in certain circumstances, limited government-wide guidance exists for agencies to use to determine whether property in warehouses is still needed and being used. Specifically, there are two sources for guidance related to assessing property in warehouses for ongoing need: ASTM’s standards for strategic warehousing and storage of property. The standard for strategic warehousing notes that entities often continue a warehouse activity largely because it is easier than going through the effort of dismantling it. It urges that entities consider whether warehousing is needed. Furthermore, the standard asserts that a sound business case should be in place to support storage of property, including a decision of whether the items need to be warehoused. The standard for storage of property notes that entities should deploy an inventory management system to track incoming and outgoing assets; such a system can help in developing performance metrics for stored items. GSA’s federal warehousing bulletin. This bulletin references the two ASTM standards identified above and discusses the importance of critically assessing the need for items in storage, but provides limited information on how to make such assessments. According to GSA officials, the use of voluntary consensus standards, such as ASTM standards, can assist agencies with property management. However, only one agency official we interviewed stated that voluntary consensus standards informed the agency’s policy; the others we interviewed were either unaware of the standards or said the standards were not relevant to agency policy or practice. The FPPMA requires agencies, in accordance with GSA guidance, to inventory and assess property. As part of such assessments, it calls for evaluations of the age and condition of the property and the extent to which the agency uses it. According to officials at the selected agencies, they are waiting for guidance from GSA before taking steps to implement FPPMA. According to GSA officials, they are in the process of developing informal guidance on minimizing and identifying excess property to meet this requirement because FPPMA did not provide GSA additional authority to issue regulations or formal guidance. In particular, GSA developed draft guidance, which incorporated principles from a new ASTM standard on identifying and reducing excess property that GSA officials expect will be issued in early 2020, and provided it to ICPM participants for review and comment in September 2019. According to GSA officials, this informal guidance will be issued in December 2019. GSA officials plan to include the guidance on the GSA website and disseminate it to ICPM participants and may provide it in hard copy at relevant GSA events. The draft guidance we reviewed encourages agencies to designate an individual to manage an agency’s asset management program and use that system to capture and provide information on property age, condition, utilization, and mission dependency on a real-time basis, among other things. The draft guidance also included some criteria agencies could use to identify excess property. However, the guidance did not provide specific approaches or practices agencies could use to assess property utilization, including property stored in warehouses. The draft guidance and an accompanying strategy document indicate that GSA will collect best practices and incorporate them into the guidance, but GSA officials did not specify what types of best practices it plans to include or provide a timeline for doing so. Including additional information on approaches or practices agencies can use to assess property use and ongoing need—such as periodic property justifications, data analytics, and utilization reviews—could assist agencies in fulfilling their FPPMA requirements. Moreover, GSA officials did not provide a documented plan or time frame for communicating the guidance beyond publishing it on GSA’s website and disseminating it to ICPM members, an approach that can limit the reach and awareness of this information to agencies government-wide. As we have previously reported, work by others has shown that inaction on unneeded government property can limit its efficient use. Conclusions As agencies continue efforts to manage their warehouse space in accordance with government-wide initiatives, improvements to how agencies assess property utilization and identify unneeded property in warehouses could enhance these efforts. The agencies in our review did not systematically assess their property for ongoing need and in some cases, retained unneeded property. More broadly, agencies across the government are operating without the benefit of government-wide guidance that could help assess their property for ongoing need in a systematic manner. With the recent enactment of FPPMA, an opportunity exists for GSA to develop and communicate guidance to help agencies assess property utilization and identify unneeded property in warehouses more efficiently that includes practices GSA identifies as being useful. Such guidance could help agencies avoid retaining property that is no longer needed and, as a result, allow them to better manage the use of their warehouse space. Recommendation The Administrator of GSA should direct the Office of Government-wide Policy (1) to incorporate into its guidance approaches or practices that agencies could use to assess utilization of and the ongoing need for property—approaches such as recommendations for periodic justifications, data analytics, and utilization reviews—and (2) to develop a plan and timelines for communicating the guidance to agencies government-wide. (Recommendation 1) Agency Comments We provided a draft of this report to GSA, DOT, DOE, and DOJ for review and comment. GSA concurred with our recommendation and provided written comments, which are reprinted in appendix II and summarized below. DOT, DOE, and DOJ each stated in an email that they had no comments on the draft report. In its written comments, GSA agreed with our recommendation and stated that it is further developing its guidance as well as a plan and timeline for dissemination of that guidance to executive agencies. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to the GSA Administrator, the Secretary of Energy, the Attorney General, and the Secretary of Transportation, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions regarding this report, please contact Lori Rectanus at (202) 512-2834 or rectanusl@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix III. Appendix I: Objectives, Scope, and Methodology This report addresses: (1) what is known about property in selected federal agencies’ warehouses and how much they spend to store this property and (2) the extent to which selected agencies assess the ongoing need for property stored in warehouses. To address both objectives, we selected three agencies for analysis—the Federal Aviation Administration (FAA) within the Department of Transportation; the Office of Science within the Department of Energy (DOE); and the Bureau of Prisons (BOP) within the Department of Justice. We limited our scope to civilian agencies because we have already done extensive work on property management within Department of Defense. At the department level, we used Reduce the Footprint data from fiscal year 2017 because they were the most current data available when we conducted the analysis to identify the top 10 departments in terms of warehouse square footage. To obtain variation among these agencies, we categorized these departments as large, medium, or small in terms of warehouse square footage and selected one from each category based on changes in square footage between fiscal years 2015 and 2017 using fiscal year 2017 Reduce the Footprint data and on the proportion of leased warehouse space to owned warehouse space using fiscal year 2017 Federal Real Property Profile (FRPP) data. Because none of the selected agencies manages property at the department level, we then selected a component within each department. For the Department of Transportation and the Department of Justice, we selected the components with the most warehouse square footage according to fiscal year 2017 FRPP data—FAA and BOP, respectively. For DOE, we used the agency’s fiscal year 2017 real property data to identify the components with the most warehouse square footage because DOE reports most information to FRPP at the department level rather than for specific offices, such as the Office of Science. We then selected the Office of Science, which had third highest amount of warehouse square footage, because of security concerns with one of the components with more warehouse square footage and because the other component with more warehouse square footage used a greater proportion of warehouse space to store nuclear and nuclear-related material. To determine what is known about property in selected agencies’ warehouses, we interviewed headquarters-level officials regarding the agencies’ property data, conducted site visits to view and photograph property stored in warehouses, and gathered information in interviews with agency officials. In selecting sites, we selected at least one site per agency that was among the largest in terms of warehouse square feet for that agency and at least one other site that was near one of the large sites, as described below: FAA: Mike Monroney Aeronautical Center in Oklahoma City, Oklahoma; Staging Area and Mobile Asset Deployment Center in Independence, Missouri; and Charles B. Wheeler Downtown airport in Kansas City, Missouri. Office of Science: Argonne National Laboratory and Fermi National Accelerator Laboratory in the Chicago area. BOP: U.S. Penitentiary Leavenworth in Leavenworth, Kansas, and Federal Correctional Institute El Reno in El Reno, Oklahoma. For each agency, we also interviewed officials at the headquarters and regional levels. Information obtained from these sites and regional officials is not generalizable to the selected agencies, and information from these agencies is not generalizable to other agencies. To determine how much selected agencies spend to store property in warehouses, as well as the numbers and square footage of these warehouses, we analyzed FRPP data from fiscal year 2018 for FAA and BOP, and DOE fiscal year 2018 real property data for the Office of Science because DOE reported most data to FRPP at the department level; both sources included information about direct costs such as rent, operations, and maintenance costs. We used FRPP data from fiscal year 2018 because that was the most recent data available when we conducted our analysis and DOE data covering the same period to be consistent. We reviewed documentation related to these data sources, interviewed knowledgeable officials, and determined that these data were sufficiently reliable for providing information about warehouse numbers, square footage, and the costs listed above. To determine the extent to which selected agencies assess the ongoing need for property stored in warehouses, we reviewed statutes, regulations, GSA guidance, our prior work, reports by federal agencies’ Offices of Inspector General, and relevant industry standards related to property storage and warehousing practices. In addition, for selected agencies we analyzed property policies and procedures for identifying and disposing of unneeded property and interviewed headquarters, regional, and site officials. We also interviewed three industry stakeholders—two property-management and one standards-setting organization—to discuss property storage and warehousing processes, practices, and standards that agencies could use to assess the ongoing need for property. We selected these organizations based on their knowledge about property management practices. Furthermore, we interviewed officials from four agencies—Census Bureau, Department of State, Internal Revenue Service, and National Aeronautics and Space Administration—that participate in the Interagency Committee on Property Management (ICPM), a committee chaired by GSA that consists of executive agency representatives interested in federal property. We invited all ICPM participants to speak with us regarding their practices for identifying unneeded property and interviewed all participants who volunteered to participate to understand how other agencies assess property for ongoing need. Finally, we reviewed FPPMA’s requirements and interviewed GSA’s Office of Government-wide Policy officials about GSA’s role in assisting agencies in identifying unneeded federal property, how FPPMA could affect GSA’s roles and responsibilities going forward, and GSA’s progress in implementing FPPMA. We conducted this performance audit from October 2018 to December 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the General Services Administration Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the individual named above, the following individuals made important contributions to this report: Nancy Lueke (Assistant Director), Rebecca Rygg (Analyst-in-Charge), Terence Lam, Malika Rice, Kelly Rubin, Patrick Tierney, Laurel Voloder, and Crystal Wesco.

Why GAO Did This Study Federal civilian agencies hold and manage billions of dollars in property that is not considered to be real property, such as vehicles, furniture, computers, and scientific instruments. Some of these items are stored in nearly 18,000 warehouses covering more than 90-million square feet. Agencies are required by law to regularly identify and dispose of unneeded items. However, GAO reported in 2018 that agencies often did not do so. The Federal Personal Property Management Act of 2018 requires agencies to use GSA guidance to assess the utilization and ongoing need for property. GAO was asked to review property stored in warehouses. This report examines: (1) what is known about property in selected agencies' warehouses and how much agencies spend to store it, and (2) the extent to which selected agencies assess the ongoing need for property stored in warehouses. GAO reviewed federal statutes, regulations, and GSA's guidance; analyzed policies from three agencies—FAA, Office of Science, and BOP—which were selected based on total warehouse square footage, among other factors; conducted site visits to agencies' warehouses; and interviewed stakeholders such as agency officials and industry groups. What GAO Found GAO found that three selected agencies stored a wide variety of property in their warehouses. For example: Federal Aviation Administration (FAA) warehouses at four main sites contained items used to build and repair aviation support systems, such as wind shear alert systems. Other sites contained tools and equipment to maintain aviation support systems or housed the systems themselves. The Department of Energy's Office of Science warehouses, located primarily at national laboratories, contained items, such as large magnets, for use in scientific experiments. Bureau of Prisons (BOP) warehouses, located mainly at federal correctional institutions, contained items, such as food, uniforms, and soap, for inmates. The above agencies reported spending approximately $50.1 million in fiscal year 2018 on warehouse rent, operations, and maintenance costs. The three selected agencies generally did not systematically assess the ongoing need for property in their warehouses and had limited guidance for doing so. For example, although two of the agencies had policies about when such an assessment should occur, none of the agencies specified how it should occur for most types of property. Instead, agencies primarily relied on agency officials' professional judgment to assess ongoing need. GAO identified instances where agencies retained unneeded property absent relevant guidance. For example, one agency site had stored obsolete computers dating back to the 1990s. While the General Services Administration (GSA) drafted guidance in response to recent legislation, this guidance does not describe approaches or practices stakeholders identified as potentially useful for assessing ongoing need for property, such as periodic retention justifications, use of data analytics, and utilization reviews. Further, while GSA officials intend to put the final guidance on GSA's website and provide it to agencies that participate in a GSA-chaired committee on property management by December 2019, GSA has not provided a documented plan or a timeline for broader dissemination. Guidance that incorporates such approaches could help agencies avoid retaining property that is no longer needed and, as a result, allow them to better manage their property and use of their warehouse space. What GAO Recommends GAO recommends that GSA incorporate approaches agencies could use to assess the ongoing need for property in GSA's guidance—such as periodic justifications, use of data, and utilization reviews—and develop a plan for communicating the guidance government-wide. GSA concurred with GAO's recommendation.

gao_GAO-20-172

gao_GAO-20-172_0

Background Federal Prosecution Roles and Immigration-Related Offenses DHS, DOJ, and the federal judiciary have different roles in the federal criminal process for immigration-related prosecutions, as shown in table 1. For this report, we define immigration-related offenses as the offenses listed in the Attorney General’s April 2017 memorandum (see table 2). Immigration-Related Prosecutions on the Southwest Border Criminal prosecution process. DHS and DOJ officials told us that DHS’s practices for referring cases for prosecution, and DOJ’s practices for prioritizing immigration-related prosecutions, vary by location along the southwest border. In general, individuals are prosecuted in the judicial district that corresponds with the location of their alleged criminal offenses. Each USAO prosecutes cases in one or more courts. USAOs coordinate with DHS and DOJ components, as well as the federal courts, to determine the types and number of cases, including immigration- related cases, each office will prosecute, according to DOJ officials. In particular, according to DOJ’s Principles of Federal Prosecution, the prosecutor has wide latitude in determining when, whom, how, and whether to prosecute for apparent violations of federal criminal law, and this broad discretion has been recognized on numerous occasions by the courts. Border Patrol officials on the southwest border told us that they receive training from the USAOs about the criminal prosecution process and that they use the prosecution priorities established by the USAO to determine whether to refer a case for criminal prosecution. In general, immigration-related cases referred to the USAO by Border Patrol follow the process described in figure 2. One-day prosecutions. In three federal judicial districts on the southwest border—Arizona, Texas Southern, and Texas Western—DOJ prosecutes defendants for improper entry in criminal proceedings that generally last one day or less, or one-day prosecutions (see figure 2). The volume of defendants prosecuted for improper entry per day in these districts varies depending on the volume of Border Patrol apprehensions and capacity limitations, among other things. Timeline of Attorney General Memoranda and Related Guidance Since 2017, there have been several federal directives related to DOJ’s prioritization of immigration-related prosecutions. They are summarized in table 3. DOJ Prioritized Immigration-Related Prosecutions in 2017 and 2018, Particularly by Increasing One- Day Improper Entry Prosecutions Beginning in April 2017, DOJ, in Coordination with DHS and Other Stakeholders, Prioritized Immigration-Related Cases Prosecutors in all five southwest border USAOs told us that, in response to the Attorney General’s 2017 directive and in coordination with DHS and other stakeholders, they took steps to prioritize immigration-related prosecutions in their respective jurisdictions. According to officials from the Office of the Attorney General, each USAO exercised its discretion in implementing the priorities identified in the memorandum. For example, some USAOs changed the threshold at which they would accept a prosecution referral for alien smuggling or illegal reentry. Officials in one USAO told us that, before the April 2017 memorandum, their office generally declined to prosecute alien smuggling cases involving fewer than six smuggled aliens. However, in response to the April 2017 memorandum, the office lowered its threshold to two smuggled aliens. Officials in another USAO said that in light of the April 2017 memorandum, they began accepting all referred illegal reentry cases that met the evidentiary standard. Previously, this office did not accept more than 150 illegal reentry defendants without a prior felony conviction per month. In particular, in response to the memoranda, all five USAOs along the southwest border prioritized improper entry referrals for prosecution. Some districts that were already prosecuting some improper entry cases, such as Arizona, were able to quickly increase such prosecutions by scaling their existing systems, according to USAO officials. Specifically, USAO officials in Arizona stated that their office began accepting referrals for first time improper entrants without aggravating circumstances in May 2017, in response to the April 2017 memorandum. In comparison, other USAOs created new processes to prosecute more improper entry cases because they were not previously accepting a significant number of such referrals. For example, prior to the Attorney General’s April 2017 memorandum, the USAO in the California Southern district did not prioritize the prosecution of improper entry cases. USAO officials in San Diego stated that in the spring of 2017, the USAO formed an immigration enforcement working group comprised of certain federal law enforcement entities in San Diego, including USMS, Border Patrol, and CBP OFO, to discuss potential actions the district might take to prioritize immigration enforcement. In May 2018, the Chief Judge in California Southern convened a criminal case management committee comprised of district and magistrate judges, court officials, USAO officials, USMS officials, and federal defenders, among others, to “identify and resolve problems” related to the increased prosecution of improper entry cases. These working groups collaborated to make decisions on issues such as the volume of improper entry cases the court could hear each day, how defendants in improper entry cases would meet with their attorneys, and how many defendants a public defender would represent in court each day. In July 2018, the San Diego court initiated a daily docket for misdemeanor improper entry cases. Regarding DOJ’s coordination with DHS, in four of the five southwest border districts, USAO officials told us that they informed local DHS partners, including local Border Patrol and OFO leadership, that their prosecution guidelines had changed in light of the 2017 memorandum and that they would accept more immigration-related cases for prosecution. As a result, Border Patrol and OFO referred more immigration-related cases to DOJ. Further, in response to the April 2018 zero-tolerance memorandum, Border Patrol issued guidance to all southwest border sectors instructing each sector to develop phased plans to refer all amenable apprehended adults to the USAO for improper entry prosecution, based on capacities of the USAO and the federal courts, and the sectors developed and implemented these plans. In general, these plans prioritized referrals of those individuals with a criminal history first, followed by those with no criminal history. For example, the plan for the Rio Grande Valley Border Patrol Sector stated that, incrementally, the Sector would increase prosecution referrals until attaining 100 percent prosecution on a timeline consistent with DOJ partners’ capacity. Regarding DOJ’s coordination with other stakeholders in the federal criminal process, including the federal courts, USMS, and public defenders, USAO officials told us that they coordinated at the local level to be able to increase immigration-related prosecutions, to the extent practicable. In particular, the federal judiciary held a border court conference in June 2018 and established a task force—including judges, public defenders, and DHS and DOJ representatives—to discuss issues related to changing prosecution priorities in southwest border districts. The task force met three times between July 2018 and April 2019. In addition, stakeholders told us they took other steps to accommodate the USAOs’ prioritization of such prosecutions. For example, Some courts added additional daily dockets or court sessions, or adjusted their use of facilities to accommodate the higher volume of cases being prosecuted. Court officials and magistrate judges we spoke with in all five southwest border districts told us that magistrate judges spent more time presiding over improper entry cases as the number of those cases increased. In McAllen, Texas, for example, court and USAO officials told us that the court added a second daily docket for misdemeanor improper entry cases in May 2018, and doubled the court’s capacity to hear such cases. In Las Cruces, New Mexico, court officials told us that there is one magistrate judge on duty each day for the docket that includes improper entry cases. Federal defenders in Las Cruces told us that stakeholders in Las Cruces, including the court, federal defenders, and USMS, met in spring 2018 and decided to use a second courtroom for magistrate judge duty—including improper entry cases—each day. One courtroom is used for an active proceeding while the other is used to meet and counsel defendants prior to their active court proceeding. In some locations, FDO told us that they developed new practices to provide representation to each defendant appearing in court. For example, the Federal Defender office in McAllen developed an “all hands on deck” process in May 2018, in which all available defenders meet individually with defendants in the courtroom before their initial appearance in court each day. In October 2018, we observed 14 Assistant Federal Public Defenders in McAllen meet with about 72 defendants during the hour before court; federal defenders we spoke with in McAllen said that the process we observed is their daily routine. In San Diego, federal defenders told us that in July 2018, they assigned a team to work full-time on improper entry cases. The team included six trial attorneys, two appellate attorneys, two legal assistants, two investigators, and one interpreter. The courts also increased their use of private defense attorneys appointed under the Criminal Justice Act and interpretation services due to the increased number of immigration-related cases. DOJ Increased Prosecutions of One-Day Improper Entry Cases in 2018, and Improper Entry Case Practices Varied Across Districts Several USAO districts were able to quickly increase the number of improper entry prosecutions in response to the Attorney General’s 2017 and 2018 memoranda, to the extent practicable, because such misdemeanor cases are less resource-intensive and less complicated to prosecute than felonies such as illegal reentry or alien smuggling, according to USAO officials in all five southwest border districts. Specifically, many improper entry cases were completed in one-day court proceedings in fiscal year 2018, and in some locations, the cases of 75 or more improper entry defendants were completed each day during a single court proceeding. In three of the five USAO districts—Arizona, Texas Southern, and Texas Western—improper entry prosecutions in fiscal year 2018 generally took place in one-day court proceedings. Based on our analysis of DOJ data, about 84 percent of the 62,000 improper entry cases filed in fiscal year 2018, or about 52,000 improper entry cases, took place in these three districts. We observed proceedings in Arizona and Texas Southern in July and October of 2018, respectively. These proceedings lasted approximately two hours, during which time 50 to 75 improper entry prosecutions were completed. In these proceedings, the initial hearing, presentation of evidence, plea, and sentencing took place during a single day—or a single morning or afternoon—in court. On the basis of our observations in Arizona and Texas Southern, as well as interviews with agency officials in Arizona, Texas Southern, and Texas Western between July 2018 and November 2018, first-time offenders without a prior criminal history typically pled guilty to the improper entry offense and were sentenced to time served. Those defendants remained in the custody of the arresting agency for the duration of the criminal court proceeding, according to Border Patrol and USMS officials at headquarters and agency officials in these three districts. At the time of our visits to the Arizona and Texas Southern districts, we observed judges sentence some defendants with a prior improper entry conviction to terms of imprisonment ranging from 10 to 180 days. The judge remanded these defendants to USMS custody to serve their sentence. In the other two USAO districts—California Southern and New Mexico— most improper entry prosecutions took place over the course of approximately one week, based on our observations of such prosecutions in California Southern and interviews with agency officials in California Southern and New Mexico in October and November 2018. Based on our analysis of DOJ data, about 16 percent of improper entry cases filed in fiscal year 2018, or about 10,000 cases, took place in these districts. After an initial appearance in court, the judge remanded the defendant to USMS custody and set a subsequent hearing for three to four days later. At the second hearing, the defendant typically pled guilty to the improper entry offense and the judge sentenced them. First-time offenders typically pled guilty to the improper entry offense and were sentenced to time served. The USAOs’ ability to increase improper entry prosecutions was also affected by different practices in the federal criminal process for improper entry cases in each of the five southwest border districts, as shown in table 4. In some locations, these practices affected the extent to which prosecutors could accept all improper entry cases referred for prosecution. According to officials from the Offices of the Attorney General and the Deputy Attorney General, DOJ contemplated such variation in its directives to federal prosecutors. Further, according to agency officials, practices for improper entry cases may change over time, depending on the priorities of various stakeholders in the federal criminal process, physical space limitations, or availability of resources such as interpreters, among other reasons. As of November 2018, Border Patrol referred nearly all single adults who could be charged with improper entry to the USAOs for prosecution in some districts, according to Border Patrol officials and Border Patrol’s operational guidance in those districts. In these locations, officials from Border Patrol, USAO, and the federal judiciary told us that they had sufficient capacity to process all such cases. In other districts, Border Patrol referred a lower percentage of single adults for prosecution for improper entry based on the ability of the USAO to accept such referrals or other factors, consistent with DHS’s May 2018 memorandum. For example, in Tucson, Arizona, the court generally allowed 75 improper entry cases per day at the time of our July 2018 visit. However, in McAllen, Texas, court officials told us that the court would hear as many improper entry cases as the USAO accepted for prosecution, which was as many as 200 cases per day, as of our October 2018 visit. At the time of our visits in July and October 2018, other considerations affecting the number of improper entry prosecutions included Border Patrol’s capacity to process case referrals (Texas Southern), restrictions on the number of daily defendants that the court could accommodate (Arizona, California Southern), and physical constraints, such as the number of seats for defendants in the courtroom (Texas Southern). In addition, public defense practices for misdemeanor improper entry cases varied across districts and, in some locations, affected the number of improper entry cases that the USAO could file each day. In California Southern and Arizona, each public defender represented a maximum of 4 or 6 defendants in court each day, respectively, in October 2018 and July 2018. In Texas Southern, one public defender may represent up to 100 defendants in court at a time, as of October 2018, according to defender office staff. Furthermore, local court rules or practices in some locations affected the number of improper entry cases that Border Patrol could refer or the USAO could file each day. For example, in California Southern, as of October 2018, the court required defendants to appear in court the next court day after their arrest. In addition, all defendants were required to undergo a medical screening for tuberculosis before their initial appearance in court. DHS Referred, and DOJ Prosecuted, More Immigration- Related Cases in Fiscal Year 2018 than in Each of the Four Prior Fiscal Years Border Patrol Referred More Cases to DOJ for Prosecution in Fiscal Year 2018 than in Each of the Four Prior Fiscal Years, and the Number of Referrals Varied by Location Border Patrol data indicate that the number of single adults referred to USAOs for prosecution more than doubled from fiscal year 2017 (about 49,700) to fiscal year 2018 (about 101,000), and was higher in fiscal year 2018 than in each of the four prior fiscal years. The total number of single adults Border Patrol apprehended varied from year to year over this time and Border Patrol data indicate that fewer single adults were apprehended in both fiscal years 2017 and 2018 than in each of the three prior fiscal years. However, the proportion of apprehended single adults that Border Patrol referred for prosecution was higher in fiscal year 2018 (38 percent) than in each of the four prior fiscal years (ranging from 20 to 24 percent) (see fig. 3). On the basis of our analysis of Border Patrol data, USAOs declined approximately 8 percent of Border Patrol’s criminal prosecution referrals in fiscal year 2018. In the four prior fiscal years, USAOs declined between 2 and 4 percent of such Border Patrol referrals. However, in fiscal year 2018, the number of cases Border Patrol referred for prosecution—and the number of cases that were accepted and prosecuted by USAOs—was also substantially higher compared to prior years, which was consistent with DHS and Border Patrol guidance to increase prosecution referrals to the extent practicable and consistent with DOJ partners’ and federal court capacity. The reasons for declinations varied and included timing and capacity-related reasons, according to Border Patrol’s data and officials. For example, defendants must generally appear before a judge within 48 hours of their Border Patrol apprehension and, according to Border Patrol officials, the remote locations of some apprehensions can make it difficult for Border Patrol to process, transport, and present defendants in court within the required timeframe. Border Patrol data indicate that apprehensions of single adults in fiscal year 2018 varied by U.S. Attorney district and, in general, Border Patrol referred a greater proportion of those apprehended for prosecution in districts with a relatively low number of apprehensions. Specifically, in the two districts with the fewest apprehensions (New Mexico and Texas Western, with about 10,000 and about 26,000 apprehensions, respectively), Border Patrol referred 80 and 75 percent of those apprehended for prosecution in fiscal year 2018. In the remaining three districts (Arizona, California Southern, and Texas Southern), each of which had more than 53,000 single adult apprehensions in fiscal year 2018, Border Patrol referred between 14 and 45 percent of those apprehended for prosecution in fiscal year 2018. According to Border Patrol officials in these three districts, various factors influenced the number of referrals to USAOs, including court capacity, availability of Border Patrol agents to prepare cases for referral, and USAO capacity to accept and prosecute cases, consistent with the Attorney General’s guidance to prioritize such prosecutions to the extent practicable. Immigration-Related Prosecutions Increased in Fiscal Year 2018, and More than Half Were for Improper Entry DOJ prosecuted more immigration-related cases—including improper entry, illegal reentry, and alien smuggling cases—in fiscal year 2018 than in each of the prior four fiscal years. Specifically, southwest border USAOs filed about 91,000 improper entry, illegal reentry, and alien smuggling cases in fiscal year 2018, compared to a prior four-year high of about 78,000 immigration-related cases filed in 2014. On the basis of our analysis of DOJ data, cases with a lead charge of improper entry comprised more than half of DOJ’s immigration-related cases filed each year from fiscal years 2014 through 2018. Further, the total number of cases filed with a lead charge of improper entry, illegal reentry, or alien smuggling increased between fiscal year 2017 and fiscal year 2018 in the five southwest border districts, consistent with the priorities in the April 2017 and April 2018 memoranda, although the magnitude of the increases varied. Figure 4 illustrates the number of cases filed by USAOs with a lead charge of improper entry, illegal reentry, or alien smuggling along the southwest border, as well as trends in such cases from fiscal years 2014 through 2018. From fiscal year 2014 through 2018, more than 95 percent of improper entry, more than 90 percent of illegal reentry, and more than 80 percent of alien smuggling cases ended in convictions. The majority of defendants for improper entry and illegal reentry cases from fiscal years 2014 through 2018 were Mexican nationals, although the proportion of defendants with nationalities other than Mexican increased in fiscal year 2018 relative to the prior four fiscal years. The majority of defendants for alien smuggling cases from fiscal years 2014 through 2018 were U.S. nationals. See appendices II and III for more detailed information on case dispositions and nationalities of defendants. Improper entry. DOJ data indicate that the total number of cases filed with a lead charge of improper entry in southwest border districts more than doubled between fiscal year 2017 and 2018, as illustrated in table 5. Figure 5 illustrates the number of improper entry cases filed by southwest border USAOs each month in fiscal years 2017 and 2018. In New Mexico and Arizona, the number of improper entry cases filed increased notably in June 2017. These districts generally did not prosecute first-time entrants for these misdemeanor offenses from 2014 until 2017, and changed their prosecution practices in response to the Attorney General’s April 2017 memorandum, according to USAO officials we spoke with in those districts. In Texas Southern, the number of improper entry cases filed increased notably in April 2018. Prior to April 2018, the USAO in McAllen allowed 40 to 50 improper entry prosecutions per day, according to USAO officials. The USAO removed this limitation in response to the Attorney General’s April 2018 memorandum. As of October 2018, this USAO accepts all prosecution referrals with sufficient evidence (on average, 100 to 200 improper entry prosecutions per day), according to officials. From fiscal year 2017 through 2018, improper entry cases filed in Texas Southern nearly tripled, from about 10,800 to about 30,100 cases. In Texas Western, the number of improper entry cases filed began to increase in March 2018, but to a lesser extent than other districts, and then decreased from July through September 2018. USAO officials attributed the increase to increased Border Patrol apprehensions and said that they accept all Border Patrol prosecution referrals, but the number of cases that the USAO receives depends on fluctuating Border Patrol apprehension numbers. In California Southern, the number of improper entry cases filed began to increase in May 2018. Prior to July 2018, California Southern did not have a court docket dedicated to prosecuting improper entry misdemeanor offenses. According to officials, following the Attorney General’s April 2018 memorandum, the San Diego district court, in coordination with the USAO, agreed to establish a daily improper entry docket with the capacity to hear initial appearances for 40 to 52 improper entry cases each day. Illegal reentry. DOJ data indicate that the number of cases USAOs filed with a lead charge of felony illegal reentry along the southwest border declined from fiscal years 2015 through 2017 before increasing by 2,669 cases from fiscal year 2017 through 2018. However, the number of illegal reentry cases filed in fiscal year 2018 (25,112) was lower than in fiscal year 2014 (31,670) or 2015 (28,480), and the magnitude of the increase in illegal reentry cases filed from fiscal year 2017 through 2018 (12 percent) was smaller than the increase in improper entry cases during the same period (130 percent). The number of cases filed with a lead charge of illegal reentry declined in Arizona each year between fiscal years 2015 and 2018, but increased or varied in other districts. Between fiscal year 2017 and 2018, illegal reentry cases filed increased most notably in Texas Western, where there were 69 percent more illegal reentry cases filed in fiscal year 2018 than in fiscal year 2017. Federal court and USAO officials in Texas Western attributed this increase in illegal reentry prosecutions to increased Border Patrol apprehensions and referrals for prosecution in fiscal year 2018. Table 6 illustrates illegal reentry cases filed, by fiscal year, from fiscal years 2014 through 2018. USAO officials attributed the changes in illegal reentry cases filed from fiscal year 2014 through 2018 to changes in prosecution practices as well as changes in the number of apprehensions. For instance, the New Mexico USAO removed a monthly limitation originally enacted in fiscal year 2016 on the number of illegal reentry cases filed they would accept following the April 2017 memorandum, according to Border Patrol and USAO officials. Other locations have varying thresholds and practices regarding accepting, charging, and prosecuting illegal reentry cases. USAO officials in New Mexico and Texas Western told us that they charge defendants with illegal reentry if the defendant has one prior deportation or one prior conviction for improper entry. Officials in three other districts told us that they generally require a more extensive criminal history—for instance, they might require multiple prior improper entry convictions—to charge illegal reentry. USAO officials in Arizona and California said that they file cases with a lead charge of illegal reentry that might ultimately end with improper entry convictions. For example, our analysis of EOUSA data indicates that of almost 12,000 illegal reentry cases filed in Arizona in fiscal year 2017, approximately 77 percent ended with an improper entry conviction and approximately 18 percent ended with an illegal reentry conviction. Figure 6 illustrates the number of cases filed with a lead charge of illegal reentry filed each month in fiscal years 2017 and 2018. Alien smuggling. DOJ data indicate that the number of cases filed with a lead charge of alien smuggling increased in four of the five southwest border districts from fiscal year 2017 through 2018. Officials from two USAO locations along the southwest border told us that they changed their thresholds for how many material witnesses (individuals being smuggled) must be present to accept an alien smuggling referral in response to the Attorney General’s April 2017 memorandum. For instance, the USAO in San Diego lowered the threshold for accepting alien smuggling referrals and, following the April 2017 memorandum, places equal priority on all alien smuggling referrals. Prior to the April 2017 memorandum, the USAO would have considered several factors when deciding whether to accept the referral, such as if there was a risk of harm to the material witnesses or whether the conviction could result in a significant term of imprisonment for the smuggler. Figure 7 illustrates the number of cases filed with a lead charge of alien smuggling each month over fiscal years 2017 and 2018. Agencies Realigned Existing Resources and Allocated Additional Resources to Help Increase Immigration-Related Prosecutions Agencies Along the Southwest Border Shifted Existing Resources to Support Increased Immigration-Related Prosecutions DOJ, DHS, and the federal judiciary realigned resources to support the prosecution priorities outlined in the April 2017 and April 2018 memoranda. Officials from USAOs, USMS, Border Patrol, federal courts, and federal defenders along the southwest border told us that they are using more personnel, physical space, or both to support increased immigration-related prosecutions than they were prior to DOJ’s prioritization of immigration enforcement in April 2017. When USAOs along the southwest border changed their prosecutorial priorities and realigned resources in response to the April 2017 and April 2018 memoranda, other agencies, such as USMS and the federal judiciary, also realigned resources to respond to and support increased immigration-related prosecutions. In some cases, these realignments affected their ability to conduct other activities. Officials from USMS and the courts told us that, as stakeholders in the federal criminal process, they are accustomed to reacting to changing conditions that may affect their operations. For example, these officials’ operations could be affected by changes in the number of Border Patrol apprehensions, changes in Border Patrol’s prosecution referral priorities, changes in the location of drug or human smuggling activity, and changing USAO prosecutorial priorities, among other things. USAOs. USAO officials in three locations stated that the more time prosecutors spend on reactive work—such as misdemeanor or felony immigration-related cases—the less time Assistant U.S. Attorneys (AUSA) have to work on other issue areas, including proactive cases that may take months or years of work to build, or civil cases. For instance, USAO officials from Texas Southern said that the high immigration caseload in McAllen affects AUSAs’ ability to prosecute other types of cases, such as Organized Crime Drug Enforcement Task Force cases, which tend to be long-term cases. According to USAO officials in San Diego, when prosecutors began accepting improper entry referrals in July 2018, there was a short-term decline in the number of prosecutions that were initiated for other cases. This decline mainly affected drug and alien smuggling cases, some of which were referred to state or local prosecutors, according to USAO officials in San Diego. As of October 2018, USAO officials in San Diego said that improper entry prosecutions were not affecting their ability to accept referrals for new felony prosecutions. USMS. According to USMS officials, each additional court docket, courtroom in use, or immigration-related defendant who appears in court requires judicial security support. USMS officials in all five southwest border locations told us that they took actions to meet the judicial security mission need, but that the increased prosecutions have strained their staff. USMS officials we spoke with in all five southwest border districts said that they reassigned deputies in fiscal year 2018 from proactive task forces (such as task forces dedicated to arresting individuals with active federal warrants) to judicial security court duty and detention security to support increased immigration-related prosecutions. In particular, USMS officials said that they assigned more deputies to judicial security court duty because of the increase in improper entry prosecutions. USMS officials we spoke with in several locations on the southwest border said that the increased judicial security duty has made it difficult for their deputies to meet their training requirements. They are concerned that the high demand for judicial security in southwest border districts may affect their ability to retain deputies. Officials from USMS in multiple locations along the southwest border told us that the increase in immigration-related prosecutions strained their existing detention space. For instance, California Southern required additional detention space for defendants in improper entry cases, and could not locate additional detention space nearby or within the judicial district. As a result, USMS officials told us in October 2018 that deputies may drive defendants to neighboring judicial districts, including California Central, Nevada, and Arizona, to detain them before and between court appearances. According to USMS officials, providing transportation for such defendants can comprise deputies’ entire shifts. Additionally, officials in the Texas Western district told us in November 2018 that due to the increase in immigration-related prosecutions in fiscal year 2018, detention facilities in Del Rio reached capacity. USMS transports prisoners up to seven hours one way to other detention facilities. Further, USMS received permission to triple-bunk prisoners (using three stacked beds, rather than two stacked beds) in Del Rio and El Paso, and to use additional temporary beds, such as cots, to house additional prisoners close to courthouses. As we have previously reported, the average daily population of USMS prisoners is directly influenced by, among other things, the activities and decisions of federal law enforcement, USAOs, and the federal judiciary. According to USMS data, the average daily population of immigration- related prisoners on the southwest border increased from 7,796 in May 2017 (a five-year low) to 11,668 in September 2018 (a five-year high). According to USMS officials with whom we spoke and documents we reviewed, in 2018, USMS sought additional detention space. In May 2018, USMS issued a public request for information to determine the availability of contractor owned and operated secure detention facilities on the southwest border. In October 2018, USMS signed intragovernmental agreements with two local detention facilities in Texas, adding approximately 655 available beds to its inventory. Additionally, USMS officials in Las Cruces told us that they had more detention space than they required for prisoners in New Mexico and that, following the April 2018 memorandum, USMS began to accept prisoners from other districts. Prior to the April 2018 memorandum, USMS in New Mexico had approximately 1,300-1,400 of their own prisoners in custody. As of November 2018, USMS in Las Cruces had approximately 1,800 prisoners in custody from New Mexico and approximately 500 prisoners in custody from neighboring districts. Border Patrol. Border Patrol agents support, and in some cases supplement, DOJ components in both prosecution and judicial security work. As of March 2019, in nine of nine southwest border sectors, Border Patrol reported that it had detailed agents to USAOs to assist with tasks like data entry and preparing court documents for immigration-related prosecutions. In addition, in seven of nine sectors, Border Patrol detailed agents to USMS locations to assist with judicial and detention security. The number of agents from Border Patrol that are detailed to assist DOJ components with immigration-related prosecutions generally varies based on the volume of prosecutions that the USAO receives and accepts and, in some sectors, based on available Border Patrol agent resources, according to Border Patrol officials. Following the Attorney General’s memoranda, Border Patrol increased the number of agents that it detailed to certain USAOs and USMS locations along the southwest border, both temporarily and on an ongoing basis, because of the increased volume of immigration-related prosecutions, according to Border Patrol officials. As of March 2019, Border Patrol sectors across the southwest border detailed from zero to four agents to perform USMS functions, and zero to five agents to perform USAO functions. The length of detail and duties assigned to Border Patrol agents detailed to USAOs and USMS vary by location, according to officials. Generally, when an assignment ends, Border Patrol agents return to their regular Border Patrol duties. Federal courts. Federal court officials we spoke with in five locations stated that they faced challenges resulting from the increased immigration caseload. For instance, court officials in Las Cruces said that, as of November 2018, staff in the clerk’s office often work on weekends to keep up with court scheduling and paperwork resulting from increased improper entry prosecutions. The Las Cruces court also implemented telework options for clerk staff to give them the option of working additional hours from home. Additionally, officials we spoke with from several courts reported that they had existing needs for judgeships, and the increasing immigration caseload placed additional strain on district and magistrate judges. For instance, the district court in Del Rio has one district judge and the number of illegal reentry prosecutions in fiscal year 2018 increased by almost 70 percent compared to fiscal year 2017. Court officials we spoke with in two locations told us that sentencing dates have been pushed out because of the increase in district judges’ caseloads. According to federal court officials in Del Rio, the district judge’s calendar is so full that, in some cases, a defendant’s sentencing might be pushed back far enough that the defendant has already served more jail time than the federal sentencing guidelines recommend by the time the defendant is sentenced. In addition, multiple court officials in multiple locations across the southwest border told us that increased immigration-related prosecutions, and particularly improper entry cases, increases strain on courtroom facilities and equipment and, in some instances, courts have to replace equipment and furniture more often. For example, in Tucson, Arizona, the improper entry courtroom can hold up to 75 improper entry defendants in restraints, such as handcuffs and/or leg restraints, at a time during morning improper entry proceedings, and court officials told us that the restraints worn by defendants cause damage to the chairs and benches in the courtroom. Defender services. FDO staff we spoke with in several southwest border districts told us that they dedicated more staff or staff time towards defendants in immigration-related cases and accommodated increased prosecutions within existing resources as of December 2018. For instance, defenders in Las Cruces stated that the court added a new docket for improper entry cases, and defense attorneys are at times scheduled to be in two courtrooms at once, and must cover for each other. Defenders in Las Cruces also told us that they have run out of physical office space for their staff. Federal defenders in McAllen said that the amount of time defense attorneys spend on improper entry interviews affects the time they can spend on felony cases. In addition, these defenders described the process of preparing 100 or more defendants for criminal proceedings each day as draining. Defenders in McAllen noted that they filed more continuances in fiscal year 2018 than in prior years as a result of the increased workload caused by the expanded improper entry docket. DOJ, DHS, and the Federal Judiciary Added Personnel to Support Increased Immigration- Related Prosecutions in the Short and Long-Term, and Tracked Some Related Expenditures EOUSA, USMS, Border Patrol, and the federal judiciary temporarily surged personnel from locations across the United States to the southwest border to support increased immigration-related prosecutions. These agencies tracked some costs associated with those temporarily detailed personnel, among other costs associated with increased immigration-related prosecutions. Additionally, EOUSA announced plans to hire new attorneys to prosecute immigration-related offenses in May 2018, both on the southwest border and in the interior of the United States. EOUSA. In May and June 2018, DOJ announced plans to permanently hire 70 new AUSAs to prosecute immigration-related offenses both at the southwest border and in the interior of the U.S. Additionally, EOUSA officials told us that they subsequently received DOJ approval to hire 13 more AUSAs to work on immigration and border security issues on the southwest border. In fiscal year 2018, EOUSA expended about $9.8 million on personnel costs associated with these prosecutors—including 35 immigration crimes prosecutors in the interior of the United States, 42 immigration crimes prosecutors in the five southwest border districts, and 6 civil condemnation AUSAs working on the southwest border. EOUSA estimated that the fiscal year 2019 continuing personnel costs associated with these prosecutors would be about $17 million. In its fiscal year 2020 Congressional Budget Justification, EOUSA requested a $23.3 million increase in funding from Congress to sustain hiring and program operations that were initially funded in fiscal year 2018, including the immigration prosecutors. EOUSA also intends to allocate a portion of these 2020 funds to USAOs around the country with demonstrable workload challenges. EOUSA and USAO officials said that these permanent AUSA positions would support immigration prosecutions on the southwest border in the long-term. USAO officials we spoke with in all five southwest border districts between July and November 2018 said that they were in the process of hiring these immigration AUSAs. While EOUSA was in the process of permanently hiring new AUSAs, EOUSA temporarily surged Special Assistant U.S. Attorneys (SAUSAs) to southwest border districts that needed more prosecutors to handle the increased immigration caseload. Some of these SAUSAs prosecuted improper entry offenses specifically and others prosecuted any immigration-related case. Specifically, EOUSA solicited attorneys from other DOJ components, the Department of Defense (DOD), and CBP to serve as SAUSAs for immigration-related offenses along the southwest border. Beginning in June 2017, DOJ detailed 12 attorneys from non- southwest border USAOs and other DOJ components to prosecute immigration-related cases in all five districts on the southwest border. In fiscal years 2017 and 2018, EOUSA expended approximately $440,000 on travel and lodging for these 12 SAUSAs. In June 2018, DOD agreed to provide military attorneys to act as SAUSAs and support immigration-related prosecutions on the southwest border. DOD detailed a total of 21 military attorneys to the southwest border for approximately six months each between June 2018 and January 2019, according to EOUSA. According to USAO officials in New Mexico, which received five military SAUSAs, and California Southern, which received five, these SAUSAs provided key support that allowed these districts to increase improper entry and illegal reentry prosecutions beginning in June 2018. In fiscal year 2018, EOUSA estimated that it expended approximately $1,186,000 on salaries, travel, and lodging for these 21 SAUSAs. In some southwest border locations, CBP regularly provides SAUSAs to add prosecutor capacity to USAOs. For example, in four locations, CBP SAUSAs are the federal prosecutors for misdemeanor improper entry cases and appear daily in court to prosecute these cases. In San Diego, CBP SAUSAs began supporting the misdemeanor improper entry docket in July 2018, when the docket began. In New Mexico, 10 part-time CBP SAUSAs supported the improper entry docket temporarily between January and July 2018, which allowed New Mexico to begin prosecuting improper entry cases with no effect to the workload of its full-time AUSAs. CBP officials also said that CBP has provided full-time SAUSAs for a six or 12-month term to some USAOs on an ongoing basis, depending on USAO request and CBP workload. USAO officials have asked CBP for additional SAUSAs in San Diego and Yuma; as of April 2019, CBP officials said that due to CBP’s workload in these locations, they have not agreed to additional SAUSAs in these locations. USMS. From June through November 2018, USMS detailed deputies from non-southwest border locations to southwest border courts to support judicial security operations. Approximately 96 deputies participated in these temporary detail rotations, which lasted two to three weeks each, over the six month period. USMS established a budget code to track additional expenditures that USMS headquarters incurred related to implementing the April 2018 memorandum. These additional expenditures included travel and lodging costs for the detailed USMS deputies and transportation costs, among others. USMS reported approximately $1,149,000 in expenditures from May through December 2018 under this budget code. In its 2020 Congressional Budget Justification, USMS requested nearly $8 million from Congress for 35 positions to address departmental priorities and initiatives, including immigration enforcement. USMS officials said that their workload, including immigration prosecutions in fiscal year 2018 surpassed previous peak levels. For instance, USMS reported more immigration-related “prisoners received” in 2018 than in each of the prior five fiscal years. Border Patrol. Border Patrol established a budget code in April 2018 to track additional expenditures directly associated with implementing the April 2018 memorandum. In particular, according to Border Patrol budget officials and documentation, Border Patrol officials were to use the budget code to track expenditures related to detainee food, supplies, and transportation. In addition, the code was to be used for Border Patrol agent overtime expenditures and any travel expenditures that could be attributed to the April 2018 memorandum. From April 2018 through December 2018, Border Patrol reported approximately $2,316,000 in expenditures under this budget code. Federal judiciary. The federal judiciary sends visiting judges from other parts of the United States to southwest border districts to assist with judge caseloads, including immigration cases. For instance, the federal judiciary approved 67 visiting judge assignments from other parts of the U.S. to New Mexico and Texas Western in fiscal years 2017 and 2018; AOUSC reported expending approximately $114,000 on travel costs for these visiting judges. Federal courts along the southwest border also expended more funds on contracted interpreter services in fiscal year 2018 than in any of the prior four fiscal years. When a defendant does not speak English, courts may have interpreters on staff and courts may use contracted interpreter services. Court officials from multiple locations along the southwest border told us that contracted interpreter services became increasingly difficult to obtain following the increase in immigration-related prosecutions. According to federal judiciary documentation, there were 100,000 more court events, or defendant appearances before a judge, in southwest border courts requiring court interpreter services in fiscal year 2018 than there were in fiscal year 2017. Expenditures for contracted court interpreters increased by over $450,000 from fiscal year 2017 to fiscal year 2018 for southwest border courts. Agency Comments We provided a draft of the sensitive report to DOJ, DHS, and AOUSC for their review and comment. DOJ, DHS, and AOUSC provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Attorney General of the United States, the Acting Secretary of the Department of Homeland Security, the Administrative Office of the U.S. Courts, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-8777 or goodwing@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix V. Appendix I: Objectives, Scope, and Methodology This appendix provides additional details on our objectives, scope, and methodology. Specifically, our objectives were to provide information on the following: 1. how the Department of Justice (DOJ) prioritized criminal prosecutions of immigration-related offenses in response to the Attorney General’s 2017 and 2018 memoranda; 2. what Department of Homeland Security (DHS) and DOJ data from fiscal years 2014 through 2018 indicate about criminal prosecutions of immigration-related offenses; and 3. resources that DOJ, DHS, and the federal judiciary used to implement increased immigration-related prosecutions. This report is a public version of the prior sensitive report that we provided to you in August 2019. DHS, DOJ, and the Administrative Office of U.S. Courts (AOUSC) deemed some of the information in the prior report as Law Enforcement Sensitive or For Official Use Only, which must be protected from public disclosure. Therefore, this report omits sensitive information about specific law enforcement, prosecutorial, and judicial practices along the southwest border, including certain courtroom security and agency staffing information. Although the information provided in this report is more limited, the report addresses the same objectives as the sensitive report and uses the same methodology. For all three objectives, we generally focused our review on the five U.S. Attorney Office (USAO) districts along the southwest border—Arizona, California Southern, New Mexico, Texas Southern, and Texas Western— because the Attorney General’s 2017 and 2018 memoranda specifically directed officials in these districts to prioritize improper entry prosecutions. Further, approximately 93 percent of all immigration-related prosecutions from fiscal years 2014 through 2018 took place in these districts. USAO districts and federal judicial districts have the same boundaries. U.S. Border Patrol (Border Patrol) sectors along the border are generally not contiguous with USAO districts. We visited three of the five districts and interviewed officials by telephone from the other two southwest border districts. Specifically, we conducted in-person site visits to Arizona in July 2018 and to California Southern and Texas Southern in October 2018. We selected these locations on the basis of several factors, including Border Patrol apprehension characteristics and DOJ prosecution practices. Specifically, to select the locations for our site visits, we considered DOJ’s history of prosecuting improper entry offenses in different locations, including whether districts implemented changes to their practices for prosecuting improper entry offenses in response to the Attorney General’s memoranda. For instance, we considered districts’ practices for prosecuting improper entry offenses and whether those practices changed in response to the April 2017 or April 2018 memoranda. In addition, we considered the number of Border Patrol apprehensions in each USAO district and changes in the number of apprehensions from fiscal years 2014 through 2018. We also considered factors such as whether DOJ, DHS, and federal court facilities are in close proximity, among other things. In the three districts we visited, we met with DOJ and federal court officials, including magistrate and district judges, to understand and observe their roles in the criminal prosecution process. We met with USAO, U.S. Marshals Service (USMS), Federal Defender Organizations (FDO), and federal court officials and observed federal criminal court proceedings in Tucson, Arizona; San Diego, California; McAllen, Texas; and Brownsville, Texas. We observed the criminal prosecution process from arrest to conviction and sentencing, including observations of district and magistrate court proceedings and USMS intake and holding facilities in federal courthouses. In addition, we observed U.S. Customs and Border Protection’s Border Patrol agents and Office of Field Operations (OFO) officers processing apprehended individuals and referring them for prosecution. We met with Border Patrol officials in Tucson, Arizona; McAllen, Texas; and San Diego, California. We met with OFO officials at ports of entry in Nogales, Arizona; San Ysidro, California; Hidalgo, Texas; and Brownsville, Texas. We also interviewed USAO, USMS, federal court, Border Patrol, and OFO officials who are involved in immigration prosecutions in Las Cruces, New Mexico in November 2018 (New Mexico district) and Del Rio, Texas in November 2018 (Texas Western district). Although the information we obtained from these site visits and interviews cannot be generalized to all locations along the southwest border, these interviews provided important insights and perspectives about immigration-related prosecutions and any process, volume, or resource changes in immigration-related prosecutions following the April 2017 and 2018 memoranda. To determine how DOJ prioritized immigration-related prosecutions, we obtained and reviewed operational guidance, policies, and memoranda describing how DOJ, DHS, and the federal judiciary implement such prosecutions along the southwest border. We also reviewed documentation to identify any changes to such practices associated with implementing the Attorney General’s April 2017 and the April 2018 memoranda. We reviewed training materials from the Executive Office of U.S. Attorneys (EOUSA) provided to some federal prosecutors regarding prosecuting immigration-related cases at a 2018 Border Security Coordinator conference and relevant U.S. Attorneys’ Bulletins from DOJ’s Journal of Federal Law and Practice, such as the July 2017 bulletin, Prosecuting Criminal Immigration Offenses, and the Justice Manual, which contains publicly available DOJ policies and procedures, including criminal prosecution procedures. In addition, we interviewed headquarters and district officials from DOJ, DHS, and the federal courts to obtain their perspectives on the Attorney General’s prioritization of immigration-related prosecutions and any changes in practices as a result of the two memoranda. Specifically, from DOJ, we interviewed officials from the Offices of the Attorney General and the Deputy Attorney General about the development and implementation of the April 2017 and April 2018 memoranda. We also interviewed officials from EOUSA about headquarters-level support to USAOs. We interviewed headquarters officials from USMS about how the Attorney General’s prioritization of immigration offenses affected USMS operations and about available data measuring such effects. from DHS, we interviewed Border Patrol and OFO headquarters officials about actions CBP components took in response to the Attorney General’s prioritization of immigration prosecutions and reviewed DHS, CBP, and Border Patrol memoranda and Border Patrol operational guidance related to the prioritization of immigration prosecutions. We also interviewed officials from U.S. Immigration and Customs Enforcement (ICE) about the effect of the Attorney General’s prioritization on ICE’s operations. from the Administrative Office of the U.S. Courts (AOUSC)—the federal judiciary agency that provides legislative, administrative, management, and program support to federal courts, among other functions—we interviewed officials in Washington, D.C. about the federal judiciary’s roles and responsibilities related to criminal immigration-related cases, including the roles of magistrate and district judges and public defenders. To determine what DHS and DOJ data indicate about prosecutions of immigration-related offenses, we analyzed record-level apprehension and prosecution referral data from Border Patrol’s Enforcement Integrated Database/e3 (e3), as well as record-level prosecution data from EOUSA’s CaseView from fiscal years 2014 through fiscal year 2018, the most recent data available at the time of our analysis. Border Patrol data. In reviewing the Border Patrol data, we determined that the majority of Border Patrol apprehensions (about 97 percent) from fiscal years 2014 through fiscal year 2018 took place along the southwest border. We excluded the small percentage of apprehensions nationwide that did not take place along the southwest border from our primary analysis, meaning that we excluded apprehensions in all districts but Arizona, California Southern, New Mexico, Texas Southern, or Texas Western from our primary analysis. We assigned each Border Patrol sector apprehension to its corresponding judicial district to maintain the judicial district as our unit of analysis for the apprehension and prosecution referrals data. For instance, if the El Paso Border Patrol sector referred a prosecution to the USAO in Las Cruces, New Mexico, we report that referral as occurring in the district of New Mexico. We matched data from e3’s apprehensions module with data from e3’s prosecutions module using an identifier that Border Patrol officials told us was unique to each apprehended individual to analyze those individuals that were and were not referred for criminal prosecution. Border Patrol’s apprehensions and prosecution referrals include individuals who are deportable and non-deportable, as determined by Border Patrol. According to Border Patrol officials, non-deportable individuals may be U.S. citizens, foreign nationals who have a valid visa, or individuals who otherwise may not be amenable to removal from the United States. We have included non-deportable individuals in our analysis because they may be referred for prosecution for immigration-related crimes, including alien smuggling. Appendix II includes information on Border Patrol apprehensions and prosecution referrals in each judicial district from fiscal years 2014 through 2018 and information on apprehensions and prosecution referrals by nationality, including U.S. citizens. We restricted our Border Patrol data analysis to apprehensions of non-juveniles who Border Patrol did not process as members of family units. In other words, we analyzed apprehensions and prosecution referrals of single adults. According to Border Patrol guidance and agency officials, e3 has system checks in place that do not allow members of family units to be referred for criminal prosecution. Prior to April 2018, Border Patrol officials said that individuals who were to be referred for prosecution were generally processed by Border Patrol as single adults whether or not they were apprehended with their minor children. In April 2018, an update to e3 allowed Border Patrol agents to separate one or more members of a family unit from that family unit and refer those individuals for prosecution. As stated previously, we included individuals that Border Patrol processed as single adults in our analysis of Border Patrol apprehensions. EOUSA data. In reviewing EOUSA record-level prosecution data from fiscal years 2014 through 2018, we determined that the majority of cases filed with an immigration-related lead charge (over 90 percent of cases with an immigration-related lead charge) took place along the southwest border. We excluded prosecutions that did not take place along the southwest border from our primary analysis; we report on them in an appendix. Additionally, we determined that improper entry, illegal reentry, and alien smuggling charges comprised approximately 99 percent of immigration-related cases filed on the southwest border from fiscal years 2014 through 2018. We excluded the other charges that the Attorney General listed in the April 2017 memorandum from our primary analysis. We analyzed EOUSA data based on the lead charge of the prosecution record. The lead charge is typically the most serious provable offense for which a defendant can be prosecuted, as determined by the USAO. We analyzed EOUSA data by fiscal year from fiscal years 2014 through 2018 to determine overall trends in immigration-related prosecutions over time. We also analyzed data by month in fiscal year 2017 and fiscal year 2018 to identify any changes in immigration-related prosecutions following the April 2017 and April 2018 memoranda. We interviewed knowledgeable USAO officials in southwest border districts level to understand how practices that they changed in response to the April 2017 and April 2018 memoranda were reflected in the data. We also analyzed the nationality of defendants based on lead charge for fiscal years 2014 through 2018, and for fiscal year 2018, to determine any changes in nationality of those prosecuted in the most recent fiscal year compared to prior fiscal years. We identified a population of defendants whose nationalities were listed as ‘unknown,’ in the EOUSA data. When USAOs are unable to determine the nationality of a defendant, officials entering the case data will list that nationality as ‘unknown.’ In appendix III, we report on the proportion of defendants with a nationality that is ‘unknown’ for alien smuggling cases because ‘unknown’ nationalities were relatively common for alien smuggling cases. We grouped the lead charges into offense categories based on the statute of the offense. We analyzed EOUSA data at the statutory level rather than by the individual charged offenses because EOUSA officials told us that USAOs may have differing data entry practices related to the level of specificity at which they enter lead charge data into CaseView. Additionally, EOUSA directed USAOs to ensure that improper entry, illegal reentry, and alien smuggling cases are entered into EOUSA’s data system on a monthly basis at the statute level in August 2017. Table 8 lists the specific offenses that we combined under their shared statute for our analysis. We assessed the reliability of Border Patrol and EOUSA data by testing for missing data and obvious errors, reviewing related documentation such as data dictionaries and guidance for entering data, and interviewing knowledgeable agency officials both at the headquarters level and in the three districts that we visited. We determined that the Border Patrol and the EOUSA data are sufficiently reliable for reporting on immigration- related prosecutions and individuals that Border Patrol apprehended and referred for criminal prosecution. To determine the resources used to implement increased immigration- related prosecutions, we obtained and reviewed DOJ, DHS, and federal judiciary documentation focused on any existing resources that agencies realigned to implement or support increased immigration prosecutions, as well as expenditures or additional personnel used to support the implementation of increased immigration-related prosecutions. For example, we reviewed memoranda of understanding between EOUSA and the Department of Defense (DOD) regarding DOD detailing attorneys to EOUSA to prosecute immigration-related offenses, as well as USMS intergovernmental agreements used to expand detention space. We also reviewed data from USMS on the unique prisoners received, average daily prisoner population, and total prisoner appearances in court to determine any changes in the volume of USMS prisoners from fiscal year 2014 through fiscal year 2018. We reviewed documentation from southwest Border Patrol sectors specifying the number of Border Patrol agents that those sectors detailed to USMS and USAO locations, as well as any changes in the number and duration of agents detailed to those locations following the April 2017 and April 2018 memoranda. To specifically identify expenditures or personnel for implementing increased immigration-related prosecutions, we reviewed agency documentation, such as documentation from expenditure tracking systems from USMS and Border Patrol. We interviewed agency budget and program officials from USMS’ Offices of Budget Formulation, Forecasting and Analysis, and General Counsel; EOUSA’s Office of Resource Management and Planning; AOUSC’s Office of the Financial Liaison and Analysis Staff; Border Patrol’s Office of Budget Execution; OFO’s Office of Budget Formulation; CBP’s Budget Office; and CBP’s Office of Chief Counsel. In instances where there was no explicit distinction between expenditures or personnel for specifically supporting immigration-related prosecutions and expenditures or personnel used to support other prosecutions, we identified the general account within which immigration-related prosecution costs would be included, and noted that those expenditures include costs for other prosecutions as well. In addition, where agencies identified that they used personnel resources to implement immigration- related prosecutions, we collected related documentation, such as expenditures for temporary details from other parts of the United States to the southwest border, as available, and spoke with district officials by telephone and during our site visits to better understand the use of these personnel resources. We also reviewed Congressional Budget Justifications for fiscal year 2020 to identify expenditures that agencies requested from Congress to support increased immigration-related prosecutions. We conducted this performance audit from May 2018 to August 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Enclosures on U.S. Southwest Border Districts This appendix provides additional detail on and characteristics of immigration-related prosecutions in the five U.S. southwest border districts: Arizona, California Southern, New Mexico, Texas Southern, and Texas Western. Each enclosure contains the following information: Description of the district. In this section, we provide a narrative description of the district, including prosecution practices in the district for improper entry, illegal reentry, and alien smuggling cases. We also provide information on the location of the federal courts in the district, U.S. Border Patrol (Border Patrol) sectors in the district, and the federal court circuit in which the district falls. Descriptions of the district reflect practices that were in place as of the date we observed prosecution practices or interviewed knowledgeable officials in the district, which generally ranged from July through November 2018. Cases filed. In this table, we show cases filed by the U.S. Attorney’s Office in the district with a lead charge of alien smuggling, improper entry, or illegal reentry from fiscal years 2014 through 2018. The lead charge is typically the most serious of the charged offenses at the time the U.S. Attorney’s Office files the case, according to Executive Office for U.S. Attorneys (EOUSA) officials. Table 9 describes these offenses. Dispositions. In this table, we show the dispositions of those cases with a lead charge of alien smuggling, improper entry, or illegal reentry from fiscal years 2014 through 2018, as of September 30, 2018, based on the year the case was filed by the U.S. Attorney’s Office. We have included a “pending” column for those cases that did not have a disposition, as of September 30, 2018. Apprehensions and prosecution referrals. In this table, we show prosecution referrals and declinations for those single adults that Border Patrol apprehended from fiscal years 2014 through 2018. In particular, this table includes single adults that Border Patrol apprehended and processed as an individual apprehension, not as a member of a family unit. The U.S. Attorney’s Office decides whether to accept or decline each case that Border Patrol refers for prosecution. We show those cases (each apprehended individual is one case) that Border Patrol referred to the U.S. Attorney’s Office for prosecution and the number of such cases that the U.S. Attorney’s Office declined to prosecute. Individuals whose immigration-related criminal cases are declined by a U.S. Attorney’s Office may be processed in administrative removal proceedings. Nationality and prosecution referrals. In this table, we show the number of prosecution referrals from Border Patrol to U.S. Attorneys’ Offices, by country of nationality. These include both referrals that U.S. Attorneys’ Offices accepted and those that U.S. Attorneys’ Offices declined. We also show, by country of nationality, the percent of individuals who were apprehended and referred for prosecution compared to all those apprehended. For example, if 100 Mexican nationals were apprehended and 50 were referred for prosecution, 50 percent of Mexican nationals apprehended were referred for prosecution. Cases filed by month. In this figure, we show the cases filed with lead charges of alien smuggling, improper entry, or illegal reentry each month from October 2016 through September 2018. We also show the timing of the Attorney General’s April 2017 memorandum, which prioritized immigration enforcement, and the Attorney General’s April 2018 memorandum, which instructed prosecutors on the southwest border to accept all improper entry referrals, to the extent practicable. Volume constraints: Yes; generally 75 improper entry cases per day in Tucson and 30 in Yuma. One-day improper entry prosecutions: No Volume constraints: Yes; daily generally 40 to 52 improper entry initial appearances in San Diego and 20 in El Centro. Alien smuggling (8 U.S.C. § 1324): According to U.S. Attorney officials, alien smuggling cases are labor intensive and require significant documentation. Percentages may not add to 100 due to rounding. Year 2014 One-day improper entry prosecutions: No Improper entry (8 U.S.C. § 1325): Las Cruces (New Mexico’s border court) resumed improper entry prosecutions for individuals with no criminal history in June 2017 after generally not prioritizing them from 2014 through 2017. In general, improper entry defendants make an initial appearance, are remanded to U.S. Marshals custody, and return to court 3 to 4 days later. At the second hearing, most plead guilty and are sentenced. Alien smuggling (8 U.S.C. § 1324): Witnesses in alien smuggling cases are generally not detained in New Mexico. Percentages may not add to 100 due to rounding. Year 2014 Volume constraints: No Alien smuggling (8 U.S.C. § 1324): In general, alien smuggling cases with sufficient evidence are accepted for prosecution; prior to 2017, the USAO generally declined referrals involving fewer than 6 smuggled aliens. Percentages may not add to 100 due to rounding. Year 2014 Volume constraints: No Alien smuggling (8 U.S.C. § 1324): Court officials in Del Rio attributed the increase in alien smuggling cases to a change in smuggler practices – from drug smuggling to human smuggling. Percent of those apprehended referred for prosecution Year Guatemala Mexico 2014 71 73 Appendix III: Nationality of Defendants This appendix provides additional detail on the nationality of defendants for improper entry, illegal reentry, and alien smuggling cases filed in U.S. southwest border federal judicial districts from fiscal years 2014 through 2018. We analyzed the nationality of defendants in cases filed with a lead charge of 8 U.S.C. § 1325 (improper entry), 8 U.S.C. § 1326 (illegal reentry after removal, or illegal reentry), and 8 U.S.C. § 1324 (alien smuggling) from fiscal year 2014 through fiscal year 2018 and for fiscal year 2018. Our analysis of Executive Office for U.S. Attorneys (EOUSA) data indicates that the majority of defendants for cases filed with a lead charge of improper entry and illegal reentry from fiscal year 2014 through 2018 were Mexican nationals. The majority of defendants in cases filed with a lead charge of alien smuggling over this time period were U.S. nationals. Improper entry (8 U.S.C. § 1325): From fiscal years 2014 through 2018, the majority of defendants in cases filed with a lead charge of improper entry were Mexican nationals. Our analysis of EOUSA data indicates that, in fiscal year 2018, the proportion of improper entry defendants who were Mexican nationals was lower than the fiscal year 2014 through 2018 time period, and the proportion of improper entry defendants who were Honduran or Guatemalan nationals was higher than the fiscal year 2014 through 2018 time period. The number of improper entry defendants who were Nicaraguan nationals increased substantially from fiscal year 2017 to 2018—from fewer than 70 defendants in fiscal year 2017 to more than 900 in fiscal year 2018. Figure 13 illustrates the nationalities of defendants with cases filed with a lead charge of improper entry, both from fiscal years 2014 through fiscal year 2018, and in fiscal year 2018. Illegal reentry (8 U.S.C. § 1326): The majority of defendants with cases filed with a lead charge of illegal reentry after removal from fiscal years 2014 through 2018 were Mexican nationals. Our analysis of EOUSA data indicates that, in fiscal year 2018, the proportion of illegal reentry defendants who were Mexican nationals was lower than in the fiscal year 2014 through 2018 time period, and the proportion of illegal reentry defendants who were Honduran or Guatemalan nationals was higher than in the fiscal year 2014 through 2018 time period. Figure 14 illustrates the nationalities of defendants with cases filed with a lead charge of illegal reentry, both from fiscal years 2014 through fiscal year 2018, and in fiscal year 2018. Alien smuggling (8 U.S.C. § 1324): The majority of defendants in cases filed with a lead charge of alien smuggling from fiscal year 2014 through fiscal year 2018 were U.S. nationals. Our analysis of EOUSA data indicates that, in fiscal year 2018, the proportion of defendants for alien smuggling who were U.S. nationals was lower than in the fiscal year 2014 through 2018 time period. Figure 15 illustrates the nationalities of defendants with cases filed with a lead charge of alien smuggling, both from fiscal years 2014 through fiscal year 2018, and in fiscal year 2018. Appendix IV: Immigration-Related Prosecutions in Non-Southwest Border Districts This appendix provides additional detail on cases filed in the 89 non- southwest border judicial districts with a lead charge of 8 U.S.C. § 1325 (improper entry), 8 U.S.C. § 1326 (illegal reentry after removal), or 8 U.S.C. § 1324 (alien smuggling) from fiscal year 2014 through fiscal year 2018. Specifically, this appendix analyzes the number of cases filed with one of these lead charges in every district but Arizona, California Southern, New Mexico, Texas Southern, and Texas Western. Our analysis of Executive Office for U.S. Attorneys (EOUSA) data indicates that illegal reentry cases comprised the majority of immigration-related offenses in non-southwest border districts. From fiscal year 2014 through fiscal year 2018, about 14 percent of cases filed with a lead charge of illegal reentry were filed in non-southwest border districts. Figure 16 illustrates the number and trends in cases filed with a lead charge of improper entry, alien smuggling, or illegal reentry in non-southwest border districts from fiscal years 2014 through 2018. Our analysis of EOUSA data indicates that cases filed with a lead charge of illegal reentry in non-southwest border districts increased by approximately 26 percent between fiscal year 2017 and fiscal year 2018. Illegal reentry cases comprised approximately 91 percent of immigration- related cases filed in non-southwest border districts from fiscal years 2014 through 2018. Table 30 illustrates the number of illegal reentry cases filed by non-southwest border district and fiscal year. Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Kathryn Bernet (Assistant Director), Mary Pitts (Analyst-in-Charge), Isabel Band, Dominick Dale, Jan Montgomery, Heidi Nielson, Hiwotte Amare, Michele Fejfar, and Eric Hauswirth made key contributions to this work.

Why GAO Did This Study In 2017 and 2018, the Attorney General directed federal prosecutors to prioritize prosecutions of immigration-related offenses, including improper entry into the United States, illegal reentry after a prior removal from the country, and alien smuggling, among other offenses. Most individuals prosecuted for such offenses are arrested by DHS's U.S. Border Patrol and referred to DOJ's USAOs for prosecution in federal court. GAO was asked to review the actions DOJ, DHS, and the federal judiciary took in response to the 2017 and 2018 memoranda. GAO reviewed (1) how DOJ prioritized prosecutions of immigration-related offenses in response to the Attorney General's memoranda, (2) what DHS and DOJ data from fiscal years 2014 through 2018 indicate about such prosecutions, and (3) resources that DOJ, DHS, and the federal judiciary used to support increased immigration-related prosecutions. GAO visited three of the five southwest border USAO districts and interviewed DOJ, DHS, and federal judiciary officials by phone from the other two districts. GAO also analyzed U.S. Border Patrol data on its arrests and prosecution referrals from fiscal years 2014 through 2018; analyzed Executive Office for U.S. Attorneys data on its prosecutions from fiscal years 2014 through 2018; and reviewed relevant laws and DOJ, DHS, and federal judiciary policies, operational guidance, and budget data. This is a public version of a sensitive report that GAO issued in August 2019. Information that DOJ, DHS, or the federal judiciary deemed sensitive has been removed. What GAO Found Department of Justice (DOJ) U.S. Attorney's Offices (USAO) in all five districts along the southwest border—Arizona, California Southern, New Mexico, Texas Southern, and Texas Western—have adopted prosecution priorities aligned with the Attorney General's prioritization of criminal immigration enforcement. In particular, all five USAOs prioritized misdemeanor improper entry cases in response to the Attorney General's 2017 and 2018 memoranda. Some USAOs, such as Arizona, were able to quickly increase such prosecutions using existing practices. In other districts, such as California Southern, USAOs had to establish new practices in coordination with other stakeholders in the federal criminal prosecution process—including the Department of Homeland Security (DHS), other DOJ components such as the U.S. Marshals Service (USMS), and the federal judiciary—before they could begin accepting a significant number of improper entry cases. Note: The lead charge is typically the most serious charged offense at the time the case is filed. The number of improper entry cases more than doubled from fiscal year 2017 (about 27,000) to fiscal year 2018 (about 62,000). In fiscal year 2018, about 84 percent of all improper entry cases filed were completed in districts with one-day improper entry court proceedings. In these proceedings, the initial hearing, presentation of evidence, plea, and sentencing took place in one day or less. DOJ, DHS, and the federal judiciary realigned resources to support the prosecution priorities outlined in the 2017 and 2018 memoranda, including personnel and physical space. In addition, agencies temporarily surged personnel to the southwest border. For example, USMS reassigned personnel from other enforcement areas to judicial security duties to support increased immigration-related prosecutions.

gao_GAO-20-224

gao_GAO-20-224_0

Background When complete, JWST will be a large, deployable space telescope, optimized for infrared observations. It is the scientific successor to the aging Hubble Space Telescope launched 29 years ago. JWST is being designed for a 5-year mission to find the first stars, study planets in other solar systems, search for the building blocks of life elsewhere in the universe, and trace the evolution of galaxies from their beginning to their current formation. JWST is intended to operate in an orbit approximately 1.5 million kilometers—or 1 million miles—from Earth. With a 6.5-meter (21.3 foot) diameter mirror, JWST is expected to operate at about 100 times the sensitivity of the Hubble Space Telescope. Its science instruments are designed to observe faint infrared sources and therefore are required to operate at extremely cold temperatures. To help keep these instruments cold, the JWST project will rely on a multi-layered, tennis court-sized sunshield to protect the mirrors and instruments from the sun’s heat. The JWST project is divided into three major segments: observatory, ground, and launch. When complete, the observatory segment will include several elements (Optical Telescope Element, Integrated Science Instrument Module, and spacecraft) and major subsystems (sunshield and cryocooler). Additionally, JWST is dependent on software to deploy and control various components of the telescope, and to collect and transmit data back to Earth. The elements, major subsystems, and software are being developed through a mixture of NASA, contractor, and international partner efforts. See figure 1 for the elements and major subsystems of JWST and appendix I for more details, including a description of the elements, major subsystems, and instruments. JWST depends on more deployment events—steps after launch that configure the observatory for its mission and place it in orbit—than a typical science mission. Due to the observatory’s large size, it is nearly impossible to perform deployment tests of the fully assembled observatory in a thermal vacuum chamber to simulate the space environment, so the verification of deployment elements is accomplished by a combination of lower level component tests in flight-simulated environments; ambient deployment tests for subsystem, element, and observatory levels; and detailed analysis and simulations at various levels of assembly. Figure 1 shows the multiple layers of integration and testing for major components of the JWST observatory. For the majority of work remaining, the JWST project is relying on two contractors: Northrop Grumman and the Association of Universities for Research in Astronomy’s Space Telescope Science Institute. Northrop Grumman plays the largest role, developing the sunshield, the Optical Telescope Element, the spacecraft, and the Mid-Infrared Instrument’s cryocooler, in addition to integrating and testing the observatory. Space Telescope Science Institute’s role includes soliciting and evaluating research proposals from the scientific community, and receiving and storing the scientific data collected, both of which are services that the Institute currently provides for the Hubble Space Telescope. Additionally, the Institute is developing the ground system that manages and controls the telescope’s observations and will operate the observatory on behalf of NASA. JWST will be launched on an Ariane 5 rocket, provided by the European Space Agency. History of Cost Growth and Schedule Delays The JWST program has a history of significant schedule delays and project cost increases, which resulted in both the 2011 and 2018 replans. Prior to approving the project’s development, cost estimates for JWST ranged from $1 billion to $3.5 billion, with expected launch dates ranging from 2007 to 2011. Due to early technical and management challenges, contractor performance issues, and low levels of cost reserve, the JWST program experienced schedule overruns, launch delays, and cost growth. The program underwent a replan in September 2011 and then a rebaseline; further, Congress placed an $8 billion cap on the formulation and development costs for the project. However, in June 2018, after a series of launch delay announcements due to technical and workmanship issues identified during spacecraft element integration, NASA notified Congress that it had revised the JWST program’s cost and schedule estimates again. NASA estimated that it would now require $828 million in additional resources over the program’s lifecycle and 29 more months beyond the estimates agreed to in the 2011 rebaseline to complete the project. Since the project’s costs and schedule were baselined in 2009, costs have increased by 95 percent and its launch date has been delayed by over 6.5 years. Prior to this more recent replan, NASA established an Independent Review Board (IRB) in April 2018, comprised of technical experts from outside the JWST program to evaluate all factors that may affect the successful completion of remaining mission steps. The board released its final report in May 2018 in which it made 32 recommendations that address a range of technical, organizational, and other factors. The IRB took into account varying technical and workmanship errors, human mistakes, adequacy of integration and test staff, and other considerations when it analyzed the project’s organizational and technical issues. The IRB recommended, among other actions, that the project conduct an audit to identify potential embedded design flaws; establish corrective actions to detect and correct human mistakes during integration and testing; establish a coherent, agreed-upon, and factual narrative on project status and communicate that status regularly across all relevant stakeholders; and, finally, augment integration and test staff to ensure adequate long-term staffing and improve employee morale. These recommendations also included reconsidering the proposed launch date. In March 2019, we found that NASA had considered many of the program’s risks while developing its 2018 replan schedule and cost baseline but recommended that additional analysis be completed to provide NASA and Congress with better insight into project resourcing and affordability. A Joint Cost and Schedule Confidence Level is an integrated analysis of a project’s cost, schedule, risk, and uncertainty, the result of which indicates a project’s likelihood of meeting cost and schedule targets. The project did not complete such an analysis as part of its second replan. NASA policy says this tool may be used to inform planning. Though not required by NASA policy, we recommended that one be conducted given the long history of program challenges and the significant and complex integration events that still needed to be completed. NASA agreed with our recommendation and completed this analysis in October 2019. GAO plans to conduct a separate, more detailed engagement on this analysis and its findings in the future. See appendix II for more information on this and other GAO recommendations. Schedule and Cost Reserves for NASA Projects The JWST project, like other complex development efforts we have reviewed, faces numerous risks and potential unforeseen technical challenges, which often become apparent during integration and testing. To accommodate unanticipated challenges and manage risk, projects include extra time in their schedules, referred to as schedule reserve, and extra funds in their budgets, referred to as cost reserve. Schedule reserve is allocated to specific activities, elements, and major subsystems in the event of delays or to address unforeseen risks. Each JWST element and major subsystem has been allocated schedule reserve. When an element or major subsystem exhausts schedule reserve, it may affect schedule reserve on other elements or major subsystems whose progress is dependent on prior work being finished. Cost reserve is additional funding within the project manager’s budget that can be used to address and mitigate unanticipated issues for any element or major subsystem. Goddard—the NASA center with responsibility for managing JWST— issued procedures detailing the cost and schedule reserve requirements for formulating and executing spaceflight programs. When NASA constructed its 2018 replan for the JWST project, it took into account the remaining integration and test activities planned prior to launch, known technology challenges that presented risks to schedule, as well as potential future risks. The project’s replan reflected a planned schedule reserve above the level indicated by Goddard policy, which would have been approximately 5 months at that time. Instead, the new schedule included a total of 293 days or 9.6 months of schedule reserve, with approximately 6 months of this reserve to be managed at the project level and the remainder held by the program at NASA headquarters. Following the replan, the project and the contractor worked toward a launch date in November 2020, which would have required none of the schedule reserve managed at the NASA headquarters level. However, the committed launch date under the replan, where all available schedule reserve is utilized, is now March 2021. JWST’s Use of Award Fees NASA’s cost-plus-award-fee contract with Northrop Grumman has spanned approximately 17 years, during which time there have been significant variances in performance. These types of contracts are suitable when uncertainties in scope of work or cost of services prevent the use of contract types where prices are fixed. Award fee contracts provide contractors the opportunity to obtain additional fee beyond the costs charged to the government for enhanced levels of performance in areas identified in the contract’s award fee plan. Award fees may be used when key elements of performance cannot be defined objectively, and, as such, require the project officials’ judgment to assess contractor performance. For JWST’s contract with Northrop Grumman, these areas include cost, schedule, technical, and business management and are established in the contract’s award fee plan, which allows for the award of a scaled fee based on assessed performance. This plan has been revised over the life of the contract to incentivize performance in certain areas, but it has always required Northrop Grumman to meet a minimum standard to receive any award fee. Over the course of the JWST contract, nearly $250 million dollars will have been available to Northrop Grumman through this incentive. We have found that when NASA and the contractor have made revisions to fee evaluation criteria to focus on certain aspects of performance, the contractor has been responsive to the new criteria during its work on the JWST project. Little Margin for Error Remains with Challenging Integration and Test Work Ahead Though the JWST project has made significant progress since our last report in March 2019, technical challenges have required the use of most of the project’s available schedule reserve. According to NASA officials, the contractor has found ways to replenish reserve, but NASA is still reviewing some of these methods and the project continues to work through significant integration and testing events with less than a quarter of the schedule reserve allotted to it in June 2018. The technical challenges have resulted in prolonged employment of the contractor workforce, which is the primary driver for increased costs. NASA Has Completed Key Testing and Integration Steps and Continues to Address Known Risks Following the June 2018 replan, the project has achieved a number of integration and testing milestones and has taken steps to address previously identified technical challenges. Since our March 2019 JWST report, the program has completed testing on the individual component elements of the observatory and has integrated them to start observatory level testing, the last of five phases of integration and testing. Leading up to observatory integration, the project completed thermal vacuum testing of the spacecraft element in May 2019. This testing helped to ensure that JWST hardware will function properly in the vacuum of space and withstand significant temperature variations during deployment and operation, and provided data to corroborate modelling on which the observatory’s mission is based. Further, the project completed the last major testing milestone for optical telescope and science instrumentation elements—deployment of the secondary mirror assembly—in August 2019. This secondary mirror focuses the light collected by the 18 hexagonal primary mirrors of JWST into a beam and directs it toward scientific equipment aboard the observatory. Integration of the observatory components was completed in August 2019, and the program has deployed the sunshield as part of observatory integration and testing. NASA has also taken steps to address challenges noted in our previous reports. For example, In February 2018, we found that Northrop Grumman planned to modify the design of the sunshield’s membrane tensioning system in response to a risk of a cable snagging during deployment. NASA approved this redesign in May 2019 and employed a new approach to cable management that involves modification and replacement of certain cable clips and routing cables differently to manage slack that could cause snags. We found in March 2019 that the project office identified concerns that trapped air in the folded sunshield membrane could put too much stress on the observatory when the launch vehicle fairing depressurizes—the fairing is the part of the rocket that encapsulates JWST during flight. NASA, Northrop Grumman, European Space Agency, and European vendors responsible for operating and producing the launch vehicle have worked together to study this issue and have designed vents for the fairing that will mitigate the risk of damage to JWST. The new fairing vent design is expected to be tested aboard a rocket planned to launch in the spring of 2020. Technical Challenges Have Significantly Reduced Schedule Margin, with Considerable Integration and Testing Ahead Despite the major accomplishments of the past year, the program has identified new technical issues that present risk for meeting the 2018 replan’s schedule requirements. Multiple technical issues have contributed to the use of schedule reserve since the June 2018 replan, but two identified in March and April 2019 have had the most significant effect. The program identified two significant anomalies during pre-testing events for the spacecraft element’s thermal vacuum testing, which first delayed thermal vacuum testing and then required additional time for investigation and implementation of solutions. Specifically, a traveling wave tube amplifier and a command and telemetry processor had errant powering issues during testing. These are important components of the observatory’s communication systems that enable JWST to send large amounts of science data and telemetry to the ground segment at high speed. Though the anomalies occurred at the same time and were both power- related, NASA does not believe they are related and has initiated separate review boards to determine solutions. The amplifier failure is attributed to workmanship issues on the part of a subcontractor. As of October 2019, the exact cause of the processor anomaly remained under investigation, but the electrical problem had been isolated to faults within specific circuit cards. NASA has taken steps to address the risks presented by both anomalies: it has received replacement amplifiers and has upgraded and tested an engineering model processor to replace the faulty one aboard the observatory if necessary. As a result of technical issues discovered since the June 2018 replan, the JWST program has had to use significantly more schedule reserve than it planned to and has been working towards the replan’s formally committed launch date of March 2021. As of October 2019, the project had used 224 days of schedule reserve, or about 76 percent of the total project and program-held schedule reserve incorporated into the June 2018 replan. All project-held schedule margin was used by March 2019, a point at which the project would have retained approximately 4 months of reserve according to its original plan. At one point since our March 2019 report, the project had as little as 18 percent of its total schedule reserve left, but contractor-led corrective action plans regained time through found efficiencies. As a result of these challenges, the project’s reserve fell below what is indicated by Goddard policy. NASA determined in May 2019 that the November 2020 launch date that the project had hoped to achieve was no longer feasible, and switched focus to meeting the committed launch date of March 2021. Figure 2 shows the level of planned reserve for JWST, reserve indicated by Goddard policy, and the project’s actual use of schedule reserve. Since then, however, the JWST project has determined that the March 2021 launch readiness date may not be feasible either, based on a detailed assessment of risks, costs, and schedule. In October 2019, the project completed a joint cost and schedule confidence level analysis in response to a GAO recommendation made in a previous report on the JWST program. Because of schedule delays resulting from technical challenges coupled with remaining risks faced by the project, the analysis assessed only a 12 percent confidence level for the project’s ability to meet the March 2021 launch readiness date. NASA typically establishes its cost and schedule baseline commitments at 70 percent confidence level. According to the analysis, this 70 percent baseline confidence level is associated with a July 2021 launch date. The project does not currently intend to change the launch readiness date in response to this analysis alone, but plans to assess the feasibility of the launch readiness date again in spring 2020 after significant technical tasks are completed. NASA and Northrop Grumman have a plan to recover schedule reserve but certain portions of the plan remain under technical review. Following the amplifier and processor anomalies, Northrop Grumman developed a corrective action plan to recover schedule reserve, and the contractor and NASA continue to look for ways to gain efficiencies. In June 2019, Northrop Grumman suggested a number of potential schedule optimization steps that were reviewed by NASA management. Northrop Grumman has begun to be implement some of these steps. If all steps are taken, the contractor estimates 65 days of project schedule will be recovered, nearly doubling the amount of reserve available to the project when the anomalies were discovered. Among the efforts described in this corrective action plan are to streamline aspects of vibration testing and to modify build and repair schedules so that a major panel on the spacecraft will only have to be opened once. Combined, these two steps would save an estimated 46 days. However, officials noted that for the plan for a single panel opening to remain viable, corrective actions for the amplifier and processor replacements would need to remain on schedule. The project continues to review some of Northrop Grumman’s proposed efficiencies, but more than half of these schedule savings have already been incorporated into the schedule reserve forecast. The project also continues to identify and monitor risks that could potentially result in further use of schedule margin. As suggested by the IRB, the project has led a number of audits looking for embedded risks. As of November, most of the audits planned have been completed and NASA identified some new risks. The following are some of the risks the project is monitoring that could affect schedule: The project found that certain bolts, determined to be deficient on another Northrop Grumman program, were used during the construction of the observatory. A study of this issue found that the bolts used did not meet specifications and could pose a mechanical strength risk. The unused bolts have been identified and isolated, but 501 were installed in the observatory. NASA is performing strength testing to determine if the bolts are strong enough, but some of the deficient bolts may need to be replaced, pending the findings of these tests. The project reported in August 2019 that grounding straps on the spacecraft’s momentum flap came loose during vibration testing. This flap will act as balance against solar pressure that could cause unwanted movement of the observatory while in orbit. Observatory- level vibration testing cannot begin until the flap is removed, repaired, and replaced aboard the spacecraft. In September 2019, the project found that a non-explosive actuator on one of its membrane retention devices did not fire as planned. These devices, which help to unfurl the sunshield of the spacecraft, are supposed to be electrically redundant, but only one of the two mechanisms used to fire the actuator worked during the test. The program reports that there are approximately 180 actuators on the JWST and the failure of any one of these actuators could result in the total loss of JWST science mission objectives. If the redundancy for the actuators is reduced, it would have a major impact on system reliability. The project is evaluating whether it needs to replace certain membrane retention devices that may not be able to withstand the coupled pressure placed upon them by the launch and newly designed fairing ventilation. Testing in the past did not account for all aspects of the pressures placed upon this hardware during launch and spaceflight. The project indicated that it is completing an analysis to determine if stronger devices need to be installed. The JWST project office reviews and reports on these and other risks monthly. As of October 2019, the project is tracking 50 risks—three more than when we last reported on JWST—of which 12 continue to be assessed as moderate concerns. Of the 50, 23 have been assessed to be at acceptable levels of risk but continue to be monitored should changes affect their status. For example, the risk associated with cabling within the sunshield was elevated in October 2019 when the project found that further testing was needed to ensure slack did not present an unacceptable threat to the spacecraft during deployment. Finally, nine of the 50 risks currently tracked by the project are related to the more than 300 single points of failure aboard the observatory. The project must conduct significant integration and testing activities in the coming months that could present further challenges. Our previous work on major NASA acquisition programs found that integration and testing is the phase when challenges are most likely to be found and schedules can slip. The science elements and the spacecraft have only recently been integrated. NASA will have to manage seven top-level integration and testing steps between October 2019 and December 2020 to include observatory-level vibration testing, sunshield deployment and stow, and electrical testing and repairs. Currently, this will all have to be completed with a diminished amount of schedule reserve. Northrop Grumman and NASA officials we interviewed agreed that no other major complication, such as those on the scale of the traveling wave tube amplifier and command and telemetry processor anomalies, can happen without putting the March 2021 launch date in jeopardy. Technical Challenges May Also Drive Additional Costs for the Project As we found in March 2019, changes to JWST project’s life-cycle cost estimate are principally driven by schedule extension, which requires keeping the contractor’s workforce longer than expected to complete integration and testing. We also found that NASA’s cost estimate for the 2018 replan was based on a more gradual workforce reduction schedule than previously used by the Northrop Grumman. NASA continues to forecast an overall reduction in contractor and government workforce following the project’s launch readiness date with continued, steady support by the Space Telescope Science Institute during remaining development and post-launch phases of the program (see figure 3). The program reports that cost reserve is generally sufficient for planned work but technical challenges could cause workforce costs to increase. The cost and schedule analysis completed by the project in October 2019 indicated that the project will not exceed the cost commitment established in the 2018 replan even if launch is delayed further by a few months. According to officials, funding is sufficient to continue work even if the launch date slips 3-4 months past the March 2021 launch date. However, the technical issues identified during integration and testing activities have required the contractor workforce to remain engaged, instead of drawing down as planned. Rather than see a temporary drop in contractor work hours as hardware deliveries were completed ahead of observatory-level testing and integration activities, the project has maintained contractor workforce levels to address the issues described above. The contractor now forecasts approximately 15 percent more workforce hours between 2019 and 2022, the year following launch (see figure 4). Approximately $133 million in cost reserve funding will be used by the project over the next 2 fiscal years to accommodate increasing workforce retention costs. NASA Has Addressed Recommendations and Sustained Oversight Improvements Since 2018 Since the June 2018 replan, NASA has taken steps to improve the JWST project by implementing Independent Review Board (IRB) recommendations, pursuing other oversight initiatives, and continuing to incentivize contractor performance through the use of award fees. NASA addressed all IRB recommendations even though the agency did not always agree with the IRB on the specific steps needed to address the recommendations. Further, NASA has sustained, and in some cases expanded, the oversight initiatives that were started prior to our last report. The cost plus award fee contract used for JWST development efforts provides the project with a means to incentivize contractor performance related to cost, schedule, technical, and business management goals. Since the 2018 replan, Northrop Grumman’s award fee evaluations have improved but remain below its average for the contract. NASA Has Completed Its Implementation of IRB Recommendations NASA assessed all IRB recommendations as closed in October 2019. The IRB made 32 recommendations covering a range of topics from improving communication with stakeholders to identifying embedded problems. NASA implemented its recommendation to establish March 2021 as the committed launch date for JWST through the June 2018 replan. Responsibility for implementing the remaining 31 recommendations was split among headquarters, the program office, and the project. The 10 headquarters- and program-level recommendations covered high-level recommendations dealing with entities outside of the project or communication between upper-level NASA management and the project. The remaining 21 recommendations were implemented at the project level and included lower-level actions related to assessing, preparing for, and improving day-to-day work. NASA assessed most recommendations as implemented prior to an IRB follow-up assessment, but the IRB found that more work was required for some to completely align with the board’s intent. In February 2019, the IRB found that the steps NASA took for approximately one-third of its recommendations were either inadequate or needed additional work, with the remainder found to be appropriate. Specifically, the IRB categorized 21 of the recommendation responses as appropriate, eight responses as appropriate with additional work needed, and three responses as inadequate. The IRB’s monitoring of the JWST project ended with the February 2019 follow up (see figure 5). Though NASA agreed with the intent of all the IRB recommendations, it took a different approach than described in the IRB report when implementing the three recommendations where the agency’s response was assessed to be inadequate. NASA conducted additional work for the majority of recommendations assessed by the IRB to be incomplete. However, NASA determined that a few of the IRB recommendations managed at the headquarters level should not be implemented the way they were delineated in the IRB report. Specifically, The IRB found that the JWST reporting structure was complex, confusing, and ineffective. The IRB made two recommendations for NASA to update its reporting chain. The IRB believed the Science Mission Directorate Associate Administrator should have responsibility of the entire JWST program and the Goddard Space Flight Center Director should be responsible for all aspects of the JWST project. The IRB asserted that restricting the involvement of the Goddard director will reduce the probability of JWST success. NASA agreed that it is important to have clear organizational roles and responsibilities but had a difference of opinion about the best course of action. In November 2018 and July 2019, NASA announced updates to the JWST reporting structure. However, both times it reduced the role of the Goddard director in favor of more direct line of accountability from the JWST program to the Science Mission Directorate Associate Administrator and the NASA Associate Administrator. NASA asserts that these changes will provide more clear accountability for program performance and allow for expedited decision making. The IRB recommended that NASA’s Launch Services Program should have accountability for the JWST launch. NASA has taken actions to increase the involvement of the Launch Services Program but NASA maintains that it is not prudent or possible for the Launch Services Program to be accountable for the launch because the European Space Agency is contributing the launch vehicle and managing the launch. The IRB recognized the unique circumstances of using an international launch vehicle but continued to assert the importance of Launch Services Program accountability. A minority of IRB members were of the opinion that NASA took appropriate action. NASA Has Sustained Key Oversight Improvements and Made Additional Improvements Since 2018 NASA has sustained increased oversight and involvement with Northrop Grumman following the announcement of an anticipated cost cap breach in March 2018. Previously reported improvements included both the implementation of IRB recommendations and the pursuit of self-initiated activities, like greater NASA on-site coverage and Northrop Grumman’s culture change campaign designed to shift focus toward quality assurance. Our March 2019 report, provided examples of these changes and initiatives. Table 1 below provides a summary of our previous report findings and the current status of the changes NASA and Northrop Grumman made in providing oversight and ensuring quality. Since we last reported in March 2019, NASA has made additional oversight changes to further enhance communication with and oversight of the contractor. Most of these changes emphasize greater involvement of NASA specialists in meetings and reviews. NASA officials reported that its increased presence with the contractor has had positive effects for both ensuring project outcomes and increasing morale of the government and contractor workforce. For example, NASA integration and testing leadership is present and embedded in Northrop Grumman’s meetings— directly participating in planning sessions, reporting, and reviews of failures and anomalies. As a result, the project was able to plan for early integration of the observatory and completed key integration activities without being the primary driver of the project’s schedule. According to officials, expanded participation has helped to ensure more realistic exercises that include procedural concerns as well as engineering considerations. NASA officials said that the increased participation has allowed NASA input to be incorporated early—potentially reducing issues in the future. Further, NASA officials believe that the consistent presence of NASA personnel has improved morale—an item highlighted by the IRB—and helped foster greater unity of effort between government and contractor workforces. Contractor Award Fee Has Fluctuated over Time, but Performance Has Improved Recently NASA has regularly assessed contractor performance through award fee assessments since the beginning of the contract in 2002. Award fee documentation over the course of the Northrop Grumman contract indicates that contractor performance was assessed as below its average before periods of significant cost and schedule growth. On average, Northrop Grumman has been rated as very good with about three-fourths of evaluations assessing its performance as either excellent or very good. For the award fee evaluations that fall below Northrop Grumman’s average score, cost performance has contributed to the majority of these dips and schedule performance has contributed to almost half. In particular, schedule performance has reduced the contractor’s overall evaluation for all award fee periods since April 2017. The latest dip below the contractor’s average preceded lifecycle cost growth of $828 million and schedule growth of nearly 2.5 years (see figure 6). Since our March 2019 report, Northrop Grumman’s ratings have improved but remain below its average. For the award fee period from October 2017 through March 2018, Northrop Grumman received an unsatisfactory rating, which resulted in the contractor receiving no award fee for the first and only time in the life of the contract. The unacceptable rating was driven by cost and schedule performance—including the anticipation of breaching the $8 billion congressional cost cap established in response to the 2011 rebaseline. In the following two periods, Northrop Grumman has improved its evaluation, but schedule performance remains a concern. During the last award fee period assessed, NASA was internally managing to a November 2020 launch date. Shortly after the award fee period ended, the project found it could no longer support the November 2020 date and began managing to the March 2021 launch date. Agency Comments and our Evaluation We are not making recommendations in this report. We provided a draft of this report to NASA for comment. NASA provided technical comments that, among other things, clarified implementation of schedule recovery steps and updated progress on observatory repairs. We incorporated suggested technical changes as appropriate. We are sending copies of this report to the appropriate congressional committees, the NASA Administrator, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions on matters discussed in this report, please contact me at (202) 512-4841 or chaplainc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Elements and Major Subsystems of the James Webb Space Telescope (JWST) Observatory Appendix II: Status of Previous GAO Recommendations on Management of the James Webb Space Telescope Program In its previous reports on the James Webb Space Telescope (JWST), the GAO has made several recommendations. These recommendations are listed below. Comments reflect the status of the program at the time GAO closed the recommendation. Appendix III: List of Independent Review Board (IRB) Recommendations Appendix III: List of Independent Review Board (IRB) Recommendations Telescope (JWST) launch success at the same level of responsibility they have for U.S. launches, or the National Aeronautics and Space Administration (NASA) should contract with Aerospace Corporation for similar accountability. 2 The Goddard Space Flight Center (Goddard) and Northrop Grumman Project Offices should be established as consistent and factual source of all JWST mission status 3 Communications of status and details appropriate for stakeholders need to be presented clearly and frequently. 4 NASA headquarters should be responsible for developing a “communication plan” (messaging strategy) for JWST. 5 Communicating complexity, risk, and science return for JWST is critically important. 6 Use the same criticality and assessment charts for all JWST reporting. 7 NASA should implement a JWST reporting structure where the Science Mission Directorate Associate Administrator has responsibility for the entire JWST program and the Goddard Space Flight Center Director is responsible for all aspects of the project. 8 NASA should revise NASA policy directives to be consistent with the recommendation. 9 Assure consistent, sustained and meaningful engagement of the Science Working Group (SWG). 10 Appoint an executive committee of NASA-selected members of the SWG to act as conduits to the broader community on mission challenges. 11 NASA should designate a Commission Manager. 12 NASA should implement sunshield hardware and simulation elements to aid in sunshield anomaly identification and resolution. 13 Northrop Grumman should establish corrective actions in1) processes, 2) training, 3) personnel certification, 4) discipline to ensure individual accountability and 5) a failure-proof “safety net” through a robust testing, analysis, and inspection process. 14 Goddard and Northrop Grumman should conduct an audit including forensic engineering, hardware pedigree assessment, drawing checks, etc., to identify potential embedded problems. 15 Goddard should conduct an audit of the JWST project residual risk, reviewing the objective evidence of (a) the completed Test As You Fly and Single Point Failures mitigation plans, and (b) failure corrective action effectiveness to determine the “as built” residual risk. 16 The project should reconcile the “as built” residual risk with the expected “as designed” residual risk. 17 NASA should define security requirements and plan for JWST transport to launch site. 18 Develop contingency operations and sparing plan for spacecraft/launch site operations. 19 Develop “pathfinder” JWST simulator and contamination protection systems for integration “dry runs.” 20 Assess shipping vessel contamination environment and develop contingency plans for off-nominal shipping operation. 21 It is critically important that Goddard JWST Project Office maintain responsibility and provide adequate support to ensure Space Telescope Science Institute (STScI) mission operations readiness 22 The Project should review all simulators/testbeds and required usage against pre-launch tests and rehearsals, post-launch deployment anomaly resolution, fault isolation, and correction. 23 The Goddard JWST Project Office should develop a staffing plan that meets the needs of integration and test and operational readiness. 24 The project should develop and approve a transition plan that defines the level of mission operations responsibility for STScI as a function of time with independent gate reviews at transition points. 25 Management should unambiguously emphasize the priority of mission success to “working level” personnel. 26 Employees must feel empowered to stop or slow down if the pace or procedures can jeopardize mission success. 27 NASA should assess “top ten” mission success enhancements and implement where appropriate. 28 Responsible Design Engineers should be involved and responsible for their element through the successful commissioning of the observatory. 29 The project should augment integration and test staff; this is critically important to execute the integration and test program. 30 Augment integration and test staff to achieve more realistic work schedules. 31 Implement strategies for improving team morale, such as periodic science lectures for Northrop Grumman personnel and families. 32 The Webb IRB recommends the launch date be established as March 2021 (based upon the Project’s 5/18 assessment of the impact of the membrane cover assembly acoustic anomaly). Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Raj Chitikila (Assistant Director), Christina Cota-Robles, Carrie Rogers, Ajani Skeete, Jay Tallon, and Thomas Twambly made key contributions to this report. Assistance was also provided by Hannah Brookhart, Brian Bothwell, Lorraine Ettaro, Emile Ettedgui, Laura Greifner, Kaelin Kuhn, Christine Pecora, Roxanna Sun, and Alyssa Weir.

Why GAO Did This Study JWST, a large, deployable telescope, is one of NASA's most complex projects and top priorities. Problems discovered during integration and testing caused multiple delays that led NASA to replan the project in June 2018. Now estimated to cost $9.7 billion, the project's costs have increased by 95 percent and its launch date has been delayed by over 6.5 years since its cost and schedule baselines were established in 2009. Prior to the replanning process, an independent review board assessed the project and made recommendations to improve performance and oversight. Conference Report No. 112-284 included a provision for GAO to assess the project annually and report on its progress. This is GAO's eighth report. This report assesses the extent to which (1) the project is executing within its revised cost and schedule targets and (2) NASA has implemented and sustained key improvements to performance and oversight established following the June 2018 replan. GAO reviewed relevant NASA policies, analyzed NASA and contractor data, and interviewed NASA and contractor officials. What GAO Found The National Aeronautics and Space Administration's (NASA) James Webb Space Telescope (JWST) project has made significant progress since GAO's last report in March 2019, such as completing testing of the observatory's individual elements and integrating them together in August 2019. However, new technical challenges have required the project to use more schedule reserve—extra time set aside in the project's schedule to accommodate unforeseen risks or delays—than planned. As of October 2019, the project had used about 76 percent of its available schedule reserve and no longer plans to launch in November 2020 (see figure). The project is now managing to a March 2021 launch date but estimates only a 12 percent likelihood that this date will be achieved. NASA plans to reassess the launch date in the spring of 2020. The project used much of the schedule reserve in April 2019 to address issues with two components needed to transmit science data to ground control. The contractor has been able to mitigate some of the schedule loss and continues to look for new efficiencies. Technical challenges also resulted in longer employment of the contractor workforce than planned, which could result in additional cost increases. NASA continues to monitor multiple, other risks that could place further schedule and cost strains on the project. Since NASA replanned the project again in June 2018, the agency has taken steps meant to improve performance and oversight. NASA has addressed all recommendations from an independent review board, but in doing so sometimes took actions that differed from those outlined in the board's report. NASA has sustained, and in some cases expanded, oversight initiatives following the revised cost and schedule commitments that, in many cases, were designed to enhance communication between the government and the contractor. What GAO Recommends GAO is not making any new recommendations at this time. GAO has made several recommendations to NASA on the management of this project in previous reports and NASA has agreed with and taken action on many of them. Most recently, in March 2019, GAO recommended that NASA complete a joint cost and schedule confidence level analysis for JWST. NASA concurred and completed the analysis in October 2019 to support a key project review.

gao_GAO-20-34

gao_GAO-20-34_0

Background Authorized Tobacco User Fee Amounts The Tobacco Control Act specifies the total amount of tobacco user fees that FDA is authorized to assess and collect each fiscal year (beginning with fiscal year 2009) and stipulates those fees must be used for FDA’s tobacco regulation activities. FDA collected about $4.5 billion in tobacco user fees from fiscal year 2010 through fiscal year 2018, according to FDA budget documents, and has ongoing authority to assess and collect $712 million from tobacco manufacturers and importers annually starting in fiscal year 2019. See table 1 for the total user fees the Tobacco Control Act authorized FDA to collect, by fiscal year. All of FDA’s activities related to regulating tobacco products—including activities aimed at preventing youth use of tobacco products, educating the public about tobacco products and the risks associated with their use, and issuing regulations on the marketing and advertising of tobacco products—are funded through tobacco user fees, as required by the Tobacco Control Act. FDA’s Center for Tobacco Products (CTP), which was established by the act, is responsible for executing FDA’s tobacco regulation responsibilities. Within CTP, the two main offices involved in carrying out FDA’s tobacco user fee responsibilities are the Office of Management and the Office of Compliance and Enforcement. CTP’s Office of Management staff duties include—but are not limited to—calculating individual tobacco manufacturer’s and importer’s market share quarterly within each tobacco product class, as well as completing FDA’s year-end reconciliation process to ensure its market share calculations for each fiscal year are based on complete and accurate data. CTP’s Office of Compliance and Enforcement staff are involved in FDA’s efforts to implement and enforce the Tobacco Control Act by (1) informing tobacco manufacturers and importers that they must pay the required quarterly tobacco user fee, if they have not done so, by the due date, and (2) working to obtain voluntary compliance, or taking advisory or enforcement actions, when manufacturers or importers continue to fail to comply with the user fee requirements. FDA’s Office of Financial Management is responsible for calculating the quarterly assessments for each tobacco product class, and for activities related to the billing and collection of tobacco user fees. For example, FDA’s Office of Financial Management generates quarterly invoices for individual manufacturers and importers based on CTP’s market share calculations. Additionally, this office processes tobacco user fee payments received and works with CTP’s Office of Compliance and Enforcement to help collect user fee payments from tobacco manufacturers and importers that do not pay a quarterly assessment by the due date. Tobacco Control Act Requirements for Assessing and Collecting Tobacco User Fees The Tobacco Control Act establishes requirements regarding the calculation, billing, and collection of tobacco user fees. Calculation. For each fiscal year, total user fees are to be allocated in 1. Class allocation. The amount of total user fees for a fiscal year (e.g., $635 million for fiscal year 2017) is allocated among the different tobacco product classes subject to user fees; this allocation is based on each class’s share of the gross domestic volume of tobacco products introduced into the U.S. market. 2. Individual allocation. The amount of user fees allocated to each manufacturer or importer is proportional to its market share within a given class of tobacco products. For example, a manufacturer with 50 percent of the cigarette market would be required to pay 50 percent of user fees allocated for the cigarette product class. The act specifies that no manufacturer or importer of tobacco products shall be required to pay a user fee in excess of the percentage share of such manufacturer or importer. See figure 1 for a summary of the tobacco user fee allocation process under the Tobacco Control Act. Notifications to each manufacturer or importer of the amount of its quarterly tobacco user fee assessments are to be sent at least 30 days before the end of the quarter for which the assessment is made. Collection. Tobacco user fee payments are due the last day of each quarter. If a manufacturer or importer does not pay its user fee assessments by the last day of the relevant quarter, the act states that tobacco product shall be deemed adulterated. FDA Regulations and Processes to Calculate, Bill, and Collect Tobacco User Fees Since the enactment of the Tobacco Control Act, FDA has issued several final rules (regulations) regarding its process to calculate, bill, and collect tobacco user fees, including the following: In 2014, FDA issued a final rule requiring tobacco manufacturers and importers to submit to FDA the information needed to calculate individual tobacco user fees, starting with fiscal year 2015. This rule applied to the four classes of tobacco products FDA initially regulated: cigarettes, snuff, chewing tobacco, and roll-your own tobacco. Fiscal year 2015 was the first year for which FDA obtained the data directly from manufacturers and importers to calculate individual tobacco user fee assessments. In 2016, FDA issued a final rule extending FDA’s regulatory authority to all tobacco products, including pipe tobacco and cigars (but excluding accessories of newly deemed products). Using its deeming authority, FDA issued another final rule requiring that pipe tobacco and cigar manufacturers and importers submit to FDA the information required to calculate user fees for these tobacco product classes. FDA began collecting tobacco user fees from the pipe tobacco and cigar classes in fiscal year 2017. See appendix I for a timeline of events related to FDA tobacco product user fees. FDA’s process for calculating, billing, and collecting user fees involves five steps. First, FDA collects the data needed to calculate the quarterly user fee allocations for each tobacco product class and, within each class, for individual manufacturers and importers. For its quarterly class allocation calculations, FDA collects data on the total volume (units) of tobacco products introduced into the U.S. market for each tobacco product class from the Department of the Treasury’s Alcohol and Tobacco Tax and Trade Bureau (TTB)—these data are published on the TTB website. FDA also collects data from individual manufacturers and importers on the volume of and federal excise taxes paid for their tobacco products introduced into the U.S. market in each product class. Tobacco companies submit these data to FDA as part of required monthly report submissions. Second, FDA uses the TTB data it collected to calculate the quarterly class allocations. Third, FDA calculates the user fees owed by individual manufacturers or importers within a given product class, based on their market share in each tobacco product class and the quarterly class allocation it previously calculated. Fourth, FDA bills—that is, generates and mails user fee invoices to—tobacco product manufacturers and importers each quarter. Fifth, FDA collects user fee payments. User fees that are not received by FDA by the last day of the quarter are considered late, and are subject to financial charges beginning 30 days past the invoice due date and for each 30-day period that the assessment remains unpaid. Figure 2 shows the steps in FDA’s process to calculate, bill, and collect user fees. Amount of Tobacco User Fees Assessed by Product Class From fiscal year 2015—the first year that FDA obtained data directly from manufacturers and importers to calculate user fee assessments—through fiscal year 2017—the most recently available data at the time of our analysis—FDA assessed and collected about $1.8 billion in tobacco user fees. During this time, the vast majority of the total user fees assessed and collected each fiscal year were from manufacturers and importers of cigarettes. See figure 3 for user fees that FDA assessed, by product class, for fiscal years 2015 through 2017. FDA Has A Process for Administering Tobacco User Fees but Has Not Completed a Key Activity to Ensure Completeness and Accuracy FDA’s Process Is Designed to Ensure User Fee Calculations, Billings, and Collections Are Complete, Accurate, and Timely FDA’s process is designed to ensure the quarterly user fees it calculates, bills, and collects each fiscal year are complete and accurate. This process is also designed to ensure user fee invoices are billed to tobacco manufacturers and importers in a timely manner and to help the agency ensure user fee payments are collected in a similar manner. Additionally, FDA has designed procedures to retroactively adjust its quarterly individual user fee calculations to include relevant excise tax data not reported to FDA at the time these calculations were completed. The agency’s year-end reconciliation process is designed to make these adjustments to ensure that the user fees assessed for a given fiscal year are complete and accurate. Calculation. FDA’s process related to its quarterly individual user fee calculations includes procedures to ensure its individual quarterly user fee assessments are complete and accurate. Tobacco manufacturers and importers provide monthly reports to FDA on the volume of and excise taxes paid on tobacco products introduced into the U.S. market, and those data are reviewed by CTP’s Office of Management for accuracy. If CTP identifies incomplete data or inaccurate reporting, it will contact the appropriate manufacturer or importer in an attempt to resolve discrepancies (e.g., differences between what the company reported to FDA and the supporting document it provided) prior to calculating individual market share for the quarterly billing cycle. However, according to agency officials, if the team is unable to resolve any discrepancies by the time it must submit market share percentages to FDA’s Office of Financial Management for the quarterly billing process, it uses the potentially incomplete or inaccurate data for its market share calculations. FDA officials stated that the agency may make adjustments to individual market shares and resulting user fees based on late or amended data it receives from manufacturers and importers after that data is received. While this is an option, FDA generally relies on its year-end reconciliation process to make all adjustments resulting from late or amended data received at one time, according to FDA officials. Billing: FDA’s process related to quarterly tobacco user fee billing includes procedures to ensure the invoices it creates for individual tobacco manufacturers and importers are complete and accurate—based on CTP’s market share percentages calculated using the monthly excise tax data submitted to FDA by manufacturers and importers—and mailed in a timely manner. For example, FDA’s billing procedures provide for quarterly user fee assessments to be calculated automatically in FDA’s Tobacco Billing Portal. Collection: FDA’s process related to the collection of quarterly tobacco user fees includes procedures to help it ensure quarterly tobacco user fee payments received are complete, accurate, and timely recorded. FDA has also designed mechanisms to identify and collect payment from tobacco manufacturers and importers who do not pay their invoices by the quarterly user fee due date (i.e., the last day of the applicable fiscal year quarter). For example, FDA has an internal system that is designed to generate alerts to warn staff of unpaid invoices that are approaching 30, 60, and 90 days past due so FDA can issue notification letters to inform the tobacco manufacturers and importers that their invoices are overdue and provide instructions for making a payment. (See appendix II for additional information on FDA’s process for the calculation, billing, and collection of tobacco user fees.) Outside of its tobacco user fee calculation, billing, and collection cycle, FDA’s procedures state that FDA will review TTB data to develop a list of current tobacco permit holders that may be subject to user fees. According to FDA officials, reviewing this list helps the agency ensure it has included all manufacturers and importers within relevant tobacco product classes in its individual quarterly user fee calculations. FDA procedures state that CTP’s Office of Management contacts the permit holders that have not reported monthly data to FDA, if identified, to inform them that (1) they are required to report monthly data to FDA for purposes of making user fee market share calculations, and (2) the permit holder may be required to pay quarterly tobacco user fees as a result of these data. User Fee Adjustments: FDA has also designed procedures to retroactively adjust its quarterly individual user fee calculations to include relevant excise tax data that were misreported or not reported to FDA at the time these calculations were completed. Individual quarterly user fee assessments are based on the market share of manufacturers and importers within each tobacco product class. As a result, FDA needs to recalculate all individual market share percentages within a given class of tobacco products if it receives new or amended data related to the excise taxes paid by manufacturers and importers in that class, to ensure compliance with the Tobacco Control Act. According to FDA’s procedures, FDA may recalculate its individual quarterly market share percentages to include changes identified by late or amended data submissions from individual tobacco manufacturers and importers, and FDA will recalculate market shares to include changes identified during its year-end reconciliation process. Late or amended data submissions. According to FDA’s procedures, FDA can receive data from tobacco manufacturers and importers that did not previously submit monthly data to FDA and were therefore excluded from FDA’s initial quarterly market share calculations. FDA can also receive late or amended data from tobacco manufacturers and importers that previously reported incomplete or inaccurate monthly data to FDA. According to FDA, in some instances, these late or amended data are data that FDA had requested during its monthly review process, but were received after FDA completed its quarterly market share calculations. According to FDA, companies may also voluntarily provide updated reports that the company itself determined were a correction to previously submitted data. Year-end reconciliation based on annual tax records from TTB and U.S. Customs and Border Protection (CBP). FDA’s procedures state that it will make annual adjustments to user fees for each fiscal year as part of its year-end reconciliation process. FDA’s procedures state that, by FDA request, TTB and CBP will provide an annual report listing the tobacco excise taxes paid by each manufacturer and importer subject to the tobacco user fee requirement. FDA officials stated that FDA submits an annual request to TTB and CBP for their records of the excise taxes paid by each tobacco permit holder in the six relevant tobacco product classes for the prior fiscal year. As of July 2019, FDA officials stated that because TTB and CBP have up to 3 years to update and finalize their data files, CTP plans to update its procedures to include two reconciliation processes for each fiscal year. According to FDA officials, the first reconciliation, the year-end reconciliation process, would begin immediately following the end of a fiscal year, and the second reconciliation would occur 3 years after that fiscal year ends. FDA’s Year-End Reconciliation Process Following the close of each fiscal year, the Food and Drug Administration’s (FDA) Center for Tobacco Products (CTP) initiates the year-end reconciliation process by requesting official records from the Alcohol and Tobacco Tax and Trade Bureau (TTB) and U.S. Customs and Border Protection (CBP). This process is designed to ensure that the tobacco user fees assessed that year are complete—that is, that all manufacturers and importers subject to user fees were assessed user fees—and accurate—that is, that the user fees assessed each quarter were based on accurate market share information. As designed, the year-end reconciliation includes steps for FDA to compare the information the agency used to calculate quarterly user fees with independent information obtained from TTB and CBP on the individual tobacco manufacturers and importers who paid tobacco excise taxes (to ensure FDA has a complete list of those who should pay user fees) and the amounts paid (to ensure FDA used the right amounts to calculate market share and user fees). According to FDA officials, the year-end reconciliation is designed to identify and make any needed corrections to its individual market share calculations based on findings of new, amended, or missing excise tax payments using the annual tax data provided by TTB and CBP (see sidebar). For example, FDA officials stated that FDA will use the data obtained from TTB and CBP to help identify tobacco manufacturers or importers that should have been assessed user fees but were not, due to the companies not reporting monthly data to FDA as required (non- reporters). This process also enables FDA to identify and address fraudulent reporting by tobacco manufacturers and importers who knowingly failed to submit or submitted false information in monthly forms sent to FDA. If FDA recalculates its quarterly market share percentages based on findings of new or amended excise tax data, FDA’s procedures specify that FDA will then make necessary adjustments to individual tobacco user fee assessments. FDA officials stated that FDA can apply necessary market share adjustments to individual user fee assessments in a subsequent quarterly invoicing cycle, or after the agency completes its year-end reconciliation process based on annual tax data from TTB and CBP. According to FDA officials, in order to limit the need to re-invoice companies multiple times outside of the regular billing cycle, FDA prefers to send the adjusted invoices out once the year-end reconciliation process is complete. However, the officials stated that they can make changes outside of the year-end reconciliation. For example, the officials reported making adjustments to individual user fee assessments for the cigar class once, for the first quarter of fiscal year 2017. In that instance, after receiving updated reports from two cigar companies that had initially reported incorrect excise tax data to FDA, FDA officials stated that the agency (1) recalculated the market share of cigar manufacturers and importers based on the amended data FDA received from both companies and (2) made necessary adjustments to the market share percentages and associated user fees for that class for that quarter. FDA Has Not Completed Its Year-End Reconciliation Process to Ensure User Fees Are Based on Complete and Accurate Data since Fiscal Year 2015 According to FDA, the agency has not completed its year-end reconciliation process since completing the reconciliation for fiscal year 2015—the first year that FDA obtained data directly from manufacturers and importers to calculate user fee assessments. FDA designed the year- end reconciliation process to ensure the agency’s individual user fee calculations are based on complete and accurate data and accurately reflect the market share of each tobacco manufacturer and importer. FDA procedures state that FDA will conduct an annual adjustment for each fiscal year using data received from TTB and CBP for individual manufacturers and importers. According to FDA officials, the agency has been unable to complete the reconciliation process for fiscal years 2016 through 2018 because it identified problems with the quality of data it had initially received from TTB and CBP for those years. FDA officials stated that the agency has worked with TTB and CBP, and officials believe they have determined the reasons for the data problems. Specifically, FDA officials said that changes in both the TTB and CBP internal data systems affected the data fields that FDA needs to complete the reconciliation process. As of July 2019, FDA officials had received revised excise tax and volume data for fiscal years 2016 and 2017 from CBP and TTB. They also received revised 2018 data from CBP and had requested, but not yet received, revised 2018 data from TTB. FDA officials said that once they have received the remaining 2018 data and determined that the data from both agencies are of sufficient quality, they will be able to perform the annual reconciliation process for those fiscal years. According to FDA officials, before they can be certain the data are of sufficient quality, the agency needs to modify its internal data system to accommodate a new CBP data format, and then run the data through the updated system. As of July 2019, FDA projects these modifications to its data system will be completed by the end of calendar year 2019. Once the modifications are finished, FDA projects it will complete the reconciliation process for fiscal year 2016 within 3 to 6 months, and then complete the reconciliation for fiscal years 2017 and 2018 in 3- to 6- month intervals consecutively after that. While FDA has identified the steps to perform the year-end reconciliation process for fiscal years 2016 through 2018, it could also face delays in the future, because it does not have reasonable assurance that it will receive quality data from TTB and CBP in a timely manner to complete the reconciliation process for future years. According to FDA officials, their efforts to obtain the data they need from TTB and CBP have focused on fiscal years 2016 through 2018, and they have not determined procedures or time frames for obtaining data from TTB and CBP for future years. However, according to FDA officials, the agency was considering possible actions for obtaining data in future years. One possible option the agency was exploring was the possibility of FDA gaining direct access to CBP’s and TTB’s data systems to obtain the data needed for the year- end reconciliation. According to officials, as of July 2019, CBP had offered this direct access to its data, and the officials expect to pursue this option with TTB officials for similar access. In addition, the agency reported efforts to schedule meetings with TTB and CBP to discuss establishing memorandums of understanding, or other written agreements, that would establish expectations—such as time frames and data format—with the agencies to obtain the quality data needed for the year-end reconciliation. As of September 2019, FDA reported it had scheduled a meeting with CBP officials and was working to schedule a meeting with TTB officials, but the agency had not yet determined procedures or time frames for obtaining the needed data from these agencies for future years. Federal internal control standards call for agencies to use quality information to achieve their objectives. As part of this standard, agencies obtain relevant data from reliable sources in a timely manner and process these data into quality information that supports their internal control system. Federal internal control standards also call for agencies to externally communicate the quality information necessary to achieve its objectives. As part of this standard, agencies communicate quality information externally through reporting lines so that external parties can help the entity achieve its objectives and address related risks. For example, information communicated includes significant matters relating to risks, changes, or issues that impact the agency’s internal controls. Consulting with TTB and CBP, determining procedures and time frames for FDA to receive the quality data it needs in future years, and documenting them in a written agreement would help to address this risk. Without completing the year-end reconciliation process in a timely manner, FDA cannot ensure that the data it uses to calculate individual user fees are complete and accurate. Until it works with TTB and CBP and resolves this issue, FDA is at increased risk that user fees may not be properly assessed on individual tobacco manufacturers and importers based on their market share of each tobacco product class. Conclusions FDA collects user fees from tobacco manufacturers and importers for its tobacco regulation activities—including important activities such as educating the public about the risks associated with the use of tobacco products and preventing youth use of these products. The agency has designed a process with several steps for assessing these fees, including a year-end reconciliation, to ensure that the calculations are complete and accurate—that is, that all companies subject to user fees pay them, and that no companies are assessed fees in excess of their market share. However, for several years, FDA has faced serious delays obtaining the quality data it needs from TTB and CBP to complete the year-end reconciliations, according to FDA. Until FDA consults with these agencies to determine and document the procedures and time frames that will allow FDA to obtain the quality data it needs to complete this key step in a timely manner, the agency risks repeating these delays. Without performing its year-end reconciliation, FDA is at increased risk of allowing some companies—such as those who did not report information to FDA or who did not report accurate information—to not pay their required share of user fees, while other companies pay too much. Recommendation for Executive Action The Commissioner of FDA should consult with TTB and CBP to determine and document—for example in Memorandums of Understanding or other written agreements—procedures and time frames for FDA to receive quality data from TTB and CBP that will allow FDA to complete its reconciliation process in a timely manner. (Recommendation 1) Agency Comments We provided a draft of this product to HHS for comment. In its comments, reproduced in appendix III, HHS generally agreed with our recommendation. The agency commented that it recognized GAO’s thorough review of FDA’s tobacco user fee program and stated that it is critically important for FDA to have a tobacco user fee collection program that is accurate, complete, and predictable. FDA also stated that it has prioritized making the necessary enhancements to its internal data system to accommodate the new format of TTB and CBP data files, and that these changes are on track to be completed by the end of 2019. HHS also provided technical comments, which we incorporated as appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees, the Secretary of Health and Human Services, FDA Commissioner, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or DeniganMacauleyM@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Timeline of Events Related to Food and Drug Administration Tobacco Product User Fees Appendix II: Additional Information on FDA’s Process Related to Tobacco User Fee Calculation, Billing, and Collection This appendix provides additional information on the Food and Drug Administration’s (FDA) process, which is designed to ensure the quarterly user fees it calculates, bills, and collects each fiscal year are complete and accurate and that user fee invoices are billed to and collected from tobacco manufacturers and importers in a timely manner. Calculation. FDA’s process related to its quarterly user fee calculations include procedures to ensure the tobacco product class and individual manufacturer and importer allocations are accurate. According to FDA’s procedures, at the start of each fiscal year, FDA Office of Financial Management, Division of User Fees staff (1) calculate the percentage share for each tobacco product class and (2) enter these percentage shares into FDA’s User Fee System, which automatically calculates quarterly class allocations for each tobacco class. According to FDA officials, FDA’s Office of Financial Management, Division of User Fees staff, as well as Center for Tobacco Products’s (CTP) Office of Management, User Fee Management Team staff, review the class allocation calculations to verify the percentage shares were accurately calculated for each tobacco product class before entering the class percentages into FDA’s User Fee System. Prior to calculating individual market share percentages that are the basis for individual user fees, the User Fee Management Team within CTP’s Office of Management reviews the monthly data reported to FDA by tobacco manufacturers and importers for accuracy. According to FDA’s procedures, the CTP User Fee Management Team checks to ensure that the volume and excise tax data reported on each FDA form 3852 are accurate based on the accompanying supporting documents. According to FDA procedures, if the CTP User Fee Team identifies incomplete or inaccurate monthly reports, it contacts the appropriate tobacco manufacturers or importers to request the missing documentation or an amended FDA form 3852 and tries to resolve any inaccuracies prior to calculating individual market share for the quarterly billing cycle. Billing. According to FDA’s procedures, the CTP User Fee Management Team submits market share percentages to the FDA Office of Financial Management, Division of User Fees in the month prior to the date that invoices are to be issued. For example, for the first quarterly invoicing cycle (October through December), the CTP User Fee Management Team would submit market share percentages on November 15 and invoices would be mailed by the Division of User Fees by December 1. Using the market share data, the Division of User Fees calculates the quarterly user fee amount assessed to individual manufacturers and importers within each tobacco product class as part of its quarterly invoicing process. FDA officials stated that, prior to creating quarterly invoices, the Division of User Fees reviews CTP market share data to ensure it received all necessary data. Prior to mailing quarterly invoices to individual tobacco manufacturers and importers, FDA officials stated that division staff verifies that the invoices created are complete and accurate by comparing the invoice information to the CTP market share data. Collection. FDA’s Office of Financial Management, Division of User Fees utilizes different mechanisms to identify and notify tobacco manufactures and importers who do not pay their invoices by the quarterly user fee due date (i.e., the last day of the applicable fiscal year quarter). According to FDA’s procedures, the Division of User Fees uses a program within FDA’s User Fee System—referred to as the Dunning Tracker—to track relevant invoice data, including the date user fee payments are due and the amounts owed. The Dunning Tracker is designed to generate alerts to warn division staff of unpaid invoices that are approaching 30, 60, and 90 days past due so they can issue Dunning notification letters—which inform the tobacco manufacturers and importers that their invoices are overdue and provide instructions for making a payment. The Dunning notification letters also inform tobacco manufacturers and importers of the amount of additional charges assessed based on the number of days that the payment is late. According to FDA officials, division staff verify that a Dunning notification letter is issued for each tobacco manufacturer or importer with an outstanding invoice and that the appropriate charges have been assessed. According to FDA officials, the Division of User Fees also maintains an arrears list—a list of tobacco manufacturers and importers who have not paid their quarterly user fees on time. FDA’s procedures provide that the Division of User Fees will share the arrears list with the CTP Office of Compliance and Enforcement to assist that office’s efforts to obtain compliance with the user fee requirements. FDA officials stated that the office monitors the arrears list and takes enforcement action when appropriate. The officials said that the office will first issue information letters, separate from the Dunning notification letters, to each tobacco manufacturer and importer on the arrears list to try to obtain voluntary compliance on the user fee payments owed. FDA officials stated that if the office is unable to obtain compliance after it issues the information letter, it may take further action, such as notifying the delinquent company that all tobacco products manufactured and imported by it are adulterated. Agency officials told us that, in 2014, FDA notified three individual tobacco manufacturers that all the tobacco products they manufactured were adulterated due to these companies’ failure to pay their tobacco user fees. According to FDA’s procedures, the Division of User Fees refers delinquent debt to the Department of Health and Human Services (HHS) Program Support Center when outstanding invoices reach 90 days past due. The Program Support Center will pursue collection efforts per its standard procedures and issues two reports each month to the Division of User Fees to inform it of which debts have been collected and which are uncollectable. Appendix III: Comments from the Department of Health and Human Services Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Kim Yamane (Assistant Director), Matthew Byer (Analyst in Charge), Sam Amrhein, Julie Flowers, Jackie Hamilton, Derry Henrick, Vikki Porter, and LaDonna Towler made key contributions to this report.

Why GAO Did This Study Tobacco use causes more than 480,000 deaths each year, according to the Department of Health and Human Services (HHS). To protect the public, the Family Smoking Prevention and Tobacco Control Act granted FDA, an agency within HHS, authority to regulate tobacco products. To fund FDA's tobacco regulation activities—such as those aimed at preventing youth use of tobacco products—the act authorizes FDA to assess and collect a specified total amount of user fees from tobacco manufacturers and importers each fiscal year. The total amount of user fees are to be allocated based on the individual manufacturers' and importers' market share in six FDA-regulated tobacco product classes. GAO was asked to review FDA's tobacco user fees. This report examines FDA's process for the calculation, billing, and collection of these fees. GAO reviewed the relevant law and regulations, as well as FDA policies and procedures, and interviewed FDA officials. What GAO Found In fiscal year 2017, the latest data available at the time of our analysis, the Food and Drug Administration (FDA) assessed about $635 million in user fees to tobacco manufacturers and importers of six classes of FDA-regulated tobacco products—cigarettes, snuff, chewing tobacco, roll-your-own tobacco, pipe tobacco, and cigars. (See figure.) FDA has a process that is designed to ensure accurate calculation, billing, and collection of tobacco user fees. However, the agency has not completed a key step in this process—its year-end reconciliation—since doing so for fiscal year 2015. FDA procedures provide that the agency will conduct this year-end reconciliation annually after receiving necessary data from the Department of the Treasury's Alcohol and Tobacco Tax and Trade Bureau (TTB) and U.S. Customs and Border Protection (CBP). FDA relies on this year-end reconciliation to ensure that its user fee calculations are based on complete and accurate data—that is, that all manufacturers and importers subject to tobacco user fees were assessed fees correctly, based on accurate market share data. Incomplete or inaccurate data for one manufacturer or importer affects the market share—and the user fee amount—for all other manufacturers and importers in its product class. FDA has not completed this year-end reconciliation in recent years because of delays in obtaining the quality data it needs from TTB and CBP. While FDA has reported receiving most of the data for fiscal years 2016 through 2018 and has plans for completing the reconciliation for those years, the agency faces a risk of repeating delays in its reconciliation efforts in the future because it does not have reasonable assurance that it will receive quality data in a timely manner moving forward. Until FDA consults with TTB and CBP to determine and document the procedures and time frames that will allow FDA to obtain the quality data it needs to complete this key step in a timely manner, the agency risks repeating these delays. What GAO Recommends GAO is recommending that FDA consult with TTB and CBP to determine and document procedures for FDA to obtain quality data so the agency can complete its annual reconciliation process in a timely manner. HHS agreed with GAO's recommendation.

gao_GAO-20-312

gao_GAO-20-312_0

DOD Finalized Its Organizational Strategy and Guidance on Cross- Functional Teams but Removed Specific Implementation Steps from the Strategy and Has Not Implemented Remaining Section 911 Requirements DOD has completed three additional statutory requirements of section 911 since our August 2019 report, but has not completed three remaining requirements, as shown in table 2. We previously reported that DOD had completed four of the statutory requirements, specifically awarding a contract for a study to determine how to best implement cross-functional teams, providing the results of the study to Congress, establishing any cross-functional teams to address critical department objectives and outputs, and reporting to Congress on the establishment of the cross- functional teams. Thus, in total, DOD has completed seven of the 10 statutory requirements. For more detail on all 10 statutory requirements, see appendix II. DOD Finalized Its Organizational Strategy but the Strategy Lacks Specific Implementation Steps On October 29, 2019, the Secretary of Defense approved DOD’s organizational strategy. In preparing the strategy for review and approval, OCMO obtained input on the draft organizational strategy from other DOD and OSD components, consistent with a recommendation from our February 2018 report that OCMO obtain stakeholder input on the development of the organizational strategy. We found that the strategy addresses key requirements of section 911, including identifying critical objectives that would benefit from the use of cross-functional teams and providing for the appropriate use of these teams. As part of the organizational strategy, DOD also identified the actions it has taken to streamline the organizational structure and processes of the Office of the Secretary of Defense, another requirement of section 911. For example, the strategy states that DOD has delegated authority to approve certain global force management actions to the Chairman of the Joint Chiefs of Staff and delegated certain acquisition oversight functions to the military departments. Further, consistent with our recommendation from our February 2018 report that the CMO address how the department will promote and achieve a collaborative culture, the organizational strategy includes a short reference to our leading practices for mergers and organizational transformations. However, while the approved organizational strategy cites the leading practices, it does not include specific implementation steps that explain how DOD will follow these practices. Earlier drafts of the organizational strategy that we had reviewed included more specific implementation steps, but those steps were removed during the internal DOD review process. As we reported in August 2019, a January 2019 draft of the organizational strategy included practical implementation steps DOD planned to take to advance a collaborative culture, each of which were shown to align with our leading practices for mergers and organizational transformations. For example, consistent with the leading practice for establishing a coherent mission and integrated strategic goals to guide the transformation, the January 2019 draft proposed that the CMO develop an implementation plan with goals and milestones for its efforts to implement the organizational strategy, communicate those goals and milestones, and report periodically on the achievement of the goals. However, in place of these specific steps, the approved organizational strategy simply lists these leading practices and makes a broad statement that DOD is committed to further incorporating and institutionalizing these practices at every opportunity. An OCMO official told us these implementation steps were removed as the OCMO prepared the draft for department-wide coordination and submission to the Secretary of Defense for review and approval. According to that official, OCMO officials made this change because the Secretary and Deputy Secretary of Defense were newly confirmed, and OCMO officials did not want to commit them to a specific course of action. That official also told us that DOD might use its senior leadership forums, such as the Deputy’s Management Action Group (DMAG), to monitor implementation of the organizational strategy and identify opportunities to improve collaboration, including implementation of our leading practices. However, the official acknowledged that any plan to use such forums for monitoring implementation has not been finalized. As we stated in making our February 2018 recommendation that the department address how it would promote and achieve a collaborative culture, section 911 identified several outcomes that DOD should achieve to advance such a culture. We also noted that DOD could use our leading practices for mergers and organizational transformations to address how the department will advance a culture that is collaborative, team-oriented, results-driven, and innovative. We further stated that DOD would be better positioned to transform and meet its mission if it incorporated these leading practices in its organizational strategy as a way to better articulate how the department will achieve the outcomes that advance a collaborative culture across DOD and address the requirements of section 911. Specific implementation steps like those included in earlier drafts of the organizational strategy offered the department a clear path forward for pursuing the goals of section 911 and promoting a collaborative culture. Absent these steps, such as developing an implementation plan with goals and milestones, it is less clear how DOD intends to implement the organizational strategy and assess progress toward its goals. Identifying and documenting specific implementation steps to advance a collaborative culture—such as those OCMO included in earlier drafts of the organizational strategy—is necessary to fully address the requirements of section 911. DOD Finalized Guidance on Cross-Functional Teams and Plans More Detailed Guidance to Fulfill All Section 911 Requirements On December 12, 2019, the Secretary of Defense approved DOD’s guidance on cross-functional teams. We found that this two-page guidance addresses most, but not all, of the 911 requirements and leading practices for cross-functional teams. Specifically, it addresses in whole or in part six of the seven section 911 requirements and six of the eight leading practices. The Secretary-approved guidance also directs the CMO to develop more detailed implementing guidance. It will be important for the CMO to develop and issue this detailed implementing guidance to fully address section 911 requirements and our leading practices for effective cross- functional teams, consistent with a recommendation in our February 2018 report. According to an OCMO official, OCMO plans to use previously drafted terms of reference as the basis for the CMO’s more detailed implementing guidance. Based on our review, when the Secretary of Defense approved guidance is considered along with the draft terms of reference expected to serve as detailed implementing guidance, both documents will fully address all section 911 requirements and leading practices for effective cross-functional teams. We will monitor the department’s progress in issuing this guidance as part of our normal process of assessing DOD’s efforts to implement our recommendations. DOD Has Not Addressed Requirements for Training and Analysis Training for Cross-Functional Teams Members and Their Supervisors DOD has not approved its curriculum for training for cross-functional team members and their supervisors. In February 2018, we reported that DOD’s draft curriculum for cross-functional team members and their supervisors addressed the section 911 requirements for that training. We reported in August 2019 that DOD had provided required training using its draft curriculum to members of the EMSO team—DOD’s only established section 911 team at the time—but had not provided training to their supervisors. According to DOD’s comments on our August 2019 report, DOD expected the draft curricula for training for cross-functional team members and their supervisors to have been approved simultaneously with the issuance of the Secretary’s guidance on cross-functional teams. According to OCMO officials, however, DOD has contracted for the delivery of the required training for cross-functional team members and their supervisors. One of those officials also told us OCMO now expects that training to be completed in 2020. Another OCMO official told us that the OCMO has been further refining the draft curriculum based on feedback from the members of the EMSO team and external experts before submitting the curriculum for review and approval. Training for Presidential Appointees DOD has not provided required training on cross-functional teams and related subjects to presidential appointees and the curriculum has not been approved. Section 911 required presidentially appointed, Senate- confirmed officials to receive training on leadership, modern organizational practice, collaboration, and the operation of cross- functional teams within 3 months of their appointment or to receive a waiver from the President of the United States. As of October 2019, 23 of 36 such positions had been filled and the officials had been in their positions for more than 3 months; none had received the statutorily mandated training. According to DOD’s comments on our August 2019 report, DOD expected the draft training curricula for presidential appointees to have been approved simultaneously with the issuance of the Secretary’s guidance on cross-functional teams. According to OCMO officials, however, DOD has contracted for the delivery of the required training for presidential appointees. One of those officials also told us they now expect that training to be provided in 2020. An OCMO official told us that OCMO has been further refining the draft curriculum and discussing possible venues for providing this training for presidential appointees, including one of the weekly meetings that the Deputy Secretary of Defense has with all Office of the Secretary of Defense presidential appointees. Analysis of Successes and Failures of Cross-Functional Teams DOD has not completed the required analysis of the successes and failures of its cross-functional teams. Section 911 requires DOD’s analysis to be completed with support from external experts in organizational and management sciences within 18 months of the establishment of the first cross-functional team under section 911. Because the first cross-functional team was established in August 2017, this analysis was due in February 2019. According to OCMO officials, DOD has contracted with an organization to help develop the analysis. One of the officials also told us DOD expects the analysis to be completed in 2020. Another OCMO official told us in December 2019 that work on the assessment, including a survey and structured interviews, was underway, and that an initial draft report was expected by the end of December 2019. DOD’s Electromagnetic Spectrum Operations Cross-Functional Team Is Continuing Its Work and DOD Has Expanded the Number of Teams DOD’s Electromagnetic Spectrum Operations Cross-Functional Team Is Continuing to Work toward Its Mission but Resource Issues Remain DOD’s EMSO team is continuing to work toward its mission to develop requirements and specific plans to improve EMSO capabilities across the department and to achieve operational superiority. The team is developing 13 initiatives in four areas—governance, organization, capabilities and gaps, and training and readiness. In addition, the team issued a report required by section 1053 of the John S. McCain National Defense Authorization Act for Fiscal Year 2019, which included the mandated assessments of the electronic warfare capabilities of the Russian Federation and the People’s Republic of China in consultation with the Director of the Defense Intelligence Agency. Section 1053 also required the team to, among other things, update the department’s Electronic Warfare Strategy in coordination with the Electronic Warfare Executive Committee. According to an EMSO official, the team is developing a new strategy, which is scheduled for completion in June 2020. The EMSO team is continuing to demonstrate leading practices for effective cross-functional teams, similar to what we reported in August 2019. Specifically, at that time, we reported that the EMSO team was demonstrating leading practices for effective cross-functional teams, such as a well-defined team structure and well-defined team goals. Based on the results of our recent survey of individual team members, most team members believe the team is demonstrating open and regular communication, an inclusive team environment, has an empowered cross-functional team leader, and has well-defined team goals. However, our survey results show that team members responded less favorably to questions about senior management support than to questions about the other leading practices. Specifically, less than half of the respondents agreed with the statements that DOD’s senior leadership provides the team with sufficient resources for its work, supported the team as a priority, and supported the team’s goals and objectives. In their survey responses and during interviews, team members expressed their concerns about the lack of resources, such as funding and sufficient office space to perform their work. According to team officials, they are continuing to work with the OCMO to resolve the team’s resource issues. In our August 2019 report, we stated that the team’s progress was negatively affected by funding delays resulting from disagreements among senior leadership over the responsibility for funding the team. The disagreement had been resolved for fiscal year 2019, but had not yet been resolved for future fiscal years. We recommended, and DOD concurred, that the CMO and EMSO cross- functional team clarify roles and responsibilities for providing administrative support and funding for the team beyond fiscal year 2019 in accordance with the memorandum establishing the team. According to EMSO team officials, however, funding for future years has not been identified. The team has discussed its funding needs with the OCMO and staff from the Office of the Deputy Secretary of Defense, but there is still no clarity regarding responsibility for funding the team. The team’s budget submission as part of DOD’s fiscal year 2021 budget process was withdrawn because the amount requested was smaller than the amounts typically reviewed in the process. According to an EMSO team official, the OCMO is providing funding for the team incrementally on a quarterly basis, and is facing challenges with funding the team’s request for a contract due to the department operating under a continuing resolution. According to another EMSO team official, the team is maintaining the status quo with its current funding and is not considering any additional initiatives. We also reported in our August 2019 report that, according to a team official, while the team had its own office space, the space did not have the level of security required to allow the team to work on a third of its initiatives. Team officials have since told us the team plans to move to an appropriately secure space in early 2020. We will continue to monitor DOD’s progress to secure resources and office space for the team as part of our normal process of assessing DOD’s efforts to implement our recommendations. DOD Has Expanded the Number of Cross- Functional Teams In its approved organizational strategy, DOD identified two existing task forces to expand the number of cross-functional teams. First, the Secretary of Defense established the Close Combat Lethality Task Force in February 2018 to develop, evaluate, recommend, and implement improvements to U.S. squad-level infantry combat formations and strengthen the combat, lethality, survivability, resiliency, and readiness of infantry squads. Second, the Secretary of Defense established the Protecting Critical Technology Task Force in October 2018 to address concerns over the security of the department’s critical technology and the loss of classified information and controlled unclassified information that puts DOD’s investments at risk and erodes the lethality and survivability of U.S. forces. According to an OCMO official, the OCMO updated the organizational strategy and designated these two task forces as cross- functional teams as a result of input from DOD senior leadership during their review of the draft organizational strategy. Based on our review of the documents used to establish the two task forces, we found that they would meet only some of the requirements we reviewed for cross-functional teams as mandated by section 911. For example, we found that the documentation for the Close Combat Lethality Task Force, as it was established, shows that the task force has a clearly established objective; is directed to develop recommendations such as policy changes and investment decisions; and is directed to make decisions on cross-functional issues—some of the key section 911 requirements. Similarly, we found that the documentation for the Protecting Critical Technology Task Force, as established, shows that the director of the task force has the authority to select the membership from across the department, another key requirement. However, based on the documents we reviewed, we found that the teams would not meet other requirements. For example, we found that the documentation for both teams did not ensure that those team members and leaders who are supervisors receive training in elements of successful cross-functional teams. According to an OCMO official, the OCMO will ensure that these task forces identified as cross-functional teams meet the requirements of section 911. For example, the OCMO will provide the required training. DOD’s newly issued guidance on cross-functional teams could help ensure that existing and any new cross-functional teams meet section 911 requirements. In addition, it could help provide these existing and any new teams with the information, direction, and authority they need to comply with mandated requirements for cross-functional teams. For example, section 911 permits the Secretary to delegate to cross- functional teams any decision-making authority that the Secretary considers appropriate to achieve the objectives of the teams; DOD’s guidance delineates the decision-making authority of cross-functional teams. Conclusions More than 3 years after the passage of the National Defense Authorization Act for Fiscal Year 2017, DOD has begun to take key steps to address the requirements of section 911 and to promote a more collaborative culture in the department, including issuing its organizational strategy and making greater use of cross-functional teams under the act. Even as it has taken these steps, challenges for the departments’ ongoing implementation of section 911 remain. The department has still not addressed key requirements to help promote a collaborative culture and, according to officials, still has not identified responsibility for funding one of its cross-functional teams established under section 911. Further, specific implementation steps that would have offered the department a clear path forward for pursuing the goals of section 911 and promoting a collaborative culture at DOD were removed from DOD’s approved organizational strategy—a disappointing development. Identifying and documenting specific implementation steps to encourage a collaborative culture is necessary to fully address the requirements of section 911 and encourage such a culture. Recommendation for Executive Action The Secretary of Defense should ensure that the Chief Management Officer identify and document specific implementation steps to advance a collaborative culture, consistent with our leading practices for mergers and organizational transformations. Agency Comments We provided a draft of this report to DOD for review and comment. In its written comments, which are reproduced in Appendix IV, DOD concurred with our recommendation. DOD also provided additional information on the steps that DOD has taken or plans to take to advance a collaborative culture, such as the Secretary and Deputy Secretary of Defense’s use of DOD’s senior governance forums to encourage collaboration across the department. DOD also stated that it plans to incorporate policies based on best practices for cultivating a collaborative organizational climate into the CMO’s guidance on the implementation of cross-functional teams as well as future National Defense Strategies and National Defense Business Operations Plans. We are sending copies of this report to the appropriate congressional committees and to the Secretary of Defense and Chief Management Officer. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2775 or fielde1@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this are listed in appendix V. Appendix I: Prior GAO Reports on the Department of Defense’s (DOD) Implementation of Section 911 of the National Defense Authorization Act (NDAA) for Fiscal Year 2017 Section 911 of the NDAA for Fiscal Year 2017 included a provision for us—every 6 months after the date of enactment on December 23, 2016, through December 31, 2019—to submit to the congressional defense committees a report. Each report is to set forth a comprehensive assessment of the actions that DOD has taken pursuant to section 911 during each 6-month period and cumulatively since the NDAA’s enactment. Table 3 identifies our five prior reports on DOD’s implementation of section 911 and the status of the 11 recommendations from those reports. Appendix II: Summary of Requirements in Section 911 of the National Defense Authorization Act for Fiscal Year 2017 Section 911 of the National Defense Authorization Act for Fiscal Year 2017 requires the Secretary of Defense to take several actions. Table 4 summarizes these requirements, the due date, and the date completed, if applicable, as of December 2019. Appendix III: Leading Practices for Implementing Effective Cross-Functional Teams In February 2018, we reported on eight leading practices for implementing effective cross-functional teams. Table 5 identifies these leading practices and their related key characteristics. Appendix IV: Comments from the Department of Defense Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Margaret Best (Assistant Director), Daniel Ramsey (Analyst-in-Charge), Sierra Hicks, Alexa Kelly, Richard Powelson, and Paulina Reaves made key contributions to this report. Other contributors included Tracy Barnes, Arkelga Braxton, Michael Holland, Ned Malone, Judy McCloskey, Jeremy Rogers, Ron Schwenn, and Sarah Veale.

Why GAO Did This Study DOD has had longstanding organizational and management challenges that hinder collaboration. Section 911 of the NDAA for Fiscal Year 2017 directed the Secretary of Defense to, among other things, issue an organizational strategy that identifies critical objectives that span multiple functional boundaries, establish cross-functional teams to support this strategy, and provide related guidance and training. The NDAA for Fiscal Year 2017 also included a provision for GAO periodically to assess DOD's actions in response to section 911. GAO has issued a series of reports since June 2017 and made a number of recommendations to DOD. This report assesses the extent to which DOD has made progress in (1) implementing the requirements of section 911 and (2) establishing cross-functional teams. GAO reviewed documentation, interviewed cross-functional team members and other DOD officials, and compared DOD's actions to section 911 requirements and leading practices for cross-functional teams. What GAO Found Since GAO's August 2019 report, the Department of Defense (DOD) has taken actions to complete three additional statutory requirements of section 911 of the National Defense Authorization Act (NDAA) for Fiscal Year 2017, but has not completed three remaining requirements. These requirements are intended to support cross-functional teams and to promote department-wide collaboration (see table). Cross-functional teams rely on individuals with different types of expertise to work toward a common, well-defined goal, and are thought to deliver better and faster solutions to complex and fast-moving problems. DOD's approved organizational strategy addresses key requirements of section 911, including identifying critical objectives that would benefit from the use of cross-functional teams and providing for the appropriate use of these teams. However, the strategy did not include practical, specific implementation steps to guide DOD's efforts to advance a collaborative culture, which had been included in earlier draft versions of the strategy. These steps had aligned with GAO's leading practices for mergers and organizational transformations. Specific implementation steps like those included in earlier drafts of the organizational strategy offered DOD a clear path forward for pursuing the goals of section 911 and for promoting a collaborative culture. Absent identifying and documenting specific implementation steps, it is less clear how DOD intends to implement the organizational strategy and assess progress toward its goals. DOD's existing cross-functional team charged with improving electromagnetic spectrum operations and defending its communication systems from attacks is continuing its work by issuing a statutorily mandated report, among other efforts, but DOD has not clarified responsibility for funding the team. GAO will continue to monitor DOD's progress toward providing such support to the team as GAO recommended in August 2019. In addition, DOD has designated the Close Combat Lethality Task Force and the Protecting Critical Technology Task Force as new cross-functional teams, although they meet only some of the section 911 requirements. DOD officials said they will ensure that the newly designated teams meet these requirements, including providing required training. What GAO Recommends In this report, GAO recommends that DOD identify and document specific implementation steps to advance a collaborative culture, consistent with GAO's leading practices. GAO also reiterates the importance of addressing its prior recommendations. DOD concurred with GAO's recommendation.

gao_GAO-19-512

gao_GAO-19-512_0

Background In the early 2000s, the Navy conceived of a new small surface combatant concept known as LCS. This ship was intended to offer the Navy an affordable, flexible platform that would be able to swap out surface warfare, anti-submarine warfare, or mine countermeasure mission packages to provide for one of those mission needs. As we found in multiple reports, the Navy’s vision for LCS evolved significantly over time in response to diminished capability expectations and significant cost and schedule growth. In 2014, the Secretary of Defense directed the Navy to evaluate alternatives to LCS, citing survivability and lethality concerns. This represented the beginning of the Navy’s pursuit of a solution to address LCS shortcomings and the evolving threat environment acknowledged by the department. The Navy initially envisioned quickly fielding a frigate—referred to as the FF program—based on a minor modified LCS design. The ship was expected to provide a more lethal and survivable multi-mission ship capable of simultaneous surface and anti-submarine warfare, with a planned contract award for the lead ship in 2018. In 2016, we found that the Navy’s planned upgrades for FF did not significantly improve certain survivability areas and lacked capabilities that were prioritized by fleet operators, such as the ship’s range of travel without refueling. Then, in April 2017 we found the Navy’s aggressive FF acquisition schedule increased risk to the government because it included a commitment to buy ships in advance of adequate knowledge. In May 2017, the Navy announced it was revising its frigate plans and began pursuing FFG(X). Shipbuilding Best Practices In 2009, we identified commercial shipbuilding best practices that could be adapted for use by the Navy. We found that successful shipbuilding programs have sound business cases built on attaining critical levels of knowledge at key points in the shipbuilding process before significant investments are made, as shown in figure 1. Regardless of the differences between Navy and commercial shipbuilding, knowledge attainment is crucial to success. Executable business cases use realistic cost and schedule targets to meet performance and quality expectations by balancing inherent uncertainties in acquisition programs. A solid business case provides for the resources necessary to mitigate challenges, such as immature technologies and design requirements. The greater the potential for challenges to occur, the more time and money should be factored into the business case to address them. The Navy has previously agreed, in principle, that knowledge should be attained prior to key milestones to better ensure ships are built to established cost, schedule, quality, and performance standards. Navy Shipbuilding Acquisition Framework In general, the Department of Defense (DOD) acquires new weapon systems, such as Navy surface combatants, through a management process known as the Defense Acquisition System. Under this system, programs typically complete a series of milestone reviews and other key decision points that authorize entry into a new acquisition phase. To execute shipbuilding acquisition programs, the Navy uses the acquisition processes included in the DOD Instruction 5000 series, as well as acquisition instructions established by the Secretary of the Navy. The Navy’s guidance supports a seven-gate review process intended to ensure that requirements align with acquisition plans, and to improve collaboration among stakeholders. Figure 2 provides an overview of the notional framework for Navy shipbuilding acquisition programs described by the DOD and Navy guidance. This acquisition framework includes decision reviews and milestones at key junctures in the acquisition cycle. The Milestone Decision Authority is the individual responsible for determining what events and documentation requirements will apply to an acquisition program, as well as providing approval for a program to proceed to the next acquisition phase. The acquisition framework and Milestone Decision Authority’s purpose is to support careful assessment of a program’s readiness to proceed to the next stage of acquisition activities. The gates and milestones that will be included in an acquisition program’s schedule can be customized based on its circumstances and needs. We have previously found that shipbuilding programs typically have different decision points than other DOD weapon systems. For example, Milestone B for ship programs usually occurs after development of ship specifications and system diagrams is well under way and is typically aligned with the decision to authorize the start of detail design. While Milestone C generally represents the decision to start production for weapon systems, several of the Navy’s more recent shipbuilding programs either do not include a Milestone C review or changed the sequencing of the review to occur after delivery of the lead ship. Programs can receive approval to tailor the requirements for information that must be developed to support this process and to have the decision- making authority delegated to other individuals for acquisition decisions and approvals. Navy Expects That FFG(X) Requirements Will Provide a More Capable Small Surface Combatant, but at Increased Cost The Navy expects that its current plans for FFG(X) will result in a small surface combatant with considerable capability improvements compared to LCS. To achieve this increased capability, the Navy is committing to construct a larger, more expensive ship than LCS. To help refine FFG(X) requirements and identify opportunities for cost savings, the Navy used a conceptual design phase, in which it awarded $75 million in contracts to industry. FFG(X) Requirements Reflect Limitations of LCS and Evolution in Capability Needs The Navy’s FFG(X) requirements represent the department’s recognition of its need for a more capable small surface combatant and the limitations of LCS. For LCS and its mission packages, the Navy has devoted nearly $28 billion (constant fiscal year 2018 dollars) to develop and buy a ship that has fallen far short of demonstrating it can meet the minimum level of capability defined at the beginning of the program. Specifically, LCS was designed with reduced survivability requirements as compared to other surface combatants. Over time the Navy lowered several survivability and lethality requirements further and removed some design features— making the ships less survivable in their expected threat environments and less lethal than initially planned. As shown in figure 3, the Navy arrived at its FFG(X) plans after spending several years developing and evaluating a variety of inputs to address problems with LCS and emerging capability needs. The Small Surface Combatant Task Force study report maintained the Navy’s need for 52 small surface combatants, which was revalidated in the Navy’s 2016 Force Structure Assessment. In recognition of LCS’s shortcomings, the Navy significantly reduced the total number of LCS, and began planning for the new frigate based on minor modifications to an LCS design—referred to as FF—to fulfill the 52-ship need. While the FF program was developing its acquisition plans and moving toward a contract award for the lead ship scheduled for 2018, the maritime operating environments continued to rapidly evolve, becoming increasingly complex and contested. In recognition of this, the Chief of Naval Operations directed the Navy to conduct another study, increasing air defense and survivability beyond the FF baseline. In response, the Navy convened a Frigate Requirements Evaluation Team from January to June 2017. The purpose of this team was to build upon FF requirements by analyzing options for air defense and vulnerability upgrades to help determine top-level mission requirements that would yield a more capable frigate. The results of this review led the Navy to cancel its FF acquisition plans and focus on meeting increased requirements through a new FFG(X) Guided Missile Frigate program. Both the FF and FFG(X) requirements reflect the 2015 Small Surface Combatant Task Force report findings that identified a need for increased capabilities for small surface combatants to address evolving threats. As we reported in June 2016, an FF based on a minor modified LCS only partially fulfilled the small surface combatant capabilities that the task force identified as most valued by the fleet. In particular, FF requirements supported a multi-mission ship with some of the fleet’s highest priority mission capabilities, such as surface and anti-submarine warfare, but did not provide air warfare capability. For FFG(X), the Navy maintained the FF requirements and added local air defense as a capability. Table 1 outlines the requirements evolution that the Navy undertook to support a more lethal and capable small surface combatant. Requirements Drive Higher FFG(X) Cost than for Previous Small Surface Combatants To achieve the increased capability expectations for FFG(X), the Navy committed to acquiring a larger, more expensive ship than LCS or the previously planned FF. Figure 4 provides average shipbuilding cost estimates for the three different ships, with costs shown in same-year dollars for comparison. Although the FFG(X) requirements have been finalized, the Navy plans to make final cost and capability tradeoffs through the process of evaluating proposed designs before selecting which one will be built. The Navy Used Conceptual Design Phase to Better Understand Ship Requirements and Associated Costs In an effort to focus on the relationship between requirements and cost, the Navy undertook a conceptual design phase for FFG(X), which enabled industry to inform requirements and identify opportunities for cost savings. In February 2018, the Navy competitively-awarded FFG(X) conceptual design contracts valued at nearly $15 million each to five industry teams. These 16-month contracts were intended to enable industry to mature parent ship designs—designs for FFG(X) that are based on ships have been built and demonstrated at sea—and help refine technical and operational program requirements. The purpose of the conceptual design phase has parallels with the purpose of pre-contractual negotiations in commercial shipbuilding. As we previously have reported, these pre-contractual practices minimize ship buyer risk prior to awarding construction contracts by developing the ship concept and specifications based on negotiations between the ship buyer and the shipyard. The practices include specifying the expected performance and the major equipment on the ship. As part of these activities, commercial shipbuilders and ship buyers analyze one or more ship concepts to identify areas of potential risk and either mitigate these risks or remove the risky elements from the ship before signing a contract. Figure 5 provides an overview of the industry teams and shipyards participating in the FFG(X) conceptual design. Each industry team performed ship development, ship design, workforce planning, and shipyard improvement planning, among other activities, in support of FFG(X) requirements refinement and cost reduction efforts. Industry teams updated the Navy regularly on their design progress and technical approach to fulfill requirements through monthly technical exchange meetings and two design review meetings. Navy officials stated that these meetings were intended to provide information to support the program’s Preliminary Design Review in May 2019 and mitigate risk prior to the Navy’s release of its request for proposal in June 2019 for the FFG(X) detail design and construction competition. Our prior work on shipbuilding best practices emphasizes the importance of having a full understanding of the effort needed to design and construct a ship before awarding a contract for ship construction in order to reduce cost and schedule risk. Navy and industry officials stated that the conceptual design phase facilitated dialogue and information sharing that helped ensure FFG(X) requirements were more fully understood by industry and the government. Specifically, industry officials noted that communication and activities during conceptual design improved their understanding of the impetus for specific Navy requirements, allowing industry the opportunity to get clarification on the intent of some requirements, propose less costly alternatives, and get government feedback on the proposed alternatives. It also improved their understanding of the linkages between FFG(X)’s approved capability requirements and system specifications. In particular, industry officials told us that one-on-one opportunities with the Navy aided knowledge sharing and provided them with a means to ask questions without concern that disclosing such information could jeopardize their competitive position. They emphasized that in other cases where the request for proposals process is their primary means for communicating with the Navy (as opposed to having a conceptual design phase), submitting questions about requirements or system specifications can be challenging because those inquiries are available to the public. As a consequence, contractors may opt to infer more about the intent of requirements to avoid compromising their competitive interests. The conceptual design phase included a formal cost savings effort, with the Navy seeking proposals internally and from industry participants to reduce cost through requirement and system specification refinement. To support this effort, Navy officials stated they established a Frigate Affordability Board to review potential cost reduction measures submitted by both contractors and government that responded directly to Navy requirements and specifications. Navy officials said the Board—co- chaired by the Program Executive Office for Unmanned and Small Combatants and the Naval Sea System Command’s Naval Systems Engineering Directorate, as well as the Chief of Naval Operations’ Surface Warfare Directorate—assessed the potential cost and capability trade-offs of these proposed changes to requirements, and accepted or declined them. Before going to the Board, relevant Navy subject matter experts reviewed the technical and requirements implications of cost reduction measures. The program office subsequently worked with Navy engineering and requirements officials to balance cost with capabilities. If the program office, Navy engineers, and requirements officials could not reach agreement on the appropriate cost and capability mix, then their different positions were presented to the Board. For cost reduction initiatives submitted by industry, the Navy provided feedback on the Board’s decision, and incorporated fully or partially accepted cost reduction initiatives into the FFG(X) system specifications. Navy officials said they informed all industry teams of any changes to the specifications on a monthly basis. Navy officials also stated that industry submitted about 350 cost reduction ideas, with roughly 60 percent partially or fully accepted by the Navy. They estimated $86 million in savings per ship (constant year 2018 dollars) based on changes made in response to the cost reduction measures submitted by industry or government-initiated cost savings measures influenced by engagement with industry. Streamlined Acquisition Approach Accelerates Planned FFG(X) Schedule, but Reduces Knowledge Available for Key Program Decisions In an effort to accelerate the time between FFG(X) acquisition planning and the fielding of ships, the Navy streamlined the program’s acquisition approach and leveraged knowledge obtained from industry during the conceptual design phase. While the program may benefit from the streamlining efforts, the acquisition approach for FFG(X) required the Navy to submit its budget request for lead ship construction before the program had a comprehensive understanding of the potential ship designs and cost. Recent Navy policy changes have created some uncertainty for Navy cost estimation activities by altering roles and responsibilities within the Navy for completing component cost positions and independent cost assessments. Navy Streamlined FFG(X) Program Acquisition Approach in an Effort to Accelerate Fielding of Ships As permitted by DOD and Navy policy, the Navy has streamlined the FFG(X) acquisition approach to move from planning to ship delivery and fielding quicker than in a more traditional acquisition program. The accelerated schedule reflects the Navy’s desire to field a minimum of 52 small surface combatants, which the Navy’s long-range shipbuilding plan states will be achieved by fiscal year 2034. Navy officials stated that the significant amount of knowledge that already existed to inform the program’s early activities and the use of parent designs helped enable the streamlined approach for FFG(X). For example, Navy officials cited previous efforts by the Small Surface Combatant Task Force and the Frigate Requirements Evaluation Team to determine appropriate ship requirements, as well as activities performed in support of the FF frigate acquisition plan that immediately preceded the shift to FFG(X). The Navy also leveraged industry input received from a request for information in 2017 to understand cost drivers and the potential shipbuilders’ abilities to meet top level FFG(X) requirements and incorporate Navy-defined equipment into ship designs. Figure 6 provides a high-level schedule of key activities for the program. To support its decision to pursue an accelerated acquisition schedule, the Navy used the previously discussed conceptual design phase as well as its decisions to limit FFG(X) to parent ship designs and minimize technology development. Navy officials noted the use of parent designs is allowing the program to proceed at a much faster pace from early assessment of capability options to detail design and construction contract award. They added that the parent designs provided a higher- fidelity design baseline from which the conceptual design industry teams incorporated Navy systems and other requirements. Use of parent designs is consistent with our best practices work in shipbuilding, which has found that commercial shipbuilders use previous ship designs to the extent possible. Doing so can reduce technical, schedule, and cost risk in building a ship as compared to a “clean sheet” new ship design. FFG(X) program officials noted the latter approach can take up to 9 years to complete an analysis of alternatives and move through the acquisition process to construction contract award. Navy officials said the program also used opportunities available as an Acquisition Category (ACAT) 1B program to shorten the approval timeline for specific acquisition requirements. For an ACAT 1B program, the head of the DOD component is generally the Milestone Decision Authority but, as appropriate, may delegate approval authorities to lower level offices under its jurisdiction. In the case of FFG(X), the Assistant Secretary for the Navy for Research, Development, and Acquisition serving as the Milestone Decision Authority delegated specific approval authorities to the Program Executive Office for Unmanned and Small Combatants. These approval authorities applied to the program’s life cycle sustainment plan, independent logistics assessment, program protection plan, and a compliance schedule addressing environmental considerations. The Navy also took advantage of opportunities to alter or waive some significant early acquisition activities. For example, the Milestone Decision Authority waived the formal Analysis of Alternatives and Affordability Analysis, decided not to conduct a Milestone A review, and deferred the full “Should-Cost” Analysis to later in the acquisition process. Table 2 defines the purpose of these DOD acquisition program elements and provides an overview of the Navy’s actions related to them. As the first major milestone for many major acquisition programs, Milestone A is a review by the Milestone Decision Authority of key program documents that support the materiel solution and risk reduction. We have previously found that DOD officials place a high value on the information developed for some of these documents, including the Analysis of Alternatives, Affordability Analysis, and Should-Cost Analysis. The Navy’s decision to not conduct a Milestone A review also eliminated a formal opportunity to bring the broad set of FFG(X) stakeholders within the Navy and the Office of the Secretary of Defense together at a relatively early stage to assess the program’s acquisition strategy and affordability and feasibility, as well as technical, cost, and schedule risks. Further, it reduced the FFG(X) acquisition approach to a single milestone decision point—Milestone B—for the broader group of DOD stakeholders to evaluate program progress and readiness to proceed to the detail design and construction contract award planned in July 2020. In the absence of Milestone A, the Navy’s Gate 3 review for FFG(X) provided an opportunity to communicate the program’s progress toward developing requirements and acquisition expectations, albeit to a more limited audience than typically would participate in a Milestone A. In particular, the Navy used Gate 3 to discuss top-level requirements changes and receive capability development document approval from the Chief of Naval Operations. It also included cost discussion related to FFG(X) affordability within the overall Navy shipbuilding portfolio. The gate’s participants included officials from the Navy and the Office of Cost Assessment and Program Evaluation (CAPE) within the Office of the Secretary of Defense. The Navy’s Gate 4 conducted in February 2019 focused on a review of the FFG(X) system specification before the draft detail design and construction request for proposal release. Gate 4 documentation for FFG(X) indicates that participants were limited to stakeholders from the office of the Deputy Assistant Secretary of the Navy for Ships; Naval Sea Systems Command Cost and Design Directorates; Program Executive Office for Unmanned and Small Combatants; the FFG(X) program office; and the Chief of Naval Operations Surface Warfare Directorate. This excludes a number of key stakeholders that Navy guidance calls on to attend and certify gate reviews, such as the Assistant Secretary of the Navy (Financial Management and Comptroller) and the testing community. As a result, the Navy would not have received insight from several key stakeholders during the Gate 4 review for acquisition activities, such as the program life cycle cost estimate development and release of the draft request for proposal. These activities are generally relevant to this gate review, as Navy guidance notes program affordability as a focus and the Navy’s streamlining documentation indicates that the gate was focused on reviewing the FFG(X) system specification before releasing the draft request for proposal. Navy officials noted that stakeholders have regularly received insight into FFG(X) activities through other prior program reviews and will have additional opportunities to review program costs and sustainment plans leading up to Milestone B. We also found that some key stakeholders did not provide formal approval for the initial FFG(X) life cycle sustainment plan that was approved in March 2019. Specifically, only FFG(X) program officials and the Program Executive Officer for Unmanned and Small Combatants— the delegated approval authority—signed the plan. However, as stated in DOD guidance, representatives from the relevant sustainment command and the Program Executive Office for Integrated Warfare Systems are key stakeholders that should provide their signed concurrence when approving the life cycle sustainment plan. The FFG(X) life cycle sustainment plan is a key document outlining the Navy’s plans to address the program’s sustainment needs and costs, as typically around 70 percent of a weapon system program’s total cost is in the sustainment phase after procurement. Navy officials stated that the plan has been reviewed by the independent logistics assessment team members that are evaluating the FFG(X) program’s integrated product support activities, and noted that the Program Executive Office for Integrated Warfare Systems has separate life cycle sustainment plans for government furnished equipment systems included in the FFG(X) design. Navy officials also said that FFG(X) sustainment plans would be reviewed by stakeholders as part of Gate 5 and the Milestone B independent logistics assessment. Budget Request for FFG(X) Lead Ship Preceded the Completion of Key Cost Estimation Activities That Should Inform Funding Decisions The FFG(X) acquisition approach required the Navy to submit its nearly $1.3 billion budget request for lead ship construction before the program had established a comprehensive understanding of the potential ship designs and estimated cost for the program. Our shipbuilding and acquisition best practices call for resource decisions to be timed to align with the availability of requisite cost, schedule, and technical knowledge in order to inform key program decisions. Navy officials stated that they had sufficient knowledge to inform key program decisions based on cost estimation and conceptual design efforts that had previously been completed. Navy officials said this included development of an FFG(X) cost estimate by November 2018 to support a realistic budget request for the lead ship. However, at the time of the Navy’s fiscal year 2020 budget request to fund detail design and the lead ship, the Navy had not completed its component cost position, which will formalize the life cycle cost expectations for FFG(X). Further, CAPE had not completed the independent cost estimate for the program. The GAO Cost Estimating and Assessment Guide says that comparing the component cost position with an independent cost estimate to validate methodologies produce similar results reinforces the credibility of a cost estimate. In addition to key cost estimating best practices that had not been completed, the Navy had not received final design review information from the industry teams participating in the conceptual design phase before requesting lead ship funds from Congress. Figure 7 reflects the budget request timeline for the FFG(X) detail design and lead ship contract award, as well as notable cost and design-related program activities that were planned to be completed after the request. The considerable cost growth that we have previously reported is common to many shipbuilding programs, as well as challenges in deviating from shipbuilding plans once a program has begun procuring ships, emphasize the importance of having a strong understanding of program expectations to back the initial procurement decision for FFG(X). Given the timing of the Navy’s budget request for lead ship funding, Congress faces a decision on whether to authorize funding for FFG(X) detail design and lead ship based on a budget request that was not informed by key cost and design information. If Congress authorizes and appropriates FFG(X) funding as the Navy requested in March 2019, it will be critical that the Navy demonstrate the program’s acquisition program baseline reflects the results of the component cost position and independent cost estimate before awarding the detail design and construction contract. Doing so before the contract award will help ensure a more reliable acquisition program baseline upon which future costs and variances are measured and funding decisions are made. Further, it will help mitigate remaining risk that stems from the Navy not being able to account for the actual FFG(X) design and associated estimated cost for ship construction until after the planned July 2020 contract award. Specifically, as currently planned, the Navy’s budget requests for fiscal years 2020 and 2021—which are intended to fund the first 3 ships—will be made before the Navy has agreed to contract pricing for FFG(X). Navy officials stated that they have completed a robust program life cycle cost estimate. They noted that the estimate was informed by Navy modeling of a notional ship design that leveraged data received from industry during conceptual design and reflected ship design elements needed to meet program requirements Navy officials also said that, as of May 2019, some additional work remains for the cost estimate to account for training and military construction considerations, as well as address any needed changes related to the final industry design reviews for the conceptual design phase. They also said that the program life cycle cost estimate informed the Gate 4 review in February 2019, and an updated version of the estimate will provide a basis for the Navy’s efforts to establish the component cost position in October 2019. As of the issuance of this report, we have requested the program life cycle cost estimate from the Navy, including the estimate’s criteria and underlying assumptions, but have not yet received this information. Recent Navy Changes in Cost Estimation Policy and the FFG(X) Program Schedule Create Uncertainty for Remaining Cost Estimation Activities Recent policy changes by the Navy related to cost analysis and estimation have created some uncertainty for Navy cost estimation activities going forward. Specifically, a March 2019 Secretary of the Navy instruction for acquisition program cost analysis shifts the Naval Center for Cost Analysis’s role and responsibilities for Navy cost estimation to the Navy’s systems commands. Previously, the Naval Center for Cost Analysis—organizationally residing completely outside of the systems command structure—would provide an independent cost assessment of the program life cycle cost estimate. The Naval Center for Cost Analysis and the acquisition program, in coordination with the relevant systems command, would discuss and adjudicate any differences between the program life cycle cost estimate and the independent cost assessment to produce the Navy’s component cost position. This independent cost assessment by the Naval Center for Cost Analysis was an important verification of the program office estimates, which were often found to be too optimistic, prior to the Navy finalizing its component cost position. The Navy’s recent changes for cost estimation and analysis may pose a risk of overly optimistic estimates carrying forward in programs. Navy officials stated that they believe Naval Sea Systems Command cost estimators can provide an independent cost estimate, as they are intended to provide technical support to acquisition programs independent of programmatic authority and report to a separate chain of command. However, as stated by the Naval Sea Systems Command, the collective mission of its organizations is to build, buy, and maintain the Navy’s ships. Based on this, we believe, as do CAPE officials with whom we spoke, that shifting independent cost assessment activities to the systems commands diminishes the Navy’s ability to independently verify a program life cycle cost estimate. As a result, the program life cycle cost estimate essentially will become the component cost position based on the lack of additional cost estimation input, such as what the Naval Center for Cost Analysis previously provided. Furthermore, CAPE officials stated that having a systems command execute cost analysis responsibilities for an acquisition program within the same system command effectively eliminates the Navy’s capacity to perform independent cost estimates for its programs based on their shared overarching mission. This position is consistent with the GAO Cost Estimating and Assessment Guide, which states that an independent cost estimate should be conducted by an organization independent of the acquisition chain of command. The Director of CAPE is required to conduct or approve independent cost estimates and cost analyses for all major defense acquisition programs. As noted by CAPE officials, CAPE has previously delegated certain cost estimation responsibilities to the Naval Center for Cost Analysis. With the recent Navy policy changes, CAPE may no longer choose to delegate independent cost estimation activities to Navy cost estimators. For FFG(X), CAPE intends to complete an independent cost estimate to verify the Navy’s component cost position. These plans include site visits and data collection from the shipyards participating in the conceptual design contracts. CAPE confirmed that the final independent cost estimate will reflect the content of the winning proposal, indicating that any FFG(X) proposals that the Navy receives from contractors not involved in the conceptual design phase will be evaluated to ensure the independent cost estimate accounts for those cost and design plans. CAPE officials also stated that their timeline for finalizing the independent cost estimate for FFG(X) is tied to when the Navy decides on the winning proposal for detail design and construction and communicates this information to CAPE. Specifically, CAPE’s final independent cost estimate will reflect only the winning FFG(X) design, so completion of the estimate will occur after the Navy informs CAPE about the FFG(X) design for which it intends to pursue a contract award. CAPE officials said that because the Navy’s decision may not be made in advance of the planned February 2020 Milestone B review for FFG(X), CAPE would likely just provide input to support the milestone and complete the independent cost estimate after that review. The Navy Has Taken Steps to Reduce Design and Technical Risk, but Technology Integration and Testing Will be Key to Meeting Program Expectations The Navy’s decision to pursue a parent ship design for FFG(X) was intended to reduce design uncertainty for the program. The Navy’s planned use of existing technologies for the ship’s mission and combat systems also supports reduced technical risk, though further maturation of some key systems and successful integration and testing will be critical to demonstrate the ship provides required capability within cost and schedule expectations. Use of a Parent Ship Design Was Intended to Increase Design Certainty Adopting a parent design requirement for FFG(X) provided the conceptual design industry teams with a proven baseline ship design. This enabled them to focus on incorporating modifications to meet the Navy’s specific FFG(X) requirements rather than designing a new ship. The Navy did not set any limitations on the extent contractors could modify or deviate from the parent design. However, Navy officials stated they actively reviewed parent design modifications through contract deliverables, technical exchange meetings, and design reviews with industry teams. The design reviews included an interim report in October 2018 and a final report in May 2019 from each industry team on their design progress. FFG(X) program officials noted that the design maturity reviews provided sufficient information to support the Navy’s decision that the designs were mature enough to release the request for proposals for the detail design and construction contract award. In addition, some industry officials told us that the conceptual design work on parent designs enabled them to develop more mature and refined designs than typical for this stage of the shipbuilding acquisition process. They also noted that continuing work in response to the pending competition should move at least some design elements closer to a detail design-level of maturity, and may provide the Navy with greater confidence in the contract proposals it receives from industry. Technology Re-Use Should Reduce Some Risk, but Integration and Testing Remain to Demonstrate Critical Systems The FFG(X) program’s design concept requires the use of many existing, more mature combat and mission systems to reduce technical risk. As stated in the approved acquisition strategy for FFG(X), the program has a requirement for all integrated systems to have achieved maturity of a technology readiness level (TRL) 6 or higher. TRL 6 is defined by GAO as the capability to produce a prototype system in a production-relevant environment. Program officials confirmed that, as of May 2019, many but not all FFG(X) integrated systems were at TRL 6 or higher. For selected key systems planned for FFG(X), Navy officials stated they will have achieved TRL 7 or higher by the planned July 2020 detail design and construction contract award. Doing so would be consistent with our acquisition best practices, which include maturing new key ship technologies into actual system prototypes and demonstrating them in a realistic environment—achieving a TRL 7—before the award of the contract for lead ship design and construction. This practice helps reduce the likelihood of costly design changes later. Technology Re-Use Many of the systems planned for FFG(X) have been demonstrated and are in use on other Navy ship classes, which helps the program fulfill capability needs while avoiding developmental risks. Table 3 provides an overview of some of the key existing systems planned for the ship. In addition to the systems that have been utilized by other Navy ships, the FFG(X) program plans to incorporate some systems that are still in development, such as the Enterprise Air Surveillance Radar (EASR) and a new version of the Aegis Weapon System. Navy officials stated that EASR—a complex radar system expected to provide long-range detection and engagement of advanced threats— is critical to FFG(X)’s air and surface warfare missions. It is a scaled down version of the Navy’s Air and Missile Defense Radar that is in production and scheduled for initial integration with the Aegis combat system on a DDG 51 Flight III destroyer in fiscal year 2020. In early 2019, the Navy began testing a full-scale, single-face EASR array engineering developmental model—the full system planned for FFG(X) will have three array faces—at a land-based test site to further demonstrate its functionality. The Navy expects to complete land-based testing of the EASR engineering development model by February 2020. The Navy also plans to integrate a rotating version of EASR and a fixed-face version on other ship classes prior to integrating the radar on the lead FFG(X). The Navy’s results from planned EASR developmental testing at the land-based site will be integral to achieving a TRL 7 and reducing risk prior to the start of FFG(X) detail design. The Navy is developing a new version of the Aegis Weapon System— FFG(X)’s combat management system—to coordinate radar and weapons system interactions from threat detection to target strike. For example, the system will support the ship’s ability to employ the Naval Strike Missile for over-the-horizon offensive capability as well as a 32- cell vertical launch system to employ missiles for air defense. The Aegis Weapon System for FFG(X) will leverage the Aegis common source software that supports the combat systems found on the Navy’s DDG 51-class destroyers and CG 47-class cruisers. Navy officials noted that they anticipate at least 70 percent of the Aegis Weapon System software for FFG(X) will be common to the Aegis software used for DDG 51 Flight III ships. Rigorous testing of the Aegis Weapon System with EASR will be critical for FFG(X), as the radar and combat management system must work in concert for the ship to detect, track, and assess possible targets. Given the radar and software commonalities, the risk level for both of these FFG(X) systems should be reduced once the DDG 51 Flight III radar and Aegis system baseline, upon which the FFG(X) integrated system is based, have been demonstrated through testing on a ship beginning in 2022. Specific to the Aegis Weapon System for FFG(X), software development is expected to run from fiscal year 2022 to late fiscal year 2024. The system’s integration and testing with EASR is scheduled to occur through fiscal year 2024. Integration and Testing While the Navy is planning to use many already mature systems on FFG(X), integration and testing of those systems will be critical to demonstrate systems fit and work together as intended on the ship. The Navy completed a technology readiness assessment in spring 2019 to identify potential technical risks, and concluded that FFG(X) does not have any critical technology elements. DOD generally defines a critical technology element as one that may pose major technological risk during development. Navy officials who completed the assessment stated that they reviewed about 150 systems as part of their activities and found none composed of new or novel technologies for which the Navy has insufficient knowledge to demonstrate maturity. The assessment noted one technology—the New Advanced Integrated Line-of-Sight Equipment System (nAILES) multi-coupler for antennas—as a watch item. The Navy would like to utilize nAILES for FFG(X), but according to Navy officials, it is not considered a critical technology because the Navy has identified alternative, proven technologies that will be used to meet the ship’s needs if nAILES is not available for use. The findings of the technology readiness assessment are consistent with the FFG(X) program’s decision to use existing systems that do not require technological innovation to deliver desired capability. However, the findings do not necessarily equate to the program having no technology risk for planned systems. For example, the Aegis Weapon System for FFG(X) did not qualify under the parameters of the technology readiness assessment as a critical technology element. Still, as already discussed, the Aegis Weapon System will carry technical risk for several years until the Navy completes development and demonstrates the system works as intended for FFG(X). The Next Generation Surface Search Radar is another system that is relatively mature—FFG(X) program officials confirmed in May 2019 it is nearing a TRL 6—but requires further development to reduce risk. The FFG(X) test and evaluation master plan and independent technical risk assessment are significant documents yet to be completed that will help to further define risks and plans to address them. The test and evaluation master plan serves to outline the program’s integrated test program and master schedule of major test events or phases. Navy officials expect the test plan to be approved in December 2019 to support the Milestone B decision. They noted that the plan may need to be updated once the FFG(X) design is selected based on the additional information that will be available to inform test planning. The independent technical risk assessment is intended to categorize risks that cover a broad range of factors, including technology maturity, integration needs, and testing. If these factors are not sufficiently accounted for, a program is likely to have difficulty meeting cost, schedule, and performance objectives. An official from the Office of the Under Secretary of Defense for Research and Engineering who is participating in the technical assessment for FFG(X) stated they plan to complete their work to identify any risks in March or April 2020. The official added that at this early stage of their activities, the potential for integration risks associated with the FFG(X) combat system is an area of interest because of the extensive number of existing systems that will need to be integrated into the new ship design. Navy test officials as well as DOD systems engineering and test officials noted potential advantages and risks related to FFG(X) program’s plans for using existing technologies. Similar to what we previously discussed about the use of a parent design, the officials stated that the use of existing systems can increase understanding of the ship and its systems, which may help the FFG(X) program achieve its planned accelerated timeline between development and delivery. However, systems engineering and test officials also indicated that, regardless of maturity, challenges typically arise when DOD takes systems from other platforms and attempts to integrate and use them in new ways on a new platform. They cautioned that programs like FFG(X) that plan to use a lot of government-furnished equipment or non-developmental systems often underestimate the amount of integration challenges they will face. The officials told us this may occur because of overconfidence that the maturity of systems demonstrated through use on other platforms eliminates most technical risk, whereas experience confirms that it is always challenging to get systems to fit and work together as intended on a new platform. Officials from the office of the Director, Operational Test and Evaluation said that the parent design approach for FFG(X) may enable the Navy to reduce some developmental testing activities; however, operational testing expectations would largely be unaffected because there will still be substantial integration to be completed and tested in order to demonstrate mission capabilities. Contracting Plans for FFG(X) May Help Mitigate Some Risk, and Use of Warranties Could Potentially Further Reduce Costs The draft FFG(X) request for proposal indicates that the Navy plans to use a fixed-price incentive contract to help control ship costs and special performance incentive fees. In addition, the Navy plans to use guarantees with limited liability for the shipbuilder to correct defects after ship deliveries. Our prior work has found that using comprehensive ship warrantees instead of guarantees could reduce the Navy’s financial responsibility for correcting defects. Use of Fixed-Price Incentive Contract Provides Benefits, but Planned Contract Structure Results in the Navy Absorbing More Cost Risk After completion of a full and open competition for FFG(X) detail design and construction, the Navy plans to use a fixed-price incentive contract in combination with additional special performance incentive fees to procure the lead and follow-on ships. As we have previously reported, full and open competition allows all responsible sources—or prospective contractors that meet certain criteria—to submit proposals for a contract. The use of competition in contracting is a critical tool for achieving the best possible return on investment for taxpayers. Competitively awarded contracts can save the taxpayer money, improve contractor performance, and promote accountability for results. The fixed-price incentive contracting approach for FFG(X) is intended to incentivize the contractor to control costs and meet performance requirements. This contracting strategy represents a significant departure from previous surface combatant programs in which the Navy negotiated cost-reimbursement contracts for construction of the lead ship. Under cost-reimbursement contracts, the Navy assumes the cost risk because the shipbuilder is reimbursed for its allowable incurred costs to the extent prescribed in the contract, regardless of whether the work is performed to the exact level desired by the Navy. For example, our prior work found that the Navy’s decisions to accept the first two LCS in incomplete, deficient conditions complied with federal acceptance provisions, largely due to the cost- reimbursement type contracts in place to construct these ships. Fixed-price incentive contracts provide an incentive for the shipbuilder to control costs in order to maximize profit. Fixed-price incentive contracts generally include a profit adjustment formula referred to as a shareline, as well as a price ceiling, target cost, and target profit. The structure of the shareline establishes how cost overruns or underruns in relation to a target cost are shared between the government and shipbuilder. For example, the 70/30 shareline that the Navy is planning for FFG(X) lead ship overruns means that the government pays 70 percent of cost and the shipbuilder pays 30 percent when the cost exceeds the target cost up to the price ceiling. Generally, the shareline functions to decrease the shipbuilder’s profit as actual costs exceed the target cost. The price ceiling is generally the maximum the government will pay under the contract and is typically negotiated as a percentage of the target cost. The target cost generally informs the shareline and price ceiling. Given the unknowns associated with design and construction, the Navy plans to account for these unresolved risks by assuming responsibility for cost growth above DOD recommended guidance. As we reported in March 2017, when the Navy assumes a greater share of cost overruns above the target cost, accepts a higher price ceiling, or both, the fixed- price incentive elements may not provide sufficient motivation for the shipbuilders to control costs. Figure 8 depicts the how risk changes as the Navy departs from a 50/50 shareline for cost overruns and a ceiling price of 120 percent. As we previously noted, for the FFG(X) lead ship the Navy is planning to have a shareline of 70/30 for target cost overruns. The Navy also plans to have a 60/40 target cost overrun shareline for the second ship, and a 50/50 overrun shareline for the remaining seven ships included in the detail design and construction contract award. Based on this plan, the first two FFG(X) ships will depart from DOD’s guidance recommending a 50/50 point of departure for negotiations between the government and shipbuilder for cost overruns up to the price ceiling. This results in more cost risk to the government for two ships in the detail design and construction contract. The Navy’s planned price ceiling for the 10 ships included in the contract award may deviate from DOD’s guidance recommending a ceiling price set at 120 percent of target cost as a point of departure for fixed price incentive contracts. Specifically, Navy officials stated that the maximum ceiling price could be as high as 125 percent for all of the ships. However, Navy officials stated that the request for proposal will provide incentive for industry to propose the minimum price ceiling that sufficiently accounts for the proposal’s level of risk, meaning that industry may propose price ceilings below 125 percent. The Navy also plans to include options for a special performance incentive fee for each of the FFG(X) ships, which will be established for the final request for proposal. These incentives have the potential to increase shipbuilder profitability. FFG(X) Plan for Guaranty Use Is More Robust than Recent Shipbuilding Programs, but Use of Warranties Could Provide More Value to the Government As outlined in the FFG(X) draft detail design and construction request for proposal and confirmed by program officials, each frigate will have a guaranty period that commences at ship delivery and is expected to end 18 months after delivery. Navy officials stated the guaranty is intended to formalize a period of responsibility during which the shipbuilder must correct defects, with the cost to the government and the contractor based on the contract terms (cost shareline and price ceiling) associated with the ship. During the guaranty period, the shipbuilder would be required to correct all defects for which it is responsible, with proposals required to include a minimum limitation of liability of $5 million per ship. Once the total cost to correct identified defects reaches $5 million, the government would pay the full cost to correct any additional guaranty period defects. The $5 million minimum limitation of liability planned for FFG(X) has a higher dollar value and covers a longer period of time than other recent shipbuilding programs. For example, we previously found that for the Navy’s LPD 25 amphibious transport dock construction, the contract initially included a $1 million limitation of liability. Navy officials stated that the final request for proposal also will include a provision allowing industry to propose a higher liability limit, up to and including no limitation of liability. Navy officials said that any additional liability amount proposed beyond the $5 million guaranty will be assessed as part of the technical evaluation criteria used to select the winning FFG(X) design. We found in March 2016 that the use of a guaranty did not help improve cost or quality outcomes for the Navy and Coast Guard ships we reviewed. We also found that commercial ship buyers and Coast Guard officials stated that warranties foster quality performance because the shipbuilder’s profit erodes as it spends money to correct deficiencies after delivery, during the warranty period. We further reported that the Coast Guard has improved cost and quality by requiring the shipbuilder to pay to repair defects by following Federal Acquisition Regulation warranty provisions. For example, the Coast Guard paid up front for the Fast Response Cutter warranty. The cost of the warranty amounted to 41 percent of the total defect correction costs. Although this ship does not have the size and advanced systems planned for FFG(X), it serves to demonstrate the potential value to the government presented by the use of warranties. The Coast Guard also used a fixed-price incentive contract with a warranty on its Offshore Patrol Cutter—a ship of comparable size to FFG(X). The first Offshore Patrol Cutter has a 2-year warranty, and follow-on ships will have 1-year warranties. The Coast Guard pays a set amount for these warranties, and in return, the shipbuilder must fix all applicable defects identified within the agreed-upon time period regardless of cost. Rather than using guarantees for the FFG(X) contract to provide for the correction of defects, the Navy could help control costs to the government through the use of warranties. Under warranties, the government generally receives a contractual right for the correction of all defects for which the shipbuilder is responsible at the shipbuilder’s expense. The use of warranties is typically not mandatory, but federal and defense acquisition regulations instruct contracting officers to consider various factors when deciding whether a warranty is appropriate for an acquisition. The regulations also instruct contracting officers to use a warranty when it is practicable and cost-effective to do so. We previously found that, unlike a warranty, the Navy almost exclusively paid for defects that were the shipbuilder’s responsibility under a guaranty because of the contract type and terms in contracts that we reviewed. Such conditions limit the incentive to discover every deficiency during the guaranty period, and may negatively affect quality improvements over time. The Navy’s FFG(X) plans suggest that the Navy may be prematurely discounting warrantees as a mechanism to improve ship quality and decrease cost to the government. Navy officials told us that mandating that industry propose a warranty could result in additional costs to the government because the initial cost of the ship could be raised substantially to include the cost of the warranty. Additionally, Navy officials said a requirement for warranty pricing could serve to limit industry participation in the FFG(X) competition if offerors are unwilling to accept the risk associated with a warranty and unable to provide reasonable pricing. The Navy provided no analysis to support these claims and confirm a clear understanding of whether a warranty could provide greater value than the $5 million guaranty the Navy is proposing for FFG(X). As part of the competitive proposal process for FFG(X) detail design and construction, the Navy could maintain its plans to require a guaranty but also seek ship warranty pricing. The full and open competition for the FFG(X) contract award may increase the potential for receiving warranty pricing that provides a cost-effective alternative to the Navy’s guaranty plans. By limiting the request for proposal to guarantees, the Navy misses an opportunity to obtain information on what comprehensive warranty coverage against defects would cost, and use it to evaluate whether warranties could further reduce risk for the FFG(X) program. Conclusions As the Navy approaches the Milestone B review for FFG(X), it is critical that funding and other major programmatic decisions are fully informed by the knowledge necessary to support them. This is especially important to help ensure that the FFG(X) program does not face some of the same cost, schedule, and performance shortfalls that have been faced by the LCS program. The Navy’s fiscal year 2020 budget request to authorize and appropriate funding for the lead frigate was developed and submitted without the benefit of key cost and design information, such as the independent cost estimate and the final results from conceptual design. As a result, it is necessary that the Navy provide Congress with a clear understanding of FFG(X) cost expectations, including CAPE’s independent cost estimate, prior to awarding the detail design and construction contract. This will help ensure that the FFG(X) program is grounded in cost and design expectations that reflect the specific aspects of the ship that the Navy selects for construction. With the start of the planned $20 billion FFG(X) procurement approaching, the Navy has limited time left to position the government to obtain the best deal possible to fix any deficiencies discovered upon delivery of the first 10 ships. The Navy’s guaranty plan for FFG(X) offers some improvements compared to recent shipbuilding programs, but does not offer the degree of coverage that could potentially be provided by a warranty. The competitive qualities of the FFG(X) acquisition approach present an opportunity for the Navy to, at a minimum, obtain warranty pricing from industry so that the program may use that input to evaluate whether a warranty would be a cost-effective means of reducing the government’s cost risk. Recommendations for Executive Action We are making two recommendations to the Secretary of the Navy: Ensure that the Assistant Secretary of the Navy for Research, Development, and Acquisition provides to Congress the finalized independent cost estimate prior to award of the detail design and construction contract and demonstrates that the estimate is consistent with the fiscal year 2020 budget request for the lead ship. (Recommendation 1) Ensure that the Assistant Secretary of the Navy for Research, Development, and Acquisition directs the FFG(X) program office to request pricing for warranties for the lead ship and the nine follow-on ship options planned for FFG(X) as part of the detail design and construction request for proposals. (Recommendation 2) Agency Comments and Our Evaluation We provided a draft of this report to DOD for comment. DOD provided written comments, which have been reproduced in appendix I. In responding to the draft report, DOD concurred and described the actions it planned to take to address our two recommendations. In response to the second recommendation to request pricing for warranties for the lead ship and the nine follow-on ship options planned for FFG(X) as part of the detail design and construction request for proposals, DOD acknowledged that the Navy will receive guaranty rather than warranty pricing, but stated that the solicitation allows industry to propose a higher limitation of liability amount, up to an unlimited limitation of liability, in its guaranty pricing for FFG(X). While this could allow for a better value to the government than has been typical for recent shipbuilding programs, permitting higher limitation of liability guaranty pricing but not requesting warranty pricing from offerors means the Navy will not have complete information on whether a warranty could be more cost-effective than a guaranty. Our prior work found that the use of Federal Acquisition Regulation warranty provisions improved shipbuilding program cost and quality outcomes. As a result, we maintain our belief that the FFG(X) program office should implement this recommendation by seeking warranty pricing as part of the detail design and construction request for proposals. The full and open competition for the FFG(X) contract award may increase the potential for receiving warranty pricing that provides a cost-effective alternative to the Navy’s guaranty plans. DOD stated that modifying the solicitation to incorporate a warranty pricing component would cause an unacceptable delay to the FFG(X) program, but did not provide an analysis to support this assertion or specify the extent of delay associated with adding a warranty pricing request. The current FFG(X) schedule has roughly 10 months between the request for proposals deadline and the contract award, and the program originally had been planning for the solicitation period to end in December 2019 before moving the deadline to September 2019 shortly before its release. We recognize the substantial effort the proposal development and review process requires, but we continue to believe that the government would benefit from adding a request for warranty pricing to the detail design and construction solicitation. While DOD stated that the Navy will support the recommendation after award by requesting pricing for an unlimited warranty before exercising the first ship option, doing so would eliminate any potential warranty pricing advantages that would occur as a result of the competitive conditions that currently exist for the current detail design and construction contract. In addition to DOD’s written response to the report, DOD officials and industry representatives associated with the FFG(X) conceptual design activities provided separate technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of Defense, and the Secretary of the Navy. This report will also be available at no charge on GAO’s website at http://www.gao.gov. If you or your staff members have any questions regarding this report, please contact me at (202) 512-4841 or oakleys@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to the report are listed in appendix II. Appendix I: Comments from the Department of Defense Appendix II: GAO Contact and Staff Acknowledgments GAO Contact Shelby S. Oakley, (202) 512-4841 or oakleys@gao.gov. Staff Acknowledgments In addition to the contact named above, the following staff members made key contributions to this report: Diana Moldafsky (Assistant Director), Lori Fields, Kurt Gurka, Stephanie Gustafson, Chad Johnson, Jennifer Leotta, Sean Merrill, Miranda Riemer, Jillena Roberts, Hai Tran, and Alyssa Weir.

Why GAO Did This Study In response to the shortcomings of the Navy's Littoral Combat Ship program and evolving threats, the Navy began the FFG(X) program. With FFG(X), the Navy intends to deliver a multi-mission ship that will provide anti-surface, anti-submarine, and air warfare capabilities. DOD approved FFG(X) requirements in February 2019.The Navy plans for a competitive contract award to support final FFG(X) design and construction. The program is expected to cost over $20 billion for 20 ships. The House report accompanying the National Defense Authorization Act for Fiscal Year 2019 included a provision for GAO to review the FFG(X) program. This report addresses, among other things, the FFG(X) acquisition approach and contracting plans. GAO reviewed requirements, acquisition, design, and cost-related documentation. GAO interviewed Navy and other defense officials, and conducted industry site visits to each shipyard participating in FFG(X) conceptual design activities. GAO also leveraged prior GAO reports and best practices guides. What GAO Found The Navy undertook a conceptual design phase for the FFG(X) Guided Missile Frigate program that enabled industry to inform FFG(X) requirements, identify opportunities for cost savings, and mature different ship designs. The Navy also streamlined the FFG(X) acquisition approach in an effort to accelerate the timeline for delivering the ships to the fleet. As shown in the figure, however, the Navy has requested funding for the FFG(X) lead ship even though it has yet to complete key cost estimation activites, such as an independent cost estimate, to validate the credibility of cost expectations. Department of Defense (DOD) cost estimators told GAO the timeline for completing the independent cost estimate is uncertain. Specifically, they stated that this estimate will not be finalized until the Navy communicates to them which FFG(X) design is expected to receive the contract award. GAO-identified best practices call for requisite cost knowledge to be available to inform resource decisions and contract awards. The Navy plans to use a fixed-price incentive contract for FFG(X) detail design and construction. This is a notable departure from prior Navy surface combatant programs that used higher-risk cost-reimbursement contracts for lead ship construction. The Navy also plans to require that each ship has a minimum guaranty of $5 million to correct shipbuilder-responsible defects identified in the 18 months following ship delivery. However, Navy officials discounted the potential use of a warranty—another mechanism to address the correction of shipbuilder defects—stating that their use could negatively affect shipbuilding cost and reduce competition for the contract award. The Navy provided no analysis to support these claims and has not demonstrated why the use of warranties is not a viable option. The Navy's planned use of guarantees helps ensure the FFG(X) shipbuilder is responsible for correcting defects up to a point, but guarantees generally do not provide the same level of coverage as warranties. GAO found in March 2016 that the use of a guaranty did not help improve cost or quality outcomes for the ships reviewed. GAO also found the use of a warranty in commercial shipbuilding and certain Coast Guard ships improves cost and quality outcomes by requiring the shipbuilders to pay to repair defects. The FFG(X) request for proposal offers the Navy an opportunity to solicit pricing for a warranty to assess the cost-effectiveness of the different mechanisms to address ship defects. What GAO Recommends GAO recommends that the Navy provide Congress with the independent cost estimate for FFG(X) prior to the detail design and construction contract award and seek ship warranty cost information from industry as part of the request for proposal process. While DOD generally concurred with GAO's recommendations, it did not agree to update its request for proposal to solicit ship warranty pricing. GAO continues to believe this is an essential element of the recommendation, as discussed in the report.

gao_GAO-19-433

gao_GAO-19-433_0

Background Nursing homes are required to keep residents safe from harm, but when abuse is alleged, a combination of federal, state, and local agencies—as well as the nursing homes themselves—play a role in investigating. Federal Oversight of Nursing Homes Federal laws establish minimum requirements nursing homes must meet to participate in the Medicare and Medicaid programs, including standards for the quality of care. These standards cover a variety of categories, such as resident rights, quality of care, and quality of life. In 2016, CMS finalized a comprehensive update to its nursing home standards to reflect new requirements and align requirements with current clinical practices, among other things. The changes were implemented in three phases, starting November 28, 2016. The federal government and the states share oversight responsibility for the nation’s nursing homes, with specific activities occurring at the national, regional, and state levels. CMS central office. At the national level, the CMS central office oversees the federal standards nursing homes must meet to participate in the Medicare and Medicaid programs. Primarily through its State Operations Manual, the office establishes the responsibilities of CMS’s regional offices and state survey agencies in ensuring that federal quality standards for nursing homes are met. CMS regional offices. CMS’s 10 regional offices oversee state activities and report back to the CMS central office the results of their efforts. Specifically, regional offices use the State Performance Standards System to evaluate state surveyors’ performance on factors such as the frequency and quality of state surveys. State survey agencies. Under agreement with CMS, a state survey agency in each state assesses whether nursing homes meet CMS’s standards, allowing them to participate in the Medicare and Medicaid programs. State survey agencies assess nursing homes using (1) recurring standard surveys and (2) as-needed investigations. Standard surveys. State survey agencies are required by federal law to perform unannounced, on-site standard surveys of every nursing home receiving Medicare or Medicaid payment at least every 15 months, with a statewide average frequency of every 12 months. These surveys are a comprehensive assessment designed to determine whether nursing homes are complying with Medicare and Medicaid quality standards. Investigations. In addition to standard surveys, state survey agencies are required by federal law to investigate (1) complaints submitted by residents, family members, friends, physicians, and nursing home staff; and (2) “facility-reported incidents,” including incidents involving abuse of residents, that are self-reported by the nursing homes. State survey agencies review the information provided through these complaints and incidents and determine if an on-site investigation is required. During this unannounced investigation, the state surveyors assess available evidence to determine whether the allegation can be substantiated. These investigations offer the state survey agency the opportunity to identify and correct care problems in a more timely manner than through the standard surveys. If a surveyor determines that a nursing home violated a federal standard during a survey or investigation, then a deficiency code specific to that standard is cited. For instance, one deficiency code for abuse of residents encompasses mental/verbal, sexual, or physical abuse; while a few additional deficiency codes encompass abuse-related issues, such as a failure by the nursing home to train staff on issues related to abuse. Cited deficiencies are then classified into categories according to scope (the number of residents potentially affected) and severity (the potential for or occurrence of harm to residents). (See table 1.) State survey agencies are required to enter data about deficiencies into CMS’s survey database. For most deficiencies, the nursing home is required to prepare a plan of correction, and, depending on the scope and severity of the deficiency, surveyors may re-visit the facility to ensure that the nursing home has implemented its plan and corrected the deficiency. In any instances where surveyors substantiate the occurrence of resident abuse, the state survey agency is required to refer the case to three entities: 1) local law enforcement; 2) the MFCU, if appropriate; and 3) the state’s nurse aide registry or other applicable professional licensure authority. When nursing homes are cited with deficiencies, federal enforcement actions—or penalties—can be imposed to encourage homes to make corrections. In general, enforcement actions: (1) may be initially recommended by the state survey agency, (2) are transferred to the CMS regional office for review, (3) are imposed by the same CMS regional office, and (4) are implemented—that is, put into effect. Depending on the scope and severity of the deficiency cited, the CMS regional office may impose certain enforcement actions so that they are implemented immediately. However, for other enforcement actions, the regional office may provide the nursing home with an opportunity to correct the deficiencies, which, if corrected before the scheduled effective date, can result in the penalty not being implemented. Penalties include directed in- service training, fines known as civil money penalties, denial of payment, and termination from the Medicare and Medicaid programs, among others. (See fig. 1.) Reporting and Investigation of Abuse by Nursing Homes When a nursing home becomes aware of an incident of alleged resident abuse, the home must: immediately report the allegation to the state survey agency and then conduct an investigation of the alleged incident. Specifically, the process is as follows: The nursing home must immediately report alleged abuse to the state survey agency. After notifying the state survey agency, the nursing home is also required to conduct its own investigation and submit its findings in a written report to the state survey agency within 5 working days of the incident. Depending on the severity of the circumstances, the state survey agency may visit the nursing home to investigate the incident or wait until the nursing home submits its report. Depending on the content of the report, the state survey agency may request the home conduct additional work or the state survey agency may investigate further on its own. If the state survey agency opts not to investigate further, it may still review the manner in which the home conducted its investigation during the state survey agency’s next scheduled standard survey. If a state survey agency determines that a nurse aide is responsible for abuse, the agency must add this finding to the state’s nurse aide registry—a registry that each state is required to maintain that lists all individuals who have satisfactorily completed approved nurse aide training and a competency evaluation program in that state. Nursing homes are prohibited from employing a nurse aide with a finding of abuse on the nurse aide registry. Further, if there is a reasonable suspicion that a crime has occurred that results in serious bodily injury, federal law requires certain covered individuals at the nursing home to immediately report to law enforcement in addition to the state survey agency. Before employing a nurse aide, nursing homes are required to check each relevant state’s registry to verify that the nurse aide has passed a competency evaluation. All nursing homes must also verify with the relevant state board of licensing the professional credentials of the licensed personnel, such as registered nurses, whom they hire. Other State and Local Agencies That May Investigate Abuse in Nursing Homes In addition to state survey agencies, there are other state and local agencies that may be involved in investigating abuse in nursing homes. These other state and local agencies that investigate abuse in nursing homes are generally focused on the different aspects of the specific alleged abuse incident, in contrast to the state survey agency, which focuses on the safety of individual residents, as well as on the facility’s policies and procedures for preventing and effectively addressing abuse. These other state and local agencies include: Adult Protective Services. In some states, Adult Protective Services’ investigators are trained to provide protection and intervention for older adults in nursing homes and can play a valuable role in helping to protect residents from abuse. Ombudsmen. Long-term care ombudsmen, who serve as advocates for nursing home residents, may also investigate abuse complaints made by or on behalf of residents. Local law enforcement. Law enforcement may also play a role in investigating alleged nursing home resident abuse. Specifically, local police departments may learn of suspected instances of resident abuse and conduct criminal investigations. MFCU. The state MFCUs typically learn of abuse allegations through referrals from state survey agencies, which CMS requires if abuse is substantiated. If, after investigating an allegation, the MFCU decides that there is sufficient evidence to press criminal charges, it may prosecute the case itself or refer the matter to the state’s attorney general or a local prosecutor. More Abuse Deficiencies Were Cited in Nursing Homes from 2013 through 2017; Physical and Mental/Verbal Abuse and Staff Perpetrators Were Most Common Abuse Deficiencies Cited and the Number of Nursing Homes Involved More than Doubled from 2013 through 2017, with the Largest Increase in Severe Cases Our analysis of CMS data found that from 2013 through 2017, abuse deficiencies cited in nursing homes became more frequent, with the largest increase in severe cases. While abuse deficiencies are relatively rare—they comprise less than 1 percent of the total deficiencies in each of the years we examined—they became more common over the 5-year period. Specifically, the number of abuse deficiencies cited more than doubled—from 430 in 2013 to 875 in 2017 (a 103.5 percent increase). This trend for the abuse deficiencies is in contrast to the trend for all deficiencies, which decreased about 1 percent between 2013 and 2017. At the state level, 32 states had more abuse deficiencies cited in 2017 than 2013, six states had a consistent number, and the remaining 13 had fewer. (See app. III for additional data on abuse deficiencies by state.) Furthermore, abuse deficiencies cited in 2017 were more likely to be categorized at the highest levels of severity—deficiencies causing actual harm to residents or putting residents in immediate jeopardy—than they were in 2013. Specifically, 42.6 percent of the 875 abuse deficiencies were categorized as causing actual harm or posing immediate jeopardy to residents in 2017, compared to 31.9 percent of the 430 abuse deficiencies in 2013. (See fig. 2.) In examining the types of survey or investigations conducted to identify abuse deficiencies, we found that, from 2013 to 2017, the majority (about two-thirds in each year) were identified through either a complaint investigation or facility-reported incident investigation. In contrast, for all types of deficiencies, we found the inverse—the vast majority were identified through a standard survey. This demonstrates the unique and significant role that complaint and facility-reported incident investigations have in identifying abuse deficiencies, because they allow for the identification and correction of abuse in a more timely manner than a standard survey. In fact, for the deficiencies for which we were able to identify the source, the percentage of abuse deficiencies identified through facility-reported incident investigations increased from 42.3 percent of the 430 abuse deficiencies in 2013 to 47.4 percent of the 875 abuse deficiencies in 2017. Conversely, for all types of deficiencies, a very small percentage resulted from facility-reported incident investigations—about 5 percent or less each year. (See fig. 3.) We found that enforcement actions—or penalties—were imposed and implemented by CMS infrequently each year in response to abuse deficiencies, and that fines were the most common type of implemented penalty. Specifically, for each year from 2013 through 2017, we found that about one-third of abuse deficiencies had an enforcement action imposed but not implemented, and less than 8 percent of abuse deficiencies had enforcement actions that were implemented against the nursing home. This was fairly consistent over the 5-year period. For example, in 2017, of the 875 abuse deficiencies cited, 275 (31.4 percent) resulted in enforcement actions that were imposed but not implemented and 65 (7.4 percent) had enforcement actions that were implemented against the nursing home. Furthermore, for abuse deficiencies cited at the most severe levels—that is, those causing actual harm or immediate jeopardy to residents—a smaller percentage of the deficiencies had an enforcement action imposed but not implemented compared to all abuse deficiencies, but a larger percentage were implemented. For example, in 2017, 373 of the 875 abuse deficiencies were cited at the most severe levels; of those, 81 (21.7 percent) resulted in enforcement actions that were imposed but not implemented, and 51 (13.7 percent) were implemented against the nursing home. Regardless of the severity, the predominant reason that CMS did not implement imposed enforcement actions was because the nursing home came into compliance prior to the implementation date of the penalty. For implemented enforcement actions, fines—known as civil money penalties—were overwhelmingly the most common type of penalty implemented against nursing homes with abuse deficiencies, increasing from 69.6 percent of the 23 abuse deficiencies with implemented enforcement actions in 2013 to 83.1 percent of the 65 in 2017. Denial of payments for new Medicare and Medicaid admissions—another financial penalty—was the second most common type of implemented enforcement action, but decreased from 34.8 percent in 2013 to 13.8 percent in 2017. Mandatory termination is the most severe enforcement action as it ends all payments for Medicare and Medicaid residents; it is implemented very rarely, with only one abuse deficiency resulting in mandatory termination of the nursing home across all 5 years. (See fig. 4.) In addition, we found the number of nursing homes with abuse deficiencies also more than doubled over the 5-year period. In 2013, 394 nursing homes (2.7 percent of all surveyed nursing homes) had at least one abuse deficiency compared to 821 nursing homes (5.6 percent of all surveyed nursing homes) in 2017. A nursing home may have more than one abuse deficiency cited in a single year, such as from a standard survey early in the year and then a complaint investigation later in the year. We found that in 2013, of the 394 nursing homes that had a total of 430 abuse deficiencies cited, 85 of the homes had two or more abuse deficiencies that year. In 2017, of the 821 nursing homes that had 875 total abuse deficiencies cited, 155 had two or more that year. Further, across the 5-year period, we found that a small proportion of all nursing homes with abuse deficiencies had them in multiple consecutive years. Specifically, across all years, 2,214 total unique nursing homes (13.6 percent of all surveyed nursing homes) had at least one abuse deficiency. A small portion of these nursing homes had at least one abuse deficiency in multiple consecutive years, indicating potential patterns in abuse at these nursing homes. Specifically, 185 of the 2,214 nursing homes with abuse deficiencies over the 5-year period—8.4 percent—had an abuse deficiency in any 2 consecutive years. In addition, 25 of the nursing homes—1.1 percent—had an abuse deficiency in 3 or more consecutive years. (See fig. 5.) Finally, we analyzed a selection of characteristics, including ownership type and bed size, for these nursing homes that had abuse deficiencies cited in multiple years and compared them to homes that had abuse deficiencies cited in a single year and surveyed homes that did not have any abuse deficiencies. We found that the nursing homes differed. For example, while for-profit organizations—the largest ownership group accounting for 67.9 percent of all surveyed nursing homes—owned 66.9 percent of nursing homes without any abuse deficiencies cited over the 5- year period, they accounted for 78.6 percent of nursing homes that had abuse deficiencies cited in 2 or more years. In addition, nursing homes designated as Special Focus Facilities—a CMS program that provides increased oversight to homes with consistent poor performance— constituted 2.5 percent of all surveyed nursing homes compared to 1.9 percent of nursing homes without abuse deficiencies and 10.1 percent of nursing homes with abuse deficiencies cited in 2 or more years. (See table 2.) Physical and Mental/Verbal Abuse Occurred Most Often, Followed by Sexual Abuse, and Staff Were More Often Perpetrators of Abuse in 2016 and 2017 Our analysis of a representative sample of CMS narrative descriptions— written by state surveyors—associated with abuse deficiencies cited in 2016 and 2017 found that physical and mental/verbal abuse occurred most often in nursing homes, followed by sexual abuse. Further, staff were more often the perpetrators of the deficiencies cited as abuse than were residents or others. (See fig. 6.) Physical abuse, which CMS defines as hitting, slapping, punching, biting and kicking residents, was present in about 46 percent (+/- 5 percent) of the abuse deficiency narratives. Mental/verbal abuse, which CMS defines as verbal or nonverbal conduct that can cause a resident to experience humiliation and fear, among other things, was present in about 44 percent (+/- 5 percent) of the abuse deficiency narratives. Sexual abuse, which CMS defines as non-consensual sexual contact with a resident, was present in about 18 percent (+/- 5 percent) of the abuse deficiency narratives. Staff, which includes those working in any part of the nursing home, were perpetrators in 58 percent (+/- 5 percent) of abuse deficiency narratives, followed by resident perpetrators (30 percent +/- 5 percent) and other types of perpetrators (2 percent +/- 5 percent). Other types of perpetrators can include family members of residents or other visitors. Further, our analysis of the narratives found that sexual abuse perpetrated by residents (39 percent) occurred more frequently within our sample than sexual abuse perpetrated by staff (10 percent) or others (17 percent). When staff were the perpetrators of abuse, we found within our sample that mental/verbal abuse was the most common type of abuse (60 percent), while physical abuse was most common in situations where residents (59 percent) or others (67 percent) were the perpetrators. For examples of the different types of abuse and perpetrators from our analysis, see table 3 below. Within our sample of narratives, mental/verbal abuse was less likely to be categorized by surveyors as severe compared to physical and sexual abuse. Specifically, we found in our sample that the proportion of mental/verbal abuse (30 percent) categorized by state surveyors as severe—defined as actual harm or immediate jeopardy—was smaller than the proportion of physical (40 percent) and sexual abuse (58 percent) categorized as severe. In addition, we found that most of the mental/verbal (88 percent), physical (91 percent), and sexual abuse (77 percent) narratives in our sample were categorized by surveyors as “isolated” in scope. Stakeholders Identified Resident Characteristics and Staffing Inadequacies as Risk Factors for Abuse, and Underreporting as among the Challenges to Investigating Abuse Some Resident Characteristics and Inadequacies in Staffing, Training, and Staff Screening Can Increase Risk of Abuse Stakeholder groups in most of the five states we interviewed—including state survey agencies, Adult Protective Services, law enforcement, MFCUs, ombudsmen, and nursing home administrators and clinical staff—identified risk factors for abuse in nursing homes that included resident characteristics, such as residents with infrequent visitors, and nursing home staffing characteristics, such as insufficient staffing levels. (See table 4 for a description of these risk factors.) Officials we interviewed from national organizations with knowledge of abuse in nursing homes also noted some of these same risk factors. Resident characteristics. Stakeholders in each of our five selected states noted that residents who do not have frequent visitors, are cognitively impaired, or mixed with widely different age groups may be at an increased risk for abuse. Residents who do not have frequent visitors. Stakeholders in four of the five states said that residents without regular visitors, such as family, may be at an increased risk for abuse because regular visitors could notice and report potential warning signs of abuse, such as changes in their behavior or physical appearance. Residents who are cognitively impaired. Stakeholders in each of the five states said that cognitively impaired residents may be especially vulnerable to abuse because they often cannot speak or may have difficulty recalling recent events, and they are therefore less likely to be able to remember or describe what happened. In addition to noting that cognitively impaired residents may be at an increased risk of abuse, some stakeholders said that some cognitively impaired residents may be more likely to be perpetrators of abuse as their condition can have behavioral symptoms, such as physical aggressiveness. Residents mixed with widely different age groups. Stakeholders in four of the five states also noted that elderly nursing home residents who are mixed with widely differing age groups, such as young adults with mental illness, may be at a higher risk for incidents of abuse due to the different characteristics of these groups. Combining these two populations, which have differing needs, can also be challenging for staff. For example, staff may have more experience caring for elderly residents with complex needs, such as dementia, and they may not have the necessary skills or training to care for needs of younger residents, who require other types of complex care. This can create a stressful environment for staff, which is a risk factor for staff as potential perpetrators of abuse. Two stakeholders noted that younger residents who may have mental illness can have conflicts with older and frailer residents, potentially leading to abusive incidents between residents. Nursing home staffing characteristics. Stakeholders we interviewed in each of our five selected states noted that nursing homes with insufficient staffing, inadequate staff training, and inadequate staff screening may be at risk for abuse. Nursing homes with insufficient staff. Stakeholders in each of the five states said that nursing homes with insufficient staff may be at risk for abuse because there may not be enough staff attending to the needs of residents. Stakeholders noted that nursing homes have faced challenges hiring and retaining qualified staff and that, as a result, existing staff can feel overworked, stressed, or exhausted, which can lead to abusive behaviors. Staffing issues are not just risk factors for staff as perpetrators of abuse, but they can also limit a staff member’s ability to identify and report abuse. For example, insufficient staffing may mean that there are not enough available staff to notice signs of abuse in a timely fashion, such as noticing a resident’s bruises before they heal. Nursing homes with inadequate staff training on abuse. Inadequate staff training on abuse was noted by stakeholders we interviewed in four of the five states as a risk factor for abuse because; for example, staff may not know how to diffuse challenging situations with residents and identify and report abuse. As previously noted, recognizing abuse can be challenging and, even when abuse is identified, it is often not reported. Officials from all of the nursing homes that we visited said that they provide training to their staff on abuse, including on defining abuse, identifying or detecting different types of abuse, and reporting abuse. Staff members we spoke with at one nursing home said that, not only are they trained to look for physical signs of abuse, such as bruising, but they are also trained to observe changes in behavior that may be warning signs for abuse, such as a resident suddenly withdrawing from group activities. Staff at another nursing home said that they are also taught to ask another staff member for assistance when they are feeling frustrated or stressed by caring for a particular resident. In contrast, staff at another nursing home noted the challenges of not having these types of resources and said they are needed at their facility. Nursing homes with inadequate staff screening. Stakeholders in three of our five states said that inadequate staff screening can be a risk factor for abuse. Some stakeholders said that a thorough background screening can be time consuming. Further, because staff screening through background checks and the nurse aide registry is not coordinated across the country, there are gaps that could enable individuals who committed crimes in one state to obtain employment at a nursing home in another state, a concern that we previously reported. Staff from a nursing home we visited said the prevention of abuse “starts with hiring the right staff” and noted the importance of conducting background checks and checking references for prospective employees. Underreporting of Abuse, Cognitive Impairment of Victims, Lack of Nursing Home Cooperation, and Lack of Agency Coordination Pose Challenges for Abuse Investigations The key challenges for abuse investigations most frequently identified by stakeholder groups in the five states we reviewed were underreporting of abuse, cognitive impairment of victims, lack of cooperation from nursing homes, and lack of agency coordination. (See table 5 for a description of these challenges.) Officials we interviewed from national organizations with knowledge of abuse in nursing homes also noted some of these same challenges. Underreporting of abuse. Stakeholders in each of the five states in our review noted that abuse in nursing homes may be underreported because residents or their families feel uncomfortable or fear retaliation from nursing home staff. For example, residents who were sexually abused may feel ashamed or embarrassed to report these incidents. In addition, residents may fear retaliation by the nursing home staff on whom they depend, which might include substandard care, exclusion from activities, or even eviction from the home. A fear of retaliation can also extend to nursing home staff, who may witness abuse by another staff member, but may be afraid to report it out of fear that they will lose their jobs or that they will face retaliation from co-workers. This underreporting creates challenges for investigators, who are unable to investigate if they do not know that abuse has occurred. Cognitive impairment of victims. Stakeholders in each of the five states in our review said that victims with cognitive impairment may not be able to give statements regarding the abuse or may not be considered reliable witnesses. For example, residents with dementia may not be able to remember the details of an abusive incident, and their memory of the details may deteriorate over the course of an investigation. Or, residents with dementia may report abuse that stems from traumatic memories from an incident that occurred earlier in their lives. One stakeholder said this can be a challenge for investigations because they do not know how much they can rely on a cognitively impaired resident’s statement, making it difficult for them to corroborate an abuse allegation. However, one stakeholder noted that, while it can be difficult to interview abuse victims with cognitive impairment, it is important to treat their allegations seriously and with credibility. One law enforcement stakeholder noted that interviews with these victims require special training. Lack of cooperation from some nursing homes. Stakeholders in each of the five states in our review said that some nursing homes may withhold, alter, or make it difficult for investigatory agencies to gain access to necessary, timely, or accurate information about alleged abuse. This may be, for example, because they may fear adverse publicity, litigation, or penalties from the state or CMS. In addition, as noted previously, nursing home staff may be fearful of losing their jobs. Stakeholders said that nursing home staff who witnessed abuse may be intentionally vague when interviewed by investigators; for example, by saying they cannot recall an incident. Some stakeholders also noted that nursing homes may delay investigators’ access to patient records, or they may even alter patient records in order to fill in information that should have been documented but was not at the time of the incident. One stakeholder we interviewed noted that the problem is not necessarily widespread—that some nursing homes are open about sharing information while others can be more difficult. Another stakeholder noted that a nursing home’s cooperation can sometimes depend on the seriousness of the allegation. Lack of agency coordination. Stakeholders in three of the five states in our review said that having multiple agencies involved in investigations, such as the state survey agency, law enforcement, the ombudsman, and, in some states, Adult Protective Services, can create challenges, including coordinating investigations and notifying one another about investigation outcomes. One stakeholder said they sometimes begin an investigation without realizing another investigatory agency has already started its own investigation. Further, stakeholders in two of the five states in our review said that CMS does not allow state survey agencies to share important investigatory information with law enforcement. (We discuss this issue in more detail later in this report.) CMS’s Ability to Ensure Nursing Home Residents Are Free from Abuse May Be Limited by Gaps in Oversight We found that CMS: (1) cannot readily access data on the type of abuse or type of perpetrator, (2) has not provided guidance on what information nursing homes should include in facility-reported incidents, and (3) has numerous gaps in its referral process that can result in delayed and missed referrals to other entities. Together, these gaps affect critical points in CMS’s oversight of abuse in nursing homes including the prevention, identification, and timely investigation of abuse. Information on Abuse and Perpetrator Types Is Not Readily Available CMS cannot readily access information on abuse or perpetrator type in its datasets and, as a result, lacks key information critical to understanding and appropriately addressing nursing home abuse with its oversight. Specifically, in two of CMS’s datasets—complaints/facility-reported incidents and deficiencies—agency officials told us they do not require the state survey agencies to record abuse and perpetrator type. As a result, we found that CMS’s data do not readily support CMS’s understanding of the types of abuse and perpetrators that are most prevalent in nursing homes. CMS officials told us they believe that the majority of abuse is committed by nursing home residents, and that physical and sexual abuse were the most common types; officials said they based this current understanding of abuse and perpetrator types on professional experience, literature, and ad hoc analyses of deficiency narrative descriptions. However, our review of a representative sample of abuse deficiency narratives from 2016 and 2017 found that staff were more often the perpetrators of deficiencies cited as abuse than residents or others, and that physical and mental/verbal abuse occurred most often in nursing homes, followed by sexual abuse. CMS officials noted that some incidents resulting from resident altercations—particularly those that do not show a willful intent to harm—may not have been cited as an abuse deficiency by some state survey agencies and may have been cited as other deficiencies not specified as abuse. This may have contributed to the difference between CMS’s understanding of the prevalence of resident to resident abuse and what their abuse deficiency data show. If CMS required information on abuse and perpetrator type to be recorded, the agency would have a better understanding of abuse in nursing homes. However, CMS officials told us they do not currently require the state survey agencies to specify abuse and perpetrator type because they consider the surveyor’s job to be identification and documentation of noncompliance. Additionally, CMS officials told us they have not conducted a systematic review to gather information on abuse and perpetrator type. This is inconsistent with federal internal control standards directing management to use quality information to achieve program objectives. Without the systematic collection and monitoring of specific abuse and perpetrator data, CMS lacks key information and, therefore, cannot take actions—such as tailoring prevention and investigation activities—to address the most prevalent types of abuse or perpetrators. Facility-Reported Incidents Lack Key Information All of the state survey agencies we spoke to told us that facility-reported incidents can lack key information that can cause potential delays in abuse investigations. Specifically, officials from each of the five state survey agencies told us that the facility-reported incidents they receive from nursing homes can lack key information that affects their ability to effectively triage incidents and determine whether an investigation should occur and how soon. Two state survey agencies we spoke with said they sometimes have to conduct significant follow-up with the nursing homes to obtain the information they need to prioritize the incident for investigation—follow-up that delays and potentially negatively affects investigations. For example, one state survey agency told us that a facility reported abuse involving two residents but did not initially report that the residents were injured, and that the facility did not file an addendum to the facility-reported incident to indicate resident injury. As a result of this incomplete information, the state survey agency did not properly prioritize this incident response. Despite federal law requiring nursing homes to self-report allegations of abuse, and covered individuals to report reasonable suspicions of crimes against residents, CMS has not provided guidance on what information should be included in these reports. Our review of CMS’s State Operations Manual found that CMS does not have guidance related to the information that nursing homes or covered individuals should report to the state survey agencies or local law enforcement; in contrast, it does contain guidance on the type of information members of the public should include in a complaint about nursing home quality to the state survey agency—and CMS makes a standardized complaint template form available on its website. The lack of guidance on the information that state survey agencies should collect on facility-reported incidents is inconsistent with federal internal control standards directing management to use quality information to achieve program objectives. CMS could outline basic information requirements that states must include on incident forms used by nursing homes and covered individuals to ensure the state survey agency is receiving the information it needs to accurately and quickly triage these incidents. CMS officials told us in November 2018 that they have efforts underway to examine guidance related to the information state survey agencies need to appropriately triage these facility-reported incidents and are developing a facility-reported incident template. Until the guidance and template are in place, these facility-reported incidents may lack key information that can cause potential delays in abuse investigations. Gaps Exist in CMS Process for State Survey Agency Referrals to Law Enforcement and MFCUs CMS requires state survey agencies to make referrals to law enforcement and, if appropriate, to MFCUs when abuse is substantiated; however, we found numerous gaps in CMS’s referral process that can result in delayed and missed referrals. (See table 6.) Timing of abuse referrals. We found CMS’s requirements for when state survey agencies should report abuse to law enforcement and MFCUs lag behind the federal requirements for when covered individuals should make such referrals, and, as a result, referrals may be significantly delayed. Specifically, federal law requires covered individuals to immediately report reasonable suspicions of a crime against a resident that results in serious bodily injury to law enforcement and the state survey agency. Conversely, state survey agencies do not have to report suspicions of crime identified on complaints submitted to, and surveys conducted by, the state survey agency until the abuse has been substantiated—a process that can often take weeks or months. Officials from one law enforcement agency and two MFCUs that we interviewed told us the delay in receiving referrals limits their ability to collect evidence and prosecute cases—for example, bedding associated with potential sexual abuse may have been washed and wounds may have healed. This is consistent with the findings of our 2002 report, where we recommended that CMS should ensure that state survey agencies immediately notify law enforcement or MFCUs when nursing homes report allegations of physical or sexual abuse. One state survey agency in our review established more stringent guidelines than CMS by requiring the surveyors to notify law enforcement and the MFCU promptly upon receiving a complaint of abuse. CMS officials told us their state survey agency reporting requirements are based on a March 2002 policy. This is inconsistent with standards for internal control, which state that management should communicate quality information externally so that external parties can help the entity achieve its objectives. Tracking of abuse referrals. In addition to delays in referring cases to law enforcement and MFCUs, CMS officials also told us that CMS does not conduct oversight to ensure that state survey agency referrals to law enforcement and the MFCUs are occurring as required for substantiated abuse, and, as a result, CMS cannot ensure that state survey agencies are complying with reporting obligations. For example, an official from one of the five state survey agencies we interviewed said they had never made a referral to law enforcement or the MFCU, despite having substantiated allegations of abuse. The state survey agency official told us that they do not refer cases to law enforcement, and that law enforcement referrals are the responsibility of the nursing home. This is incompatible with CMS guidelines requiring that substantiated abuse be referred to law enforcement; however, CMS officials told us that they do not track whether state survey agencies make referrals to law enforcement and the MFCUs. This is inconsistent with federal standards for internal control, which state that management should establish and operate monitoring activities to monitor the internal control system and evaluate the results. Definition of substantiated abuse. We found confusion among some state survey agencies about CMS’s definition of what it means to substantiate an allegation of abuse—a challenge because substantiation is a trigger in the investigation process, and CMS requires state survey agencies to make referrals to law enforcement and staff registries when abuse is substantiated by evidence. As a result, there is a potential for substantiated abuse to not be reported and, subsequently, not referred to law enforcement or MFCUs for criminal investigation. Two of the five state survey agencies in our review told us they believed they could not substantiate an allegation unless they could also cite a federal deficiency. This is inconsistent with CMS’s guidance, which says that state survey agencies can substantiate that an allegation occurred without citing a federal deficiency and that, subsequently, these substantiated allegations must be referred to law enforcement and staff registries. For example, according to CMS guidance, if the state survey agency investigated and found evidence that a resident was abused, but the nursing home had taken preventive actions against the deficient practice, the state survey agency would then substantiate that the abuse occurred, but not cite a deficiency. However, state survey agencies may decide not to substantiate an abuse allegation verified by evidence if they believe no deficiency should be cited, such as if the nursing home had taken preventive action against the deficient practice, which could result in that abuse going unreported and not referred to law enforcement, MFCUs, or staff registries. Because substantiation of abuse is a critical trigger in abuse investigations, confusion around its interpretation could prevent these important next steps. CMS officials told us they are aware that the state survey agencies have varying interpretations of what it means to substantiate abuse. According to federal standards for internal control, management should internally communicate quality information to achieve the entity’s objectives. Information sharing. We also found that CMS’s guidance on state survey agency referrals contained in its State Operations Manual does not specify what incident information can be shared with local law enforcement, either in response to local law enforcement’s request for information or when the state survey agency refers substantiated findings of abuse to local law enforcement. As a result, both state survey and law enforcement agencies expressed confusion and frustration about what information can be shared and said delays have occurred that can impede law enforcement investigations. Officials from two state survey agencies told us that CMS does not allow them to share any information with law enforcement without a written request. For example, officials from one state survey agency said that they cannot share the name of the resident abused or the time when the incident occurred. One state survey agency said that information sharing can be uneven, and told us that law enforcement is required to share information with the state survey agencies, but the state survey agencies do not share their investigatory information with law enforcement. Officials from another state survey agency wrote to CMS notifying CMS of a change in their state survey agency protocol that would make the referral process timelier by providing un-redacted survey records of substantiated abuse to local law enforcement. However, in CMS’s 2017 written response to the survey agency, CMS told them that all written requests for these records must continue to be forwarded to CMS for processing in accordance with the federal Privacy Act. When we asked CMS officials what information state survey agencies can share with law enforcement in a referral, CMS explained that scenarios for requesting information can vary, and that CMS does not prescribe a specific method as it depends on the needs of the investigation. This lack of guidance is inconsistent with federal standards for internal control, which state that management should internally communicate quality information to achieve the entity’s objectives. Conclusions While nursing home abuse is relatively rare, our review shows that abuse deficiencies cited in nursing homes are becoming more frequent, with the largest increase in severe cases. As such, it is imperative that CMS have key information critical to understanding abuse and that the agency’s oversight of nursing homes is strong. We found weaknesses in both CMS’s understanding of abuse and in its oversight that need to be addressed. Specifically, because CMS cannot readily access information on abuse or perpetrator types in its data, it lacks key information critical to taking appropriate actions to address the most prevalent types of abuse and perpetrators. In addition, CMS has not provided guidance on what information should be included in facility-reported incidents, contributing to a lack of information for state survey agencies and, subsequently, delays in their investigations. This lack of guidance related to facility- reported incidents is important in light of our findings that abuse deficiencies are identified most commonly through facility-reported incidents. We also found other gaps in CMS’s process related to ensuring timely referrals of abuse to law enforcement, tracking abuse referrals, defining abuse substantiation, and sharing information with law enforcement. These gaps affect CMS’s oversight of abuse in nursing homes—including the prevention, identification and timely investigation of abuse—and may limit CMS’s ability to ensure that nursing homes meet federal requirements for residents to be free from abuse. Recommendations for Executive Action We are making the following six recommendations to the administrator of CMS: Require that abuse and perpetrator type be submitted by state survey agencies in CMS’s databases for deficiency, complaint, and facility- reported incident data, and that CMS systematically assess trends in these data. (Recommendation 1) Develop and disseminate guidance—including a standardized form—to all state survey agencies on the information nursing homes and covered individuals should include on facility-reported incidents. (Recommendation 2) Require state survey agencies to immediately refer complaints and surveys to law enforcement (and, when applicable, to MFCUs) if they have a reasonable suspicion that a crime against a resident has occurred when the complaint is received. (Recommendation 3) Conduct oversight of state survey agencies to ensure referrals of complaints, surveys, and substantiated incidents with reasonable suspicion of a crime are referred to law enforcement (and, when applicable, to MFCUs) in a timely fashion. (Recommendation 4) Develop guidance for state survey agencies clarifying that allegations verified by evidence should be substantiated and reported to law enforcement and state registries in cases where citing a federal deficiency may not be appropriate. (Recommendation 5) Provide guidance on what information should be contained in the referral of abuse allegations to law enforcement. (Recommendation 6) Agency Comments We provided a draft of this product to HHS for review and comment. In its comments, reproduced in appendix IV, HHS concurred with our six recommendations and identified actions it is taking to implement them. Specifically, HHS said that it will: (1) look into options for requiring state survey agencies to record data on abuse and perpetrator type so that HHS may assess trends in these data; (2) develop guidance that includes a list of standardized data elements to be included when nursing homes report facility-reported incidents and guidance specific to the reporting and tracking of facility-reported incidents involving abuse; (3) require state survey agencies to immediately refer complaints to law enforcement if a reasonable suspicion of a crime against a resident has occurred and share relevant survey information; (4) consider how to implement mechanisms for tracking law enforcement referrals; (5) identify opportunities to clarify in guidance situations where citing a federal deficiency may not be appropriate, but reporting the abuse is still required; and (6) develop a list of standardized elements that should be included when reporting an abuse allegation to law enforcement. HHS also provided technical comments, which we incorporated as appropriate. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the Secretary of HHS and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7114 or at dickenj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix V. Appendix I: Additional Detail on Analysis of Centers for Medicare & Medicaid Services’ (CMS) Data This appendix describes our scope and methodology for determining the trends and types of abuse occurring in nursing homes in recent years. For this examination, we reviewed CMS guidance and analyzed data from 2013 through 2017, which represented the most recent data for a 5-year period at the time of our review. Specifically, we first reviewed the CMS State Operations Manual’s Appendix PP that was in effect during our period of review to determine which federal standards and deficiency codes were relevant to resident abuse. We focused our analysis on the deficiency code to be used by state surveyors when a nursing home fails to keep a resident free from abuse, which encompasses mental/verbal, sexual, or physical abuse. Surveyors can also use other deficiency codes for abuse-related issues, such as a failure by the nursing home to train staff on issues related to abuse, either in conjunction with an abuse deficiency or without an abuse deficiency. Since these abuse-related deficiency codes do not necessarily represent incidents of abuse, but do represent situations where a nursing home’s inadequate policies could leave residents vulnerable to abuse, we conducted a limited analysis on the trends of these deficiencies, which is described in appendix II. For our analysis, we identified abuse deficiencies cited by surveyors in all 50 states and Washington, D.C., between 2013 and 2017, using data provided by CMS from its Certification and Survey Provider Enhanced Reports system. Specifically, we calculated the number of abuse deficiencies cited each year and determined how many of these abuse deficiencies were at each level of severity—no actual harm with a potential for minimal harm, no actual harm with a potential for more than minimal harm, actual harm, and immediate jeopardy—for each year. We compared the results for abuse deficiencies with the results for all types of deficiencies in each year. To avoid over-counting deficiencies, deficiencies that were for the same violation on the same day for the same facility were counted as a single deficiency. We then tracked (1) the origin of these abuse deficiencies and (2) enforcement actions implemented against nursing homes with these abuse deficiencies. Origin of abuse deficiencies. To identify trends in the origin of those abuse deficiencies—that is, whether the deficiency originated from a standard survey, complaint investigation, or a facility-reported incident investigation—we analyzed data provided by CMS from its Automated Survey Processing Environment Complaint/Incident Tracking System. Specifically, we matched the deficiencies with the complaint/incident data using provider number, survey date, and deficiency code. We found that some deficiencies were the result of a combination of complaints, facility-reported incidents, surveys, or all three. We counted those deficiencies as originating from each relevant category. Enforcement actions. To identify trends in the enforcement actions imposed and implemented against nursing homes with abuse deficiencies, we analyzed data provided by CMS from its Automated Survey Processing Environment Enforcement Manager. Specifically, we matched the deficiencies with the enforcement data using provider number, survey date, case identification number, and deficiency code. To avoid over-counting, deficiencies that share the same code and case identification number were counted as a single deficiency. For each year, we determined how many of the abuse deficiencies resulted in enforcement actions imposed or implemented, the severity of the abuse deficiencies with enforcement actions, and the types of enforcement actions implemented. We then examined these abuse deficiencies to determine the number of nursing homes that had abuse deficiencies, as well as the number of homes with repeated abuse deficiencies cited across the 5 years and the characteristics of those homes. We also determined the proportion of surveyed nursing homes in a given year that had an abuse deficiency. Nursing homes that had repeated abuse deficiencies. Since a nursing home can have more than one abuse deficiency cited in a given year, we determined the number of surveyed nursing homes each year that had at least one abuse deficiency, both nationally and by state. For each of those nursing homes, we determined if the home had an abuse deficiency repeated in multiple years and in two or more consecutive years. Nursing home characteristics. We attempted to identify commonalities among homes with multiple years of abuse deficiencies, homes with only a single year with an abuse deficiency, and surveyed homes without any abuse deficiencies throughout the 5- year period. Specifically, we matched deficiency data to CMS’s publicly available Provider of Services files and the Nursing Home Compare Provider Information files for each nursing home; and we examined bed size, non-profit or for-profit status, Five-Star Quality Rating System overall rating, Special Focus Facility designation, and urban or rural location. Finally, because abuse and perpetrator type are not readily identifiable in CMS’s data, we identified this information by reviewing the narratives written by surveyors that describe the substantiated abuse. Specifically, we obtained 1,557 narrative descriptions written by state surveyors for abuse deficiencies cited in 2016 and 2017 provided by CMS from its Automated Survey Processing Environment database. From that universe of abuse deficiency narratives, we selected a randomly selected representative sample of 400 narratives, and each narrative was reviewed by two separate reviewers who independently analyzed the text of each narrative to determine the abuse and perpetrator type according to the definitions that CMS implemented on November 28, 2017, in its State Operations Manual. Any disagreements between the two reviewers were resolved by a third independent reviewer. (See table 7.) For those narratives where the abuse type could not reasonably be categorized under an existing CMS definition, reviewers had the option to mark narratives as “other.” Furthermore, we analyzed the scope and severity for each narrative within our sample. CMS’s abuse deficiency code also included involuntary seclusion in the time period we examined and is defined in its November 22, 2017, guidance as “separation of a resident from other residents or from her/his room or confinement to her/his room (with or without roommates) against the resident’s will, or the will of the resident representative.” Our analysis of the narrative descriptions found that 3 percent of the abuse deficiency narratives in our sample were attributable to involuntary seclusion. We were unable to categorize the abuse and perpetrator type for about 11 percent of the deficiency narratives in our sample, because we determined the narrative description did not meet CMS’s abuse definition. We assessed the reliability of each of the datasets by checking for missing values and obvious errors and discussed them with CMS officials who were knowledgeable about the data. In the course of this assessment, we found some data limitations. Specifically, CMS officials told us that some state survey agencies may not have entered all facility- reported incidents into the Automated Survey Processing Environment Complaint/Incident Tracking System, while other state survey agencies did. We also found underreporting, as noted in our 2019 report, where the Oregon state survey agency was not entering all abuse-related complaints or facility-reported incidents into this same database—a problem that could exist in other states. In addition, CMS officials told us that it is possible there are additional incidents that may not have been represented in the abuse deficiency data during the period of our review. Specifically, CMS officials noted that some incidents resulting from resident altercations—particularly those that do not show a willful intent to harm—may not be cited as an abuse deficiency by some state survey agencies. We therefore consider the number of abuse deficiencies that resulted from complaints or facility-reported incidents to be a conservative estimate. After reviewing the possible limitations of these data, we determined the data were sufficiently reliable for the purposes of this reporting objective. Appendix II: Trends in Abuse-Related Deficiencies This appendix describes trends in abuse-related deficiencies over the 5- year period from 2013 through 2017. We reviewed Centers for Medicare & Medicaid Services (CMS) guidance that was in effect during this period of review to determine which federal standards and deficiency codes were relevant to resident abuse. For the report, we focused our analysis on the deficiency code cited when state surveyors substantiate incidents of abuse, but there are also deficiencies that surveyors can cite for abuse-related issues, such as a failure by the nursing home to train staff on issues related to abuse, either in conjunction with an abuse deficiency or without an abuse deficiency. Since these abuse-related deficiencies do not necessarily represent incidents of abuse, but do represent situations where a nursing home’s inadequate policies could leave residents vulnerable to abuse, we also conducted a limited analysis on the trends of these deficiencies. Specifically, we analyzed CMS data to identify the number of abuse-related deficiencies cited in each year in all 50 states and Washington, D.C., and determined how many were cited at each level of severity—no actual harm with a potential for minimal harm, no actual harm with a potential for more than minimal harm, actual harm, and immediate jeopardy. We also tracked the source of these abuse-related deficiencies—that is, whether the deficiency originated from a standard survey, complaint investigation, or a facility-reported incident investigation. Finally, we compared the results for abuse-related deficiencies with the results for all types of deficiencies cited by surveyors in each year. From 2013 to 2017, we found that abuse-related deficiencies became slightly more common with a resulting increase in severity. Specifically, abuse-related deficiencies increased by about 9.9 percent over the 5-year period, from 4,899 deficiencies cited in 2013 to 5,383 deficiencies cited in 2017, but peaked in 2016 with 5,687 deficiencies. This increasing trend for abuse-related deficiencies is in contrast to the slight decrease in all deficiencies cited over the same period, but not nearly as high as the 103.5 percent increase in abuse deficiencies. In addition, the proportion of abuse-related deficiencies cited at the highest levels of severity— deficiencies causing actual harm to residents or putting residents in immediate jeopardy—fluctuated throughout the 5-year period. Specifically, about 6.1 percent of the 4,899 abuse-related deficiencies in 2013, about 5.6 percent of the 5,278 abuse-related deficiencies in 2015, and about 7.8 percent of the 5,383 abuse-related deficiencies in 2017 caused actual harm or immediate jeopardy. (See fig. 7.) We also found that over half of the abuse-related deficiencies each year were cited by surveyors as a result of standard surveys, and the rest were cited by surveyors as a result of either complaint or facility-reported incident investigations. This falls between what we found for abuse deficiencies—the majority were a result of either complaint or facility- reported incident investigations—and all types of deficiencies—the vast majority were a result of standard surveys. Over the 5 years, similar to abuse deficiencies and all types of deficiencies, the percentage of abuse- related deficiencies that resulted from standard surveys decreased while the percentage that resulted from both complaint and facility-reported incident investigations increased. Specifically, over the 5-year period, the percentage of abuse-related deficiencies resulting from standard surveys decreased by about 8.8 percentage points, complaint investigations increased by about 3.6 percentage points, and facility-reported incident investigations increased by about 5.3 percentage points. (See fig. 8.) Appendix III: State Information on Abuse Deficiencies Tables 8 and 9 provide state-level data on abuse deficiencies and the nursing homes that had abuse deficiencies cited in consecutive years. Appendix IV: Comments from the Department of Health and Human Services Appendix V: GAO Contact and Staff Acknowledgments GAO Contact John E. Dicken, (202) 512-7114 or dickenj@gao.gov. Staff Acknowledgments In addition to the contact named above, Karin Wallestad (Assistant Director); Sarah-Lynn McGrath and Kathryn Richter (Analysts-in-Charge); Luke Baron; Summar Corley; Zosha Kandel; and Julianne Flowers made key contributions to this report. Also contributing were Laurie Pachter, Jennifer Whitworth, and Vikki Porter.

Why GAO Did This Study Nursing homes provide care to about 1.4 million nursing home residents—a vulnerable population of elderly and disabled individuals. CMS, an agency within the Department of Health and Human Services (HHS), defines standards nursing homes must meet to participate in the Medicare and Medicaid programs. GAO was asked to review abuse of residents in nursing homes. Among other objectives, this report: (1) determines the trends and types of abuse in recent years, and (2) evaluates CMS oversight intended to ensure residents are free from abuse. GAO reviewed CMS's policies, analyzed CMS data on abuse deficiencies from 2013 through 2017, the most recent data at the time of our review, and interviewed officials from CMS and state survey agencies in five states, as well as other key stakeholders in those states such as ombudsmen and law enforcement officials. The states were selected for variation in factors such as number of nursing homes and role of other state agencies in abuse investigations. What GAO Found The Centers for Medicare & Medicaid Services (CMS) is responsible for ensuring nursing homes meet federal quality standards, including that residents are free from abuse. CMS enters into agreements with state survey agencies to conduct surveys of the state's homes and to investigate complaints and incidents. GAO analysis of CMS data found that, while relatively rare, abuse deficiencies cited in nursing homes more than doubled, increasing from 430 in 2013 to 875 in 2017, with the largest increase in severe cases. GAO also reviewed a representative sample of abuse deficiency narratives from 2016 through 2017. Physical and mental/verbal abuse occurred most often in nursing homes, followed by sexual abuse, and staff were more often the perpetrators of the abuse deficiencies cited. CMS cannot readily access information on abuse or perpetrator type in its data and, therefore, lacks key information critical to taking appropriate actions. GAO also found gaps in CMS oversight, including: Gaps in CMS processes that can result in delayed and missed referrals. Federal law requires nursing home staff to immediately report to law enforcement and the state survey agency reasonable suspicions of a crime that results in serious bodily injury to a resident. However, there is no equivalent requirement that the state survey agency make a timely referral for complaints it receives directly or through surveys it conducts. CMS also does not conduct oversight to ensure that state survey agencies are correctly referring abuse cases to law enforcement. Insufficient information collected on facility-reported incidents. CMS has not issued guidance on what nursing homes should include when they self-report abuse incidents to the state survey agencies. Officials from all of the state survey agencies in GAO's review said the facility-reported incidents can lack information needed to prioritize investigations and may result in state survey agencies not responding as quickly as needed. What GAO Recommends GAO is making six recommendations, including that CMS: require state survey agencies to submit data on abuse and perpetrator type; require state survey agencies to immediately refer to law enforcement any suspicion of a crime; and develop guidance on what abuse information nursing homes should self-report. HHS concurred with all of GAO's recommendations and identified actions it will take to implement them.

gao_GAO-20-27

gao_GAO-20-27_0

Background In the Telecommunications Act of 1996 (the 1996 Act), Congress specified that consumers in “rural, insular, and high-cost areas” should have access to telecommunication rates and services that are “reasonably comparable” to consumers in urban areas. The 1996 Act altered the federal mechanism for funding universal service by requiring telecommunications carriers and other entities providing interstate telecommunications service to contribute to the USF, unless exempted by FCC. The carriers generally pass these costs on to customers, sometimes in the form of a line item on customers’ telephone bills. USF provides financial support to carriers through four different programs, each targeting a particular group of telecommunications carriers or consumers. The high-cost program provides support to both wireline and wireless carriers that provide telecommunications services in areas that carriers would otherwise not serve and where there is no competition from other providers. These are typically rural or remote areas where the customer base is relatively small and the cost of installing infrastructure is high. The high-cost program has been the largest USF program based on disbursements and has been particularly important to rural areas. High- cost support is intended to offset the carriers’ higher costs, thereby allowing them to provide services and rates that are reasonably comparable to those that consumers in lower-cost—generally urban— areas receive. In 2009, Congress required FCC to develop a broadband plan to ensure that every American has access to broadband capability, including a detailed plan for providing this service at affordable rates. In response, an FCC task force issued the National Broadband Plan in 2010, which recommended reforming USF so it could support both telephone and broadband service. FCC’s USF Transformation Order of 2011 emerged in response to this recommendation and provided USF support to carriers for broadband capable networks. The order required carriers that receive support to meet broadband-speed and quality-deployment requirements. Through the USF Transformation Order, FCC adopted a framework to transition high-cost carriers from traditional cost-accounting support to incentive-based support mechanisms, using forward-looking broadband cost models and competitive bidding. FCC’s forward-looking cost models use historical data to project the future financial needs of carriers providing telecommunications services. According to FCC, rate-of-return carriers receive about $2.5 billion in annual support from the high-cost program to support service deployments in these carriers’ 1,078 rate-of- return service areas, which FCC refers to as “study areas.” FCC has allowed rate-of-return carriers to choose, on a voluntarily basis, one the following mechanisms to receive USF support: Traditional cost-accounting support mechanism. This method retroactively provides support to carriers for costs already incurred, based on cost studies, including financial statements these companies provide each year. At the time of our review, according to FCC officials, FCC guaranteed these companies recovery of eligible deployment costs, plus a return of 10.25 percent on regulated investment costs. According to FCC’s OIG officials we interviewed, many carriers contract with telecommunications accountants to navigate the complicated process of determining which costs are reimbursable by the high-cost program and file the associated documentation with USAC. According to FCC, as of September 2019, there were approximately 437 study areas served by rate-of- return carriers receiving support through this mechanism. Model-based support mechanism. This method is aimed at providing a level of support to carriers based on modeled forward- looking costs and revenues of an efficient carrier to serve an area with voice and broadband Internet. According to FCC officials, in developing the model, FCC: had experts peer-review the model’s methodology; demonstrated how different inputs affect model support and sought stakeholder feedback on the reasonableness of how these inputs affected support levels; publicly released the model’s methodology; and used historical deployment cost and revenue data to develop the model’s inputs and assumptions. As of September 2019, FCC officials told us that rate-of-return companies serving 641 study areas were receiving support through this mechanism (or almost 60 percent of all 1,078 rate-of-return carriers’ study areas). FCC determines overall policy and issues the regulations that govern the high-cost program, while FCC’s Wireline Competition Bureau in particular implements FCC’s policies and programs regarding rate-of-return carriers. State governments play a role in implementing the federal high- cost program, as do a not-for-profit corporation (USAC) and an association (NECA). As shown in table 1, FCC, USAC, and NECA have responsibilities for the high-cost program to ensure payments to rate-of- return carriers are made properly. FCC has the following audit and oversight procedures for the high-cost program: Carrier self-certification. Carriers submit cost and line count data directly to NECA. Carrier self-certification is the primary tool for ensuring that carriers use high-cost program support consistent with program rules. USAC uses these data to qualify carriers for the program and also to calculate the amount of support carriers are eligible to receive. Carrier audits. Audits of carriers receiving high-cost program support are the primary tool used to oversee carrier activities, and audits may be conducted by USAC, state regulators, or FCC’s OIG. USAC primarily relies on assessments from the Payment Quality Assurance Program and Beneficiary and Contributor Audit Program that occur after disbursements have been made to detect improper payments, which may include fraud. Carrier data validation process. All cost data that the carriers submit to NECA for purposes of high-cost support are subject to several electronic validations, which focus on ensuring that all required data are reported and that the data ranges are consistent with information reported in previous years. In addition, NECA compares the reported cost data with financial records supporting carriers’ audited financial statements to identify any discrepancies and to require corrections when discrepancies are discovered. Carriers’ broadband deployment verification. Since 2018, USAC has performed carrier broadband deployment verifications by obtaining broadband location data to monitor whether a carrier’s broadband deployment meets FCC requirements. Carriers receive verification reports from USAC that reflect the results of the verification process. Whistleblower process. USAC maintains a whistleblower log that is shared with FCC. Through whistleblower complaints, USAC may identify instances of potentially fraudulent activity. FCC has identified three rate-of-return carriers that received at least $34 million in improper payments from the high-cost program in prior years. Two such cases were described above. In the third case, a rate-of-return carrier self-reported to NECA and USAC what it represented to be the costs and revenues of providing its telecommunications service; as discussed previously, NECA and USAC rely upon the accuracy and completeness of the carrier’s reporting to calculate the carrier’s support. An FCC OIG investigation later revealed that the carrier had manipulated FCC’s accounting rules by including the costs of a nonregulated, commercial mobile radio service in the information it submitted to NECA, thus inflating the amount of high-cost program support the carrier received. FCC eventually determined that the carrier owed the federal government almost $7 million in support overpayments received between 2005 and 2010. A petition for reconsideration is pending. As there is a finite amount of funding for the high-cost program, compensating carriers for improper, ineligible, and inflated costs they claim means less program funds are available for deploying service to the areas the program was designed to serve. Federal internal control standards, along with GAO’s fraud risk framework, OMB guidance, and the Fraud Reduction and Data Analytics Act of 2015 have placed an increased focus on the need for federal program managers to take a strategic approach to managing improper payments and risks, including the risk of fraud. GAO’s fraud risk framework provides comprehensive guidance for conducting fraud-risk assessments and using the results to develop a robust fraud risk management strategy. This framework also describes overarching concepts and leading practices for establishing an organizational structure and culture that are conducive to fraud risk management, designing and implementing controls to prevent and detect potential fraud, and monitoring and evaluating fraud risk management activities. The leading practices in the fraud risk framework are organized into four components—commit, assess, design and implement, and evaluate and adapt—as depicted in figure 1. FCC Adopted Several Funding Reforms to Enhance Carriers’ Accountability, but Not All Reforms Are Mandatory FCC Reforms Are Intended to Improve the Accountability of Rate-of- Return Funding FCC, in various orders, has adopted several funding and other reforms specific to rate-of-return carriers. As described below, the reforms we reviewed were designed to (1) control the carrier and high-cost program expenditures, (2) incentivize efficient broadband deployment, and (3) ensure carriers’ compliance with the high-cost program. Control Carrier and High-Cost Program Expenditures Prior to the 2011 USF Transformation Order, rate-of-return carriers primarily received high-cost support based on their actual costs. Under the old rules, carriers faced no FCC-imposed limits and, according to FCC, had no incentive to be more efficient. FCC adopted the reforms described in figure 2 to control the program’s expenditures. As shown in figure 2 above, FCC’s reform effort related to eliminating support to areas with competition has been ongoing since 2011. According to FCC officials, FCC relied on its broadband deployment data to identify competitively served areas, but we have previously reported that FCC’s broadband deployment data are not always accurate. In August 2017, FCC initiated a proceeding to review the Form 477—the principal tool FCC uses to gather data on communications services, including broadband services—to help inform its policy making. According to FCC, a goal of this proceeding was to enable FCC to collect better and more accurate information on the Form 477. In August 2019, FCC adopted an order based on the proceeding that, among other things, established requirements for collecting geospatial broadband-coverage maps from internet service providers. According to the order, FCC will require the service providers to submit granular maps of the areas where they have broadband-capable networks; FCC intends that these broadband-deployment maps will enable FCC to precisely target scarce universal service dollars to where broadband service is lacking. Incentivize Efficient Broadband Deployment According to FCC, one of the USF’s core principles since 2011 has been to ensure that support is provided in the most efficient manner possible, recognizing that ultimately American consumers contribute to programs like the high-cost program. FCC adopted the reforms described in figure 3 to advance its long-standing objective of adopting incentive-based policies to spur additional broadband deployment, while preserving additional funding in the high-cost program for other reforms. According to FCC, the prior cases of carriers’ abuses of USF support for unrelated purposes prompted FCC to issue more specific rules for compliance and reporting obligations. Accordingly, FCC adopted reforms described in figure 4 to improve accountability and transparency of the high-cost program. FCC’s Model-Based Support Reform May Reduce Fraud Risks, but It Is Voluntary and Not All Carriers Received Model- Based Support Of the reforms we reviewed, one reform in particular—the development of a model-based support mechanism—shows promise in reducing fraud risk, according to stakeholders from federal and state government, industry, and accounting firms we contacted. Stakeholders said the model-based support mechanism is less prone to fraud risks and is a more efficient support mechanism than traditional cost-accounting support. In particular, unlike the traditional cost-accounting-support mechanism, model-based support does not rely on carrier-submitted data to determine support amounts. Instead, the model uses, among other things, a combination of historical cost data and other data, such as expected customer revenue, to determine support amounts. Since there are no data provided by carriers in the process of determining support amounts, there is no means by which carriers can provide falsified information to fraudulently receive excess support. The carriers involved in the previously described improper payments cases were receiving support from the traditional cost-accounting support mechanism. On the other hand, stakeholders told us FCC’s traditional cost-accounting support mechanism is complex and difficult to audit, and that such weaknesses make it prone to fraud risks. For example, USAC officials told us it is time consuming to detect inflated costs associated with carriers’ affiliate company transactions. The traditional cost-accounting support mechanism also requires that carriers separate costs based upon the type of service with which the cost was associated. According to FCC’s OIG officials and representatives from accounting firms we contacted, determining whether a carrier has overly attributed costs to eligible services is difficult. For instance, determining if labor costs are properly being allocated between eligible and ineligible services requires looking at each employee’s timesheet. According to USAC, it also faces challenges auditing traditional cost-accounting support payments due to limited expertise and capacity to address the complexity of the audits. USAC officials noted that this issue has been exacerbated by audit staff turnover. According to USAC officials and some stakeholders we contacted, auditing carriers receiving traditional cost-accounting support is also difficult due to the extensive documentation requirements for this type of support, requirements that often entails hundreds of pages of financial information per carrier. USAC officials told us that a single audit can take over 1,000 hours to complete, and USAC officials told us they only completed 10 audits of carriers that received support on a traditional cost-accounting basis in fiscal year 2018. As previously noted, FCC allows carriers to choose which funding mechanism is best suited for their company. FCC officials told us they developed the model-based funding mechanism in consultation with carriers and industry stakeholders. However, according to FCC officials, the model’s use is not mandatory because some carriers do not believe that the model would accurately reflect their specific costs. FCC officials said the agency does not have plans to assess the accuracy of the model’s cost estimates or require carriers to receive model-based support. FCC officials told us they did not have plans to assess the model. FCC officials told us they had not planned to do so because in May 2019 FCC had just made available model-based support to the remaining legacy carriers, and FCC was still in the process of evaluating next steps. Planning for and conducting such an assessment would enable FCC to demonstrate the validity of the model and its reliability in accounting for the costs of broadband deployment. Federal internal- control standards state that management should use quality information to make informed decisions and evaluate program performance in achieving key objectives. Furthermore, according to FCC’s strategic plan, FCC must ensure its USF programs, including those for the high- cost program, are well managed, efficient, and fiscally responsible, and the National Broadband Plan says that FCC should move rate-of-return carriers to incentive-based regulation mechanisms, such as model-based support. Yet because a substantial number of rate-of-return study areas— 437—continue to receive traditional cost-accounting support, and the carriers that provide service in these areas cannot be effectively audited, significant fraud risks remain for the high-cost program. By assessing the model, FCC would have greater assurance that it is producing reliable cost estimates and be better positioned to determine whether to make its use mandatory. FCC Has Taken Steps to Manage Fraud Risks, but Its Efforts Do Not Fully Align with Leading Practices Managers of federal programs are responsible for managing fraud risks. Implementing effective fraud risk-management processes is important to help ensure that federal programs fulfill their intended purpose and funds are spent effectively. GAO’s fraud risk framework is aligned with federal internal-control standards related to assessing fraud risk. It focuses on preventive activities, which generally offer the most cost-efficient use of resources since they enable managers to avoid a costly and inefficient “pay-and-chase” model, which refers to the practice of detecting fraudulent transactions and recovering funds after fraudulent payments have been made. As discussed previously, our fraud risk framework consists of four components—commit, assess, design and implement, and evaluate and adapt—each of which includes overarching concepts and leading practices for carrying them out. We found that FCC has implemented some policies and procedures related to managing fraud risk for the high-cost program. For example, according to a memorandum of understanding between FCC and USAC, FCC requires USAC to alert, as appropriate, FCC’s OIG and Enforcement Bureau about potential instances of fraud. However, as detailed in appendix II, FCC’s efforts do not fully align with some elements of the fraud risk framework. In particular, we found deficiencies in FCC’s efforts related to the following three overarching concepts and one high-level component: creating a structure with a dedicated entity to manage fraud risk activities (overarching concept within the commit component); planning regular fraud-risk assessments tailored to the program and assessing these risks to determine the program’s fraud risk profile (two overarching concepts within the assess component); and designing and implementing an antifraud strategy for the program (the design and implement component). Creating a structure with a dedicated entity to lead fraud risk- management activities. Leading practices for managing fraud risk include demonstrating management’s commitment to combating fraud and designating an entity to design and oversee fraud risk-management activities. According to GAO’s fraud risk framework, an entity should lead these activities by serving as the repository of knowledge on fraud risks and controls, managing the fraud-risk assessment process, leading fraud- awareness activities, and coordinating antifraud initiatives. According to FCC officials, FCC has steering committees for each of the four USF programs, including the high-cost program. According to FCC officials, the steering committees allow in-depth discussions about each program, including on operational issues such as current spending levels and information technology systems, as well as improper payments and other issues. However, fraud risk is but one of many responsibilities of these steering committees, and they do not fill the role of a dedicated fraud risk- management entity, as called for by the fraud risk framework. In August 2019, FCC officially launched a Fraud Division—comprising existing FCC staff who investigate and prosecute fraud—within its Enforcement Bureau. However, FCC told us the scope of the new division’s operations is limited to investigations, so the Fraud Division does not fill the role of a dedicated fraud risk-management entity. Planning regular fraud-risk assessments tailored to the program and determining the fraud risk profile. An effective antifraud entity tailors the approach for carrying out regular fraud-risk assessments of its programs. According to GAO’s fraud risk framework, the approach should, among other things: fully consider the specific fraud risks the agency or program faces, analyze the potential likelihood and effects of fraud schemes, and document prioritized fraud risks. According to FCC officials, FCC has annually worked with USAC high- cost program staff to identify and assess some risks facing the high-cost program, some of which are fraud risks, but has not planned regular fraud-risk assessments that are tailored to the high-cost program in accordance with GAO’s fraud risk framework. FCC officials also told us that they adopted a tool originally developed by another agency that was used to evaluate risks facing that agency’s loan and grant programs, not just fraud risks. Using that tool as a model, FCC created a risk assessment document that included fraud risk as one of nine categories of risks across the high-cost program. Based on our discussions with FCC officials, however, the document does not constitute a fraud-risk assessment that takes into account changes to the program or operating environment. Furthermore, the risk assessment document does not constitute a fully tailored risk assessment because it does not identify and assess the fraud risks stakeholders we interviewed described as inherent to the high-cost program, detailed below. Risk caused by the complexity of the high-cost program’s cost- allocation rules. Officials from three out of four accounting firms, FCC’s OIG, and a state utility commission we contacted singled out the specific fraud risk caused by what they described as confusing and subjective rules governing the process carriers use to separate eligible and ineligible costs. Risks related to oversight challenges. Stakeholders identified several oversight challenges as significant in that they could contribute to fraud risks for the program, such as: financial mismanagement within carriers that allowed companies to submit potentially fraudulent information to USAC and NECA, and that a telecommunications accountant told us contributed to previous instances of alleged fraud; USAC’s audit personnel challenges that were due to attrition and limited resources and expertise and that were identified by officials from FCC, USAC, FCC’s OIG, and an accounting firm; and deficiencies identified by FCC’s OIG in NECA’s internal controls over payments to carriers, data validation, and the appropriateness of NECA’s role validating carriers’ cost information. In addition, FCC’s OIG officials told us of oversight challenges related to carriers’ reporting, including that it is difficult for USAC to detect when carriers improperly report rates billed for services provided by an affiliate of the company or report incorrect labor rates. Furthermore, we found FCC had not identified and assessed risks to determine the fraud risk profile for the high-cost program, as called for in the fraud risk framework. A fraud risk profile is the summation of key findings and conclusions from a fraud-risk assessment, including the analysis of the types of internal and external fraud risks, their perceived likelihood and effects, managers’ risk tolerance, and the prioritization of risks. FCC officials said they consider the risk of fraud to be low in the high-cost program, and FCC includes fraud risk in its current risk assessment process. Since FCC believes the fraud risk is low for the high-cost program, FCC has not deemed it necessary to conduct a separate fraud-risk assessment of the program. For example, FCC provided us with documentation related to its Enterprise Risk Management activities that identifies risks USAC faces to achieving its corporate objectives. However, while FCC considers fraud risks as part of these activities, the document does not specify the fraud risk tolerance for the program or constitute a fraud risk profile. Without conducting regular fraud-risk assessments to gauge the likelihood and effects of the inherent fraud risks described above, and potentially others, FCC cannot determine or document the program’s fraud risk profile. Furthermore, FCC has no assurance that it has fully considered important fraud risks, determined its tolerance for risks that could be lower priorities, or made sound decisions on how to allocate resources to respond to fraud risks. Not doing so could result in FCC compensating carriers for improper, ineligible, and inflated costs, such as in the previously discussed cases of identified fraud. By regularly assessing fraud risks to determine a fraud risk profile, FCC could better determine the extent to which it has designed and implemented adequate fraud-prevention controls. Designing and implementing an antifraud strategy for the program. Managers who effectively manage fraud risk develop and document an antifraud strategy that describes the program’s activities for preventing, detecting, and responding to the fraud risks identified during the fraud-risk assessment. FCC and USAC have established mechanisms to enhance the oversight of USF programs, mechanisms that can also help mitigate fraud risks for the high-cost program, including: In fiscal year 2016, USAC implemented a risk-based selection method for conducting audits to identify the entities with the greatest risk. USAC forwards potential fraud, waste, and abuse cases to FCC. FCC’s OIG established a hotline that can be used to report potential fraud, and USAC established a Whistleblower Alert mechanism to inform USAC of possible instances of universal service support being misapplied or mismanaged, or when carriers are potentially violating laws, rules, or regulations. USAC shares this information with FCC. FCC and USAC formed a working group tasked with developing a data-analytics tool designed to share USAC high-cost program data with FCC. FCC’s documentation of the tool states that once developed, the tool will help FCC’s Enforcement Bureau in its fraud detection activities across all USF programs. FCC officials described the development of the tool as technically challenging and said there is no established timeline for implementing the tool. FCC officials said that FCC has regular, informal interactions concerning fraud risk with USAC and, to a lesser degree, with NECA and that FCC has confidence that USAC’s improved audit processes are identifying issues appropriately. However, FCC has not specifically designed or implemented an overall strategy to mitigate fraud risks across the high-cost program. An FCC official said FCC believes its existing antifraud activities are adequate. The FCC official said FCC considers the risk of fraud in the high-cost program to be low because USAC audits have revealed that carriers’ financial reporting errors occur at a low rate and therefore do not indicate that a large amount of fraud is occurring. The official said FCC’s fraud risk-management practices are based on federal internal-control standards and are woven into existing FCC mechanisms. Given that FCC has not conducted a fraud-risk assessment that is tailored to the high-cost program and that the deceptive nature of fraud makes it difficult to measure in a reliable way, it is unclear how FCC officials reached the conclusion that the program’s risk of fraud is low. Also, in the absence of an antifraud strategy, FCC has little assurance that it has the specific control activities to prevent and detect high-cost program fraud and that the types of misconduct that previously occurred are not widespread. The improper payment activity discussed previously was caught after USAC provided support to these carriers, and it was not prevented or detected through any strategic fraud risk-management effort on FCC’s part. Furthermore, the FCC’s OIG has expressed significant concerns about such abuses by rate-of-return carriers and about the effectiveness of USAC’s auditing processes in helping prevent improper payments to these carriers. As noted above, USAC’s high-cost program audits can take over 1,000 hours and USAC faces human capital challenges that have diminished its audit capacity. In addition, while NECA’s data validations and reviews of high-cost support provide opportunities to identify input and data-reporting errors, they do not specifically address whether or not the data provided by carriers are reasonable or if the money carriers spent addresses the intended purposes of the high-cost program. Designing and implementing an antifraud strategy that conforms to leading practices would help FCC effectively manage and respond to the fraud risks identified during the fraud-risk assessments. Conclusions Given the continuing importance of deploying telecommunications services in difficult-to-serve areas, effective oversight for rate-of-return carriers is an important component for helping ensure that the high-cost program’s finite funds are used properly to meet the intent of the program. Overall, the traditional cost-accounting mechanism that FCC uses to provide support to a substantial number of rate-of-return carriers is complex, prone to fraud risks, and presents auditing challenges that FCC has not fully addressed. By following leading practices from GAO’s fraud risk framework, FCC could better ensure that it is addressing and strategically targeting the most significant fraud risks facing the high-cost program. Furthermore, FCC’s model-based support mechanism has now been in use by some rate-of-return carriers for several years and stakeholders agree that it is less prone to fraud risks. However, FCC has not assessed the extent to which the model is producing reliable cost estimates. By conducting such an assessment, FCC would have greater assurance that it is producing reliable cost estimates and be better positioned to determine whether to make its use mandatory. Recommendations for Executive Action We are making the following five recommendations to FCC: The Chairman of FCC should ensure that FCC’s Office of Managing Director follows the leading practices in GAO’s fraud risk framework related to a dedicated entity’s management of its antifraud activities, such as serving as the repository of knowledge on fraud risks and coordinating antifraud initiatives. (Recommendation 1) The Chairman of FCC should plan regular fraud-risk assessments tailored to the high-cost program and assess these risks to determine the program’s fraud risk profile, as provided in GAO’s fraud risk framework. (Recommendation 2) The Chairman of FCC should design and implement an antifraud strategy for the high-cost program with specific control activities, based upon the results of fraud-risk assessments and a corresponding fraud risk profile, as provided in GAO’s fraud risk framework. (Recommendation 3) The Chairman of FCC should assess the model-based support mechanism to determine the extent to which it produces reliable cost estimates. (Recommendation 4) The Chairman of FCC should consider whether to make use of the model-based support mechanism mandatory depending on the results of the assessment. (Recommendation 5) Agency Comments We provided a draft of this report to FCC for review and comment. In FCC’s written comments, reprinted in appendix III, FCC described actions it would take to implement our recommendations. FCC also provided technical comments, which we incorporated as appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees, the Chairman of the FCC, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-2834 or vonaha@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Key Elements of the Fraud-Risk Assessment Process for Creating a Fraud Risk Profile To help managers combat fraud and preserve integrity in government agencies and programs, we identified leading practices for managing fraud risks and organized them into a conceptual framework called the Fraud Risk Management Framework (the framework). As described in the background section of this report and depicted visually in figure 1, the framework encompasses control activities to prevent, detect, and respond to fraud, with an emphasis on prevention, to help managers achieve the objective of mitigating fraud risks. The second of four framework components—Assess—calls for specific actions managers should take to achieve the objective of mitigating fraud risks. Specifically, managers should plan regular fraud-risk assessments and assess these risks to determine a fraud risk profile. Figure 5 illustrates the key elements of the fraud-risk assessment process that lead to the creation of a program’s fraud risk profile. Appendix II: Assessment of the Federal Communications Commission’s Antifraud Efforts for the Universal Service Fund High- Cost Program We developed a data collection instrument to structure our assessment of the Federal Communications Commission’s (FCC) antifraud efforts for the high-cost program related to the commit, assess, and design and implement components of GAO’s fraud risk framework. For our assessment, we determined the extent to which FCC had implemented the leading practices within each component, as summarized in table 2. We did not assess FCC’s approach against leading practices in the “evaluate and adapt” component of the framework because we determined that FCC had not adopted fraud risk management leading practices within the first three components, and therefore it was premature for us to assess whether FCC was evaluating and adapting its use of these practices. Appendix III: Comments from the Federal Communications Commission Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact: Staff Acknowledgments: In addition to the individual named above, Sally Moino (Assistant Director); Sean Standley (Analyst in Charge); Philip Farah; Camilo Flores; Mark Goldstein; Gary Guggolz; Hannah Hubbard; Josh Ormond; Ben Licht; Rebecca Shea; Andrew Stavisky; and Michelle Weathers made key contributions to this report.

Why GAO Did This Study The Universal Service Fund's high-cost program provides financial support to telecommunications carriers in areas where the cost to provide broadband is high. Through this program, FCC provides about $2.5 billion in annual support payments to rate-of-return carriers. The manner in which FCC currently provides the support payments to some of these carriers is prone to fraud risks. A prior case involved a rate-of-return carrier that received at least $27 million in improper payments from the program. GAO was asked to review funding reforms and fraud controls FCC has implemented for rate-of-return carriers. This report examines the extent to which FCC: (1) has implemented funding reforms specific to rate-of-return carriers, and (2) is managing fraud risks for the high-cost program in accordance with leading practices. GAO reviewed FCC's and USAC's procedures, relevant regulations, and guidance, and assessed these documents against applicable criteria, including federal internal-control standards, FCC's strategic plan, and GAO's fraud risk framework. GAO interviewed FCC and USAC officials, in addition to industry and other stakeholders representing a variety of non-generalizable viewpoints. What GAO Found The Federal Communications Commission (FCC) has implemented several funding reforms for small, rural telecommunications carriers—referred to as “rate-of-return carriers”—receiving high-cost program support. These reforms are aimed at controlling the program's expenditures and incentivizing efficient broadband deployment. According to FCC's strategic plan, FCC must ensure the high-cost program is well managed, efficient, and fiscally responsible. One of the reforms that GAO reviewed established a funding mechanism for the carriers whereby FCC determines the level of financial support to provide the carriers based on cost and revenue estimates produced by a model. Stakeholders told GAO that this model-based funding mechanism is less prone to fraud risks than the traditional cost-accounting funding mechanism, which reimburses carriers for their reported costs. However, FCC did not make use of this reform mandatory and a substantial number of rate-of-return carriers continue to receive support from the traditional funding mechanism. FCC officials said they developed the model-based funding mechanism in consultation with industry stakeholders. However, FCC officials said they did not have plans to assess the accuracy of cost estimates from the model, which has been in use for several years, or require carriers to receive model-based support as a way to reduce fraud risks. By assessing the model, FCC would have greater assurance that it is producing reliable cost estimates and be better positioned to determine whether to make its use mandatory. FCC has some policies and processes in place to manage fraud risks for the high-cost program. For example, the Universal Service Administrative Company (USAC)—the not-for-profit corporation that administers the program—reviews and audits rate-of-return support payments and forwards potential fraud cases to FCC's Office of Inspector General and Enforcement Bureau for further investigation. FCC is also developing a data-analytics tool to help detect fraud, and in August 2019 launched a new Fraud Division to focus on investigating fraud in the Universal Service Fund's programs. However, FCC's efforts do not fully align with some elements of GAO's fraud risk framework, including: designing and implementing an antifraud strategy for the program. Without regular fraud-risk assessments of the high-cost program, FCC has no assurance that it has fully considered important fraud risks, determined its tolerance for risks that could be lower priorities, or made sound decisions on how to allocate resources to respond to fraud risks. Not doing so could result in FCC compensating carriers for improper, ineligible, or inflated costs. Furthermore, in the absence of an antifraud strategy, FCC has little assurance that it can prevent or detect the types of documented rate-of-return carrier misconduct that have previously occurred. Designing and implementing an antifraud strategy that conforms to leading practices would help FCC effectively manage and respond to the fraud risks identified during the fraud-risk assessments. What GAO Recommends GAO is making five recommendations, including that FCC should assess the model-based support mechanism and consider making its use mandatory, and implement an antifraud strategy for the high-cost program. FCC stated it would take steps to implement these recommendations.

gao_GAO-19-482T

gao_GAO-19-482T_0

Status of Major Space Systems DOD space systems support and provide a wide range of capabilities to a large number of users, including the military services, the intelligence community, civil agencies, and others. These capabilities include positioning, navigation, and timing; meteorology; missile warning; and secure communications, among others. Space systems can take a long time to develop and involve multiple segments, including space, ground control stations, terminals, user equipment, and launch, as figure 1 below shows. DOD satellite systems are also expensive to acquire. Unit costs for current DOD satellites can range from $500 million to over $3 billion. The associated ground systems can cost over $6 billion to develop and maintain and the cost to launch a satellite can climb to well over $100 million. Table 1 provides highlights of the current status of DOD’s major space programs. As the table shows, DOD is also in the beginning phases of acquiring several constellations of new satellites and ground processing capabilities—including for missile warning, protected communications, space-based environmental monitoring, and space command and control. We have work underway to assess the Air Force’s space command and control development efforts and examine DOD’s analysis of alternatives for wideband communication services. For a more complete description of these major space programs, see appendix I. In addition, DOD is exploring alternatives for acquiring wideband satellite communications as well as funding development of new launch vehicles as it pursues a new acquisition strategy for procuring launch services. Our prior work has shown that many major DOD space programs have experienced significant cost increases and schedule delays. For instance, the total program cost for the Advanced Extremely High Frequency (AEHF) satellite program, a protected satellite communications system, has grown 117 percent since the program’s original cost estimate and its first satellite was launched more than 3.5 years late. For the Space Based Infrared System (SBIRS), a missile warning satellite program, the program cost grew 265 percent from its original estimate and the launch of the first satellite was delayed roughly 9 years. Both programs moved to the production phase where fewer problems tend to surface, and where there is typically less risk of significant cost and schedule growth. A more recent major satellite program, Global Positioning System (GPS) III, has seen an almost 4-year delay due to technical issues and program cost growth of about 32 percent. Cost and schedule growth has also been a challenge for satellite ground systems and user equipment. Ground system delays have been so lengthy, that satellites sometimes spend years in orbit before key capabilities can be fully exploited. For example, The command and control system for GPS III satellites, known as the Next Generation Operational Control System, or OCX, is approximately 5 years behind schedule. As a result, the Air Force has had to start two separate back-up efforts to modify the current ground system to ensure the continuity of GPS capabilities and to make anti- jamming capabilities available via Military Code, or M-code, until OCX is delivered. Our ongoing review of GPS includes an assessment of OCX schedule risk and potential impacts on OCX delivery, acceptance, and operation. We expect to issue our report on GPS in spring 2019. Development of GPS user equipment that can utilize the M-Code signal has lagged behind the fielding of GPS M-code satellites for more than a decade, due to prolonged development challenges. In December 2017, we found that while DOD had made some progress on initial testing of the receiver cards needed to utilize the M-code signal, additional development was necessary to make M-code work with the over 700 weapon systems that require it. We also found that DOD had begun initial planning to transition some weapon systems to use M-code receivers, but significantly more work remained to understand the cost and schedule of transitioning to M-code receivers across DOD. Further, in December 2017, we found that multiple entities were separately maturing their own receiver cards. We recommended that DOD assign responsibility to a single organization to collect test data, lessons learned, and design solutions so that common design solutions are employed and DOD could avoid duplication of efforts. DOD concurred with the recommendation, but has not yet taken action on it. We have previously reported that over 90 percent of the capabilities to be provided by Mobile User Objective System communications satellites—currently, five satellites are in orbit, the first of which launched in 2012—are being underutilized because of difficulties with integrating the space, ground, and terminal segments and delays in fielding compatible user terminals. Largely because of technical and management challenges, the Joint Space Operations Center Mission System (JMS) Increment 2 program—intended to replace and improve upon an aging space situational awareness and command and control system—was almost 3 years behind schedule and 42 percent over budget before the Air Force stopped development work last year. Last month, we reported that operational testing in 2018 found that JMS Increment 2 was not operationally effective or suitable due, in part, to missing software requirements, urgent deficiencies that affected system performance, and negative user feedback. Cost and schedule growth in DOD’s space programs is sometimes driven by the inherent risks associated with developing complex space technology; however, over the past 10 years we have identified a number of other management and oversight problems that have worsened the situation. These include making overly optimistic cost and schedule estimates, pushing programs forward without sufficient knowledge about technology and design, and experiencing problems in overseeing and managing contractors, among others. We have also noted that some of DOD’s programs with operational satellites, such as SBIRS, were also exceedingly ambitious, which in turn increased technology, design, and engineering risks. While SBIRS and other satellite programs provide users with important and useful capabilities, their cost growth has significantly limited the department’s buying power at a time when more resources may be needed to protect space systems and recapitalize the space portfolio. Challenges Facing Acquisitions of New Space Systems DOD faces significant challenges as it replenishes its satellite constellations. First, DOD is confronted with growing threats in space, which may require very different satellite architectures and acquisition strategies. Second, DOD is in the midst of planning major changes to its leadership for space. While these changes are designed to streamline decision-making and bring together a dispersed space workforce, they could cause some disruption to space system acquisition programs. Third, in fiscal year 2016, Congress required DOD to establish guidance to speed up acquisition timeframes by streamlining acquisition processes and oversight for certain acquisitions. GAO is examining DOD’s application of streamlining to its weapons programs. For space, challenges with past streamlining efforts may offer some lessons learned. And fourth, DOD may face resource and capacity challenges in taking on multiple space acquisitions at one time. For example, our work and other reports point to potential gaps in the space acquisition workforce and ongoing difficulties managing software development. Growing Threats to Satellites Require New Approaches According to Air Force Space Command and others, U.S. space systems face intentional and unintentional threats that have increased rapidly over the past 20 years. These include radio frequency interference (including jamming), laser attacks, kinetic intercept vehicles, and ground system attacks. Additionally, the hazards of the already-harsh space environment (e.g., extreme temperature fluctuations and radiation) have increased, including numbers of active and inactive satellites, spent rocket bodies, and other fragments and debris. According to a February 2019 Defense Intelligence Agency report, China and Russia in particular are developing a variety of means to exploit perceived U.S. reliance on space-based systems and challenge the U.S. position in space. The report also states that Iran and North Korea have demonstrated some counterspace capabilities that could pose a threat to militaries using space-based services. In response, recent governmentwide and DOD strategic and policy guidance have stressed the need for U.S. space systems to be survivable or resilient against such threats and DOD has taken steps to be more resilient in some of its new programs. As we found in October 2014, one way to do this is to build more disaggregated systems, including dispersing sensors onto separate satellites; using multiple domains, including space, air, and ground to provide full mission capabilities; hosting payloads on other government or commercial spacecraft; or some combination of these. With capabilities distributed across multiple platforms, rather than centralized onto just a few satellites, it may be more difficult for an adversary to target all assets to attack full system capabilities, and if an attack does take place, the loss of one smaller satellite or payload could result in less capability loss than damage to, or loss of, a large multifunctional satellite. In addition to disaggregation, DOD could make satellites more maneuverable and build in defense capabilities to protect themselves as a means to increase survivability. We also found in October 2014 that some of these options could have beneficial impacts on acquisition. For example, acquiring smaller, less complex satellites may require less time and effort to develop and produce. This may be in part due to improved requirements discipline, as more frequent production rates may allow program managers to delay new requirements to the next production cycle instead of incorporating them into ongoing timelines midstream. Building more, less-complex satellites might also provide DOD the opportunity to use commercial products and systems that have already been tested in the market. At the same time, however, addressing the need to make satellites more resilient could introduce complications. For example, DOD may need to acquire higher quantities of satellites, which may make it more difficult to manage acquisition schedules. In addition, potentially more development and production contracts may result in more complexity for program offices to manage, requiring increased oversight of contractors. Adding more satellites and new technologies may also complicate efforts to synchronize satellite, terminal, and ground system schedules, limiting delivery of capabilities to end users. Our work has also found potential barriers to making satellites more resilient. For example, in October 2014, we found that disaggregation could require DOD to make significant cultural and process changes in how it acquires space systems—for instance, by relying on new contractors, relinquishing control to providers who host government payloads on commercial satellites, using different contracting methods, and executing smaller but more numerous and faster-paced acquisition programs. It will likely require DOD to be more flexible and agile when it comes to satellite acquisitions, especially with regard to coordinating satellite delivery with interdependent systems, such as user equipment. Yet, as we have previously found, DOD’s culture has generally been resistant to changes in space acquisition approaches, and fragmented responsibilities have made it very difficult to coordinate and deliver interdependent systems. Senior leaders have recognized the need to change the space acquisition culture, and as discussed below, changes are being made to space leadership and acquisition approaches. More recently, in July 2018, we found that two factors have contributed to DOD’s limited use of commercially hosted payloads. First, DOD officials identified logistical challenges to matching government payloads with any given commercial host satellite. For example, most of the offices we spoke with cited size, weight, and power constraints, among others, as barriers to using hosted payloads. Second, while individual DOD offices have realized cost and schedule benefits from using hosted payloads, DOD as a whole has limited information on costs and benefits of hosted payloads. Further, the knowledge DOD obtained is fragmented across the agency—with multiple offices collecting piecemeal information on the use of hosted payloads. The limited knowledge and data on hosted payloads that is fragmented across the agency has contributed to resistance among space acquisition officials to adopting this approach. We recommended, and DOD concurred, that the department bolster and centralize collection and analysis of cost, technical, and lessons learned data on its use of hosted payloads. Lastly, in October 2018, we found that DOD faced mounting challenges in protecting its weapon systems—satellites and their ground systems included—from increasingly sophisticated cyber threats. We reported that this was due to the computerized nature of weapon systems, DOD’s late start in prioritizing weapon system cybersecurity, and DOD’s nascent understanding of how to develop more secure weapon systems. In operational testing, DOD routinely found mission-critical cyber vulnerabilities in systems that were under development, yet program officials GAO met with believed their systems were secure and even discounted some test results as unrealistic. Using relatively simple tools and techniques, testers were able to take control of systems and operate largely undetected, due in part to basic issues such as poor password management and unencrypted communications. DOD has recently taken several steps to improve weapon system cybersecurity, including issuing and revising policies and guidance to better incorporate cybersecurity considerations. Further, in response to congressional direction, DOD has also begun initiatives to better understand and address cyber vulnerabilities. Space Leadership Changes Are a Positive Step, But Have Some Risk We and others have reported for over two decades that fragmentation and overlap in DOD space acquisition management and oversight have contributed to program delays and cancellations, cost increases, and inefficient operations. For example, in February 2012 we found that fragmented leadership contributed to a 10-year gap between the delivery of GPS satellites and associated user equipment. The cancellations of several large programs over the past 2 decades were in part because of disagreements and conflicts among stakeholders. In July 2016, in response to a provision of a Senate Report accompanying a bill for the National Defense Authorization Act for Fiscal Year 2016, we issued a report that reviewed space leadership in more depth and concluded that DOD space leadership was fragmented. We identified approximately 60 stakeholder organizations across DOD, the Executive Office of the President, the Intelligence Community, and civilian agencies. Of these, eight organizations had space acquisition management responsibilities; eleven had oversight responsibilities; and six were involved in setting requirements for defense space programs. At the same time, many experts stated that no one seemed to be in charge of space acquisitions. Our report highlighted the pros and cons of various options to reorganize space functions recommended in prior congressionally-chartered studies. The issue has taken on more importance in recent years, as DOD has realized satellites are highly vulnerable to attacks and needs to make dramatic changes in space system architectures and operations. We have found that leadership has not been focused enough to overcome interagency rivalries and resistance to change, and it has not been able to get concurrence on future architectures. The President’s Administration and DOD have taken significant steps to change space leadership. Most recent is the President’s Space Policy Directive-4, issued on February 19, 2019, and DOD’s subsequent legislative proposal submitted on March 1, 2019, to establish a United States Space Force as a sixth branch of the United States Armed Forces within the Department of the Air Force. The Policy Directive states that this is an important step toward a future military department for space and that the Space Force will (1) consolidate existing forces and authorities for military space activities, as appropriate, to minimize duplication of effort and eliminate bureaucratic inefficiencies; and (2) not include the National Aeronautics and Space Administration, the National Oceanic and Atmospheric Administration, the National Reconnaissance Office, or other non-military space organizations or missions of the United States Government. According to the Policy Directive, the Space Force would include the uniformed and civilian personnel conducting and directly supporting space operations from all DOD Armed Forces, assume responsibilities for all major military space acquisition programs, and create the appropriate career tracks for military and civilian space personnel across all relevant specialties. Pertaining to organization and leadership, the Policy Directive states that there should be a civilian Under Secretary of the Air Force for Space, to be known as the Under Secretary for Space, appointed by the President, and establishes a Chief of Staff of the Space Force, who would serve as a member of the Joint Chiefs of Staff. Furthermore, the Policy Directive states that as the Space Force matures, and as national security requires, it will become necessary to create a separate military department, to be known as the Department of the Space Force. This department would take over some or all responsibilities for the Space Force from the Department of the Air Force. The Policy Directive requires the Secretary of Defense to conduct periodic reviews to determine when to recommend that the President seek legislation to establish such a department. Our past work has identified fragmentation in space leadership, but because implementation has not yet occurred, it remains to be seen whether this policy directive and proposed legislation would resolve these issues. In implementing these changes there are many complexities to consider. For example, because space capabilities are acquired and used across the military services and defense agencies, it will be important to address many details on how to implement a Space Force among these equities. Our past work suggests that without close attention to the consequences of the compromises that will inevitably have to be made to carve out a new force structure from existing space functions, there is risk of exacerbating the fragmentation and ineffective management and oversight the Space Force is intended to address. For instance, in March 2019, DOD established the Space Development Agency to unify and integrate efforts across DOD to define, develop, and field innovative solutions. But it is unclear how this new organization will mesh with the Air Force Space and Missile Systems Center, which acquires satellites, the Defense Advanced Research Projects Agency, which creates breakthrough technologies and capabilities, and similar organizations. Moreover, even if changes are implemented effectively, they are only a first step toward addressing space acquisition problems. As we discuss below, programs will still need to embrace acquisition best practices, such as using demonstrable knowledge to make decisions. Our prior work has found that they will also need to be open to flexible and innovative approaches, and work effectively with a very wide range of stakeholders, including those that will not be part of the Space Force, such as the intelligence agencies, civilian space agencies, the current military services, as well as entities within the Office of the Secretary of Defense who help oversee and manage acquisitions. Senior leaders have acknowledged that additional changes are needed and have taken steps to help bring them about, such as the restructuring of the Air Force’s Space and Missile Systems Center, which is designed to break down stovepipes and streamline acquisition processes. Past Streamlining Efforts Offer Lessons Learned DOD is managing a number of new space acquisition programs using a new authority, established under Section 804 of the National Defense Authorization Act for Fiscal Year 2016, which is to provide a streamlined alternative to the traditional DOD acquisition process. Specifically, the programs—which include follow-on missile warning and protected communications satellites, among others—will be exempted from the acquisition and requirements processes defined by DOD Directive 5000.01 and the Joint Capabilities Integration and Development System. Instead, program managers are encouraged to use a tailored approach to documentation and oversight to enable them to demonstrate new technologies or field new or updated systems within 2 to 5 years. We have ongoing work looking across the military departments at how middle-tier acquisition authority is being implemented, including for the Air Force’s space acquisition programs, and plan to issue a report later this spring. GAO and others have highlighted lessons learned from past efforts to streamline, specifically with an approach adopted for space systems in the 1990s known as Total System Performance Responsibility (TSPR). TSPR was intended to facilitate acquisition reform and enable DOD to streamline its acquisition process and leverage innovation and management expertise from the private sector. Specifically, TSPR gave a contractor total responsibility for the integration of an entire weapon system and for meeting DOD’s requirements. We found in May 2009 that because this reform made the contractor responsible for day-to-day program management, DOD did not require formal deliverable documents—such as earned value management reports—to assess the status and performance of the contractor. As a result, DOD’s capability to lead and manage the space acquisition process diminished, which magnified problems related to unstable requirements and poor contractor performance. Further, the reduction in DOD oversight and involvement led to major reductions in various government capabilities, including cost- estimating and systems-engineering staff. This, in turn, led to a lack of technical data needed to develop sound cost estimates. Best practices that we identified in the aftermath of TSPR include retaining strong oversight and insight into programs; using quantifiable data and demonstrable knowledge to make decisions to proceed, not allowing development to proceed until certain thresholds are met, empowering program managers to make decisions on the direction of the program but also holding them accountable for their choices, and canceling unsuccessful programs. Similarly, in its study of TSPR programs, the Defense Science Board/Air Force Scientific Advisory Board Joint Task Force emphasized the importance of managing requirements, sufficiently funding programs, participating in trade-off studies, and assuring that proven engineering practices characterize program implementation, among other actions. See appendix II for a more complete list of the best practices we have identified for developing complex systems. DOD May Face Resource and Capacity Challenges in Taking on Multiple Programs at One Time DOD is simultaneously undertaking new major acquisition efforts to replenish its missile warning, protected communications, GPS, and weather satellites. At the same time, it is boosting efforts to increase space situational awareness and protect space assets. It is also helping to fund the development of new launch vehicles, and it is considering additional significant acquisitions in wideband satellite communications and in support of missile defense activities. While there is increased attention within DOD on funding for space and building the Space Force, such widespread acquisition activities could still pose resource challenges. For example: Funding requests for space system modernization have in the past 10 years represented a small percentage (3.9 to 5 percent) of total weapon system modernization funding DOD requested. Space is competing with ships, aircraft, and the nuclear triad, among other programs for funding. This can be challenging, because over the past 2 years, DOD has begun over 9 new space acquisition programs to recapitalize current space capabilities and enhance system resiliency. In the past, we have found that it has been difficult for DOD to fund multiple new space programs at one time, particularly when it was concurrently struggling with cost overruns and schedule delays from its legacy programs. For example, OCX system development challenges have resulted in a $2.5 billion cost increase and approximate 5-year delay to the system becoming operational— using more resources for a longer time—at a cost to other programs. It is unclear whether DOD has a sufficient workforce to manage multiple new space programs. We issued a report last month that found DOD did not routinely monitor the size, mix, and location of its space acquisition workforce. We collected and aggregated data from multiple DOD space acquisition organizations and found that at least 8,000 personnel in multiple locations nationwide were working on space acquisition activities at the end of 2017. Echoing concerns raised in our prior work, we also found that DOD had difficulty attracting and retaining candidates with the requisite technical expertise. Officials from the Air Force’s Space and Missile Systems Center were concerned that there are not enough experienced mid- level acquisition personnel and also expressed concern that the bulk of military personnel assigned to program management positions were more junior in rank than the Center was authorized to obtain. We recommended that DOD (1) identify the universe of its space acquisition programs and the organizations that support them, and (2) collect and maintain data on the workforce supporting these programs. DOD concurred with our first recommendation but not the second. Software is an increasingly important enabler of DOD space systems. However, DOD has struggled to deliver software-intensive space programs that meet operational requirements within expected time frames. Although user involvement is critical to the success of any software development effort, we found in our report issued last month on DOD software-intensive space programs that the programs we reviewed that experienced cost or schedule breaches often did not effectively engage users to understand requirements and obtain feedback. Program efforts to involve users and incorporate feedback frequently did not match plans. The lack of user engagement has contributed to systems that were later found to be operationally unsuitable. The programs we reviewed also faced challenges in delivering software in shorter time frames, and in using commercial software, applying outdated tools and metrics, as well as having limited knowledge and training in newer software development techniques. DOD acknowledged these challenges and is taking steps to address them, including identifying useful software development metrics and ways to include them in new contracts. We recommended, and DOD concurred, that the department ensure its guidance addressing software development provides specific, required direction on the timing, frequency, and documentation of user involvement and feedback. Moreover, it should be noted that software development has been a struggle for other non-space weapons programs as well. The Defense Innovation Board recently reported that the department’s current approach to software development is broken and is a leading source of risk to DOD—it takes too long, is too expensive, and exposes warfighters to unacceptable risk by delaying their access to the tools they need to assure mission success. Chairman Cooper, Ranking Member Turner, and Members of the Subcommittee, this concludes my statement. I am happy to answer any questions that you have. GAO Contact and Staff Acknowledgments If you or your staff have any questions about this statement, please contact me at (202) 512-4841 or chaplainc@gao.gov. Contacts for our Offices of Congressional Relations and Public Affairs may be found on the last page of this statement. Individuals who made key contributions to this statement include Rich Horiuchi, Assistant Director; Erin Cohen (Analyst in Charge); Emily Bond; Claire Buck; Maricela Cherveny; Susan Ditto; Burns C. Eckert; Laura Hook; and Anne Louise Taylor. Key contributors for the previous work on which this statement is based are listed in the products cited. Appendix I: Status of Major Department of Defense Space Acquisitions Appendix II: Best Practices GAO Has Identified for Space and Weapons Systems Acquisitions Our previous work on weapons acquisitions in general, and space programs in particular, identified best practices for developing complex systems. We summarize these best practices in table 3, below. Related GAO Products DOD Space Acquisitions: Including Users Early and Often in Software Development Could Benefit Programs. GAO-19-136. Washington, D.C.: March 18, 2019. Defense Space Systems: DOD Should Collect and Maintain Data on Its Space Acquisition Workforce. GAO-19-240. Washington, D.C.: March 14, 2019. Weapon Systems Cybersecurity: DOD Just Beginning to Grapple with Scale of Vulnerabilitie. GAO-19-128. Washington, D.C.: October 9, 2018. Military Space Systems: DOD’s Use of Commercial Satellites to Host Defense Payloads Would Benefit from Centralizing Data. GAO-18-493. Washington, D.C.: July 30, 2018. Weapon Systems Annual Assessment: Knowledge Gaps Pose Risks to Sustaining Recent Positive Trends. GAO-18-360SP. Washington, D.C.: April 25, 2018. Global Positioning System: Better Planning and Coordination Needed to Improve Prospects for Fielding Modernized Capability. GAO-18-74. Washington, D.C.: December 12, 2017. Space Launch: Coordination Mechanisms Facilitate Interagency Information Sharing on Acquisitions GAO-17-646R. Washington D.C.: August 9, 2017 Satellite Acquisitions: Agencies May Recover a Limited Portion of Contract Value When Satellites Fail. GAO-17-490. Washington, D.C.: June 9, 2017 Space Acquisitions: DOD Continues to Face Challenges of Delayed Delivery of Critical Space Capabilities and Fragmented Leadership. GAO-17-619T. Washington, D.C.: May 17, 2017. Defense Acquisitions: Assessments of Selected Weapon Programs. GAO-17-333SP. Washington, D.C.: March 30, 2017. Global Positioning System: Observations on Quarterly Reports from the Air Force. GAO-17-162R. Washington, D.C.: October 17, 2016. Defense Space Acquisitions: Too Early to Determine if Recent Changes Will Resolve Persistent Fragmentation in Management and Oversight. GAO-16-592R. Washington, D.C.: July 27, 2016. Evolved Expendable Launch Vehicle: DOD Is Assessing Data on Worldwide Launch Market to Inform New Acquisition Strategy. GAO-16-661R. Washington, D.C.: July 22, 2016 Defense Weather Satellites: DOD Faces Acquisition Challenges for Addressing Capability Needs. GAO-16-769T, Washington, D.C.: July 7, 2016. Defense Weather Satellites: Analysis of Alternatives is Useful for Certain Capabilities, but Ineffective Coordination Limited Assessment of Two Critical Capabilities. GAO-16-252R. Washington, D.C.: March 10, 2016. Space Acquisitions: Challenges Facing DOD as it Changes Approaches to Space Acquisitions. GAO-16-471T. Washington, D.C.: March 9, 2016. Space Acquisitions: GAO Assessment of DOD Responsive Launch Report. GAO-16-156R. Washington, D.C.: October 29, 2015. Space Situational Awareness: Status of Efforts and Planned Budgets. GAO-16-6R. Washington, D.C.: October 8, 2015. GPS: Actions Needed to Address Ground System Development Problems and User Equipment Production Readiness. GAO-15-657. Washington, D.C.: September 9, 2015. Evolved Expendable Launch Vehicle: The Air Force Needs to Adopt an Incremental Approach to Future Acquisition Planning to Enable Incorporation of Lessons Learned. GAO-15-623. Washington, D.C.: August 11, 2015. Defense Satellite Communications: DOD Needs Additional Information to Improve Procurements. GAO-15-459. Washington, D.C.: July 17, 2015. Space Acquisitions: Some Programs Have Overcome Past Problems, but Challenges and Uncertainty Remain for the Future. GAO-15-492T. Washington, D.C.: April 29, 2015. Space Acquisitions: Space Based Infrared System Could Benefit from Technology Insertion Planning. GAO-15-366. Washington, D.C.: April 2, 2015. Defense Acquisitions: Assessments of Selected Weapon Programs. GAO-15-342SP. Washington, D.C.: March 12, 2015. Defense Major Automated Information Systems: Cost and Schedule Commitments Need to Be Established Earlier. GAO-15-282. Washington, D.C.: February 26, 2015. DOD Space Systems: Additional Knowledge Would Better Support Decisions about Disaggregating Large Satellites. GAO-15-7. Washington, D.C.: October 30, 2014. U.S. Launch Enterprise: Acquisition Best Practices Can Benefit Future Efforts. GAO-14-776T. Washington, D.C.: July 16, 2014. 2014 Annual Report: Additional Opportunities to Reduce Fragmentation, Overlap, and Duplication and Achieve Other Financial Benefits. GAO-14-343SP. Washington, D.C.: April 8, 2014. Defense Acquisitions: Assessments of Selected Weapon Programs. GAO-14-340SP. Washington, D.C.: March 31, 2014. Space Acquisitions: Acquisition Management Continues to Improve but Challenges Persist for Current and Future Programs. GAO-14-382T. Washington, D.C.: March 12, 2014. Evolved Expendable Launch Vehicle: Introducing Competition into National Security Space Launch Acquisitions. GAO-14-259T. Washington, D.C.: March 5, 2014. The Air Force’s Evolved Expendable Launch Vehicle Competitive Procurement. GAO-14-377R. Washington, D.C.: March 4, 2014. Space Acquisitions: Assessment of Overhead Persistent Infrared Technology Report. GAO-14-287R. Washington, D.C.: January 13, 2014. Space: Defense and Civilian Agencies Request Significant Funding for Launch-Related Activities. GAO-13-802R. Washington, D.C.: September 9, 2013. Global Positioning System: A Comprehensive Assessment of Potential Options and Related Costs is Needed. GAO-13-729, Washington, D.C.: September 9, 2013. Space Acquisitions: DOD Is Overcoming Long-Standing Problems, but Faces Challenges to Ensuring Its Investments are Optimized. GAO-13-508T. Washington, D.C.: April 24, 2013. Satellite Control: Long-Term Planning and Adoption of Commercial Practices Could Improve DOD’s Operations. GAO-13-315. Washington, D.C.: April 18, 2013. Defense Acquisitions: Assessments of Selected Weapon Programs. GAO-13-294SP. Washington, D.C.: March 28, 2013. Launch Services New Entrant Certification Guide. GAO-13-317R. Washington, D.C.: February 7, 2013. Evolved Expendable Launch Vehicle: DOD Is Addressing Knowledge Gaps in Its New Acquisition Strategy. GAO-12-822. Washington, D.C.: July 26, 2012. Space Acquisitions: DOD Faces Challenges in Fully Realizing Benefits of Satellite Acquisition Improvements. GAO-12-563T. Washington, D.C.: March 21, 2012. Space and Missile Defense Acquisitions: Periodic Assessment Needed to Correct Parts Quality Problems in Major Programs. GAO-11-404. Washington, D.C.: June 24, 2011. Space Acquisitions: Development and Oversight Challenges in Delivering Improved Space Situational Awareness Capabilities. GAO-11-545. Washington, D.C.: May 27, 2011. Space Acquisitions: DOD Delivering New Generations of Satellites, but Space System Acquisition Challenges Remain. GAO-11-590T. Washington, D.C.: May 11, 2011. Global Positioning System: Challenges in Sustaining and Upgrading Capabilities Persis., GAO-10-636. Washington, D.C.: September 15, 2010. Defense Acquisitions: Challenges in Aligning Space System Components. GAO-10-55. Washington D.C.: October 29, 2009. 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.

Why GAO Did This Study DOD space systems provide critical capabilities that support military and other government operations. They can also be expensive to acquire and field, costing billions of dollars each year. As DOD seeks to replenish its satellite constellations, it faces a number of challenges to ensuring funds are used effectively. Because space-based capabilities are fundamental to U.S. national security and civilian activities, it is essential that DOD manage its space system acquisitions carefully and avoid repeating past problems. This statement provides an update on DOD's space acquisitions, focusing on challenges facing acquisitions of new space systems. This statement is based on GAO reports issued over the past 10 years on DOD space programs. In addition it draws on recent work performed in support of GAO's 2019 annual reports on the progress of major defense acquisition programs as well as duplication, overlap, and fragmentation across the federal government, among other sources. What GAO Found DOD is simultaneously undertaking new major acquisitions to replenish its missile warning, protected communications, navigation, and weather satellites. At the same time, it is boosting efforts to increase space situational awareness and protect space assets. Such widespread acquisition acitivites could face a wide range of resource and management challenges that GAO has reported on, including: Growing threats to satellites. Threats to satellites from both adversaries—such as jamming and cyber attacks—and space debris are increasing. DOD is making changes to how it designs its space systems to increase the resilience and survivability of space capabilities. But it has been challenged in adopting new approaches, such as using commercial satellites to host payloads, and in prioritizing cybersecurity for all of its weapon systems. For hosted payloads, GAO recommended, and DOD concurred, that the department bolster and centralize collection and analysis of cost, technical, and lessons learned data. Implementing leadership changes . DOD is planning major changes to leadership for space. It recently proposed legislation to establish a United States Space Force—initially to be housed within the Department of the Air Force—that would, according to the President's Space Policy Directive, consolidate existing military space activities and minimize duplicative efforts across DOD. GAO found in July 2016 that changes are needed to reduce fragmentation that has negatively affected space programs for many years. But open questions remain about governance as new programs get underway and whether the changes themselves may result in further fragmentation. For example, it is unclear at this time how the new Space Development Agency will mesh with organizations currently involved in testing and acquiring new space technologies. Having the right resources and know-how. While there is increased attention on funding for space and building the Space Force, new programs can still face resource challenges. DOD has begun over 9 new space programs at a time when it is also seeking increased investments in ships, aircraft, and the nuclear triad, among other programs. Moreover, it is unclear whether DOD has a sufficient workforce to manage its new programs. GAO issued a report last month that found DOD does not routinely monitor the size, mix, and location of its space acquisition workforce. Further, DOD has difficulty attracting and retaining candidates with the requisite technical expertise. GAO recommended that DOD collect and maintain data on its space acquisition workforce. DOD did not concur, but GAO maintains that DOD should have better information on such personnel, especially in light of its proposal for establishing the Space Force. GAO also found in March 2019 that selected software-intensive space programs often did not effectively engage users to understand requirements and obtain feedback. GAO recommended, and DOD concurred, that the department ensure its guidance addressing software development provides specific, required direction on the timing, frequency, and documentation of user involvement and feedback. What GAO Recommends Past GAO reports have recommended that DOD adopt acquisition best practices to help ensure cost and schedule goals are met. DOD has generally agreed and taken some actions to address these recommendations.

gao_GAO-20-24

gao_GAO-20-24_0

Background Approximately 51,000 drinking water systems and 15,000 public wastewater systems provide clean and safe water to communities nationwide. About 9,000 drinking water systems provide service to 92 percent of the total population, or approximately 273 million people nationwide. The remaining 8 percent of the population is served by small systems that according to the American Society of Engineers frequently do not have the financial, managerial, and technical capabilities necessary to meet state and federal requirements for safe drinking water, such as limits in the levels of specific contaminants in drinking water. Drinking water and wastewater facilities include infrastructure such as tanks, pipes, pumps, and buildings that contain electrical, chemical, and mechanical equipment to treat and test water. The infrastructure is often built to last for over 50 years or longer, depending on the equipment. Many utilities in the country were built decades ago and therefore have existing and aging infrastructure that they must operate and maintain. Utilities generally develop long-term capital plans to identify the infrastructure they will need to replace or rebuild in the future. Utilities generally use historic records of seasonal precipitation, runoff, water temperature, and snow pack levels to determine how their systems should be designed and operated. According to the Water Research Foundation’s 2014 study, utilities have designed their infrastructure based on the expectation that future climate conditions will remain the same and have used historical climate or other data within a 100-year range. Generally, the study reported that utility infrastructure is designed and operated to convey or treat water up to a specific threshold amount based on these historic records. As they plan to rebuild or replace their infrastructure, utilities employ or contract with engineers to ensure that their infrastructure treats and transports water appropriately to meet standards under the Safe Drinking Water Act or the Clean Water Act. Under the Safe Drinking Water Act, EPA, among other actions, sets standards to protect the nation’s drinking water from contaminants, such as lead and arsenic. The Clean Water Act generally prohibits the discharge of pollutants from “point sources”—such as discharge pipes from industrial facilities and wastewater treatment facilities—without a permit. Drinking water and wastewater infrastructure remain the largest financial investment by communities nationwide, according to the Water Research Foundation’s 2014 Study. To pay for operations, maintenance, repair, and replacement of their infrastructure, drinking water and wastewater utilities generally raise revenues by charging their customers for the services they provide. In addition, the federal government invests in drinking water and wastewater infrastructure, as we reported in September 2017. In 2017, the most recent year for which data were available, state and local governments spent approximately $109 billion on their drinking water and wastewater infrastructure, according to Congressional Budget Office data. During the same time period, the federal government spent approximately $4 billion on drinking water and wastewater infrastructure. Climate Information Provided by the Federal Government Agencies across the federal government, such as NOAA and the National Aeronautics and Space Administration, collect and manage many types of climate information and provide technical assistance to make this information more meaningful to federal, state, local, and private decision makers. Decision makers from all levels of government and the private sector use different types of climate information in their planning processes to reduce the potential impacts of climate change. To be useful, climate information must be tailored to meet the needs of each decision maker, such as an engineer responsible for building a bridge in a specific location, a county planner responsible for managing development in a large region, or a federal official managing a national-scale program. Decision makers also need climate information at different timescales corresponding to the short-, medium-, or long-term nature of their planning processes. A 2011 World Meteorological Organization report stated that decision makers need access to expert advice and support to help them select and properly apply climate information. According to a 2010 National Research Council report on making informed decisions about climate change and our November 2015 report on climate information, most decision makers need a basic set of information to understand and make choices about how to adapt to climate change. The set of information includes the following: Information and analysis about observed climate conditions. This includes information on, for example, temperature, precipitation, drought, storms, and sea level rise and how they may be changing in a local area. Information about observed climate impacts and vulnerabilities. This includes site-specific and relevant information on environmental, social, and economic impacts and vulnerabilities, resulting from observed changes in the climate against which past and current decisions can be monitored, evaluated, and modified over time. Projections of what climate change may mean for a local area. This includes, for example, projections based on easily understandable best- and worst-case scenarios with confidence intervals and probability estimates and examples of potential climate impacts. The primary source is NOAA’s online Climate Explorer, which provides climate projections in a range of climate variables relevant to decision makers for every county in the contiguous United States, enabling users to compare historical climate observations under two possible climate change scenarios that could occur this century. Information on the economic and health impacts of climate change. Observed and projected local impacts must be translated into costs and benefits, as this information is needed for many decision-making processes. Agencies across the federal government collect and manage many types of climate information, including observational records from satellites and weather monitoring stations on temperature and precipitation, among other things; projections from complex climate models; and other tools to make this information more meaningful to decision makers. Federal Planning for Critical Infrastructure Resilience Presidential Policy Directive 21 directs federal agencies to work with owners and operators and state, local, tribal, and territorial entities to manage risks and strengthen the security and resilience of critical infrastructure against all hazards. The directive, issued in 2013, identifies 16 critical infrastructure sectors whose assets, systems, and networks—either physical or virtual—are considered so vital to the United States that their incapacitation or destruction would have a debilitating effect on the nation’s security, economy, and public health or safety. One of the sectors is the Water and Wastewater Sector. The directive established a national policy on critical infrastructure security and resilience and made DHS the lead agency to coordinate the overall federal effort to promote security and resilience of the nation’s critical infrastructure. The directive assigned protection responsibilities to selected federal government agencies and departments, called Sector Specific Agencies, and designated EPA as the Sector Specific Agency for the Water and Wastewater Sector. As the Sector Specific Agency, EPA organized a Water and Wastewater Government Coordinating Council, including federal, state, and local decision makers. In turn, water utility owners and operators organized the Water and Wastewater Sector Coordinating Council. EPA and the councils work together and are responsible for planning and implementing the sector’s security and resilience activities. Presidential Policy Directive 21 also directed the DHS to update the National Infrastructure Protection Plan to provide a framework for how federal, state and local decision makers and private sector stakeholders can coordinate to improve the security and resilience of critical infrastructure. The DHS updated the National Infrastructure Protection Plan in 2013 and EPA issued the Water and Wastewater Sector-Specific Plan in 2015. In 2016, the Water and Wastewater Government Coordinating Council and the Water and Wastewater Sector Coordinating Council chartered the Water and Wastewater Sector Strategic Roadmap Work Group to review key threats and vulnerabilities of the sector, identify gaps in the sector’s capabilities relative to the key threats and vulnerabilities, and develop priorities and associated actions to address those gaps. In 2017, the work group issued the report, Roadmap to a Secure and Resilient Water and Wastewater Sector (Roadmap), in part, to help inform utilities’, industry groups’, and government agencies’ planning processes and to support collaboration and leverage resources between stakeholders in the sector. The resulting report identified weather-related disasters, such as floods and earthquakes, and long-term climate-related hazards, such as drought and sea level rise, as among the most significant risks to drinking water and wastewater infrastructure. GAO Work on Climate Information and Disaster Resilience Our previous work on climate change found that the federal government could improve the way that it provides information to facilitate more informed local infrastructure adaptation decisions. In November 2015, we reported that federal agencies could help local infrastructure decision makers by providing the best available climate-related information and by clarifying federal sources of technical assistance for incorporating climate- related information into their planning. In November 2015, we found that federal efforts to provide climate information could be improved by incorporating key organizational and data elements, including (1) a focused and accountable organization; (2) authoritative data that define the best available information for decision makers; and (3) technical assistance to help decision makers assess, translate, and use climate information in planning. We recommended that the Executive Office of the President direct a federal entity to develop a set of authoritative climate change projections and observations and create a national climate information system with defined roles for federal and nonfederal entities. The Executive Office of the President neither agreed nor disagreed with the recommendations and, as of May 2018, had not implemented them. Our previous work on natural disasters found that disaster costs are a key source of federal fiscal exposure. In our July 2015 report on Hurricane Sandy, we found that there was no comprehensive, strategic approach to identifying, prioritizing, and implementing investments for disaster resilience, which increases the risk that the federal government and nonfederal partners will experience lower returns on investments or lost opportunities to strengthen critical infrastructure. We recommended that the Mitigation Framework Leadership Group—an interagency group chaired by FEMA that organizes mitigation efforts across the federal government and assesses the effectiveness of mitigation strategies— establish an investment strategy to identify, prioritize, and guide federal investments in disaster resilience and hazard mitigation-related activities and make recommendations to the President and Congress on how the nation should prioritize future disaster resilience investments. The Mitigation Framework Leadership Group agreed and issued the National Mitigation Investment Strategy in August 2019. In September 2018, we reported that four near-sequential disasters in 2017—Hurricane Harvey, Hurricane Irma, Hurricane Maria, and the California wildfires—created an unprecedented demand for federal disaster response and recovery resources and that Hurricanes Harvey, Irma, and Maria ranked among the top five costliest hurricanes on record. As of June 2018, Congress had appropriated over $120 billion in supplemental funding for response and recovery related to the 2017 hurricanes and wildfires. In October 2019, we issued a Disaster Resilience Framework that identifies federal actions and opportunities to enhance and promote disaster and climate change resilience nationwide focusing on three principles where the federal government can influence decision-making. First, the framework states that federal action can help ensure that decision makers at all levels of government and across industrial sectors can access, understand, and use information on current and future disaster risk. As part of this, federal agencies can use risk reduction strategies, such as providing technical assistance to help decision makers use climate information in their infrastructure investment decisions. Second, the framework stated that federal agencies can help decision makers use risk reduction strategies and prioritize all types of risk. For example, federal agencies can ensure that federal programs and policies that support disaster risk reduction are well coordinated. Third, the framework stated that federal agencies can provide decision makers at all levels of government and across sectors with incentives to make long- term, forward-looking risk reduction investments and remove barriers to such investments. Potential Climate Change Impacts Could Have Various Effects on Drinking Water and Wastewater Infrastructure, and the Type and Severity of the Effects Will Vary by Region Projected increases in the frequency, severity, and duration of extreme temperature changes or precipitation events, as well as rising sea levels, are among the potential impacts of climate change that may affect drinking water and wastewater infrastructure. The type and severity of these potential impacts on drinking water and wastewater infrastructure will vary by region. EPA, the USGCRP, NOAA, and other federal agencies have identified a variety of potential climate change impacts that may affect drinking water and wastewater infrastructure, as well as other critical and interconnected industries. EPA’s Adaptation Strategies Guide for Water Utilities (Guide) identifies five general categories of climate change impacts that can affect drinking water and wastewater utilities: ecosystem changes, droughts, floods, water quality degradation, and changes in service demand and use. Within these five general categories, EPA has identified specific climate change impacts that may affect drinking water and wastewater infrastructure systems. For example, degraded water quality from decreased stream flows may lead to higher treatment costs and the need for capital improvements to treat wastewater before discharging it from wastewater treatment facilities to meet more stringent regulatory requirements. Additionally, projected sea level rise can lead to saltwater intrusion in coastal groundwater aquifers and in estuaries. This may degrade water quality and increase treatment costs for drinking water treatment facilities or require new desalination facilities to treat water supplies with higher salt content. According to the Fourth National Climate Assessment, compound extreme events—the combination of two or more hazard events or climate variables (e.g., extreme rainfall and storm surge) that occur simultaneously or consecutively that lead to an extreme impact—have a multiplying effect on the risk to drinking water and wastewater infrastructure systems. Compound extreme events can also increase the risk of cascading infrastructure failure since some infrastructure systems rely on others and the failure of one system can lead to the failure of interconnected systems. This includes a water infrastructure system relying on the energy sector for power to operate pump stations and drinking water and wastewater treatment facilities. For example, during Hurricane Sandy in 2012, extreme rainfall coincided with high tides creating a storm surge. Hurricane Sandy caused power outages and flooding at eight of New York City’s 14 wastewater treatment facilities and 42 of the city’s 96 pumping stations. Further, power outages and flooding of wastewater treatment facilities and the large influx of floodwater in the sewer system resulted in the release of approximately 562 million gallons of untreated and diluted sewage into local waterways, as shown in figure 1. The Fourth National Climate Assessment states that drinking water and wastewater infrastructure in every region in the United States are sensitive to weather- and climate-related events and noted that the effects of such events will vary in severity and type by region, meaning different measures will be required to make infrastructure more resilient. The Fourth National Climate Assessment established 10 climate regions to better address the risks and needs of specific regions across the United States. Further, EPA’s Guide states that the type and severity of potential climate impacts on utilities will vary by region, and identifies the impacts that have the greatest likelihood of affecting utilities in the different regions and along the U.S. coast. For example, in the Southwest, increased duration and intensity of drought may stress water supplies and increase water demand for agricultural uses, increase energy requirements to treat and cool drinking water and wastewater effluent, and require investments in new water sources and options for reusing water. In the Northwest, increased water temperatures, as well as wildfires that create increased nutrient runoff, may degrade drinking water quality from higher levels of harmful toxins and algal blooms, and require drinking water utilities to develop increased treatment capabilities. The interactive map in figure 2 displays the 10 climate regions as established in the Fourth National Climate Assessment and the U.S. coasts, the most relevant potential climate change impacts for each region and the coast, and examples of the potential effects on drinking water and wastewater utilities, according to EPA’s Guide. One Federal Program Is Designed to Provide Technical Assistance to Water Utilities for Climate Resilience, but Options Exist for Coordinating Additional Technical Assistance One federal program—EPA’s Creating Resilient Water Utilities initiative— is designed to provide technical assistance to drinking water and wastewater utilities for planning climate resilient infrastructure, although the 15 selected utilities used a mix of sources, including other federal programs, to obtain technical assistance with understanding climate impacts and designing resilient infrastructure. To provide additional technical assistance for climate resilience, selected experts generally supported the option of developing a coordinated network of technical assistance providers including federal and state agencies, universities, consultants, and industry groups. EPA’s Creating Resilient Water Utilities Initiative Is Designed to Provide Technical Assistance to Water Utilities for Climate Resilience Our review of the programs federal agencies’ used to provide technical assistance to 15 selected utilities to help make drinking water and wastewater infrastructure more climate resilient found that one program— EPA’s Creating Resilient Water Utilities initiative (CRWU)—was specifically designed to provide drinking water, wastewater, and stormwater utilities with the practical tools, resources, training, and technical assistance needed to increase resilience to extreme weather events. The initiative provides web-based tools and resources in the form of an interactive guide, a case studies map, a risk assessment tool, climate scenario projection maps, and storm surge inundation maps to help drinking water and wastewater utilities understand potential long- term risks and options to enhance their resilience to climate impacts, including extreme weather events. Furthermore, CRWU provides direct utility technical assistance and training through workshops and onsite exercises with utilities. As part of the initiative, EPA developed the Climate Resilience Evaluation and Assessment Tool (CREAT), a web- based application to assist drinking water and wastewater utilities in understanding potential climate change impacts and assessing the related risks to their systems. EPA also developed a Resilient Strategies Guide for Water Utilities, a web-based interactive guide to help drinking water and wastewater utilities identify resilience strategies to prepare for droughts, protect water quality, build flood protections, preserve ecosystems, maintain service levels, improve energy efficiency, implement green infrastructure, and conserve water. In addition, from 2010 through 2013, EPA collaborated with the Water Research Foundation and NOAA to publish the results of a series of workshops assessing the information and tools necessary to incorporate climate risks into utility planning. As part of a pilot program to help EPA develop CREAT, most of the drinking water and wastewater utilities we reviewed used CREAT to conduct climate risk assessments of their systems. Utility representatives said the tool was a helpful starting point for thinking about potential climate risks and vulnerable infrastructure qualitatively. For example, Bozeman Water and Sewer (Bozeman, Montana) used CREAT to assess potential consequences of drought, water quality changes, and wildfires on their drinking water assets and operations to better understand their systems’ vulnerabilities and start thinking about potential resilience measures. Keene Public Works (Keene, New Hampshire) also used CREAT to assess potential climate change impacts from extreme precipitation events on their water supplies and drinking water system and evaluate the performance and costs of additional short-term and long-term resilience measures. However, representatives from a few drinking water and wastewater utilities said they used additional assistance from consulting firms to help them use CREAT, and to complete assessments on the current and future climate risks to their infrastructure systems. A few other federal agencies have been involved in efforts to help utilities incorporate climate resilience into their planning, but their programs were not specifically designed to provide technical assistance to water utilities. NOAA’s Regional Integrated Science Assessments (RISA) program and the National Center for Atmospheric Research (NCAR) worked with utilities and EPA, through an effort called the Water Utility Climate Alliance which aims to enhance the quality and accessibility of regional climate change data to help improve water resource planning, develop adaptation strategies, and assist overall decision-making for water-related policies. The alliance, which was formed in 2007 and includes 12 of the nation’s largest drinking water utilities, provides leadership and collaboration on climate change issues affecting the country’s water agencies. The alliance collaborates with member agencies, federal agencies, industry groups, academia, and consulting firms to provide workshops on planning for climate change uncertainty for drinking water and wastewater sector professionals. Representatives from the New York City Department of Environmental Protection (New York City, New York) and the San Diego County Water Authority (San Diego, California), stated that through their membership in the alliance, they have used technical assistance from NOAA’s RISA program research teams and the Water Research Foundation to manage their climate risks. Specifically, in 2010, four Water Utility Climate Alliance members, including the New York Department of Environmental Protection (New York City, New York), contributed to a pilot project to better understand how climate change might affect their water systems through collaboration between climate experts and utilities, with the goal of improving the process of producing climate information utilities need for decision-making. Two RISA research teams, the Consortium on Climate Risk in the Urban Northeast at Columbia University and the Pacific Northwest Climate Impacts Research Consortium at Oregon State University, provided technical assistance on climate information and modeling to support the effort. In 2013, three Water Utility Climate Alliance members, including the San Diego County Water Authority (San Diego, California), contributed to a research study to increase the adaptive capacity of water utilities in planning for and responding to pressures that may result from climate change, particularly related to the demand for water. The Water Research Foundation led the study. We found that other federal programs offer technical assistance, but the assistance is either not targeted to drinking water and wastewater utilities or it is not specific to climate impacts. For example, San Diego Public Utilities (San Diego, California) worked with the Bureau of Reclamation to assess the region’s water supply and demand, determine the potential effects from climate change impacts within the region, and explore alternatives for addressing future water management challenges. Utilities in Estes Park, Colorado and Iowa City, Iowa worked with FEMA after flood events to develop long-term recovery plans that made their river pipeline crossings stronger and moved a wastewater treatment plant from the floodplain, respectively. In addition, several of the selected utilities worked with NOAA or the U.S. Geological Survey to collect data necessary for planning efforts, including monitoring weather and storms, rainfall levels from stream gauges, and salt water intrusion into water supplies. Houston Water (Houston, Texas) also used NOAA’s Atlas 14 Precipitation-Frequency Atlas to update its floodplain regulations and redefine the amount of rainfall it takes to qualify as a 100-year or 1,000- year flood event (see fig. 3 for pictures of flooded infrastructure). To date, federal efforts to provide technical assistance to help drinking water and wastewater utilities manage climate risks have been small- scale or pilot efforts to develop tools and information. For example, EPA’s CRWU has developed a number of tools and guides for utilities and has provided training and assisted a number of utilities, but the number of utilities that EPA helped directly is small—about 50—and EPA does not have the resources to provide assistance to all utilities, according to EPA officials. Similarly, the Water Utility Climate Alliance’s membership consists of 12 utilities in large metropolitan areas, and has focused on large utilities when developing examples of how drinking water utilities can plan for climate risks according to Water Utility Climate Alliance representatives; however, these alliance members are large enough to have in-house climate expertise and have established relationships with federal or university-based climate services providers. According to industry group officials, the majority of the 70,000 utilities across the country are small and do not have resources to work with consultants or research climate information. While water utilities used federal technical assistance, we found that almost all of the selected drinking water and wastewater utilities, regardless of size, used a mix of technical assistance providers including consultants, industry groups, academia, or federal programs to help them plan for resilience projects, as shown in appendix IV. Most of the selected utilities said they used a mix of assistance because they needed help understanding what climate information and climate models were appropriate to use for their regions and locales. For example, Anacortes Public Works (Anacortes, Washington) worked with the Skagit Climate Science Consortium—a nonprofit organization—to conduct a climate risk assessment for their drinking water system. Anacortes Public Works used the initial climate risk assessment to implement projects that will increase their resilience to the most significant effects from climate- related impacts, flooding, and increased sediment levels in their water supply (see fig. 4). Anacortes Public Works plans to work with the consortium again to better understand how rising sea levels and increasing salinity levels will affect their drinking water supply in the future. A few utilities said that technical assistance efforts should be a collaborative process between the utilities using climate information to make decisions and the scientists providing the technical assistance to ensure that climate information and models are what drinking water and wastewater utilities need to plan for climate resilience. Selected Experts Stated That Additional Technical Assistance and a Network of Technical Assistance Providers Could Help Utilities Enhance Climate Resilience All 10 of the selected experts we interviewed said that drinking water and wastewater utilities need additional technical assistance to manage climate risks. Specifically, these experts stated that utilities need technical assistance to use key climate information to incorporate climate resilience into their planning and operations. This information includes the following: forward-looking climate information and models to identify vulnerabilities to specific geographic regions; potential climate change impacts on regional and local socioeconomic and demographic trends for utility users; hydrologic information on the movement, distribution, and quality of water at the local, regional, and/or watershed level; and estimates of benefits and costs of incorporating resilience into utility projects. According to several of the selected experts we interviewed, such information is provided through a mix of sources, depending on what is available, and all sources are needed. Several selected experts also said that the utilities could obtain forward-looking climate information and models from federal agencies, such as NOAA, and could obtain information on potential climate change impacts from CREAT. In addition, several experts stated that they could obtain local socioeconomic and demographic data, hydrologic information, and benefit-cost information from industry sources, universities, and consultants. Several of the experts we interviewed also said that such assistance is not a one-time event, but requires consistent and continuous collaborative efforts between utilities and technical assistance providers. For example, several experts said that utilities need technical assistance on an ongoing basis to reevaluate their planning and operations regularly given the uncertainty associated with the severity of some potential climate risks. In addition, several experts said that individual utilities need help understanding which climate information and analytical tools are appropriate for assessing the climate risks specific to their regions or localities, and how to use them to manage climate risks to their infrastructure. Almost all of the experts said that small and rural utilities would need additional technical assistance to collect and use the information necessary to enhance their resilience to climate change impacts. Specifically, several experts said that, as opposed to many large utilities, small utilities lack the technical capacity to use climate information and do not have the financial resources to hire consultants or develop the internal expertise necessary to manage climate risks to their drinking water and wastewater infrastructure. Further, most of the selected experts we interviewed stated that a network of providers would be needed to provide assistance to water utilities. This is consistent with what we and others have previously reported. For example, we reported in November 2015 that clearly organized technical assistance would improve federal climate information efforts by helping different types of decision makers—ranging from those who can define their needs to those who have limited experience using climate information—access, translate, and use climate information. We also found that key stakeholders and relevant studies generally called for a system of nonfederal technical assistance providers, with federal leadership to help federal, state, and local decision makers, including utility decision makers, use climate information. In addition, a 2014 task force of state, local, and tribal leaders stated that the greatest need for enhancing climate resilience is often not the creation of new data or information, but assistance and tools for decision makers, including utility managers, in navigating the wide array of resources already available. Further, in August 2019, the National Mitigation Investment Strategy recommended that the federal government increase investment in hazard mitigation by building the capacity of communities to address their risks, including climate-related risks. To implement the recommendation, the strategy said that the federal government should create a professional network to encourage collaboration and information sharing across different levels of government and the water and wastewater sector, and that the federal government and its nonfederal partners should work together to develop a pool of skilled mitigation professionals. The following is a list of options for providing a network of technical assistance providers that selected experts we interviewed discussed, as well as the advantages and disadvantages of each. Existing utility technical assistance providers. A strengthened and expanded network of existing federal technical assistance providers, including EPA’s Environmental Finance Centers, USDA’s Rural Utilities Service, the National Rural Water Association, and the Rural Community Assistance Partnership, could help consolidate climate information and provide technical assistance to utilities to improve their resilience. Most experts said that a network of existing utility technical assistance providers would have the advantage of established relationships with communities and utilities or could ensure that small and rural utilities obtain needed information and assistance to improve their resilience to climate change. However, several experts said that the network may lack the expertise necessary to effectively identify or develop climate information and planning tools to provide the technical assistance necessary to meet the specific needs of utilities to improve their resilience. See appendix V for additional information on these programs. Existing federal climate services providers. A strengthened and expanded network of existing federal climate services providers, such as USDA’s Climate Hubs, Interior’s Landscape Conservation Cooperatives, Interior’s Climate Science Centers, and NOAA’s RISA program could provide technical assistance to utilities to improve their resilience. Several experts said that a network of existing federal climate services providers would have a good understanding of the available climate information and would, for example, be best positioned to develop the specific tools and guidance necessary to provide the technical assistance utilities need to improve their resilience. In contrast, several experts said that federal climate services providers may not have the established relationships with utilities necessary to understand and tailor technical assistance to the needs of individual utilities. In addition, one expert said that the climate services providers may not have the funding to provide these services to utilities in a comprehensive way. See appendix V for additional information on these programs. Universities and university-based research centers. A new network of academic or university-based technical assistance providers, such as NCAR, organized by state, region, or watershed could provide technical assistance to all types of utilities to improve their resilience. According to several experts, this option would be advantageous because many universities and centers already have the technical capacity to use climate information to provide risk assessment and planning tools necessary to provide technical assistance to utilities at the local or regional level. Several experts also said that it would be cost-effective to expand this option because some universities and centers are already providing technical assistance. However, several experts said that without a clear shift in federal incentives to prioritize the applied research necessary to provide the technical assistance that utilities need, universities and centers are unlikely to provide sustained assistance nationwide. Similarly, several experts said that federal coordination would be needed to ensure that the universities and centers were consistently providing information, planning tools, and assistance that meet the specific needs of utilities. See appendix V for additional information on these programs. Industry groups and private-engineering consultants. A new network of nonfederal industry and nonprofit groups, such as the American Water Works Association and the Association of Municipal Water Utilities, could provide technical assistance to utilities to improve their resilience. Several experts said that this option would be advantageous because it could leverage existing relationships, for example, to strengthen information sharing between utilities regarding the best available climate information and approaches to resilience planning. In addition, several experts said that industry groups and private engineering consultants would have a better understanding of utility operations and management when compared to other options for providing technical assistance. In contrast, half of the experts said that this network would need additional federal oversight and coordination. For example, several experts said that there would need to be a certification process for industry groups and private consultants to ensure that the technical assistance being provided to utilities was sufficient and transparent. In addition, several experts said that the network would not be effective unless it was coordinated among stakeholders from, for example, the private sector; industry groups; and federal, state, and local governments. A network of utilities. A network of utilities, similar to the Water Utility Climate Alliance, could consolidate and update information and provide technical assistance for all types of utilities to improve their resilience. Similar to a network of industry groups and consultants, several experts said a network of utilities could help coordinate and strengthen information sharing between utilities on best practices and lessons learned from resilience planning. However, several other experts said that it would be difficult to develop and expand a network of utilities that was capable of providing technical assistance to utilities of different sizes or geographic locations. One expert also said that utilities that provide technical assistance would need to be certified by the federal government, academics, or industry groups to ensure the technical assistance being provided to utilities was sufficient. When asked how they would design a network to provide technical assistance, most experts supported an approach in which federal agencies organized a network of technical assistance providers for drinking water and wastewater utilities, a network that would include federal and state agencies, universities, consultants, and industry groups. For example, one expert said that EPA and other relevant federal agencies could provide guidance and leadership for a network of (1) university and federal climate services providers that would assess the risks that potential climate impacts pose to utilities and (2) utility technical assistance providers, including consultants and industry groups, to help utilities apply those assessments to their infrastructure to make it more resilient. Another expert said that existing networks of universities, industry groups and consultants, or utilities would not be as effective unless they were part of a larger networked effort with clear leadership that provides continuous technical assistance to utilities. Similarly, several experts stated that it was important that the network be a collaboration of different technical assistance providers to be able to tailor the technical assistance to the needs of different types of utilities, in different locations, with differing technical capabilities. For example, one expert said that universities, industry groups, and federal programs have different levels of resources and expertise in different regions of the country and a coordinated network could help utilities identify the sources of technical assistance in their regions or localities. Further, another expert said that it was important that the network have the capability to help utilities understand and respond to climate risks that other types of infrastructure create. Specifically, the expert said that while EPA has a role in regulating drinking water and wastewater infrastructure, the agency does not regulate larger-scale infrastructure, such as dams and reservoirs that need to be operational to reduce risks to utilities. Under Presidential Policy Directive 21, EPA, as the Sector Specific Agency for the water and wastewater systems sector, is to work to enable efficient information exchange between federal agencies and infrastructure owners and operators, and to implement an integration and analysis function to inform planning and operational decisions regarding critical infrastructure. In addition, one of the key activities of the Water Sector Government Coordinating Council, which EPA chairs, is to facilitate information sharing between federal, state, and local decision makers on critical infrastructure protection. This is consistent with our disaster resilience framework, which states that federal efforts should improve the availability of authoritative, understandable, and comprehensive information on disaster risks and risk reduction strategies to help entities effectively assess their climate risks, determine what viable alternatives are available to increase resilience to those risks, and better understand and measure the impact of resilience strategies. Our framework also states that federal efforts can help by providing technical assistance and capacity building to nonfederal partners. To date, however, federal efforts to provide technical assistance to drinking water and wastewater utilities do not provide the ongoing technical assistance that according to experts utilities need to plan and build climate resilient infrastructure. In addition, current efforts may not be widespread enough to provide comprehensive coverage of the drinking water and wastewater utilities across the nation. The 2017 Roadmap shows actions for the short term (2 years) and midterm (5 years), but it does not include actions such as developing guidance on technical assistance, building networks of technical assistance providers, or developing other methods to help utilities build capacity to manage their climate change risks and plan for resilient infrastructure. According to EPA officials we interviewed, the agency has worked within its existing authorities and available resources to prioritize developing voluntary guidance, tools, training, and webinars that utilities can use to identify potential risks from climate change and plan to improve their resilience. Further, EPA officials said that while the agency has collaborated closely with key federal, state, and local decision makers; industry groups; and utilities in its role as chair of the Water Sector Government Coordinating Council, the council has focused on other short-term threats to utilities, such as disasters and terrorism, and has not assessed how it could develop and coordinate a network to effectively provide the technical assistance that utilities need to enhance their climate resilience. By identifying and engaging existing technical assistance providers in a network to help drinking water and wastewater utilities incorporate climate resilience into their projects and planning on an ongoing basis, EPA would have more reasonable assurance that climate information was effectively exchanged among federal agencies and infrastructure owners and operators. Supporting the need for a broader collaborative approach, several of the selected utilities are already members of organizations that coordinate and collaborate among members and various technical assistance providers, including federal agencies, to understand the potential climate impacts for their regions, use similar climate models, and share best practices for projects to enhance climate resilience. For example, the Southeast Florida Regional Climate Change Compact is a decade-old partnership between Miami-Dade, Broward, Monroe, and Palm Beach Counties to coordinate mitigation and adaptation activities across county lines in response to the effects of climate change, including sea level rise, flooding, and economic and social disruptions. The compact and its partners work with various federal, state, regional, municipal, nonprofit, academic, and private sector partners to provide technical assistance and support for utilities in southeast Florida to help the region identify emerging issues and all move in one direction for resilience planning efforts. The supporting federal agencies include NOAA, EPA, and the Army Corps of Engineers. Another example is Charleston Water (Charleston, South Carolina), a member of the Charleston Resilience Network—a collaborative group of public, private, and nonprofit organizations in the region that work to increase resilience of communities, critical infrastructure, and the economy to natural disasters and chronic coastal hazards, such as rising sea levels. The network provides a forum to share science-based information, educate stakeholders, and enhance long-term resilience planning decisions. The network also works to provide consistent information for planning decisions. The federal agencies that advise the Charleston Resilience Network include NOAA, DHS, and the Army Corps of Engineers. The Federal Agencies Do Not Consistently Provide Financial Assistance for Projects That Could Enhance Climate Change Resilience and Limit Future Fiscal Exposure The four selected federal agencies in our review provide broad financial assistance to help drinking water and wastewater utilities plan and build infrastructure projects. The agencies have taken some actions to promote climate resilience when providing financial assistance for water infrastructure projects, but they do not consistently include the consideration of climate resilience when funding such projects. Most selected experts we interviewed suggested that requiring the consideration of climate change risks in the planning and design of all federally funded water and wastewater infrastructure projects could help enhance climate resilience and limit future federal fiscal exposure. Four Federal Agencies Provide Financial Assistance for Projects But Do Not Consider Climate Resilience Consistently The four federal agencies we reviewed have nine programs that provide broad financial assistance, through loans or grants, for drinking water and wastewater infrastructure (see table 1). However, federal programs generally do not have selection criteria or requirements for utilities to incorporate climate resilience in the planning and design of projects that receive federal financial assistance. Each of the programs used different selection criteria for providing financial assistance to drinking water and wastewater utilities. EPA’s Drinking Water State Revolving Fund, Clean Water State Revolving Fund, and Water Infrastructure Finance and Innovation Act (WIFIA) programs generally provide financial assistance to projects that address the most serious risks to human health and ensure compliance with the Safe Drinking Water Act or Clean Water Act. Other programs, such as FEMA’s Public Assistance and Hazard Mitigation Grant programs provide financial assistance to repair or replace infrastructure damaged during natural disasters, or to enhance disaster resilience against future damage. HUD’s Community Development Block Grant-Disaster Recovery funding is used for, among other things, projects to help cities, communities, and states recover from presidentially-declared disasters or enhance disaster resilience of damaged infrastructure, especially in low- income areas. USDA provides financial assistance for drinking water and wastewater infrastructure in small and rural communities. According to EPA, FEMA, HUD, and USDA officials we interviewed, drinking water and wastewater utilities can use financial assistance from their programs to pay for projects that, in addition to other benefits, can help enhance climate resilience. We have previously reported that the federal government invests billions of dollars annually in infrastructure—such as roads, bridges, and wastewater infrastructure—but faces increasing risks from climate change. When the climate changes, infrastructure—typically designed to operate within past climate conditions—may not operate as well or for as long as planned, leading to economic, environmental, and social impacts. We have also reported that some federal agencies have made efforts to manage climate change risk within existing programs and operations—a concept known as mainstreaming—and these efforts may convey some climate resilience benefits. For example, an agency planning to build a seawall to protect a coastal facility might build it higher to account for rising sea level projections, but may not track this spending as related to climate change. Representatives of several of the drinking water and wastewater utilities we reviewed reported using selected federal financial assistance programs to help fund projects for fiscal years 2011 through 2018 that, in addition to other benefits, enhanced their climate resilience. For example, Iowa City Public Works used financial assistance from HUD Community Development Block Grant-Disaster Recovery funding and FEMA’s Public Assistance grant program to increase their resilience to floods by relocating a flood-prone wastewater treatment facility after flooding in 2008, as shown in figure 5. Similarly, as of December 2018, Houston Water was working with FEMA to use Public Assistance grants and Hazard Mitigation grants to increase the utility’s resilience to floods and extreme storm events when rebuilding the wastewater infrastructure damaged by Hurricane Harvey in 2017, according to Houston Water representatives. In addition, the San Diego Public Utilities Department received an EPA WIFIA loan to increase its resilience to droughts by building a new recycled wastewater treatment facility that will provide an additional source of drinking water and reduce the need for water imported from the Colorado River Basin (see app. VI for details on completed and ongoing infrastructure projects that utilities undertook to enhance their climate resilience, according to selected drinking water and wastewater utility representatives). The remaining selected utilities relied on other sources of funding such as municipal bonds and funds raised primarily through user rates and fees for fiscal years 2011 through 2018 to enhance their climate resilience (see app. VII for details on the financial assistance drinking water and wastewater utilities used for infrastructure projects). However, making the nation’s drinking water and wastewater infrastructure resilient will be expensive, costing anywhere from $448 billion to $944 billion, including operations and maintenance through 2050, according to a 2009 Association of Metropolitan Water Agencies study, the most recent such study. These costs would likely be in addition to the EPA-estimated $774 billion in costs for replacing and repairing existing infrastructure over the next 20 years. According to representatives of several of the selected utilities in our review, additional financial assistance will be necessary to enhance the resilience of drinking water and wastewater infrastructure. Representatives from several utilities said they would not be able to make the necessary upgrades to incorporate climate resilience into their drinking water or wastewater systems without additional grant assistance. Based on estimates from one of the selected utilities, the costs to enhance their resilience will be high. For example, in 2013, the New York Department of Environmental Protection estimated that it would cost about $315 million to build the protective measures necessary to make its wastewater treatment facilities and pump stations resilient to future flood projections. Officials from EPA, FEMA, HUD, and USDA said that federal agencies have taken action to change program requirements or selection criteria to provide financial assistance for projects that enhance climate resilience. However, according to federal officials, some federal agencies are providing financial assistance to utilities for projects that do not consider climate resilience in their planning and design consistently. In addition, federal officials stated that their ability to require that climate resilience be incorporated in the projects they fund is limited by requirements specific to their programs. Examples of their efforts, and limited authorities, include the following: EPA. EPA provides annual grants to states to capitalize their state- level drinking water and wastewater state revolving fund programs. The states use the revolving funds to provide low-cost loans or other financial assistance to communities for, among other things, a wide range of drinking water and wastewater infrastructure projects. According to EPA officials, states establish program criteria and do not consider climate resilience consistently in planning and designing projects that receive financial assistance from state revolving fund programs. Specifically, EPA officials said that despite agency efforts to promote climate resilience, states have discretion in setting project funding criteria and priorities for their state revolving fund programs, and that the agency does not have the authority to require that states prioritize projects that incorporate climate resilience. EPA continued to encourage the states to incorporate resilience planning in their priority systems. In documents released in May 2016, September 2016, and June 2017, the EPA described the types of climate resilience projects eligible for drinking water and clean water state revolving fund assistance. The September 2016 document also describes how programs can encourage resilient infrastructure through financial incentives. According to fiscal year 2015 data that EPA provided, 17 state clean water revolving fund programs have created additional financial incentives that utilities could use to fund climate resilience projects, and only New York’s program requires that climate risks from sea level rise be incorporated into the projects that receive financial assistance. In addition, utilities have discretion in whether to incorporate climate resilience into their state revolving fund project applications, and EPA cannot require utilities to incorporate climate resilience into the planning and construction of projects that states fund, according to EPA officials. Similarly, while EPA manages the WIFIA program and its application process and criteria, EPA officials said that the 2018 and 2019 program guidance did not prioritize protection against the impacts of climate change in its selection criteria, and that the agency does not require that applicants incorporate climate resilience into project planning and design. FEMA. FEMA’s Public Assistance Grant Program provides grants to state, tribal, territorial, and local governments, and nonprofits that can be used to repair and replace damaged infrastructure, including drinking water and wastewater infrastructure. In addition, FEMA’s Pre- Disaster Mitigation and Hazard Mitigation Grant Programs can provide financial assistance to states, communities, or tribes that can be used to reduce the risks to drinking water and wastewater infrastructure from future disasters. FEMA officials said they have developed guidance for states and communities to incorporate climate resilience into the planning for projects funded by all three programs. However, officials said that states and utilities do not consider climate change resilience consistently in planning and designing of projects that use financial assistance from FEMA. Specifically, according to FEMA officials, funding through the Public Assistance Program and the Hazard Mitigation Grant Program is limited to states and localities with a presidentially-declared disaster and generally is not provided for projects that incorporate climate resilience into their planning and design. In addition, according to FEMA officials, states and localities have discretion over the projects they choose to submit for funding and FEMA cannot require them to incorporate climate resilience into the planning and construction of projects that states fund without a change to program requirements. HUD. HUD provides grants to states and local governments through its Community Development Block Grant program to fund housing; economic development; neighborhood revitalization; and other community development activities, including drinking water and wastewater infrastructure. In addition, HUD can provide grants that can be used for reconstruction of drinking water and wastewater infrastructure to help communities recover from presidentially declared disasters through its Community Development Block Grant program. According to HUD officials, the agency has taken action to encourage states and local governments to incorporate climate resilience planning in the projects they fund after disasters. Officials also said that HUD provides guidance on how financial assistance requirements for states and entitlement communities can be waived so that states and communities can use Community Development Block Grant funding for disaster recovery and resilience in presidentially-declared disaster areas. In addition, in 2016, HUD finalized rules requiring states and localities to consider incorporating resilience to natural hazard risks and climate change into their planning documents for Community Development Block Grant funding in low- and moderate-income communities. However, officials said that states do not consider climate change resilience consistently when planning and designing projects using financial assistance from HUD. Specifically, according to HUD officials, the agency can only directly provide financial assistance to projects that enhance climate resilience using Community Development Block Grant-Disaster Recovery Grants if climate change resilience is specified in disaster relief appropriations language. Further, states and localities have discretion regarding whether to incorporate climate resilience into their project applications, and HUD cannot require them to incorporate climate resilience into the planning of projects that receive financial assistance, according to HUD officials. USDA. USDA’s Rural Utilities Service provides grants and loans for drinking water, wastewater, and stormwater projects in rural areas— defined as any area not in a city or town with a population in excess of 10,000 inhabitants. According to USDA officials, the agency has promoted climate resilience planning through its Water and Waste Disposal Program by requiring small and rural utilities to complete planning and vulnerability assessments for natural disasters. In addition, USDA officials said the agency has collaborated with EPA to develop guidance and training through the Sustainable Rural and Small Utility Management Initiative to help small and rural utilities create plans for improving their sustainability, including planning to help make the utilities resilient to potential climate impacts. According to USDA officials, utilities have discretion in whether to incorporate climate resilience into their Water and Waste Disposal project applications, and USDA cannot under its current regulations require them to incorporate climate resilience into the planning and construction of projects that receive financial assistance. As a result, according to officials, utilities do not consider climate resilience consistently when planning and designing projects that receive financial assistance from USDA. Most Selected Experts Said That Requiring Climate Resilience in Federal Projects Would Help Address Future Climate Impacts and Limit Future Federal Fiscal Exposure According to most selected experts, requiring the consideration of climate risks in projects that receive financial assistance will help limit the future fiscal exposure of the federal government and help enhance the climate resilience of drinking water and wastewater infrastructure. Specifically, most of the experts we interviewed said that a federal requirement that potential climate impacts be considered and, if necessary, incorporated into the design of all new drinking water and wastewater infrastructure projects that receive federal financial assistance, should be a high or very high priority for the federal government. Several of the experts said that this option would be advantageous because it could help ensure more effective and efficient use of federal dollars on drinking water and wastewater infrastructure. For example, several experts said that this option would help ensure that infrastructure funded by the federal government incorporated climate risks during the planning stages, helping avoid expensive retrofits or the abandonment of federally funded infrastructure that was not climate resilient. Several other experts said that such a federal requirement could help make consideration of future climate risks to enhance resilience a standard industry practice within the drinking water and wastewater sector. Several of the selected utilities said that a federal requirement for potential climate impacts to be considered and, if necessary, incorporated into the design of all new drinking water and wastewater infrastructure projects that receive federal financial assistance, would be moderately to extremely effective in helping utilities enhance their resilience. These selected utilities also said that it would be at least moderately feasible to implement. Several of the selected utilities are already required to consider some potential climate risks in the planning and design of their drinking water and wastewater infrastructure. For example, according to representatives from the Miami-Dade County Water and Sewer Department, Miami-Dade County adopted an ordinance requiring that potential climate risks be considered in the design of county-funded infrastructure. According to the same officials, this requirement has shifted the culture of the Miami-Dade County Water and Sewer Department to emphasize potential future climate change risks in the planning and design of all of the county’s drinking water and wastewater infrastructure. Representatives from a few selected utilities also said that a requirement could make it easier to access federal financial assistance programs for projects that enhance climate resilience. Several selected experts cautioned that many utilities do not have the climate information and technical capacity to carry out such requirements or that the uncertainty of the available climate science would make it difficult to implement for some utilities. In addition, several experts said that such a requirement may force utilities with limited funding to prioritize planning and investment in projects to improve climate resilience over more pressing concerns, such as repairing and replacing damaged or obsolete infrastructure. Several selected utilities said that it will be difficult to implement these new requirements, but added that additional technical and financial assistance could help. For example, representatives from Cottage Grove Public Utilities said that the federal government will need to provide additional financial and technical assistance opportunities for small and medium-sized public utilities that do not have the capacity to plan, implement, and fund large climate resilience projects. However, if the federal agencies do not require the incorporation of climate resilience into the projects that receive financial assistance, they may continue to fund drinking water and wastewater infrastructure projects that may be damaged or incapacitated by future floods, drought, water quality problems, and other climate change impacts. This increases the risk that critical infrastructure will not be well protected and drinking water and wastewater utilities will not be able to continue operations that provide critical public health and environmental services to the public. EPA and other federal, state, local, and sector-level officials, recognizing the need to incorporate climate resilience into drinking water and wastewater infrastructure, have taken action to promote climate resilience but generally do not require it to be incorporated in these projects. Specifically, the 2017 Roadmap calls for the Water Government Coordinating Council and the Water Sector Coordinating Council to promote eligibility criteria for financial assistance programs to support resilience activities by 2019. In addition, in a 2019 report, EPA’s Environmental Finance Advisory Board recommended that EPA create a coordination group to set priorities and reduce gaps in funding predisaster resilience for drinking water and wastewater infrastructure, and that EPA consider expanding the state revolving fund program to include financial assistance for flooding and storm-related damages. Further, the National Mitigation Investment Strategy, issued in draft in January 2018, and finalized in August 2019, states that successful mitigation of natural hazard risks requires shared priorities, consistent approaches, aligned funding, expanded incentives, and coordination between the federal government and nonfederal partners. It also states that the federal government and nonfederal partners should look at risk and resilience consistently by, for example, having similar requirements for assessing risk and rebuilding for long-term resilience. It emphasizes the need to focus on critical infrastructure in communities, such as drinking water and wastewater infrastructure. Incorporating climate resilience likely decreases the risk that water and wastewater infrastructure, some of which is paid for with federal financial assistance, will fail during extreme events. According to the National Research Council, as the climate changes and historical patterns—in particular, those related to extreme weather events—no longer provide reliable predictions of the future, infrastructure designs may underestimate the climate-related impacts to infrastructure over its design life, which can range as long as 50 to 100 years. In April 2013, we reported that according to one set of commonly used design standards, wastewater treatment plant components are typically designed for 25-, 50-, or 100-year storms. We reported that changes in characteristics of strong storms—for instance, a storm that historically occurred once every 100 years may occur every 50 years in the future—could cause wastewater management systems to be overwhelmed more frequently. Incorporating climate resilience into drinking water and wastewater infrastructure projects also likely decreases the risk that the federal government will need to pay to repair and replace damaged facilities. In our previous work, we said that building resilience can help reduce the federal fiscal exposure. As we reported in April 2013, such resilience means reducing potential future losses rather than waiting for an event to occur and paying for recovery afterward. We said that enhancing resilience can create additional up-front costs, but can also reduce potential future damage from climate-related events that—given expected budget pressures—would otherwise constrain federal programs. In 2018, the National Institute of Building Sciences found that every dollar spent on infrastructure hazard mitigation to enhance resilience to wind- and flooding-related disasters resulted in 7 to 8 dollars in avoided future losses, respectively. This potential can be considered in light of recent costs that the federal government incurred to address losses. In particular, from fiscal year 2011 through fiscal year 2018, we estimate that FEMA’s Public Assistance program and HUD’s Community Development Block Grant-Disaster Recovery Grants have obligated at least $2.3 billion and at least $1.4 billion, respectively, in federal disaster recovery funding on drinking water and wastewater infrastructure-related projects. Conclusions Drinking water and wastewater utilities face challenges in using climate information to identify actions that they can take to enhance their climate resilience. At the moment, utilities obtain technical assistance and use climate information from a mix of sources and that assistance is not organized to help ensure more comprehensive coverage of the more than 70,000 drinking water and wastewater utilities across the nation. As designated lead agency for the resilience and security of the drinking water and wastewater sector and as chair of the Water Sector Government Coordinating Council, EPA is tasked with coordinating federal and sector efforts to provide the information and assistance that state and local decision makers—including utilities—need to enhance their climate resilience. The councils have identified a number of actions to support the drinking water and wastewater sector, but EPA, other federal agencies, and the water and wastewater sector, have not assessed how they could organize a network of technical assistance providers to effectively provide the assistance that utilities need to enhance their resilience to climate change. By identifying existing technical assistance providers and engaging them in a network to help drinking water and wastewater utilities consider climate resilience in the planning and design of projects on an ongoing basis, EPA, as chair of the Water Sector Government Coordinating Council, would have more reasonable assurance that climate information was effectively exchanged among federal agencies and infrastructure owners and operators. In recognition of the federal interest in protecting the health and economic benefits that clean and safe water provide, federal programs provide funding to support drinking water and wastewater infrastructure. In 2013, Presidential Policy Directive 21 identified the water and wastewater sector as critical infrastructure, with important implications for protecting and investing in that sector. Federal agencies such as EPA, FEMA, HUD, and USDA provide financial assistance to help ensure the long-term success of drinking water and wastewater utilities. These agencies have taken action to promote climate resilient infrastructure projects with the financial assistance they provide, but their abilities to ensure that projects receiving financial assistance are resilient are limited. To enable agencies to further drive climate resilient investments by drinking water and wastewater utilities, changes would be needed to programs that EPA, FEMA, HUD, and USDA administer to require that climate resilience be incorporated into planning for projects that receive federal financial assistance. Such changes could help ensure that drinking water and wastewater infrastructure projects that receive federal financial assistance adequately address risks from climate change and ensure that utilities carry out their critical operations. Such changes could also help limit the fiscal exposure to the federal government for future recovery costs. Matter for Congressional Consideration We are making the following matter for congressional consideration: Congress should consider requiring that climate resilience be incorporated in the planning of all drinking water and wastewater projects that receive federal financial assistance from programs that EPA, FEMA, HUD, and USDA administer. (Matter for Consideration 1) Recommendations for Executive Action We are making one recommendation to EPA: The Director of Water Security of EPA, as Chair of the Water Sector Government Coordinating Council, should work with the council to identify existing technical assistance providers and engage these providers in a network to help drinking water and wastewater utilities incorporate climate resilience into their projects and planning on an ongoing basis. (Recommendation 1) Agency Comments and Our Evaluation We provided a draft of this report to EPA, DHS, HUD, NOAA, and USDA for review and comment. EPA provided written comments, which are reproduced in appendix VIII. The other four agencies did not provide comments on our draft report. EPA and USDA provided technical comments, which we incorporated as appropriate. In its written comments, EPA neither agreed nor disagreed with our recommendation that the Administrator, as Chair of the Water Sector Government Coordinating Council, should work with the council to identify existing technical assistance providers and engage these providers in a network to help drinking water and wastewater utilities incorporate climate resilience into their projects and planning on an ongoing basis. The agency noted in its technical comments that the Director of Water Security is the chair of the Water Sector Council, not the administrator. We made this change in the report. In its written response, EPA made two points related to the recommendation. First, it stated that it will, consistent with our recommendation, continue to work with its wide-ranging, existing technical assistance providers and coordinate with its stakeholders to identify additional providers as applicable. We agree with this approach and highlighted several of these efforts in our report. For example, EPA noted that it provides annual training to over 5,000 water utilities, state officials, and federal emergency responders on how to become more resilient to natural or manmade incidents that could endanger water and wastewater services. Second, in response to the part of the recommendation that EPA engage the providers in a network, the agency noted that states serve as a coordinating entity under its Small System Training and Technical Assistant grants. Further, EPA also noted that the providers work with states to identify the systems in greatest need of assistance and identify the training topics of greatest need for small public water systems. We agree that this could be a helpful approach, but note that EPA remained silent on how it plans to work with the states and the water and wastewater sector to develop a network of technical assistance providers. Our report showed that utilities obtain technical assistance from a number of different sources and that they could benefit from a larger network with continuous technical assistance. The Water Sector Coordinating Council functions as a forum to coordinate members of existing networks, and to ensure they have the most current and relevant information as they provide assistance to utilities. As EPA works with its wide-ranging technical assistance providers, consistent with our recommendation, we would encourage it to also work with the Water Sector Coordinating Council to ensure the coordination of the different networks that exist in the water and wastewater sector. As agreed with your offices, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees; the Administrator of the Environmental Protection Agency; and the Secretaries of Homeland Security, Housing and Urban Development, Commerce, and Agriculture. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staffs have any questions about this report, please contact me at (202) 512-3841 or gomezj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff members who made key contributions to this report are listed in appendix IX. Appendix I: Objectives, Scope, and Methodology The objectives of our review were to examine (1) the potential impacts of climate change and the effects of these impacts on drinking water and wastewater infrastructure; (2) technical assistance selected federal agencies provided to selected utilities to help make drinking water and wastewater infrastructure more resilient to the impacts of climate change, and options experts identified for providing additional technical assistance to utilities; and (3) financial assistance federal agencies provided to selected utilities to help make drinking water and wastewater infrastructure more resilient to the impacts of climate change, and options experts identified for providing additional financial assistance to utilities. For the first objective, we reviewed the Fourth National Climate Assessment; the Environmental Protection Agency’s (EPA) Adaptation Strategies Guide for Water Utilities, Climate Resilience Evaluation and Awareness Tool Methodology Guide, and Climate Scenarios Projection Map, and the U.S. Climate Resilience Toolkit, which the Department of Commerce’s National Oceanic and Atmospheric Administration (NOAA) manages and hosts with oversight from the U.S. Global Change Research Program. Based on our review of these sources, we first identified different categories of potential climate change impacts, and how those impacts may vary in the different climate regions identified in the Fourth National Climate Assessment. For both the second and third objectives, we reviewed the efforts of and interviewed five federal agencies and 15 drinking water and wastewater utilities. We reviewed our previous reports to identify agencies that provide technical assistance or financial assistance, or both, to drinking water and wastewater utilities and identified five agencies: EPA, NOAA, the Department of Homeland Security’s (DHS) Federal Emergency Management Agency (FEMA), the Department of Housing and Urban Development (HUD), and the Department of Agriculture’s (USDA) Rural Utilities Service. For the second and third objectives, we also selected a nongeneralizable sample of 15 drinking water and wastewater utilities in 13 communities using a stratified purposeful sampling approach. We selected utilities to obtain variation in their size and climate region to capture similarities and differences among utilities. We classified utilities into small, medium, and large utilities based on the sizes of the populations that they serve. We defined small utilities (serving populations of 10,000 or less), medium utilities (serving populations of 10,001 to 999,999), and large utilities (serving populations of 1 million or more) for this report to capture utilities with the greatest resources available for climate resilience efforts. In order to ensure geographic diversity, we selected small, medium, and large utilities from different climate regions identified in the Fourth National Climate Assessment. Because this was a nonprobability sample, the findings related to the 15 utilities cannot be generalized to all drinking water and wastewater utilities but provide illustrative examples of how the selected utilities used federal technical assistance and financial assistance. Further, for the second and third objectives, we selected 10 experts in the climate change and disaster fields to interview about options for providing additional technical and financial assistance to drinking water and wastewater utilities. To identify experts on the resilience of water infrastructure to climate change, we searched Elsevier’s Scopus database for peer-reviewed articles published from January 2003 through September 2018 searching titles, abstracts, and keywords for “drinking water” or “wastewater” in close proximity to terms such as “infrastructure,” “climate change,” and “resiliency.” We identified approximately 300 studies from this search, identified the relevant studies from that group, and then found an additional eight studies from their citations. We reviewed the abstracts of these studies and found 96 that were within the scope of our objectives. To develop a list of potential experts, we extracted the names of the authors of these studies and the names of authors cited in these studies using the Python programming language and the Scopus Application Programming Interface. Next, we used statistical software to calculate the number of times that each author cited every other author. Using these calculations, we arrayed the authors into a network graph, in which authors who frequently cited each other were situated closer together and authors who did not cite each other were situated further apart. We analyzed this network using social network analysis techniques. Specifically, to measure each author’s prominence in the network, we calculated the number of times that each author was cited in the articles written by other authors in the network. To divide the network into groups, we used an algorithm known as hierarchical clustering. This algorithm allowed us to identify groups of authors who cited each other frequently and who cited authors in the rest of the network infrequently. We sorted authors by group and by the number of times they were cited. For the most frequently cited authors in the largest groups in the network, we examined biographical details and publication details via web searches, such as their geographic location and the relevance of their publications to our research topic. We selected a final list of 15 frequently cited experts who were primarily from the largest clusters in the network, who were based in North America, whose research was topically relevant, and who were still active in the field. Eight of these experts agreed to be interviewed and we included them in our final sample. We supplemented this list with two experts who served as lead authors for the water chapter of the Fourth National Climate Assessment. While these 10 experts are prominent researchers and correspond to a range of major fields of research on the topic, their views do not represent the views of all experts on the resilience of drinking water and wastewater infrastructure to climate change. To examine the first part of the second objective, the technical assistance selected federal agencies provided to selected utilities, we reviewed relevant laws, regulations, and planning guidance about programs that can provide technical assistance to drinking water and wastewater utilities to help enhance climate resilience for each selected federal agency. We also interviewed federal officials at each agency. To examine the first part of the third objective, the financial assistance selected federal agencies provided to selected utilities, we reviewed project eligibility criteria and appropriation amounts for EPA’s Clean Water State Revolving Fund, Drinking Water State Revolving Fund, and Water Infrastructure Finance and Innovation Act Programs; HUD’s Community Development Block Grant Program and Community Development Block Grant-Disaster Recovery Fund; and USDA’s Water and Wastewater Disposal Program for fiscal years 2011 through 2018. We also interviewed federal officials at each agency. As part of analyzing the federal financial assistance to drinking water and wastewater utilities, we estimated FEMA’s pre- and post-disaster spending to help such utilities recover from natural disasters. To identify federal disaster recovery and hazard mitigation obligations on drinking water and wastewater infrastructure, we analyzed federal financial assistance that FEMA’s Public Assistance, Hazard Mitigation, and Pre- disaster Mitigation Programs provide for disaster recovery for drinking water and wastewater infrastructure. Specifically, using a list of search terms associated with drinking water and wastewater infrastructure, we queried FEMA’s disaster recovery spending database to identify a list of drinking water and wastewater infrastructure disaster recovery and hazard mitigation projects funded from fiscal years 2011 through 2018. After we queried FEMA’s disaster recovery spending database, we manually reviewed records from a stratified sample to ensure that each project was related to water and wastewater infrastructure. We reviewed all 25 records with the highest obligated amounts, 15 records in which a project was associated with more than one site, and 35 records in which a project was associated with just one site. We chose this sample design to ensure that we were capturing projects with the highest dollar amounts as well as all other projects, while also ensuring that if one site in a project was water related, the rest of the sites under the project were also water related (manual review showed that if one site in a project was water related, 98 percent of the other sites in the project were also water related). After manual review, we generated an estimate of total obligated funds from the ratio of number of projects that we reviewed that were related to water and wastewater infrastructure to the total number of projects in our sample. The estimate we used was the lower bound of a 95 percent confidence interval. We chose this estimate in order to give a conservative estimate of the amount that FEMA’s public assistance program has obligated. The relative error was 0.07. To assess the reliability of the disaster recovery obligations data, we (1) performed electronic testing for errors in accuracy and completeness, (2) reviewed related documentation about the data and the system that produced them, (3) interviewed agency officials knowledgeable about the data, and (4) worked closely with agency officials to identify and resolve data discrepancies before conducting our analyses. We determined that the data were sufficiently reliable for the purposes of our reporting objectives. To examine what technical assistance and financial assistance selected drinking water and wastewater utilities used for the second and third objectives, we provided a short questionnaire and interviewed utility representatives from the 15 selected drinking water and wastewater utilities to understand what technical and financial assistance they used to enhance their climate resilience for fiscal years 2011 through 2018. In the questionnaire and interviews, we discussed their efforts to plan for climate resilience and the technical and financial assistance they used for such efforts, which could include the five agencies we selected to review or other federal and nonfederal entities we did not review, but knew could potentially be sources of technical and financial assistance for utilities based on our prior work. Specifically, the federal agencies we did not review, but included in our questionnaire were: NOAA, the Department of Defense’s U.S. Army Corps of Engineers, and the Department of the Interior’s Bureau of Reclamation (Reclamation). To examine the second parts of the second and third objectives, the options experts identified for providing additional technical and financial assistance to utilities, we conducted semistructured interviews with the 10 climate change and disaster resilience experts. To develop the semistructured interview documents, we assessed the content of the 96 articles identified in our literature review to develop a list of actions that the federal government could take to make drinking water and wastewater infrastructure more resilient to the effects of climate change. The articles used to develop this list of actions were identified by searching resources such as Agricola, ProQuest’s Environmental Databases, Policyfile, Harvard’s Think Tank Search, and Scopus. We searched for both peer-reviewed articles and reports from nonprofits and think tanks published between January 2003 and September 2018 searching titles, abstracts and keywords for “water” in close proximity to “climate change,” “utilities,” and terms such as “project,” “program,” “policy,” or “recommendation.” We asked the 10 experts about the list of actions during our interviews (see table 2). We conducted semistructured interviews with the 10 selected experts and asked the experts to rate the effectiveness of the nine actions we provided for making drinking water and wastewater infrastructure more resilient to the impacts of climate change, describe the advantages and disadvantages of each action, and describe how the actions could be implemented. We also asked experts to rate the administrative feasibility and cost of the actions. Finally, we asked the experts if any additional actions should be added to our list. We then analyzed the results of our interviews to identify five options to provide technical assistance and developed a follow-up questionnaire. The questionnaire asked the 10 selected experts to rate the effectiveness of the five options for providing additional technical assistance, describe the advantages and disadvantages of each option, and describe how the options could be implemented (see table 3). We also asked experts to rate the overall effectiveness, administrative feasibility, and cost of the options. We also requested written responses from the 15 selected utilities on the 5 technical assistance options and the 4 financial assistance options identified in our interviews with experts. We conducted this performance audit from October 2017 to January 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Examples of the Most-Relevant Potential Climate Change Impacts and Their Potential Effects on Drinking Water Utilities by Region Table 4 corresponds with figure 2 in the report, which is an interactive figure and contains the text for drinking water utilities that is not accessible to readers of print copies of this report. As readers scroll over the water-drop icons in the figure, separate pop-up boxes appear describing specific regional impacts. Appendix III: Examples of the Most-Relevant Potential Climate Change Impacts and Their Potential Effects on Wastewater Utilities Table 5 corresponds with figure 2 in the report, which is an interactive figure and contains the text for wastewater utilities that is not accessible to readers of print copies of this report. As readers scroll over the water- drop icons in the figure, separate pop-up boxes appear describing specific regional impacts. Appendix IV: Technical Assistance Providers That Selected Drinking Water and Wastewater Utilities Used Table 6 provides additional information on the selected drinking water and wastewater utilities and the sources of technical assistance they used for climate resilience planning for fiscal years 2011 through 2018. Appendix V: Federal Programs That Provide Technical Assistance The following federal programs have the potential to help drinking water and wastewater utilities, in particular smaller utilities that do not have the resources to conduct climate risk assessments and plan for measures to help make their drinking water and wastewater infrastructure more resilient to climate change impacts. Several of the federal efforts we reviewed provide general assistance with planning and operating drinking water and wastewater infrastructure. Specifically: Environmental Protection Agency’s (EPA) Environmental Finance Centers. The Environmental Finance Centers provide targeted technical assistance to, and partner with states and the private sector to help manage the costs of environmental financing. Environmental Finance Centers can provide technical assistance for financing drinking water and wastewater infrastructure and its operations and maintenance. EPA’s Training and Technical Assistance for Small Systems Grants. EPA’s Training and Technical Assistance to Small Systems grants provide funding to nonprofit organizations to provide training and technical assistance to small public water systems, small wastewater systems, and private well owners, located in urban and rural communities in the U.S. and its territories. According to EPA officials, training and technical assistance to small systems facing drought, flooding, and other weather-related challenges is an eligible activity for the grants. Department of Agriculture’s (USDA) Rural Water and Wastewater Technical Assistance and Training Program. USDA’s Rural Water and Wastewater Technical Assistance and Training Program provides grants to nonprofits such as the National Rural Water Association and the Rural Community Assistance Partnership to provide training and technical assistance to small and rural utilities for operating, managing, and financing drinking water and wastewater infrastructure. USDA’s Rural Water and Wastewater Circuit Rider Program. USDA contracts with a qualified national organization, through its Circuit Rider program, to provide technical assistance to rural water and wastewater systems to provide technical assistance to rural utilities for operating, managing, and financing water and wastewater infrastructure. Circuit riders also provide critical assistance in disaster response and recovery. The circuit rider contract was awarded to the National Rural Water Association in fiscal year 2019. Other federal efforts help decision makers use climate information in existing planning processes. Specifically: USDA Climate Hubs. USDA established regional Climate Hubs to deliver science-based knowledge and practical information to farmers, ranchers, and forest landowners to support decision-making related to climate change. Department of the Interior’s (Interior) Landscape Conservation Cooperatives. Interior developed a network of collaborative Landscape Conservation Cooperatives composed of federal, state, local, and tribal governments; nongovernmental organizations; universities; and interested public and private organizations to, manage large landscapes such as national forests, grasslands, and wetlands. As part of this program, the groups develop and provide the science and technical expertise needed to apply climate data in natural resources decision-making. U.S. Geological Survey’s Climate Adaptation Science Centers. Climate Adaptation Science Centers partner with natural and cultural resource managers to provide science that helps fish, wildlife, ecosystems, and the communities they support adapt to climate change by, among other things, providing climate, water, and ecosystem information to decision makers. National Oceanic and Atmospheric Administration’s (NOAA) Regional Integrated Sciences and Assessments (RISA) Program. NOAA’s RISA program supports a network of 11 regional research teams that work with public and private decision makers to identify and provide specific climate information and models to identify risks and adaptation options to increase resilience to climate variability and change. One area of emphasis for the RISA teams is conducting research on climate and water management issues while engaging with a range of water management organizations, including some water utilities. National Center for Atmospheric Research (NCAR). NCAR carries out interdisciplinary research on adaptation to climate change by generating scenarios of projected climate change, developing scientific tools and methods for analyzing current and future vulnerability, and conducting integrated analyses of climate change impacts and adaptation. An important component of NCAR’s program is the integration of decision makers and users of climate information, including water utilities, into its research activities. NCAR provides the atmospheric research community in academia, government, and the private sector with the shared resources necessary to conduct their research. Appendix VI: Examples of Ongoing and Completed Drinking Water and Wastewater Capital Improvement Projects to Enhance Climate Resilience Table 7 presents examples of drinking water and wastewater capital improvement projects to enhance climate resilience, according to utility representatives, from fiscal years 2011 through 2018. Appendix VII: Types of Financial Assistance Used by Selected Drinking Water and Wastewater Utilities on Capital Improvement Projects to Enhance Climate Resilience Table 8 presents additional information on financial assistance used by utilities we reviewed for capital improvement projects to enhance their climate resilience for fiscal years 2011 through 2018. Appendix VIII: Comments from the Environmental Protection Agency Appendix IX: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Susan Iott (Assistant Director), Micah McMillan (Analyst-in-Charge), Jim Ashley, Mark Braza, Colleen Candrl, Caitlin Cusati, John Delicath, David Dornisch, Kathryn Godfrey, Holly Halifax, Karen Howard, Rob Letzler, Jon Melhus, Patricia Moye, Eve Nealon, Sam Portnow, Dan Royer, Kiki Theodoropoulos, Joe Thompson, Seyda Wentworth, and Melissa Wolf provided key contributions to this report.

Why GAO Did This Study Human health and well-being require clean and safe water, according to the Water Research Foundation. The Fourth National Climate Assessment states that the potential impacts of extreme weather events from climate change will vary in severity and type and can have a negative effect on drinking water and wastewater utilities. GAO's previous work on climate change and resilience to extreme weather and disasters has shown how the federal government can provide information and technical and financial assistance to promote and enhance climate resilience. In 2015, GAO reported that enhancing climate resilience means taking action to reduce potential future losses by planning and preparing for climate-related impacts, such as extreme rainfall. This report examines federal technical and financial assistance to utilities for enhancing climate resilience, and options experts identified for providing additional assistance, among other things. GAO reviewed relevant federal laws, regulations, and guidance from four federal agencies—EPA, FEMA, HUD, and USDA—and interviewed federal officials, representatives from 15 water utilities selected for diversity of size and geography, and 10 experts selected to represent different views. What GAO Found Four federal agencies—the Environmental Protection Agency (EPA), the Federal Emergency Management Agency (FEMA), and the Departments of Housing and Urban Development (HUD) and Agriculture (USDA)—provide technical and financial assistance (e.g., loans and grants), to drinking and wastewater utilities. Technical assistance. EPA provides technical assistance to drinking water and wastewater utilities to enhance their infrastructure's resilience to climate change. However, according to EPA officials, EPA's program is small and cannot assist utilities nationwide. All of the selected experts GAO interviewed stated that utilities need additional technical assistance on an ongoing basis to manage climate risks, and most experts said that organizing a network of existing technical assistance providers, including federal and state agencies, universities, and industry groups, would be needed to provide such assistance. Under a presidential policy directive, EPA is to work to enable efficient information exchanges among federal agencies and to help inform planning and operational decisions for water and wastewater infrastructure. By identifying existing technical assistance providers and engaging them in a network to help utilities incorporate climate resilience into their infrastructure projects on an ongoing basis, EPA would have better assurance that climate information was effectively exchanged among federal agencies and utilities. Financial assistance. Federal agencies have taken some actions to promote climate resilience when providing financial assistance for water infrastructure projects, but agencies do not consistently include the consideration of climate resilience when funding such projects. Most selected experts suggested that federal agencies should require that climate information be considered in the planning of water infrastructure projects as a condition of providing financial assistance. Moreover, representatives from several utilities said that such a requirement could be an effective and feasible way to help enhance utilities' climate resilience. A requirement would ensure that utilities consider climate resilience in planning for water infrastructure projects and potentially limit future fiscal exposures. For example, from fiscal years 2011 through 2018, the federal government provided at least $3.6 billion in disaster recovery financial assistance for drinking water and wastewater infrastructure related projects (see figure). What GAO Recommends GAO recommends that EPA identify technical assistance providers and engage them in a network to help water utilities incorporate climate resilience into infrastructure projects. Also, Congress should consider requiring that climate resilience be considered in planning for federally funded water infrastructure projects. EPA neither agreed nor disagreed. GAO believes the recommendation is still warranted.

gao_GAO-20-422

gao_GAO-20-422_0

Background BLS Consumer Price Indexes BLS currently produces a number of different price indexes to estimate price inflation (see table 1). In line with its strategic plan, BLS aims to make these estimates as accurate as possible, meaning that they reflect the average level of price inflation for a selected group of consumers. The accuracy of a price index can be assessed in multiple ways, such as the extent to which the index applies appropriate formulas to data that are complete and drawn from sufficiently large samples covering the relevant group of people. BLS bases its collection of these data on the population covered by the Consumer Price Index for All Urban Consumers (CPI-U). BLS then uses data collected for the CPI-U to produce three other price indexes. After introducing the CPI-U as its primary, or headline, index, BLS maintained a separate data collection for the CPI-W from 1978 through 1980 but found little difference between data for CPI-W and CPI- U. According to BLS, as a result of this and budgetary issues, BLS stopped collecting separate data for the CPI-W in 1981 and began using CPI-U data to derive the CPI-W. To create the CPI-U, BLS chooses a sample of outlets (e.g., stores or internet sites) at which the CPI-U population shops (see fig. 1 for more information on how BLS creates price indexes). BLS then collects price data at these outlets for goods and services the CPI-U population buys and uses the data to develop basic, or elementary, indexes for each good and service. BLS combines the elementary indexes into a single, aggregated index by applying a set of expenditure weights—factors that determine, for example, whether a change in the price of apples or mobile phone service has a larger effect on total inflation (see fig. 2). These expenditure weights reflect the proportion of spending consumers direct to each good or service. To develop expenditure weights, BLS directs the Census Bureau to gather data about the relative importance of each purchase within the target population’s “market basket” of consumer goods and services. The Census Bureau collects these data in the Consumer Expenditure Survey, a nationwide household survey conducted by BLS to determine how consumers spend their money that also contains demographic data about the households surveyed. BLS uses 2 years of Consumer Expenditure Survey data to calculate the expenditure weights, in part so the sample sizes are large enough to produce accurate weights. From data collected to produce the CPI-U, BLS derives two subpopulation indexes—indexes that focus on the spending patterns of a portion of the population of all urban consumers: the CPI-W and the CPI- E. To produce these subpopulation indexes, BLS adjusts the relative importance of price changes in each good and service through a process sometimes referred to as “reweighting,” meaning BLS develops alternate sets of expenditure weights that reflect the spending patterns of the subpopulation. For example, since medical care comprises more of the CPI-E subpopulation’s total expenditures (about 12 percent) than of the CPI-U population’s total expenditures (about 9 percent), the CPI-E gives more weight to medical care than the CPI-U. BLS also creates a “chained” index using the same data for the entire CPI-U population but changing the formula used to combine indexes for each good and service, known as elementary or basic indexes, into a single aggregated index. This formula captures how consumers shift spending among different types of goods and services as prices change (see text box). In contrast, the other indexes assume that consumers keep purchasing various categories of goods and services in the same proportions over a 2-year period regardless of price changes. What is a chained price index? A chained price index uses a formula that is believed by some economists to better approximate a cost-of-living index by more accurately accounting for changes in consumption patterns in response to relative price changes. They contend that such a formula reduces the potential for overstating inflation relative to the other indexes BLS produces, which assume consumers keep buying goods and services in the same proportions no matter their price. Like the other three indexes BLS produces, the Chained CPI-U reflects consumers’ ability to adapt to changing prices by choosing among closely related goods and services as prices change, for example purchasing a different type of apple because it is on sale. However, unlike the other three indexes, the Chained CPI-U further reflects consumers’ ability to choose among all available goods and services as prices change, such as taking a train to work instead of driving when the price of gasoline rises, and purchasing headphones to listen to music during the commute. We previously reported that, were federal retirement benefits to be indexed to the Chained CPI-U, SSA and other agencies would need to determine whether to base retirement COLAs on final data that may be outdated or preliminary data that may be inaccurate. This is because the data needed to use a superlative index formula only become available after a significant time lag. This lag delays issuance of final monthly estimates for the Chained CPI-U by up to 1 year. Additionally, the chair of a panel convened at the request of BLS to examine issues in measuring the cost of living cautioned that chained indexes may not accurately reflect the way people with varying incomes substitute goods and services. For example, retirees with lower incomes might not have the same ability as retirees with higher incomes to substitute other goods and services when the prices of needed medical care or prescription drugs rise. BLS receives input on its processes from several sources. For example, BLS receives advice and recommendations from several advisory committees that variously focus on technical issues and the needs of users of BLS statistics. BLS also periodically receives input on its price indexes through external commissions and panels. For example, in May 1995, the U.S. Senate created the Advisory Commission to Study the Consumer Price Index, commonly referred to as the “Boskin Commission,” after its chairman, Michael J. Boskin. In December 1996, the Boskin Commission released its final report identifying sources of bias in the production of CPIs that the commission concluded were causing the indexes to overstate inflation. BLS also receives input on its price indexes through public comment. For example, in May 2019, the Office of Management and Budget issued a request for public comments on the various price indexes produced by BLS and BEA. Social Security Retirement Benefits While there are a number of federal retirement benefit programs, Social Security is by far the largest provider of indexed retirement and disability benefits in the United States, paying out over $1,047 billion in retirement and disability benefits in 2019. Social Security was established in 1935 to provide for the general welfare of older Americans by, among other things, establishing a system of federal old-age benefits, including a retirement program. To determine a worker’s initial retirement benefit, Social Security indexes the worker’s earnings to an average wage index. According to SSA, this ensures that a worker’s future benefit reflects the general rise in the standard of living that occurred during his or her working lifetime. Since 1975, Social Security has also indexed retirement benefits after the initial benefit level has been set to a CPI. According to SSA, this ensures that benefits are not eroded by inflation over time. When SSA began indexing benefits, CPI-W was the only national CPI available, and SSA continues to use the CPI-W to determine COLAs. As we have previously reported, the Social Security program faces financial difficulties that, if not addressed, will affect its long-term stability. In April 2020, SSA projected that Social Security’s retirement program trust fund will be unable to pay full benefits in 2034. We have also reported that, according to projections by SSA and the Congressional Budget Office, use of an alternate index to determine COLAs would have less effect on Social Security’s long-range finances than some other options for addressing the program’s finances, such as changing the taxation of earnings or raising the retirement age. That said, we found that, according to SSA projections, using an alternate CPI to calculate COLAs would affect Social Security’s finances in different ways. Specifically, using the CPI-E would increase expected COLAs and thus program costs and using the Chained CPI-U would decrease expected COLAs and thus program costs, while using the CPI-U would result in little change to either. National Income and Product Accounts (National Accounts) Produced by BEA, the National Accounts are a set of statistics on U.S. production, income, consumption, investment, and saving. Among these are Gross Domestic Product, a measure of the goods, services, and structures produced across the economy, and the Personal Consumption Expenditures index, a measure of consumer inflation similar to CPIs, but constructed using different methods and data sources and covering different populations and transactions. Data collected by BEA to produce the National Accounts differ in a number of ways from those collected by BLS to produce CPIs. For example, while CPIs focus on the expenditures of households in urban areas, the National Accounts also include expenditures on institutional populations, such as individuals living in nursing homes. Further, while CPI expenditure data are based on the recollection of consumers, National Accounts expenditure data primarily reflect the records of the businesses that serve consumers. In other words, to collect data on the quantity of goods and services consumed, BLS surveys consumers about how much they bought, whereas BEA surveys companies about how much they sold. The National Accounts are produced primarily from data collected by federal government agencies. These data include both “statistical” data collected from federal statistical agencies, such as the Census Bureau, as well as “administrative” data collected by federal agencies as a byproduct of administering their programs. For example, BEA uses sample data generated by the Internal Revenue Service in processing tax returns to estimate corporate profits. BEA supplements these statistical and administrative data collected by federal agencies with data obtained from trade associations, businesses, international organizations, and other sources. BLS Faces Challenges Developing Consumer Price Indexes, but Has Made Limited Use of Data Collected by the Federal Government That May Help It Improve the Indexes’ Accuracy and Timeliness BLS Faces Challenges Related to the Accuracy and Timeliness of CPIs, Among Others BLS faces a number of challenges related to the accuracy and timeliness of CPIs, as well as challenges related to measuring inflation for older Americans. Some of these challenges may have implications for federal retirement benefit adjustments. Accuracy According to BLS officials and documentation, BLS is unsure if the data sources it uses to produce the CPI-U are adequate to produce accurate subpopulation estimates—specifically, the CPI-E and CPI-W. For the CPI-E, BLS has not evaluated the adequacy of the CPI-U data it uses to measure inflation for the 62-and-older subpopulation. Specifically, BLS has not evaluated the extent to which CPI-U data represent the outlets where members of this older subpopulation shop, the prices they pay, or the mix of goods and services they purchase. BLS considers the CPI-E an experimental index, in part, because of the relatively small sample size within the Consumer Expenditure Survey used to create the expenditure weights for this subpopulation, which account for the mix of goods and services the subpopulation purchases. According to BLS documentation, the expenditure weights for the CPI-U rely on about 65,000 household interviews, which are collected quarterly over 2 years. In contrast, the expenditure weights for subpopulation indexes use about one-third or less of that: 21,000 interviews for the CPI-E and 16,000 for the CPI-W. For the CPI-W, BLS has not evaluated the adequacy of using CPI-U data since 1980, but the relative sample size used to calculate the expenditure weights for the CPI-W subpopulation has been shrinking in part because of declining response rates and demographic shifts away from the occupations included in the CPI-W. For example, occupations in the CPI- W include blue-collar jobs such as clerical, sales, laborer, and construction jobs. BLS officials and documentation indicate that as a result of these demographic shifts and the subsequent shrinking sample size within the Consumer Expenditure Survey, the accuracy of the CPI-W expenditure weights may be deteriorating. A core element of BLS’s mission is to provide accurate products. Moreover, standards of internal control call for agencies to obtain relevant data from reliable internal and external sources to meet information requirements for meeting their objectives. For BLS, this could include obtaining relevant data from reliable sources for producing CPIs. BLS officials said they have not evaluated the adequacy of the existing data because it is costly to undertake a full evaluation, but there may be cost- efficient ways to do so. BLS also has not evaluated different methods to conduct a cost-efficient analysis. Without taking actions to understand available options for a cost-efficient solution, BLS lacks reasonable assurance that adjustments to Social Security and other retirement benefits are based on indexes that reflect what they are intended to reflect. Specifically, benefits could be subject to adjustment based on potentially inaccurate information. Most experts we interviewed identified potentially cost-efficient methods to evaluate the adequacy of existing data for subpopulation indexes. For example, five experts we interviewed, including some on BLS advisory groups, suggested that BLS may be able to use existing data to examine the adequacy of using Consumer Expenditure Survey data for the CPI-E. Specifically, one expert suggested that BLS could compare expenditure patterns for the older subpopulation in the Consumer Expenditure Survey to those in third-party data. Another expert added that the overall prices older Americans pay may not be significantly different than the prices the general population pays. For example, gas stations generally charge the same price to each customer regardless of age, so this expert said that it may not be worthwhile for BLS to collect separate price data for older Americans. Another expert indicated that, while it might not be possible to link expenditures and demographics (such as age) for all CPI categories using third-party data, it may be possible for certain categories such as groceries, which are a sizeable portion of the older population’s expenditures. Another suggested that to improve subpopulation indexes, BLS could shift resources from cost savings realized from other ongoing projects. BLS officials acknowledged some potentially cost-efficient methods could exist to evaluate the adequacy of existing data for subpopulation indexes. For example, they said that a recent change in survey methodology will enable them to connect demographic information with information on where people shop beginning in 2019. The ability to make this connection should allow them to determine whether certain subpopulations shop at the same or different outlets and could help them determine the adequacy of their outlet sample selection. According to agency officials, BLS advisory groups could weigh in on such issues, but BLS has not asked the advisory groups to do so nor do the advisory groups have any recent or ongoing research on indexes for subpopulations such as older Americans. BLS officials added that obtaining transaction and demographic data from credit card companies could help, but cautioned that companies may be unwilling to share these data. BLS is currently undertaking a project to improve how it estimates its subpopulation indexes, CPI-E and CPI-W, in part by examining changes to the formulas used to apply expenditure weights. As part of its justification for the project, BLS expressed concerns about the decrease in the relative sample size for the CPI-W population in the Consumer Expenditure Survey and reiterated the importance of the CPI-W in adjusting federal retirement benefits. This project is a step in the right direction but does not fully address the question of whether the CPI-U data are adequate to produce CPI-W and CPI-E. In 2009, BLS began another project to address measurement error in and households’ willingness to respond to the Consumer Expenditure Survey, which is primarily conducted to create expenditure weights for CPIs. According to agency documents, the survey faces increasing costs and declining response rates. One particular goal of the project is to reduce error due to underreporting. For example, BLS is currently testing replacing a paper record of household expenditures with an online form with the goal of more accurately capturing expenditures and maintaining response rates. The project is ongoing and BLS expects to implement changes in stages through and beyond 2022. According to agency officials, the project was not designed to address subpopulation indexes, but instead was designed to address broader issues with the accuracy of the Consumer Expenditure Survey. Timeliness and Relevance BLS also faces challenges regarding the timeliness and relevance of CPIs. In particular, most CPIs are published using expenditure data that can be up to 4 years old, and, in this dynamic economy, as expenditure data age, they become less relevant to present-day expenditure patterns. Most of BLS’s price indexes, including the CPI-U, CPI-E, and CPI-W, rely on 2 years of expenditure data and the data require additional time to be collected and processed for use, referred to as a lag. For example, the CPIs produced from January 2014 to December 2015 used expenditure data from 2011 through 2012. BLS officials said reducing the lag could enable more timely use of expenditure data for CPIs but would not be possible without a significant change to the use or design of the Consumer Expenditure Survey. Another of BLS’s indexes, the Chained CPI-U, aims to incorporate current-period expenditure data, which may be most relevant for current- period price changes, but as we reported in 2019, the data are subject to revision and BLS produces the final, revised Chained CPI-U with a 10 to 12 month delay. BLS officials told us they do not currently have timely enough expenditure data to produce the Chained CPI-U without this delay. We found in our 2019 report that if the Chained CPI-U were to be used to calculate Social Security or other federal retirement benefit COLAs, it could result in permanent differentials stemming from measurement error that would have a larger effect on people who receive benefits longest or have lower incomes. Other Challenges BLS also faces several other challenges measuring inflation for older Americans, several of which BLS is examining in the subpopulation project discussed above. Large purchases. BLS is examining how to treat large purchases that are acquired in one time period but used throughout many time periods, such as owner-occupied housing and durable goods. BLS’s current approach to owner-occupied housing is to calculate what it would cost to rent a similar home. In part, because many seniors own their homes, BLS is considering instead calculating how much it costs to own and occupy the home (e.g., by including mortgage interest payments but not the purchase price of the home). Definition of average. BLS is also examining whether a subpopulation index should represent the average expenditures of all households (as its CPIs currently do) or the expenditures of an average household. The current approach of representing the average expenditures of all households is simpler because the index can be constructed from information on average expenditures. The alternate approach of representing expenditures of an average household is more complicated because it gives each household equal weight, and requires first constructing a price index for each household, then an averaging of those indexes. According to BLS, the current approach tends to give more relative weight to the purchasing behavior of higher-income households, whereas the alternate approach may be more appropriate for a subpopulation index, such as the one used to adjust Social Security benefits. For example, taking the average of all expenditures tends to reflect the more expensive purchases typically made by higher-income households. In contrast, measuring the average household’s expenditures may better represent expenditures made by a particular subpopulation, such as recipients of federal benefits programs like Social Security. User needs. BLS is also examining how to define the subpopulation of interest to meet the needs of its users, such as the Social Security Administration. Specifically, CPI-E is based on households headed by someone age 62 or older and the CPI-W is based on households with particular occupations, and BLS is examining whether other definitions could meet user needs. For example, BLS said it plans to contact stakeholders to ask about whether expanding the CPI-W to include all labor force participants (thereby increasing sample size) would meet user needs. Quality change vs. inflation. A further challenge for all price indexes is determining what portion of the price change is due to changes in quality as opposed to inflation, according to eight of the nine experts we interviewed. BLS has several methods to adjust for quality changes. For example, if an older television is replaced with a new model with an increased price, BLS analysts collect information on the characteristics of those televisions and conduct an analysis to determine how much of the price change is due to a change in quality (e.g., the new television has additional features). The remainder of the price change is attributed to inflation. While accounting for quality change is a challenge for all price indexes, four of the nine experts we interviewed said it may be particularly difficult when measuring inflation for older populations. According to these experts, this is because older populations tend to consume more medical care goods and services, for which quality changes are particularly difficult to measure. BLS Has Taken Steps to Incorporate Alternative Data Sources into CPIs, but Has Made Limited Use of Other Data Currently Collected by the Federal Government Alternative data. To improve its price indexes, BLS is exploring the use of alternative data sources, such as “big data” obtained directly from companies, from third parties, or from the internet (see text box below). For example, BLS recently purchased a large private dataset to use in an experimental index for new vehicles. According to BLS, big data may lead to methodological improvements and cost savings in the CPIs. Notably, some big data may provide “real-time” expenditure data that could potentially be used to capture consumer behavior in response to relative price changes, thereby addressing substitution bias. According to agency officials and most experts we spoke with, big data may be promising but incorporating them in the CPIs requires additional considerations and adjustments to the processes BLS currently has in place. For example, the data may not be consistently available with the information needed to produce CPIs. Additionally, big data are not always free and some companies may be reluctant to share these data. What is “big data?” Big data encompass a number of very large data sets that can be structured or unstructured and have the potential to be mined for information. Web-scraped data and scanner data are two prominent types of big data relevant for consumer price indexes. Web- scraped data are price data collected on goods sold online. Scanner data include price and quantity data on sales of goods obtained by scanning bar codes for goods, such as at electronic points of sale in retail outlets. Advances in technology have allowed large amounts of data to be collected and stored easily and could be used in consumer price index construction. In addition to big data, BLS currently uses some administrative data collected by the federal government to improve inflation estimates for certain goods and services. For example, BLS obtains information from the Department of Energy on household consumption averages for electricity and piped gas service. It also uses administrative data from the Centers for Medicare & Medicaid Services about which facilities provide adult home care. According to BLS officials, they are unable to use some administrative data (e.g., certain federal tax data) because of current law. Other data collected by the federal government (National Accounts data). While BLS is exploring numerous alternative data sources, BLS has not fully explored the potential to update expenditure weights on a more frequent basis using supplementary data from the National Accounts in years when the most current biennial weights using Consumer Expenditure Survey data are not available. As discussed earlier, BLS typically requires 2 years of data from the Consumer Expenditure Survey to produce expenditure weights, which have a lag. In contrast, National Accounts data comprise administrative and statistical data representing the whole economy, many of which have a large sample size and are available on an annual basis. Standards of internal control call for agencies to obtain relevant data from reliable internal and external sources in a timely manner to meet information requirements for meeting their objectives. For BLS, this could include obtaining relevant data from reliable sources for producing CPIs. As part of its strategic plan, BLS maintains goals to improve the accuracy and timeliness of BLS data and to ensure relevance in an ever-changing economy. Without adequately exploring the potential of using National Accounts data to supplement Consumer Expenditure Survey data, BLS may be missing an opportunity to move closer towards those goals. Over time, expenditure survey data lose their accuracy and relevance to the present-day expenditure patterns of consumers, which can introduce bias in measures of inflation used to adjust federal retirement benefits. For example, the longer the time period between expenditure weight updates, the longer the delay to include new products in the expenditure patterns reflected in the CPIs. This delay could become increasingly important because of the rapid development in new technology, such as smart phones. Of the 15 publications we reviewed, six discussed ways to improve the CPI and four of these suggested more timely expenditure weight updates could make the CPIs more accurate and relevant. For example, a 2009 working paper by BLS staff found that more frequent weighting may offer better representation of current price change, as well as a closer approximation to a cost-of-living index. In particular, the authors simulated updating expenditure weights annually, which resulted in slower inflation increases that the authors posited are a closer approximation to a cost-of-living index. While these improvements may not be currently possible given the lag in Consumer Expenditure Survey data, the authors conclude that further examination of the weighting issue is a potentially fruitful avenue of research. The three other studies similarly indicated that more timely weight updates would result in more relevant CPIs, for example by better reflecting changes in consumer spending patterns. BLS officials acknowledged that updating the weights more frequently would make the index more relevant, though they did not believe using the Consumer Expenditure Survey to do so was practical in part because they said it would require additional costs to increase the sample size. In 2002, BLS increased the frequency of its weight updates from every 10 years to every 2 years, which they said was an improvement but required a sample size increase in Consumer Expenditure Survey. As previously described, the Consumer Expenditure Survey faces increasing costs and declining response rates and, according to agency officials, obtaining a large enough sample to update weights annually would require a 50 to 100 percent increase in sample size, for example, to avoid an increase in sampling error. Indeed, three studies we reviewed suggested that it can be challenging to obtain enough responses for household surveys such as the Consumer Expenditure Survey, indicating that alternate data sources may become more important. In contrast, BLS officials acknowledged that National Accounts data could provide useful supplementary information if the expenditure survey is not providing timely enough data. However, BLS officials said they have not explored using National Accounts data, in part because they have not examined the effects of altering the expenditure weights in about 10 years. BLS officials expressed concern that National Accounts data can be subject to revision. According to the Bureau of Economic Analysis (BEA), the revisions do not reflect errors but are driven by the incorporation of more complete source data. BLS officials also noted that some National Accounts data are adjusted by the CPI, so BLS would have to remove the CPI’s effect in order to use National Accounts data in the CPI. Moreover, the supplementary use of National Accounts data could also help address some of the concerns with measurement error in household surveys, according to some literature we reviewed. Specifically, National Accounts data could be used to address underreporting due to recall bias, the difficulty some survey respondents have recalling infrequent purchases, or underreporting of certain goods that may be seen as socially undesirable, such as tobacco and alcohol. For example, according to a recent Brookings Institution report, the National Accounts data used for the BEA’s Personal Consumption Expenditure index weights are mostly based on business surveys and administrative data and thereby avoid the reporting biases inherent in the Consumer Expenditure Survey. BLS’s Technical Advisory Committee recommended using administrative data to address such underreporting in fiscal year 2016, as did a National Academy of Sciences report in 2013. While BLS has taken steps toward increased use of administrative data, BLS has not fully implemented the Technical Advisory Committee recommendation as of March 2020. Selected Countries Use Various Strategies, Such As Obtaining Data from Alternative Sources and Bolstering Collaboration with Stakeholders, to Update Their Indexes for Retirement Benefits Use of Retiree-Specific and Chained Price Indexes for Adjusting National Pension Benefits Is Relatively Uncommon Our review of Organisation for Economic Co-operation and Development (OECD) countries’ national pension systems revealed that it is relatively uncommon to use a retiree-specific index (i.e., a CPI for the older subpopulation) for the purpose of adjusting national pension benefits. Of the 36 OECD countries, 27 have national pension programs in which indexation is based, at least in part, on prices after initial benefits have been set, similar to Social Security in the United States (see app. I). Most OECD countries use their primary measures of inflation to adjust national pension benefits, according to reports and documents about the retirement systems in these countries. Of the 27 countries using prices to adjust national pension benefits, we found evidence in 10 that the national statistical agency produces an index for the older subpopulation. Each of these 10 countries generally uses the same price information for the older subpopulation index as the main CPI but reweights the price information based on the expenditures for that subpopulation, rather than gathering new information that is unique to that group (see text box). A similar approach is used for the CPI-E in the United States. However, of these 10 countries, only four countries use the index for the older subpopulation to adjust their national pension benefits (Australia, Czech Republic, Hungary, and the Slovak Republic). The others produce the subpopulation index for research or other purposes, but do not use it for pension benefit adjustments. Agency officials in all three of our case study countries (Australia, New Zealand, and the United Kingdom) said they generally saw a value in having a primary index for macroeconomic purposes, such as inflation targeting, and a subpopulation index that could be used for other purposes, such as indexation of benefits. Methods for Validating Use of Existing CPI Data in Subpopulation CPIs In the three case study countries we selected for review, each national statistical agency relied upon different approaches to validate the use of existing data from the primary (main) CPI in the subpopulation CPI. Agency officials indicated that some of the methods for validating the use of existing CPI data for the subpopulation CPIs were cost efficient. Australia agency officials said they validated the use of existing data in the index for the older subpopulation in part by both researching whether pensioners pay different prices or shop at different outlets and cross-checking some data from industry sources. Officials said they expected that pensioners and the general population generally pay the same prices for most items and included different prices in the index for the older subpopulation for those items known to be discounted for pensioners. To get a better sense of the older population’s expenditures, they also increased the sample size of the expenditure survey from about 7,000 households to about 10,000 households to include more pensioners. New Zealand agency officials said they validated the use of existing data in part by using existing expenditure data to confirm that goods and services most important to the older subpopulation were adequately represented in the data. They also said they consider the coverage of the subpopulation group when determining the make-up of the CPI basket. Since older people may shop at different stores than the general population, New Zealand’s statistical agency also developed separate outlet weights for the older subpopulation, which more accurately reflect the different mix of outlets, or stores, frequented by this group, according to agency officials. Overall, officials said they found that using subpopulation-specific outlet data instead of general CPI outlet data had very little impact on the index for the older subpopulation. United Kingdom agency officials said they validated the use of existing data by organizing expenditure data from the household survey into categories that align with national expenditure data, which allowed them to generate bigger samples than exist in the household survey data. As a result of the larger sample, their statistical agency said they were able to achieve more precise estimates for the index for the older subpopulation. It is also relatively uncommon for a country to produce a chained index for the purpose of adjusting national pension benefits. Of those 27 OECD countries that are using price indexation, five of them produce a chained index (Australia, Canada, the United States, the United Kingdom, and New Zealand). However, none of the OECD countries use the chained index to adjust their national pension benefits. In our three case study countries, the statistical agencies used the chained index as an analytical tool to measure bias in the CPI or for comparative purposes. Officials we spoke with said that the delay required to produce a chained index made it impractical to use the index to adjust benefits. While some of the stakeholders in selected case study countries indicated it could be theoretically possible to create a chained CPI for the older subpopulation, we did not identify any countries with such an index during this review. Selected Countries Are Supplementing CPI Data with Other Government- Collected Data to Help Bolster Gaps in Information While government-collected data are often collected for reasons other than the production of the CPIs, the three selected case study countries are using government-collected data to help fill the gaps in data they collect expressly for the CPI (see table 2). According to agency officials in the three selected countries, use of this government-collected data improves accuracy of the CPIs and can be a relatively affordable way to supplement data collected for the CPI. National Accounts, key sources of government-collected data, are typically used for national summary measures like the Gross Domestic Product. However, all three of the selected countries are also using relevant consumption data from National Accounts to supplement their CPI data, which agency officials in Australia said is in-line with recommendations from the International Labour Organization (see text box). Australia, New Zealand, and the United Kingdom are all using their National Accounts data to supplement expenditure survey data in their CPIs, while New Zealand is also using another form of government- collected administrative data to improve its CPIs. International Guidance for Calculating CPIs and Subpopulation Indexes The International Labour Organization produces a manual that provides an overview of issues that national statistical offices can consider when making decisions on how to deal with the various problems in the compilation of Consumer Price Indexes. Researchers from many countries’ national statistical agencies, universities, and international organizations (such as the World Bank, International Monetary Fund, and Organisation for Economic Co-operation and Development) are involved in creating the manual. The manual also establishes international conventions, such as a suggestion that countries regularly evaluate the use of average wages as opposed to price indexes (and vice versa). Last published in 2004, an update to the manual is scheduled to be released in 2020. The upcoming revised manual is expected to elaborate on the use of National Accounts data and alternative data sources to develop expenditure weights. Australia. Australia’s statistical agency uses consumption data from their National Accounts to update the CPI expenditure weights more frequently than officials said was previously possible. Using this data has helped reduce substitution bias, meaning that the data better reflect changes in consumer purchases in response to price changes. Previously, Australia updated its expenditure weights every 6 years, when its household expenditure survey was released. In other words, the CPI was previously calculated assuming that consumers’ expenditure patterns did not change for 6 years. As a result, the CPI did not account for substitution patterns to different goods and services over significant periods of time, leading to bias in the CPI. In 2018, the Australian statistical agency incorporated National Accounts data in the CPI in those years when the expenditure survey was not conducted, allowing the expenditure weights to be updated annually to reflect what statistical agency officials described as more timely and relevant consumption patterns and to improve the accuracy of the data. According to Australian statistical agency officials, they did not have the budget to increase the frequency of their household expenditure survey, which they said is very costly. Instead, officials said they researched alternative ways that would allow for more frequent reweighting and settled on using the National Accounts data in between survey years to update the weights annually. This approach does not require a budget increase because the National Accounts data are already produced. Australian officials said more frequent weighting helped reduce substitution bias in their CPIs by about 0.2 percentage points per year, which can have a large impact on benefits over time. By incorporating consumption data from the National Accounts, Australian statistical agency officials said they can generate more timely and relevant CPI measures, including the subpopulation indexes. Australia’s index for the older subpopulation, called the Pensioner and Beneficiary Living Cost Index, also benefits from more frequent updates of the expenditure weights and subsequent reduction in substitution bias in the CPI, according to agency officials. Agency officials said that despite not having demographic information in the National Accounts, their methods have made use of this consumption data fit for purpose for the subpopulation indexes, and the subpopulation indexes are as methodologically sound as the primary CPI. New Zealand. New Zealand’s statistical agency also uses National Accounts data to estimate expenditure weights for insurance services, which are relatively difficult to measure in survey data, according to agency officials. Specifically, the expenditure weights for health and life insurance are based on data from the National Accounts. United Kingdom. In the United Kingdom, annual spending data from the National Accounts are the main source for CPI expenditure weights, as stakeholders noted that the National Accounts spending data are more precise and timely than their household expenditure survey. According to statistical agency officials, household expenditure data are ultimately obtained by organizing the United Kingdom’s expenditure survey data into categories that align with the National Accounts and scaling up these data to the National Accounts data. Officials said this method allows the United Kingdom’s statistical agency to achieve larger sample sizes, and thus smaller variances and more precision in estimates for subgroup indexes. United Kingdom officials said that their National Accounts estimates are more accurate and comprehensive than their household expenditure survey, which has a smaller sample size of nearly 6,000 households. Having more accurate expenditure data and weights leads to a more accurate and relevant primary index for pension benefits, as well as a more accurate subpopulation index, according to agency officials. The National Accounts data also help the United Kingdom adjust for any potential underreporting of particular goods in the household expenditure survey, such as alcohol, further increasing the accuracy and relevance of the dataset, according to officials. Collecting prices directly from the source is more accurate than relying on someone to recall how much they spent on items, according to one stakeholder. Government agencies from selected countries also produce other administrative data that can be useful in measuring the CPI. For example, New Zealand’s statistical agency partnered with the Ministry for Business, Innovation, and Employment to use its tenancy bond database, which covers approximately 85 percent of all rental housing units in the country. These data facilitated a new way to measure rent in their CPI. Moreover, this partnership enabled New Zealand’s statistical agency to create an index of rent prices monthly, instead of quarterly, which resulted in a more accurate and timely depiction of what people are spending on rent and a more accurate indexation of benefits overall. According to agency officials, the transition to these administrative data replaced the CPI survey of landlords, and in doing so it lowered respondent burden, increased the timeliness of the rental component of New Zealand’s CPI, and improved population coverage. In all of our case study countries, various data are used to measure housing prices (see text box). Housing and the Consumer Price Index Measuring the change in housing prices for CPI is widely acknowledged by experts to pose methodological and data challenges. In response, national statistical agencies have developed a variety of approaches to address the measurement of owner-occupied housing costs, both in the primary CPI and subpopulation indexes. Officials in the national statistical offices of the case study countries said that one of the factors underlying the approach to housing is whether the measure should reflect inflation in the economy overall or inflation as experienced by households. In Australia and New Zealand, the primary CPI includes price changes stemming from the purchase of a new home but not via mortgage interest payments (known as the acquisitions approach), while the subpopulation index excludes the purchase of a new home but includes mortgage interest (referred to as outlays or payment approach). In the United Kingdom, there are two versions of the primary CPI: one that uses “rental equivalence” (a calculation of what the owner would pay in rent for an equivalent house) and one that excludes owner-occupied housing costs. In addition, the United Kingdom’s subpopulation index uses a payments approach. Selected Countries Are Using Alternative Big Data Sources to Get More Data in a More Timely Way Officials in our selected case study countries said they are using alternative big data sources, such as web-scraping data and transactional (scanner) data to help them more accurately index their national pension benefits (see table 3). These officials said that these alternative data sources allow countries to obtain a higher volume of data and more accurate data to incorporate into their CPIs, subsequently making the indexation of benefits more accurate. Electronic price data obtained from a retailer, whether through the retailer’s website or through scanner data the retailer shares with the national statistical agency, reflects accurate and timely data on the price and quantity of goods and services sold. Electronic price data can be an improvement over data collected in household expenditure surveys, for example, as several experts and agency officials in one case study country noted that household expenditure surveys suffer from recall bias, resulting in less accurate spending data. The three selected countries are at different stages of incorporating scanner data into their CPI. Officials at the national statistical agencies in all three of our case-study countries stated that they are primarily focused on incorporating scanner data from grocery stores into their CPI. Using grocery store data is possible, in part, because these countries contain a relatively small number of stores that dominate grocery sales, according to agency officials, which is a difference from the United States. Australia. According to stakeholders, the Australian statistical agency developed a formula that incorporates a chained formula into a portion of the CPI using high- frequency scanner data from the country’s dominant grocery stores, which provides timely price and expenditure data on food items for their indexes. Integrating this type of high- frequency data is not easy, they said, since the traditional CPI formulas are not built to handle the volume of data that scanner data produce. However, in consultation with academics and statistical agencies from around the world, Australia was able to develop a chained formula that uses an innovative statistical method, known as a multilateral approach, to incorporate the scanner data. As a result, the portion of the CPI for which Australia has scanner data (about one-sixth of the CPI, comprised mostly of food and other grocery data) is based on a chained formula. Incorporating these data allows the country to include all of the products available in the datasets, rather than a small sample of products, leading to a more accurate calculation of food prices and a more accurate index overall, for both the general population and the older subpopulation, according to agency officials. Stakeholders in Australia noted that the international price statistics community has since reached a consensus that multilateral methods are the most effective way to capitalize the full amount of information provided in scanner data, and they said that the forthcoming update of the International Labour Organization’s CPI Manual is expected to recommend this method as well. New Zealand. New Zealand’s statistical agency is working towards incorporating more scanner data, primarily from its two large supermarket chains, in the production of the country’s CPIs, which will help achieve a more accurate index for both the general population and the older subpopulation, according to agency officials. New Zealand started using retail scanner data to supplement its expenditure data in its CPI in 2006, and in 2014 New Zealand incorporated direct measurement from scanner data for consumer electronics products into its CPI. Officials from the national statistical agency said they hope to expand their use of this type of big data in the near future. They have already received the data from supermarkets, whose goods account for roughly 20 percent of the goods and services in the CPI, but they have not yet integrated the data into their CPIs. Agency officials said they expect to integrate this in the next year. New Zealand’s statistical agency officials said they have a goal to obtain scanner data for other CPI components soon as well, such as fuel. United Kingdom. In the United Kingdom, agency officials said improvements in technologies have resulted in new alternative sources for price data that could be used in the compilation of their price indexes in the near future. The United Kingdom’s statistical agency is currently exploring both scanner data and online price data. The agency currently has several streams of research looking into the expanded use of alternative data, including research studying the feasibility of moving away from collecting prices manually towards using electronic means wherever feasible and efficient. The agency is now receiving web-scraped data from an online source that captures prices from online sales of goods like clothing. The United Kingdom’s statistical agency is also continuing to engage with retailers on receiving scanner data covering areas such as clothing and groceries, targeting some of the largest retailers from which the agency currently manually collects prices. These data sources may provide a more efficient way to capture the increase in online expenditures that has occurred over the last decade, and will likely continue to occur. These new data are initially being used for research work, but over time the web-scraped online prices and scanner data will be used when calculating primary inflation indexes, according to agency officials. The research done by the United Kingdom’s statistical agency into grocery store items has also enabled officials there to explore different methods of collecting web scraped prices in-house. The officials said this has led to wider benefits for the agency in general, with an increase in knowledge and experience that has contributed to the success of other big data projects. Selected Countries Collaborate and Consult with National Stakeholders and Experts When Implementing Changes to Their CPIs Our selected case-study countries use committees with stakeholders and advisory panels, including academic researchers with subject matter expertise, to implement innovative changes to their CPIs (see table 4). The statistical agencies in these three countries have shown a willingness to act on recommendations that came out of these collaborative efforts. These countries are also seeking input from the international statistical community, which country officials said has led to positive developments in their CPIs. Australia. Australia’s statistical agency has taken a variety of approaches to collaborate with external stakeholders, which agency officials said has led to positive changes to their CPIs, and thus indexation of benefits over the years. According to agency officials, Australia’s collaborative efforts include: conducting regular reviews and seeking stakeholder input every 6 years with the release of the expenditure survey; convening workshops with stakeholders including both academics and users (e.g., the agencies that distribute benefits); participating in international conferences to receive feedback on changes to the country’s CPI and subpopulation indexes; partnering with methodology experts in other agencies such as the Treasury and central bank, occasionally by obtaining staff on detail; and commissioning reports that research and review measures to strengthen the financial security of seniors. These reviews and associated collaborative efforts have helped the agency learn more about the issues it faces and have helped trigger changes that will improve the accuracy of the nation’s CPI, according to agency officials. For example, as discussed above, agency officials said that a 2011 CPI review revealed concern by the Reserve Bank of Australia and others that the infrequent reweighting was resulting in bias in the CPI that affected inflation targeting by the central bank, as well as benefit expenditures. This review helped spur innovations, such as including the incorporation of scanner data into the nation’s CPI, which delivered positive results with respect to more timely and relevant data being used to estimate inflation. Australia’s statistical agency officials said they sought extensive input from key governmental stakeholders, a number of academic experts, as well as international experts to research how to best incorporate scanner data into their CPI, which agency officials noted was necessary to facilitate the integration of high-frequency scanner data into the CPI. They also conducted numerous bilateral and multilateral consultations with key stakeholders in the government that use CPI data, including the Reserve Bank of Australia, the Treasury, Department of Finance, Department of Social Services, and State Treasuries. Australian statistical agency officials suggested that consulting with users of the data frequently was an important part of implementing changes to the measurement of the CPI and subpopulation indexes. New Zealand. New Zealand’s statistical agency has also used CPI advisory committees composed primarily of external stakeholders who make use of the agency’s CPIs. For example, in 2013 New Zealand’s statistical agency convened a committee to independently review the methods and practices used to compile the CPI and make recommendations, for example, about how additional indexes should be measured. The committee also incorporated public submissions on the scope and uses of the CPI, for example, from nongovernmental organizations and interest groups such as retiree advocacy groups. The committee then released a report recommending the creation of additional CPIs that are designed for microeconomic purposes, such as the indexation of retirement benefits, to better reflect changes in the purchasing power of the incomes of particular subgroups of the population, like the older subpopulation. The committee also recommended that New Zealand’s statistical agency review the sample size and collection methods of their expenditure survey to improve the reliability of expenditure estimates of the required population subgroups so that the estimates could eventually be of high enough quality to be published, which they subsequently were. According to officials, the committee’s report helped lead to the creation of New Zealand’s subpopulation indexes. Moreover, the committee recommended that the statistical agency try to use retail scanner data to measure price change and stated that the method aligns with international best practices. New Zealand’s statistical agency recognized these best practices and the international consensus that multilateral methods are the preferred way to incorporate big data. Indeed, it has started to use these methods in the rental prices data and it plans to continue to research implementing these methods further. United Kingdom. The United Kingdom has also developed advisory panels on consumer prices to provide independent advice to the National Statistician, which officials said has allowed the United Kingdom’s statistical agency to learn more about challenges with the nation’s CPIs and to find possible solutions. Similar to the United States, the United Kingdom has advisory groups on technical issues, as well as on the uses of price indexes. The reports published by various advisory groups have raised technical issues with the Retail Price Index (RPI), which is the United Kingdom’s longest running measure of inflation. These technical issues resulted in the RPI being higher than the CPI. Ultimately, agency officials said consultations and advisory panel input helped lead to the RPI being decertified as a national statistic (see text box). The United Kingdom’s statistical agency also hosted numerous meetings and a collaborative workshop about the conceptual foundations of its subpopulation indexes, which are currently being developed. According to agency officials, obtaining input from internal and external stakeholders has been critical to developing solutions to indexation challenges. The United Kingdom’s Experience Changing Price Index Used for Pension Adjustments Changing the index used for benefit adjustments can be difficult, as switching price indexes can involve tradeoffs. For example, public and private pension benefits in the United Kingdom have traditionally been indexed by the Retail Price Index (RPI), the oldest index in the United Kingdom. The United Kingdom recently switched indexation of certain government benefits, including pension benefits, from the RPI to the slower-growing CPI. This is expected to result in lower payouts from the government. In contrast, the government continued using the faster- growing RPI for some provisions, such as student loan interest rates, that resulted in higher payments to the government. Stakeholders suggested that having multiple measures of inflation can create incentives for the government to use different indexes for its own budgetary advantage, with pensioners receiving lower benefit adjustments and students facing relatively higher loan payment adjustments. The United Kingdom’s experience highlights that changing the index for benefits may result in advantages and disadvantages for different groups and thus may be politically difficult, according to agency officials. Conclusions Federal retirement programs like Social Security have relied upon a subpopulation price index to adjust benefits since automatic cost-of-living adjustments were first enacted almost 45 years ago. This index estimates changes in purchasing power for wage earners as opposed to changes in the standard of living or some other type of measurement. In recent years, numerous legislative proposals have been suggested to change this index from one that measures the purchasing power of wage earners to one that targets some different population, for example one solely focused on the elderly. Much of the debate over using a different index has centered on the ability (i.e., the accuracy) of the indexes to capture changes in the cost of living for a particular group in society. BLS is unsure whether the data sources it currently uses are adequate to produce accurate CPI-E and CPI-W subpopulation indexes on a timely basis, according to BLS officials and documentation. While the CPI-E is experimental and not used by federal programs, the CPI-W is used to adjust billions of dollars of Social Security and other federal retirement program benefits. It is therefore critical that the measurement be as accurate as possible. However, ensuring the measurement’s accuracy may require a reexamination of the underlying data used to produce the subpopulation indexes. BLS has not evaluated the adequacy of existing data because it is costly to undertake a full evaluation, according to agency officials. But experts we interviewed, including some on BLS advisory groups, indicate there may be cost-efficient ways to conduct such a review. Although the experiences of other countries may not be directly applicable, other countries have found ways to evaluate the use of existing data for their subpopulation indexes, and officials in all three of our case study countries expressed the view that some of these methods were cost efficient. Absent BLS evaluating the adequacy of the existing data it uses to produce its subpopulation indexes, BLS will continue to be uncertain if its subpopulation indexes are accurate and it may not learn of potential areas for improvement. In addition, BLS currently relies on the Consumer Expenditure Survey to produce expenditure weights that measure the mix of goods and services consumers purchase and, because of survey shortcomings and processing lags, the weights reflect spending patterns that can be up to 4 years out of date. Although BLS has taken other steps to improve the accuracy, timeliness, and relevance of data used in the CPIs, BLS has not fully explored the potential to update expenditure weights on a more frequent basis using annual data from the National Accounts, which are currently collected in part to measure Gross Domestic Product. While not specifically designed for use in CPIs, the National Accounts data may provide BLS an opportunity to supplement Consumer Expenditure Survey data in the intervening years. Moreover, some literature we reviewed indicated that the use of National Accounts data has the potential to mitigate measurement error in the Consumer Expenditure Survey, thereby increasing accuracy. Without adequately exploring the potential of such an option, BLS may be missing an opportunity to improve its CPIs. Recommendations for Executive Action We are making the following two recommendations to the Department of Labor: The Secretary of Labor should ensure that BLS explores cost-efficient ways to evaluate the data sources currently used to produce subpopulation indexes, such as by engaging more directly with other stakeholders or seeking input from its advisory groups and other knowledgeable entities about approaches to expand data collection in a cost-efficient manner. (Recommendation 1) The Secretary of Labor should ensure that BLS explores the use of already collected National Accounts data to produce more accurate, timely, and relevant CPIs. (Recommendation 2) Agency Comments and Our Evaluation We provided a draft of the report to the Department of Labor, the Social Security Administration, and the Department of State for their review and comment. We also sent an informational copy to the Bureau of Economic Analysis. The Department of Labor and the Social Security Administration provided technical comments, which we have incorporated where appropriate. In an email, the Department of State said it had no comments on the report. The Department of Labor also provided written comments, which are reproduced in appendix III and discussed below. In its written comments, the Department of Labor stated that BLS continually improves its measures according to a guiding principle to provide accurate, objective, relevant, timely, and accessible information. The Department of Labor agreed with the first recommendation to explore cost-efficient ways to evaluate the data sources currently used to produce subpopulation indexes and stated that it would continue to investigate improvements to subpopulation indexes. The Department of Labor disagreed with the second recommendation to explore the use of National Accounts data in the construction of its indexes, stating that the National Accounts data are not a replacement for Consumer Expenditure Survey data. While we agree that the National Accounts data are not a wholesale replacement for the Consumer Expenditure Survey data, we believe that it would be useful to examine National Accounts data as an augmenting, alternative source of data that could supplement or enrich the Consumer Expenditure Survey. Such an effort could potentially lead to more accurate, timely, and relevant CPIs. Although the Department of Labor stated that the Consumer Expenditure Survey is a continuous survey and that data are received quarterly, most CPIs still rely on expenditure weights based on Consumer Expenditure Survey data that are up to 4 years out-of-date. In addition, the Consumer Expenditure Survey faces increasing costs and declining response rates. The Department of Labor stated in its comments that it is exploring ways to accelerate the data collection and processing time and that it periodically investigates the frequency of updating expenditure weights. We commend the Department of Labor for considering these efforts, and we maintain that they could take further action to explore additional opportunities for improvement. For example, the Department of Labor could research the extent to which there are instances or categories for which the National Accounts data could be used to produce more up-to- date expenditure weights than the Consumer Expenditure Survey. As we noted in our report, Department of Labor officials told us they periodically examine National Accounts expenditure data to explore differences with the Consumer Expenditure Survey data, not to explore supplementary use of alternative data. While it cannot be ensured that every expenditure data point in the National Accounts will be of use for producing CPIs, we maintain that further exploring the National Accounts expenditure data as a complement to the Consumer Expenditure Survey data may provide opportunities for BLS to improve the accuracy, timeliness, and relevance of its CPIs. We are sending copies of this report to the Secretary of Labor, the Commissioner of Social Security, and the Secretary of State. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7215 or jeszeckc@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: National Pension Indexation Formulas in the 36 OECD Countries Appendix II: Additional Information about Selected Case Study Countries Appendix II: Additional Information about Selected Case Study Countries technical corrections as necessary. We note also that the fact that a legal feature was successful in one or more of the countries we visited, which may have significantly different cultures, histories, and legal systems than the United States, does not necessarily indicate that it would be successful in the United States. Appendix III: Comments from the Department of Labor Appendix IV: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments Charles A. Jeszeck, (202) 512-7215 or jeszeckc@gao.gov In addition to the contact named above, Michael Collins (Assistant Director), Laura Hoffrey (Analyst in Charge), Emilio Fonseca, Kathleen McQueeney, Tom Moscovitch, and Julie Miller made key contributions to this report. Also contributing to this report were Deborah Bland, Alicia Cackley, Charles Ford, Sarah Gilliland, Susan Irving, Kelsey Kreider, Sheila McCoy, Jessica Orr, Oliver Richard, Joseph Silvestri, Almeta Spencer, Curtia Taylor, Frank Todisco, Walter Vance, Adam Wendel, and Sirin Yaemsiri.

Why GAO Did This Study In the United States, federal retirement programs typically include cost-of-living adjustments based on a CPI that measures inflation for a subpopulation of workers. This includes Social Security, which provides benefits for more than 60 million older Americans, workers with disabilities, and their families. As the life expectancy of Americans continues to increase, more Americans will be subject to these adjustments, so it is critical for them to be accurate. GAO was asked to review U.S. and international efforts to measure the cost of living for older populations. This report examines (1) key issues that BLS faces in measuring the cost of living for older Americans; and (2) the experiences of other countries that developed alternate methods of adjusting retirement benefits. GAO reviewed pertinent literature; assessed BLS efforts to measure inflation; conducted case studies in three countries—Australia, New Zealand, and the U.K.—with a variety of CPIs, which GAO selected based on expert referral and document review; and interviewed agency officials and experts. What GAO Found The U.S. Bureau of Labor Statistics (BLS) faces accuracy, timeliness, and relevancy challenges developing consumer price indexes (CPI) for subpopulations of blue-collar workers and older Americans. For example, the CPI for these workers is used to adjust federal retirement benefits for inflation, including Social Security. BLS has not evaluated the extent to which its existing data are adequate to produce CPIs that reflect what these subpopulations pay, where they shop, and what they purchase. Officials cite budgetary reasons for not having done this, but there may be cost-efficient methods for evaluating the adequacy of these data. Without an evaluation, federal retirement benefits could be subject to adjustment based on potentially inaccurate information. Additionally, BLS has made limited use of certain data already collected by the federal government—such as National Accounts data on U.S. production and consumption—that could be used to increase the accuracy, timeliness, and relevancy of CPI calculations that reflect the mix of goods and services consumers purchase. Without adequately exploring the potential of using these data, BLS may be missing an opportunity to improve its CPIs. Reports about the retirement systems in the 36 Organisation for Economic Co-operation and Development countries indicate that most use their primary measures of inflation to adjust government retirement benefits. In addition, all three of GAO's case study countries (Australia, New Zealand, and the United Kingdom, or U.K.) have a variety of CPIs, including for subpopulations, and they filled information gaps in their CPIs with National Accounts and other data. For example, Australia and the U.K. use National Accounts data annually to update their calculations of the mix of goods and services consumers buy, thereby making the CPIs more relevant and accurate. All three countries also collaborated with stakeholders—such as other agencies—to implement changes, for example by gathering input on the design of subpopulation CPIs. What GAO Recommends GAO recommends that BLS explore cost-efficient ways to evaluate the data currently used to produce subpopulation indexes, and explore the use of National Accounts data to produce more accurate, timely, and relevant CPIs. BLS agreed with the first recommendation but disagreed with the other. GAO continues to believe both recommendations are warranted, as discussed in the report.

gao_GAO-19-572

gao_GAO-19-572_0

Background Administration of USDA’s Nutrition Education Programs USDA administers its nutrition education programs through multiple agencies in two mission areas—Food, Nutrition, and Consumer Services and Research, Education, and Economics (see fig. 1). Within the Food, Nutrition, and Consumer Services mission area, FNS oversees nutrition assistance programs with nutrition education components, such as SNAP, WIC, and child nutrition programs. For SNAP-Ed and WIC, the FNS national office develops program policies and guidance and works with the FNS regional offices to provide technical assistance to state agencies. The FNS regional offices also review SNAP-Ed and WIC state plans. The Team Nutrition initiative is administered by FNS national officials who also work on child nutrition programs. FNS staff develop Team Nutrition materials, training resources, and guidance and provide assistance to state agencies and local entities overseeing the child nutrition programs. Within the Research, Education and Economics mission area, NIFA national officials oversee EFNEP, in part by providing program guidance, reviewing grant recipient plans, and conducting some monitoring and oversight of local implementing entities. The NIFA national office, together with the FNS national office, administers the FINI program. Although NIFA has primary responsibility over the grant award process, FNS has been overseeing an independent evaluation of program efforts. Interventions for USDA’s nutrition education programs are provided through varied local entities and settings. For example, land-grant universities may provide SNAP-Ed and EFNEP interventions, while local health clinics may provide WIC interventions. USDA’s programs also provide nutrition education in varied settings, ranging from grocery stores to hospitals (see fig. 2). Sometimes multiple nutrition education programs operate in the same setting. For example, SNAP-Ed may provide classes for students while Team Nutrition may distribute teacher training materials and nutrition education curricula to the same school. Structures, Target Populations, and Types of Education Most of USDA’s nutrition education programs target interventions to low- income populations with varied characteristics, as shown in table 1, and the programs also differ in how nutrition education fits into their structures. For example, SNAP-Ed and EFNEP are primarily focused on providing nutrition education to participants, while Team Nutrition provides nutrition education to both child nutrition program implementers and participants. WIC provides benefits for food and referrals to health and other social services, as well as nutrition education, including breastfeeding promotion and support, to participants. FINI provides benefits for purchasing healthy foods and may include additional nutrition education programming. Programs also provide nutrition education through various intervention methods, ranging from direct education, such as cooking demonstrations, classes on healthy eating, and one-on-one counseling, to social media campaigns and efforts to change policies, systems, or environments. SNAP-Ed provides direct education through a variety of nutrition educators, although its interventions also may involve social marketing and policy, systems, and environmental changes (PSE). PSE is intended to shape policies, practices, and physical environments to support and improve nutrition education, physical activity habits, and obesity prevention efforts. In fiscal year 2018, approximately 76 percent of SNAP-Ed interventions included direct education, whereas 54 percent included PSE, according to USDA data. EFNEP primarily provides direct education through paraprofessionals, also known as peer educators. Paraprofessionals typically live locally in the community, which allows them to recruit and receive referrals for new participants. University and locally-based professional staff train and supervise the paraprofessionals. In addition, EFNEP has incorporated PSE interventions in recent years. For example, USDA provides PSE training for EFNEP program implementers, as one step toward adopting the PSE approach. WIC programs also provide direct education, such as counseling and group discussions, and, according to federal regulations, are allowed to use other intervention methods as long as they are easily understood by participants and bear a practical relationship to participant nutritional needs, household situations, and cultural preferences. For example, WIC programs may conduct demonstrations or grocery store tours to help consumers understand how to read nutrition labels or shop on a budget. Team Nutrition creates and disseminates web-based and hard-copy educational materials to child nutrition program implementers in part to educate child nutrition program participants. For example, Team Nutrition provides curricula, posters, tools, guides, recipes, and cookbooks for schools and child care sites. Team Nutrition also provides annual grants to enhance nutrition education intervention efforts in schools and child care settings, as well as training for program implementers through its partnership with the Institute of Child Nutrition. FINI supports healthy eating choices by incentivizing the purchase and consumption of fruits and vegetables. For example, some FINI programs provide vouchers redeemable for qualifying fruits and vegetables. Further, according to USDA officials, a FINI program may partner with another USDA nutrition education program, such as SNAP-Ed or EFNEP, to provide nutrition education. Other USDA Nutrition Education Efforts USDA agencies also provide nutrition education through other research and guidance directed at the general public: USDA’s Center for Nutrition Policy and Promotion (CNPP), within FNS, works with the U.S. Department of Health and Human Services to develop the Dietary Guidelines for Americans, dietary guidance linking scientific research to the nutrition needs of consumers. CNPP also takes the lead on consumer nutrition education, including MyPlate, which translates the Dietary Guidelines for Americans for consumers. USDA’s Economic Research Service (ERS) conducts research and issues publicly available reports related to promoting the purchase and consumption of healthy, economical foods. ERS also provides data relevant to the nutrition of U.S. households and communities. USDA’s Agricultural Research Service (ARS) serves as a repository for publicly available nutrition education information and data. ARS manages the website Nutrition.gov, the Historical Dietary Guidance Digital Collection, and the FoodData Central data system, which provides food nutrient data for consumers. USDA Collects Information on Nutrition Education Participation and Expenditures, but Faces Challenges Assessing Effectiveness for One of Its Largest Programs Information on Participation Includes Those Receiving Direct Education and Other Measures of Program Reach According to USDA data, 3.8 million and 436,000 people participated in direct education interventions for SNAP-Ed and EFNEP, respectively, in fiscal year 2018. Direct education participation in these two programs, which are focused primarily on nutrition education, has decreased in recent years. Between fiscal years 2010 and 2018, SNAP-Ed direct education participation declined by 33 percent and EFNEP declined by 28 percent. Program officials we spoke with noted some factors that may in part explain these trends. For example, USDA officials said direct education has been less of a focus in SNAP-Ed in recent years, as the department has encouraged programs to use policy, systems, and environmental change interventions and social marketing, in addition to the traditional direct education, following implementation of the Healthy Hunger-Free Kids Act of 2010. USDA officials said that all WIC recipients are offered nutrition education, and therefore they report that 6.9 million people were offered nutrition education through the program in fiscal year 2018. Although officials consider this to be the best proxy for WIC nutrition education participation, more than 5.2 million of these WIC recipients were infants or children ages 5 and under. In addition, WIC recipients do not need to participate in nutrition education to receive the program’s food benefits. As a result, USDA’s proxy overcounts the number of people who participated in WIC nutrition education. For Team Nutrition, USDA tracks the reach of its nutrition education using the volume of materials distributed. Between fiscal years 2012 and 2018, Team Nutrition distributed around 5.1 million of its hard-copy materials, such as curricula, technical assistance and training tools, and other materials, to child nutrition program implementers, including schools and day care providers. Further, from March 2014 through fiscal year 2018, there were about 11 million unique views of Team Nutrition materials hosted on USDA’s Team Nutrition website. Additionally, USDA is collecting participation data for FINI through the FINI National Evaluation. The evaluation is ongoing and FINI participation data will be available after it concludes, according to USDA officials. USDA Collects Annual National Expenditure Data for All Programs, but Detailed Data Are Limited USDA’s data show that nearly $907 million was expended on nutrition education programs in fiscal year 2017, the most recent year for which complete data are available, with $826 million expended on two programs—WIC and SNAP-Ed (see fig. 3). Specifically, states expended $422 million on WIC nutrition education and nearly $404 million on SNAP-Ed in that year. Further, grantees expended $51 million on EFNEP, $16 million on Team Nutrition, and $13 million on FINI in fiscal year 2017. USDA has total annual expenditure data at a national level for its nutrition education programs, but it does not have detailed information on how the funding is expended that can be routinely analyzed in its two programs with the largest expenditures—WIC and SNAP-Ed. Since both programs allow states to use various types of nutrition education interventions, information on spending by type of intervention may help USDA compare costs, and with additional information, potentially assess the cost effectiveness of various nutrition education interventions. For WIC, USDA collected detailed information on nutrition education spending at the local agency level in 2016 through a survey and analyzed the costs associated with different types of nutrition education interventions. In contrast, USDA collects information on SNAP-Ed local implementing agency expenditures in narrative annual reports that make it difficult to assess spending by type of nutrition education intervention. USDA Has Taken Steps to Evaluate the Effectiveness of its Programs, and Faces Ongoing Challenges Assessing SNAP-Ed Through studies and data collection, USDA has gathered some information on the effectiveness of its nutrition education interventions. For example, in 2018, USDA completed the WIC Nutrition Education Study, which assessed WIC nutrition education in both descriptive and evaluative ways (see text box). Additionally, USDA officials said a new study is underway looking at how the WIC nutritional risk assessment tailors the benefit package participants receive, including the nutrition education offered. USDA has also funded various grants and cooperative agreements that have evaluated WIC nutrition education to some extent, according to USDA officials. Findings from WIC Nutrition Education Study: Phase II Report This 2018 study was designed to address research questions about the impact of the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC)’s nutrition education on participant nutrition and physical activity behaviors, among other things, in six pilot sites. Although pilot sites’ nutrition education practices varied, the study did not find significant differences in reported exposure to nutrition education, such as the number of contacts with an educator or receipt of materials to reinforce nutrition messages during visits, or significant differences in participant outcome behaviors, among participants by site. For EFNEP, USDA regularly collects participant data to assess the effectiveness of EFNEP interventions nationwide, and several studies have also assessed the cost effectiveness of EFNEP. Specifically, EFNEP participants take standardized food and physical activity questionnaires and provide information on their dietary consumption in the past 24 hours before and after participating in an intervention, such as a class. EFNEP administrators use this information to measure participant behavior change and also report it to USDA through EFNEP’s data reporting system (see text box). USDA is then able to aggregate these data at the national level and use them to assess the effectiveness of EFNEP interventions nationwide. Further, several studies have evaluated EFNEP cost effectiveness, including one which used national data to estimate EFNEP cost effectiveness by state. Outcomes Reported on Expanded Food and Nutrition Education Program (EFNEP) Participant Questionnaires Questionnaires are administered when adult EFNEP participants enter the program and again when they exit the program to measure behavior change in core areas, including diet quality and physical activity, food safety, food resource management, and food security. For fiscal year 2018, the majority of participants reported improvements in diet quality (92 percent), food resource management (80 percent), food safety (79 percent), and physical activity (78 percent). Further, almost half of participants reported improvement in food security (47 percent). In addition to WIC and EFNEP, USDA is currently collecting information from its grantees on FINI effectiveness as part of its forthcoming FINI National Evaluation. An interim evaluation report found a positive, but modest, impact of FINI on monthly household fruit and vegetable expenditures, but no measurable impact on adults’ daily fruit and vegetable consumption. Previously, USDA assessed the effectiveness of the Healthy Incentives Pilot, which was a predecessor to FINI. The pilot tested the impact of making fruits and vegetables more affordable for SNAP participants and found that participants consumed almost one- quarter of a cup more fruits and vegetables per day than non- participants. USDA has also taken steps to support evaluation of the effectiveness of SNAP-Ed interventions both through its own research and the development of an evaluation framework. In 2012 and 2013, USDA reviewed selected SNAP-Ed interventions to identify potential models of effective SNAP-Ed interventions and impact evaluations. Specifically, USDA evaluated five interventions aimed at increasing fruit and vegetable consumption in preschool or elementary-age children, one intervention aimed at increasing fruit and vegetable consumption in low-income seniors, and one intervention aimed at increasing low-income women’s knowledge of healthy eating choices. Also in 2013, the FNS Western Regional Office began an effort to develop the SNAP-Ed Evaluation Framework (Framework), which was finalized in 2016 and fully adopted for national use. The Framework was developed as a way to evaluate program interventions, and with the intention of encouraging use of policy, systems, and environmental change interventions, according to USDA officials. All states are currently using the Framework to evaluate SNAP-Ed program interventions, according to USDA officials; however, because the Framework allows for myriad ways to measure outcomes, information reported by states on the effectiveness of SNAP-Ed interventions varies widely. Within states, SNAP-Ed implementing entities can select from 51 indicators and various outcome measures in the Framework to evaluate their interventions. Although USDA has identified 7 of the Framework’s 51 indicators as priority indicators, and encouraged states to use these, each indicator has multiple outcome measures and data collection methods associated with it (see text box). Therefore, even if the same indicators are selected to evaluate the effectiveness of different SNAP-Ed interventions, each state may select different outcome measures and data collection methods, and report different information on effectiveness. In our prior work, we found that agencies that seek to manage an excessive number of performance measures may risk creating a confusing excess of data that will obscure rather than clarify performance issues. Elements of a Selected Medium Term Change Indicator: Healthy Eating The medium term change indicator for healthy eating acts as a priority indicator among the 51 indicators included in the Supplemental Nutrition Assistance Program Education (SNAP-Ed) Evaluation Framework. SNAP-Ed programs may use this indicator to measure healthy eating behavioral changes reported by SNAP-Ed participants before and after participation in a series of direct nutrition education classes. Within this indicator, programs may select from various outcome measures and data collection tools: Programs may select from 13 outcome measures to assess the participants on this indicator. Some options include eating more than one kind of vegetable, drinking water, and using MyPlate to make food choices. To assess these outcome measures, programs may select from 11 surveys and other data collection tools compiled by the U.S. Department of Agriculture, such as a food behavior checklist for adults and a beverage and snack questionnaire for older youth. In addition, USDA receives information on states’ evaluations of effectiveness that is not easily analyzed nationwide. Although states report information on SNAP-Ed interventions to USDA in a data system, including information on participation, demographic characteristics of direct education participants, and types of education interventions, the data system is not structured to allow states to report information on intervention effectiveness, including cost effectiveness. Instead, USDA uses SNAP-Ed state plans and annual reports to collect information on state efforts to evaluate program effectiveness, among other things. However, in their plans and reports, states identify the Framework indicators they use and describe their evaluation efforts and outcomes in narrative form, limiting USDA’s ability to aggregate evaluation information across states or interventions, according to USDA officials. One local SNAP-Ed official said her state’s most recent annual report was approximately 60 pages long, highlighting the magnitude of the narrative information some states provide. While USDA officials acknowledged these challenges, they said a narrative report is used to accommodate the differences among SNAP-Ed programs. USDA officials said that because the Framework is still relatively new, they are working to determine both how to assist states’ efforts to use it to evaluate SNAP-Ed effectiveness and to ensure these evaluations provide USDA with useful information for assessing these programs. Further, USDA officials said they are currently in the process of determining future SNAP-Ed reporting protocols to improve program implementation and impact. Federal internal control standards state that agencies should use relevant, quality information from reliable sources to inform decision- making and evaluate performance in achieving key objectives. Without information that can be compared across states or easily aggregated or reviewed nationwide, USDA is unable to assess the effectiveness of interventions used across the country to determine whether SNAP-Ed is achieving program goals. USDA Lacks a Formal Coordination Mechanism and Does Not Fully Leverage Internal Expertise for Its Nutrition Education Efforts Coordination of Nutrition Education Efforts Is Limited USDA’s national office does not have a formal coordination mechanism for department-wide nutrition education efforts; however, the department has taken some steps to coordinate efforts related to nutrition. For example, USDA convened staff from various program offices in November 2017 for a two-day Intra-Departmental Nutrition Workgroup Meeting. The focus of the meeting was not specifically nutrition education, but included a discussion of current and potential USDA efforts to encourage healthy food choices for certain age groups. The department also has a few committees that address nutrition issues, including the Human Nutrition Coordinating Committee and the Interagency Committee on Human Nutrition Research. Although these committees do not focus on nutrition education, they convene USDA officials and other federal partners on a regular basis. Despite the lack of a focus on nutrition education in these meetings, USDA officials who participated said these opportunities were useful for sharing related information with staff from across the department. USDA has also taken some steps to coordinate efforts across nutrition education programs that have an intersection of target populations, though this has not consistently occurred at the federal level, according to USDA officials. For example, in recent years, WIC officials collaborated with Team Nutrition officials on the development of infant feeding and breastfeeding resources for use in child care settings to ensure consistent messaging. However, USDA officials reported that other programs with similar target populations have not coordinated. For example: USDA officials told us WIC and SNAP-Ed officials have limited interaction, although both programs serve low-income families with young children and coordination could help reinforce key messaging from each program. Several regional SNAP-Ed officials said that they had limited involvement with Team Nutrition, although both programs may serve students in schools and sharing resources could help maximize program impact. Both SNAP-Ed and EFNEP focus on providing nutrition education to similar populations and are delivered by land-grant universities, yet there is limited coordination between the two programs. Regional officials who work on SNAP-Ed reported limited familiarity with EFNEP and said they have learned about EFNEP efforts intermittently through state and local officials, rather than from the national office. Similarly, representatives of the two land-grant universities we spoke with who solely administer EFNEP had limited information regarding SNAP-Ed efforts, though they expressed interest in coordinating efforts to maximize both programs’ reach and avoid duplication of effort. In the absence of formal coordination mechanisms from USDA headquarters, other efforts have developed to help coordinate nutrition education programs nationwide, though USDA national office involvement is limited. Association of SNAP Nutrition Education Administrators: Representatives of SNAP-Ed state implementing agencies formed the Association because they lacked a mechanism to communicate with FNS national office staff or one another on topics related to nutrition education, according to a representative of this group. Officials from FNS’s national and regional offices formally participate in the group’s annual conferences and other activities, but this representative told us that members of the group would appreciate more opportunities to interact directly with these officials. SNAP-Ed Program Development Group: Land-grant universities established this separate SNAP-Ed-focused workgroup to strengthen SNAP-Ed programs and nutrition networks at the state, regional, and national levels, and identify linkages between SNAP-Ed and the land- grant university system’s broader outreach, education, and research mission. SNAP-Ed officials from FNS’s national office do not regularly participate in this group, yet the NIFA administrator of EFNEP sits on the group’s leadership committee. Food, Nutrition, and Consumer Services Nutrition Council: This group convenes national and regional staff in the Food, Nutrition, and Consumer Services mission area on nutrition-related topics and is currently led by regional officials, although the group was previously led by both national and regional officials. FNS officials told us the Nutrition Council has not regained momentum at the national office level since leadership transitioned to the regional office level, and one regional official with leadership responsibilities on the Council told us the group would benefit from more leadership support from FNS national office staff. State Nutrition Action Councils (SNACs): At the state level, SNACs are primarily comprised of state representatives from FNS programs and develop statewide cross-program nutrition education plans. FNS’s national office has supported SNACs as a model for coalescing state programs around nutrition education and obesity prevention efforts but has delegated leadership of the SNACs to the regions, who work directly with state agencies. USDA does not have a dedicated individual or entity with leadership responsibility for nutrition education, and program staff who work on nutrition education are currently focused on their individual programs, according to USDA officials. Although FNS has a senior nutrition advisor who supports national and regional officials who work on FNS programs, the advisor’s role does not encompass department-wide coordination on nutrition education. Further, program staff whose responsibilities include nutrition education serve the needs of their individual programs and lack formal communication channels with one another, according to USDA officials. Previously, from 1998 through 2008, USDA had a centralized Nutrition Services Staff that served as a formal coordinating entity for FNS and held cross-program nutrition education meetings, which were useful for information sharing, according to USDA officials. In 2008, this division, which had been comprised largely of nutritionists, was dissolved, with its staff with nutrition expertise largely dispersed to individual program offices. According to national and regional officials, in recent years, coordinating nutrition education has not been a priority for USDA, and there has been a loss of staff resources dedicated to nutrition education in the department overall. National and regional officials said it is hard to find time to coordinate across nutrition education efforts because they face competing priorities and increased workloads, at times because staff with nutrition education expertise have left employment with USDA and not been replaced. According to regional officials, a voluntary group of FNS national and regional officials who meet to discuss nutrition issues has experienced diminishing participation in recent years, in part due to these reasons. Regional officials and land-grant university officials said that more formal coordination mechanisms to provide leadership and promote cross- department coordination and information sharing on nutrition education could help increase efficiency, maximize the use of federal resources, and avoid potential duplication of effort. One regional official said she regularly reaches out to a colleague to obtain information on other FNS nutrition education programs, but a centralized tool could provide this information quicker and more efficiently. Another regional official said she compiled information on USDA nutrition education grant opportunities for states in her region, but it would be helpful if this information were centrally compiled by the national office. Regional officials and land- grant university officials we spoke with also said formal collaboration mechanisms, such as a document or tool with information on all of USDA’s nutrition education efforts, examples of best practices for coordination, or an annual meeting to encourage information sharing, would be useful. Federal internal control standards state that agencies should communicate quality information across reporting lines to enable personnel to perform key roles in achieving objectives, and management should set the tone at the top and throughout the agency to ensure priorities are understood by all stakeholders. In our prior work, we reported that effective coordination can help reduce overlap and duplication, and we found that sustained leadership is an essential element to developing collaborative working relationships. We also identified leading practices that federal agencies can use to enhance the effectiveness of their collaborative efforts, such as agreeing on roles and responsibilities and establishing policies and procedures to work across organizational boundaries. USDA has acknowledged the importance of nutrition education coordination for maximizing the reach and potential impact of federal nutrition education and nutrition assistance programs in some of its program regulations and guidance, and this emphasis is consistent with new federal requirements. For example, FNS’s SNAP-Ed plan guidance directs states to coordinate SNAP-Ed activities with other national, state, and local nutrition education, obesity prevention, and health promotion initiatives and interventions, such as WIC and EFNEP. In our 2004 review of USDA’s nutrition education efforts, we found that increased coordination, such as sharing curricula, lessons learned, and data collection tools across efforts, could help USDA’s nutrition education programs make more efficient and effective use of resources. Consistent with this focus, the Agriculture Improvement Act of 2018 (Farm Bill) requires USDA to submit an annual report to Congress that includes an evaluation of the level of coordination between SNAP-Ed, EFNEP, and other USDA nutrition education programs. USDA Has Not Fully Leveraged Its Expertise for Nutrition Education Some USDA nutrition experts are in agencies disconnected from the nutrition education programs (see fig. 4), yet these agencies play a significant role in developing and compiling dietary guidance, research, and other information related to nutrition education (see table 2). Despite their role in developing and compiling research and information related to nutrition education, consultation with these experts by the program offices is limited, according to USDA officials, possibly because they are located in separate agencies. For example, Although CNPP leads a cross-cutting committee that reviews nutrition education materials developed by USDA program staff to ensure materials are consistent with the Dietary Guidelines for Americans,CNPP officials noted they have been infrequently consulted by program officials while materials are under development or activities are being implemented. This may have been in part related to organizational structure, as until recently CNPP and FNS were separate agencies that individually reported to the Office of the Under Secretary for Food, Nutrition, and Consumer Services, according to USDA officials. Some nutrition education program staff also told us they currently use the core nutrition messages on USDA’s website when developing nutrition education materials—messages that CNPP officials noted were developed in 2010 and have not been updated to reflect the latest edition of the Dietary Guidelines for Americans. This approach may lead to inefficiencies in the development of nutrition education materials. Although ERS conducts nutrition research, nutrition education program officials were not always aware of or using ERS resources, possibly because most of the programs reside in a different USDA mission area. A prior working group attempted to bridge the organizational divide between ERS and some of USDA’s other agencies and offices that work on nutrition education by assisting efforts to share information, but the group has since dissolved. Currently, some national and regional officials we spoke with who work on nutrition education programs had limited awareness of ERS’s nutrition education research. For example, some program officials in the national office were unsure whether ERS did work related to nutrition education and learned of ERS research through automated email updates. Further, one regional official learned of ERS data on food insecurity, which can help states meet federal requirements for targeting nutrition education services to local areas based on their level of need, through a meeting with an outside agency. Nutrition education program officials were also generally unaware of ARS’s efforts related to nutrition education. Specifically, USDA nutrition education program officials we spoke to said they had little direct contact with ARS officials and were generally unaware of ARS efforts related to nutrition education. Further, regional officials who work on SNAP-Ed had not used or distributed ARS resources to state officials and also seemed generally unaware of ARS’s nutrition education efforts. USDA lacks a mechanism for systematically integrating its internal nutrition expertise into its nutrition education programs, which may inhibit the effectiveness of the department’s efforts. Federal internal control standards state that agencies should use quality information from reliable internal sources, among others, to inform decision-making. Further, in our prior work, we found that identifying and addressing needs by leveraging resources is a leading practice for collaboration. Nutrition education program officials are missing opportunities to benefit from relevant expertise within USDA but outside their program offices. Failing to leverage its own internal expertise hinders USDA’s development of nutrition education materials that are informed by the latest nutrition guidance and research. Conclusions Poor nutrition contributes to costly chronic diseases that are among the leading causes of death for Americans, and USDA’s nutrition education programs and related efforts strive to educate Americans on nutrition and improve their dietary choices. Because USDA’s nutrition education programs are primarily targeted to low-income adults and children, who may receive federally-funded nutrition assistance benefits, these programs also have the potential to improve the likelihood that recipients will spend those benefits to obtain foods that have a positive impact on their health. However, in order to reach these goals, USDA needs to ensure that its programs are effectively educating participants to maximize the impact of the federal investment in nutrition education. Although USDA has some information on the effectiveness of its nutrition programs, without improvements to how USDA gathers information on the effectiveness of SNAP-Ed interventions nationwide, USDA will be unable to ensure one of its largest investments in nutrition education is meeting its goals. The 2018 Farm Bill included a requirement for USDA to begin reporting annually on the level of coordination between its nutrition education programs, and USDA has acknowledged the importance of coordination and information sharing to maximize nutrition education programs’ impacts. However, the department currently lacks a formal mechanism to ensure this occurs. As a result, USDA risks missing opportunities to increase efficiency, maximize the use of federal resources, and avoid potential duplication of effort. In addition, without coordination between nutrition education program officials and others with nutrition expertise in the department, programs will develop nutrition education materials that fail to fully leverage the latest nutrition guidance and research, possibly missing opportunities to effectively influence the dietary choices of their target populations in the process. Recommendations for Executive Action We are making the following three recommendations to USDA: 1. The Administrator of FNS should improve how FNS gathers information on the effectiveness of SNAP-Ed interventions, in order to ensure that these interventions are meeting program goals. (Recommendation 1) 2. The Secretary of Agriculture should direct the Under Secretaries for Food, Nutrition, and Consumer Services and for Research, Education, and Economics to develop a formal mechanism, such as a designated individual or group of individuals, for providing cross-department leadership for USDA’s nutrition education efforts and facilitating cross- program information sharing. (Recommendation 2) 3. The Secretary of Agriculture should direct the Under Secretaries for Food, Nutrition, and Consumer Services and for Research, Education, and Economics to identify and implement mechanisms to fully leverage the department’s nutrition expertise for its nutrition education efforts. (Recommendation 3) Agency Comments We provided a draft of this report to USDA for review and comment. In its comments, reproduced in appendix II, USDA generally agreed with our recommendations. USDA also noted that FNS has efforts underway to comply with the 2018 Farm Bill requirement that the department report annually on the level of coordination between its nutrition education programs. USDA also provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the Secretary of Agriculture, congressional committees, and other interested parties. In addition, this report will be available at no charge on the GAO website at https://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-7215 or larink@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix III. Appendix I: Objectives, Scope, and Methodology Our report examines the extent to which the U.S. Department of Agriculture (USDA) (1) has information on participation, expenditures, and effectiveness for its nutrition education programs; and (2) coordinates its nutrition education efforts and leverages internal nutrition expertise for these efforts. The scope of our review includes five federal programs that provide nutrition education: Supplemental Nutrition Assistance Program Education (SNAP-Ed), the Expanded Food and Nutrition Education Program (EFNEP), the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), Team Nutrition, and the Food Insecurity Nutrition Incentive (FINI) Grant Program. Among USDA programs that provide nutrition education, four of these received the greatest amount of federal funding for nutrition education in fiscal year 2018—WIC, SNAP-Ed, EFNEP, and Team Nutrition. In addition, we included FINI because it is a grant program in which nutrition education can be a component, and the program’s goal is to incentivize healthy eating. We also reviewed USDA efforts that provide nutrition education through nutrition-related research and guidance directed at the general public. In addition to the methods discussed below, to address both of our research objectives, we reviewed relevant federal laws, regulations, and guidance, as well as our prior work on USDA nutrition education efforts and leading practices for collaboration. We interviewed officials from relevant USDA agencies, including the Food and Nutrition Service (FNS) and the National Institute of Food and Agriculture (NIFA), which oversee the nutrition education programs described in this report. We also interviewed officials from other USDA agencies overseeing nutrition- related research and guidance, including the Agricultural Research Service, the Center for Nutrition Policy and Promotion, and the Economic Research Service. Additionally, we interviewed officials from the seven FNS regional offices, including officials who work on SNAP-Ed and the Child and Adult Care Food Program. We also interviewed representatives of selected organizations knowledgeable about USDA’s nutrition education efforts. We assessed USDA’s efforts to collect information on its nutrition education programs, coordinate its nutrition education efforts, and leverage internal nutrition expertise against GAO’s standards for internal controls in the federal government. Participation Data To address the first objective, we analyzed USDA data on nutrition education participation. Two of the nutrition education programs, SNAP- Ed and EFNEP, collect data on direct education participation. We analyzed SNAP-Ed total direct education participation data for fiscal years 2010 through 2018 collected through SNAP-Ed’s data reporting system, the Education and Administrative Reporting System (EARS). We analyzed EFNEP total direct education participation data for fiscal years 2010 through 2018. These data are reported through the Web-based Nutrition Education, Evaluation and Reporting System (WebNEERS), an integrated data collection system, sponsored by NIFA, and used at the county, state, and federal levels. To assess the reliability of the SNAP-Ed and EFNEP participation data, we interviewed FNS and NIFA officials and reviewed relevant documentation. We determined that these data were sufficiently reliable for the purpose of reporting the number of direct education participants in SNAP-Ed and EFNEP. We also reviewed available USDA data on the number of people reached by nutrition education efforts in SNAP-Ed and EFNEP other than through direct participation. SNAP-Ed collects information on the number of people reached by nutrition education efforts that are not direct education, such as policy, systems, and environmental change interventions and social marketing. However, states face challenges with tracking individuals reached by these education interventions, and these data are likely to include duplicate records of individuals, according to USDA officials. Therefore, we concluded that these data were not sufficiently reliable for the purpose of reporting the number of people indirectly reached by SNAP-Ed. EFNEP also collects information on indirect education reach. This information tracks other family members of adults who participated in direct education who therefore may also benefit from the information shared, according to USDA officials. To assess the reliability of these data, we interviewed NIFA officials and reviewed relevant documentation. We determined that these data were sufficiently reliable for the purpose of describing the number of people indirectly reached by EFNEP. Because USDA officials consider the total number of WIC participants to be the best proxy for WIC nutrition education participation, as all WIC participants are offered nutrition education, we analyzed WIC total participation, and participation by women, infants, and children, for fiscal years 2010 through 2018. These data are reported on the FNS- 798/798A Financial Management and Participation Report form, which contains programmatic and financial data reported by state agencies, Indian Tribal Organizations, and U.S. territories through the Food Programs Reporting System (FPRS). To assess the reliability of these data, we interviewed FNS officials and reviewed relevant documentation. We determined that these data were sufficiently reliable for reporting the number of WIC participants offered WIC nutrition education. Data were unavailable on participation for Team Nutrition—a program which provides training and technical assistance to child nutrition program operators, and creates and disseminates materials for child nutrition program participants. As a proxy measure for program reach, we analyzed data on nutrition education materials disseminated to participants and the online views and downloads of nutrition education materials. To assess the reliability of these data, we interviewed FNS officials and reviewed relevant documentation. We determined that these data were sufficiently reliable for the purpose of reporting the number of Team Nutrition materials disseminated. Expenditure Data To address the first objective, we also analyzed WIC, SNAP-Ed, EFNEP, Team Nutrition, and FINI total nutrition education expenditure data for fiscal year 2017, the most complete data available as of April 2019. Like WIC participation data, WIC expenditure data are reported on the FNS- 798/798A Financial Management and Participation Report form through FPRS. Federal SNAP-Ed and Team Nutrition expenditure data are reported on the SF-425 form, which state agencies submit quarterly, also through FPRS. USDA tracks nationwide expenditures for EFNEP and FINI through NIFA’s payment system, Automated Standard Application for Payments; grants management system, Cooperative Research, Education, and Extension Management; and financial management system, Financial Management Modernization Initiative. To assess the reliability of these data, we interviewed officials from FNS and NIFA and reviewed relevant documentation. We determined that these data were sufficiently reliable for the purpose of reporting nationwide expenditures for these five programs. Program Evaluations and Additional Program Data To determine what information USDA has on the effectiveness of its nutrition education programs, we reviewed relevant program evaluations from USDA issued within the last 10 years. We selected these evaluations based on information we obtained from USDA and other knowledgeable officials through interviews and relevant documents. To provide additional context on program operations for SNAP-Ed and EFNEP, we reviewed various program data. We analyzed the EARS data on the total number of SNAP-Ed implementing agencies, including the number of land-grant universities that were implementing agencies, and the types of education provided by SNAP-Ed programs in fiscal year 2018. For EFNEP, we reviewed data on participant outcomes, reported through WebNEERS, for fiscal year 2018. To assess the reliability of these data, we interviewed officials from FNS and NIFA and reviewed relevant documentation. We determined that these data were sufficiently reliable for the purpose of our reporting objectives. Interviews with Land-Grant University Representatives To gain the perspective of officials involved in the implementation of nutrition education efforts, we interviewed representatives of four land- grant universities. Land-grant universities are the sole provider of EFNEP and one of the main providers of SNAP-Ed. We judgmentally selected a non-generalizable sample of four land-grant universities based on various criteria, including the recommendations of knowledgeable officials, geographic dispersion, and other factors, such as the percentage of the university’s state population in poverty. Two of the universities we selected solely administer EFNEP and two administer both EFNEP and SNAP-Ed. We gathered information from these land-grant university representatives on how they provide nutrition education through their programs and the extent to which they coordinate with other SNAP-Ed programs in their county and state, as well as with other USDA nutrition education programs. We also gathered information on support they receive from the USDA national office for coordination, if any; their perspectives on challenges USDA faces to coordinating nutrition education across its programs, if any; and their views on opportunities for USDA to improve coordination across nutrition education programs. Information collected from the land-grant university representatives cannot be generalized to all land-grant universities nationwide. We conducted this performance audit from December 2018 to July 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Comments from the U.S. Department of Agriculture Appendix III: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Rachel Frisk (Assistant Director), Kristen Jones (Analyst-in-Charge), and Sara Rizik made key contributions to this report. Also contributing to this report were Monika Gomez, Stacy Ouellette, Almeta Spencer, Rachel Stoiko, Curtia Taylor, Walter Vance, Sarah Veale, and Adam Wendel.

Why GAO Did This Study The Centers for Disease Control and Prevention reports that many Americans' diets lack adequate sources of good nutrition and that this contributes to costly chronic health conditions. USDA funds and administers a variety of nutrition education efforts, which aim to help educate Americans on nutrition and improve their dietary choices. GAO was asked to review these efforts. This report examines the extent to which USDA (1) has information on participation, expenditures, and effectiveness for its nutrition education programs; and (2) coordinates its nutrition education efforts and leverages internal nutrition expertise for these efforts. GAO reviewed relevant federal laws, regulations, guidance, and GAO's prior work on nutrition education and leading practices for collaboration; analyzed USDA data on nutrition education participation in fiscal year 2018 and expenditures in fiscal year 2017, the most recent year with complete data available; and reviewed program evaluations and available outcome data for fiscal year 2018. GAO also interviewed USDA officials and representatives of relevant organizations. What GAO Found The U.S. Department of Agriculture (USDA) administers five key programs that provide nutrition education and has information on participation, expenditures, and effectiveness for most of these programs. USDA tracks the number of participants in direct education, such as classes and counseling, as well as other measures of program reach. For example, Supplemental Nutrition Assistance Program Education (SNAP-Ed), one of USDA's largest nutrition education programs, served 3.8 million participants through direct education in fiscal year 2018. USDA also collects nationwide expenditure data for all of its nutrition education programs, which totaled nearly $907 million in fiscal year 2017—the most recent year with complete data available. In addition, USDA collects some information on the effectiveness of most of its nutrition education programs; yet information USDA collects from states on SNAP-Ed effectiveness cannot be easily aggregated or reviewed. States provide this information in narrative reports, which hinders USDA's ability to assess the effectiveness of interventions used across the country and determine whether SNAP-Ed is achieving its goals. USDA does not have a formal coordination mechanism for its nutrition education efforts and does not fully leverage the department's nutrition expertise. According to USDA officials, coordinating nutrition education efforts has not been a priority in recent years, and the department does not have a dedicated individual or entity with leadership responsibility for nutrition education. This has resulted in limited coordination across USDA's nutrition education programs, including programs with similar target populations. GAO previously reported that effective coordination can help reduce overlap and duplication. In its absence, USDA's nutrition education programs are missing opportunities to share information and avoid duplicating efforts. Further, some USDA nutrition experts are not located in agencies or offices overseeing the nutrition education programs, and possibly because of this, program staff consult these experts on a limited basis, if at all. Failing to leverage its internal expertise hinders USDA's development of nutrition education materials that are informed by the latest nutrition guidance and research and may reduce the effectiveness of these efforts. What GAO Recommends GAO is making three recommendations to USDA, including that USDA improve how it gathers information on SNAP-Ed effectiveness, develop a formal mechanism for coordinating nutrition education across the department, and take steps to fully leverage the department's nutrition expertise for its nutrition education efforts. USDA generally agreed with GAO's recommendations.

gao_GAO-19-281

gao_GAO-19-281_0

Background In 2017, three major hurricanes made landfall in the United States and historic wildfires struck California. According to FEMA, the 2017 hurricanes and wildfires collectively affected 47 million people—nearly 15 percent of the nation’s population. See figure 1 for a timeline of these major disasters. Overview of Federal Disaster Response and Recovery When disasters hit, state and local entities are typically responsible for disaster response efforts. The Stafford Act establishes a process by which the Governor of the affected state or the Chief Executive of an affected Indian tribal government may request a presidential major disaster declaration to obtain federal assistance. According to the DHS National Response Framework—a guide to how the federal government, states and localities, and other public and private sector institutions should respond to disasters and emergencies—the Secretary of Homeland Security is responsible for ensuring that federal preparedness actions are coordinated to prevent gaps in the federal government’s efforts to respond to all major disasters, among other emergencies. The framework also designates FEMA as the lead agency to coordinate the federal disaster response efforts across 30 federal agencies. The National Response Framework identifies 14 emergency support functions that serve as the federal government’s primary coordinating structure for building, sustaining, and delivering disaster response efforts across more than 30 federal agencies. Each function defines specific mission areas—such as communication, transportation, and energy—and designates a federal department or agency as the coordinating agency. For example, provision of assets and services related to public works and engineering, such as temporary roofing or power, are coordinated by USACE within DOD. See Appendix II for more information about emergency support function responsibilities across the federal government. FEMA’s Response Directorate coordinates disaster response efforts through mission assignments—work orders that it issues to other federal agencies to direct them to utilize their authorities and the resources granted to them under federal law in support of direct assistance to state, local, tribal, and territorial governments. Mission assignments are authorized by the Stafford Act, and agencies may fulfill these assignments through federal contracts. FEMA made 1,515 mission assignments for the 2017 hurricanes and California wildfires, and total obligations for these mission assignments were more than $7.8 billion as of January 2018, according to FEMA. See figure 2 for a depiction of the mission assignment process under a notional scenario of removing derelict marine vessels—boats and ships damaged during a hurricane and that are determined to be inoperable. The National Response Framework states that when an Emergency Support Function is activated in response to an incident, the primary agency for that emergency support function is responsible for executing contracts and procuring goods and services as needed, among other things. For example, DOD and USACE are the coordinators for Emergency Support Function 3—public works and engineering—and as part of this role, these agencies are responsible for emergency contracting support for lifesaving and life-sustaining services. As such, during the 2017 disasters, USACE obligated funds on contracts in support of its assigned mission of public works and engineering by restoring the electrical grid in Puerto Rico following Hurricane Maria and removing debris following the California wildfires. FEMA’s Contracting Workforce In its role as the lead coordinator of federal disaster response efforts across federal agencies, FEMA’s contracting workforce plays a key role in post-disaster contracts. FEMA’s contracting efforts are supported by its contracting workforce within FEMA’s Office of the Chief Procurement Officer (OCPO). In our prior work, we found that FEMA’s contracting workforce had grown significantly since Hurricane Katrina, but the agency struggled with attrition at times. While the majority of FEMA’s contracting workforce is located in headquarters, contracting officers are also located in each of FEMA’s 10 regional offices. See figure 3 for the location of FEMA’s 10 regional offices as well as the states and territories for which each one is responsible in terms of fulfilling National Response Framework duties. In addition, FEMA can deploy members of its Disaster Acquisition Response Team (DART), a group whose primary purpose is to support contract administration for disasters. There are two DART teams under FEMA’s Expeditionary branch, each comprised of contracting officers, contracting specialists, and quality assurance specialists. Figure 4 shows how FEMA’s contracting workforce is organized. In headquarters, FEMA’s contracting officers support a variety of functions, such as contracting for information technology needs, activities to prepare for and mitigate disasters, and disaster response. In the field, the disaster and field operations division manages contracting for disaster response efforts including: Logistics: delivering goods and services to support disaster survivors and communities, including life-sustaining commodities such as meals, blankets, and electricity generators, Response: coordinating capabilities needed immediately following a disaster, such as air and ground evacuation services and emergency sheltering, and Recovery: primarily supporting rebuilding efforts, including technical assistance programs. Regional Contracting Officers Regional contracting officers serve as the first response for contracting if a disaster occurs in their region. During a disaster, the regional offices can request additional contracting support from headquarters if needed. Contracting officers are typically located in each regional office’s mission support division, which provide essential administrative, financial, information technology, and acquisition support for the region. Each region is headed by a Regional Administrator who reports directly to the head of FEMA, the FEMA Administrator. In response to a 2009 DHS Inspector General Report, FEMA created a formal agreement to establish a new role for FEMA’s OCPO to oversee regional contracting staff. The Inspector General report found that regional contracting officers only reported to their respective supervisor in the region—who usually are not contracting officers—with no formal link to FEMA’s OCPO. The Inspector General recommended that only contracting officials should manage the technical performance of contracting officers. The report stated that having the contracting officer’s performance and career advancement controlled by someone who is not a contracting professional was an internal control risk and created a potential conflict-of-interest situation for the contracting officer. A subsequent 2011 agreement between the regions and headquarters states that a FEMA OCPO official will be the contracting officers’ performance reviewer and that the regional supervisors will continue to manage regional contracting officials’ day-to-day activities. As a result, regional contracting officers have a dual reporting chain to both FEMA OCPO in headquarters and to their supervisor within the region. In September 2015, we identified challenges with how the agreement was being implemented, particularly in that it heightened the potential for an environment of competing interests for the regional contracting officers. Specifically, we found that being physically located in a regional office where their regional supervisor is not a contracting professional gave contracting officers less standing to resist requests to perform duties outside of a contracting officer’s responsibilities or to resist pressure from program officials to make certain decisions. Further, we found that FEMA had not updated its 2011 agreement, even though the agreement states that FEMA OCPO and the regions will revisit it each year. We recommended that the FEMA Administrator direct FEMA OCPO and the regional administrators to revisit the 2011 agreement to, among other things, add details about the extent of operational control headquarters and regional supervisors should exercise to minimize potential competing interests experienced by regional contracting officers, and further detail headquarters and regional supervisors’ roles and responsibilities for managing regional contracting officers to improve coordination and communication. We also recommended, and FEMA agreed, that it establish a plan to review this agreement on an annual basis. As of January 2019, FEMA had not implemented these recommendations. Joint Field Offices After a major disaster is declared, FEMA establishes a joint field office, a temporary office through which it coordinates disaster response and recovery efforts with state and local governments and organizations. Once the need for disaster response and recovery ends and a joint field office is closed, the contracts supporting the disaster are returned to the cognizant regional contracting office. Post-Katrina Emergency Management Reform Act Contracting Requirements Congress enacted the Post-Katrina Emergency Management Reform Act of 2006 (PKEMRA) after shortcomings were identified in preparation for and response to Hurricane Katrina—one of the largest and most destructive natural disasters in U.S. history, which hit the Gulf Coast in 2005. PKEMRA included several provisions related to contracting, including: Contracting preference for local vendors. PKEMRA amended the Stafford Act to provide a contracting preference for local vendors. Specifically, for contracts or agreements with private entities, the provisions of the act state, in part: in general, for major disaster assistance activities, agencies shall provide a preference, to the extent feasible and practicable, to organizations, firms, and individuals residing or doing business primarily in the area affected by the major disaster or emergency; they may be set aside for local vendors, which means that only vendors residing or primarily doing business in the declared disaster area are allowed to compete for an award; those not awarded to local vendors shall be justified in writing in the contract file. After the enactment of PKEMRA, changes were made to the FAR to implement provisions regarding the award of set-aside contracts to local vendors. Figure 5 displays the steps a contracting officer must take to implement the preference for awarding post-disaster contracts to a local vendor based on related laws and regulation. Use of noncompetitive contracts using the urgency exception. Agencies are generally required to use full and open competition— achieved when all responsible sources are permitted to compete— when awarding contracts. The Competition in Contracting Act of 1984 recognizes that full and open competition is not feasible in all circumstances and authorizes contracting without full and open competition under certain conditions, such as in cases with an unusual and compelling urgency and the government would be seriously injured unless the agency is permitted to limit the number of sources from which it solicits offers (“urgency exception”). When DHS awards disaster contracts non-competitively based on the urgency exception, PKEMRA, as implemented in the Homeland Security Acquisition Regulation, restricts the period of performance to 150 days, unless the Head of Contracting Activity determines that exceptional circumstances apply. For other uses of the urgency exception, the FAR’s period of performance limit is generally no more than one year. Generally, exceptions to full and open competition must be supported by written justifications that contain sufficient facts and rationale to justify use of the specific exception. Depending on the proposed value of the contract, the justifications require review and approval at successively higher approval levels within the agency. Use of advance contracts. PKEMRA requires FEMA to establish advance contracts, which are typically needed to quickly provide life- sustaining goods and services, such as tarps and meals, in the immediate aftermath of disasters. While not required under PKEMRA, USACE also establishes advance contracts for supplies and services (e.g., generators for its temporary power mission) using its independent statutory authorities for emergency management, such as Section 5 of the Flood Control Act of 1941. In addition, DLA has an interagency agreement with FEMA to provide disaster commodities and services, including fuel. As such, DLA also has some advance contracts in place. In December 2018, we found that FEMA and USACE were the primary users of advance contracts. Federal Agencies Obligated at Least $5 Billion through Post- Disaster Contracts as of June 2018, but More Comprehensive Data on Disaster Contracting Obligations Would Enhance Transparency As of June 30, 2018, federal agencies obligated at least $5 billion through post-disaster contracts to support disaster response and recovery efforts after hurricanes Harvey, Irma, and Maria and the 2017 California wildfires. USACE and FEMA awarded over three quarters of the reported obligations on post-disaster contracts. However, data on post-disaster contracting are not comprehensive due to changes in the criteria for establishing and closing a NIA code and DHS’s inconsistent implementation of the criteria for closing codes. Specifically, we found DHS closed the codes for Hurricanes Harvey and Irma less than a year after the storms hit, compared to prior hurricanes when the NIA codes remained open for at least 5 years. Federal Agencies Obligated at Least $5 Billion through Post- Disaster Contracts for the 2017 Disasters As of June 30, 2018, federal agencies obligated at least $5 billion through post-disaster contracts in response to the three 2017 hurricanes and the California wildfires. Data on obligations for the California wildfires are limited to those contracts identified by two selected agencies in our review—FEMA and USACE—because no NIA code was established in FPDS-NG to track contracts specifically for the wildfire events at a government-wide level. The obligations on post-disaster contracts accounted for more than half of the $9.5 billion in contract obligations on contracts related to the three hurricanes and the 2017 California wildfires, with the remainder of the dollars obligated on advance contracts. See figure 6 for details on post-disaster and advance contract obligations by event. FEMA and USACE accounted for more than three quarters of the total obligations on post-disaster contracts for the three hurricanes. Because there was no NIA code for the 2017 California wildfires, we cannot identify government-wide obligations in FPDS-NG and, therefore, do not know which agencies had the highest contract obligations for the two wildfire events. Figure 7 provides details on known obligations on post-disaster contracts, by agency. About 63 percent of the obligations on post-disaster contracts, or $3.1 billion, was for services. See figure 8 for a breakdown of services and products by 2017 disaster. Five services across the 2017 disasters comprised nearly 80 percent of total obligations for services on post-disaster contracts. Contracts for repair and maintenance services comprised 38 percent of total obligations on post-disaster contracts for services, largely driven by the $1 billion obligated to support the power restoration effort in Puerto Rico following Hurricane Maria. Following Hurricanes Harvey and Irma, agencies primarily awarded post-disaster contracts for management support functions, such as call center services. See figure 9 for the top post- disaster contract services across the three hurricanes and the California wildfires. Of the $1.8 billion agencies obligated on goods through post-disaster contracts, 28 percent was on contracts for subsistence, such as food and water. Nearly 30 percent, or more than $530 million, of all obligations on post-disaster contracts for goods was on contracts for electric wire and power distribution equipment, almost all of which was for the power mission in Puerto Rico following Hurricane Maria. See Figure 10. Across all three hurricanes and the California wildfires, we found that the competition rate—the percentage of total obligations reported under competitive contracts—was about 75 percent for post-disaster contracts. This is an increase from the past since we previously found that the competition rate in the immediate aftermath of Hurricane Katrina was about 53 percent. Contracting for disaster relief and recovery efforts presents unique circumstances in which to solicit, award, and administer contracts. Under the FAR, agencies are generally required to use full and open competition when soliciting offers, with some exceptions. As discussed earlier, an agency may award a contract without full and open competition, for example when the need for goods and services is of such an unusual and compelling urgency that the federal government faces the risk of serious financial or other type of loss, unless the agency is permitted to limit the number of sources from which it solicits offers (“urgency exception”). When using the urgency exception, the FAR requires agencies to request offers from as many potential sources as practicable. Based on FPDS-NG data, we found that about 47 percent of obligations on post-disaster contracts were on contracts citing the urgency exception, with 63 percent of those obligations on contracts coded in FPDS-NG as using “limited competition.” Among our selected contracts, we also found that contracting officers implemented the urgency exception to seek offers from as many sources as possible in different ways. Of the 11 contracts in our sample that cited the urgency exception, five included abbreviated award time frames in the justification documentation. More Comprehensive Data Could Provide Increased Transparency on Disaster Contracting The full extent of disaster contracting—for both advance and post- disaster contracts—related to the 2017 disasters is unknown due to changes in the criteria for establishing and closing a NIA code in FPDS- NG and DHS’s inconsistent implementation of the updated criteria for closing codes. The NIA code data element in FPDS-NG was established following landfall of several major hurricanes in 2005 to enable consistent tracking of emergency or contingency-related contracting. Contracting officers select the applicable NIA code in FPDS-NG when entering related contract information into the system. Officials at GSA—the agency responsible for operating and maintaining FPDS-NG—stated there is little to no cost or administrative burden associated with establishing or maintaining a NIA code. Based on a memorandum of agreement (the agreement), GSA, DHS, and DOD are jointly responsible for determining when a NIA code should be established and closed. DHS delegated its role, on behalf of civilian agencies for disaster or emergency events, to its Office of the Chief Procurement Officer (DHS OCPO), and DOD, on behalf of military departments and defense agencies for contingency operations, delegated its role to the Defense Contract and Pricing office. The agreement outlines criteria DHS and DOD should consider in making determinations to establish and close a NIA code. We identified changes in the criteria for establishing and closing a NIA code between a June 2012 agreement and a June 2018 update that superseded and replaced it. According to DHS OCPO officials, the agencies updated the agreement to incorporate lessons learned (such as adding that events should have a procurement impact as criteria for establishing a NIA code), and because it had not been revisited in 6 years. See table 2 for criteria from the agreements, changes in 2018, and examples of potential implications of those changes that we identified related to emergency or disaster events. The June 8, 2012 agreement criteria applied to the establishment of NIA codes for the 2017 disasters, while the June 1, 2018 updates applied to determinations to close or extend the NIA codes after this date for the 2017 disasters. DHS OCPO requested that a NIA code be established for each of the 2017 major hurricanes (Harvey, Irma, and Maria). However, the codes for Harvey and Irma closed on June 30, 2018, less than a full year after the hurricanes hit. The code for Maria was scheduled to close on December 15, 2018, and in August 2018 we began raising questions about the planned or actual NIA code closures for the three 2017 hurricanes. Since December 2018, DHS OCPO provided two additional extensions for Maria, with the code now valid through June 15, 2019, about 21 months after that hurricane made landfall. In contrast, the NIA code for Hurricane Sandy, which made landfall in October 2012, remained open until December 2017, more than 5 years after the disaster. The NIA code for Hurricane Katrina, which made landfall in August 2005, remained open until August 2018, 13 years after the disaster. We observed that DHS OCPO requested NIA codes for Hurricanes Florence and Michael in 2018, although we did not review the data associated with those events. After we sent this report to the agencies for comment on February 15, 2019, the agencies allowed the codes for Florence and Michael to expire, on March 15, 2019 and April 12, 2019, respectively. DHS OCPO officials offered several different rationales to support their decision to close the NIA codes for the 2017 hurricanes and cited the changes to the criteria in the 2018 agreement for closing the codes. However, we found that these rationales were inconsistent with the criteria in the agreement, did not consider key user needs, and did not fully explain the decisions to close these codes. For example: DHS OCPO officials told us that NIA codes for disasters should be closed when agencies no longer use the special emergency procurement authority such that the procurement thresholds—such as the simplified acquisition and micro purchase thresholds—return to the general (non-emergency) procurement thresholds in the FAR. Further, when FEMA requested to keep the codes open, DHS OCPO questioned why agencies would need to continue tracking with a NIA code after the thresholds had returned to general procurement thresholds. DHS officials stated that the updated agreement put an emphasis on this criterion; however, our analysis indicated that was not consistent with 2018 agreement, which includes multiple criteria and is not limited to this factor. Further, the agreement does not provide additional emphasis on one criterion over others. DHS OCPO officials stated that the purpose of the NIA code is to track federal procurement related to response, not recovery efforts. However, both the 2012 and 2018 agreements specifically state that the NIA code is intended to track disaster response and recovery efforts. Further, according to the National Response Framework and National Disaster Recovery Framework, we found that there are no clear lines of distinction between the start and end date of these two efforts, and often these stages of the process overlap. Additionally, FEMA officials from the Recovery Support Function Leadership Group’s Program Management Office stated that they use the NIA code to track government-wide contracting related to recovery efforts. The Recovery Support Function Leadership Group, an interagency body chaired by FEMA, tasked the Program Management Office with providing accountability and transparency of projects and outcomes for the 2017 disasters, among other things. DHS OCPO officials pointed to the Digital Accountability and Transparency Act of 2014 as providing alternatives to FPDS-NG. The Digital Accountability and Transparency Act of 2014 required improvements in the quality of data on federal spending, including disaster spending, by making data more accessible and transparent, such as by improving the quality of data submitted by federal agencies to USASpending—an online tool that tracks federal grant, loan, contract, and other awards. However, we found that USASpending provided some information on contract obligations using disaster response and recovery funds but does not separate obligations by disaster event. Further, our prior work on the Digital Accountability and Transparency Act of 2014 has found limitations with the data agencies provide, notably the completeness and accuracy of data. Specifically, we found that agencies routinely provided award descriptions in an abbreviated way and lacked clarity needed to compare data across the federal government. Moreover, we found inconsistencies in agencies’ ability to track contract actions by disaster. While FEMA has the capacity to provide contract information by disaster through a centralized contract tracking tool, USACE officials stated that they use a decentralized tracking process where they reach out to the districts and centers to identify and track disaster contracts without a NIA code. Prior to the June 30, 2018 decision to close the NIA codes for Harvey and Irma, DHS OCPO officials told us they found that the number of actions FEMA was making for these events had decreased. Our analysis of the NIA codes showed that components across ten departments, including within DHS and DOD, were executing contracts related to Harvey and Irma in June 2018. When we requested supporting documentation and analysis, DHS OCPO officials provided some correspondence with FEMA but did not provide government-wide data analysis to identify what other agencies were awarding and executing contracts related to these events. DHS OCPO officials stated they also sought input from DOD through the Defense Pricing and Contracting Office on whether to keep the codes open. According to DHS officials, DOD deferred to DHS on the decision because DHS was responsible for establishing the codes. Further, DOD officials did not provide evidence that would allow us to determine whether they assessed which defense components were executing contracts related to these events or sought the input of the components that were doing so, such as USACE and the Navy. FPDS-NG—a public, government-wide database of federal procurements—offers a resource the federal government can use to create recurring and special reports for key users, such as the President, Congress, executive agencies, and the general public. The NIA code in FPDS-NG provides consistent tracking and government-wide visibility into contracting related to disaster events through a publicly available database. Without clear criteria for establishing and closing NIA codes that consider the needs of data providers and users, such as FEMA, and the high visibility of the event being tracked and a mechanism to ensure consistent implementation of these criteria, insight into disaster contracting may be limited. Additionally, federal internal control standards state that management should use quality information, communicate quality information internally, and communicate quality information externally to achieve objectives. Management should accomplish this by considering appropriate methods for communicating externally, such as to the President, Congress, and the general public. As noted above, the 2018 agreement no longer includes the 2012 criteria that a NIA code can be closed if the NIA no longer has high visibility and there is no other interest in the NIA code. In our discussions with officials, DHS OCPO could not provide a rationale for these changes and the rationale is also not included in the updated agreement. Prior to DHS OCPO’s decision to close the codes for Hurricanes Harvey and Irma, a senior FEMA procurement official requested that they remain open, in part because of the high visibility of these events. As such, this official stated that there will be continued interest in the 2017 hurricanes including inquiries from Congress, which will require agency officials to pull data for interested parties, as that data can no longer be tracked and identified through public databases, such as FPDS-NG and USASpending. DHS OCPO officials denied FEMA’s request, pointing to the criteria in the 2018 agreement, which does not include consideration of the visibility of the event or key user needs. As the federal agency responsible for coordinating disaster response and recovery, FEMA is well positioned to understand the level of national and political interest in tracking procurement information for a disaster or emergency event. Yet, it is unclear why neither the 2012 nor the updated 2018 agreements included a role for or consideration of key users, such as FEMA and Congress. Further, as noted above, FEMA program officials expressed concern over closing the Harvey and Irma codes because they had planned to use the codes to assess recovery efforts for the 2017 disasters. As we have previously reported, it can take years to fully account for federal contract obligations related to response and recovery after a hurricane. Once a NIA code is closed, there is no publicly available, government-wide system to track contract obligations for specific events. Moreover, DHS OCPO officials were unable to provide data analysis conducted using available data from prior events to determine historical patterns in federal contracting obligations for disasters prior to closing the codes for Hurricanes Harvey and Irma. Figure 11 illustrates the lack of insight we have into disaster contracting activities related to the 2017 hurricanes, in comparison to what we know about prior storms with high federal procurement obligations. Further, using the description field in FPDS-NG, we found that between July 1 and September 30, 2018, after the NIA codes were closed, agencies obligated at least $136 million on contracts for Hurricane Harvey and $123 million on contracts for Hurricane Irma. While this provides some important insights regarding the continued contracting activity related to these hurricanes, the description field in FPDS-NG cannot be relied on to provide a full picture. Some agencies may include event- specific information in the description field; however, we found that, for the 2017 hurricanes, about 65 percent of contract obligations linked to a NIA code did not include event-specific information in the description. Without reopening the NIA codes for Hurricanes Harvey and Irma, and, to the extent practicable, retroactively populating the NIA codes for contract actions supporting response and recovery for these hurricanes during the period they were closed, decision makers are missing important information to understand the procurement impact of these disasters. Retroactively entering NIA code information is not unprecedented. For example, based on our analysis, the NIA codes for the 2005 hurricanes were established in October 2005, and contracting officers retroactively entered data for contracts related to these events which occurred as early as August of that year to enable full insight into contracting for these disasters. Challenges in Planning Post- Disaster Contracts Hindered Response and Recovery Efforts Based on the contracts we reviewed and officials we spoke with responsible for the planning of these contracts, we found that agencies experienced challenges planning for post-disaster contracts, especially when it came to contracting with local vendors. Additionally, FEMA also experienced challenges with requirements development—in that program officials did not always provide well-defined or sufficiently specific requirements for post-disaster contracts. However, FEMA has taken steps to address its challenges with requirements development, but it is too soon to tell the extent to which these steps will address the challenges we identified. Agencies We Reviewed Experienced Challenges Contracting with Local Vendors Steps to Implement Local Vendor Preference, as Outlined in the Post-Katrina Emergency Management Reform Act and the Federal Acquisition Regulation (FAR) Step 1: Identify the set-aside area in accordance with FAR § 26.202-1—Local Area Set-Aside and § 6.208—Set-asides for Local Firms During a Major Disaster or Emergency Step 2: Conduct market research to determine whether there are qualified vendors in the set-aside area. Step 3: Issue a solicitation that provides for local vendor preference to the extent feasible and practicable either through the use of a set-aside or an evaluation preference. Step 4: Review offers based on evaluation criteria in the solicitation. If using a local area set-aside, review information from potential vendors to determine if they reside or primarily do business in the set-aside area in accordance with FAR § 52.226-3—Disaster or Emergency Area Representation. Step 5: Award contract to qualified vendor. If the vendor selected is not local or no qualified vendors are in the set-aside area, justify the decision in writing. determine that a vendor resides or primarily does business in the local justify in writing awards that they made to vendors outside the set- aside area. Some Officials We Interviewed Were Not Consistently Aware of the Regulatory Definition of Local Area For the contracts we reviewed, contracting officials at FEMA correctly identified the local area for six set-aside contracts across the three hurricanes, and USACE correctly identified the local area for two set- aside contracts in Puerto Rico. However, based on the interviews we conducted during our review, USACE contracting officials were not consistently aware of the specific regulation for doing so and did not correctly identify the local area for two other USACE contracts awarded in support of the California wildfires. When awarding a local area set-aside or using an evaluation preference for local vendors, FAR § 26.202-1 states that a major disaster area can span several counties in several contiguous states, but need not include all the counties in the disaster area, and cannot extend beyond the counties designated in a Presidential disaster declaration. Figure 12 provides an example of a disaster declaration that depicts which counties could be included in the set-aside area. For all six local area set-aside FEMA contracts—awarded in response to Hurricanes Harvey, Irma, and Maria—we reviewed, FEMA officials defined the local area in accordance with regulation. This was an improvement from what we previously found. Specifically, in 2015, we found that FEMA contracting officers were confused about the definition of the set-aside area and recommended that the FEMA Administrator provide new or updated guidance to ensure all contracting officers are aware of requirements concerning contracting with local vendors, among other things. DHS concurred, and FEMA updated its annual disaster contracting webinar training to reiterate the requirement and clarify how to determine the geographic area using the disaster declaration. For the two local area set-aside USACE contracts awarded, officials responsible for those contracts told us that when awarding these contracts, they were not aware of the regulatory requirements for defining the geographic area of the local area set-aside. However, as the presidential disaster declaration for Hurricane Maria included the entire island of Puerto Rico, the local set-aside area covered the entire island. As a result, officials met the set-aside area requirement in accordance with regulation, even though they noted that they were not familiar with the requirement at the time. Officials told us they became aware of the regulation after conducting research pursuant to a protest related to the use of local vendor preference. We also reviewed two other USACE contracts that were used to support the debris removal mission following the California wildfires. Contracting officials stated that they conducted market research on the availability of local contractors, and they ultimately did not find qualified local firms. However, based on a review of contract file documentation, we found that USACE officials did not identify the local area in accordance with regulation for these contracts. Instead they used congressional districts that overlapped with impacted areas to identify the local area. We found that the areas USACE identified included areas outside of the geographic area defined by the presidential disaster declaration for the California wildfires. Contracting officials responsible for these debris removal contracts stated they were not aware of a policy or regulation for how to identify the geographic area for a local area set-aside, but that their office had internally determined the use of congressional districts impacted by a disaster to be the preferred method. A senior USACE official told us that there is no agency supplemental guidance or related training regarding the use of local vendor preference for contracts supporting disaster recovery and response, only that they expect USACE contracting officials to comply with the FAR. Without additional guidance or related training, contracting officers may be unaware of how to define the geographic area for a local area-set aside in accordance with regulation and may miss opportunities to support improving the local economies of disaster impacted areas by giving preference in awarding contracts to local vendors to the extent feasible and practicable, per the Stafford Act. Regulation for Determining Whether a Vendor Resides or Primarily Does Business in the Set-Aside Area Presents Challenges Despite contracting officers having a high degree of discretion to determine that an offeror qualifies as a “local firm,”—that is, a firm that resides or primarily does business in the designated set-aside area— contracting and legal officials at both FEMA and USACE told us they were unsure what or how much information is sufficient to determine that an offeror qualifies as a local firm under the FAR. After contracting officials have identified the geographic boundaries of the local “major disaster or emergency area” and included required clauses in the solicitation and issued it as a local area set-aside, offerors must represent in their offer that they reside or primarily do business in the set-aside area. Specifically, FAR § 52.226-3(c) outlines two criteria a contracting officer should use to determine whether an offeror is to be considered “local.” If an offeror does not meet these first two criteria, FAR § 52.226- 3(d) provides eight additional criteria contracting officers may consider to make this determination (see sidebar). under FAR § 52.226-3(c) An offeror is considered to reside or primarily do business in the set-aside area if, during the last 12 months, 1) the offeror had its main operating office in the area; and 2) that office generated at least half of the offeror’s gross revenues and employed at least half of the offeror’s permanent employees. If the offeror does not meet the criteria under FAR § 52.226-3(c) consider other factors listed in FAR § 52.226-3(d) including: 1) Physical location(s) of the offeror’s permanent office(s) and date any office in the set-aside area(s) was established; 2) Current state licenses; 3) Record of past work in the set-aside area(s); 4) Contractual history the offeror has had with subcontractors and/or suppliers in the set-aside area; 5) Percentage of the offeror’s gross revenues attributable to work performed in the set-aside area; 6) Number of permanent employees the offeror employs in the set-aside area; 7) Membership in local and state organizations in the set-aside area; and 8) Other evidence that establishes the offeror resides or primarily does business in the set-aside area. Of the eight local area set-aside contracts we reviewed, two were impacted by bid protests—which is when an offeror challenges an award or proposed award of a contract or a solicitation—related to the FAR criteria for determining that an offeror qualifies as a local firm. The following protests show examples of the criteria agencies reviewed to determine whether a firm resided or primarily did business in a set-aside area. FEMA contract for food: In a protest of the award of a contract for food on the basis that FEMA improperly determined the protester failed to meet the requirements in FAR§ 52.226-3(d), the protester stated it met the requirements of FAR § 52.226-3(d), because it had (1) done past work in the set-aside area; (2) maintained a warehouse in the set-aside area; (3) maintained a contractual history with subcontractors in the set-aside area; and (4) maintained a current state license and filed a franchise tax return. FEMA denied, the protest stating that the evidence the protester provided was not sufficient to qualify as “residing or primarily doing business” in the local area. USACE Blue Roof contract: To support the Blue Roof mission— which provides temporary blue plastic roofs for disaster-impacted residences to prevent further damage and allow homeowners to arrange for permanent repairs—following Hurricane Maria in Puerto Rico, contracting officials awarded two post-disaster contracts. In a protest of the awards filed with GAO, the protestor argued, among other things, that one of the awardees did not meet local firm criteria in FAR § 52.226-3(c). USACE had assessed information on the awardee, including its local business address in the System of Award Management and other documentation of prior work in Puerto Rico, prior to award and determined that the awardee met Stafford Act criteria for award to a local vendor. USACE officials told us that, after the protest was filed, they further assessed information on the awardee in question and determined that it was a subsidiary of a larger national company. According to USACE officials, in order to quickly continue work on the Blue Roof mission, which had increased in scale, USACE negotiated pricing with the protestor while the protest was ongoing and made a third award under the solicitation. The protestor withdrew the protest. Contracting and legal officials at FEMA and USACE described difficulty in determining whether a vendor resides or primarily does business in the local set-aside area and cited a lack of clarity and different interpretations of the FAR. Based on conversations with the agencies’ legal officials, we found that USACE and FEMA applied the eight criteria in FAR § 52.226- 3(d) differently. FEMA officials told us that in determining whether a firm is local, if the first two criteria are not met, they evaluate an offeror’s information related to the eight criteria in FAR §52.226-3(d) to see if the first two criteria can be met with this additional information. They added that they look to see if the firm’s main operating office is in the set-aside area and if that office generated at least half of the offeror’s gross revenues and employed at least half of its permanent employees, but stated that the eight criteria do not need to be met within the last 12 months. Alternatively, USACE officials told us that in determining if a firm is local, if the first two criteria are not met, they evaluate an offeror’s information against the eight criteria in FAR § 52.226-3(d) independent of the two criteria described under FAR § 52.226-3(c). Legal officials at both USACE and FEMA stated that the FAR criteria should be clarified. Further, agencies’ varying application of the criteria increases the risk that an offeror may be considered local by some agencies, but not others. FEMA legal officials told us that contracting officers have been instructed to ask offerors for information on a local firm status in post-disaster solicitations. USACE legal officials explained that it is not always clear what specific information or documents provide the necessary information to meet the criteria under FAR § 52.226-3. For example, it may not be clear what documentation adequately demonstrates the number of permanent employees the offeror employs in the set-aside area, or the percentage of the offeror’s gross revenue earned in the set-aside area. The Office of Federal Procurement Policy provides overall direction of government-wide procurement policies, regulations, procedures, and forms for executive agencies. However, Office of Federal Procurement Policy staff told us that they have not provided additional guidance or clarification related to this FAR clause. Federal internal control standards state that management should use quality information to achieve objectives. Management should accomplish this by identifying information requirements, collecting relevant data from reliable sources, and processing data into quality information to be communicated internally and externally. Without clarifying guidance, contracting and legal officials will likely continue to have varying interpretations on how to implement the FAR criteria for determining that an offeror qualifies as a local firm. Some Agencies We Reviewed Did Not Consistently Write Justifications for the Use of Non-Local Vendors When contracts for major disaster or emergency assistance activities are not awarded to local vendors, the Stafford Act, as implemented in the FAR, requires that the decision be justified in writing in the contract file. Contracting officers at three of the four agencies included in our review— FEMA, USACE, and the Coast Guard—did not consistently justify in writing the award of selected contracts to non-local vendors. Specifically, 12 of the 14 contracts in our review that were not awarded to local vendors did not contain the required written justifications in the files (see table 3). DLA included written justifications for the use of non-local vendors, as required. After the 2017 disasters, FEMA identified the absence of justifications for the use of non-local vendors as an area for improvement. According to FEMA officials, they subsequently released guidance and a pre-solicitation memorandum to assist contracting officers in identifying what documentation related to local vendor preference is required in a contract file. FEMA officials told us they expect these steps will improve compliance with the requirement to document the justification for using non-local vendors going forward. While the Coast Guard provided a memorandum ahead of the 2017 disaster response that addressed the use of local vendors, it did not reference the requirement under the Stafford Act, as implemented in the FAR, to justify in writing the use of non-local vendors. A senior USACE official told us the agency had not issued any guidance to address requirements for contracting with local vendors and was not aware of any guidance issued at the department level. USACE legal officials noted the lack of written justification may be due to abbreviated timeframes under which post-disaster contracts are awarded. However, we found that USACE contracts included consolidated justification documents outlining rationales for the use of limited competition or abbreviated solicitation timeframes, but they did not include justifications for the use of non-local vendors. Without additional guidance or tools, contracting officials may not be aware that they are required to include written justifications for the use of non-local vendors in contract files, and federal agencies are at risk of not complying with the Stafford Act requirement to do so. FEMA Has Begun to Address Challenge with Requirements Development for Post- Disaster Contracts Contracting officers responsible for the FEMA contracts we selected and senior procurement officials stated that during disaster response they received post-disaster requirements packages that were lacking in technical specificity or were otherwise deficient, but FEMA has begun to address this challenge. Program officials communicate contract requirements to contracting officers through requirements documents that include, among other items, a statement of work describing goods or services to be provided by an offeror, market research, and an independent government cost estimate. Contracting officials explained that when they received deficient documents, they had to conduct additional work to refine the requirements before soliciting for the contract—such as spending time assisting program officials to develop the required documentation. This additional work may add time to already tight award time frames for post-disaster contracts. When compared to large dollar value acquisitions, post-disaster contracts are awarded on significantly abbreviated time frames. For example, among the 12 FEMA contracts we assessed, time frames between the submission of a resource request and award date ranged from 1-26 days. This is faster than suggested; FEMA’s Procurement Administrative Lead Time guidance suggests preparation time frames of 60-300 days for new procurements based on the nature and value of an action. We found instances where FEMA program offices provided inaccurate or untimely estimates of the quantities of goods or services needed for the contracts we reviewed, in some cases leading to additional time and efforts spent to meet the need. For example: After Hurricane Harvey, FEMA awarded contracts to supply a food bank. Officials told us the initial requirement from the food bank through the program office to the contracting officer was expressed in terms of “truck loads” but did not specify, for example, how large the truck should be, or how many pallets should be loaded per truck. FEMA ultimately awarded three contracts to meet the post-disaster need—the first contract had a period of performance of 4 days and, according to FEMA officials, was intended to meet initial needs for food while the program and contracting officials determined the full scope of the requirement. The second contract—a $37 million contract with a period of performance of 52 days—was intended to fulfill the remaining requirement. However, due to miscommunication of the requirement as documented in the contract files and according to a program official responsible for the contracts, FEMA needed to award a third contract for an additional 2.5 months and $23 million to meet the need. Due to the value of the contracts, FEMA deemed that the subsequent contract required a new solicitation and award, rather than a modification to the existing contracts, thereby increasing the time and effort required of procurement personnel to meet the post- disaster need for food. In response to Hurricane Maria, FEMA awarded four post-disaster contracts for self-help tarps—which are used to cover small areas of roof damage. Of these contracts, two were terminated for convenience, both of which were included in our sample. The terminations were due in part to a national supply shortage. FEMA officials told us that under one of the contracts included in our review, at the request of the Commonwealth of Puerto Rico through program officials, FEMA ordered 500,000 40-foot-by-40-foot tarps, which differ from the size of the tarps normally ordered and stocked by the agency. Due to the supply shortage, FEMA received none, but officials noted that the impact of not receiving the tarps was minimal because the agency had initially overestimated the total number of tarps needed. Since the 2017 disasters, FEMA has started to address the issues with requirements development. Specifically, in 2018, FEMA officials told us the agency used portfolio managers in the field to assist with developing requirements for disaster response. Previously, in 2017, portfolio managers told us they supported the National Response Coordination Center but did not deploy to the disasters. Organizationally housed within FEMA’s OCPO, portfolio managers we spoke with told us they provide general templates for and guidance on acquisition documents for program officials to use and are primarily responsible for supporting steady-state acquisitions included in FEMA’s Master Acquisition Planning Schedule. Additionally, portfolio managers told us they provide informal, optional, “brown bag” training sessions for program officials. FEMA OCPO officials told us that they receive more requests for portfolio manager assistance than they can support, as the portfolio management section only maintains up to six staff. FEMA OCPO officials noted, however, that the agency expected to award an acquisition support contract to expand portfolio management capabilities. While the use of portfolio managers is an important step, it is too soon to tell the extent to which the use of portfolio managers in the field will address FEMA’s challenges with requirements development for post-disaster contracts. Agencies in Our Review Have Identified Some Lessons Learned in Disaster Response, but Interagency Contracting Coordination and FEMA Workforce Challenges Remain The agencies we reviewed each have a process for identifying lessons learned following a disaster, and we found they used these processes for the 2017 disasters. While agencies have identified actions they plan to take in response to the lessons they found following the 2017 disasters, additional challenges remain. Specifically, the agencies in our review encountered interagency contracting coordination challenges during the mission assignment process. Further, FEMA identified disaster contracting workforce shortages. Selected Agencies Have Processes for Identifying Lessons Learned FEMA, USACE, Coast Guard, and DLA each have processes for identifying lessons learned within their agencies through after-action reports. These reports identify lessons learned and areas for improvement and may be completed following a training exercise or a real-world event. Through these processes, agencies identified lessons learned during the 2017 disasters. Table 4 lays out each agency’s practice or requirement for identifying lessons learned and key findings— those related to contracting and mission assignments during the 2017 disasters. FEMA Established Interagency Lessons Learned Group but the Coast Guard and USACE Could Enhance Information Sharing FEMA has also taken steps to identify interagency lessons learned by leading the Emergency Support Function Leadership Group and developing a mechanism to regularly report to the Secretary of Homeland Security. This group consists of the national emergency support function coordinators from each of the functions (such as transportation and firefighting), along with FEMA headquarters and regional officials. This body of senior officials is tasked with coordinating responsibilities and resolving operational and preparedness issues relating to interagency response activities in support of the National Response Framework. According to its charter, the group is required to carry out post-incident and after-exercise critiques, and perform substantive reviews of after- action reports, with recommendations for federal interagency partners to address shortfalls. Following the 2017 disasters, in May 2018, the Emergency Support Function Leadership Group identified 19 corrective actions, including improvements to mission assignment submission documents. Federal internal control standards state that communicating internally is key to an entity achieving its objectives. Further, as part of this communication, management should receive quality information about the entity’s operational processes that flows up the reporting lines from personnel to help management achieve the entity’s objectives. FEMA officials stated that there are processes, such as data calls, in place to solicit input from agencies. However, we noted, and FEMA officials agreed, that there is no formal reporting mechanism to the leadership group, and that it is up to the representatives from these agencies to raise issues for the group’s consideration. However, this is not consistently happening within the Coast Guard because it does not have a formal reporting process for soliciting input from officials directly involved in responding to these disasters to share with the Emergency Support Function Leadership Group. Coast Guard officials stated that they actively collect input during and immediately after an event or incident response, and that Coast Guard responders are able to provide input and issues through their chain of command at any time, but there is no formal process for reporting to the interagency group. During the course of our review, USACE officials did not provide information that indicated they had a formal reporting process for soliciting input from officials directly involved in responding to these disasters to share with the Emergency Support Function Leadership Group. Some senior level USACE officials responsible for the agency’s public works and engineering mission stated that they were unsure of the process for raising concerns to the Emergency Support Function Leadership Group and that officials were sometimes hesitant to raise issues to the group. However, in response to our draft report, USACE stated it has a formal process called the USACE Remedial Action Program for soliciting input from officials directly involved in the agency’s response and recovery following a disaster. As discussed later, we will follow up with USACE as part of our recommendation follow-up process. While Emergency Support Function Leadership Group member agencies may raise issues to the group, additional opportunities exist within these agencies to enhance the lines of communication from responders to the senior officials that comprise this leadership group. For example, some of the interagency challenges we identified in our review were not identified by this group, such as challenges in managing state and local expectations of federal response, which is discussed in more detail below. Also, USACE officials told us that some of the interagency challenges they cited following the 2017 disasters related to the mission assignment process were still present during the response to Hurricane Florence, which struck the Carolina coast in 2018. Formal processes for Emergency Support Function agencies—such as the Coast Guard and USACE—to solicit and share input from officials directly involved in the response and recovery efforts would help ensure the Emergency Support Function Leadership Group does not miss additional opportunities to improve disaster response. Interagency Contracting Coordination Challenges within the Mission Assignment Process Remain As the federal disaster coordinator, FEMA obtains requirements from states and localities and tasks the appropriate federal agencies, based on their emergency support function, through the mission assignment process. The agency assigned to a specific mission is then responsible for fulfilling those requirements, and may use contracts to do so. For example, the Coast Guard fulfills its pollution mitigation mission by executing contracts, and utilizes its own workforce to execute its search and rescue mission. USACE officials we spoke with raised concerns about the mission assignment process for the debris removal and power restoration missions related to the 2017 disasters. Specifically, USACE officials noted concerns about coordination between state, local, and federal partners for the contracts we reviewed. USACE debris removal mission: In December 2018, we found that USACE and California state officials reported different expectations related to USACE’s debris removal contracts following the wildfires, such as what structures would be removed from private properties and what levels of soil contamination would be acceptable. USACE removed more than 2.2 million tons of debris from more than 4,500 properties following the northern California wildfires. Due to the size and scope of this mission, USACE used both its advance contracts and additional post-disaster contracts for debris removal. According to USACE officials, they relied on FEMA, the lead for coordinating federal disaster response, to manage communication with states and localities and to identify and manage expectations about the scope of work to be performed using their debris removal contracts. USACE officials cited challenges with communicating to state and local officials what the agency was permitted to do under its mission assignment. For example, USACE officials told us that local officials believed that USACE would replace soil removed as part of its debris removal efforts; however, this was not part of the mission assignment from FEMA. Further, officials added that different environmental standards created confusion regarding what types of soil should be removed. For example, Napa County officials said that USACE’s mission required them to ensure that no contaminated soil remained on the properties, without regard for the naturally occurring levels of arsenic and asbestos in Napa area soil. As a result, Napa County officials said that USACE removed more soil than was necessary. However, following discussions with Napa County officials, USACE obtained site-specific samples from some properties to understand pre-existing contamination levels prior to further debris removal. USACE power restoration mission: Hurricane Maria destroyed much of the electricity grid in Puerto Rico, leaving millions without power and resulting in the longest blackout in U.S. history. To restore power to its 3.3 million people, Puerto Rico requested federal assistance with its power grid. To coordinate this effort across all stakeholders, FEMA established a unified command structure—which included the federal agencies, the Puerto Rican government and its contractors, and utility companies providing mutual assistance. According to FEMA officials, this structure allowed stakeholders to target priority work, ensure crews could access the work areas, and identify the needed materials. USACE officials stated that they received direction from FEMA and had limited direct interaction with Puerto Rican officials. However, despite this structure, USACE officials noted that changing direction from FEMA contributed to inefficiencies in contract management. For example, the scope of power restoration work Puerto Rico was requesting changed several times—such as from transmission work to distribution. These changes necessitated adjustments in contractor workforce configurations and contributed to idle time and equipment, according to officials. FEMA’s mission assignment policy designates a Federal Disaster Recovery Coordinator as the person responsible for facilitating disaster recovery coordination and collaboration among federal, state, local, tribal, and territorial governments; the private sector; and voluntary, faith-based, and community organizations. However, neither FEMA’s mission assignment policy nor its guide—which provides guidance on how to open and close mission assignments—provide additional details on how that coordination is to take place. Further, FEMA’s Response Directorate—the office that oversees the mission assignment process— was unable to identify at what level this coordination should occur. USACE and Coast Guard officials also noted that the mission assignment process does not account for other contracting considerations, such as demobilization, which occurs when contractor personnel leave the work site and return to their headquarters. According to USACE and Coast Guard officials, demobilization is required to be completed by the end of the contract’s period of performance; therefore, contracting officers need to know when the mission will end so that they can build adequate time for demobilization into the contract. Coast Guard pollution mitigation mission: Under this mission, the Coast Guard is responsible for responding to threats to public health, welfare, or the environment caused by actual or potential oil and hazardous materials incidents. Coast Guard officials told us that mission timing and the length of requirements were not communicated by FEMA in a timely manner. They told us that they contacted FEMA multiple times to determine if its mission assignment would be continued, but they did not receive an answer until shortly before the end of a contract’s period of performance. As a result, officials told us they were unsure whether they would need to demobilize contractors before completing the work, which created uncertainty about the availability of subcontractors. A FEMA Response Directorate official stated that these issues are coordination and planning concerns that should be worked out in advance between FEMA and the mission assigned agency. Ultimately, FEMA extended the Coast Guard’s mission assignment for pollution mitigation following Hurricane Maria four times. Figure 13 depicts the number of times Coast Guard’s mission was extended by FEMA. USACE power restoration mission: USACE officials cited similar challenges during the power restoration mission in Puerto Rico following Hurricane Maria. For example, USACE officials stated they typically begin planning for demobilization as soon as a mission begins. However, in this instance, officials did not know the eventual end date in order to plan for demobilization activities. Officials added that demobilization may take about 30 days, but USACE cannot extend contracts or obligate funds without a FEMA mission assignment extension. For example, if the mission assignment is scheduled to end on June 30, contracting officials would need to direct the contractor to begin demobilization as early as May 31. Officials stated that a mission assignment extension or option period of 30 days beyond the anticipated mission end date would facilitate demobilization and reduce any undue burden or concern around demobilization efforts. FEMA’s mission assignment guide does not provide a process or mechanism to follow up on the status of a mission once it is assigned. A FEMA official stated that the Response Directorate is responsible for informing their leadership of expiring mission assignments and contacting the mission-assigned agency to make them aware of the impending expiration, but that there is no standard time frame for doing so. Further, the official stated that, in some cases, FEMA may be performing this work a few days before a mission is set to expire. However, officials at USACE and Coast Guard told us they are dependent upon FEMA to reissue, clarify, or extend mission assignments. Further, the FEMA official told us that contracting considerations—such as the time needed for a contractor to mobilize and demobilize—are not necessarily built into the period of performance of a mission assignment. FEMA identified issues related to the mission assignment process, both during the 2017 disasters and following Hurricane Sandy in 2012. For example, in its 2013 Hurricane Sandy After-Action Report, FEMA found that the mission assignment process was not optimally set up to quickly surge resources to the field in a large-scale incident. To address these challenges, FEMA convened an Executive Steering Committee to update the mission assignment process, among other actions, and subsequently updated its mission assignment policy in 2015. Following the 2017 disasters, the Emergency Support Function Leadership Group identified challenges related to the mission assignment process and made recommendations to: (1) ensure response officials are properly trained on their department or agency’s statutory authorities and FEMA’s mission assignment process, and (2) develop specific recommendations to the FEMA Response Directorate on ways to reform mission assignment submission documents. These recommendations have been assigned to working groups within the Emergency Support Function Leadership Group, which plans to track the status until they are implemented. While these actions may improve the mission assignment process, they do not specifically address the issues we identified related to coordination and contracting. While the emergency support functions lay out agencies’ general responsibilities, agencies are dependent upon FEMA’s mission assignment process to further define how to perform their roles. Federal internal control standards state that management should implement control activities through its policies. These control activities include periodically reviewing policies, procedures, and related control activities for continued relevance and effectiveness in achieving the entity’s objectives or addressing related risks. Further, these standards also state that communicating internally and externally are key to achieving an entity’s objectives. As part of its internal controls, entities should evaluate the methods to communicate quality information throughout and outside of the entity on a timely basis. While FEMA revised its mission assignment guide in 2017, it still does not require FEMA to lay out coordination responsibilities in detail when assigning a mission. Without a mission assignment policy and related guidance that better incorporates contracting considerations, such as demobilization, and requires FEMA to clearly define coordination responsibilities with federal, state, and local stakeholders during the mission assignment process, federal agencies may encounter challenges fulfilling their assigned missions and may not fulfill their disaster response and recovery missions efficiently. FEMA Identified Contracting Workforce Shortages, but Has Not Fully Assessed Its Needs During the 2017 disasters, FEMA leveraged contracting staff from its regions, headquarters, and the DART teams—FEMA’s deployable contracting workforce. However, FEMA’s after-action report and officials we spoke with cited workforce shortages as a continuing challenge for disaster response and recovery. For example, officials we spoke with in several regional offices stated that there are only one to three contracting officers per region. Further, information provided by FEMA OCPO shows that eight of FEMA’s 10 regional offices have only one permanent full- time contracting official. Some of FEMA’s regional offices have additional contracting staff through FEMA’s Cadre of On-Call Response/Recovery Employees, but this varies from region to region. Regional offices are responsible for managing post-disaster contracts, even if regional procurement staff were not involved in the initial award of those contracts, according to FEMA officials. As noted in table 4 above, FEMA’s after-action report recommended increasing contract support capacities; however, it did not provide a specific plan to do so. According to FEMA officials, the agency’s workforce needs have not been assessed since a FEMA workforce analysis pilot conducted in 2014. We have identified several key principles that strategic workforce planning should address, including: determining the critical skills and competencies that will be needed to achieve current and future programmatic results, and developing strategies that are tailored to address gaps in the number, deployment, and alignment of human capital approaches for enabling and sustaining the contributions of all critical skills and competencies. Further, in our review of FEMA’s 2014 analysis, we found that FEMA evaluated contracting workforce needs, but did not specifically consider contracting workforce needs in the regional offices or address DART employees. The analysis was based on 5 years of workload data and conducted at the task or activity level, such as performing market research prior to making a contract award. However, the analysis did not prioritize skills or mission needs, nor did it identify critical competencies. In September 2018, FEMA procurement officials told us that, based on the 2014 analysis, they planned to hire 57 additional contracting staff. Officials noted that FEMA’s general operation funding does not support these additional hires, thus the agency plans to hire these staff as Stafford Act employees for 2-year appointments using disaster funding. While this is an important step, it is unclear when these staff will be hired or how they will be allocated across FEMA OCPO. For example, as of July 2018, FEMA OCPO had 72 vacant positions, including key leadership positions and contracting specialists. Without assessing its current contracting workforce needs—including staffing levels, mission needs, and skill gaps—and developing a plan to address these gaps that includes time frames, FEMA will not know whether it has the appropriate number of contracting officials with the key skills needed to meet its mission and is not likely to be well-positioned to respond to future disasters. Conclusions Contracting during a disaster can pose a unique set of challenges as officials face a significant amount of pressure to provide life-sustaining goods and services to survivors as quickly as possible. Given the scale and consecutive nature of the 2017 disasters, disaster contracts— particularly post-disaster contracts—played a key role in the response and recovery efforts. In these situations, it is important that the federal government be accountable for the contracting decisions it makes and the money it obligates, support the local economy and survivors as effectively as possible, and implement lessons learned before the next disaster strikes. Regarding accountability for the contracting decisions it makes and dollars obligated following disasters, without the ability to track disaster contracts using a NIA code in FPDS-NG, agencies, Congress, and the public lack full insight into post-disaster contracts. Providing clear criteria for establishing and closing the NIA code that accounts for the needs of users and consistently implementing these criteria will help ensure insight into high-visibility disaster events. Further, the ability to identify and track contracting dollars for disasters through a publicly available database, such as FPDS-NG, can reduce the burden on agencies to provide these data for interested parties, including Congress and other users, and offer a resource for historical data across major disasters. To help meet the needs of the local economy as effectively as possible, using a contracting preference for vendors in a disaster-affected area is an important component to early recovery efforts. Without guidance or training to ensure contracting officers are aware of the regulatory definition of the local area, agencies may miss opportunities to provide financial support to local vendors. Additionally, without clarifying how contracting officers determine whether offerors reside or primarily do business in a disaster area for the purposes of a local area set-aside, contract officials will remain uncertain on how to implement related FAR criteria. Similarly, guidance and tools to help ensure contracting officials are aware of the requirement to provide preference to the extent feasible and practicable to local vendors, including the need for written documentation on the use of non-local vendors for post-disaster contracts, will help ensure agencies comply with the requirement to do so. Taken together, these actions could enhance compliance with the Stafford Act provisions related to the award of contracts to local businesses in the disaster area, which could help jump-start the local economy. With regards to implementing lessons learned before the next disaster strikes, large scale disasters, like those that occurred in 2017, require effective coordination across emergency support function agencies. Given the Emergency Support Function Leadership Group’s responsibility to identify gaps or seams in the federal government’s efforts to respond to disasters, it is essential that the group have accurate and up-to-date information. Formal processes for soliciting and sharing information to communicate lessons learned to this group would help enhance agencies’ abilities to identify and address weaknesses in disaster response. Further, incorporating contracting considerations, such as demobilization, into the mission assignment policy, could enhance federal agencies’ ability to fulfill their disaster response and recovery missions efficiently. Lastly, without an assessment of FEMA’s contracting workforce needs, FEMA is at risk of not having a sufficient contracting workforce during a disaster. Recommendations for Executive Action We are making a total of 10 recommendations, including one to DHS, one to the Office of Federal Procurement Policy, two to FEMA, three to the Army, two to the Coast Guard, and one to GSA (in coordination with DOD and DHS). The Administrator of the General Services Administration, in coordination with the Secretaries of Defense and Homeland Security, should jointly revisit and assess the extent to which the criteria in the 2018 NIA code Memorandum of Agreement, including criteria for closing NIA codes, meet long-term visibility needs for high visibility events and account for the needs of users, such as FEMA, other agencies, and the Congress. At a minimum, the agreement should include criteria that take into account the roles of the federal agencies involved in response and recovery and provide a process that ensures consistent consideration and implementation of the criteria. (Recommendation 1) Until the NIA code Memorandum of Agreement between the General Services Administration and the Departments of Defense and Homeland Security is revised, the Secretary of Homeland Security should, in coordination with the Department of Defense and the General Services Administration, keep the existing NIA code for Hurricane Maria open, reopen the other NIA codes established for 2017 and 2018 hurricanes (Hurricanes Harvey, Irma, Florence, and Michael), and request that agencies retroactively enter NIA codes for contract actions for Hurricanes Harvey and Irma made after June 30, 2018, for Hurricane Florence made after March 15, 2019, and for Hurricane Michael made after April 12, 2019 into FPDS-NG to adequately capture contract obligations, to the extent practicable. (Recommendation 2) The Secretary of the Army should direct the Commanding General of the U.S. Army Corps of Engineers to provide guidance or related training to ensure contracting officers are aware of the regulatory definition of “local area”. (Recommendation 3) The Administrator of the Office of Federal Procurement Policy should provide additional clarification on how contracting officers should determine whether offerors reside or primarily do business in a disaster area for the purposes of a local area set-aside contract. (Recommendation 4) The Commandant of the Coast Guard should provide guidance and tools for contracting officials to use to ensure requirements concerning contracting with local vendors, including justification requirements for the use of non-local vendors, are consistently met. (Recommendation 5) The Secretary of the Army should direct the Commanding General of the U.S. Army Corps of Engineers to provide guidance and tools for contracting officials to use to ensure requirements concerning contracting with local vendors, including justification requirements for the use of non- local vendors, are consistently met. (Recommendation 6) The Secretary of the Army should direct the Commanding General of the U.S. Army Corps of Engineers to establish a formal process to solicit input from officials directly involved in the agency’s response and recovery following a disaster and to share that input with the Emergency Support Function Leadership Group. (Recommendation 7) The Commandant of the Coast Guard should establish a formal process to solicit input from officials directly involved in the agency’s response and recovery following a disaster and to share that input with the Emergency Support Function Leadership Group. (Recommendation 8) The FEMA Administrator should take the lead to work together with the Coast Guard and the U.S. Army Corps of Engineers to revise the mission assignment policy and related guidance to better incorporate consideration of contracting needs, such as demobilization, and to ensure clear communication of coordination responsibilities related to contracting. (Recommendation 9) The FEMA Administrator should assess its workforce needs—including staffing levels, mission needs, and skill gaps—for contracting staff, to include regional offices and DART; and develop a plan, including timelines, to address any gaps. (Recommendation 10) Agency Comments and Our Evaluation We provided a draft of this report to DOD, DHS, GSA, and OMB for review and comment. In written comments provided by DOD, DHS, and GSA (reproduced in appendixes III, IV, and V), as well as an email response from OMB, the agencies concurred with nine of the 10 recommendations. They generally provided steps they plan to take to address these recommendations. As discussed further below, USACE described actions it stated were sufficient to fully address the seventh recommendation, the steps described by FEMA would not fully meet the intent of the tenth recommendation, and DHS did not concur with our second recommendation. In response to the seventh recommendation as written in our draft report—to establish a formal process to solicit input from officials directly involved in the agency’s response and recovery following a disaster and to share that input with the Emergency Support Function Leadership Group—in its comments, USACE concurred and stated it has a formal process and it considered the recommendation completed. USACE noted that its Remedial Action Program solicits input from officials involved in response and recovery efforts and added that USACE shares findings from this program with the Emergency Support Function Leadership Group throughout the year and annually during the senior leaders seminar. During the course of our review, USACE did not provide information that indicated that they had such a formal process. As part of our recommendation follow-up process, we will request documentation regarding the process and how it solicits and shares information to the Emergency Support Function Leadership Group. In response to the tenth recommendation that FEMA assess its workforce needs—including staffing levels, mission needs, and skill gaps—for contracting staff, to include regional offices and DART; and develop a plan, including timelines, to address any gaps, FEMA stated that its Office of the Chief Component Procurement Officer assesses its workforce on an annual basis, with the last assessment conducted in January 2019. FEMA also noted that it entered into a contract for acquisition support services and plans to hire Cadre of On-Call Response and Recovery employees to provide dedicated support during disasters. Following FEMA’s response, we requested and received the FEMA Office of the Chief Component Procurement Officer’s 2019 workforce assessment. As with FEMA’s 2014 workforce analysis, the 2019 assessment calculated the number of employees needed based on the estimated time to complete a task. However, the assessment did not include an analysis of mission needs or skill gaps, and the assessment provided does not specify whether it includes the needs of regional offices and DART. FEMA estimates that it will implement this recommendation in September 2019, and we will continue to monitor FEMA’s planned efforts through our recommendation follow-up process. DHS did not concur with the draft report’s second recommendation regarding NIA codes. In its response, with regards to extending existing NIA codes and reinstating expired NIA codes, DHS stated that it is bound by the memorandum of agreement with GSA and DOD, unless or until all three signatory agencies agree to revise or suspend the agreement. We recognize that all three agencies are bound by the agreement, and also recommended in the first recommendation that GSA, DOD, and DHS jointly revisit the agreement. GSA concurred with this recommendation in its written comments reproduced in Appendix V. In an email sent from an official within DOD’s Defense Pricing and Contracting Office, DOD concurred. DHS did not respond to our first recommendation. As such, we have revised the second recommendation to state that DHS take action in coordination with DOD and GSA. We also note that the memorandum of agreement states that extending expiring or already expired NIA code end date is appropriate, in part, when two or more agencies do not have a reasonable alternative method of identifying and internally tracking those emergency acquisitions. We discuss in our report how once the NIA code is closed, there is no publicly available, government-wide system to track contract obligations for specific events. We also discuss how, using the description field (which does not provide a full picture) in FPDS-NG, agencies obligated more than $250 million on contracts for Hurricanes Harvey and Irma during the three months after the NIA codes for these two hurricanes were closed. Given this, we continue to believe DHS should consider reopening the codes for Hurricanes Harvey and Irma, in coordination with DOD and GSA. Moreover, in its response to the second recommendation DHS further stated that FEMA’s Office of the Chief Component Procurement Officer (who is not currently a party to the memorandum of agreement), believes the recommendation to extend the NIA codes for 2018 Hurricanes Michael and Florence goes beyond the scope of this audit. While the main focus of this report is the 2017 hurricanes and California wildfires, we discuss Hurricanes Florence and Michael in this draft with respect to the NIA codes, as the same issues and concerns we raised apply regardless of the year of the hurricane. However, after we sent the draft to the agencies for comment, the agencies let the codes for Hurricanes Florence and Michael expire on March 15, 2019 and April 12, 2019, respectively. We therefore revised the second recommendation to recommend that the codes for Hurricanes Florence and Michael should be reopened (rather than kept open). In its written comments, DHS also stated that neither DHS nor FEMA can unilaterally direct other agencies to retroactively enter FPDS-NG data for Hurricanes Harvey and Irma. We acknowledge this and have revised the recommendation to recommend that DHS request, rather than direct, other agencies to retroactively enter the information, to the extent practicable. As we state in the report, the NIA codes for the 2005 hurricanes were established in October 2005, and contracting officers retroactively entered data for contracts related to these events to enable full insight into contracting for these disasters. DHS further stated that retroactively entering data into FPDS-NG is not practical and places an unreasonable burden on contracting staff, and that the draft did not support the case that there were any benefits to be gained. We recognize that there is some burden associated with the recommendation, thus we recommended that DHS request agencies take action to the extent practicable. In terms of benefits, the report identifies benefits in terms of providing decision makers with important information to understand the procurement impact of such disasters. DOD and DHS also provided technical comments, which we incorporated as appropriate. We are sending copies of this report to the appropriate congressional committees, the Secretary of Defense, the U.S. Army Corps of Engineers Director of Contracting, the Director of the Defense Logistics Agency, the Secretary of Homeland Security, the Administrator of the Federal Emergency Management Agency, the Federal Emergency Management Agency’s Chief Procurement Officer, the Commandant of the Coast Guard, the Administrator of the General Services Administration, the Director of the Office of Management and Budget, and the Administrator of the Office of Federal Procurement Policy. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-4841 or makm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix VI. Appendix I: Objectives, Scope, and Methodology This report specifically addresses the use of post disaster contracts and: (1) assesses the extent to which federal agencies obligated funds on post-disaster contracts in response to the 2017 major disasters; (2) assesses the extent to which selected agencies experienced challenges in the planning process for selected post-disaster contracts; and (3) describes selected agencies’ lessons learned as a result of the 2017 major disasters and assesses the extent to which they have taken action to address them. To identify the extent to which federal agencies obligated funds on post- disaster contracts in response to the 2017 disasters, we reviewed Federal Procurement Data System-Next Generation (FPDS-NG) data through June 30, 2018, the most recent and complete data at the time of our review. We adjusted the obligation data to constant fiscal year 2018 dollars using the Fiscal Year Gross Domestic Product price index. We identified hurricane obligations using the national interest action (NIA) code, as well as the contract description. Data on obligations for the California wildfires is limited to those contracts, if any, identified by the agencies with the highest obligations on post- disaster contracts for the hurricanes—the Federal Emergency Management Agency (FEMA), U.S. Army Corps of Engineers (USACE), Defense Logistics Agency (DLA), and the U.S. Coast Guard (Coast Guard)—because no NIA code was established in FPDS-NG. Coast Guard officials stated that they did not execute any contracts in response to the 2017 California wildfires. DLA officials stated that they maintain contracts, which for the most part provide inventory replenishment for DLA and the U.S. Forest Service within the U.S. Department of Agriculture, but they were unable to provide data on contracts awarded or executed specifically for the two wildfire disasters in the scope of our review. Therefore, our analysis only captures obligations for FEMA and USACE reported contracts related to the 2017 California wildfires. To determine which obligations were made through the use of post- disaster contracts versus advance contracts, we reviewed documentation provided by FEMA and USACE identifying the advance contracts they have in place and that were used in support of the 2017 disasters. We analyzed the FPDS-NG data against these contracts to identify obligations on post-disaster contracts and compared these to obligations on advance contracts by disaster. We analyzed competition procedures used and the types of goods and services procured for post-disaster contracts. In addition to advance contracts for disaster response, agencies can leverage other existing contract vehicles. For example, to respond to its pollution mitigation functions under emergency support function 10, the Coast Guard awards task orders off of its portfolio of basic ordering agreements. For the purposes of this report, post-disaster contracts include all contract awards and orders that were not identified by FEMA or USACE as advance contracts. To assess the extent to which disaster contract obligations can be tracked through FPDS-NG using the NIA code, we identified prior hurricane events with the highest contract obligations from 2005 through September 2018. We analyzed the data to determine when the highest level of federal contract obligations occurs following a hurricane. We also assessed the process for establishing and closing a NIA code. Specifically, we reviewed the criteria in the 2012 and 2018 memorandums of agreement between DHS, DOD, and the General Services Administration, and interviewed officials involved in the process. We assessed the reliability of FPDS-NG data by reviewing existing information about the FPDS-NG system and the data it collects— specifically, the data dictionary and data validation rules—and performing electronic testing. We also compared FPDS-NG data to the contract files in our review. Specifically, to review our selected post-disaster contracts for data reliability, we compared items such as, the extent competed, the use of a local area set-aside, NIA code, and termination status, based on the contract information and the information in FPDS-NG. Based on the steps we took, we determined the FPDS-NG data were sufficiently reliable for the purposes of describing agencies’ post- disaster contract obligations. To assess the extent to which agencies experienced challenges in the planning of selected post-disaster contracts, we reviewed relevant laws and regulations, including the Post-Katrina Emergency Management Reform Act (PKEMRA), the Federal Acquisition Regulation (FAR), the Robert T. Stafford Disaster Relief and Emergency Assistance Act (Stafford Act), as well as agency policy and guidance. We identified a non-generalizable sample of 23 post-disaster contracts from the four agencies with the highest post-disaster obligations based on FPDS-NG data as of March 31, 2018—DHS’s FEMA, DOD’s USACE, DOD’s DLA, and DHS’s Coast Guard. We selected contracts across the four major 2017 disasters included in our scope (Hurricanes Harvey, Irma, and Maria, as well as the California wildfires) based on four selection criteria—(1) contracts using the urgency exception to full and open competition; (2) contracts using a local area set-aside; (3) contracts awarded to small businesses; and (4) contracts terminated for cause or convenience. Our goal in this selection was to ensure we selected a range of contracts within each of these four criteria so as to assess the extent to which these contracts implemented certain laws and regulations. Specifically, we selected contracts based on the use of urgency and local area set-asides in order to assess agencies’ implementation of relevant PKEMRA, Stafford Act and FAR criteria for post-disaster contracts. Because the obligations for local area set-aside contracts was low across all federal agencies, about 5 percent of total post-disaster obligations, we selected contracts that were awarded to small business vendors as a proxy to identify other awards to local vendors. Finally, we selected terminated contracts to assess additional challenges related to post- disaster contracts, such as the availability of contracted services and supplies and the requirement setting process. Based on these criteria, we selected 12 FEMA, 7 USACE, 2 DLA, and 2 Coast Guard contracts. Findings based on information collected from the 23 contracts cannot be generalized to all post-disaster contracts. Additional details on our selected contracts can be found in table 5. To assess how agencies used the urgency exception to full and open competition, we reviewed selected contracts for the inclusion of a justification and approval for other than full and open competition including sole source justifications and exclusion of sources justifications. To assess the extent to which agencies provided preference to local vendors for post-disaster contracts, we reviewed selected contract files for the use of a set-aside or an evaluation preference listed in the contract solicitation, and the inclusion of justifications for contracts not awarded to local vendors. Additionally, we reviewed applicable agency guidance and interviewed contracting and senior procurement officials across all four agencies regarding their use of local area set-asides, including the means by which they define the geographic set-aside area and determine that an offeror primarily resides or does business in the set-aside area. We also met with officials from the Office of Management and Budget’s Office of Federal Procurement Policy to discuss relevant FAR criteria. To assess how FEMA program offices develop and deliver requirements packages for use by contracting officers and the extent to which those packages are sufficiently specific to allow contracting officers to issue a contract solicitation, we interviewed contracting, program, and senior procurement officials responsible for the contracts in our selection sample. We discussed the specificity of initial versus final requirements, the nature of requirements changes, the process of requirements development, and training provided to program officials regarding the requirements development process. We also reviewed new post-disaster awards at FEMA to determine time frames between resource request to award on average for post-disaster contracts. We compared these findings to relevant agency guidance on acquisition planning. To describe lessons learned selected agencies identified related to the use of post-disaster contracts and assess the extent to which agencies have taken action to address them, we reviewed available completed after-action reports from the 2017 and prior disasters, including the Hurricane Sandy FEMA After-Action Report, the 2017 Hurricane Season FEMA After-Action Report, USACE’s Temporary Emergency Power Mission After Action Review for Hurricane Matthew, USACE’s Puerto Rico After Action Review, USACE’s Northern California Wildfires Debris Removal Mission After Action Review, the Coast Guard’s 2017 Hurricane Season Strategic Lessons Learned After Action Report, and the Defense Logistics Agency’s 2017 Hurricane After Action Meeting papers. We also reviewed findings from the Emergency Support Function Leadership Group related to interagency lessons learned. As part of our review, we identified requirements for agencies to document or practices agencies use to document lessons learned following a disaster, agency specific and interagency lessons learned specific to post-disaster contracts and mission assignments, and recommendations or actions planned by the agencies to address them. We reviewed federal internal control standards and the Emergency Support Function Leadership Group charter and the standard operating procedures for its Preparedness Evaluation/Corrective Action Working Group. To describe challenges related to coordination with state and local officials on the use of post-disaster contracts, we interviewed FEMA, USACE, DLA, and Coast Guard officials. To obtain perspectives and examples from state and local government officials involved in disaster response, we interviewed officials in California on the use of federal contracts. We also met with state and local officials in Texas, Florida, Puerto Rico, and the U.S. Virgin Islands to discuss the federal response to the 2017 hurricanes more broadly. The information gathered from these officials is not generalizable to all officials. To describe challenges related to the mission assignment process, we interviewed FEMA, USACE, and Coast Guard officials, including officials from FEMA’s Response Directorate and the contracting officials from USACE and the Coast Guard that awarded the contracts these agencies used to fulfill their missions. We also reviewed the mission assignment documents, where FEMA assigned USACE and Coast Guard missions and laid out their responsibilities. To assess workforce challenges, we reviewed DHS’s 2014 workforce assessment, which identified gaps in FEMA’s contracting workforce. We also obtained information from FEMA on its current contracting workforce in headquarters, regional offices, Disaster Assistance Response Team, and joint field offices. We also interviewed FEMA contracting officials to obtain their perspectives and experiences during the 2017 disaster season. We conducted this performance audit from March 2018 to April 2019 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Emergency Support Functions Responsibilities across the Federal Government The National Response Framework identifies 14 emergency support functions (ESF) and designates a federal department or agency as the coordinating agency for each function. ESFs are the federal government’s primary coordinating structure for response, and under this structure, the Federal Emergency Management Agency (FEMA) acts as the federal coordinating agency. Appendix III: Comments from the Department of Defense Appendix IV: Comments from the Department of Homeland Security Appendix V: Comments from the General Services Administration Appendix VI: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Janet McKelvey (Assistant Director), Katherine Trimble (Assistant Director), Caryn E. Kuebler (Analyst in Charge), Lindsay Taylor, and Sarah Tempel were principal contributors. In addition, the following people made contributions to this report: Emily Bond, Lorraine Ettaro, Suellen Foth, Julia Kennon, Carol Petersen, Sylvia Schatz, Alyssa Weir, and Robin Wilson.

Why GAO Did This Study Federal contracts play a key role in timely response and recovery efforts following disasters. While federal agencies, such as FEMA and USACE, may have advance contracts in place for obtaining goods and services following disasters, agencies may also award post-disaster contracts. GAO was asked to review the federal government's response to three major hurricanes in 2017, as well as the 2017 California wildfires. This report addresses, among other objectives, the extent to which (1) federal agencies obligated funds on post-disaster contracts in response to the these events, and (2) selected agencies experienced challenges in the planning of selected contracts. GAO analyzed data from the Federal Procurement Data System-Next Generation; selected a non-generalizable sample of 23 post-disaster contracts based on factors such as if the contract was set aside for award to a local contractor; reviewed federal regulations and agency guidance; and interviewed agency officials. What GAO Found Following hurricanes Harvey, Irma, and Maria and the 2017 California wildfires, federal agencies obligated at least $5 billion in post-disaster contracts—which are awarded after disasters hit— to support disaster response and recovery efforts. The U.S. Army Corps of Engineers (USACE) and the Federal Emergency Management Agency (FEMA) comprised over three-quarters of reported post-disaster contract obligations as of June 30, 2018 (see figure). However, the full extent of post-disaster contracting related to the 2017 disasters is unknown due to the Department of Homeland Security's (DHS) inconsistent implementation of the criteria for closing a national interest action (NIA) code. This code allows agencies to track data on contract actions related to national emergencies, providing government-wide insight into response and recovery efforts. DHS closed the codes for Harvey and Irma on June 30, 2018, less than a year after those hurricanes hit. In contrast, the codes for prior hurricanes were open for at least five years, with Katrina remaining open for 13 years. Based on a review of 23 contract files from FEMA, USACE, the Defense Logistics Agency, and the Coast Guard, GAO identified challenges in the planning of selected contracts. For example, GAO found USACE officials were not consistently aware of the regulation that defines “local area.” GAO also found that contracting officers at FEMA, USACE, and the Coast Guard did not consistently write justifications for awards to non-local vendors outside the disaster area, as required. FEMA developed guidance to address this, but the Coast Guard and USACE have not issued guidance or tools to address this requirement. Without addressing planning challenges, agencies may miss opportunities to award contracts to local businesses in the disaster area to the extent feasible and practicable, which could help jump-start the local economy. What GAO Recommends GAO is making 10 recommendations, including that DHS reopen NIA codes for Hurricanes Harvey and Irma; USACE provide guidance on the local area definition; and the Coast Guard and USACE provide guidance to ensure contracting requirements for the use of non-local vendors are met. Agencies concurred with 9 recommendations. DHS did not agree that NIA codes should be reopened. GAO continues to believe DHS should do so, to the extent practicable, as discussed in the report.

gao_GAO-19-596

gao_GAO-19-596_0

Background The boards that oversee the Social Security and Medicare trust funds are technically separate entities under the Social Security Act, but the same set of trustees have served as board members for each trust fund and the boards meet concurrently. For each board, four of the trustees are ex officio, i.e. members by virtue of their office and position: the Secretary of the Treasury, Secretary of Labor, Secretary of Health and Human Services, and the Commissioner of Social Security. The remaining two trustees are members of the public, nominated by the President, confirmed by the Senate, and the public trustees must not be from the same political party as one another. The public trustee positions were established in 1983; they have been vacant since 2015. The boards issue two separate Trustees reports each year, one on the Social Security trust funds and one on the Medicare trust funds. Under the Social Security Act, these reports are due by April 1 of each year. The Trustees reports provide information on the present and projected statuses of the trust funds, including their projected balances over the next 10 years (short-term), the next 75 years (long-term), and the assumptions and methods used to make these projections. The reports provide estimates of the projected costs and incomes of the trust funds, and any dates that the boards project the trust funds’ reserves to become depleted, among other information about the programs. Because projections are inherently uncertain, the reports include three projection scenarios: intermediate, low-cost, and high-cost alternatives, along with other information about uncertainty. The intermediate scenario is based on assumptions that reflect the boards’ best estimate of future experience. The low-cost scenario makes assumptions that are relatively more favorable with respect to the projected statuses of the trust funds, while the high-cost scenario does the opposite. For example, the low-cost scenario assumes more workers will pay into the trust funds and fewer beneficiaries will receive benefits, while the high-cost scenario assumes fewer workers and more beneficiaries. Agency Officials and Trustees Follow a Collaborative Process Each Year to Develop the Trustees Reports Agency Officials and Trustees Determine the Economic, Demographic, and Programmatic Assumptions Prior to Drafting the Trustees Reports Officials in the Social Security Administration Office of the Chief Actuary (SSA OCACT) and the Centers for Medicare & Medicaid Services Office of the Actuary (CMS OACT) work with other agency officials and trustees to develop assumptions and draft and revise the Trustees reports in an annual cycle, according to agency officials and the board meeting minutes and report development schedules we reviewed (see fig. 1). At the end of each cycle, the boards have established a working group that is largely responsible for overseeing the day-to-day development of the next year’s report. This working group consists of officials in the four agencies that are led by the ex officio trustees (Treasury, DOL, HHS, and SSA), officials from SSA OCACT and CMS OACT, and the public trustees, when confirmed. All of the working group’s discussions and agreements are subject to the approval of the boards. The Secretary of the Treasury serves as the Managing Trustee and Chairperson of the boards. Treasury staff has historically coordinated the report development process, including organizing the development schedule and hosting the working group and boards’ meetings. The reports are drafted by SSA and CMS. SSA OCACT and CMS OACT officials we interviewed said they work to update the assumptions—both long-term and short-term—they propose as the basis for the trust fund projections in the reports. Assumptions are the demographic, economic, and program-specific factors that the actuaries use to model the future financial status of the trust funds (see appendix I). For each assumption, SSA OCACT and CMS OACT go through a process of updating the values for the 75-year projection period as needed and use those updated values as inputs in their models to project future costs and income for the trust funds. As a part of this process, the working group discusses issues that inform the assumptions proposed by SSA OCACT and CMS OACT. The work on assumptions is divided according to the specializations and expertise of the two actuarial offices, and is developed by staff from a range of disciplines, including actuaries, demographers, and economists. SSA OCACT develops the demographic and economic assumptions that are common to both reports, including rates for fertility, mortality, and growth in gross domestic product. SSA OCACT also prepares the programmatic assumptions for Social Security, such as the numbers of retirement and disability beneficiaries and the anticipated income into the trust funds from payroll taxes. CMS OACT prepares the programmatic assumptions that are specific to Medicare, such as the number of Medicare beneficiaries and expected growth in health care costs. SSA OCACT and CMS OACT officials update assumptions and revise their methodologies based on recent data, if a change is warranted. For example, the 2017 Social Security Trustees report projected an increase in the total fertility rate. Information collected in the subsequent year showed that fertility rates had not risen as expected, so officials reduced the fertility rate assumptions for the 2018 report. SSA OCACT officials told us that they look at both the reasonableness of the assumptions individually and in the aggregate, as some assumptions interrelate. According to agency officials we interviewed, the assumptions generally undergo gradual or no changes from year to year, unless there are significant policy changes. SSA OCACT and CMS OACT also update their models by incorporating more recent data into them. For example, in the 2017 Social Security Trustees report, the model for projecting average age benefit levels of retired worker and disabled worker beneficiaries who are newly entitled to benefits used a sample of these beneficiaries from 2013. In the 2018 report, this model was updated to use a sample from 2015. The Working Group Discusses and Works Toward Consensus on Assumptions The working group considers and works towards consensus on the assumptions proposed by SSA OCACT and CMS OACT. Members of the working group meet periodically to discuss the assumptions and come to an agreement on the values for them. In these meetings, the working group often hears presentations from internal or external experts on specific topics. For example, in one meeting, SSA staff led a presentation and discussion on Disability Insurance, and DOL staff led a presentation and discussion on how globalization might affect long-term economic trends. To inform their discussions, the working group may also review reports from technical panels or invite panel members to discuss their findings and recommendations at a working group meeting. For example, in September 2012, the working group discussed a Medicare technical panel recommendation that the board continue to present alternative projections in which average Medicare spending per beneficiary rises faster than the current law baseline; the working group and board agreed to implement this recommendation. Throughout the working group’s activities, the members representing the ex officio trustees generally serve as a liaison between the trustee for their agency and the working group. When confirmed, public trustees participate directly on the working group. After consideration, the working group finalizes long-term assumptions at a fall board meeting. The long-term assumptions serve as the basis for the short-term assumptions and the 75-year trust fund projections. For each long-term assumption, the boards set the “ultimate value”, i.e. the constant rate or number that is projected to be met in a particular year (within 10 years in most cases) and then continued through the remainder of the 75-year projection period. For example, for the 2019 Trustees reports, the boards set the ultimate value for the annual change in covered earnings as a percent of total labor compensation for each year beginning in 2028 and continuing through 2093. In most cases, according to agency officials we interviewed, the working group achieves consensus on the assumptions before the fall board meeting. However, when the working group is unable to reach consensus, the boards settle any outstanding issues and tend to either make no changes or incremental changes over time to avoid major swings in year-to-year projections, according to some agency officials we interviewed. Once the long-term ultimate values are set by the boards, the working group then discusses the short-term assumptions that bridge the gap between current data and the ultimate values. The working group first considers and works toward agreement on the short-term economic assumptions and then the health assumptions. Short-term economic assumptions can vary during the early years of the projection period. The projection of Medicare’s HI Trust Fund depletion date is based on detailed short-term growth rate assumptions for individual types of Medicare services, such as inpatient hospital care. SSA OCACT and CMS OACT Draft the Trustees Reports Once the assumptions are set, officials at SSA OCACT develop the projections that determine the actuarial status and then draft the Social Security Trustees report, and officials at CMS OACT do the same for the Medicare Trustees report. The reports include information on and values of the assumptions, projected financial statuses of the trust funds and programs, actuarial analyses and estimates, and technical information on the methodologies and projections. In addition, the reports note changes to the assumptions, methodology, and projections from prior reports, and explain the implications for the trust funds. The reports also include statements of opinion by the relevant agency’s Chief Actuary regarding whether the techniques and methodologies used are generally accepted within the actuarial profession and whether the assumptions used and the resulting actuarial estimates are reasonable. The Working Group Comments and Develops Consensus on Reports When the drafts are completed, SSA OCACT and CMS OACT circulate them to the working group for comments and agreement. According to one former public trustee, these comments are mostly related to the presentation of the information, such as word choices, as members have previously agreed to the assumptions. SSA OCACT or CMS OACT officials respond to these comments, and make revisions to the reports in several rounds, engaging with the working group for comment on each new version of the reports. As with the earlier round when the working group worked toward consensus on the assumptions, the working group members that represent the ex officio trustees can brief the trustee from their agency and bring any input back to the working group to help ensure that the trustees agree with the reports. The Boards Approve and Issue Trustees Reports The final drafts of the Trustees reports are presented and approved at the annual spring meeting of the boards. Under the boards’ bylaws, members of the boards must be present at these meetings to approve the reports. During the meeting, agency officials provide an overview of the reports to the trustees and other attendees, and explain changes in the overall projections from the previous year’s reports. For those trust funds with an estimated depletion date, agency officials explain the estimated dates of depletion and the potential implications for beneficiaries. After any discussion, the trustees sign the reports and the boards formally issue them to Congress. Public Trustees Can Play Unique Roles in Developing and Presenting the Trustees Reports Public trustees, when confirmed, play unique roles as members of the boards and also the working group that develops the Trustees reports. Former public trustees we interviewed said their role was to represent the public in the report development process, independent of the ex officio trustees and other agency officials in the administration. To become members of the board, public trustees must be nominated by the President and confirmed by the Senate, and the public trustee cannot both be from the same political party. Those we interviewed stressed the importance of not allowing personal and political opinions to influence their work on the Trustees reports. As a result, according to both agency officials and former public trustees, having public trustees in place lends credibility to the reports. Former public trustees stated that they worked closely with their counterpart public trustee to coordinate their comments and input to the working group. Historically, public trustees sometimes questioned or encouraged changes to some assumptions used in the reports, according to former public trustees and some agency officials. When in place, public trustees regularly attend working group meetings, whereas ex officio trustees do not. According to the former public trustees we interviewed, they saw part of their role as facilitating conversations as leaders and moving the group towards consensus on assumptions. Additionally, former public trustees and some agency officials said trustees are more hesitant to change the assumptions in the reports when there are no public trustees in place, out of concern that any change could be viewed as politically motivated. For example, in 2017 the boards discussed whether or not to change the long-range real interest rate assumption from the rate used in the previous year’s Trustees reports. The boards decided to keep the assumptions unchanged, in part because there were no public trustees in place. When they are in place, public trustees can also help communicate the message of the Trustees reports to policy makers and the public. As an example, the Trustees reports can be technical and difficult to understand; to address this, the public trustees introduced a summary of the reports in 1991, which presented the reports’ findings in a way that is more accessible to the general public. Former public trustees said they were able to inform policy makers on the contents of the reports through congressional testimony and direct conversations with congressional staff. One former public trustee reported that he was a resource for the media, spending hours on the phone providing his perspective and explaining the reports’ implications for policy decisions. In addition, public trustees published a separate message that allowed them to present what they believe to be the main idea of the reports. The Boards Have Frequently Missed the Statutory Deadline and Have Not Effectively Managed the Report- Development Schedule Trustees Reports Have Been Issued Late in Part to Allow More Time for Updating Reports and Due to Scheduling Issues The boards issued the Trustees reports to Congress after the April 1 statutory deadline in 17 of the 25 years from 1995 to 2019, including every year from 2009 through 2019 (see fig. 2). Since 2009, the boards have issued the reports at least 2 months late six times; they only issued the reports this late one time in the 14 years from 1995 to 2008. Agency officials and former public trustees provided a number of reasons why the Trustees reports have been late in recent years. Agency officials and former public trustees said they may delay reports in order to include the impact of late-breaking legislation or policy changes on the assumptions or data. SSA OCACT told us that this decision is based on (1) if the policy change results in substantial changes to assumptions and (2) if the policy change affects a policy that is directly governing a trust fund. For example, agency officials and former public trustees stated that the Patient Protection and Affordable Care Act (PPACA), enacted on March 23, 2010, significantly contributed to the 2010 reports being issued August 5, 2010, over 4 months past the deadline. PPACA significantly affected many of the factors that were the basis for the Medicare Trustees report projections, such as reducing projected Medicare expenditures through various policy changes, including a change to the payment formula for the Medicare Advantage program—the private health plan alternative to traditional Medicare. According to one former public trustee, if the boards had issued a report that did not reflect the changes made by PPACA, it would not have been applicable to the current outlook of the Medicare trust funds and therefore not as useful to Congress and the public. Agency officials have also reported that there have been instances of waiting for more complete or recent data sets to become available before calculating the actuaries’ projections. According to CMS officials, a tradeoff exists between updating data and meeting the deadline. For example, Treasury officials told us that because the working group decided that CMS OACT should not wait for January 2019 Medicare Advantage enrollment data, the 2019 Trustees reports were issued earlier (April 22) than they would have been if they had waited for the complete end of year data, as they had in previous years. Agency officials and public trustees also cited difficulties in scheduling the spring board meetings as a factor that contributed to delays in issuing the Trustees reports. The boards’ bylaws require the annual reports to be adopted by a majority of the trustees who are present and voting. However, sometimes Treasury staff experienced difficulty scheduling the meeting. According to Treasury officials responsible for scheduling the meeting, they generally wait until the first drafts of the Trustees reports are completed before they schedule the spring board meeting to avoid having to reschedule the meeting if the draft reports are provided after the working group’s internal deadline. For the last 15 years (2005-2019), report development schedules from SSA OCACT indicated that the draft reports were provided to the working group after the internal deadline 12 times. In the other 3 years, the report development schedules did not show the actual date that the draft reports were provided. As a result of scheduling the meeting later in the process, Treasury staff has sometimes not been able schedule a meeting that all of the Trustees can attend prior to the statutory deadline of April 1. Other challenges that contribute to delays include government shutdowns and staff having conflicting concurrent responsibilities, according to some agency officials or former public trustees. When the government shut down for 11 business days in October 2013, the board meeting minutes show that it affected the timelines for the 2014 Trustees reports. However, according to HHS, government shutdowns have never materially delayed the release dates for the Trustees reports. Some former public trustees and one agency official we spoke to stated that agency officials involved in the report process sometimes had other duties competing for their time, which could result in delaying their work on the Trustees reports, while other agency officials stated this was not a factor. The Process for Managing the Schedule for Developing the Trustees Reports Does Not Reflect Best Practices Agency officials’ scheduling process is inconsistent with GAO’s guide on best practices for schedule management. Agency officials and former public trustees said they attempted to meet the statutory deadline each year, but did not believe issuing the report after the deadline created serious negative consequences. As a result, agency officials and former public trustees involved with developing the reports in recent years said they developed a schedule designed to meet the deadline knowing it would most likely not be met. Several agency officials and former public trustees described the schedule as “ambitious” and difficult to achieve. If the schedule is unrealistic from the start of the process, and if involved parties view it as an unlikely goal, rather than the expected outcome, then the schedule does not serve as a useful tool for managing the timely development of the Trustees reports. In addition to designing an unrealistic Trustees reports schedule, agency officials did not always document actual progress in meeting scheduled dates or modify the schedule in a way that would allow them to overcome early setbacks. Treasury officials, who organize the schedule for developing the Trustees reports, stated that the initial proposed schedule is updated only once during the report process, after the first drafts of the reports are completed. According to best practices, the schedule should be updated regularly with actual progress and remaining work. Without doing so, it could be difficult to respond to actual events while still meeting set deadlines. Treasury has not regularly archived the final version of the schedule with the dates that milestones were actually met. According to best practices, the final iteration of the schedule that was actually followed should be archived and used to inform and improve future schedules. Treasury officials were able to provide us with the archived, updated schedules for only 6 of the last 25 years, and these schedules were incomplete. While these updated schedules showed the actual dates that the draft reports were provided and the planned dates for later milestones, they did not include the dates for milestones before the draft reports were provided or the actual dates for the later milestones, including the reports’ issuance dates. Further, although the boards have regularly missed the statutory deadline, the initial report-development schedules have not significantly changed in recent years. Based on the proposed schedules for the Trustees reports that are presented in the spring board meeting minutes we reviewed, the initial schedules for each milestone in the report development process, such as obtaining agreement on assumptions or circulating drafts, has not significantly changed in the last 6 years, although the schedules have consistently proven difficult to meet (see table 1). Without recording the actual report production schedule that was followed, participating officials do not have the historical data that would assist them in making meaningful and effective changes to future schedules. Finally, according to best practices, it is important that stakeholders, including decision makers, have access to information on the progress of the project. Agency officials and former public trustees stated that they do not have a policy or practice of informing Congress of delays or changes to the schedules for the Trustees reports, even in years when the board issues the reports months after the deadline. Given this, Congress, the recipient of the Trustees reports, remains uninformed of the reports’ release date or the factors contributing to a delay in any given year. This uncertainty may hinder Congress from planning legislative sessions in advance that would use the findings of the Trustees reports. For example, congressional committees of jurisdiction may be hindered in scheduling hearings on or around the time of the reports’ release date and having access to the latest data from the reports to inform their oversight. Conclusions Trustees and agency officials set out at the start of each report cycle with a schedule to meet the statutory deadline to issue the Social Security and Medicare reports each year by April 1. However, over the past 25 years, they have mostly issued the reports after the deadline. Some of the factors contributing to the boards delivering the reports late may seem reasonable to agency officials and trustees. For example, investing time to make the report consistent with new legislation impacting Social Security or Medicare programs, or waiting for end of year data to be available, may make the report more useful than if it contained older information. However, other factors related to the management of the schedule for developing the reports, such as not formally tracking the reports’ progress or adjusting the schedule based on lessons learned in prior years, may have contributed to delays. Taking steps to improve the management of the report-development schedule would better position the trustees and agency officials to anticipate and plan for scheduling the spring boards meeting and to meet the statutory deadline in future years. Additionally, recognizing that there may continue to be instances in which the issuance of the reports will be delayed, establishing a policy to inform Congress of potential delays and factors contributing to those delays would enhance Congress’s ability to conduct oversight and make decisions about these important programs. Recommendations for Executive Action We are making the following two recommendations to the Secretary of the Treasury: The Secretary of the Treasury, as Chairperson of the Boards of Trustees, should work with the other trustees to take steps—in consultation with the chief actuaries of SSA and CMS—to improve the management of the report development schedule in order to provide the Trustees reports to Congress by the statutory deadline. These steps could include regularly updating the schedule using actual progress and archiving the final iteration of the schedules. (Recommendation 1) The Secretary of the Treasury, as Chairperson of the Boards of Trustees, should work with the other trustees to establish a policy to inform Congressional committees of jurisdiction when the trustees determine that the reports are expected to miss the issuance deadline. This outreach should include 1) the factors that are contributing to delays, and 2) the reports’ expected issuance dates. (Recommendation 2) Agency Comments and Our Evaluation We provided a draft of this report to the Secretary of the Treasury, the Secretary of Health and Human Services, the Secretary of Labor, and the Commissioner of Social Security for review and comment. Treasury and SSA provided formal written comments, and both agencies agreed with our recommendations. (See appendixes II and III.) Treasury, SSA, and HHS provided technical comments, which we incorporated as appropriate. DOL had no comments. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies to the appropriate congressional committees, the Secretary of the Treasury, the Secretary of Health and Human Services, the Secretary of Labor, the Commissioner of Social Security, and other interested parties. In addition, the report will be available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact Elizabeth Curda at (202) 512-7215 or curdae@gao.gov or James Cosgrove at (202) 512-7114 or cosgrovej@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: Assumptions Discussed in the 2019 Social Security and Medicare Trust Fund Reports Appendix II: Comments from the Department of the Treasury Appendix III: Comments from the Social Security Administration Appendix IV: GAO Contacts and Staff Acknowledgments GAO Contacts Staff Acknowledgments: In addition to the contact named above, Mark Glickman (Assistant Director), Gregory Giusto (Assistant Director), Paul Schearf (Analyst-in- Charge), Christie Enders, and Samuel Gaffigan made key contributions to this report. Additional assistance was provided by Bill Boutboul, Juana Collymore, Robert Dacey, Alex Galuten, Yvette Gutierrez, Janice Latimer, Emei Li, Sheila R. McCoy, Art Merriam, Mimi Nguyen, Stacy Ouellette, Oliver Richard, Joseph Silvestri, Dawn Simpson, Ardith Spence, Almeta Spencer, Frank Todisco, and Walter Vance.

Why GAO Did This Study The Social Security Act requires boards of trustees to issue reports to Congress by April 1 each year on the financial status of the Social Security and Medicare trust funds. Policymakers and others can use these reports to understand the programs' finances, conduct oversight, and consider legislative proposals for the programs. GAO was asked to review the timeliness of these reports. This report (1) describes how the boards of trustees develop the annual Trustees reports, and (2) examines the extent to which the boards of trustees have provided the reports to Congress by the April 1 deadline since 1995, and what factors account for any delays. GAO reviewed boards of trustees meeting minutes from 1995-2018, working group agendas from 2011-2018, and report development schedules and the annual Trustees reports from 1995-2019; as well as relevant federal law. GAO also interviewed agency working group officials from SSA and CMS; the Departments of Health and Human Services, Labor, and the Treasury; and eight former public trustees who served since 1995. What GAO Found Annual reports on the status of Social Security and Medicare trust funds are developed through a collaboration between agency officials and trustees, which include relevant Cabinet members and public members nominated by the President (if confirmed). Offices of the Chief Actuaries from the Social Security Administration (SSA) and the Centers for Medicare & Medicaid Services (CMS) submit data and draft reports to a working group of agency officials representing trustees and any public trustees. The working group reviews the information and, after gaining consensus, submits it to the boards of trustees for final approval. The boards of trustees send the final reports to Congress. The trustees missed the April 1 statutory deadline for submitting the reports to Congress in 17 of the 25 years from 1995 to 2019, and have issued them more than 2 months late in 6 of the last 10 years (see figure). According to agency officials and former public trustees GAO interviewed, factors that may account for delays include late-breaking changes to assumptions or data, and difficulty scheduling the boards' meetings. Additionally, contrary to GAO's guide on best practices for project schedules, officials have not taken steps to update the report-development schedules to reflect actual progress, maintained a formally documented baseline schedule to incorporate lessons learned from prior years, or notified Congress of their progress. Without taking steps to improve report-development schedule management, these trust fund reports will likely continue to be untimely, missing the April 1 statutory deadline. Also, without improved efforts to keep congressional committees informed, Congress will be unaware of when the reports will be issued, potentially hindering oversight of the trust funds. What GAO Recommends GAO recommends that Treasury take steps to work with the other trustees to improve schedule management for developing the annual Trustees reports, and to establish a policy to inform congressional committees of jurisdiction about expected delays in issuing the reports. Treasury agreed with the recommendations.

gao_GAO-20-250

gao_GAO-20-250_0

Background Within CBP, Border Patrol is responsible for securing U.S. borders and apprehending individuals arriving at the border between ports of entry. Also within CBP, OFO is responsible for inspecting travelers and cargo seeking to enter the United States through ports of entry and encounters or apprehends individuals determined to be inadmissible to the country. Upon apprehension of individuals at or between ports of entry, Border Patrol agents and OFO officers generally decide whether to (1) place apprehended adults and family units into expedited removal proceedings, or (2) initiate full immigration proceedings, according to CBP officials. If agents or officers place individuals into expedited removal proceedings, CBP will transfer them to DHS’s U.S. Immigration and Customs Enforcement (ICE) for longer-term detention (see appendix II for more information on eligibility, screening standards, and possible screening outcomes for credible and reasonable fear cases). Noncitizen adults and family units may make a fear claim in CBP custody at any point after apprehension, and during the pendency of their expedited removal proceedings in ICE custody (see appendix III for data on apprehensions of noncitizens placed into expedited removal who claimed fear of returning to their country, along with other characteristics of their cases). ICE is generally responsible for referring any fear claims to USCIS for a fear screening after individuals enter detention. If USCIS makes a negative determination and the determination is either not reviewed by an immigration judge, because the noncitizen has declined immigration judge review, or, if reviewed, is upheld by a reviewing immigration judge, ICE is then responsible for removing the person from the country. Within USCIS, the Refugee, Asylum, and International Operations Directorate (RAIO) is to provide, among other things, services for people who are fleeing oppression, persecution, or torture or facing urgent humanitarian situations. RAIO is made up of two divisions: the Asylum Division and the International and Refugee Affairs Division. USCIS’s Asylum Division is responsible for, among other responsibilities, adjudicating affirmative asylum applications—that is, claims made at the initiative of the individual who files an application for asylum with USCIS—and screening credible and reasonable fear cases. As of March 2019, USCIS had 546 asylum officers on board and eligible to screen credible and reasonable fear cases (out of 745 authorized positions). Asylum officers screen cases at the Asylum Prescreening Center in Arlington, Virginia, and eight asylum offices nationwide. USCIS established the Asylum Pre-Screening Center in fiscal year 2016 to provide additional support for the credible and reasonable fear caseload. As of April 2019, the Asylum Pre-Screening Center and the Arlington asylum office together had jurisdiction over 27 ICE detention centers across the United States. EOIR is responsible for conducting immigration court proceedings, appellate reviews, and administrative hearings to fairly, expeditiously, and uniformly administer and interpret U.S. immigration laws and regulations. As of September 30, 2019, 442 immigration judges presided over EOIR’s 63 immigration courts nationwide. In addition to removal proceedings, immigration judges also conduct certain other types of hearings, such as the review of negative credible fear determinations. Table 1 provides additional information about DHS’s and DOJ’s roles in the credible and reasonable fear processes. In July 2019, USCIS made several changes to its credible fear screening processes in response to an interim final rule implementing a new mandatory bar to asylum, known as the “third country transit bar.” Under the interim final rule, noncitizens who enter, attempt to enter, or arrive in the United States across the southern land border on or after July 16, 2019, after transiting through at least one country outside their country of citizenship, nationality, or last lawful habitual residence en route to the United States, must be found ineligible for asylum unless they demonstrate that they fall under an exception to the third country transit bar. USCIS’s and EOIR’s Credible and Reasonable Fear Caseloads Generally Increased from Fiscal Years 2014 through 2018, and a Majority of USCIS Screening Outcomes were Positive USCIS’s Credible and Reasonable Fear Caseload Nearly Doubled from Fiscal Years 2014 through 2018 As shown in table 2, USCIS’s credible and reasonable fear caseloads nearly doubled from fiscal year 2014 (over 56,000 referrals to USCIS) to fiscal year 2018 (almost 109,000 referrals)—the most recent full year of USCIS data available at the time of our analysis. From fiscal year 2014 through the first two quarters of fiscal year 2019, referrals to USCIS for credible fear screenings comprised about 89 percent of all credible and reasonable fear referrals. The number of referrals for credible fear screenings in the first two quarters of fiscal year 2019 alone was larger than the total number of referrals in each of fiscal years 2014 and 2015. Referrals for reasonable fear screenings also increased from fiscal years 2014 to 2018, and comprised between 9 and 15 percent of all referrals during that time period. Appendix III contains additional information on the characteristics of credible and reasonable fear applicants from fiscal year 2014 through March 2019. A Majority of Credible and Reasonable Fear Referrals to USCIS from Fiscal Year 2014 through the First Two Quarters of Fiscal Year 2019 Resulted in Positive Determinations As shown in figure 1, USCIS asylum officers made positive determinations in about 71 percent of all credible and reasonable fear screenings between fiscal years 2014 and the first two quarters of fiscal year 2019. The remaining credible and reasonable fear screenings were almost evenly divided between negative determinations and administrative closures (approximately 14 percent each) with a small remainder of screenings pending resolution (0.1 percent). Individually, from fiscal year 2014 through the first 2 quarters of fiscal year 2019, USCIS asylum officers made positive determinations in nearly 77 percent of all credible fear screenings; officers made positive determinations in about 30 percent of reasonable fear screenings. Regarding credible fear screenings specifically, the percentage of positive determinations ranged from about 73 to 80 percent of total credible fear cases completed each year from fiscal year 2014 through the first two quarters of fiscal year 2019 (see fig. 2). Regarding reasonable fear screenings, as shown in figure 3, outcomes for reasonable fear cases from fiscal year 2014 through the first two quarters of fiscal year 2019 were generally split evenly each year among positive determinations (from 28 to 32 percent), negative determinations (from 29 to 35 percent), and administrative closures (from 35 to 42 percent). EOIR Reviewed Over 50,000 USCIS Credible Fear Decisions from Fiscal Year 2014 through the First Three Quarters of Fiscal Year 2019; Immigration Judges Upheld Most Decisions EOIR’s credible and reasonable fear workload increased by about 16 percent—from about 8,100 reviews to about 9,400 reviews each year— between fiscal year 2014 and fiscal year 2018. According to EOIR data, from fiscal year 2014 through the third quarter of 2019 (the most recent data available at the time of our analysis), EOIR’s immigration judges, at the noncitizens’ requests, reviewed about 55,000 cases in which USCIS asylum officers made a negative credible or reasonable fear determination (see figure 4). Approximately 10 percent of these reviews were for individuals detained at the Karnes, Dilley, or Berks family residential centers. As shown in figure 4, immigration judges upheld USCIS’s negative credible and reasonable fear determinations in 77 percent of all reviews judges conducted from fiscal year 2014 through the third quarter of fiscal year 2019. During this time period, immigration judges vacated (or overturned) 22 percent of USCIS’s negative determinations—meaning, judges found that those individuals had a credible or reasonable fear, as appropriate. As a result, individuals found to have a credible fear were to be placed in full removal proceedings and individuals found to have a reasonable fear were to be placed into more limited removal proceedings to consider the applicants’ eligibility for withholding of removal or deferral of removal. Immigration judges upheld 45 percent of USCIS’s negative determinations and vacated 54 percent of USCIS’s negative determinations for individuals in ICE’s Dilley, Karnes, or Berks family residential centers. In addition, EOIR publicly reports data on the outcomes of removal cases across immigration courts that originated with a positive credible fear determination. EOIR reported that, from fiscal years 2014 through March 2019, immigration judges completed about 135,000 cases that began with a positive credible fear determination. Individuals in about 75,800 of the completed removal cases filed applications for asylum (56 percent). In about 59,200 of the completed removal cases (44 percent), individuals did not file an asylum application. However, as previously described, individuals who have received positive credible fear determinations may apply for other forms of relief or protection besides asylum, such as withholding of removal, and those applications are not represented in the statistics on EOIR’s website. Further, EOIR officials told us that, for data reporting purposes, each member of a family who receives a Notice to Appear before an immigration judge is counted as one EOIR removal case and each removal case may or may not include an asylum application. However, for a number of immigration applications before the court, including asylum and the related screening for credible fear, a spouse or child (defined as an unmarried natural or legally adopted child under 21 years of age) may be included as a dependent on a principal’s application and derive lawful immigration status from the principal applicant if the application is granted. As previously discussed, individuals detained in family residential centers—including individuals who could be eligible dependents for credible fear screening and asylum application purposes—comprise a substantial proportion of those who receive positive credible fear determinations. As such, according to EOIR officials, each family member would not be expected to file a separate asylum application. For example, a mother and her two children whose removal cases originated with a positive credible fear screening would comprise three removal cases in EOIR’s publically reported data, but it is likely that only the mother’s case would include an application for asylum, with her children as dependents on that application. For those removal cases in which the noncitizen applied for asylum, immigration judges granted asylum in about 19,300 cases (25 percent of the 75,800 completed removal cases with an asylum application). USCIS Has Policies and Procedures for Overseeing Credible and Reasonable Fear Screenings, but Gaps in Training, Quality Assurance, and Family Processing Exist USCIS Has Policies and Procedures for Managing and Overseeing the Credible and Reasonable Fear Screening Process, Including Requiring Supervisory Review of All Cases USCIS has developed various policies and procedures related to managing and overseeing credible and reasonable fear cases in accordance with the regulations governing credible and reasonable fear screenings, including setting requirements for interview procedures, background and security checks, and supervisory review. In particular, USCIS has a Credible Fear Procedures Manual and a Reasonable Fear Procedures Manual that outline the procedures officers are to follow in screening these cases. Interview procedures. As of July 2019, an asylum office is to wait a minimum of one full calendar day from the applicant’s arrival at an ICE detention facility before conducting a credible fear interview; an asylum office is to wait 48 hours after an initial orientation on the reasonable fear process before a reasonable fear interview, according to USCIS policy. However, both credible and reasonable fear interviews generally occur at least 48 hours after the applicant’s arrival at a detention facility, according to USCIS officials. Asylum officers may conduct credible and reasonable fear interviews either in-person or on the phone. Asylum officers are to arrange the assistance of an interpreter, generally connected over the phone, if the applicant is unable to proceed effectively in English pursuant to regulation. Asylum officers are to verify and document that applicants have received and understood information regarding the credible or reasonable fear process before they begin asking substantive questions during the interview about the applicant’s claim. According to USCIS documents and officials, during the interview, asylum officers are to elicit all information relevant to a credible or reasonable fear claim, and regulation requires they conduct interviews in a non- adversarial manner. For example, asylum officers are to ask applicants questions to determine whether they can establish a credible or reasonable fear of persecution based on their race, religion, nationality, membership in a particular social group, or political opinion. In addition, asylum officers are to ask applicants questions to determine whether they can establish a credible or reasonable fear of torture if returned to their home country. During our observations of in-person and telephone interviews, we observed asylum officers asking questions to ensure they fully explored any aspect of the claim related to a protected ground that could result in a positive determination. For example, we observed asylum officers asking applicants separate questions about each protected ground, even if the applicant had not previously expressed they were harmed because of their political beliefs or race. USCIS policy notes the applicant’s credibility is dependent on various factors such as comparing information provided during the interview with that previously provided in the applicant’s sworn statement to Border Patrol or OFO when initially apprehended. If asylum officers identify an issue with the applicant’s credibility, they are to inform the applicant of the concerns and ask the applicant for his or her perspectives. During our site visits, we observed asylum officers questioning applicants on inconsistencies, in a non-adversarial manner, between information provided during the interview as compared to the applicant’s sworn statements to Border Patrol agents upon apprehension. At the end of the interview, asylum officers are to provide a verbal summary of the material facts of the applicant’s claim, and provide an opportunity for the applicant to make any corrections or additions. We observed asylum officers providing such summaries in all but one of the interviews that we observed in full. According to USCIS policy, asylum officers are to record key information about the applicant’s claim, as well as specific details of the determination, on required forms that serve as the official record of the credible or reasonable fear screening. In addition, asylum officers use a “checklist” to record more detailed legal analysis related to the applicant’s claim. Asylum officers also generally type notes during interviews in a question and answer format, capturing each question and follow-up question they ask, and each response the applicant provides. We observed asylum officers documenting interviews in this way during all of the interviews where we observed the asylum officer in person. Background and security checks. USCIS policy requires asylum officers to ensure certain background and security checks are conducted. If security checks or information discovered during the interview raises concerns related to fraud, public safety, or national security, asylum officers are to refer the case to USCIS’s Fraud Detection and National Security Directorate (FDNS) for assistance. FDNS officials told us the short time frames in the credible and reasonable fear process, among other factors, make direct involvement in individual cases less likely than in other caseloads at USCIS, such as affirmative asylum cases. As such, the scope and extent of FDNS investigations into credible and reasonable fear cases is limited relative to other USCIS caseloads. FDNS data indicate that asylum officers referred approximately 1,400 total credible and reasonable fear cases to FDNS between fiscal years 2017 and 2018. Of those, 13 cases resulted in a formal finding, called a Statement of Finding. FDNS officials told us referrals from asylum officers on credible and reasonable fear cases typically result in FDNS conducting research related to an applicant’s criminal history or travel patterns. FDNS may refer this information, in turn, to ICE to reference in the applicant’s removal proceedings, as appropriate. In contrast, according to FDNS officials, a fraud referral in the affirmative asylum context may result in a more formal finding of fraud in a Statement of Finding. Supervisory review. USCIS oversight of credible and reasonable fear cases includes a required supervisory review of each case after an asylum officer makes a positive or negative determination. USCIS officials said supervisors are to review cases for legal sufficiency and accuracy, including a review of the screening checklist and the asylum officer’s supporting interview notes. According to officials, supervisors are to communicate the results of their review to the asylum officer informally (e.g., via email or in-person discussion) for small issues, such as an administrative error, or through a formal write-up for larger issues, such as if the asylum officer’s legal analysis was insufficient and requires a second interview with the applicant. USCIS Provides Initial Training on Credible and Reasonable Fear to New Asylum Officers and Asylum Offices Are Required to Provide Ongoing On-the-Job Training USCIS oversight of credible and reasonable fear cases includes basic training for new asylum officers and ongoing training for incumbent officers at asylum offices; these trainings include information specific to credible fear and reasonable fear screenings. As of the time of our review, the initial training program for asylum officers is comprised of two main components: Distance Training. New asylum officers participate in 3 weeks of self- paced RAIO Directorate and Asylum Division distance training in their respective asylum offices. During distance training, asylum officers are expected to participate in webinars, read the training materials and complete exercises and quizzes in preparation for residential training. The Asylum Division distance training includes course readings on credible and reasonable fear, and observations of credible and reasonable fear interviews. Residential Basic Training. Asylum officers participate in a 6-week residential basic training program, which includes 3 weeks of training in issue areas common across USCIS’s Refugees, Asylum, and International Operations Directorate, as well as three weeks of Asylum Division-specific training. In the first 3-week session, courses include classroom instruction, practical exercises, and interviewing exercises on a variety of topics and skills relevant to multiple areas of USCIS’s work, such as on affirmative asylum and refugee adjudications. The legal topics and skills covered in this initial training include eligibility for asylum, an applicant’s nexus to protected grounds, and eliciting testimony, among others. The second 3-week session focuses on division-specific policy, procedure, and law related to asylum adjudications and screenings. For example, the 3-week session includes training on the affirmative asylum process, and multiple mock affirmative asylum interviews, among others, as well as 2 days of training specific to credible and reasonable fear cases. These 2 days include practical exercises; one mock credible fear interview exercise; and formal presentations on interviewing skills and security checks in a credible fear context, forms required for credible and reasonable fear, and on the Convention against Torture. At the end of the 6-week residential training course, new asylum officers must pass final exams with a score of at least 70 percent. We reviewed a version of the exam and found that it included questions specific to credible and reasonable fear screenings. Asylum Division officials said the 9 combined weeks of distance and residential basic training constitute the minimum amount of formal training required for asylum officers to effectively screen credible and reasonable fear cases. However, Asylum Division officials said it is important for individual asylum offices to provide additional, on-the-job training to new officers assigned to screen credible and reasonable fear cases, specifically. Asylum officers screen credible and reasonable fear cases under shorter time frames and with less corroborating documentation compared to affirmative asylum cases. As such, Asylum Division officials told us that officers accustomed to adjudicating affirmative asylum cases may need to adjust to the shorter time frames required in credible and reasonable fear cases. For example, some asylum offices have developed formal presentations on local policies and procedures, or provide officers with an opportunity to observe other officers conducting credible or reasonable fear interviews and gradually increase the number of cases they screen per day. Given their caseloads, the Houston and Arlington asylum offices provide 3 and 4 weeks of additional credible and reasonable fear training for new asylum officers, respectively. By comparison, the San Francisco and Newark asylum offices provide 1 week of training on credible and reasonable fear procedures for new asylum officers and Los Angeles provides 2 days of such training, according to officials. For incumbent asylum officers, USCIS policy requires asylum offices to allocate four hours per week for formal or informal training. The training can range from classroom instruction by a training officer, to individual study time that asylum officers can use to review case law, research country conditions affecting asylum applicants, or read new USCIS procedures and guidance. Individual asylum offices design their weekly training programs based on the types of cases their office generally receives, according to Asylum Division officials. The Asylum Division requires training officers to track the date and topic of each weekly training session and report that information to Asylum Division headquarters on a quarterly basis. Our analysis of fiscal year 2018 quarterly training reports for all asylum offices and sub-offices indicates that offices with larger credible and reasonable fear caseloads generally provided more weekly trainings on these topics. For example, Houston and Arlington conducted seven or more weekly training sessions on credible and reasonable fear screenings in fiscal year 2018. By comparison, two offices with smaller credible and reasonable fear caseloads—Newark and New York—conducted one or fewer weekly sessions on credible and reasonable fear (see app. III for credible and reasonable fear workload data by asylum office). In addition to this training program for asylum officers, USCIS trains officers from outside the Asylum Division to screen credible and reasonable fear cases, including refugee officers and others. Refugee officers receive some of the same basic training as asylum officers, as they participate in the same RAIO distance training and RAIO Directorate residential training. Refugee officers do not participate in Asylum Division distance training or residential training. As a result, USCIS provides refugee officers with 3 days of training on screening credible fear cases before they can begin screening cases. We reviewed training materials for the refugee officer training, and found the sessions are similar to Asylum Division residential training sessions on credible fear screening. In addition, some materials provide information and guidance on the differences between adjudicating refugee cases and screening credible fear cases. Officials said refugee officers generally screen credible fear cases, including at the family residential centers, only if they are detailed to the Houston and Arlington asylum offices. Both Houston and Arlington provide refugee officers detailed to their offices with 1-2 weeks of additional training on credible fear screening, similar to the procedural training they provide to new asylum officers. At both offices, trainings include formal presentations or exercises on legal concepts and procedures specific to credible fear, credible fear interview observations, and a gradual increase in the number of cases refugee officers screen each day. Pre-departure Training for USCIS Asylum Officers Screening Family Units at ICE Family Residential Centers Is Inconsistent Across Asylum Offices Although all new asylum officers receive basic training on the credible and reasonable fear screening process and may also receive on-the-job training in their home offices, not all offices provide additional pre- departure training to asylum officers before they begin screening cases for family units at ICE family residential centers. Credible fear screenings at ICE’s family residential centers, in particular, represent a significant percentage—about 34 percent—of all credible fear cases asylum officers screened from fiscal year 2014 though the second quarter of fiscal year 2019. As discussed previously, asylum offices with relatively small credible and reasonable fear local caseloads generally provide less on- the-job training throughout the year on credible and reasonable fear. However, almost all asylum offices send officers to the family residential centers in Texas for in-person interviews, including those offices with small credible and reasonable fear caseloads at the local level. Asylum Division officials said they require asylum offices to send a specific number of asylum officers—a number in proportion to the size of the office—with the largest offices sending the most officers to the family residential centers each year. For example, in fiscal year 2018, Newark, Los Angeles, Houston, and Chicago sent the most officers to the family residential centers, as shown in figure 5 below. At least two asylum offices provide pre-departure training to asylum officers being sent to ICE’s family residential centers. To support officers who are more accustomed to adjudicating affirmative asylum cases, the Los Angeles asylum office provides pre-departure training for officers before they travel to the family residential centers. In Los Angeles, officers observe credible and reasonable fear interviews and gradually increase to a full caseload of credible or reasonable fear cases at their home office, according to officials. In San Francisco, officers receive pre- departure training highlighting procedures unique to the family residential centers or to processing family units in credible and reasonable fear. Specifically, San Francisco pre-departure training includes a formal presentation on family residential center procedures, including discussion of challenges officers may experience, according to officials. By comparison, officials from the Chicago and New York offices told us they do not provide formal pre-departure training but rather direct or recommend that officers review Asylum Division guidance and procedures on family processing independently before they travel. Officials from two other offices told us they rely on the training asylum officers may receive throughout the year related to credible and reasonable fear, which can vary, as previously discussed. Asylum officers also noted inconsistent pre-departure training prior to their temporary duty during our February 2019 visits to two Texas family residential centers in Dilley and Karnes. For example, some asylum officers we interviewed said they screened credible and reasonable fear cases at their home office in preparation for their assignment. Others said they reviewed procedures independently on family processing in credible and reasonable fear cases. Officers from one asylum office said they relied primarily on an email from USCIS support staff located at the family residential centers to learn about screening family cases. Asylum officers are to review the procedures on family processing in credible and reasonable fear before they arrive at the family residential centers, according to officials. However, there is no minimum amount of pre-departure training, or required content for such training, that all asylum offices are to provide before officers begin screening family units. Asylum Division officials acknowledged that training on screening of family units for credible and reasonable fear varies by asylum office and noted that offices have been given discretion to determine what, if any, pre-departure training to provide on screening family units. Arlington and Houston asylum office officials stated that inconsistent asylum officer training on credible and reasonable fear cases negatively impacts efficiency at the family residential centers. Specifically, these officials noted that asylum officers who typically adjudicate affirmative asylum applications benefit from training on key differences between credible fear, reasonable fear, and affirmative asylum. For example, officials said more training could reduce administrative errors in applicants’ paperwork, and changes needed during supervisory review, both of which occur more often for officers with less experience and training, according to officials. Further, officials said asylum officers less experienced in credible and reasonable fear may not be able to handle a full caseload at the family residential centers when they first arrive. As a result, Houston officials said they may spend the first week of a 2-week assignment providing additional support to inexperienced officers as they gradually increase to a full caseload. Standards for Internal Control in the Federal Government states management should demonstrate a commitment to recruit, develop, and retain competent individuals. The standards also note that competence is the qualification to carry out assigned responsibilities, and requires relevant knowledge, skills, and abilities, which are gained largely from professional experience, training, and certifications. As previously noted, Asylum Division officials told us additional training for asylum officers before they begin screening cases at the family residential centers is important. Officials also said their intention is to balance such training against the need for rapid deployment, in some cases. Although additional training may not be feasible before every deployment, providing asylum officers additional pre-departure training before they begin screening credible and reasonable fear cases for family units would better prepare officers and help ensure efficient and effective case processing at ICE’s family residential centers. USCIS Conducts Various Quality Assurance Reviews of Credible and Reasonable Fear Cases, but Does Not Document Results in a Consistent Manner USCIS relies primarily on two quality assurance reviews for assessing quality of credible and reasonable fear cases, but does not document the results of one of these reviews in a consistent manner. Annual, Asylum Division-wide reviews. The Asylum Division conducts division-wide quality assurance reviews on a random sampling of credible fear, reasonable fear, or affirmative asylum cases selected proportionally from asylum offices nationwide. To do so, the Asylum Division works in collaboration with the RAIO Directorate. The reviews occur each year, and rotate between a sampling of credible fear cases, reasonable fear cases, and affirmative asylum cases. For credible and reasonable fear reviews, USCIS randomly selects a specified number of cases after supervisory review, but before officers serve determinations to the applicant. Reviewers use a checklist to identify and track quality issues arising in each reviewed case, such as accurate data entry, appropriate legal analysis, asylum officer notes that reflect a skilled interview, and others. The review process for each case includes two lines of review. If the two reviewers come to different conclusions, they discuss any differences in their reviews and reach consensus about how to score the case. USCIS records the results of these reviews in a document that lays out the numbers and percentages of errors in areas covered in the review checklist. For example, for the 2018 review of credible fear cases, the document states asylum officer notes did not reflect a skilled interview in an estimated 58 percent of cases. For most of these cases, the reason for the error that reviewers noted was insufficient follow-up questions. USCIS officials said while the sample across asylum offices is generalizable with respect to the credible or reasonable fear caseloads nationwide, the samples taken from each asylum office are not large enough to draw conclusions about trends at individual asylum offices. Officials said they rely on periodic reviews to identify trends by asylum office. Periodic, asylum office reviews. In addition to the Asylum Division-wide quality assurance reviews, the Asylum Division began conducting periodic reviews at asylum offices in November 2017. As of November 2019, the Asylum Division had conducted periodic reviews of credible and reasonable fear cases or affirmative asylum adjudications at some asylum offices, as well as a review of credible and reasonable fear cases at the family residential centers. For periodic reviews, the Asylum Division selects cases over a period of several weeks. For example, based on the Asylum Division’s draft standard operating procedures for the periodic reviews, an asylum office may send two to four credible and reasonable fear cases every day for several weeks to reach the required total number of cases. According to the draft standard operating procedures, Asylum Division reviewers are to use a reviewer checklist, modeled off the checklist used for the Asylum Division-wide reviews, as a starting point for what factors to review. However, USCIS does not document the results of the periodic reviews in a consistent manner. We reviewed the reports resulting from six periodic reviews conducted at the Arlington, Chicago, Miami, and Houston asylum offices, the New Orleans sub-office, and the family residential centers. We found that all reports included information on strengths and weaknesses, and some reports further organized analysis into additional categories. For example, some reports had analysis on details related to procedures, eliciting testimony, and issues related to fraud detection and national security. Other reports included analysis of specific trends in persecution cases and in Convention against Torture cases. Some reports also included analysis on legal sufficiency, applicant country of origin, determination outcomes, and others. According to the draft standard operating procedure, reviewers are to note trends, common errors, and collect samples to create a deliverable, such as a short report or other deliverables, for the asylum office at the end of the review. However, the Asylum Division has not provided guidance on what specific information is important to include in reports resulting from periodic reviews in order to track trends within an asylum office over time, or across asylum offices. Standards for Internal Control in the Federal Government states management should establish and operate monitoring activities to oversee the internal control system and evaluate the results. Management should document the results of ongoing monitoring and separate evaluations to identify internal control issues, and should use this evaluation to determine the effectiveness of the internal control system. Asylum Division officials told us the primary purpose of the periodic reviews is to collect information about current, office-specific trends, and provide timely support in the form of training sessions and other guidance. Further, the Asylum Division historically has not used the periodic reviews to compare one office to another, though they have sometimes noted issues from these reviews requiring similar guidance across multiple offices. Documenting the results of periodic reviews in a consistent manner would help the Asylum Division identify trends and provide support across asylum offices. The draft standard operating procedures for the periodic review provides general directions for reviewers to share information on trends to asylum office personnel, such as strengths, weaknesses, and other developing trends. However, the draft standard operating procedures do not specify requirements for documenting the results of these reviews. Asylum Division officials told us the periodic review standard operating procedures are in draft form, and that they may provide more specific guidance on aspects of the reviews in the future. However, officials also said they are not planning any changes or additions to the standard operating procedure as of September 2019. More specific guidance on requirements for documenting results would better position USCIS to track trends in a consistent manner for credible and reasonable fear reviews within and across asylum offices. USCIS Does Not Systematically Record Case Outcomes When Screening Family Members for Credible Fear By regulation, dependents, specifically a spouse or child, of a noncitizen (referred to as the “principal applicant”) can be included in the applicant’s credible fear determination if the dependent (1) arrived in the United States concurrently with the principal applicant, and (2) desires to be included in the principal applicant’s determination. However, any noncitizen may have his or her credible fear determination made separately, if he or she expresses such a desire. USCIS policy is to include any dependents on a principal applicant’s credible fear determination if the principal applicant receives a positive determination, resulting in both the principal applicant and any dependents being issued a Notice to Appear for full removal proceedings. For example, USCIS may process credible fear cases together for family units detained at ICE’s family residential centers, including children as dependents on a parent’s case, or issuing a Notice To Appear for the parent and children in the interest of family unity (see figure 6). We observed asylum officers at the family residential centers asking principal applicants whether they were apprehended with any family members. If yes, asylum officers asked for the names and dates of birth of those family members, and recorded the information in their typed notes. For parents who received a positive determination, we observed asylum officers including the child on the parent’s case as a dependent. Further, if a parent receives a negative credible fear determination, and his or her child receives a positive credible fear determination, USCIS may issue a Notice to Appear to the child as a positive credible fear determination and to the parent in the interest of family unity. In that case, because a parent could not be a “dependent” of a child under the regulation, USCIS policy is to use its discretion to issue a Notice to Appear to both the child receiving a positive determination and the parent he or she arrived with, in the interest of family unity, even though the parent initially received a negative determination. Issuing the parent and child a Notice to Appear places them into full removal proceedings where they can apply for multiple forms of relief or protection before an immigration judge, including asylum, rather than being expeditiously ordered removed in accordance with the expedited removal process. The exercise of this discretion to issue a Notice to Appear to both the child receiving a positive determination and the parent he or she arrived with in the interest of family unity is limited to cases in which the children are under the age of 18 because, according to the policy, family unity interests are more compelling when the child is a minor. USCIS data indicate that asylum officers screened more than 141,000 credible fear cases at ICE’s four family residential centers between fiscal years 2014 and the first two quarters of 2019 (see table 3). In addition, USCIS data indicate that positive credible fear determination rates are higher at the family residential centers—87 percent compared with the nationwide rate of 77 percent from fiscal year 2014 through the second quarter of fiscal year 2019 (see app. III for data on reasonable fear cases screened at ICE’s family residential centers). Asylum officers are to record individual case outcomes for all family members in USCIS’s automated case management system. However, Asylum Division officials said their system does not allow asylum officers to record whether an individual receives a credible fear determination as a principal applicant, dependent, or in the interest of family unity. Instead, asylum officers are to record positive determinations in the USCIS case management system for both (1) dependents on the basis of the principal applicant’s positive case, and (2) parents with negative determinations, on the basis of their child’s positive case. USCIS does record more specific information related to outcomes for family units in the family members’ individual hardcopy alien files, but this information is not readily available in an automated manner. USCIS’s case management system allows asylum officers to record family relationships—that is, officials stated that asylum officers are to record who is a principal applicant, and who is a spouse, child, parent, or sibling of the principal applicant. According to USCIS officials, asylum officers are to record a parent who receives a positive credible fear determination as the principal applicant, and record any children as a child. Further, asylum officers are to link known family members’ cases in the system, but adding a description of the family relationship is up to asylum officers’ discretion. However, the system does not allow officers to record whether an applicant’s determination stems from his or her own case, or from a family member’s case. As a result, USCIS does not maintain automated data in a readily accessible manner on outcomes for family members in a manner that indicates whether (1) an eligible family member received a positive determination as a dependent on a principal applicant’s positive case or (2) whether a parent was issued a Notice to Appear based on his or her child’s positive determination, after the parent received a negative determination. Standards for Internal Control in the Federal Government states that management should process obtained data into quality information that supports the internal control system. Quality information is appropriate, current, complete, accurate, accessible, and provided on a timely basis. Although USCIS data indicate that positive credible fear determination rates are higher at the family residential centers compared to rates of positive credible fear determinations across all detention facilities, USCIS officials stated the higher rates result from the ability to (1) include children under 18 on a parent’s positive credible fear determination, and (2) record all family members as positive in the system when USCIS uses its discretion in the interest of family unity. USCIS officials told us that systematically recording all outcomes of credible fear screenings for family members in a more complete manner would require changes to their case management system; according to Asylum Division officials, they are continually exploring options to improve the system’s capabilities. Without complete data in its case management system on all outcomes of credible fear screenings at family residential centers, USCIS is not well-positioned to report on the scope of either the agency’s policy for family members who are treated as dependents, pursuant to regulation, or USCIS’s use of discretion in the interest of family unity. USCIS and EOIR Have Processes for Managing Credible and Reasonable Fear Workloads, but USCIS Does Not Have Complete Data on Case Delays USCIS Makes Staffing Allocation Decisions Based on National and Local Staffing Models USCIS manages its credible and reasonable fear workloads using national- and local-level staffing models to inform staffing allocation decisions. Specifically, USCIS has an agency-wide staffing model to allocate staff to different workload categories, including credible and reasonable fear workloads, for each upcoming fiscal year. Asylum Division headquarters officials stated they collaborate with USCIS’s Office of Performance and Quality, USCIS administrative offices, and local asylum offices to develop the national staffing model. Headquarters officials said the national staffing model is intended to allocate staff for each workload category for the upcoming fiscal year. They begin working on the staffing model in June for any given year in anticipation of resource decisions the agency will make before a new fiscal year begins, usually in September or October. According to officials, the national staffing model for credible and reasonable fear is based on historical case receipt and workload data, historical staffing data and future staffing workload forecasting data, bi-weekly reports on staffing and workload data, and observations from asylum offices submitted to headquarters, among other things. The Asylum Division also maintains staffing models that guide local staffing deployment, according to headquarters officials. Asylum offices make local staffing allocation decisions in collaboration with the Asylum Division at headquarters. Headquarters officials stated they consider several factors in allocating staff specifically for the credible and reasonable fear workloads for local asylum offices. Such factors include workload projections, available facilities and planned facilities projects, and existing workforce and vacancy levels (table 4 shows the number of asylum officers authorized and on board for each asylum office in March 2019). In addition, headquarters officials said they work with local asylum offices to review the number of credible and reasonable fear case receipts and current staffing allocations by asylum office on a daily basis. Headquarters officials stated they change staffing allocation as necessary to address changes in credible or reasonable fear case receipts. In the Houston and Arlington asylum offices, in particular, officials stated they review headquarters data on workload projections to assign personnel to the credible and reasonable fear workloads for their offices. For example, officials in the Houston office said they look at the projected credible and reasonable fear workload to determine the number of officers they may need. Houston office officials said they assign officers based on officer availability, considering factors such as leave or training schedules. Similarly, a senior official responsible for staffing in the Arlington office said they look at the projected numbers of credible and reasonable fear cases, as well as the location of the cases, to determine the target number of officers assigned to a specific workload. Once they set targets for the number of officers needed, a senior official responsible for staffing in the Arlington office said they assign personnel based on a number of factors, including officer preferences, seniority, and locations with the greatest need. Although USCIS uses national and local staffing models for determining staffing needs and allocating staff at and across field offices, senior Asylum Division officials stated that predicting future workload for credible fear cases is challenging. Moreover, headquarters officials told us that USCIS’s credible fear workload projections have been off by as much as 50 percent when comparing projected and actual credible fear workload volume in recent years. For example, the USCIS projections for credible fear cases in fiscal year 2015 were 78,485, but actual case receipts totaled 48,052. More recently in fiscal year 2018, USCIS projected 70,000 credible fear case receipts, but actual case receipts totaled 99,035. Headquarters officials stated that a variety of external factors— unpredictable changes in country conditions, and CBP and ICE decisions to either place individuals in expedited removal or issue Notices to Appear before an immigration judge—make it difficult to project this workload. Furthermore, headquarters officials stated the volume of credible fear cases can fluctuate on a weekly basis, while reasonable fear projections have been fairly accurate, since the number of reasonable fear cases has remained relatively stable in recent years. To manage its workload of credible and reasonable fear cases, USCIS relies on a flexible workforce to respond to fluctuations in cases, in addition to asylum officers who generally screen credible and reasonable fear cases. For example, USCIS pulls asylum officers from affirmative asylum adjudications and uses overtime hours to handle surges in credible and reasonable fear case receipts, according to officials. Senior Asylum Division officials stated they do not receive staffing increases to account for lost or stopped work in other workload categories, such as affirmative asylum, that result from surges in credible fear case receipts. They stated surges in credible and reasonable fear case receipts may require immediate staff redeployment from the affirmative asylum workload. As a result, asylum offices have sometimes canceled planned affirmative asylum interviews and have prioritized credible and reasonable fear screenings over affirmative asylum cases, which have significantly contributed to the current backlog in pending affirmative asylum cases, according to headquarters officials. As previously discussed, asylum offices across the country also send officers on details to ICE’s family residential centers to conduct credible and reasonable fear screenings. In addition, the Asylum Division headquarters tracks the number of asylum officers assigned to the credible and reasonable fear workload, among other workload categories such as affirmative asylum, through biweekly reports received from local asylum offices. Headquarters officials told us they use the reports to respond to specific requests for information about Asylum Division staffing allocation. For example, Congress may request information on the number of USCIS personnel working on credible fear cases for a particular time period, according to headquarters officials, so they maintain these reports to fulfill such requests. Specifically, with regard to the biweekly reports, asylum offices record the number of asylum officers assigned to credible and reasonable fear cases for each day in the 2-week pay period. The resulting biweekly reports are spreadsheets with 15 tabs, one tab for each day in a pay period and one tab summarizing the pay period, with 26 separate spreadsheets for each year per asylum office. Headquarters officials stated the biweekly reports are manually compiled and may contain errors, but the biweekly reports have historically provided the overall number of personnel performing credible and reasonable fear work for any particular date or pay period. As of October 2019, headquarters officials said they are developing automated software that will track information similar to that collected in the biweekly reports, which will allow more systematic analysis of the staffing data that the current biweekly reports contain. USCIS Monitors Credible and Reasonable Fear Processing Times to Help Manage Its Workload USCIS sets and monitors timeliness goals for completing credible and reasonable fear cases. Monitoring timeliness goals for credible fear cases. USCIS monitors credible fear processing times by setting timeliness goals for completing credible fear cases and those goals have changed over time. USCIS regulation does not require that credible fear cases be completed in a specific time frame; however, Asylum Division headquarters officials said they have used timeliness goals to help monitor their credible fear workload. In addition, case delays may occur for credible fear cases (discussed further below). Specifically, from fiscal year 2009 through the first quarter of fiscal year 2018, USCIS used a 14-day goal to monitor credible fear case processing times. In other words, USCIS monitored the extent to which officers completed credible fear cases within 14 calendar days of USCIS receiving referral documents from ICE and created an electronic file for the case in their case management system. According to our analysis, USCIS completed at least 81 percent of credible fear cases in 14 or fewer days for each fiscal year from 2014 to 2017—the last full fiscal year under the 14-day goal (see table 5). In February 2018, USCIS lowered its credible fear processing time goal to 10 days. USCIS completed 68 percent of credible fear cases in 10 or fewer days between February and September 2018. Monitoring timeliness requirements for reasonable fear cases. USCIS monitors reasonable fear processing times by setting a 10-day goal. Pursuant to regulation, asylum officers are to conduct reasonable fear interviews and make a determination within 10 days of receiving a referral from CBP or ICE with an indication that the individual has made a fear claim, absent exceptional circumstances. Additionally, a 2015 settlement agreement in the Alfaro-Garcia v. Johnson case (“Alfaro- Garcia” Settlement Agreement) requires USCIS to achieve an average national reasonable fear determination period of no more than 10 court days (i.e. business days), calculated on a monthly basis, for cases in which individuals are detained by DHS. For reasonable fear cases subject to this settlement agreement that take longer than 20 court days to complete, asylum offices are to notify the Chief of the Asylum Division in writing and provide an explanation for the delay. Further, USCIS must provide class counsel in the Alfaro-Garcia case a notice and remedial plan of action for cases that exceed 20 days that are subject to that settlement agreement. Consistent with USCIS policy and the Alfaro-Garcia settlement agreement, officers may pause the clock for reasonable fear cases— and thus case processing times—in the following limited circumstances: the applicant or the applicant’s representative requests to defer the reasonable fear interview; the applicant refuses to participate in the reasonable fear interview or accept service of a reasonable fear determination; or exceptional circumstances. USCIS pauses processing times for detained reasonable fear cases by recording the dates when the case was paused and when processing resumed, once the basis for pausing the clock no longer exists. Asylum Division headquarters officials said pauses in reasonable fear case processing times are separate from case delay reasons, but case delays may occur for reasonable fear cases. In our case processing time analysis of USCIS data, we excluded approximately 13 percent of reasonable fear cases that had at least one pause in case processing time from our analysis because, in conducting our analysis, we could not systematically confirm the appropriate order of dates for those cases. As shown in table 6, our review of USCIS data for cases that did not include pauses found that USCIS completed at least 91 percent of reasonable fear cases within 10 or fewer court days from fiscal year 2016 to the second quarter of fiscal year 2019. USCIS Does Not Have Complete Data on Reasons for Case Delays Although the Asylum Division monitors overall processing times for credible and reasonable fear cases, it does not collect comprehensive data in its case management system on some types of case delays. For example, USCIS tracks whether cases are delayed for certain reasons related to the individual—such as if he or she has a medical condition that prevents the asylum officer from conducting the interview, if the individual requests that the interview be rescheduled, or if the individual is detained in a remote location. In addition, USCIS’s system can track if cases are delayed for logistical or resource constraints. Specifically, asylum officers may select “lack of resources” as one case delay reason in the system. However, this field in the system does not allow officers to distinguish more specific types of delays—such as a lack of space in detention facilities for officers to screen fear cases, telephones not working properly, and other types of delays—which officers told us occur on a regular basis. Asylum officers we interviewed in the Arlington and Houston offices stated that logistical delays could affect the number of credible or reasonable fear cases they can complete each day. Specifically, some asylum officers said they have experienced delays up to 30 minutes waiting for phone lines to work properly at detention facilities. Moreover, supervisors we interviewed in the Arlington office stated telephone and interpreter delays could add 20 or 30 minutes per case, resulting in a cumulative delay that could affect an officer’s productivity for any given day. Moreover, supervisors in the Arlington office said it is challenging to identify the appropriate number of cases to assign to officers because the number depends on whether or not disruptions occur. Asylum officers in Arlington said they are expected to conduct a certain number of credible or reasonable fear screenings per day, but expectations for completing their assigned cases may be tempered by circumstances such as interpreter availability or if there are issues at the detention facility, including physical space shortfalls or difficulty in locating the individual at the facility. Similarly, asylum officers in Houston said they are expected to complete a certain number of credible or reasonable fear cases per day, but supervisors understand that they may face logistical challenges such as interpreter or telephone issues. In addition to system limitations in tracking case delay reasons, Asylum Division headquarters officials said their case management system does not have the capability to track how long case delays may last. Our analysis of USCIS data from fiscal year 2014 through the second quarter of fiscal year 2019 indicates that 21,528 credible fear cases and 6,724 reasonable fear cases had delays. USCIS’s system can calculate the number of days for each credible and reasonable fear case—in other words, the total processing time for each case—and the system can produce daily reports noting these overall processing times. However, officials in the Houston office told us they must investigate individual cases on an ad hoc basis to understand how long cases have been delayed during processing. Specifically, officials in the Houston office said they maintain weekly “late reports” using information from USCIS’s case management system that show pending credible and reasonable fear cases with the longest processing times and that they must spend time researching cases on the report to determine the length of the delays. Standards for Internal Control in the Federal Government state that management should obtain data on a timely basis so that they can be used for effective monitoring. These standards also state that management should process the obtained data into quality information that supports the internal control system. As previously discussed, USCIS’s case management system does not track specific logistical reasons for any delays in credible and reasonable fear cases, which affect the number of cases an officer can complete in a day. Furthermore, USCIS’s system can calculate the number of processing days for each credible and reasonable fear case. However, the system cannot track how long a case delay lasts. Headquarters officials said they evaluate the usefulness of their system, and consider options for improvements or changes, on an ongoing basis. However, as of October 2019, they stated they did not have plans for significant changes to the system to track more specific case delay reasons. Collecting additional information in its automated case management system on case delays would provide USCIS with more readily available information and analyzing such data could help USCIS identify case delay reasons relevant in the current environment for officers conducting fear screenings and better position USCIS to mitigate the reasons for the delays and improve efficiency in case processing. EOIR Has Processes to Manage the Credible and Reasonable Fear Review Workload and Is Developing a Tool to Monitor Adherence to Required Review Processing Times EOIR has developed processes for immigration courts and judges to help manage its workload related to credible and reasonable fear reviews. As previously discussed, in the event of a negative outcome of their credible or reasonable fear screening, noncitizens can request a review of USCIS’s negative determination by an immigration judge. The Immigration and Nationality Act, as amended, and regulation require that such reviews occur within certain time frames. Specifically, immigration judge reviews of negative credible fear determinations are to be conducted no later than 7 days after referral from USCIS, to the maximum extent practicable, and immigration judge reviews of negative reasonable fear determinations are to be conducted within 10 days of referral, in the absence of exceptional circumstances. EOIR officials told us that increased resources, beginning in fiscal year 2015, and a faster process for hiring immigration judges have allowed EOIR to increase the number of immigration judges. As of September 30, 2019, EOIR reported that it had 442 immigration judges on board, including 173 judges hired in fiscal year 2018 and fiscal year 2019. EOIR reports that the number of immigration judges has increased each year from fiscal year 2015 through fiscal year 2019. EOIR officials told us that they plan to hire an additional 100 judges in fiscal year 2020. Additionally, EOIR officials told us that they prioritize credible and reasonable fear reviews and that these reviews can generally be accommodated within EOIR’s existing resources—specifically, by finding efficiencies within judges’ existing schedules to add credible or reasonable fear review hearings or by conducting hearings via video teleconferencing (VTC). EOIR officials also said that credible and reasonable fear reviews for individuals in ICE’s family residential centers comprise a small portion of EOIR’s overall workload. According to EOIR officials, each ICE detention facility is assigned to the jurisdiction of an immigration court, and the workload for credible and reasonable fear reviews is managed locally by the court to which each detention facility is assigned. ICE officers are to initiate the immigration judge’s review by filing a request with the appropriate immigration court. Some courts are co-located with ICE detention facilities in which the detainee requesting the credible or reasonable fear review is housed. EOIR officials said that reviews in those locations are typically heard in person by immigration judges assigned to that facility, and that the court finds room in the judge’s regular calendar to hear credible and reasonable fear reviews. For individuals in detention facilities without a co-located immigration court, including ICE family residential centers, immigration judges typically conduct credible and reasonable fear reviews via VTC. Judges conducting credible or reasonable fear reviews via VTC may be located in any immigration court in the United States. According to EOIR officials, the Assistant Chief Immigration Judge for each court is responsible for managing the court’s workload, including seeking support from judges outside the court in circumstances where there are too many cases for the court’s assigned judges. EOIR officials told us that the use of VTC technology—which is available in all courtrooms—provides flexibility to the courts in balancing workloads related to credible and reasonable fear reviews, among other workloads. In addition, EOIR officials stated that judges’ credible and reasonable fear workload is impacted, in particular, by immigration enforcement priorities and USCIS credible or reasonable fear determinations. For example, if DHS places more noncitizens into expedited removal proceedings who subsequently express fear or intent to apply for asylum, EOIR’s related workload might increase. In addition, because immigration judges do not review USCIS’s positive credible fear determinations, if USCIS’s screenings result in more negative determinations, EOIR’s caseload related to credible or reasonable fear reviews might increase. As of January 2018, EOIR has performance measures that include timeliness goals for credible and reasonable fear reviews, and these timeliness goals align with the required credible and reasonable fear review time frames. However, EOIR data we reviewed indicate that about 30 percent of credible and reasonable fear reviews are not completed within the required time frames. Specifically, EOIR’s memorandum on Case Priorities and Immigration Court Performance Measures states that 100 percent of credible fear reviews should be completed within seven days of an asylum officer’s negative determination and that 100 percent of reasonable fear reviews should be completed within 10 days of the filing of a negative reasonable fear determination. Further, according to EOIR officials, courts are to assign credible and reasonable fear reviews to a judge within 48 hours of receipt of the request from ICE, and immigration judges are to complete such reviews within 24 hours after they are assigned. EOIR officials said their automated case management system maintains data on the date when courts receive a request from ICE for an immigration judge review, the date the review is assigned to a judge, and the date the review takes place. EOIR headquarters officials told us that they monitor the extent to which judges are completing reviews within 24 hours after they are assigned using an automated immigration judge performance dashboard, which allows officials to review this performance measure for all judges combined, for individual courts, or for individual judges. Further, EOIR officials told us that if courts are scheduling credible and reasonable fear reviews within 48 hours after receipt and judges are completing reviews within 24 hours after they are assigned, they expect that EOIR should be meeting the required time frames (7 days after ICE’s referral for credible fear reviews and 10 days after ICE’s referral for reasonable fear review) for conducting credible and reasonable fear reviews. EOIR data we reviewed indicate that, from fiscal year 2014 through June 2019, approximately 28 percent of credible fear and 36 percent of reasonable fear reviews exceeded the required time frames, as shown in table 7 below. As previously discussed, the Immigration and Nationality Act and regulation allow for some flexibility with regard to the required credible and reasonable fear review time frames. Specifically, credible fear reviews are to be completed within 7 days, to the maximum extent practicable, and reasonable fear reviews are to be completed within 10 days, absent exceptional circumstances. EOIR officials we spoke with said there are a variety of court, judge, or applicant-related reasons that reviews could exceed the required time frames. For example, case file documentation sent from USCIS to the court may be incomplete. Further, a detention facility may have a medical quarantine that restricts court proceedings for a certain period of time. EOIR headquarters officials told us that, as of October 2019, they review weekly reports that include the median processing times for completed credible and reasonable fear reviews. For example, according to one weekly report from October 2019, the median completion time for credible fear reviews was 7 days. These reports also include information about the average and median number of days pending per case, for those credible and reasonable fear reviews that are not complete. For example, the weekly report we reviewed from October 2019 showed that EOIR had 553 pending credible fear reviews that week, with a median of 7 days pending and an average of 18 days pending. While these weekly reports allow EOIR headquarters officials to monitor some information about their credible and reasonable fear workload, they do not provide information to EOIR officials about the proportion of EOIR’s credible and reasonable fear reviews that are completed within the required time frames, or whether any reviews are delayed for reasons within the limits set out in the law or regulation. EOIR officials said they plan to implement an automated court operations dashboard in early 2020 which is to, among other things, monitor court performance against the performance goals EOIR established in January 2018, including the credible and reasonable fear performance goals. This automated dashboard is to be similar to the immigration judge performance dashboard, which EOIR implemented in early 2019. According to EOIR, the court operations dashboard is intended to operationalize EOIR’s performance measures—including completion of 100 percent of credible fear reviews with 7 days and 100 percent of reasonable fear reviews within 10 days—by providing court staff with daily alerts and warning notices to help court administrators prioritize the scheduling of cases based on the performance measures. This prioritization, combined with EOIR’s monitoring of judge performance to ensure that credible and reasonable fear reviews are completed within 24 hours after they are scheduled, should provide EOIR officials with sufficient information to monitor EOIR’s adherence to the required credible and reasonable fear review time frames. Because implementation of the court operations dashboard is planned for early 2020, it is too soon to know if EOIR will use the dashboard to monitor adherence to the required credible and reasonable fear review time frames or if it will help EOIR understand reasons for delays in those cases that take longer than 7 or 10 days. Conclusions The number of credible and reasonable fear cases has increased since fiscal year 2014, and USCIS policies and procedures require completion of those cases within short time frames. The Asylum Division provides training for credible and reasonable fear cases to new asylum officers in basic training, given the differences between these screenings and affirmative asylum adjudications. However, not all offices provide additional training on screening such cases at the family residential centers. Ensuring that all asylum offices provide such training, in addition to basic training for new officers, would better prepare them to screen those cases efficiently and effectively. In addition, USCIS relies on its periodic quality assurance reviews to assess the quality of credible and reasonable fear cases across asylum offices. Developing and implementing more specific guidance on requirements for documenting the results of its periodic quality assurance reviews would better position the agency to track trends for credible and reasonable fear reviews across asylum offices. USCIS data show that positive credible fear determination rates are higher at the family residential centers than they are nationwide, in part because USCIS’s automated case management system does not track whether an individual receives a credible fear determination as a principal applicant, dependent, or in the interest of family unity. Without systematically recording credible fear determinations involving family members, USCIS may not have complete data on credible fear determination rates, and the agency may not be in a position to report on the scope of its policy for family members in the credible fear process. Asylum officers have experienced logistical delays that can affect the number of credible and reasonable fear cases they complete each day. Although USCIS tracks some of these delays in its case management system, the system does not distinguish between specific reasons for logistical case delays, such as telephones nor working properly or lack of space at detention facilities for officers to screen cases. Furthermore, USCIS’s system can calculate the number of processing days for each credible and reasonable fear case. However, the system cannot track how long case delays last. By collecting and analyzing additional information on case delays, including specific reasons for delays and how long they last, USCIS can identify relevant case delays for officers conducting fear screenings. Moreover, analyzing specific case delay information could help USCIS mitigate reasons for case delays and improve efficiency in case processing. Recommendations for Executive Action We are making the following four recommendations to USCIS: The Director of USCIS should ensure that, in addition to USCIS’s basic asylum officer training, all asylum offices provide pre-departure training on the credible and reasonable fear processes before their officers begin screening cases at the family residential centers. (Recommendation 1) The Director of USCIS should develop and implement more specific guidance on requirements for documenting results of Asylum Division periodic quality assurance reviews. (Recommendation 2) The Director of USCIS should ensure asylum officers systematically record in USCIS’s automated case management system if individuals receive credible fear determinations as principal applicants, dependents, or in the interest of family unity, pursuant to regulation or USCIS policy. (Recommendation 3) The Director of USCIS should collect and analyze additional information on case delays, including specific reasons for delays and how long they last, that asylum officers may face when screening credible and reasonable fear cases. (Recommendation 4) Agency Comments We provided a draft of this report to DHS and DOJ for review and comment. DHS provided formal, written comments, which are reproduced in full in appendix IV. DHS also provided technical comments, which we incorporated as appropriate. DOJ told us they had no comments on the draft report. DHS concurred with our recommendations and described actions planned or underway to address them. For example, regarding our recommendation that all USCIS asylum offices provide officers with pre-departure training on credible and reasonable fear before they officers begin screening cases at family residential centers, DHS stated that USCIS plans to develop a standardized pre-departure training and provide this training to all asylum officers prior to their deployment to the family residential centers. In addition, regarding our recommendation that USCIS ensure that asylum officers record in their automated case management system if individuals receive credible fear determinations as principal applicants, dependents, or in the interest of family unity, DHS noted USCIS will explore ways to modify its case management system to ensure that asylum officers record such data and train officers on any subsequent system changes. We are sending copies of this report to the appropriate congressional committees, the Acting Secretary of Homeland Security, the Attorney General, and other interested parties. In addition, the report is available at no charge on GAO’s website at https://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-8777 or gamblerr@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. Key contributors to this report are listed in appendix V. Appendix I: Objectives, Scope, and Methodology We were asked to review processes for screening noncitizens who arrive at the southwest border expressing an intention to apply for asylum, a fear of persecution or torture, or a fear of return to their country, and the resources needed to carry out these screenings within applicable time frames by the Department of Homeland Security’s (DHS) U.S. Citizenship and Immigration Services (USCIS) and Department of Justice’s (DOJ) Executive Office for Immigration Review (EOIR). This report discusses (1) what USCIS and EOIR data show about the credible fear and reasonable fear processes, (2) the extent to which USCIS has policies and procedures for overseeing credible fear and reasonable fear screenings, and (3) the extent to which USCIS and EOIR have processes for managing their respective credible fear and reasonable fear-related workloads. To address all three objectives, we interviewed USCIS headquarters personnel from the Refugee, Asylum, and International Operations Directorate (RAIO) and RAIO’s Asylum Division who are responsible for managing USCIS’s credible and reasonable fear screening processes. We also interviewed officials from USCIS’s Fraud Detection and National Security Directorate (FDNS), which is responsible for leading USCIS’s efforts to detect and deter immigration benefit fraud and help detect national security and public safety threats. We conducted site visits at two of USCIS’s eight asylum offices—Houston, Texas and Arlington, Virginia—in April 2019. We selected these asylum offices based on the relatively large size of their credible and reasonable fear caseloads in fiscal year 2018—the most recent, complete data available at the time of our review. During these visits, we conducted in-person, semi-structured interviews with asylum officers, supervisory asylum officers, training officers, FDNS immigration officers, and asylum office management. During these interviews, we discussed topics related to data quality, supervisory review, training, quality assurance, family processing, and resource allocation. While the views expressed in these interviews do not represent those of all Houston and Arlington asylum office officials, they provide valuable insights from stakeholders who have experience with credible and reasonable fear policies and procedures. In addition, we collected written responses from the remaining six asylum offices on the same topics. Further, we conducted site visits to U.S. Immigration and Customs Enforcement (ICE) adult detention centers and family residential centers. Specifically, we visited single adult detention facilities in San Diego, California (September 2018), and Port Isabel and Pearsall, Texas (October 2018 and February 2019, respectively). We selected these ICE single adult facilities based on their geographic proximity to various CBP field locations we visited (discussed below). In addition, in February 2019, we visited ICE’s Enforcement and Removal Operations field office in San Antonio, Texas, as well as ICE’s South Texas Family Residential Center in Dilley, Texas, and Karnes County Residential Center in Karnes, Texas. We selected these two ICE family residential centers for field visits because they accounted for more credible and reasonable fear referrals to USCIS than any other single adult detention facility or family residential center. We also selected them to examine unique aspects of ICE and USCIS processing of credible and reasonable fear claims made by members of family units. During these visits to USCIS asylum offices and ICE detention facilities, we observed USCIS asylum officers conducting credible or reasonable fear screenings of single adults and family unit members either in person or via telephone. In total, we observed more than 20 credible and reasonable fear interviews across our site visits. Our observations are not generalizable to all USCIS asylum offices conducting credible and reasonable fear screenings, but provided us the opportunity to learn more about how USCIS personnel conduct interviews, make fear determinations, process these cases, and coordinate with ICE officials. For additional context about how noncitizens are apprehended at the border, processed into expedited or full immigration removal proceedings, transferred to ICE, and ultimately referred to USCIS for credible and reasonable fear screenings, as appropriate, we interviewed headquarters personnel from DHS’s U.S. Customs and Border Protection’s (CBP) Office of Field Operations (OFO) and U.S. Border Patrol (Border Patrol) who are responsible for apprehending noncitizens at or between U.S. ports of entry. In addition, we conducted site visits at CBP facilities in California and Texas from September 2018 to October 2018. In California, we visited Border Patrol’s San Diego sector headquarters and Imperial Beach station, and OFO’s San Ysidro port of entry. In Texas, we visited CBP’s Central Processing Center and McAllen Border Patrol station in McAllen, Texas; Border Patrol’s Fort Brown, Weslaco, and Harlingen stations; and OFO’s Hidalgo and Brownsville ports of entry. During these visits, we interviewed Border Patrol and OFO officials and observed how CBP personnel processed apprehended individuals and, as appropriate, documented whether those individuals expressed an intention to apply for asylum, a fear of persecution or torture, or a fear of return to their country. To select these locations, we assessed CBP data on Border Patrol and OFO apprehensions along the southwest border and targeted specific locations that saw the greatest increase in the number of apprehensions of individuals from fiscal year 2016 to 2017. As noted previously, we also considered the geographical proximity of multiple CBP and ICE facilities to maximize observations. Our observations during site visits are not generalizable to all Border Patrol and OFO operations along the southwest border, but provided us the opportunity to learn more about policies and procedures for processing noncitizens into removal proceedings and documenting any fear claims. To address the first objective, we obtained and analyzed data and documentation from USCIS and EOIR. Regarding USCIS, we analyzed record-level data from USCIS’s automated case management system from fiscal year 2014 through the second quarter of fiscal year 2019 (March 2019)—the most recent time period for which complete data were available at the time of our review. We analyzed these data to identify the number, characteristics, and outcomes of credible and reasonable fear cases. According to USCIS officials, USCIS’s system creates a unique number, or “case ID” for each case. USCIS officials told us that a previous system used a different identifier for each case—the individual’s Alien number (or “A-number”)—and did not use a “case ID” field. USCIS transitioned from its previous system to its current system in February 2018 and, according to USCIS officials, cases originally opened prior to the transition to the new system may have been re-opened under the same “case ID” number in the new system. As part of our data reliability testing, we checked for unique “case ID” numbers by searching for duplicate values and determined the data did not have duplicate values for “case ID” numbers. For our analysis of USCIS data specifically for ICE detention facilities and family residential centers, we only included credible and reasonable fear cases for detained individuals. To assess the reliability of USCIS data, we completed a number of steps, including (1) performing electronic testing for obvious errors in accuracy and completeness, such as running logic tests; (2) reviewing existing information about the data and the systems that produced them, such as relevant training materials for USCIS officers who use the system; and (3) discussing data entry issues and data limitations with USCIS officials. We determined the data were sufficiently reliable to describe the number, outcomes, and characteristics of credible and reasonable fear cases. Regarding EOIR, we reviewed data on immigration judge reviews of credible and reasonable fear cases posted on its public website. Specifically, we reviewed EOIR data on credible and reasonable fear reviews from fiscal year 2014 through June 2019—the most recent, complete data available at the time of our review. We also obtained and analyzed summary data from EOIR on credible and reasonable fear reviews for those individuals detained in ICE’s family residential centers. We analyzed the data to determine the outcomes of all credible and reasonable fear reviews and compared the outcomes of all reviews with the outcomes of reviews at ICE’s family residential centers. Finally, we reviewed EOIR data on the outcomes in immigration court for those completed removal cases that began with a positive credible or reasonable fear determination. In addition, we interviewed immigration judges and other court personnel serving both detained and nondetained dockets from EOIR’s Otay Mesa Immigration Court and San Diego Immigration Court in California, and from EOIR’s Harlingen Immigration Court in Texas. We also observed two immigration judge reviews of negative credible fear determinations. Our observations are not generalizable to all immigration judge reviews, but provided us the opportunity to learn more about EOIR’s processes. We interviewed EOIR officials about their data entry and management practices for credible and reasonable fear reviews. We determined that the data EOIR provided, much of which they report publicly on their website, are sufficiently reliable for analyzing the number and duration of credible and reasonable fear reviews that are received, completed, and pending. To provide additional context on the numbers, characteristics, and outcomes of CBP apprehensions, we obtained and analyzed record-level data on all apprehensions by Border Patrol and OFO from fiscal year 2014 through the second quarter of fiscal year 2019. We also obtained and analyzed record-level data on ICE detentions from fiscal year 2014 through fiscal year 2018 (see app. III for the results of our analyses). To assess the reliability of Border Patrol, OFO, and ICE data, we completed a number of data reliability steps, including (1) performing electronic testing for obvious errors in accuracy and completeness, such as running logic tests; (2) reviewing existing information about the data and systems that produced them, such as relevant training materials for Border Patrol agents and OFO, and ICE officers who use agency data systems; and (3) discussing data entry issues and data limitations with Border Patrol, OFO, and ICE officials. We also received demonstrations on the data systems from Border Patrol, OFO, and ICE officials at headquarters and in the field. As described below, we determined that the data are sufficiently reliable for providing information on the numbers, characteristics, and outcomes of CBP apprehensions and ICE detentions. Border Patrol data. For our analysis of Border Patrol data, we used “apprehensions” as our unit of analysis, instead of the number of individuals apprehended, because an individual may have been apprehended multiple times in the same year. We identified a small number of Border Patrol apprehension records that had the same date of apprehension and unique identifier (“A-number”). It is possible that these apprehension records represented one apprehended individual that Border Patrol agents processed as two apprehensions. These records comprised less than one percent of the more than 2.3 million apprehension records we analyzed. We included these apprehension records in our analysis because Border Patrol considers them unique apprehensions and because their small number did not materially affect our analysis. In addition, Border Patrol did not systematically track family relationships in its data systems until fiscal year 2016, as we have previously reported. Therefore, our analysis of Border Patrol apprehensions of family unit members processed under expedited removal proceedings is for fiscal years 2016 through the first two quarters of 2019. Further, according to Border Patrol officials, Border Patrol did not record reasonable fear cases in its automated data system before April 2016. Therefore, we are reporting the number of reasonable fear cases recorded by the Border Patrol in its automated system from fiscal year 2017 (the first full year for which Border Patrol recorded this information in its system) through the second quarter of fiscal year 2019. We did not include the 860 reasonable fear cases that Border Patrol recorded in its automated system for fiscal year 2016, since this number represents only partial-year data. According to Border Patrol officials, prior to April 2016, these reasonable fear cases would likely have been recorded under other case dispositions in their automated system, such as one indicating the reinstatement of a prior removal order. We determined that Border Patrol data are sufficiently reliable to describe the numbers and demographic characteristics of individuals and family unit members apprehended from fiscal year 2014 through the second quarter of fiscal year 2019. OFO data. For our analysis of OFO data, we used “apprehensions” as our unit of analysis, instead of the number of individuals apprehended. We excluded approximately 13 percent of all apprehension records (including single adults, unaccompanied alien children, and parents and children that arrived as part of a family unit) from our analyses because we could not confirm an A-number for those apprehensions. Among the apprehension records missing an A-number, 44 percent were cases in which OFO officers paroled the individuals and, according to OFO officials, officers are not required to assign an A-number to these individuals. In addition, 47 percent of the records with a missing A- number were cases that involved the individual withdrawing their application for admission into the United States, in which OFO officers have discretion whether to assign an A-number. According to OFO officials, additional records with missing A-numbers may be due to data entry errors or problems with the data system saving this information in the database that OFO used to pull the data. Finally, we collapsed 182,266 apprehension records into 86,597 apprehension records because we determined that they were duplicate records for the same individual and the same apprehension, based on factors such as alien number, birth date, and date and time of apprehension. As a result, we determined that we could not present precise figures for analyses that include OFO data and instead provided approximations throughout the report. We rounded all data and figures on OFO apprehensions down to the hundreds place and described relevant data using modifiers such as “at least” because of possible missing information. In addition, according to OFO officials, OFO does not capture information in its automated data system on individuals who were processed under expedited removal with a reasonable fear claim. OFO officials stated that, since OFO has historically processed a relatively small number of such apprehensions, it does not collect automated data on reasonable fear claims. With the previously-described modifications, we determined that OFO data are sufficiently reliable to generally describe the numbers and demographic characteristics of individuals and family unit members apprehended from fiscal year 2014 through the second quarter of fiscal year 2019. ICE data. To report on ICE detentions of adults and family unit members, we obtained and analyzed ICE detention data from fiscal years 2014 to 2018, the most current data available at the time of our review. The ICE data we obtained contained information on whether adults and family members booked-in to an ICE detention facility had a fear claim recorded in ICE’s data system as of the date our data were pulled. Specifically, we divided our analysis of ICE detention data into two parts. First, we obtained data on all individuals (all adults and children without consideration of any family relationships) detained from fiscal year 2014 through fiscal year 2018. Second, we obtained data specifically on family unit members apprehended by CBP and housed at the four ICE family residential centers from fiscal year 2014 through fiscal year 2018. Regarding our analysis of family unit members who made a fear claim in one of ICE’s family residential centers, we excluded less than one percent of all detention records from our analyses because we could not confirm a unique identifier for the individual. In addition, for individual family unit members who were detained more than once in a fiscal year, we included the most recent record for the individuals in our analyses to report on the most recent information available about each individual. This accounted for less than one percent of all detention records in our time period of analysis. We determined that the data were sufficiently reliable to describe the numbers of individuals (adults and family unit members) who were apprehended by CBP and recorded by ICE as having made a credible fear claim. To address our second objective, in addition to our aforementioned interviews and site visits, we reviewed relevant laws and regulations governing the credible and reasonable fear screening process. We collected and analyzed documentation on key USCIS oversight mechanisms related to credible and reasonable fear screenings— supervisory review, asylum officer training, and quality assurance reviews. In particular, we reviewed the Credible Fear Procedures Manual, and the Reasonable Fear Procedures Manual, standard operating procedures, training and quality assurance records and materials, and guidance on conducting credible and reasonable fear screenings for families in ICE detention. Specifically, we reviewed USCIS asylum officer basic training materials from RAIO and the Asylum Division, and training materials for officers from outside the Asylum Division who screen credible and reasonable fear cases. In particular, we reviewed USCIS Asylum Division quarterly training reports for fiscal year 2018 and used them to analyze the weekly training activities in each asylum office for each week of the reporting quarter. We compared RAIO and Asylum Division training materials with federal internal control standards related to developing competent individuals qualified to carry out assigned responsibilities. We also reviewed documents associated with the quality assurance reviews that the Asylum Division conducted, including those reviews conducted in collaboration with RAIO. Specifically, we reviewed standard operating procedures, reviewer checklists, and resulting reports and analysis for three RAIO nationwide reviews of credible and reasonable fear cases and for the six periodic reviews of credible and reasonable fear cases the Asylum Division conducted at asylum offices and at the family residential centers between November 2017 and May 2018. We compared these policy documents and their role in providing oversight of the credible and reasonable fear process against federal internal control standards related to ongoing monitoring activities and evaluation of results. We also reviewed USCIS standard operating procedures, requirements, and training materials for processing family members, and corresponding data on applicant family relationships. We then compared the procedures, requirements and data against federal internal control standards related to obtaining high quality data. To address our third objective, we reviewed USCIS and EOIR documents and data, and interviewed relevant officials to evaluate the extent to which USCIS and EOIR have process for managing their respective credible and reasonable fear-related workloads. USCIS. In particular, we reviewed USCIS documentation and spoke with officials from Asylum Division headquarters and local asylum offices regarding the Asylum Division’s staffing allocation model for the credible and reasonable fear workload. In addition, we obtained and analyzed record-level data from USCIS’s automated case management system to identify processing times and case delays for credible and reasonable fear cases between fiscal year 2014 through the second quarter of fiscal year 2019 (March 2019). We included cases that had a fear determination that was served or an administrative closure for both detained and nondetained individuals. In this report, we present information on both credible and reasonable fear case receipts and analysis of processing times for the cases using the “clock-in” date recorded in USCIS’s automated case management system. However, while USCIS relies on the “clock-in” date to track case processing times, according to an Asylum Division official USCIS tracks and reports the number of credible and reasonable fear case receipts based on the date cases are input into, or “created” in, its automated system. According to the official, these “created” and “clock-in” dates are often the same, but can differ slightly. Therefore, the number of case receipts tracked and reported by USCIS may differ slightly from those presented in this report. Regarding credible fear cases, we determined case processing times by calculating the difference between the beginning and end dates for credible fear cases. We considered credible fear case processing times for detained individuals to begin on the day when USCIS receives referral documents and records a “clock-in” date in the automated case management system, as noted previously. For nondetained individuals, the clock starts for credible fear cases when a USCIS asylum office conducts the interview for a credible fear screening. We used the starting clock date for detained and nondetained individuals provided by USCIS for our analysis. We considered credible fear case processing times to end on the day when cases either had a fear determination that was served or an administrative closure. We included credible fear cases that had a fear determination that was served or an administrative closure for detained and nondetained individuals. We also reviewed USCIS’s publicly-reported data on credible fear processing times during this time period. Regarding reasonable fear cases, we used USCIS data to count the number of processing days and percent of cases completed in certain time intervals. We determined reasonable fear processing times by calculating the difference between the beginning and end dates for reasonable fear cases. We considered reasonable fear case processing times for detained individuals to begin on the day when USCIS receives referral documents and records a “clock-in” date in the automated case management system, as noted previously. For nondetained individuals, the clock starts for reasonable fear cases when a USCIS asylum office conducts the interview for a reasonable fear screening. We used the starting clock date for detained and nondetained individuals provided by USCIS for our analysis. We calculated reasonable fear processing times in court days by excluding weekends and federal holidays. USCIS may also pause the clock when processing reasonable fear cases in certain circumstances. We excluded approximately 13 percent of reasonable fear cases that had at least one pause in case processing time from our analysis because in conducting our analysis we could not systematically confirm the appropriate order of dates for those cases. We considered reasonable fear case processing times to end on the day when cases either had a fear determination that was served or administrative closure. We also included reasonable fear cases that were served a fear determination or received an administrative closure for detained and nondetained individuals. To identify the reasons for delays in credible and reasonable fear cases during the time period of our analysis, we identified the fields that USCIS’s case management system tracks for case delays related to the credible and reasonable fear workload. In addition, we reviewed USCIS’s manuals and documentation on its case management system. We compared USCIS’s recording and tracking of data on case delays to federal internal control standards related to obtaining data on a timely basis for management to use for effective monitoring and that data should be processed into high quality information. We determined that the USCIS data we reviewed on credible and reasonable fear processing times and case delays were sufficiently reliable for our purposes. EOIR. To evaluate EOIR’s process for managing its credible and reasonable fear-related workload, we interviewed EOIR officials about their practices to manage the credible and reasonable fear workload, including immigration judge hiring, oversight of credible and reasonable fear review processing times, infrastructure requirements for credible and reasonable fear reviews, and the use of video teleconferencing by judges to conduct credible and reasonable fear reviews. We reviewed publicly available data about EOIR’s workload and case adjudications, including data about the number of credible and reasonable fear reviews EOIR judges completed and data about judges hired from fiscal year 2014 through fiscal year 2019. We also reviewed guidance documents, such as EOIR’s 2018 Case Priorities and Performance Measures memorandum, which established performance measures for credible and reasonable fear reviews. In addition, we used EOIR data to analyze the timeliness of EOIR’s completion of credible and reasonable fear reviews and compared EOIR’s processing times for fiscal year 2014 through the third quarter of fiscal year 2019 with required time frames. By reviewing documentation on EOIR’s case management system and interviewing officials with knowledge about EOIR’s case management system and the methodology used to calculate the publicly-reported data, we determined that the EOIR data we reviewed on credible and reasonable fear review processing times and outcomes was sufficiently reliable for analyzing the number of credible and reasonable fear reviews completed and pending, and the duration of the reviews. We conducted this performance audit from November 2018 to February 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient and appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Eligibility, Screening Standards, and Possible Screening Outcomes for Credible and Reasonable Fear Cases This appendix provides detailed information on eligibility, screening standards, and possible screening outcomes for both credible fear and reasonable fear cases. Noncitizens placed into expedited removal who make fear claims will be referred to U.S. Citizenship and Immigration Services (USCIS) for a credible fear screening by an asylum officer or, if the individual has been issued a final administrative removal order after conviction for an aggravated felony or has a prior order of removal that is reinstated, and expressed a fear of return, for a reasonable fear screening. Table 8 below describes the eligibility and screening standards, as well as the potential outcomes for USCIS’s credible fear screening cases. Similarly, table 9 details eligibility, screening standards, and potential outcomes for reasonable fear screening cases. Appendix III: Data on Noncitizens Apprehended, Detained, and Screened for Credible or Reasonable Fear by Department of Homeland Security If noncitizens are placed into expedited removal proceedings instead of full removal proceedings, they are to be ordered removed from the United States without further hearing before an immigration judge unless they indicate either an intention to apply for asylum or a fear of persecution or torture or a fear of return to their country (referred to throughout this appendix as making a “fear claim”). This appendix provides information on the number and dispositions (such as full removal proceedings or expedited removal proceedings, among others) of noncitizens who were apprehended by the Department of Homeland Security’s (DHS) U.S. Customs and Border Protection’s (CBP) U.S. Border Patrol (Border Patrol) and Office of Field Operations (OFO) at or between U.S. ports of entry from fiscal year 2014 through the second quarter of fiscal year 2019. It also includes U.S. Immigration and Customs Enforcement (ICE) data on detentions of noncitizens who made a credible fear claim. For cases in which noncitizens were referred to U.S. Citizenship and Immigration Services (USCIS) for a fear screening, this appendix also provides additional information on the characteristics of these cases, including their country of origin, age, gender, whether they had representation, and location of their screenings. Case Dispositions for Border Patrol Apprehensions of Noncitizens from Fiscal Years 2014 through the First Two Quarters of Fiscal Year 2019 As shown in table 10, Border Patrol apprehensions totaled more than 2.3 million from fiscal year 2014 through the second quarter of fiscal year 2019. Further, Border Patrol data indicate that agents processed about 687,000 (or 30 percent) for full immigration proceedings and nearly 931,000 (or 40 percent) under expedited removal proceedings. For those apprehensions that agents processed under expedited removal, more than 197,000 (approximately 9 percent of total apprehensions) included a credible fear claim made in Border Patrol custody. As also shown in table 10, during fiscal years 2017 through the first two quarters of 2019, Border Patrol apprehended more than 10,000 additional noncitizens who made reasonable fear claims. As shown in figure 7, the number of Border Patrol apprehensions of individuals who were placed into expedited removal proceedings with a credible fear claim increased from more than 16,000 apprehensions in fiscal year 2014 to more than 51,000 in fiscal year 2018. These apprehensions of individuals claiming fear ranged from 3 percent to 13 percent of total apprehensions during these fiscal years. Characteristics of Noncitizens Apprehended by Border Patrol Processed Under Expedited Removal Who Claimed Fear Border Patrol data include various characteristics of each apprehension such as age, gender, and whether a noncitizen was a member of a family unit. For example, of the nearly 208,000 apprehensions processed under expedited removal with a credible or reasonable fear claim during fiscal years 2014 through the first half of fiscal year 2019, approximately 166,000 (or 80 percent) were adults age 18 and above with the remaining 42,000 (or 20 percent) encompassing children age 17 and under (see table 11). Of the nearly 208,000 apprehensions processed under expedited removal with a credible or reasonable fear claim during fiscal years 2014 through the first half of fiscal year 2019, approximately 117,000 (or 56 percent) were male and the remaining 90,000 (44 percent) were female (see table 12). As shown in table 13, for fiscal years 2016 through the first two quarters of 2019, Border Patrol apprehended nearly 456,000 noncitizens who were members of families. Of these, Border Patrol processed more than 120,000 (or 26 percent) under expedited removal proceedings. Nearly 71,000 apprehensions during this time period (15 percent of total family unit members apprehended and 59 percent of those placed in expedited removal) included a credible fear claim. Case Dispositions for OFO Apprehensions from Fiscal Years 2014 through the First Two Quarters of Fiscal Year 2019 From fiscal year 2014 through March 2019, OFO apprehensions at ports of entry totaled at least 546,900. Of these 546,900 apprehensions, OFO officers placed at least 193,500 (or 35 percent) into expedited removal proceedings. For those in expedited removal proceedings, OFO data indicate that at least 104,600 apprehensions included a credible fear claim in OFO custody (19 percent of total apprehensions). In addition, OFO issued Notices to Appear before an immigration judge for full immigration proceedings to at least 167,400 (or 31 percent) of the approximately 546,900 total apprehensions (see figure 8). As shown in figure 9, the number of OFO apprehensions in expedited removal proceedings with a credible fear claim generally increased over this time period from at least 11,600 apprehensions in fiscal year 2014 to at least 27,000 in fiscal year 2018 (the last full year of data available at the time of our analysis). In addition to this overall increase, the percentage of OFO’s total apprehensions placed into expedited removal proceedings with a credible fear claim also increased. Specifically, these apprehensions increased from about 17 percent of all apprehensions in fiscal year 2014 to about 26 percent in fiscal year 2018. Characteristics of Noncitizens Apprehended by OFO and Placed into Expedited Removal Proceedings with a Credible Fear Claim OFO apprehension data include various characteristics such as age, gender, and whether an apprehension involved a member of a family unit. For example, as shown in table 14, of the approximately 104,300 OFO apprehensions with credible fear claims, at least 78,500 (or 75 percent) were adults age 18 and above with about 25,700 (or 25 percent) of the remaining credible fear claims encompassing children age 17 and under. Also, for each year during this period, the percentage of adults versus children was generally consistent with this overall percentage with the exception of fiscal year 2019, for which the partial year’s data show that about 98 percent of those apprehensions processed under expedited removal with a credible fear claim were adults. In addition, for fiscal years 2014 through the first two quarters of fiscal 2019, at least 56,500 (or 54 percent) of these apprehensions involving a fear claim were male and at least 47,400 (or 45 percent) were female (see table 15). Also, for each year during this period, the number of males and females were almost evenly split with the exception of fiscal year 2019, for which the partial year’s data show a larger proportion of males claiming fear. As shown in table 16, for fiscal years 2016 through the first two quarters of 2019, OFO had a total of at least 144,100 apprehensions involving members of family units. Of these approximately 144,100 apprehensions, OFO placed at least 39,100 (27 percent) into expedited removal proceedings of which at least 32,900 (about 23 percent of total family unit members apprehended and approximately 84 percent of those placed in expedited removal) claimed a credible fear of returning to their country. Number of Individuals in ICE Detention with a Credible Fear Claim, Fiscal Years 2014 through 2018 The number of individuals in expedited removal proceedings detained in ICE facilities with a credible fear claim increased from fiscal years 2014 to 2018. Specifically, as shown in table 17, ICE data indicate that the number of individuals in expedited removal proceedings with a recorded credible fear claim while in ICE detention increased from about 37,000 (or 9 percent) in fiscal year 2014 to about 99,000 (or 26 percent) in fiscal year 2018. The period of greatest percentage increase was from fiscal years 2015 to 2016 when the percentage of individuals in expedited removal proceedings with a credible fear claim while in ICE custody increased from approximately 15 percent to 25 percent. For fiscal years 2014 through 2018, the majority of family unit members in ICE’s four family residential centers had a credible fear claim (81 percent), as demonstrated in table 18. The number of family unit members with a fear claim ranged from approximately 69 percent in fiscal year 2015 to 88 percent in fiscal year 2018. For fiscal years 2014 through 2018, slightly more than half of all family unit members in ICE’s four family residential centers with a credible fear claim were children under the age of 18 (55 percent). As also shown in table 19, the division between adults and children with fear claims varied little each year. Nationalities of Noncitizens Referred to USCIS for Credible or Reasonable Fear Screenings from Fiscal Year 2014 through the First Two Quarters of Fiscal Year 2019 As shown in Figure 10, the majority of credible fear cases referred to USCIS for screening from fiscal year 2014 through the first two quarters of fiscal year 2019 had applicants who were nationals of El Salvador, Honduras, Guatemala, or Mexico. Citizens of these countries accounted for 74 percent of all credible fear cases during this time period (approximately 306,000 referrals). As shown in table 20, El Salvador had the most credible fear referrals to USCIS each year from fiscal year 2014 through fiscal year 2017. However, beginning in fiscal year 2018, Honduras accounted for the most credible fear referrals to USCIS among these four countries. As shown in Figure 11, applicants from the countries of Mexico, Honduras, El Salvador, and Guatemala accounted for all but approximately 7 percent of the reasonable fear cases screened by USCIS for fiscal years 2014 through the first two quarters of fiscal year 2019. Overall, Mexican nationals accounted for the largest number of reasonable fear cases among these four countries (33 percent of total reasonable fear cases). As shown in table 21, Mexico had the most reasonable fear referrals to USCIS each year from fiscal years 2014 through the first two quarters of fiscal year 2019. Outcomes of USCIS Credible Fear and Reasonable Fear Screenings Based on the Presence of Representation at the Applicant’s Interview As table 22 shows, noncitizens making credible fear claims who had representation present at their interviews with asylum officers more often received positive determinations of fear by the asylum officer. Overall, during this time period, the number of positive determinations in cases with representation was nearly 10 percentage points greater than those without representation. As table 23 shows, similar to credible fear cases, noncitizens making reasonable fear claims who had representation present at their interviews with asylum officers more often received positive determinations of fear by the asylum officer. Overall, during this time period, the number of positive determinations in cases with representation was over 20 percentage points greater than those without representation. ICE Detention Facilities and Family Residential Centers Making the Most Credible Fear and Reasonable Fear Referrals to USCIS for Screening During Fiscal Years 2014 through the First Two Quarters of Fiscal Year 2019 As table 24 shows, two of ICE’s family residential centers (Dilley and Karnes family residential centers) accounted for the highest number of credible and reasonable fear referrals, among the top five facilities making these referrals, from fiscal years 2014 through the first two quarters of fiscal year 2019. Reasonable Fear Referrals to USCIS from ICE’s Family Residential Centers, and Related Positive Outcomes, During Fiscal Years 2014 through the First Two Quarters of Fiscal Year 2019 As shown in table 25, reasonable fear screenings for those in ICE family residential centers comprised 6 percent of all such cases referred to USCIS during this same period with the percentage of positive determinations (77 percent) higher than that for all reasonable fear cases nationwide (30 percent). Number of Credible Fear and Reasonable Fear Cases Screened, by USCIS Asylum Office, for Fiscal Years 2014 through the First Two Quarters of Fiscal Year 2019 As shown in table 26, the Houston asylum office screened two-thirds (67 percent) of credible fear cases from fiscal year 2014 through the first two quarters of fiscal year 2019. Also, over this same time period, USCIS’s Los Angeles asylum office screened the second most credible fear cases (11 percent). However, since fiscal year 2018, USCIS’s Asylum Pre- Screening Center has screened the second most credible fear cases after Houston. As shown in table 27, the Houston asylum office screened nearly half (approximately 45 percent) of reasonable fear cases from fiscal year 2014 through the first two quarters of fiscal year 2019. Also, over this same time period, USCIS’s Los Angeles asylum office screened the second most reasonable fear cases (12 percent). However, since fiscal year 2018, USCIS’s Asylum Pre-Screening Center has screened the second most reasonable fear cases after Houston. Appendix IV: Comments from the Department of Homeland Security Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments In addition to the contact named above, Kathryn Bernet (Assistant Director), Michael Harmond (Analyst-in-Charge), Hiwotte Amare, Miranda Cohen, Benjamin Crossley, Michele Fejfar, Cynthia Grant, Jan Montgomery, Heidi Nielson, Mary Pitts, Adam Vogt, and Jessica Walker made key contributions to this work.

Why GAO Did This Study Individuals apprehended by DHS and placed into expedited immigration proceedings are to be removed from the country without a hearing in immigration court unless they express an intention to apply for asylum, or a fear of persecution, torture, or return to their country. Those with such “fear claims” are referred to USCIS for a credible fear screening. Individuals who have certain criminal convictions or who have a reinstated order of removal and claim fear are referred for a reasonable fear screening. Those with negative outcomes can request a review by EOIR's immigration judges. GAO was asked to review USCIS's and EOIR's processes for fear screenings. This report examines (1) USCIS and EOIR data on fear screenings, (2) USCIS policies and procedures for overseeing fear screenings, and (3) USCIS and EOIR processes for workload management. GAO analyzed USCIS and EOIR data from fiscal years 2014 through mid-2019; interviewed relevant headquarters and field officials; and observed fear screenings in California, Texas, and Virginia, where most screenings occur. What GAO Found Data from the Department of Homeland Security's (DHS) U.S. Citizenship and Immigration Services (USCIS) and Department of Justice's Executive Office for Immigration Review (EOIR) indicate that their credible and reasonable fear caseloads generally increased from fiscal year 2014 through fiscal year 2018. USCIS's caseloads nearly doubled during this timeframe—from about 56,000 to almost 109,000 referrals for credible and reasonable fear screenings. Further, the credible fear caseload was larger in the first two quarters of fiscal year 2019 alone than in each of fiscal years 2014 and 2015. Referrals to USCIS for reasonable fear screenings also increased from fiscal years 2014 through 2018. USCIS asylum officers made positive determinations in 71 percent of all credible and reasonable fear screenings between fiscal years 2014 and the first two quarters of fiscal year 2019. The outcomes of the remaining screenings were generally split evenly (14 percent each) between negative determinations or administrative closures (such as if the applicant was unable to communicate). EOIR's caseload for immigration judge reviews of USCIS's negative credible and reasonable fear determinations also increased between fiscal year 2014 and fiscal year 2018. EOIR's immigration judges reviewed about 55,000 cases from fiscal year 2014 through the third quarter of 2019 (the most recent data available), and judges upheld USCIS's negative determinations in about three-quarters of all reviews. USCIS has developed various policies and procedures for overseeing credible and reasonable fear screenings in accordance with the regulations governing those screenings, such as interview requirements and mandatory supervisory review. USCIS provides basic training for new asylum officers and other training at individual asylum offices that includes credible and reasonable fear. The training at asylum offices includes on-the-job training for officers newly-assigned to credible and reasonable fear cases and ongoing weekly training for incumbent officers—some of which includes credible and reasonable fear. However, USCIS asylum offices do not all provide additional pre-departure training before officers begin screening families in person at DHS's family residential centers. Asylum Division officials told GAO that additional training for asylum officers before they begin screening such cases is important—in particular, credible fear screenings at these facilities represent about one-third of USCIS's caseload. Almost all USCIS asylum offices send officers to the family residential centers, including those offices with small fear caseloads at the local level. Some asylum offices provide pre-departure training to officers being sent to screen families, but such training is inconsistent across offices. By comparison, officials from the Chicago and New York offices stated they do not provide formal pre-departure training, but rather direct or recommend that officers review Asylum Division guidance and procedures on family processing independently before they travel. Officials from two other offices stated they rely on the training asylum officers may receive throughout the year related to credible and reasonable fear, which can vary. Providing pre-departure training, in addition to USCIS's basic training for new asylum officers, would help USCIS ensure that officers from all asylum offices are conducting efficient and effective fear screenings of families. Further, consistent with regulation, USCIS policy is to include any dependents on a principal applicant's credible fear determination if the principal applicant receives a positive determination, resulting in the principal and any dependents being placed into full removal proceedings with an opportunity to apply for various forms of relief or protection, including asylum. For example, a parent as a principal applicant may receive a negative determination, but his or her child may receive a separate positive determination. In the interest of family unity, USCIS may use discretion to place both the parent and child into full removal proceedings rather than the parent being expeditiously ordered removed in accordance with the expedited removal process. However, USCIS's case management system does not allow officers to record whether an individual receives a determination on his or her case as a principal applicant, dependent, or in the interest of family unity. Without complete data on all such outcomes, USCIS is not well-positioned to report on the scope of either the agency's policy for family members who are treated as dependents, pursuant to regulation, or USCIS's use of discretion in the interest of family unity. USCIS and EOIR have processes for managing their respective credible and reasonable fear workloads. For example, USCIS uses national- and local-level staffing models to inform staffing allocation decisions. USCIS also sets and monitors timeliness goals for completing credible and reasonable fear cases. Although USCIS monitors overall processing times, it does not collect comprehensive data on some types of case delays, which officers told us can occur on a regular basis. Asylum officers whom GAO interviewed stated that certain delays could affect the number of credible or reasonable fear cases they can complete each day. Collecting and analyzing additional information on case delays would better position USCIS to mitigate the reasons for the delays and improve efficiency. EOIR has developed processes for immigration courts and judges to help manage its workload that include performance measures with timeliness goals for credible and reasonable fear reviews. EOIR data indicate that about 30 percent of credible and reasonable fear reviews are not completed within the required timeframes. EOIR officials said they plan to implement an automated tool in early 2020 to monitor court performance, including the credible and reasonable fear performance goals. Because implementation of the automated tool is planned for early 2020, it is too soon to know if EOIR will use the tool to monitor adherence to the required credible and reasonable fear review time frames or if it will help EOIR understand reasons for case delays. What GAO Recommends GAO is making four recommendations, including that USCIS provide additional pre-departure training to USCIS asylum officers before they begin screening families, systematically record case outcomes of family members, and collect and analyze information on case delays. DHS concurred with GAO's recommendations.

gao_GAO-20-442

gao_GAO-20-442_0

Background Federal Improper Payment Requirements Key requirements related to improper payments during the period of our audit were included in IPIA, OMB M-18-20, and OMB Circular A-136. Federal agencies were required to take various steps regarding improper payments under IPIA and as directed by OMB M-18-20. The steps include the following: 1. Review all programs and activities and identify those that may be susceptible to significant improper payments (commonly referred to as a risk assessment), 2. For those programs and activities that agency risk assessments, OMB, or statute identifies as being susceptible to significant improper payments, agencies should develop statistically valid improper payment estimates, as well as analyze the root causes of improper payments and develop corrective actions to reduce them, 3. Report on the results of addressing the foregoing requirements. According to OMB officials, agencies are responsible for maintaining the documentation to demonstrate that these steps, if applicable, were satisfied. Figure 1 illustrates these steps, as well as the major components of conducting an improper payment risk assessment. IPIA required that agencies conduct improper payment risk assessments for all federal programs and activities at least once every 3 years and identify any program or activity that may be susceptible to significant improper payments. OMB M-18-20 provides guidance for implementing the IPIA requirements and covers agencies’ responsibilities for improper payment risk assessments, estimation, and reporting. According to the OMB guidance, agencies must institute a systematic method of reviewing all programs and activities to identify those that may be susceptible to significant improper payments. This systematic method can be a quantitative evaluation based on a statistical sample or a qualitative method, such as a risk-assessment questionnaire. Regardless of which method of review is used, IPIA required agencies to consider seven risk factors during the risk assessment. (See table 1.) OMB is also required to designate a list of high-priority programs for greater levels of oversight and review. The threshold for high-priority program determinations for fiscal year 2018 reporting and subsequent years is $2 billion in estimated improper payments, regardless of the improper payments rate estimate. In addition, OMB may determine that a program is high-priority for reasons other than exceeding the $2 billion threshold. High-priority programs are subject to additional requirements, such as submitting information about semi-annual or quarterly actions taken to reduce improper payments that can be used as a tool for tracking progress. According to OMB M-18-20, another fundamental requirement that agencies must meet is to recover any federal dollars that are a monetary loss to the government, unless legislation specifically prevents such recovery. Specifically, the Improper Payments Elimination and Recovery Improvement Act of 2012 (IPERIA) requires any program that expends at least $1 million during the year to implement payment recapture audits, if cost-effective to the agency, in order to recover improper payments. The requirement to conduct payment recapture audits applies to all agencies regardless of whether they have a program susceptible to significant improper payments. A payment recapture audit is a review and analysis of an agency’s or program’s accounting and financial records, supporting documentation, and other pertinent information supporting its payments, that is specifically designed to identify overpayments. It is not an audit that is performed in accordance with government auditing standards. OMB M-18-20 also states that for high-priority programs the agency shall report any action it has taken or plans to take to recover improper payments and intends to take to prevent future improper payments. If an agency has determined that performing payment recapture audits for any applicable program or activity is not cost-effective, a justification for that determination must be reported. Further, OMB M-18-20 states that agencies should report a justification for that determination through AFRs, Performance Accountability Reports, or in the format required through data requests from OMB. DOE Structure and Processes to Collect Information and Report on Improper Payments DOE’s 15 field Chief Financial Officers, in cooperation with DOE contracting officers, are responsible for overseeing contractor and other activities in the field, and they assist the OCFO in implementing improper payment reporting requirements. The OCFO issues Annual Payment Integrity Requirements and Guidance that transmits DOE’s instructions for meeting improper payments reporting and recapture requirements prescribed by OMB M-18-20. This guidance includes instructions for completing an attached template for reporting risk assessments and improper payments and payment recapture information. Using this template, 48 payment reporting sites provide information that is the basis for DOE’s department-wide improper payment risk assessment and reporting. These payment reporting sites consist of four types of federal entities and two types of contractors. (See appendix I for more information about DOE’s payment reporting sites.) In addition to the completed template, sites are directed to submit a signed certification that attests to the accuracy of the improper payment information and risk assessment and, if applicable, a justification for why payment recapture audits were not conducted. The OCFO completes a quality assurance checklist for each site and consolidates and reports the data as one program in DOE’s annual AFR. DOE reports on its improper payments 1 year in arrears; meaning, for example, that DOE’s fiscal year 2019 AFR included information on its improper payments identified in fiscal year 2018. In addition to reporting payment recovery information, as required by OMB Circular A-136, DOE has optionally reported some information it collected about improper payments that it identified at the time of the AFR issuance each year for fiscal years 2015 through 2019. (See table 2.) Specifically, DOE reported the amount of improper payments that had been made and identified in the preceding year—not based on a statistically valid estimate of improper payments but, rather, on reported amounts of known improper payments from individual payment reporting sites. DOE also reported improper payment rates that it calculated based on these reported amounts. DOE was not required to report a statistically valid estimate of improper payments in its AFRs because it determined it was at low risk of susceptibility to significant improper payments. See appendix II for additional details about improper payments in the data that DOE collected from the sites. In previous years, DOE reported statistical estimates of its improper payments. Specifically, DOE’s Performance and Accountability Reports and AFRs from fiscal years 2004 and 2007 through 2011 indicated that DOE used statistical sampling to produce projected improper payment estimates for certain payment categories. During these years, DOE reported estimated improper payment rates of less than 1 percent in these categories. However, in 2012 the DOE OIG determined that the estimated improper payment rate presented in DOE’s fiscal year 2011 AFR was not based on a statistical method. According to OCFO officials, DOE discontinued the use of statistical sampling to produce estimates in fiscal year 2012 because it was not required to do so, due to DOE’s determination that it was at low risk for significant improper payments. Since 2012, the DOE OIG has found DOE to be compliant with requirements for improper payment reporting and risk assessments as part of its required review. Specifically, the DOE OIG reported each year that DOE met the requirements for publishing improper payment information in its AFRs and performed the required risk assessments. According to an OIG official, the OIG is not required to perform evaluative procedures to determine the adequacy and completeness of DOE’s risk assessment and reporting in its AFR, and they have not optionally performed these procedures. The Role of Audits and Investigations in Identifying and Recovering DOE’s Improper Payments The DOE OIG and other federal agencies or external audit organizations conduct periodic incurred cost audits and assessments of DOE’s cost- reimbursement contracts. The purpose of incurred cost audits is to determine whether such incurred costs are reasonable; applicable to the contract; allowable under generally accepted accounting principles and cost accounting standards applicable in the circumstances; and not prohibited by the contract, statute, or regulation. If, as a result of these audits or assessments, improper payments are identified—such as reimbursements for costs determined to be unallowable under the contract—DOE will question these costs, indicating that there is a possibility the costs are improper. DOE may then negotiate or otherwise work with the contractor to resolve the questioned costs. Sometimes, this can result in DOE recovering funds. According to DOE’s fiscal year 2019 annual payment integrity requirements and guidance, for the purpose of improper payment reporting, a questioned cost is not deemed an improper payment until it has been determined by the contracting officer to be unallowable. In addition, investigations conducted by DOE OIG, the Department of Justice, and other federal agencies may identify potentially unallowable DOE payments. Upon their resolution, these investigations may find such DOE payments to have been improper. According to OIG officials, improper payments identified through OIG investigations may be recovered through civil or administrative processes, and some of the improper payments identified through OIG investigations may lead to government-run criminal investigations. OCFO officials told us that recovered amounts may differ from the monetary loss associated with the original payments because of fees or fines, among other reasons. IPERIA requires agencies to include all identified improper payments in their reported estimate, regardless of whether the improper payment has been or is being recovered. According to DOE’s fiscal year 2019 annual payment integrity requirements and guidance, if the terms of a settlement require repayment to DOE, then the settlement amount would be considered an unallowable cost. Furthermore, the 2019 guidance states that due to the timing of when a settlement is reached, it is not possible to report these costs as an improper payment in the current year of reporting. Beginning in fiscal year 2018, DOE reported information in its AFR on improper payments made in prior years that were identified for recapture in the current reporting year. For example, in its fiscal year 2018 AFR, DOE reported that $92.69 million in prior years’ improper payments had been identified for recapture in fiscal year 2017. Similarly, in its fiscal year 2019 AFR, DOE reported $14.18 million in prior-year payments identified for recapture in fiscal year 2018. DOE’s reporting sites generally identify prior years’ improper payments identified for recapture through audits and investigations, among other strategies. DOE did not provide information on the years in which the prior-year improper payments were made, and the prior year improper payments identified for recapture were reported separately but not included in DOE’s reported improper payment amount and rate in any of its AFRs. Improper Payment Amounts DOE Reported in Its AFRs for Fiscal Years 2015 through 2019 May Not Be Accurate or Complete The improper payment amounts that DOE reported in its AFRs for fiscal years 2015 through 2019 may not be accurate or complete and are likely understated, for two key reasons. First, we found that some DOE payment reporting sites did not correctly identify, track, and report their improper payments. Second, DOE reported improper payment amounts and rates for the current year, but did not report that the amounts and associated rates do not include a substantial amount of improper payments that may be identified in the years following the year in which the payment took place. Some DOE Payment Reporting Sites Did Not Correctly Identify, Track, and Report Their Improper Payments The information in DOE’s AFRs for fiscal years 2015 through 2019 may not be accurate or complete, in part because DOE does not ensure that payment reporting sites correctly identify, track, and report their improper payments to the OCFO. Our review of documentation and interviews with officials at the selected payment reporting sites found some instances in which payment reporting sites’ processes for identifying and tracking improper payments did not always result in accurate and complete financial reporting as required. Specifically, we identified the following errors in reporting improper payments to the OCFO at three of the 10 sites we selected for review: Officials at one site told us that they resolve a portion of their improper payments by adjusting future invoices to account for the error. Site officials told us that in such cases, they do not track or report the amounts to the OCFO as improper payments. While adjusting future invoices is an efficient way to recapture improper payments, not tracking such adjustments as improper payments results in understated improper payments reported to the OCFO. The total amount of the understatement of these improper payments is not known. In addition, the site may have overstated other improper payments. In particular, site officials told us that they were unsure whether some of the annual adjustments from its indirect cost reconciliation process were reported as improper payments, even though OCFO officials told us that such adjustments are routine and are not considered improper payments. This could have resulted in an overstatement of improper payments, but the amount overstated is unknown. Officials at another site told us that they mistakenly included almost $1 million in questioned costs in their fiscal year 2017 improper payment reporting to the OCFO. Because questioned costs are not considered improper until they are determined to be unallowable, this means that the site overstated its improper payments by almost $1 million for that year. Additionally, this site subtracted its underpayments from its overpayments for its fiscal year 2015 reporting, resulting in an understatement of improper payments. Improper payments, regardless of whether they are over- or underpayments, should be added together and not netted, as both amounts are considered improper. Officials at the site told us that these issues had been corrected as of fiscal year 2018. Officials at a third site told us that they do not closely track underpayments and cannot state with certainty that all underpayments are included in the site’s annual improper payments report. The site therefore may be understating its improper payments each year. DOE’s Financial Management Oversight order states that financial management processes must include procedures and methods for ensuring that financial managers provide accurate, relevant financial reporting to customers, such as Congress and OMB. Additionally, federal internal control standards state that management should implement control activities through policies, including documenting policies in the appropriate level of detail to allow management to effectively monitor the control activity. However, not all of the payment reporting sites have fully documented their procedures for correctly identifying, tracking, and reporting their improper payments or ensuring the quality of their data, in part because there was no requirement to do so. Specifically, officials from all 10 selected payment reporting sites we interviewed told us they have procedures for tracking their identified improper payments. However, three of the 10 selected sites had not documented their procedures and two sites had documented some of their procedures but not others, including two of the sites mentioned in the examples above. By requiring payment reporting sites to document their procedures for identifying, tracking, and reporting their improper payments to ensure the quality of their data, the OCFO could better ensure that each payment reporting site maintains consistent procedures and provides comparable information about that site’s improper payments over time. Furthermore, the OCFO cannot ensure that sites are correctly identifying, tracking, and reporting improper payments and ensuring the quality of their data because the OCFO does not have a process to monitor that sites have documented—and are implementing—procedures to do so. The OCFO has taken some steps to help ensure the quality of the improper payments data that the sites report to the OCFO. For example, OCFO officials said they confirm that sites provide accurate information by requiring sites to self-certify the accuracy and completeness of the data, but does not take steps to verify the certification. Further, four of the five contractor payment reporting sites we interviewed told us the DOE field sites that oversee the contractors review the contractors’ submissions before sending the information to the OCFO; OCFO officials told us that the field sites do not formally approve these submissions. Additionally, OCFO staff complete a quality assurance checklist for each site’s submission. The checklist contains a series of questions to determine whether a site has submitted the required documentation and whether certain elements of that documentation are complete. OCFO quality assurance reviews also include simple mathematical checks for internal consistency, such as ensuring that the amount for total identified improper payments is the same across multiple tables. These steps, however, are not sufficient to ensure that sites are correctly identifying, tracking, and reporting improper payments and ensuring the quality of their data. For example, the quality assurance checklist does not include any tests to verify the accuracy of the procedures sites used to generate that data to ensure the sites’ data are reliable. By developing a monitoring process to ensure that payment reporting sites have developed and implemented procedures for identifying, tracking, and reporting their improper payments to the OCFO and ensuring the quality of their data, the OCFO could better ensure that the information it reports about improper payments in its AFR is accurate and complete. DOE Identifies a Substantial Amount of Improper Payments in Subsequent Years That Are Not Included in Any Years’ Improper Payment Amount or Rate The amount of current year improper payments DOE reports in each fiscal year, as well as the improper payment rate DOE calculates based on this amount and reports in its AFRs, is not accurate or complete because it does not disclose that there are additional improper payments that are (1) not identified or that DOE’s OCFO is not aware of until a later date, or (2) potential improper payments that may be identified at a later date. Additionally, DOE does not conduct payment recapture audits, which may identify additional improper payments that could be recovered. DOE identifies many of its improper payments after the end of the fiscal year in which the payments occur and does not identify some improper payments until several years after they occur. These improper payments are identified through processes such as post-payment reviews, audits and assessments, and investigations that do not conclude until after the end of the fiscal year in which DOE made the payments. As a result, there is a known lag in identifying certain improper payments. The current year improper payment amount and associated rate DOE reported in its AFR excludes any improper payments that are identified after the end of the fiscal year in which the payments occurred. For example, in its fiscal year 2018 AFR, DOE reported $32.86 million of current reporting year (fiscal year 2017) improper payments, with an associated improper payment rate of 0.09 percent. In its fiscal year 2018 AFR DOE also reported $92.69 million of improper payments made earlier than fiscal year 2017; however DOE did not disclose that the amount of improper payments originally reported for any prior fiscal year had subsequently increased as a result of improper payments identified after the end of the fiscal year. While it is not possible for DOE to report on the specific amount of improper payments it has not yet identified, DOE also did not disclose in its fiscal year 2018 AFR that it expected to complete audits and investigations in subsequent years that could increase the amount of improper payments reported for fiscal year 2017. See figure 2 for categories of improper payments and the extent to which they are included in DOE’s improper payment amount and rate. Specifically, the OCFO excludes some known improper payments from the annual amount and associated rate it reports in its AFRs for the following reasons: Post-payment reviews may not conclude in the same fiscal year the reviewed payments were made. We have previously found that DOE identifies some improper payments through post-payment reviews. For example, DOE has not required its contractor payment reporting sites—most of which are M&O contractors—to submit invoices before DOE makes payments; instead DOE uses a “payments cleared funding arrangement,” which authorizes the contractors to withdraw funds directly from federal accounts. OCFO officials told us that improper payments made by DOE to contractors without such an agreement would be reported by the responsible federal site, and improper payments made by M&O contractors would be reported by the M&O contractor. DOE policies and procedures do not require that DOE site officials monitor M&O contractor withdrawals to determine the appropriateness of their incurred costs. DOE officials do not review M&O contractor withdrawal of funds to determine the appropriateness of M&O contract costs, and thus can only identify improper payments associated with these contracts through post-payment reviews of contractor costs that may occur after the end of the fiscal year. However, such post-payment reviews, such as monthly or quarterly reviews of invoices, may not identify certain improper payments— including improper payments that occurred late in a given fiscal year—leading DOE to exclude them from their annual reported improper payment amount and associated rate. For example, according to a document describing improper payments that one selected payment site reported to the OCFO, the site identified about $103,000 in fiscal year 2016 improper payments associated with travel during that same fiscal year. Additional reviews of fiscal year 2016 travel payments conducted in fiscal year 2017 identified further improper payments for travel of more than $35,000. Because the contractor identified these additional travel payments as improper through quarterly reviews that did not conclude until after the end of fiscal year 2016, this increase of about 35 percent in the site’s known improper travel payments was not included in the OCFO’s reported improper payment amount or rate for that year. In our March 2017 report, we recommended that DOE help ensure that necessary data are available to employ data analytics—which can identify improper payments more quickly than post-payment reviews can, increasing the likelihood that DOE will include them in its reported amount and rate for each fiscal year—as a tool to perform contractor cost-surveillance activities. Specifically, we recommended that DOE require contractors to maintain sufficiently detailed transaction-level cost data that are reconcilable with amounts charged to the government, including (1) cost data that, at a minimum, represent a full data population; and (2) the details necessary to determine the nature of each cost transaction. DOE disagreed with the recommendation. According to DOE officials, DOE is now developing plans to begin to use data analytics in fiscal year 2021. We continue to believe it is important for DOE to employ data analytics as a cost surveillance tool so DOE can better identify improper payments to its contractors in a timely manner and look forward to reviewing DOE’s plans and actions to address our prior recommendation. Audit coverage of DOE payments is limited, and some audits are not completed until several years after the audited payments were made. As we also found in March 2017, DOE uses incurred cost audits and assessments to identify contractors’ improper payments. However, our review of DOE OIG and other external entities’ audits and assessments of incurred costs for DOE’s 24 largest contractors for this report shows that, historically, these audits are infrequent and may occur several years after the costs have been incurred. For example, our updated analysis shows that as of September 2019, only about $25 billion—or 23 percent—of the nearly $108 billion in costs incurred during fiscal years 2014 through 2018 by DOE’s 24 largest contractors had been audited or assessed (see table 3). Although there is no requirement for how often contractors should be audited, the Contract Disputes Act of 1978 imposes a 6-year statute of limitations for the government to seek recovery of unallowable costs that could be identified through audits. According to our review of DOE reporting and documentation, known DOE improper payments amounts for a given fiscal year can increase in later years as more costs are audited. For example, one payment reporting site reported to the OCFO nearly $164,000 in improper payments made and identified in fiscal year 2017, and OCFO included this amount in the improper payment rate it reported in its fiscal year 2018 AFR. According to site documentation, the same site also identified, as the result of an audit in fiscal year 2017, an improper payment of nearly $920,000 that had occurred in a prior year. This improper payment was substantially more than the total amount of improper payments that the site reported in the fiscal year 2018 AFR. However, because the improper payment occurred prior to fiscal year 2017, the OCFO did not include it in its current year improper payment amount or rate for any fiscal year. For fiscal year 2020, DOE’s OIG has planned several assessments of costs the contractors incurred in prior fiscal years. However, contractor costs the OIG plans to review in the fiscal year 2020 planned assessments were incurred as early as fiscal year 2015. Therefore, any improper payments identified through the planned assessments will not be included in DOE’s reported improper payment rate using the current reporting methods and will instead be included in an overall lump sum amount of prior year improper payments, which has no effect on DOE’s reported improper payment rate. DOE does not track questioned costs centrally, and such costs can take several years to resolve. Audits and assessments can identify questioned costs that require additional review before they are either allowed or deemed improper. In its Semiannual Report to Congress, DOE’s OIG reported nearly $700 million of unresolved, questioned costs identified through its own audits and investigations as of September 30, 2019. Our analysis of the DOE OIG’s reporting found that a substantial portion of questioned costs the OIG identified were ultimately determined to be allowable once they were resolved; however, our analysis also found that DOE has not consistently resolved questioned costs in a timely manner. For example, some of the questioned costs that the DOE’s OIG identified—such as potential state gross receipts tax overpayments of $15.1 million that a DOE payment site made in fiscal years 2010 and 2011—have remained unresolved for nearly a decade. Large amounts of unresolved costs reported by DOE’s OIG add uncertainty about the completeness of the OCFO’s improper payment reporting. Moreover, the nearly $700 million of unresolved questioned costs that the DOE OIG reported does not include questioned costs identified through external audits of non-M&O contractors, such as those conducted by the Defense Contract Audit Agency or nongovernmental entities. Questioned costs identified through these external audits can be substantial, like those the DOE OIG has reported. For example, a 2017 incurred cost audit of a DOE contractor’s fiscal year 2010 costs conducted by an external firm identified nearly $280 million in questioned and unresolved DOE payments to the contractor. In November 2019, DOE officials told us that these questioned payments were resolved when DOE reached a settlement agreement with the contractor. DOE disallowed $34 million of the questioned costs as part of the settlement agreement, according to DOE officials. DOE’s Financial Management Oversight order states that financial management processes must include procedures and methods for ensuring that financial managers provide accurate, relevant financial reporting to customers. DOE customers include Congress and OMB. Additionally, federal internal control standards state that management should implement control activities through policies, including documenting policies in the appropriate level of detail to allow management to effectively monitor the control activity. According to OCFO officials, DOE does not have a mechanism for tracking questioned costs identified through external audits. Instead, OCFO officials said they rely on payment reporting sites to track these costs to resolution. However, the office does not require payment reporting sites to document policies for such tracking. DOE officials from two selected sites told us that their sites do not have policies for tracking questioned costs identified through external audits, including questioned costs that may later be deemed improper. As a result, the OCFO may not be aware of all potentially improper payments identified through external audits or know the status of their resolution. Without a requirement for sites to have policies to track questioned costs to their resolution, the OCFO cannot ensure that payment reporting sites are tracking—and ultimately reporting—all improper payments, and thus cannot ensure that it is including all improper payments in the amount it reports as actual in its AFRs. Investigations that identify DOE improper payments may not conclude until years after the payments were made. Investigations by DOE’s OIG, the Department of Justice, and other federal agencies can also identify DOE improper payments. However, similar to improper payments identified through audits, these improper payments—which can be substantial—may not be identified until years after they occur due to the length of time it takes to investigate and resolve criminal, civil, or administrative cases. For example, in fiscal year 2018, DOE reported $60.6 million of improper payments identified through a fiscal year 2017 settlement with a contractor. DOE made some of these improper payments as early as 2001. Also, in fiscal year 2018, a DOE payment site reported that no improper payments were made or identified in fiscal year 2017, but the site reported a $4.6 million prior-year improper payment associated with a subcontractor’s false claims that were settled with the subcontractor in fiscal year 2017. The OCFO reported these two cases, along with other DOE improper payments identified through investigations, as lump sum prior-year improper payments identified for recapture in its fiscal year 2018 AFR. However, the OCFO did not include these known improper payments in the improper payment amounts used to calculate its improper payment rates for the years in which DOE incurred the disallowed costs. Furthermore, some DOE improper payments are not reported as current or prior-year improper payments because the investigations of the payments were resolved in a manner that prevented DOE from formally considering the payments improper. For example, in 2015, DOE’s OIG reported that a company received a loan guarantee of more than $500 million from DOE after it “provided the Department with statements, assertions, and certifications that were inaccurate and misleading, misrepresented known facts, and, in some instances, omitted information that was highly relevant to key decisions in the process to award and execute” the loan guarantee. The company later declared bankruptcy and did not repay the loan. However, because DOE did not determine this payment to be improper through a legal case or any other process, the $500 million of known monetary loss was not included in DOE’s improper payments reporting in its AFR for any fiscal year. Also, in fiscal year 2017, DOE excluded a six- figure settlement with an outside party from its improper payment reporting. OCFO officials told us that they excluded payments associated with this case from their office’s reporting due to certain aspects of the settlement agreement. DOE’s Information Quality Guidelines state that information disseminated to the public, such as information on improper payments reported in DOE’s AFRs, should be presented in an accurate, complete, unbiased, and clear manner and should be useful to the intended users of the information. As previously noted, agencies with programs that are susceptible to significant improper payments—defined to include improper payments exceeding $100 million in a year—are required to develop improper payment estimates and corrective action plans. However, the OCFO cannot determine whether improper payments in a given year exceeded the $100 million threshold because the OCFO does not track information on the year that payments were made for all known improper payments for a given fiscal year—including improper payments identified in later years through resolution of questioned costs or conclusions of audits or investigations. By tracking information on the year the payment occurred for all improper payments identified, to include those identified in later years, and determining and disclosing in its AFR whether improper payments in a given year exceeded the $100 million threshold, DOE could better inform Congress, OMB, and the public about whether it has made significant improper payments. Additionally, DOE sites perform some payment recapture activities, but does not conduct payment recapture audits, which could identify additional improper payments that could be reported, and potentially recovered. As previously discussed, IPIA required any program that expended at least $1 million annually to conduct payment recapture audits, if cost-effective to the agency, or to provide justification if such audits are determined not to be cost-effective. A payment recapture audit is a review and analysis of an agency’s or program’s accounting and financial records, supporting documentation, and other pertinent information supporting its payments, that is specifically designed to identify overpayments. As such, payment recapture audits are tools to identify improper payments, in addition to an avenue for recovering those overpayments. DOE included a justification for its decision not to conduct payment recapture audits in its AFRs for fiscal years 2015 through 2019. For example, in its fiscal year 2019 AFR, DOE cited its improper payment rate of 0.09 percent and recapture rate of 97 percent to support the department’s determination that it was not cost-effective to perform payment recapture audits. DOE also cited other activities it employed to identify and recapture improper payments, such as prepayment review and approval of invoices, post-payment reviews, contractor internal audits, results of cost allowability audits of integrated contractors, and results from travel audits, among others. The OCFO fiscal year 2018 payment integrity guidance included a list of seven criteria that sites were to use to determine whether payment recapture audits are cost-effective. For fiscal year 2018, 42 of DOE’s 48 payment reporting sites submitted a justification stating that it would not be cost-effective to employ payment recapture auditors. Our review of the 42 justifications found that the quality of the justifications varied by site. We found that 40 of the 42 justifications did not demonstrate consideration of any of the seven criteria in support of their determinations that payment recapture audits would not be cost- effective. One DOE field site’s justification included three bullet points, as shown in figure 3, none of which aligned with the criteria. The OCFO uses a quality assurance checklist to review payment sites’ improper payment reports that includes verifying that the site submitted a justification and that the justification is “adequate.” The checklist does not define “adequate,” and the OCFO approved all of the justifications submitted, even those that did not demonstrate consideration of any of the seven criteria from the payment integrity guidance. DOE’s Financial Management Oversight order states that financial management processes must include procedures and methods for ensuring that financial managers provide accurate, relevant financial reporting to customers. Furthermore, under OMB M-18-20, agencies are required to recover any federal dollars that are a monetary loss to the government, unless legislation specifically prevents such recovery. By clarifying guidance to define the factors for assessing the adequacy of the justifications, and reviewing sites’ justifications for not performing or arranging for payment recapture audits, DOE could better ensure that the justifications it reports have a sound basis and that DOE is not missing opportunities to identify and recover improper payments. Additionally, DOE may be missing opportunities to recover federal dollars that are a monetary loss to the government, as required under OMB M- 18-20, because it has not evaluated whether sites could identify additional improper payments through payment recapture audits. Our analysis of information provided by DOE shows that in fiscal year 2003 the department conducted payment recapture audits and that the improper payments identified through these audits far exceeded the costs of conducting the audits. According to OCFO officials, the information on payment recapture efforts was from a payment recapture audit at one site; it was not an OCFO recovery audit program. The OCFO officials reiterated that the majority of the payment reporting sites have not performed payment recapture audits because they believe existing efforts are effective in recovering identified improper payments. However, payment recapture audits are designed to identify additional improper payments not previously identified. By evaluating whether it could identify enough additional improper payments to make payment recapture audits cost-effective, such as performing audits at a limited number of sites, DOE would have an opportunity to identify and recover additional improper payments or have better information to justify that payment recapture audits are not cost-effective. DOE’s Fiscal Year 2018 Risk Assessment May Not Provide a Reasonable Basis for Its Risk Determination In its fiscal year 2018 improper payment risk assessment, DOE assessed its risk of susceptibility to significant improper payments as low. However, DOE did not provide sufficient documentation to support how it conducted its risk assessment and made this low-risk determination. Consequently, we could not determine if the process DOE used to perform its improper payment risk assessment provided a reasonable and reliable basis for making its risk determination. DOE’s Risk Assessment Process May Not Adequately Support Its Low-Risk Determination DOE’s process to conduct its fiscal year 2018 risk assessment may not be adequate to support its low-risk determination of susceptibility to significant improper payments. DOE has a decentralized process for conducting its statutorily required improper payment risk assessment every 3 years. For fiscal year 2018, DOE developed and provided each payment reporting site with an improper payment risk assessment template to complete. DOE directed all of its payment reporting sites to consider the seven risk factors listed in IPIA, as well as four additional risk factors that DOE developed. Table 4 lists the additional DOE-developed risk factors that sites were to consider in their risk assessments. DOE’s improper payment risk assessment template included a variable scale for rating each of the risk factors. The OCFO provided guidance instructing payment reporting sites to, when populating the template, consider the site’s exposure to the risk factors and to rate them by applying a numerical score to each risk factor. Each payment site totaled its numerical scores to calculate the site’s overall level of susceptibility to significant improper payments. DOE then consolidated all of the payment site assessments into an overall department-wide risk assessment. However, DOE could not explain, and did not provide us documentation to support, its rationale for the variable scales used to score such risk factors in its fiscal year 2018 assessments—both in the 10 payment-site risk assessments we reviewed and in DOE’s department-wide risk assessment—and how the scores assigned for each risk factor affected DOE’s susceptibility to significant improper payments. As a result, we could not determine if DOE’s process for conducting its fiscal year 2018 improper payment risk assessment provided a reasonable basis for DOE’s overall risk determination. Furthermore, the OCFO weighted all of the payment reporting sites equally in terms of overall risk when it aggregated the risk ratings into an overall assessment of susceptibility to significant improper payments. However, DOE did not provide an explanation or documentation of why the sites were weighted equally in the overall department-wide improper payment risk assessment, even though the payment types and dollar amounts of outlays processed by the sites varied widely. For example, a payment site processing $3 million of outlays in fiscal year 2017 had the same weight in the aggregated assessment as a payment site processing $5.7 billion of outlays. Finally, the OCFO did not provide evidence that it considered the known lag in identifying certain improper payments as an inherent risk during its fiscal year 2018 department-wide improper payment risk assessment process. This inherent risk relates to certain limitations affecting DOE’s ability to determine the extent of improper payments until several years after they occur, such as those identified through incurred cost audits and investigations, as previously discussed. For example, in its fiscal year 2018 AFR, DOE reported that a total of $124.35 million in payments were identified for recapture during fiscal year 2017, including $31.66 million made in fiscal year 2017 and $92.69 million made in years prior to fiscal year 2017. However, DOE did not provide us documentation to support how it considered the $92.69 million in improper payments made during years prior to fiscal year 2017—which could represent an inherent risk to the department—when assessing its risk of susceptibility to significant improper payments. As discussed earlier, some of the $92.69 million of improper payments identified for recapture occurred in fiscal year 2016. Thus, the amount of fiscal year 2016 improper payments that DOE reported in its fiscal year 2017 AFR is understated. Federal internal control standards state that management should design control activities to achieve objectives and respond to risks, and should implement control activities through policies. To contribute to the effective design and implementation of such control activities, management should clearly document internal control and all transactions and other significant events in a manner that allows the documentation to be readily available for examination. Additionally, management should periodically review policies, procedures, and related control activities for continued relevance and effectiveness. Further, although OMB does not direct agencies to demonstrate how each risk factor contributes to the agency’s overall susceptibility of risk for significant improper payments, OMB M-18-20 states that if a qualitative method is used during an improper payment risk assessment, it must be designed to accurately determine whether the program is susceptible to significant improper payments. DOE may not have an adequate process to support its risk determination because it did not properly document how it developed and considered risk factors during its fiscal year 2018 risk assessment. Until DOE revises its department-level process for conducting improper payment risk assessments, it cannot ensure that the process produces a reliable assessment of whether it is susceptible to significant improper payments. Specifically, without documenting its rationale for the variable scale used to score risk factors and weighting of the payment reporting sites, and consideration of the known lag in identifying the extent of total improper payments each fiscal year to support the development of its department- level risk assessment, DOE cannot demonstrate that its process for determining its low risk of susceptibility to significant improper payments is reasonable. Addressing these issues may result in DOE determining that it is susceptible to significant improper payments, and therefore subject to additional requirements—such as developing a statistically valid estimate of improper payments and reporting on actions to reduce improper payments, including a description of the root causes, and developing corrective actions to reduce them, including program-specific improper payment reduction targets. DOE Did Not Sufficiently Review the Reasonableness of Selected Payment Reporting Sites’ Improper Payment Risk Assessments We also found that DOE’s OCFO did not sufficiently review the reasonableness of the selected payment reporting sites’ improper payment risk assessments. When we reviewed the risk assessments of the 10 selected sites, we found a lack of consistency in how the sites applied DOE guidance, as well as inadequate documentation supporting how the sites considered improper payment risk factors. Specifically, we found that the OCFO review process did not identify instances in which these sites did not adequately support certain ratings or did not adhere to DOE instructions for completing the improper payment risk assessment template. Staff from the OCFO used a quality assurance checklist to review the sites’ fiscal year 2018 improper payment risk assessments. However, the extent to which the OCFO reviewed documentation supporting payment sites’ risk assessments is unclear. Although the reviewer guidance provided in the quality assurance checklist directs reviewers to ensure that the documentation supporting the payment site’s risk rating adequately supports the risk factor being evaluated, a payment site official told us that OCFO reviewers did not consistently request to view their supporting documentation. Eight out of 10 payment reporting sites we reviewed had documentation to support that they followed DOE’s guidance to consider the results of prior GAO and DOE OIG audit reports and OMB Circular A-123-related assessment results. However, we found that two sites did not have such documentation. One site rated itself as having no significant deficiencies despite audit reports that indicated some deficiencies and findings for that site. Another payment site did not discuss the OMB Circular A-123 assessment results in its improper payment risk assessments, despite OCFO guidance to include such results when conducting improper payment risk assessments. However, quality assurance checklists completed by OCFO staff for these two sites did not indicate that documentation supporting the sites’ consideration of these prior reports and assessments in their risk assessments was missing. Further, five of the 10 payment reporting sites we reviewed did not provide sufficient explanation or documentation supporting their ratings for several of the risk factors they considered in their improper payment risk assessment, despite instructions in DOE’s guidance to do so. For example, one site cited “discussions with team lead” as the primary source of support for its ratings assigned for several risk factors. However, the site did not have documentation to support the results of these discussions and how such discussions supported the ratings for each risk factor. Federal internal control standards state that management should design control activities to achieve objectives and respond to risks, and should implement control activities through policies. To contribute to the effective design and implementation of these control activities, management should clearly document internal controls and other significant events in a manner that allows the documentation to be readily available for examination. We also found that OCFO staff did not document any potential changes to the payment sites’ risk ratings in the 10 quality assurance checklists we reviewed. However, the process to be followed in the event OCFO reviewers find that payment site risk ratings are not reasonable is unclear because DOE has not defined and documented in its policies and procedures the process for OCFO reviewers to override these risk ratings. DOE’s Financial Management Oversight order directs business units to evaluate and assess the effectiveness of their financial management oversight activities and other internal controls, such as the OCFO’s oversight of the payment reporting sites’ risk assessments. Further, the order charges the OCFO with reviewing and analyzing activities throughout DOE to evaluate the adequacy of established policies, procedures, and standards governing accounting and related reporting functions; evaluating the performance of internal controls over those functions; and recommending corrective actions as needed. In addition, according to federal internal control standards, management should also establish and operate monitoring activities to monitor the internal control system and evaluate the results. Such monitoring includes regular management and supervisory activities, comparisons, reconciliations, and other routine actions. Additionally, management should periodically review policies, procedures, and related control activities for continued relevance and effectiveness. Further, federal internal control standards state that management should use quality information to achieve the entity’s objectives. By developing, documenting, and implementing policies and procedures to require OCFO to review documentation supporting payment site risk assessments and define the process for overriding their risk determinations, DOE would enhance its ability to adequately monitor its decentralized improper payment risk assessment process and help ensure the accuracy and reliability of payment reporting sites’ risk assessments and DOE’s assessment of overall risk of susceptibility to improper payments. Conclusions DOE’s OCFO relies on its 48 payment reporting sites to provide information about improper payments that DOE reports in its AFR; however, we identified several reasons that the information in DOE’s AFRs for fiscal years 2015 through 2019 may not be accurate or complete. First, DOE’s improper payments information may not be accurate or complete because the OCFO does not require the payment reporting sites to document their procedures for correctly identifying, tracking, and reporting their improper payments. By doing so, the OCFO could better ensure that the payment reporting sites provide consistent and comparable information about their improper payments over time. Second, the OCFO cannot ensure that sites are correctly identifying, tracking, and reporting improper payments to the OCFO and ensuring the quality of their data because OCFO does not have a process to monitor that sites have—and are implementing—procedures to do so. By developing such a monitoring process, the OCFO could better ensure that the information it reports about improper payments in DOE’s AFRs is accurate and complete. Third, DOE may not be reporting additional improper payments in the form of unallowable costs claimed by some contractors because, as we have previously found, DOE policies and procedures do not require that DOE sites monitor M&O contractor withdrawals to determine the appropriateness of costs incurred by the contractor. Under this arrangement, DOE does not use prepayment reviews to determine the appropriateness of M&O contract costs and, thus, can only identify improper payments associated with these contracts through post-payment reviews that typically occur after the end of the fiscal year. We previously recommended that DOE ensure data are available to employ data analytics—which can identify improper payments more quickly than post-payment reviews can—but DOE has not fully implemented the recommendation. We continue to believe it is important for DOE to employ data analytics as a cost surveillance tool so DOE can better identify improper payments to its contractors in a timely manner. DOE only includes improper payments that occur and are identified in the same fiscal year in its reported improper payment amount and rate in the AFR. However, DOE does not identify a substantial amount of improper payments in the same fiscal year due to the known lag in identifying such payments. Audits and assessments of DOE’s contractors can identify questioned costs that require additional review before they are either allowed or deemed improper, but DOE has not consistently resolved questioned costs in a timely manner because the OCFO does not direct payment reporting sites to document policies for tracking questioned costs to resolution. Without a requirement for sites to have policies to track questioned costs to their resolution, the OCFO cannot ensure that payment reporting sites are tracking—and ultimately reporting—all improper payments, and thus cannot ensure the accuracy and completeness of improper payments reported in DOE’s AFRs. Additionally, the OCFO cannot determine whether improper payments in a given year exceeded the $100 million threshold because the OCFO does not track information about the year that payments were made for all known improper payments for a given fiscal year. By tracking and disclosing information about all improper payments identified and the year in which these payments were made in its AFR, DOE would have better information to provide to Congress, OMB, and the public about whether it has made significant improper payments. Although DOE’s sites submitted individual justifications for not completing payment recapture audits, the quality of the justifications varied and did not meet DOE requirements. By clarifying guidance to define the factors for assessing the adequacy of the justifications, and reviewing sites’ justifications for not performing or arranging for payment recapture audits to ensure that the justifications meet requirements and are supported by appropriate analysis that considers the costs and benefits of performing the audits, DOE can better ensure that the justifications it reports have a sound basis and that DOE is taking advantage of all opportunities to both identify and recover improper payments, which in turn will help reduce the monetary loss to the government. Further, DOE may be missing opportunities to recover federal dollars that are a monetary loss to the government because it has not evaluated whether sites could identify additional improper payments through payment recapture audits. DOE has concluded that based on its self-assessed low improper payment rate and recapture rate, it is not cost effective to perform payment recapture audits. By evaluating whether it could identify enough additional improper payments to make payment recapture audits cost-effective, such as performing audits at a limited number of sites, DOE would have an opportunity to identify and recover additional improper payments or have better information to justify that payment recapture audits are not cost- effective. Finally, DOE may not have an adequate process to support its risk determination because it did not properly document how it developed and considered risk factors during its fiscal year 2018 risk assessment. Until DOE revises its department-level process for conducting improper payment risk assessments, it cannot ensure that the process produces a reliable assessment of whether it is susceptible to significant improper payments. Further, the process for the OCFO to oversee the accuracy of payment site risk ratings is unclear because DOE has not defined and documented, in its policies and procedures, the process for OCFO reviewers to override a payment site’s risk ratings in the event the reviewer finds that the rating was not reasonable. By developing, documenting, and implementing department-wide policies and procedures, DOE would enhance its ability to adequately monitor its decentralized improper payment risk assessment process and help ensure that individual payment reporting sites accurately score their risk factors—leading DOE to obtain a more accurate and reliable assessment of its overall risk of susceptibility to improper payments. Recommendations for Executive Action We are making the following nine recommendations to DOE: The Office of the Chief Financial Officer should require payment reporting sites to document their procedures for identifying, tracking, and reporting improper payments to ensure they provide consistent and comparable information about their improper payments over time. (Recommendation 1) The Office of the Chief Financial Officer should develop a monitoring process to ensure that payment reporting sites document and implement procedures that will enable them to correctly identify and report improper payments to the OCFO. (Recommendation 2) The Office of the Chief Financial Officer should require payment reporting sites to document policies for tracking questioned costs to resolution. (Recommendation 3) The Office of the Chief Financial Officer should track information on the year the payment occurred for all improper payments, regardless of when they are identified, and determine and disclose in DOE’s AFR whether the department’s total annual improper payments exceeded $100 million in any given year. (Recommendation 4) The Office of the Chief Financial Officer should clarify guidance to (1) define the factors for assessing adequacy of payment reporting sites’ justifications that conducting recapture audits would not be cost-effective, and (2) require that the Office of the Chief Financial Officer review the sufficiency of these justifications against the criteria defined. (Recommendation 5) The Office of the Chief Financial Officer should evaluate whether payment reporting sites could identify enough additional improper payments through payment recapture audits to make those audits cost- effective, such as by performing audits at selected sites. (Recommendation 6) The Office of the Chief Financial Officer should revise DOE’s department- level process for conducting improper payment risk assessments to include (1) developing and documenting the rationale for the variable scale used to score risk factors and weighting of the payment reporting sites; and (2) documenting DOE’s consideration of the inherent risk associated with the lag in identifying certain improper payments subsequent to the fiscal year they occurred to ensure that the process results in a reliable assessment of whether the department is susceptible to significant improper payments. (Recommendation 7) The Office of the Chief Financial Officer should revise DOE’s department- level policies and procedures for reviewing risk assessments submitted by payment reporting sites to require a review and approval of the documentation supporting these assessments to help ensure the accuracy of the sites’ assessments. (Recommendation 8) The Office of the Chief Financial Officer should revise DOE’s department- level policies and procedures for conducting improper payment risk assessments to define the process for overriding a payment reporting site’s risk determination, when appropriate. (Recommendation 9) Agency Comments and Our Evaluation We provided a draft of this report to DOE for review and comment. DOE concurred with six of our recommendations and said that it plans to complete actions from November 2020 through December 2021 to address these recommendations. DOE did not concur with three of our recommendations; however, we believe that these recommendations remain valid. DOE’s written response is reproduced in appendix III and summarized below. In addition, DOE provided technical comments, which we incorporated as appropriate. DOE did not concur with our sixth recommendation to evaluate whether its payment reporting sites could identify enough additional improper payments through payment recapture audits to make those audits cost- effective, such as by performing audits at selected sites. In response to this recommendation, DOE stated in its comments that it has an ongoing Fraud Risk Management Working Group and that officials have developed a Fraud Risk Management and Data Analytics Implementation Plan to strengthen DOE’s capability to prevent, identify, and recover improper payments and fraud. However, DOE’s plan is still in draft form, and according to DOE’s technical comments, they will not begin using data analytics until fiscal year 2021. In addition, DOE stated in its comments that existing payment recapture activities such as pre- and post-payment reviews, contractor internal audits, use of the results of cost allowability audits of integrated contractors, and interim and close-out reviews of contracts and financial assistance awards are sufficient. As we discuss in the report, DOE determined that it does not need to conduct payment recapture audits based on justifications submitted by the reporting sites. However, most of the sites’ justifications did not include consideration of the OCFO criteria for making determinations about the cost-effectiveness of conducting payment recapture audits. We continue to believe that by evaluating whether it could identify enough additional improper payments to make payment recapture audits cost-effective, such as by performing audits at a limited number of sites, DOE would have an opportunity to identify and recover additional improper payments or have better information to justify that payment recapture audits are not cost-effective. DOE did not concur with our seventh recommendation to (1) develop and document the rationale for weighting risk factors, including the weighting of all payment reporting sites; and (2) document its consideration of the inherent risk associated with the lag in identifying certain improper payments subsequent to the fiscal year they occurred to ensure that the process results in a reliable assessment of whether the agency is susceptible to significant improper payments. Regarding the weighting of risk factors, DOE said that its risk assessment evaluates the volume and dollar amount of payments by payment category, payments subject to manual controls, and fluctuations in volume and dollar amounts. We recognize that DOE’s risk assessment template asks each site to assess its risk with regard to payment amounts and fluctuations. However, we are recommending that the OCFO document the weighting of all its risk factors, including its decision to consider as equal the risks identified by all sites—regardless of the dollar amount of outlays. While assessing the risk of improper payments at an individual site is important, it does not address the intent of our recommendation. We continue to believe that, because DOE did not properly document how it developed and considered risk factors during its fiscal year 2018 risk assessment, it cannot ensure that the process produces a reliable assessment of whether DOE is susceptible to significant improper payments. Regarding the consideration of inherent risk, DOE stated in its comments that the Payment Integrity Risk Assessment directs payment reporting sites to consider inherent risk as part of DOE’s Internal Control Program. We recognize that sites are to assess the inherent risk that an improper payment may occur. However, even if none of the sites identifies the known lag in identifying improper payments as a risk, based on our review of DOE’s AFRs, this lag is a risk to DOE as a whole. Therefore, we continue to believe that DOE should document in its risk assessment process its consideration of the known lag in identifying improper payments. Finally, DOE did not concur with our eighth recommendation to revise DOE’s department-level policies and procedures for reviewing risk assessments. Specifically, we recommended a policy revision to require OCFO review and approval of documentation submitted by payment reporting sites in support of their risk assessments to help ensure the accuracy of these sites' assessments. DOE stated in its comments that sufficient processes are in place for ensuring the accuracy of payment reporting sites’ risk assessments. DOE also stated that OCFO’s Payment Integrity Guidance instructs payment reporting sites to maintain detailed information supporting risk assessments, which is available to the OCFO and DOE’s auditors upon request, and that review and approval of the documentation occurs during periodic payment reporting site visits by OCFO staff. Further, DOE stated that as part of the OCFO’s quality assurance reviews, the OCFO evaluates the documentation used to support risk assessment ratings and directs updates to risk assessments if documentation listed does not support the stated risk ratings. As we discuss in the report, five of the 10 sites we reviewed did not provide sufficient explanation or documentation supporting their ratings for several of the risk factors. This includes one site that cited “discussions with team lead” as the primary source of support for the ratings it assigned for several risk factors. We continue to believe that by developing, documenting, and implementing policies and procedures to require the OCFO to review documentation supporting payment site risk assessments, DOE would enhance its ability to adequately monitor its decentralized improper payment risk assessment process and help ensure that individual payment reporting sites accurately score their risk factors, leading DOE to obtain a more accurate and reliable assessment of its overall risk of susceptibility to improper payments. We are sending copies of this report to the appropriate congressional committees, the Secretary of Energy, the Administrator of the National Nuclear Security Administration, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff members have any questions about this report, please contact Allison Bawden at (202) 512-3841 or bawdena@gao.gov; or Beryl Davis at (202) 512-2623 or davisbh@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix IV. Appendix I: The Department of Energy’s Payment Reporting Sites The Department of Energy (DOE) has 48 payment reporting sites that are responsible for conducting improper payment risk assessments and annually providing data on actual improper payments to DOE’s Office of the Chief Financial Officer (OCFO). The 48 sites consist of six types, four of which are types of federal entities and two of which are types of contractors. The four types of federal entities are Headquarters, DOE field sites, Power Marketing Administrations, and the Federal Energy Regulatory Commission. The two types of contractors are management and operating (M&O) contractor and non-M&O contractor. Table 5 lists the 48 payment reporting sites and provides the fiscal year 2017 outlays and improper payments data they reported to the OCFO for DOE’s fiscal year 2018 Agency Financial Report (AFR). Appendix II: Additional Details of the Department of Energy’s Fiscal Year 2017 Improper Payments Reported by Payment Reporting Sites The Department of Energy’s (DOE) Office of the Chief Financial Officer (OCFO) requires the payment reporting sites to provide some details about their improper payments that were not required to be included in the department’s Agency Financial Report (AFR) during the period under review. These details include information about how the improper payments were identified and the reasons why the payments were determined to be improper. As shown in figure 4, two methods accounted for most of the current year improper payments identified by DOE in fiscal year 2017: post-payment review (57.6 percent) and self-reporting (22.1 percent). As shown in figure 5, there was a broader range of reasons payments were determined to be improper in fiscal year 2017, although the majority (54.7 percent) were attributable to settlements as the result of litigation. Appendix III: Comments from the Department of Energy Appendix IV: GAO Contacts and Staff Acknowledgments GAO Contacts Staff Acknowledgments In addition to the contacts named above, Hilary Benedict (Assistant Director), Michelle Philpott (Assistant Director), Kathy Pedalino (Analyst in Charge),Taya Tasse (Auditor in Charge), Perry Chen, Andy Furillo, Isabella Guyott, Latesha Love, Laura Pacheco, and Farrah Stone made key contributions to this report. Also contributing to this report were Kevin Bray, John Delicath, James Kernen, Jason Kirwan, Dan C. Royer, and Anne Thomas.

Why GAO Did This Study Improper payments—payments that should not have been made or were made in an incorrect amount—are a significant problem in the federal government. Agencies are required to perform risk assessments to identify programs that are susceptible to significant improper payments. House Report 115-697 included a provision for GAO to review DOE's system for tracking improper payments. This report examines the extent to which (1) the amounts reported in DOE's AFRs for fiscal years 2015 through 2019 were accurate and complete, and (2) its fiscal year 2018 risk assessment provided a reasonable basis for its risk determination. GAO reviewed DOE's improper payment reporting for fiscal years 2015 through 2019 and its fiscal year 2018 risk assessment, and reviewed documents and interviewed officials from 10 of 48 reporting sites selected to provide a range of sites and about half of fiscal year 2018 reported improper payments. What GAO Found The improper payments amounts that the Department of Energy (DOE) reported in its annual agency financial reports (AFR) for fiscal years 2015 through 2019 may not be accurate or complete. Agencies with programs that are susceptible to significant improper payments—including those with more than $100 million of improper payments in a year—are required to report statistically valid estimates of their improper payments. DOE determined these requirements did not apply, but optionally reported information on actual improper payments it made and identified in the prior year. For example, in its fiscal year 2019 AFR, DOE reported fiscal year 2018 improper payments—such as those made to contractors for unallowable costs—totaling about $36 million, less than 0.1 percent of its outlays. However, DOE did not disclose that these amounts do not include improper payments identified through reviews, audits, and investigations completed several years after it issues its AFR (see figure). For example, as of September 2019, DOE had not audited $23.8 billion of its $38.5 billion in fiscal year 2018 outlays. Such audits may increase the improper payments in a year by millions of dollars. For example, based on a 2017 audit, DOE identified $34 million in fiscal year 2010 improper payments. DOE does not always track information on the year improper payments were made that would allow it to determine whether improper payments identified later would increase the total to more than $100 million. By tracking and disclosing such information, DOE could better inform Congress, the public, and others about whether it exceeded the $100 million threshold and should be subject to additional reporting requirements. DOE determined that its risk of significant improper payments was low in its fiscal year 2018 risk assessment. However, GAO found that the risk assessment may not provide a reasonable basis for DOE's determination. DOE did not provide sufficient documentation to support that it considered the known lag in identifying improper payments as an inherent risk, nor did it provide sufficient documentation to support its rationale for the scale it used to score risk factors or for weighting risk ratings of payment reporting sites. For example, a payment site processing $3 million of outlays had the same weight in the overall assessment as a payment site processing $5.7 billion of outlays. As a result, DOE cannot demonstrate that its low-risk determination is reasonable and that its risk assessment process produces reliable results. What GAO Recommends GAO is making nine recommendations to DOE, including to track and disclose information on improper payments identified later and determine whether these payments exceeded $100 million in any year, and to revise its risk assessment process to ensure the process has a reasonable basis and reliable results. DOE agreed with six of the recommendations, but did not agree with three recommendations, including to revise its risk assessment process. GAO maintains that the recommended actions are valid.

gao_GAO-20-417

gao_GAO-20-417_0

Background DHS and Its Components DHS has 15 components involved in achieving its broad strategic goals of countering terrorism and homeland security threats, securing U.S. borders and sovereignty, securing cyberspace and critical infrastructure, preserving U.S. prosperity and economic security, and strengthening preparedness and resilience. DHS relies on contracts to support these missions and has 10 contracting activities with authority to procure products and services within and across DHS’s components. For example, OPO within DHS’s Management Directorate is responsible for contracting for a number of DHS’s components and offices, including the Science and Technology Directorate, the Cybersecurity and Infrastructure Security Agency, and the Countering Weapons of Mass Destruction Office. See appendix II for DHS’s organizational chart, identifying operational and support components and contracting activities. FAR and DHS Guidance The FAR requires that agencies take certain steps when identifying and developing requirements that need to be addressed through the execution of a contract. For example, the FAR requires that agencies conduct market research, as appropriate, and defines market research as the process used to collect and analyze information about capabilities in the market that could satisfy an agency’s needs. While the extent of market research will vary depending on characteristics of the requirement, the FAR provides general policies and procedures for conducting market research with the goal of arriving at the most suitable approach to acquiring, distributing, and supporting supplies and services. The FAR also requires agencies to perform acquisition planning activities for all acquisitions to ensure that the government meets its needs in the most effective, economical, and timely manner possible. In addition to the FAR, DHS relies on the Homeland Security Acquisition Regulation and Homeland Security Acquisition Manual—issued by DHS’s Chief Procurement Officer to implement and supplement the FAR—to establish policies and procedures for all acquisition activities within the department. For example, together the Homeland Security Acquisition Regulation and Homeland Security Acquisition Manual provide more specific department-wide policies and procedures for implementing acquisition requirements laid out by the FAR, such as competition, acquisition planning, and market research. Contracting activities may also implement their own procedures that support and implement the FAR, Homeland Security Acquisition Regulation, and Homeland Security Acquisition Manual. DHS also has its own policies and guidance for managing its service acquisition programs. For example, DHS generally defines major acquisition programs as those with life-cycle cost estimates of $300 million or more. However, DHS’s Acquisition Management Instruction 102-01-001 identifies additional thresholds for approval of stand-alone service acquisition programs—service contracts that are not part of a larger acquisition program. Specifically, service acquisition programs with annual cost estimates of $1 billion or more, or between $100 million and $1 billion are identified as Major Level 1 or 2 acquisition programs, respectively, and generally require approval from DHS’s Chief Acquisition Officer. Service acquisition programs with annual cost estimates under $100 million can be approved at the component level in accordance with component policies and processes. As of November 2019, DHS did not have any service programs identified as Major Level 1 or 2. Service Functions Requiring Heightened Management Attention In response to the 2009 Presidential Memorandum on Government Contracting, OFPP, within OMB, issued a policy letter in September 2011 to all executive agencies—including DHS—to clarify, in part, when governmental outsourcing of services is and is not appropriate. Specifically, the letter defines inherently governmental functions, according to the definition in the Federal Activities Inventory Reform Act, as those that are so intimately related to the public interest as to require performance by federal employees, such as determining agency policy or budget requests. Additionally, it identifies categories of service functions that agencies are allowed to contract for, but that require heightened management attention, as they pose a risk to the government losing control of either its responsibility to perform inherently governmental functions or its mission and operations. Figure 1 illustrates the increasing risk related to contracting for these types of functions. The letter also provides guidance on managing the performance of closely associated with inherently governmental and critical service functions, among others. In 2010, OFPP had also identified categories of services requiring heightened management attention. The three categories of service functions requiring special or heightened management attention follow: 1. Closely associated with inherently governmental functions. The 2011 OFPP policy letter adopts a single definition of an inherently governmental function, clarifies the types of services that constitute those closely associated with inherently governmental functions, and highlights the steps that agencies must take to ensure that the contractor does not ultimately perform functions that are reserved exclusively for federal employees. The response to public comments in the 2011 OFPP policy letter—in accordance with the FAR— provides the example that aspects of acquisition planning, such as determining requirements and approving a contract strategy, are inherently governmental functions. However, contractors may be used to support acquisition planning efforts through functions such as performing market research or drafting statements of work. These supporting functions are deemed closely associated with inherently governmental functions and can be contracted for. However, the OFPP policy letter states that agencies are required to take certain steps—such as assigning a sufficient number of qualified government employees to perform contract management—to ensure, among other things, that the contractor does not perform, interfere with, or undermine the integrity of the agency’s decision-making responsibilities. 2. Critical functions. The 2011 OFPP policy letter describes critical functions that, when contracted for, pose a risk that the agency could lose control of its mission and operations. Among other things, the policy established the criteria for identifying critical functions that are internally unique to each agency based upon their mission and operations. As an example, the 2011 OFPP policy letter notes that analyzing areas of tax law that impose significant compliance burdens on taxpayers may constitute a critical function for the Internal Revenue Service’s Office of the Taxpayer Advocate. OFPP notes that when contracting for a critical function, agencies must retain sufficient internal capability either through: dedicating an adequate number of federal personnel to perform the function in-house or alongside the contractors in the event the contractor fails to perform; or ensuring federal personnel are available to oversee and manage the contractor workforce. 3. Special interest functions. Special interest functions, according to OFPP, are functions that required increased management attention due to heightened risk of workforce imbalance. Some special interest functions may also be either closely associated with inherently governmental or critical functions. According to OFPP, contracting for these functions also poses a risk that the agency can lose control of its mission and operations. In a November 2010 memo, OMB instructed agencies to identify and analyze a list of product and service codes to be deemed special interest functions. DHS, with OMB approval, has chosen 17 product and service codes to categorize as special interest functions, including policy review and development and acquisition support services. To mitigate the risk associated with contracting for special interest functions, agencies are required to analyze their contracts for special interest functions annually to ensure the mix of federal employees to contractors is appropriately balanced. For examples of functions deemed closely associated with inherently governmental, critical, and special interest, see appendix III. Evolution of DHS’s Identification of Inherently Governmental, Closely Associated with Inherently Governmental, Critical, and Special Interest Functions Since March 2019, DHS has required program officials to complete its Inherently Governmental and Critical Functions Analysis job aid for all proposed service contract requirements above the simplified acquisition threshold—currently $250,000—with a product and service code that is not included on DHS’s exemption list. The department established the Inherently Governmental and Critical Functions Analysis job aid to enable it to systematically ensure that proposed service requirements do not include inherently governmental functions and to identify those that contain functions considered closely associated with inherently governmental or critical. The job aid collects general information about the proposed service contract, such as a brief description, followed by three discrete sections to check for these three functions. Section 1. This section includes a checklist for functions that the FAR has identified as being inherently governmental, such as developing federal agency policy and determining price reasonableness of vendor bids. In order to proceed with contracting for the service, the program official has to certify that none of these functions exist within the proposed requirement. Section 2. This section includes a checklist for functions that the FAR and OMB have identified as being closely associated with inherently governmental functions, such as conducting market research or drafting statements of work. If program officials identify any functions that are closely associated with inherently governmental in the proposed requirement, the job aid includes a narrative section where the program official is expected to input information on the nature of the work to be performed by the contractor and how heightened management attention will be given. Section 3. This section requires program officials to consider whether the proposed requirement is necessary for the agency to effectively perform and maintain control of its mission, which would designate the requirement as critical. Agencies are allowed to contract for critical functions so long as the program official certifies that the agency has sufficient internal capacity to undertake the work if, for any reason, the contractor is unable to provide the service. Special interest functions are not required to be identified in the job aid. The job aid concludes with the program official’s signature and is eventually forwarded to the contracting officer as part of the overall procurement package prior to soliciting for the proposed requirement. The job aid was put in place following the March 2019 decommissioning of DHS’s prior tool—the BWAT. The BWAT was used to implement DHS’s Balanced Workforce Strategy, which focuses on achieving the appropriate mix of federal and contractor personnel. This strategy was established in October 2009 to meet the statutory requirements in the 2009 Omnibus Appropriations Act. The 2009 Omnibus Appropriations Act directed most federal agencies—including DHS—to devise and implement guidelines and procedures to ensure that, on a regular basis, consideration is given to using federal employees to perform new functions, and functions that are performed by contractors but can be performed by federal employees. The Balanced Workforce Strategy established processes to enable DHS to achieve the appropriate mix of federal employees and contractors to accomplish the department’s mission, while minimizing risk to DHS’s missions from an overreliance on contractors. DHS implemented this strategy through the BWAT—an online questionnaire completed by individual program offices for certain service contracts. The function of the BWAT was to ensure the proposed service functions are not inherently governmental, and to identify whether the functions are closely associated with inherently governmental, critical, or special interest, among others. In addition, the BWAT recommended the ratio of federal employees to contractors needed to oversee those services. This analysis was then approved by the program and reviewed by the contracting officer as part of the procurement package. According to officials from an internal DHS working group, the Balanced Workforce Strategy—and BWAT by extension—were deemed no longer necessary based on the maturation of the department’s program and contracting officials’ ability to identify inherently governmental, closely associated with inherently governmental, critical, and special interest functions without a detailed questionnaire. In addition, the software used to conduct the BWAT was not supportable and faced obsolescence issues. For additional information on the differences between the BWAT and the job aid, see appendix IV. DHS’s Planning, Programming, Budgeting, and Execution Process DHS relies on its planning, programming, budgeting, and execution process to plan for and allocate resources—including those for service contracts—across the department. DHS uses this process to develop its Future Years Homeland Security Program—a database that contains 5- year program funding plans and is used to prepare a report to Congress—and the department’s annual budget request. According to DHS guidance, at the outset of the annual planning, programming, budgeting, and execution process, the Office of Policy and Office of Program Analysis and Evaluation under the Chief Financial Officer provides resource planning guidance to the components outlining departmental priorities. Following the identification of departmental priorities, DHS guidance states that components should consider their objectives and commitments within fiscal guidance constraints, to estimate needs in their resource plans. The components then prepare their annual resource plans, based on their needs and in line with DHS priorities, which are reviewed by DHS leadership and culminate in a document reflecting the department’s resource decisions. See figure 2 for a depiction of the planning, programming, budgeting, and execution process. Beginning with the fiscal year 2017 budget request, DHS has used the common appropriation structure to organize the information in its budget requests. This common appropriation structure is comprised of four appropriation accounts: procurements, construction, and improvements; operations and support; and federal assistance. Each of these accounts has mission oriented program/project activities that correspond to the components’ different operations. For example, ICE’s fiscal year 2020 budget request includes program/project activities for the three operational directorates that accomplish its mission— Homeland Security Investigations, Enforcement and Removal Operations, and the Office of the Principal Legal Advisor. Within the component’s program/project activity accounts, service contract requirements are reflected in budget documents through object classes prescribed by OMB. OMB guidance establishes object classes as a measure for communicating resource needs in budget justifications and identifies eight object class codes for other contracted services, as shown in table 1. Our Prior Work on Service Contracts We have conducted prior work on the use of service contracts across the federal government, including how agencies have mitigated challenges overseeing and managing risks associated with service contracts that require heightened management attention, and how agencies have identified estimated service contract needs as part of agency budget requests. Specifically: In December 2011, we reported on how the Departments of Homeland Security, Transportation, and Housing and Urban Development, the United States Agency for International Development, and the National Science Foundation considered and mitigated risks associated with professional and management support service contracts—including contracts that are considered to be a special interest function and can increase the risk that contractors inappropriately influence the government’s authority, control, and accountability for decisions. We found that these agencies generally did not consider and mitigate the risks associated with selected professional and management support service contracts prior to their award. We recommended that OMB establish a deadline for agencies to develop procedures to improve their management of risks related to professional and management support service contracts. OMB agreed with our recommendation but did not establish such a deadline. In February 2016, we reported on what insights the Department of Defense had into the military department’s use of service contracts to fulfill current and future requirements, and how the department reported on service contract requirements in its annual budget requests to Congress. We found that while program offices within the military departments generally had information on current and future service contract requirements beyond the budget year, that future service requirements through the Future Years Defense Program were not identified to Department of Defense leadership in annual budget requests because there was no requirement to do so. We also found that the Department of Defense’s budget requests to Congress did not include all planned service contract needs and that its contracted services budget exhibit intended to meet certain statutory reporting requirements significantly underreported its estimated budget request for contracted services. We suggested that Congress should consider revising statutory reporting requirements to include estimated requirements beyond the budget year. In August 2018, Congress included a provision in the National Defense Authorization Act for Fiscal Year 2019 requiring the Department of Defense to include information on planned service contract requirements in the Future Years Defense Program. We also recommended that the military departments revise budgeting guidance to collect service contract information beyond the budget year, and that the Department of Defense modify its approach for reporting on service contracts in budget exhibits to ensure that certain service contract requirements are included. The department generally agreed with these recommendations, and has taken some steps to update military department budget guidance and modified its approach for reporting service contract requirements in its budget requests. In September 2019, we reported on the extent to which the National Nuclear Security Administration reports information on service contract requirements in its congressional budget justification documents and manages potential risks of service contracts that are at risk of performing inherently governmental functions. We found that the National Nuclear Security Administration did not consistently include information on all of its service contracts in budget justification materials. We also found that the agency may not be effectively managing the risks of contractors performing inherently governmental activities because contracting officers are not required to document how they will oversee contracts for services closely associated with inherently governmental functions, and the agency does not verify that planned oversight is performed. We recommended that the National Nuclear Security Administration report on all professional support services contracts with obligations as part of its budget justification materials, ensure contracting officers document plans to oversee service contracts at risk of performing inherently governmental functions, and develop a process to ensure that contracting officers are carrying out planned oversight. The National Nuclear Security Administration generally agreed with these recommendations. DHS’s Reliance on Service Contracts to Support Its Mission, Including Those in Need of Heightened Management Attention, Has Increased DHS’s Service Obligations Are over Three-Quarters of Total Contract Obligations, and Annual Service Contract Obligations Have Increased DHS obligated about $70.7 billion, or 76 percent, of its $93.7 billion in total contract obligations on services from fiscal years 2013 through fiscal year 2018. See figure 3 for details on DHS’s obligations on services and products from fiscal years 2013 through 2018. DHS annual service contract obligations increased by 40 percent from fiscal years 2013 to 2018, from about $10.5 billion to $14.7 billion. This increase in service contract obligations was largely driven by increases in Federal Emergency Management Agency and CBP service contract obligations, which grew by $2.2 billion and $927 million respectively. In fiscal year 2018, the Federal Emergency Management Agency had the highest service contract obligations, at $3.3 billion, followed by DHS headquarters organizations, and CBP. Of the Federal Emergency Management Agency’s fiscal year 2018 service contract obligations, $2.5 billion, nearly 75 percent, were identified as disaster-related. See figure 4 for additional detail on fiscal year 2018 service contract obligations by DHS component. DHS relies on a variety of services to accomplish its missions. For example, about $2.1 billion, or 14 percent of DHS’s total fiscal year 2018 service contract obligations, were for guard services to protect federal buildings or other security needs. DHS obligated about $2 billion, or 13 percent of its total fiscal year 2018 service contract obligations, towards various information technology and telecommunications services—such as satellite services and hardware and software maintenance. DHS’s five service categories with the highest amount of contract obligations in fiscal year 2018 accounted for about 40 percent of its total service contract obligations that year. See figure 5 for additional details on DHS’s top service obligations. DHS Continues to Use Service Contracts in Need of Heightened Management Attention In fiscal year 2018, 65 percent of DHS’s total service contract obligations were for services in need of heightened management attention or oversight due to being a closely associated with inherently governmental, critical, or special interest function. DHS’s obligations on contracts for these types of services increased by about 58 percent, from about $6 billion in fiscal year 2013 to $9.5 billion in fiscal year 2018. See figure 6 for additional details on the proportion of contract obligations for services in need of heightened management attention over time. Within our selected components, obligations for service contracts in need of heightened management attention increased the most from fiscal years 2013 to 2018 for contracts awarded by ICE—increasing by $732.1 million. CBP’s obligations for service contracts in need of heightened management attention increased over this time frame by $598 million. Service contracts in need of heightened management attention accounted for more than three quarters of all service contract obligations in fiscal year 2018 for DHS headquarters organizations and ICE. See figure 7 for additional detail on fiscal year 2018 contract obligations for services in need of heightened management attention by DHS component. Processes Are in Place to Identify and Develop Service Contract Requirements, but DHS Does Not Have an Approach to Consistently Address Certain Service Procurements DHS and Components in Our Review Have Guidance and Processes for Identifying and Developing Service Contract Requirements DHS has policies and guidance to identify its service and product needs and develop contract requirements. In addition to the FAR, Homeland Security Acquisition Regulation, and Homeland Security Acquisition Manual, which combined establish DHS’s acquisition regulations and contracting policies, DHS has developed additional guidance specific to identifying needs and developing contract requirements. For example, DHS’s Developing and Managing Contract Requirements Desk Guide for the Acquisition Workforce is available to program personnel as a resource for how to define requirements, including processes and required documents and templates. DHS has also developed guidance for program and contracting officials for specific activities related to the requirements development process—such as market research, acquisition planning, and source selection guides—as well as guidebooks for specific participants involved in identifying needs and developing contract requirements, such as the contracting officer’s representative. Based on DHS policies and guidance, we identified key processes DHS undertakes to identify needs and develop contract requirements for services and products. Of these key processes, assessing for inherently governmental functions is specific to DHS’s development of service requirements. In response to the 2011 OFPP policy letter’s requirements to screen service contracts for the performance of inherently governmental functions and consider how contractor employees are used to perform agency functions, DHS implemented the BWAT in 2013. As previously noted, this tool has now been replaced by the Inherently Governmental and Critical Functions Analysis job aid. These tools have been required for service contracts specifically to ensure that contractors are not performing tasks that should be reserved for federal employees. Once completed, the output from these tools are reviewed by the contracting officer and included in the procurement package. Figure 8 summarizes key processes we identified that DHS uses to identify and develop service requirements. In addition to the policies and guidance DHS has for identifying and developing service requirements, DHS components in our review have implemented additional guidance and tools. For example, USCIS has developed specific guidance to support the program office’s development of requirements, including information on how to define requirements, conduct market research, and develop a cost estimate and acquisition strategy. Further, all of the components in our review reported using tools, such as templates and checklists, to help guide program and contracting officials through the requirements development process. For example, all of the components in our review use templates for market research, acquisition plans, and requirements documents that identify what information officials should include in these documents. The components in our review also provided program and contracting officials with checklists for what documents are required in the procurement package, depending on the type of contract being solicited. Some of the components in our review maintain this information on acquisition websites that serve as repositories for DHS and component guidance, templates, and other requirements. For example, ICE’s Office of Acquisition Management’s portal provides guidance, documents, and templates by phase of the acquisition process, from acquisition planning and solicitation preparation through contract administration and close-out. DHS components in our review also relied on subject matter experts to assist in their requirements development efforts, with the level of involvement varying depending on the requirement. Specifically, officials associated with two of the eight contracts in our review stated they used integrated product teams to assist with developing their service requirements. For example, officials involved in requirements development for services at USCIS’s 135 Application Support Center locations told us they established an integrated product team with program officials, the contracting officer, cost estimators, Field Office Directorate personnel, and Office of Security and Integrity personnel. Officials from the other six contracts relied on more informal subject matter expert involvement. Component officials from three of our contracts that relied on more informal coordination methods said that when the requirement is recurring and has previously been contracted for, formal coordination through an integrated product team may not be necessary. DHS Has Not Fully Developed an Approach to Ensure Certain Service Procurements Are Clearly Defined and Consistently Reviewed DHS has established a process for reviewing the procurement strategy for certain service and product procurement actions prior to award, but has not developed an approach to ensure proposed service contract requirements are clearly defined or that it is consistently reviewing what DHS considers to be high-risk service procurement actions. In 2018, OCPO and the Office of Program Accountability and Risk Management began piloting a DHS-wide Service Requirements Review to validate, optimize, prioritize, and approve service requirements early in the development process. However, DHS discontinued these efforts before the pilot was finalized. According to DHS officials, they initiated this pilot because there had been no consistency or rigor for reviewing service contract requirements even though these contracts account for over 70 percent of DHS’s contract obligations. According to DHS documents and officials, the main objectives of the pilot were to: ensure service requirements are clearly defined and reviewed before planning how the services are obtained; assess whether the services should be provided in whole or in part by foster collaboration and opportunities to leverage efficiencies for similar service requirements to avoid duplication in services across the department; and assess whether the requirement should be managed as a service acquisition program. To accomplish these objectives, DHS identified stakeholders from within DHS’s Management Directorate to be headquarters-level reviewers for service requirements based on the type of service being contracted for. However, according to OCPO and Office of Program Accountability and Risk Management officials, the pilot was discontinued in April 2019 before any service requirements were reviewed because it was determined to be too resource intensive. According to DHS officials, the discontinuation of the Service Requirements Review pilot coincided with the implementation of the Procurement Strategy Roadmap, a separate OCPO-led initiative to review and approve the procurement strategy for all service and product acquisitions with a total estimated value over $50 million. The Procurement Strategy Roadmap requires contracting activities, along with their procurement teams, to present and discuss the procurement strategy with the DHS Chief Procurement Officer, members of OCPO, and other stakeholders as needed, prior to drafting an acquisition plan or other decision documents. According to OCPO officials, it was intended to require procurement staff to meet with OCPO officials early in the acquisition planning process, prior to the service contract requirement being finalized, to discuss how services and products would be purchased. Specifically, the Procurement Strategy Roadmap is intended to address what OCPO considered as key elements of the procurement process, such as the requirement, competition, the availability of strategic sourcing or small business options, and contract type. Following the discontinuation of the Service Requirements Review pilot in April 2019, OCPO and Office of Program Accountability and Risk Management officials discussed expanding the Procurement Strategy Roadmap to incorporate some elements of the Service Requirements Review pilot, including reviewing proposed requirements to determine if they are clearly defined and valid, when appropriate. For example, OCPO officials said they have included the Office of Program Accountability and Risk Management and the Office of the Chief Information Officer to facilitate additional DHS stakeholder involvement in some reviews, and to broaden the discussion beyond how services and products will be purchased and include what the requirement is and whether it needs to be purchased at all. However, as of February 2020, OCPO officials told us that reviewing requirements to ensure they are clearly defined and collaborating with additional DHS stakeholders to identify opportunities to leverage existing service requirements was not the intent of the Procurement Strategy Roadmap. For example, OCPO officials stated that proposed requirements may only be reviewed by additional DHS stakeholders during the Procurement Strategy Roadmap if the requirement is new, “unique,” or “high risk,” and that this decision is based on their review of the information in the Procurement Strategy Roadmap and professional judgment. When asked what constitutes a unique or high risk requirement, officials told us a proposed requirement could be high risk if it had historical procurement issues, but noted that ultimately the decision to review the requirement and whether to involve additional DHS stakeholders in that review is subjective and based on whether OCPO leadership believes other stakeholders may add value in developing and reviewing the proposed requirement. According to OCPO officials, some Procurement Strategy Roadmap requirements supporting major acquisition programs undergo separate review by DHS headquarters stakeholders in the Acquisition Review Board. However, high-dollar service acquisitions that are not associated with a major acquisition program or not above $100 million do not currently receive headquarters- level scrutiny to determine whether requirements are clearly defined or to leverage efficiencies and buying power for similar service requirements across the department. In addition, OCPO has not established a process to ensure it is consistently reviewing proposed procurement actions through the Procurement Strategy Roadmap. Our review of the fiscal year 2019 Procurement Strategy Roadmap eligible procurement actions found that OCPO subjectively waived the review for 18 of the 49 eligible actions— over 36 percent of the actions that should have been subject to a Procurement Strategy Roadmap. The waived procurement actions included three out of six Federal Emergency Management Agency actions, eight out of 16 OPO actions, four out of 12 CBP actions, and two out of six Transportation Security Administration actions. According to OCPO officials, the decision to waive a procurement action is a subjective one, made by OCPO leadership based on the initial information provided. For example, officials said the review may be waived if the procurement action is recurring or will be fulfilled using an already established DHS contract vehicle. We found, however, that the subjective decision to waive the reviews does not take into account other acquisition risks. For example, our review of the description of waived procurement actions found that 11 of the 18 actions were for services, including some for administrative and professional support and information technology services that DHS considers to be in need of heightened management attention. Two of the waived actions were for requirements that resulted in orders placed off General Services Administration Schedule contracts, despite an OCPO official telling us that these orders are expected to receive increased scrutiny to ensure that any existing DHS contract vehicles have been fully considered. Our review of waived procurement actions also found that OCPO waived several Federal Emergency Management Agency actions for disaster response activities and CBP actions for services at temporary soft-sided facilities used for holding detainees on the U.S.-Mexico border. Our prior work has noted challenges in requirements development and acquisition planning for these types of contracts. For example, in April 2019, we reported that contracting officers at FEMA were receiving requirements packages for disaster contracts that lacked technical specificity or had inaccurate estimates of the products and services needed. In March 2020, we also reported on acquisition planning, requirements development, and information sharing challenges with one of the waived procurement actions—a CBP delivery order for a soft-sided facility and services to hold and care for detainees—finding that these challenges led to CBP spending millions of dollars on services that were not ultimately needed. Federal internal control standards state that management should identify and respond to risk to achieve its objectives. OCPO officials acknowledged that the intent of the Procurement Strategy Roadmap was not to replace the Service Requirements Review that preceded it, and that expanding the scope of the Procurement Strategy Roadmap to review requirements would require additional resources. However, the department’s previous efforts to devote management attention to its growing proportion of service procurements are indicative of its concerns about its use of service contracts. While the Procurement Strategy Roadmap is not specific to services, it can provide a mechanism to address these concerns. Moreover, without documenting factors OCPO considers when waiving certain Procurement Strategy Roadmap eligible procurement actions, DHS is at risk of not consistently reviewing service procurement actions that could benefit from headquarters-level review. Given DHS’s reliance on service contracts, which accounted for 78 percent of DHS’s contract obligations in fiscal year 2018, developing a risk-based approach for reviewing proposed service requirements through the Procurement Strategy Roadmap or other means could help to improve DHS’s use of service contracts by identifying opportunities to leverage efficiencies and ensuring service requirements are clearly defined across the department. DHS and the Components in Our Review Inconsistently Planned for and Conducted Oversight of Selected Service Contracts in Need of Heightened Management Attention DHS Does Not Have a Formal Process for Planning and Updating Personnel Resources Needed for Service Contracts Requiring Heightened Management Attention DHS does not have a formal process for identifying all service requirements in need of heightened management attention or for planning, documenting, and updating the amount of federal personnel necessary to perform or oversee these requirements. In order to maintain control of their mission and performance of inherently governmental functions, part of contracting for services in need of heightened management attention—including functions that are closely associated with inherently governmental, critical, and special interest—is ensuring that agencies dedicate an adequate number of federal employees to oversee these functions. Specifically, OFPP notes that prior to contract award, for services that require heightened management attention agencies should complete an analysis that among other things and depending on the service, establishes that they can: retain sufficient capacity and capability to give heightened management attention to contractor performance or retain control of its operations; limit or guide the contractor’s exercise of discretion; ensure reasonable identification of contractors and contractor work avoid or mitigate conflicts of interest. Functions Requiring Heightened Management Attention The Office of Management and Budget’s (OMB) Office of Federal Procurement Policy (OFPP) guidance identifies three categories of service contracts requiring heightened management attention—those closely associated with inherently governmental functions, critical functions, and special interest functions. Depending on the function, these categories of service contracts may involve contractor work products that support policy development and program evaluation, and other tasks that are essential to the agencies’ ability to perform its mission. According to OMB, these contracts require management attention to ensure that they do not result in the performance of inherently governmental functions by the contractor and that agencies retain control of their mission and operations. DHS officials stated that, as of March 2019, they use the Inherently Governmental and Critical Functions Analysis, or job aid, to screen proposed service requirements to ensure that there are no inherently governmental functions and to identify functions that may be contracted for that are closely associated with inherently governmental or critical. If a function is identified as closely associated with inherently governmental or critical, program officials must certify that there is sufficient internal capacity to oversee contractor activities and maintain control of its missions and operations. Further, if a function is closely associated with inherently governmental, the job aid includes a narrative section where the program office should document mitigation strategies to ensure heightened management attention and enhanced oversight occur throughout the life of the contract. We found that the job aid does not provide a place to identify special interest functions that require heightened management attention. In addition, the job aid does not require program officials to analyze or document the expected federal personnel necessary to perform or oversee service requirements in need of heightened management attention following contract award; therefore information available for planning purposes is limited. We analyzed all nine of the 27 completed job aids that included closely associated with inherently governmental functions, and found that none included any calculation of federal oversight personnel necessary or mentioned the federal personnel who will be expected to perform oversight activities. The narrative section of the job aid instructs components to document mitigation strategies for functions identified as closely associated with inherently governmental functions. We found that two of the nine job aids identified mitigation strategies, such as noting that federal employees will ensure the contractor’s presence is announced at all meetings. However, neither provided any detail about who—such as the program manager or contracting officer representative—would be responsible for performing and overseeing the contractor employees performing the contracted functions or tasks. Program and contracting officials from ICE, CBP, and USCIS stated that analyzing and documenting the expected federal oversight personnel necessary prior to contract award with the BWAT provided visibility—both within the program and across the component—into resource needs. However, only ICE continues to analyze and document federal oversight necessary outside of the job aid. Following the BWAT’s decommissioning, ICE established a Service Contract Review Template for all service contracts above $1 million. ICE program officials are expected to complete this template with information such as a description of the requirement, the anticipated product and service code, identification of special interest functions, expected number of contractors needed, the number of federal employees available to oversee the work, and a justification for outsourcing the requirement. From this information, the template produces a recommended percentage of federal personnel necessary to perform management oversight. ICE created this new process because it wanted to ensure that it has a repeatable, documented decision-making process that helps plan oversight, such as the proper balance of federal and contractor employees and determining the reasonableness of the contract. DHS’s job aid also does not provide a process to update oversight needs if the contracted tasks or functions change throughout the life of the service contract. Officials from three of the four components in our review reported not having a formal process for updating federal oversight needs when elements of the contract change—such as an increase in the number of contractor personnel performing tasks or a change in scope. For example, component program and contracting officials told us that, although one of the service contracts in our review experienced an increase in the number of contractor personnel, they did not update planning for federal oversight personnel needs. In contrast, officials from ICE reported having a process to reevaluate federal oversight needed that is triggered by specific contract funding actions. Funding actions that trigger the process include: establishing a new contract, exercising an option on an existing contract, or adding funding to a service contract. ICE officials explained that through this process they review the service contract for changes to the number of contractors and whether the current oversight levels are sufficient. If they find that current oversight levels are no longer sufficient, ICE officials stated that they would require the program office to develop a risk mitigation strategy, such as assigning additional oversight personnel or increasing the contractor’s reporting requirements. DHS headquarters officials told us they no longer have a formal process for analyzing and documenting federal oversight requirements because the department has matured since implementing the BWAT, and program and contracting staff are aware of how to plan for federal oversight requirements for service contracts in need of heightened management attention. Specifically, during our review, DHS and component officials from OPO and USCIS stated that they rely on their program and contracting officials’ historical knowledge and professional judgment to determine and communicate oversight needs informally at the component level. However, we found a lack of understanding and inconsistencies in how oversight was analyzed and documented prior to the BWAT’s decommissioning. Specifically, 25 of the 75 required BWATs for special interest functions we reviewed either could not be provided or did not contain the information used to calculate and, therefore, plan for sufficient federal employees to conduct oversight. In addition, according to DHS documents and officials, the department plans for federal oversight personnel needs more broadly through its annual workforce planning efforts; and therefore, it is not necessary to analyze federal oversight personnel needs at the contract level. Yet we found that DHS’s fiscal year 2018 annual workforce plan focused on DHS and government-wide mission critical occupations, like Border Patrol Agents and Transportation Security Administration Officers. The plan does not address oversight needs based on services in need of heightened management attention (i.e., contracted functions that are closely associated with inherently governmental, critical, or special interest). While DHS’s workforce plan accounts for government-wide mission critical occupations, such as contracting officers and specialists, there is not the same level of consideration given to program managers, employees who serve as contracting officer’s representatives, or other program staff that are responsible for performing oversight at the contract level. According to OCPO officials, program officials completing the job aid should document in the narrative section the federal personnel responsible for ensuring the task does not become inherently governmental. However, we found that the job aid instructions do not address how program officials should analyze or document the federal personnel who will be tasked with conducting oversight. The job aid also does not include similar instructions, or provide space, to depict this information for functions identified as critical. Moreover, although there is guidance on when an initial job aid needs to be completed, there is no guidance indicating when, and under what circumstances, program and contracting officials may need to update federal oversight needs based on changes to the functions or task being performed by the contractor. Officials associated with only three of the eight contracts in our review reported receiving some training on the new job aid, but OCPO officials explained that they have not provided additional training beyond the instructions in the acquisition alert that implemented the job aid. Federal internal control standards state that agency’s management should use and internally communicate quality information to achieve the entity’s objectives. OCPO officials told us that components—such as ICE—have the discretion to establish additional processes for identifying and calculating federal oversight beyond what is required by the job aid. However, without consistently identifying all service requirements in need of heightened management attention and establishing a repeatable process across the department for analyzing, documenting, and updating the federal personnel needed to perform or oversee the requirement when changes occur, program and contracting officials lack reasonable assurance that they are dedicating an adequate number of federal employees to oversee these functions. This places DHS components at risk of inconsistently planning federal oversight necessary to ensure the department retains control of its missions and the performance of inherently governmental functions. DHS Components Are at Risk of Not Conducting Needed Oversight Tasks DHS components included in our review are at risk of not conducting the oversight tasks and safeguards necessary to ensure that, once the contract has been awarded, the contractor’s functions are performed in a way so as to not become inherently governmental, and that DHS retains sufficient internal capability to retain control of its mission for functions that are closely associated with inherently governmental, critical, or special interest. The 2010 Consolidated Appropriations Act states that agencies should have specific safeguards and monitoring systems in place to ensure the work that contractors are performing has not changed or expanded during performance to become an inherently governmental function. Additionally in 2010, OMB issued a memo that states agencies shall conduct meaningful analysis—through the annual service inventory—focused on special interest functions that require heightened management attention to ensure proper workforce balance. Based on our review of contract documentation and interviews with program and contracting officials associated with the eight contracts in our review, oversight of these service contracts in need of heightened management attention focused largely on assessing the quality of specific contractor tasks. Oversight of these contracts did not include a focus on ensuring the work of the contractors is not performed in a way so as to become inherently governmental, or that DHS retains sufficient internal capability to perform its missions. While assessing quality is important in monitoring contractor performance, it does not allow DHS to identify when tasks beyond what is detailed in the contract—including tasks that are potentially inherently governmental and require that final agency action reflects the independent conclusions of agency officials—are being performed. According to DHS’s most recent service contract inventory analysis and OCPO officials, DHS relies on well-trained contracting officer’s representatives to monitor contractor performance for inherently governmental functions. Yet one of the eight contracts in our review has not had a certified contracting officer’s representative assigned to the contract since its award in September 2018. For the remaining seven contracts, we found that their contracting officer’s representative appointment letters—which document oversight responsibilities— mentioned performing surveillance and inspections against the contract’s performance requirements. But only two of the seven letters—both from CBP—reference performing oversight tasks focused on how the contractor is completing the work. Specifically, both appointment letters stated that ongoing reviews should be completed focusing on the way work is performed and how the government is managing service acquisitions for closely associated with inherently governmental and critical functions. However, none of the letters we reviewed identified specific safeguards—such as vetting all contractor recommendations through a panel of federal employees—that federal personnel should perform to mitigate identified concerns with contractors performing closely associated with inherently governmental or losing sufficient internal capability for performing critical functions. Similarly, program and contracting officials associated with six of the eight service contracts in our review did not identify additional oversight tasks undertaken as a result of the contract requiring heightened management attention. Rather, these officials said they assess the contractor’s performance in terms of the quality of deliverables when asked about the types of oversight tasks performed. For example, DHS headquarters officials responsible for overseeing a service contract for technical support related to the development of nuclear detection technologies stated that their oversight largely focuses on tracking the completion of tasks included in the statement of work as well as available funding. These officials did not identify any additional actions taken to address the risk of contractors working in situations that permit or might permit access to confidential business or other sensitive information—a function closely associated with inherently governmental functions in need of heightened management attention. Additionally, acquisition officials from one of the DHS components in our review stated that they have previously relied on the contractor to report if they were performing work that was not specified in the contract. While performance monitoring is crucial to ensure that the contractor is meeting the terms of the contract, it alone does not provide DHS visibility into whether work is being performed that is outside the scope of the contract or inappropriate for contractors. Program and contracting officials associated with two of the eight service contracts in our review identified safeguards they have established to prevent contractors from performing inherently governmental work. For example, program and contracting officials associated with a USCIS contract awarded to assist in the preparation of Freedom of Information Act requests stated that they have safeguards in place to ensure the contractor does not approve agency responses to Freedom of Information Act requests—an inherently governmental function, according to the 2011 OFPP policy letter. Specifically, officials associated with this contract explained that they use a software program that does not allow a user without federal employee credentials to approve a request within the system. This is an example of a safeguard that can be instituted for similar service contracts when the risk of the contractor performing the inherently governmental function of approving requests is present. Additionally, officials associated with CBP’s service contract for maintaining its unmanned aircraft systems stated that they ensure that their onsite personnel do not direct contractors to perform unauthorized tasks by requiring these personnel to report directly to the program office. Contracting and program officials’ lack of focus on safeguards to mitigate risks associated with contract functions in need of heightened management attention is due, in part, to DHS not identifying—either in guidance or training that we reviewed—a list of oversight tasks that program and contracting officials can perform. DHS’s OCPO officials explained that there are unique aspects of each contract that should drive oversight needs so they have not established any required safeguards component program and contracting officials must employ. Despite the uniqueness of each contract, officials from OPO stated that it would be helpful to have a list of identified potential oversight tasks or safeguards for service contracts in need of heightened management attention to ensure they are managing the risk of the contractor performing work outside of scope. We found that at least one federal agency has such a list available. Specifically, the Department of State’s Contracting Officer’s Representative Handbook provides a list of mitigation strategies contracting officer’s representative can employ for contracts requiring risk mitigation—such as reserving final approval authority of any contractor proposed action for federal employees only. Additional strategies listed include requiring contractor affiliation be clearly displayed on all presentation material, and conducting conflict of interest reviews when contractors are performing services that involve or relate to evaluating another contractor’s performance. Without identifying what oversight tasks or safeguards component personnel can institute to prevent contractors from performing inherently governmental functions or from affecting the ability of the agency to maintain control of its mission and operations, DHS is at risk of its personnel not knowing which steps they should take to prevent that from occurring. DHS Components Consider Service Contract Requirements When Budgeting but DHS Budget Documents Do Not Provide Visibility into Details DHS components in our review consider service contract requirements when identifying their resource needs and formulating their budget justifications, but DHS headquarters and Congress have limited visibility into requested and actual service contract requirement costs. DHS uses the planning, programming, budgeting, and execution process to allocate resources—including those for service contracts—across the department. DHS’s guidance for this process that we reviewed does not provide specific instructions for how the components should consider service contract requirements when budgeting, but program officials we spoke with said that they generally provide information on specific service contract costs, among other resource needs, to their budget offices during the programming phase. Components then include these resource needs in their budget justifications, which are submitted to DHS headquarters for review before being submitted to OMB and then Congress. Based on our review of component budget justifications, components communicate service contract requirements in three primary ways, but none provide complete visibility into service contract requirements. Object Classes: Object classes are broad spending categories identified in OMB guidance. As shown earlier in table 1, there are eight object classes for other contracted services. According to component officials, once the resource needs for service contracts are identified, they are grouped into the object classes that best represent the requirement by either program or budget officials before submitting budget justifications to DHS headquarters. However, object class codes do not provide visibility into just service contract requirements. For example, budget officials at ICE and USCIS told us that aligning service requirements across object classes is not always perfect or precise. According to ICE officials, object class codes may include other expenses, such as interagency agreements. Further, USCIS officials noted that some contract requirements can apply to multiple object classes, so how requirements are communicated by object class is subjective based on program officials’ judgment. Cost Drivers: According to DHS budget officials, cost drivers identified in budget documentation represent the requirements that make up the largest costs at the program/project activity level. Service requirements may be included as a cost driver, but only if the estimated value of the contract represents a large portion of the program/project activity’s costs. For example, ICE’s budget guidance instructs the program offices to identify major requirements that add up to at least 50 percent of the program/project activity resource needs as non-pay cost drivers. Based on that guidance, in fiscal year 2020 budget documentation, one of ICE’s service contracts included in our review—for Office of the Principal Legal Advisor document management services—is identified as a cost driver. Only one other contract included in our review—from USCIS—was identified as a cost driver. Capital Investment Exhibit: According to DHS budget officials, the five contracts with the highest dollar value supporting each component’s capital investment are identified in the component’s budget documentation. Service contracts may be included in the capital investment exhibit if they meet this criteria, but the details included are vague. For example, for each contract listed, the exhibit typically includes information such as the contract number and total value, but does not categorize whether the contract is for a product or service nor consistently provide a description of the contract itself. For the contracts in our review, one of the eight—a contract for nuclear detection technology technical support—was included in the capital investment exhibit in fiscal year 2020 budget documentation. Since component budget offices submit their proposed budget requests with service contract requirements aligned into object class codes, program/project activities, and capital investment exhibits, DHS lacks visibility into the components’ requested service contract requirement needs. For example, officials from the Office of the Chief Financial Officer stated that they do not have visibility into DHS and the components’ specific service contract requirements. Rather, officials said their visibility is limited to changes in service contract requirements that are justified as part of requested increases or decreases in components’ funds. While officials from the Office of the Chief Financial Officer stated that they can request additional information from the components on service requirements if needed, officials could not identify any specific circumstances that have led to them requesting this information for their own purposes or in response to congressional interest. See figure 9 for details on how service contract requirements are communicated to DHS headquarters in budget documentation. Moreover, although DHS obligates over three-quarters of its contract spending to services, neither the Office of the Chief Financial Officer or OCPO have full visibility into or track service contract requirement costs. For example, similar to how information is portrayed in budget documents, officials from the Office of the Chief Financial Officer stated they report obligations to Congress by object class level on a quarterly basis; therefore, visibility into service contract requirement costs is limited. Further, OCPO officials stated that they also do not have a system for tracking service contract obligations reported through FPDS- NG or otherwise. In a discussion held during the course of our review, congressional requesters expressed interest in receiving additional information and visibility into DHS’s estimated service contract requirements. Members of Congress have also previously expressed interest in having increased oversight and visibility into other aspects of DHS’s proposed spending as well as into the Department of Defense’s estimated service contract requirements. For example, DHS budget officials told us that the decision to include the top five highest dollar value contracts in its capital investment exhibits was driven by congressional interest in this type of information on service contracts. In addition, in 2009, Congress began requiring the Department of Defense to identify in its budget submission the amounts requested for its service contracts for each component, installation, or activity, excluding services related to research and development and military construction. For example, the Department of Defense has two budget exhibits that provide details on estimated service contract requirements—one that details its advisory and assistance services, and another that tracks contracted services across prior fiscal years. In February 2016, we found shortfalls in the Department of Defense’s reporting of service contract requirements in its budget documents. We recommended that it modify its approach for reporting service contracts in its budget justifications to include additional service requirements. The Department of Defense agreed with this recommendation and, in February 2016, took steps to fully report on these service categories in its service contract spending exhibit accompanying the fiscal year 2017 budget request. We have also reported on challenges with congressional visibility into DHS’s major acquisition programs in budget documents. In April 2018, we found that DHS budget practices limit Congress’s visibility into costs and recommended that DHS work with Congress to include information on operations and support funding requests for major acquisitions in its annual budget justifications. DHS agreed with this recommendation and addressed it by adding an operations and support funding information display for major acquisition programs to its congressional budget justification for fiscal year 2021. Federal internal control standards state that agency’s management should communicate quality information internally and externally to inform decisions. Although detailed information on service requirements is available at the component level, DHS’s budget justifications do not provide that level of visibility. Visibility into service requirements is especially critical given that increases in DHS’s service contract obligations—particularly those in need of heightened management attention—may pose risks to DHS maintaining control over its mission. Given these increases, additional visibility into how much of DHS’s mission is being accomplished through the use of services requiring heightened management attention could inform DHS’s decision-making on the tasks it chooses to contract for, and the balance of its federal and contractor workforce. Without working with Congress to determine the format and level of detail needed to communicate service contract requirements in budget information, DHS headquarters and Congress are at risk of not having the information for sound resource planning and decision-making related to DHS’s use of service contracts. Conclusions Service contracts play a critical role in supporting DHS’s wide range of missions, but increases in service contract obligations—including significant increases in obligations for services in need of heightened management attention—necessitate DHS’s attention as it develops, reviews, oversees, and budgets for service contract needs. DHS’s recent effort to perform a headquarters-level review of certain service and product procurement actions is a positive step in improving the department’s visibility into how it is acquiring certain services and products. However, without developing a risk-based approach for reviewing certain proposed service contract requirements to ensure they are clearly defined and valid before they are procured and consistently reviewing eligible procurement actions, DHS cannot ensure it has established the rigor needed to review its service procurements. Further, changes in DHS’s processes and a lack of agency-wide guidance for planning, documenting, and updating federal oversight personnel and activities for services in need of heightened management attention have put the department at risk of not effectively addressing whether contractors are performing inherently governmental functions. These risks could pose challenges to DHS’s ability to maintain control over its mission and operations. Ensuring DHS has guidance for planning and updating the resources needed to oversee these contracts, and identifying the types of activities that federal personnel should be performing to mitigate the risks associated with these contracts are critical to DHS’s ability to address these concerns. Finally, despite the availability of information on specific service contract requirements within component program offices, DHS does not communicate most of this information in budget documentation provided to DHS headquarters or Congress, nor is DHS currently required to do so. Given that DHS’s service contract obligations—including those in need of heightened management attention—account for more than three quarters of DHS’s total annual contract obligations, DHS is missing opportunities to make more informed strategic decisions because it does not have visibility into its current or future service requirement spending for these services. Recommendations for Executive Action We are making six recommendations to the Secretary of Homeland Security: The Secretary of Homeland Security should direct the DHS Chief Procurement Officer to, in coordination with the Office of Program Accountability and Risk Management, develop a risk-based approach for reviewing service requirements—through the Procurement Strategy Roadmap or other means—to ensure proposed service requirements are clearly defined and reviewed before planning how they are to be procured. (Recommendation 1) The Secretary of Homeland Security should direct the DHS Chief Procurement Officer to document the factors the Office of the Chief Procurement Officer considers when waiving procurement actions from its Procurement Strategy Roadmap to ensure it is consistently considering potential acquisition risks in its planning—including those specific to services. (Recommendation 2) The Secretary of Homeland Security should direct the DHS Chief Procurement Officer to update the Inherently Governmental and Critical Functions Analysis to require the identification of special interest functions. (Recommendation 3) The Secretary of Homeland Security should direct the DHS Chief Procurement Officer to update the Inherently Governmental and Critical Functions Analysis to provide guidance for analyzing, documenting, and updating the federal workforce needed to perform or oversee service contracts requiring heightened management attention. (Recommendation 4) The Secretary of Homeland Security should direct the DHS Chief Procurement Officer to develop guidance identifying oversight tasks or safeguards personnel can perform, when needed, to mitigate the risk associated with contracts containing closely associated with inherently governmental functions, special interest functions, or critical functions. (Recommendation 5) The Secretary of Homeland Security should direct the DHS Chief Financial Officer to work with Congress to identify information to include in its annual congressional budget justifications to provide greater transparency into requested and actual service requirement costs, particularly for those services requiring heightened management attention. (Recommendation 6) Agency Comments and Our Evaluation We provided a draft of this report to DHS for review and comment. In its comments, summarized below and reproduced in appendix V, DHS agreed with the third and fifth recommendations and identified steps it plans to take to address them. DHS disagreed with the first, second, fourth, and sixth recommendations. DHS also provided technical comments, which we incorporated as appropriate. DHS did not agree with the first recommendation, that OCPO, in coordination with the Office of Program Accountability and Risk Management, develop a risk-based approach for reviewing service requirements through the Procurement Strategy Roadmap, or other means, to ensure that proposed service requirements are clearly defined and reviewed before planning how they are to be procured. In its response, DHS cited Instruction 102-01-001 as codifying how DHS and its components acquire and sustain services for major acquisitions. However, as noted in our report, as of November 2019 none of DHS’s services programs rose to the level of being classified as a major service acquisition. Therefore, DHS is at risk of overlooking those service contracts that are not a service acquisition program or not associated with its major acquisitions. DHS also noted the use of existing key processes that enable it to identify needs and develop contract requirements for services. While we acknowledge in our report that DHS and selected components have these processes in place, we found they were not consistently used throughout the contracts in our review, and none can serve as a replacement for the kind of risk-based headquarters-level oversight that we believe is necessary. For example, among its processes, DHS cited the use of integrated product teams as a way to facilitate comprehensive reviews of service requirements. We noted in our report, however, that according to officials only two of the eight contracts in our review used such an approach. Further, DHS stated that other existing efforts meet the primary objectives of the Service Requirements Review pilot, thus making an additional headquarters-level review of service requirements unnecessary. However, all of these efforts were also already in place when the then Under Secretary of Management directed OCPO and the Office of Program Accountability and Risk Management to undertake its December 2018 pilot program to provide consistency and rigor to reviewing service contract requirements. Therefore, we continue to believe that given the amount DHS obligates in service contracts to support its mission, establishing a risk-based approach to review service requirements prior to and in coordination with its consideration of how those requirements are to be procured will help prevent negative acquisition outcomes and the potential for wasted resources. DHS also did not agree with the second recommendation. In its response, DHS stated that OCPO’s decision to waive a Procurement Strategy Roadmap review does not mean that the Chief Procurement Officer did not consider acquisition risks, and that it is unclear what other acquisition risks we believe are not being considered. The recommendation to document the factors considered when waiving the Procurement Strategy Roadmap is intended to ensure that the department is able to consistently apply a framework and maintain institutional knowledge—particularly given the risks and challenges that vacancies in top leadership positions throughout the department could pose to addressing management issues. Waiving procurements without documentation of what acquisition risks are being considered puts the department at risk of inconsistently making those decisions and not being able to leverage Procurement Strategy Roadmap lessons learned. DHS noted in its response that the decision to waive a procurement review is based on several considerations, such as the type of service, information provided to the Chief Procurement Officer by the Head of Contracting Activity, and historical and current knowledge of the procurement, among others. However, the department offered no further insights as to: what types of services may not warrant a Procurement Strategy Roadmap; what type of information provided by the Head of Contracting Activity may indicate a review is unnecessary; or how the Chief Procurement Officer maintains the historical knowledge of procurements that may have previously experienced challenges and thus warrant a Procurement Strategy Roadmap. We continue to believe that taking the step of documenting the factors considered—such as types of services that may require additional review, or challenges with prior procurements, some of which may have been awarded years prior—will help ensure that decisions to waive Procurement Strategy Roadmaps are made consistently and transparently. DHS did not agree with the fourth recommendation, that OCPO should update the Inherently Governmental and Critical Functions Analysis job aid to provide guidance for analyzing, documenting, and updating the federal workforce needed to perform or oversee service contracts requiring heightened management attention. In its response, DHS stated that the job aid requires components to certify that they have sufficient internal capacity to oversee and manage contractor activities and maintain control of its missions and operations when the requirement is a closely associated with inherently governmental or critical function. Further, DHS stated that the job aid requires components to certify that there are an adequate number of positions filled by federal employees to manage and monitor contractors if the requirement is a critical function. As noted in our report, each component is making its own determination, in the absence of guidance, as to what factors to consider. In its response, DHS stated that OCPO will assist components with examples of analysis by reviewing what some components are doing, and sharing those examples with others. However, in the absence of guidance about what DHS expects the components to analyze and document based on those examples, DHS does not know how or whether the components are considering the federal workforce available to oversee service contracts in need of heightened management attention, or what steps, if any, the components are taking to mitigate risks if there are not enough federal personnel available to oversee the contracts after award. In its response, DHS recognized the need to provide guidance for updating the job aid, if there is a change in the contract requirement, to help ensure it has sufficient internal capacity to oversee and manage contractor activities, maintains control of its missions and operations, and has the appropriate workforce in place. We consider this to be a positive step to address part of the recommendation; however, it is unclear what considerations the components will use to update their analysis without the presence of guidance for how to analyze the federal workforce needed prior to the contract being awarded. We maintain that without guidance, DHS is at risk of inconsistent consideration of federal oversight for service contracts across its components—an action at odds with its goals of improving integration, and centralizing and coordinating its many functions to ensure that its whole is greater than its parts. Finally, DHS did not agree with the sixth recommendation, to work with Congress to identify information to include in congressional budget justifications to provide greater transparency into requested and actual service requirement costs, particularly for services requiring heightened management attention. In its response, DHS stated that it does not believe including additional information on estimated or actual service contract requirement costs is appropriate, and stated that contract information can be found in congressional budget justifications in budget object class breakouts, cost drivers, and in the Procurement, Construction, and Improvement Appropriation Capital Investment exhibit. We acknowledge these same three sources of information in our report, and note the limitations with each (either over-estimating or under- estimating service contracts) to providing visibility into DHS’s estimated or actual service contract requirements—both internally to DHS and externally to Congress. For example, as we note in the report, contracts identified in the Capital Investment Exhibit are not categorized as being for a product or service nor does the exhibit consistently provide a description of what the contract is for. In its response DHS noted limitations with our analysis comparing contract obligation data from FPDS-NG with what is reported in DHS’s budget justifications, however, after discussion with DHS officials during our review, we did not include that comparison in our report. DHS also noted in its response that the congressional budget justifications are intended to focus on the request, not on the previous or current year’s contracts. However, the recommendation that DHS work with Congress is impartial as to what type of service contract information would be useful for providing greater transparency into DHS’s service contract requirements. Rather, the recommendation is intended to address the limited visibility both DHS and Congress have into DHS service requirements—in particular the significant increases in services requiring heightened management attention—and provide a means to report on that information to improve internal and external oversight over these requirements and to allow for more informed decision-making. The need for this visibility into service contract requirements is aligned with prior recommendations GAO has made related to the need to increase visibility in DHS’s congressional budget justifications for major acquisition programs’ funding requests; recommendations that DHS has agreed with and implemented. Given that service contracts accounted for over three quarters of DHS’s contract obligations from fiscal years 2013 through 2018, we continue to believe that our recommendation to work with Congress on how to convey that information in congressional budget requests is valid. We are sending copies of this report to the appropriate congressional committees, the Secretary of Homeland Security, the DHS Chief Procurement Officer, the Commissioner of U.S. Customs and Border Protection, the Director of Immigration and Customs Enforcement, and the Director of U.S. Citizenship and Immigration Services. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you or your staff have any questions about this report, please contact me at (202) 512-4841 or makm@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed in appendix VI. Appendix I: Objectives, Scope, and Methodology You asked us to review the Department of Homeland Security’s (DHS) use of and planning for service contracts. This report addresses the extent to which DHS and selected components and offices (1) used service contracts from fiscal years 2013 through 2018; (2) identified, developed, and reviewed service contract requirements; (3) ensured oversight of service contracts requiring heightened management attention; and (4) considered service requirements in budgeting processes. To identify the extent to which DHS used service contracts, we reviewed the Federal Procurement Data System-Next Generation (FPDS-NG) data on DHS-funded contract obligations from fiscal years 2013 through 2018 adjusted for inflation using the Gross Domestic Product Price Index. We identified obligations for services using the codes associated with services in the General Services Administration’s Federal Procurement Data System Product and Service Codes Manual. We analyzed the FPDS-NG data to identify DHS service obligations compared to obligations for products, service obligations by DHS component, the types of services procured, and the proportion of service contracts for functions in need of heightened management attention—those deemed closely associated with inherently governmental, critical, and special interest functions. We assessed the reliability of FPDS-NG data by reviewing existing information about the FPDS-NG system and the data it collects— specifically the data dictionary and data validation rules—and performing electronic testing. We determined the FPDS-NG data were sufficiently reliable for the purposes of identifying DHS’s use of service contracts. We selected a non-generalizable sample of four DHS contracting activities that had high obligations for service contracts and special interest functions compared to other DHS contracting activities—U.S. Customs and Border Protection (CBP), U.S. Immigration and Customs Enforcement (ICE), U.S Citizenship and Immigration Services (USCIS), and the Office of Procurement Operations (OPO). For the purposes of this report, we will refer to these contracting activities, which include three components and one office, as components. From these components, we selected a non-generalizable sample of 100 contracts awarded in fiscal year 2018 that were above the simplified acquisition threshold and were not exempt from performing a Balanced Workforce Assessment Tool (BWAT)—a risk analysis tool used by DHS components at that time to identify the appropriate mix of federal and contractor employees. Seventy five of the 100 contracts were for special interest functions, with the remaining 25 randomly selected. From that sample, we selected eight contracts—two from each component—that were identified as requiring heightened management attention. We selected a range of contracts based on whether the contract contained functions requiring heightened management attention, the percent of recommended federal oversight, and whether the requirement was new, among other selection criteria. We conducted semi-structured interviews with program, contracting, and budgeting officials from the eight selected component contracts to identify how the selected service contract requirements were developed, overseen, and considered when budgeting. Information collected from the four components and eight contracts cannot be generalized to all components and contracts. For additional details on the contracts we selected, see table 2. To determine how DHS and selected components identified, developed, and reviewed service contract requirements prior to soliciting for a contract, we reviewed relevant documentation, including the Federal Acquisition Regulation (FAR), and DHS, CBP, ICE, USCIS, and OPO contracting policies. To determine what processes selected components have for identifying and developing service requirements, we reviewed documentation, and interviewed program and contracting officials associated with our four selected components and eight selected contracts. To determine how DHS is reviewing service contract requirements, we reviewed DHS Office of the Chief Procurement Officer (OCPO) and Office of Program Accountability and Risk Management guidance and documentation on recent DHS headquarters initiatives—the Procurement Strategy Roadmap and Service Requirements Review pilot—and federal internal control standards on risk assessment. We also interviewed officials on these efforts to identify similarities and differences, and the processes established to review certain service contracts. To determine the extent to which DHS and the selected components in our review ensured federal oversight of service contracts requiring heightened management attention, we reviewed relevant documentation and regulations including Office of Federal Procurement Policy (OFPP) memorandums, the FAR, DHS contracting policies and guidance, and federal internal control standards on information and communication and risk assessment. To understand how DHS and selected components planned and documented oversight needs we reviewed available BWATs from our non-generalizable sample of contracts that we identified as special interest functions. Using FPDS-NG data, we identified an additional 27 contracts with completed job aids that were awarded in fiscal year 2019 for special interest functions across our selected components following implementation of the Inherently Governmental and Critical Functions Analysis job aid in March 2019. We reviewed the completed job aid associated with each of these contracts to understand how the oversight planning process has changed. We interviewed OCPO and component program and contracting officials about their use of both the BWAT and the Inherently Governmental and Critical Functions Analysis job aid. To determine the extent to which DHS and selected components conducted federal oversight of service contracts requiring heightened management attention throughout the life of a service contract, we analyzed documentation—such as contracting officer’s representative appointment letters depicting oversight responsibilities and training for contracting and program officials—and interviewed officials responsible for performing oversight functions. Additionally, to understand the types of tasks oversight officials can perform to mitigate the risk of contractors performing inherently governmental functions or losing control of the department’s mission, we reviewed OCPO provided guidance and trainings and interviewed relevant officials. To determine the extent to which DHS and selected components consider service contracts when budgeting, we reviewed Office of Management and Budget (OMB), DHS headquarters, and component budgeting guidance, federal internal control standards on information and communication, and interviewed headquarters and component budget officials. To determine how service contract requirements are communicated during resource planning and budget formulation, we reviewed DHS and component budget justification documents to identify what ways service requirement information is reflected, including whether specific information on our selected contracts was visible. We compared the resources DHS reported needing and receiving in its fiscal year 2018 budget documentation with DHS’s use of service contracts as reported in FPDS-NG in the same fiscal year as a proxy for visibility of service contract requirements in DHS budgeting. We conducted this performance audit from February 2019 to May 2020 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Appendix II: Department of Homeland Security Organizational Chart Appendix II: Department of Homeland Security Organizational Chart The Office of Procurement Operations is the contracting activity for the following DHS components and offices shown above: Countering Weapons of Mass Destruction; Cybersecurity and Infrastructure Security Agency; Civil Rights and Civil Liberties; General Counsel; Office of Intelligence and Analysis; Legislative Affairs; Office of Operations Coordination; Partnership and Engagement; Office of Strategy, Policy, and Plans; Public Affairs; Chief Information Officer; Chief Financial Officer; Secretary/ Deputy Secretary; Management Directorate; Privacy; and the Science and Technology Directorate. Appendix III: Examples of Certain Functions Requiring Heightened Management Attention In September 2011, the Office of Federal Procurement Policy (OFPP) in the Office of Management and Budget (OMB) issued a policy letter to help agencies manage the performance of inherently governmental and critical functions. The guidance states contracts whose performance may involve closely associated with inherently governmental, critical, or special interest functions require heightened management attention. Specifically, guidance states that closely associated with inherently governmental functions are functions that require heightened management attention to ensure that contractor’s activities do not expand into inherently governmental functions. OMB’s response to public comments on the proposed policy letter provides examples of inherently governmental and closely associated with inherently governmental functions, as shown in table 3. In response to public comments on the proposed policy letter, OMB called critical functions core to the agency’s mission or operations. In addition, the policy letter states that critical functions, when contracted for, pose the risk that the agency can lose control of its mission and operations. Examples of work previously identified by DHS as critical functions for the department include: Intelligence services—proprietary software used to conduct deep and dark web searches on possible threats against senior officials. Risk mitigation services—supporting undercover agents’ identities Program support—immigration data integration Administrative services—working closely with agency senior leadership to conduct research, schedule and attend meetings, as well as develop policies. The policy letter states that agencies must retain sufficient internal capability to give critical functions heightened management attention by: dedicating an adequate number of qualified federal personnel to understand the agency’s requirements and perform functions alongside contractors, if necessary, in the event the contractor fails to perform; or ensure qualified federal personnel are available to oversee and manage the contractor workforce. OMB guidance also describes special interest functions as requiring heightened management attention. In a 2010 memo, OFPP issued guidance to help agencies conduct a required service contract inventory for fiscal year 2010. The guidance describes the service contract inventory as a tool to better understand how contracted services are used and whether contractors’ skills are utilized in an appropriate manner. According to the guidance, agencies should give priority consideration to special interest functions, which for fiscal year 2010 OFPP identified as the categories of professional management services and information technology support services. Special interest functions require increased management attention due to increased risk of workforce imbalance. DHS, in line with OMB guidance, has identified 17 product and service codes to categorize as special interest functions, as shown in table 4. Appendix IV: Key Differences in Tools for Evaluating Functions Requiring Heightened Management Attention The Department of Homeland Security (DHS) implemented its Balanced Workforce Strategy (BWS) in October 2009 to establish a set of processes that, when repeated on a regular basis, enables the department to achieve the appropriate mix of federal employees and contractors to accomplish the department’s mission while minimizing mission risk that may result from an overreliance on contractors. To accomplish the intended goals of the BWS, DHS instituted an online questionnaire called the Balanced Workforce Assessment Tool (BWAT). The BWAT was in place until March 2019 when DHS determined it— along with the strategy—were no longer necessary given the maturation of their acquisition workforce and their inability to support the software underlying the BWAT. In its place, DHS commissioned its Inherently Governmental and Critical Functions Analysis—known as the job aid. Appendix V: Comments from the Department of Homeland Security Appendix VI: GAO Contact and Staff Acknowledgments GAO Contact Staff Acknowledgments Marie A. Mak 202-512-4841 or MakM@gao.gov In addition to the contact named above, Penny Berrier (Assistant Director), Meghan Perez (Analyst in Charge), Erin Butkowski, Signe Janoska-Bedi, and Jacqueline Wade were principal contributors. In addition, the following people made contributions to this report: Pete Anderson, Lorraine Ettaro, Suellen Foth, Julia Kennon, Roxanna Sun, and Anne Louise Taylor.

Why GAO Did This Study DHS's spending on services—such as guard services and technology support—represents over 75 percent of its annual contract obligations. The Office of Management and Budget has recognized that some service contracts require extra management attention because they pose a risk that the government could lose control of its decisions or operations. GAO was asked to review DHS's use of and planning for service contracts. This report addresses, among other objectives, the extent to which DHS and selected components and offices use, oversee, and budget for service contracts. GAO analyzed Federal Procurement Data System-Next Generation data from fiscal years 2013 through 2018; selected non-generalizable samples of four components with high service contract obligations and eight service contracts requiring heightened management attention; and interviewed DHS officials. What GAO Found From fiscal years 2013 through 2018, the Department of Homeland Security (DHS) increased its reliance on contracts for services, particularly those in categories that may need heightened management attention, such as drafting policy documents (see figure). These services include functions that are closely associated with inherently governmental, critical, or special interest, which could put the government at risk of losing control of its mission if performed by contractors without proper oversight by government officials. GAO found that DHS and selected components do not consistently plan for the level of federal oversight needed for these contracts because there is no guidance on how to document and update the number of federal personnel needed to conduct oversight. GAO also found that program and contracting officials from six of the eight contracts GAO reviewed did not identify specific oversight activities they conducted to mitigate the risk of contractors performing functions in a way that could become inherently governmental. DHS lacks guidance on what these oversight tasks could entail. Without guidance for documenting and updating the planned federal oversight personnel needed, and identifying oversight tasks, DHS cannot mitigate the risks associated with service contracts in need of heightened management attention. Selected DHS components have information on service requirements, but budget documentation—submitted to DHS headquarters as well as to Congress—does not communicate details about most estimated or actual service contract requirements costs. Given that services account for over three-quarters of DHS's annual funding for contracts, additional insights would shed light into how much of DHS's mission is being accomplished through services, including those requiring heightened management attention. Without more visibility into this information, DHS headquarters and Congress are at risk of not having complete information for sound resource planning and decision-making, particularly as it relates to determining what proposed service contract requirements DHS should prioritize when budgeting. What GAO Recommends GAO is making six recommendations, including that DHS provide guidance for documenting and updating the federal workforce needed to oversee certain service contracts and identifying oversight tasks, and report service requirement information in budget documents to Congress. DHS agreed with two of the recommendations and did not agree with four of them. GAO continues to believe the recommendations are valid, as discussed in the report.

gao_GAO-20-54

gao_GAO-20-54_0

Background Hurricanes Irma and Maria In September 2017, Hurricane Irma struck the islands of St. Thomas and St. John, and two weeks later, Hurricane Maria struck the island of St. Croix in the USVI, causing catastrophic damage across the entire territory and neighboring Caribbean islands (see fig. 2). The storms severely damaged the USVI’s critical infrastructure, leaving many of the territory’s 107,000 residents without electricity, phone service, food, or running water. According to a September 2018 report from the USVI Hurricane Recovery and Resilience Task Force (USVI Task Force Report), the hurricanes devastated the territory’s electricity grid and telecommunications systems, shutting down both for months. Further, the storms damaged more than half of the territory’s housing units, as well as its hospitals, government buildings, schools, water and wastewater facilities, and more (see figs. 3 and 4). Overall, the report estimated that the hurricanes caused approximately $10.7 billion in total damages across the USVI. In response to the request of the Governor of the USVI, the President declared a major disaster the day after each hurricane struck the territory. Major disaster declarations can trigger a variety of federal response and recovery programs for government and nongovernmental entities and households and individuals, including assistance through the Public Assistance program and Hazard Mitigation Grant Program. Under the National Response Framework and National Disaster Recovery Framework, DHS is the federal department with primary responsibility for coordinating disaster response and recovery, and within DHS, FEMA has lead responsibility. The Administrator of FEMA serves as the principal adviser to the President and the Secretary of Homeland Security regarding emergency management. FEMA’s, States’, and Territories’ Roles and Responsibilities for Disaster Recovery Once the President has declared a major disaster, FEMA, the state or territorial government (the recipient), and local or territorial entities (the subrecipient) work together to, among other things, identify and develop projects through the Public Assistance program and Hazard Mitigation Grant Program. After a project has completed FEMA’s review process and is approved, FEMA obligates funding for the project by placing money into an account where the recipient has the authority to draw down—or expend—funding to pay for eligible work upon completion. Further, when a project has been completed, FEMA conducts a close-out process to certify that all eligible work has been completed and reconciles the actual cost incurred. If the actual cost of the completed work is less than the amount of money FEMA obligated for the project, FEMA will deobligate funding. However, if the actual cost of the completed work is greater than the amount of money FEMA obligated for the project, FEMA may reimburse the subrecipient for these additional costs. A state or territorial governor may designate a governor’s authorized representative to oversee all aspects of disaster assistance—including Public Assistance program and Hazard Mitigation Grant Program funding—to ensure the USVI’s compliance with federal regulations and FEMA requirements. Among other responsibilities, the governor’s authorized representative is to confirm that subrecipients submit complete documentation demonstrating that all eligible work completed is in accordance with program requirements. FEMA’s Public Assistance Program and Hazard Mitigation Grant Program FEMA’s Public Assistance program provides grant funding to state, territorial, local, and tribal governments, as well as certain types of private nonprofit organizations, to assist with the repair or replacement of disaster-damaged public infrastructure. To develop projects under the Public Assistance program, FEMA and USVI officials collaborate to identify and document the damage caused by a disaster to a particular system or facility. These officials then use this damage assessment to formulate the scope of work—or activities required to fix the identified damage—as well as the estimated cost of these activities. As shown in figure 5, Public Assistance grant funds are organized broadly as “emergency work” or “permanent work.” Within these areas are separate categories of work. In addition to emergency work and permanent work, the program includes category Z, which represents indirect costs, direct administrative costs, and any other administrative expenses associated with a specific project. Under the Public Assistance program’s permanent work categories, FEMA also provides grant funding for cost-effective hazard mitigation measures to reduce or eliminate the long-term risk to people and property from future natural and man-made disasters and their effects. FEMA provides this funding in conjunction with the repair of disaster-damaged facilities to enhance their resilience during future disasters. For example, this funding could be used to replace damaged wooden utility poles with composite fiberglass ones to increase the resilience of an electricity distribution system and mitigate the potential for future damage from hurricane-force winds. FEMA’s Hazard Mitigation Grant Program provides grant funding for long- term mitigation solutions to reduce the risk of loss of life and property from future disasters. Unlike mitigation measures funded through the Public Assistance program to further protect disaster-damaged infrastructure, the Hazard Mitigation Grant Program may fund measures for systems, facilities, or properties that were not damaged in the disaster. For example, program funding can be used to construct floodwater control measures—such as berms and rock linings—that did not exist prior to the disaster, or to update existing hazard mitigation plans to accurately reflect current mitigation goals. The Public Assistance Alternative Procedures Pilot Program in the USVI In July 2018, FEMA approved the use of the Public Assistance alternative procedures pilot program for permanent work projects in the territory. Unlike the standard Public Assistance program wherein FEMA will fund the actual cost of a project, the alternative procedures require awards for permanent work projects to be made on the basis of fixed-cost estimates. As a result, the recipient or subrecipient is ultimately responsible for any project costs that exceed the agreed-upon fixed-cost estimate at the time of the close-out process. However, the alternative procedures program also provides the USVI with financial incentives for the timely and cost-effective completion of work and additional flexibilities that are not available through the standard Public Assistance program. For example, the USVI may use excess grant funding for cost-effective hazard mitigation measures and, in certain circumstances, consolidate permanent work projects approved under the alternative procedures and share obligated funding across these projects. Further, section 20601 of the Bipartisan Budget Act of 2018 authorizes FEMA, when using the alternative procedures, to provide assistance to fund the replacement or restoration of disaster-damaged infrastructure that provide critical services without regard to pre-disaster condition (see fig. 6). For example, through the Act, FEMA may fund the restoration of a disaster-damaged school building—which provides a critical service—to accepted industry standards applicable to the construction of education facilities. Therefore, according to FEMA policy, if components of the school building were not up to industry standards or in poor condition prior to the 2017 hurricanes, the Act allows FEMA to fund the restoration of this building to a better condition than it was in prior to the storms. The Sheltering and Temporary Essential Power Pilot Program The Sheltering and Temporary Essential Power (STEP) pilot program is an emergency sheltering program implemented under FEMA’s emergency work authority and funded through the Public Assistance program’s category B emergency work. The program—which was created following Hurricane Sandy in 2012—allows FEMA to fund emergency, temporary repairs to make damaged homes habitable by, for example, restoring electricity to a private home and applying temporary patches to roofs and windows to protect the interior. In funding these types of repairs, FEMA’s goal is to quickly make damaged homes habitable in the short term until the homeowner could complete more permanent repairs independently through other FEMA programs or private insurance payments. Since 2012, FEMA has implemented the program as a tool in addressing the unique circumstances and challenges associated with providing safe sheltering options for disaster survivors. FEMA Obligated More Than $1.9 Billion through Two USVI Recovery Programs as of June 30, 2019, but Faced Some Challenges in Implementing Recovery Projects Status of FEMA Public Assistance Program and Hazard Mitigation Grant Program Funding in the USVI As of June 30, 2019, FEMA obligated more than $1.9 billion in grant funding for 640 projects through the (1) Public Assistance program and (2) Hazard Mitigation Grant Program in the USVI. First, FEMA obligated more than $1.8 billion in Public Assistance grant funding for 618 projects across the USVI (see fig. 7). Specifically, FEMA obligated more than $1.1 billion for emergency work projects (categories A and B), about $588.5 million for permanent work projects (categories C through G), and about $141.2 million for management costs (category Z). Of the approximately $1.8 billion FEMA obligated in Public Assistance grant funding as of June 30, 2019, the USVI had expended nearly $1.1 billion (59 percent) to reimburse subrecipients for completed work. Appendix I provides more detailed information on the status of Public Assistance grant funding in the USVI. Second, FEMA obligated about $60.6 million for 22 Hazard Mitigation Grant Program projects in the territory as of June 30, 2019. These projects are designed to fund mitigation measures to increase the longer- term resilience of the USVI’s infrastructure during future disasters. Of the $60.6 million FEMA obligated as of June 30, 2019, the USVI expended about $1.7 million (3 percent) across 5 projects. Appendix II provides more detailed information on the status of Hazard Mitigation Grant Program funding in the USVI. While these data represent the status of grant funding as of June 30, 2019, the amount of FEMA obligations and USVI expenditures for both programs will likely increase over time as additional projects are finalized and approved. Limited Availability of Local USVI Personnel and Territory’s Difficult Fiscal Situation Presented Challenges to Program Implementation FEMA, USVI officials, and contractor personnel identified challenges across three areas that affected the implementation of the Public Assistance program and the Hazard Mitigation Grant Program in the USVI. Specifically, they cited: (1) the limited availability of local staff in the USVI to implement and oversee recovery programs, (2) the inability of local construction crews to undertake the large number of recovery projects, and (3) the impact of the USVI’s difficult fiscal situation on recovery efforts. Limited availability of local staff. USVI and FEMA officials cited the limited number of local USVI personnel with the knowledge and expertise necessary to staff recovery-related positions in key USVI agencies as a significant challenge following the 2017 hurricanes. For example, USVI officials told us that the Virgin Islands Territorial Emergency Management Agency—the agency initially responsible for overseeing all aspects of both the Public Assistance program and Hazard Mitigation Grant Program in the territory—did not have enough employees on staff to effectively implement and manage these programs. Further, a senior FEMA official noted that after the storms, the USVI had only one individual responsible for managing all aspects of the Hazard Mitigation Grant Program across the territory. In addition, the limited availability of local staff in the USVI was exacerbated by the departure of qualified individuals following the hurricanes as well as competition among recovery agencies to hire qualified staff that remained in the territory, according to USVI officials. To address these challenges, the USVI hired two contractors to augment the territory’s capacity in the shorter term and established a new Office of Disaster Recovery to oversee recovery efforts in the longer term. First, in December 2017, the USVI hired two contractors to assist the territory in planning, developing, implementing, and overseeing recovery projects, among other responsibilities. Second, in February 2019, the USVI established the Office of Disaster Recovery as the primary territorial agency responsible for coordinating and overseeing all disaster recovery efforts in the USVI, including the Public Assistance program and Hazard Mitigation Grant Program. The office’s Director told us that while contractor personnel had been valuable in augmenting the USVI’s capacity, the territory was prioritizing the hiring and training of qualified local hires to replace these contractors for the longer term. Shortage of local construction crews. Due to the territory’s relatively small population, FEMA and USVI officials stated there were not enough local construction crews to address the large amount of construction work required to repair and rebuild damaged infrastructure following the 2017 hurricanes. These officials told us this construction crew shortage affected the USVI’s ability to keep Public Assistance program and Hazard Mitigation Grant Program projects proceeding on time. FEMA and contractor personnel stated that unlike in the contiguous United States, the USVI does not have neighboring states that can easily send construction crews to affected areas to augment local crews. In addition, historically, the USVI relied on Puerto Rico to supplement the territory’s capacity, but this was not an option as Puerto Rico was undergoing its own massive recovery effort as a result of Hurricane Maria. The USVI’s fiscal situation. USVI officials and contractor personnel stated that the challenging fiscal situation in the territory directly affected its ability to effectively implement recovery programs. Specifically, USVI officials told us that the territory’s financial condition made it difficult to provide initial funding to reimburse subrecipients for completed work prior to drawing down funding from the account holding FEMA-obligated money. These officials stated this process was problematic because instead of funding all eligible projects as quickly as possible to move the recovery forward, the USVI had to prioritize certain recovery projects over others based on the availability of funding. Further, USVI contractor personnel told us that the territory often does not have the cash on hand necessary to provide these reimbursements to subrecipients, which can result in delays in paying subrecipients and contractors. According to USVI officials, pursuing projects under the Public Assistance alternative procedures program may help to address these issues by providing more flexibility regarding when and how projects are funded. Challenges Implementing the Public Assistance Alternative Procedures Program Have Delayed Recovery Projects in the USVI The Public Assistance alternative procedures program provides the USVI with financial incentives and new flexibilities in implementing recovery projects that are not available through the standard Public Assistance program. However, FEMA and USVI officials stated that implementing the alternative procedures program in the USVI presented challenges that affected recovery efforts and delayed the obligation of funding for permanent work projects. Specifically, they cited challenges in (1) developing accurate fixed-cost estimates for program projects and (2) implementing the new flexibilities authorized by section 20601 of the Bipartisan Budget Act of 2018. Senior USVI officials told us that due to these challenges and the financial risk associated with the use of fixed- cost estimates, the USVI is planning to take a cautious approach in pursuing alternative procedures projects. As established in FEMA guidance, USVI officials have a deadline of March 2020 to finalize the fixed-cost estimates for such projects for inclusion in the alternative procedures program. Fixed-cost estimates. As the USVI is financially responsible for any actual costs that exceed the fixed-cost estimate for any given alternative procedures project, ensuring the accuracy of these estimates is critical due to the USVI’s already difficult fiscal situation. However, USVI officials told us that developing fixed-cost estimates that accurately forecast the future costs of completing large, complex permanent work projects in the remote island territory is difficult given the unique circumstances that influence construction costs in the USVI, such as the limited availability of local resources and the need to import construction materials and labor. To address this challenge, in October 2018, FEMA asked an independent contractor to analyze whether a USVI-specific “cost factor” should be incorporated into FEMA’s process for developing fixed-cost estimates to ensure the actual costs of implementing permanent work projects in the territory were captured. According to FEMA officials, the independent contractor determined that a cost factor was appropriate for use in the USVI and the contractor proposed several options. However, territorial officials contended that these proposals did not sufficiently or accurately capture the unique circumstances that influence construction costs in the territory. Further, USVI officials stated that ensuring the accuracy of the cost factor was critical given the significant financial risk using fixed-cost estimates posed to the USVI. Since incorporating a cost factor into the process for developing fixed- cost estimates increases the base cost for any given permanent work project—and therefore the amount of funding FEMA obligates—FEMA officials explained the USVI had an incentive to delay the obligation of projects until FEMA finalized this factor. As a result, FEMA officials told us in May 2019 that obligations for permanent work projects had been mostly on hold since October 2018 while the contractor worked to develop the USVI-specific cost factor. As the USVI is reliant on federal recovery funding to reimburse subrecipients for completed work, this delay in obligations directly affected the USVI’s ability to move recovery projects forward. In May 2019, the contractor proposed a new cost factor, which FEMA approved on an interim basis pending further analysis. In July 2019, FEMA officials told us that while additional analyses are required to ensure its final process for developing fixed-cost estimates in the USVI accurately captures construction costs, using this interim cost factor in the meantime allows FEMA and USVI officials to move forward with the development and final approval of alternative procedures projects. In August 2019, a senior USVI official told us the territory plans to begin using the interim cost factor, where appropriate, to keep projects progressing forward. However, she stated that the USVI questioned whether the interim cost factor did, in fact, sufficiently capture the actual costs of construction in the USVI. Given the uncertainty around these fixed-cost estimates, USVI officials told us the territory will need to balance the potential flexibilities provided by the alternative procedures program with the financial risk posed by cost overruns when deciding whether to use the alternative procedures or the standard Public Assistance program for any given permanent work project. We are currently assessing FEMA’s process for developing cost estimates for projects under both the standard and alternative procedures programs, and plan to report our results in early 2020. The Bipartisan Budget Act of 2018. While FEMA and USVI officials told us that section 20601 of the Bipartisan Budget Act presented a valuable opportunity to advance the USVI’s recovery, they also reported challenges with implementing the new flexibilities authorized by the Act, which made developing eligible permanent work projects difficult. For example, USVI officials stated that, at times, they were unclear about the implementation process for key components of the Act and thus ensuring subrecipients understood the process was difficult. Further, FEMA officials in the USVI told us that initially, they had difficulty obtaining clarification from FEMA headquarters regarding how to implement key provisions of the Act, such as the process for identifying and incorporating relevant industry standards for specific alternative procedures projects. As a result, permanent work projects that were eligible to use the flexibilities provided by the Act remained on hold until FEMA could clarify the process for implementing the Act and pertinent industry standards could be approved. In addition, the Bipartisan Budget Act was signed into law in February 2018 and applies exclusively to federal disaster assistance to the USVI and Puerto Rico. As a result, FEMA officials faced the challenge of interpreting the Act’s language and appropriately implementing its provisions for the first time. For example, the Act allows for a new process for determining whether a disaster-damaged facility is eligible to receive funding to (1) repair the existing facility or (2) replace the facility with a new structure. Under the standard Public Assistance program, this determination is calculated using the “50 percent rule”—if the cost of repairing the disaster-related damage sustained by the facility exceeds 50 percent of the cost of replacing it, FEMA may fund the replacement of the facility. In contrast, the Act does not provide a similar cost estimating process for use in developing fixed-cost estimates through the alternative procedures program. In September 2018, FEMA issued guidance for implementing section 20601 of the Bipartisan Budget Act through the Public Assistance alternative procedures program, which provides that critical services infrastructure—such as medical and educational facilities—is eligible for replacement “if repair is feasible, but replacement is more prudent.” FEMA officials in the USVI told us that since the agency did not have further guidance or criteria on the appropriate process for evaluating repair or replacement under this new standard, they were responsible for developing the agency’s first justification to support the replacement of a hospital in St. Croix based on their interpretation of the new standard. These officials also stated that since their rationale justifying the facility’s eligibility for replacement was the first of its kind and would set a precedent for future projects, they submitted it to FEMA headquarters for review. In May 2019, FEMA officially approved the replacement of this hospital through the alternative procedures program. For more information on how the Public Assistance program and the Bipartisan Budget Act are affecting recovery efforts at this facility, see appendix III. The Additional Supplemental Appropriations for Disaster Relief Act of 2019, which was signed into law in June 2019, provides additional direction to FEMA regarding the implementation of section 20601 of the Act. Among other things, this legislation includes a provision directing FEMA to change its process for determining whether a disaster-damaged facility is eligible for repair or replacement. FEMA evaluated this and other provisions of the Act and, in September 2019, issued an updated policy to provide clear guidance moving forward, according to agency officials. The USVI Governor and senior territorial officials stated that due to the challenges outlined above and the financial risk posed by exceeding fixed-cost estimates, the USVI plans to take a cautious approach in implementing the Public Assistance alternative procedures program. Specifically, the Governor told us the territory will most likely pursue alternative procedures projects that are simple, have clear scopes of work, and do not include high levels of uncertainty to reduce the financial risk of potential cost overruns. USVI officials added that if they are not comfortable with the fixed-cost estimate for any given alternative procedures project, the territory has the option to pursue the project under the standard Public Assistance program. Under the standard program, the USVI cannot take advantage of the flexibilities and financial incentives provided by the alternative procedures and the Bipartisan Budget Act, but FEMA would reimburse the USVI for the actual cost—including any cost overruns—of all work completed in accordance with a project’s approved scope of work, thereby mitigating the territory’s financial risk. The USVI is ultimately responsible for deciding whether the benefits provided through the alternative procedures program and the Bipartisan Budget Act outweigh the financial risk associated with agreeing to fixed-cost estimates for permanent work projects. Since the territory has until March 2020 to finalize these fixed-cost estimates, it remains too early to determine the extent to which the alternative procedures program will play a role in the USVI’s long-term recovery strategy. FEMA Discontinued the STEP Pilot Program After Expanding It in the USVI, but Has Not Evaluated Options for Providing Future Emergency Sheltering Assistance FEMA Expanded the STEP Pilot Program in the USVI In October 2017, FEMA authorized the STEP pilot program in the USVI in response to the widespread damage to homes that displaced residents and overwhelmed sheltering and temporary housing resources in the territory. Through the program, FEMA funded minimal, temporary protective repairs (or “Phase I” repairs) to private homes to allow residents a safe place to shelter. For example, Phase I emergency repairs included applying temporary patches to roofs and windows to protect the interior from outside weather conditions and ensuring a functional kitchen and bathroom and safe sleeping area. According to FEMA documentation, the intent of these minimal temporary repairs was to quickly make damaged homes habitable in the short term until homeowners could complete more permanent repairs independently through other FEMA programs or using private insurance payments. In August 2018, FEMA expanded the STEP pilot program to include the “permanent” repair or replacement of damaged roofs (or “Phase II” work)—the first time in its history that FEMA authorized such work through this pilot program. Phase II work funded more permanent work on USVI residents’ damaged roofs—either by repairing damages to the existing roof or replacing it with a new one. In addition, Phase II work included incorporating roof hardening measures, such as installing hurricane clips to the roof berms, to increase the resiliency of the roofs against hurricane-force winds. Figure 8 provides two examples of USVI homes that participated in Phase II of the STEP pilot program. FEMA expanded the STEP pilot program to address the USVI’s unique, longer-term sheltering needs. Specifically, as the 2018 hurricane season arrived, FEMA was faced with the challenge of ensuring adequate sheltering options were available to USVI residents in the event that another hurricane struck the territory. The following factors contributed to FEMA’s decision to expand the STEP pilot program: Infeasibility of other sheltering programs. Alternate sheltering options were not viable in the USVI due to the unique circumstances in the territory. For example, the Transitional Sheltering Assistance program—where FEMA funds non-congregate sheltering (typically in hotels or motels) for displaced residents who cannot safely return to their homes—was not a feasible option as there was only one operating hotel in the USVI capable of sheltering disaster survivors. Further, FEMA officials told us that temporary housing units—such as manufactured housing units or recreational vehicles—could not be deployed to supplement the territory’s available housing stock due to logistical challenges, including the prohibitive costs of shipping these units to the remote territory and the limited availability of space to install them. Operation Blue Roof caused additional damage to homes. According to FEMA officials, FEMA’s decision to allow homes that had received temporary blue tarps as an emergency roofing measure through Operation Blue Roof to be eligible for the STEP pilot program led to expanding the scope of allowable work funded by the program. FEMA and USVI officials told us this change was implemented to address several issues with the blue tarps installed on homes, including the temporary nature of the tarps—which had a post-installation lifespan of only 30 days—and the need to fix the damage caused by installing the blue tarps on undamaged sections of roofs. FEMA officials stated that expanding the STEP pilot program to conduct more permanent roof repairs on these homes helped to ensure homeowners were able to safely shelter in the event of another hurricane. Shortage of construction crews. As previously discussed, FEMA and USVI officials cited the limited number of construction crews available to implement recovery work as a challenge, including for the STEP pilot program. Specifically, this challenge made it difficult for private homeowners to independently hire qualified contractors to conduct permanent repairs to their homes, according to FEMA officials. Therefore, these officials explained that using Phase II of the STEP pilot program to manage contractors in an official capacity made it more likely that necessary permanent repairs would be completed in a timely manner. Evacuation was not an option. When requested, FEMA is responsible for providing safe sheltering options following a disaster and, in the absence of feasible local sheltering options, FEMA is responsible for evacuating residents to a safe location outside the potentially affected area. However, according to FEMA documentation, developing and executing a plan to evacuate the USVI’s more than 100,000 residents in the event of another hurricane was impractical. Given these factors and the risk of another hurricane, FEMA officials determined that authorizing Phase II roof repairs or replacements of a permanent nature represented an appropriate solution to ensure eligible program participants could safely shelter in their homes. The STEP pilot program in the USVI officially ended on April 15, 2019. FEMA reported that 7,381 homes ultimately received repairs through the program. Specifically, 6,372 homes received Phase I temporary repairs and 1,631 homes received Phase II roof repairs or replacements of a permanent nature. In addition, 622 homes received both Phase I and Phase II repairs. According to FEMA officials, the agency is now conducting the close-out process for the STEP pilot program in the USVI, which includes reviewing the paperwork for each participating home to ensure all work was completed in accordance with both the home’s approved scope of work and overall programmatic requirements. FEMA Has Decided Not to Use the STEP Pilot Program During Future Recovery Efforts, but Has Not Evaluated Its Options for Providing Similar Emergency Sheltering Assistance In May 2019, FEMA’s Chief Counsel stated that FEMA had decided to discontinue the STEP pilot program due to significant challenges and lessons learned from prior experiences implementing the program. Specifically, FEMA stated that while FEMA had implemented the STEP pilot program within its authority pursuant to Section 403 of the Stafford Act, the agency was no longer “comfortable from a legal, policy, or pragmatic perspective” with implementing the STEP pilot program following future disasters. FEMA cited two main challenges in implementing the program in the USVI and elsewhere: (1) limiting the program’s scope to provide only minimal, emergency repairs, as intended, and (2) completing these emergency repairs in a timely manner. First, FEMA stated that in multiple iterations of the STEP pilot program—including in the USVI—FEMA officials had “succumbed to the pressure” from state and territorial leaders to expand the scope of allowable repairs under the program to conduct more extensive repairs. For example, in the USVI specifically, although expanding the program to authorize permanent roof repairs was legally supportable and represented an earnest effort to meet the territory’s needs, the expansion did “push the boundaries of appropriateness” and increased FEMA’s risk of interfering with the agency’s authority to provide assistance through other FEMA programs. Second, FEMA stated that the lengthy process for delivering the STEP pilot program—including in the USVI—undercut the program’s stated intent of providing emergency sheltering within 3 to 4 months following a disaster. For example, while FEMA authorized the program in the USVI in October 2017, initial repairs did not begin until March 2018 and eligible work was not completed until April 2019—18 months after the program’s authorization. According to FEMA, completing STEP pilot program repairs took longer than intended across most instances of the program’s implementation due to the amount of time required to develop disaster- specific program guidance, hire a large number of construction crews to undertake the repair work, obtain the necessary permissions from homeowners, and ultimately complete the repairs. Given the STEP pilot program’s protracted period of implementation, FEMA stated the agency had not been successful in ensuring that program repairs provided disaster survivors with emergency shelter in a timely manner. FEMA’s decision to discontinue the STEP pilot program following future disasters raises questions about how the agency plans to address the emergency sheltering needs of disaster survivors in the future—especially in communities that face challenges and circumstances similar to those the program was specifically designed to address. Since implementing it in 2012, FEMA used the STEP pilot program to supplement other FEMA sheltering programs and provide necessary additional capacity to help address the emergency sheltering needs of disaster-affected communities, as described below. In certain cases, the program provided assistance to more disaster survivors than other relevant FEMA programs, including the Transitional Sheltering Assistance program and the provision of temporary housing units. FEMA implemented the STEP pilot program in the following locations: Louisiana: FEMA authorized the STEP pilot program to supplement other federal programs implemented in Louisiana following severe storms and flooding in 2016. Specifically, the Transitional Sheltering Assistance program was not a viable option for most survivors, partially due to the limited availability of hotels and motels in the affected area, and FEMA ultimately used this program for approximately 4,300 households. In addition, FEMA deployed temporary housing units for approximately 4,600 households. FEMA also funded repairs through the STEP pilot program for nearly 11,000 homes. Texas: FEMA authorized the STEP pilot program to supplement other federal programs in Texas following Hurricane Harvey in 2017. Specifically, FEMA determined that implementing the STEP pilot program provided a useful option since the number of displaced survivors significantly exceeded the available capacity for sheltering survivors in local hotels and motels. FEMA used the Transitional Sheltering Assistance program for approximately 55,000 households and deployed temporary housing units for more than 3,500 households. FEMA supplemented these programs by funding repairs through the STEP pilot program for approximately 15,700 homes. Puerto Rico: FEMA authorized the STEP pilot program to address the unique emergency sheltering needs in Puerto Rico following Hurricanes Irma and Maria in 2017. Specifically, FEMA determined that approximately 80 percent of the island did not have power and would not have it restored for an extended period of time. As a result, FEMA implemented the STEP pilot program to, among other repairs, reconnect homes to a functioning electricity grid or, as necessary, fund the installation of generators if the grid could not be restored in a timely manner. Further, similar to what occurred in the USVI, FEMA authorized the STEP pilot program to repair homes that participated in Operation Blue Roof to address the roof damage caused by the installation of the blue tarps. FEMA funded repairs through the STEP pilot program for nearly 108,500 homes in Puerto Rico. In addition, due to the lack of available hotels and motels in Puerto Rico, participation in FEMA’s Transitional Sheltering Assistance program was limited to approximately 7,000 households, most of which were relocated to hotels and motels in the contiguous United States. Further, as detailed below, FEMA implemented its new Voluntary Agencies Leading and Organizing Repair program for the first time in Puerto Rico to conduct repairs to approximately 4,600 homes. North Carolina: FEMA authorized the STEP pilot program to address the particular emergency sheltering needs in North Carolina following Hurricane Florence in 2018. FEMA amended the STEP pilot program to allow both contracted construction crews and voluntary organizations to conduct the repairs, and ultimately funded repairs through the program for approximately 2,200 homes. In addition, FEMA used the Transitional Sheltering Assistance program for more than 870 households and deployed temporary housing units for approximately 650 households. Overall, FEMA authorized the STEP pilot program following 8 declared disasters since 2012 and obligated approximately $2.6 billion to fund repairs to more than 167,000 disaster survivors’ homes, according to FEMA documentation. While the program may not have provided repairs as rapidly as FEMA intended, these repairs nonetheless played a significant role in ensuring these disaster survivors could safely shelter in their homes. Since discontinuing the program, FEMA has not evaluated its options for addressing the emergency sheltering needs of disaster survivors. FEMA stated that it continues to support the use of congregate sheltering and will consider authorizing the Transitional Sheltering Assistance program, among other options, to address disaster survivors’ needs following future disasters. However, as detailed above, FEMA used the STEP pilot program for the specific purpose of providing necessary additional capacity to supplement these and other federal programs. Further, in certain cases, the STEP pilot program was used when implementing these other programs was unfeasible, such as in the USVI and New York where the particular circumstances on the ground made using the Transitional Sheltering Assistance program or deploying temporary housing units impractical. FEMA officials also told us the agency could utilize voluntary organizations to a greater extent than in the past to conduct the same types of repairs provided through the STEP pilot program. Specifically, FEMA officials stated that the Voluntary Agencies Leading and Organizing Repair program—which was implemented in Puerto Rico in response to the 2017 hurricanes and used to repair about 4,600 homes— could be used for this purpose following future disasters. However, given the program’s limited implementation, it is too early to determine the extent to which it represents a feasible solution for addressing emergency sheltering needs, or is capable of providing assistance on as large a scale as other federal programs, such as the STEP pilot program that funded repairs on more than 100,000 homes in Puerto Rico alone. Further, in North Carolina—where FEMA amended the STEP pilot program to allow both hired construction crews and voluntary organizations to conduct the repairs—the number of homes repaired by the construction crews—about 2,000—far exceeded the number of homes repaired by voluntary organizations—about 150—which raises questions about the ability of voluntary organizations to undertake the large volume of repairs necessary following disasters. Standards for Internal Control in the Federal Government states that management should identify, on a timely basis, significant changes to internal conditions that have already occurred, including changes to the entity’s programs or activities. Further, management should identify, analyze, and respond to risks related to achieving the entity’s defined objectives. FEMA has not assessed how its decision to discontinue the STEP pilot program will affect its ability to provide emergency sheltering assistance following future disasters. While FEMA officials told us they plan to use other sheltering programs when the next disaster strikes, these programs may not be sufficient in addressing the emergency sheltering needs of disaster survivors, especially in communities where implementing such programs is not feasible. FEMA officials also stated that given the agency’s decision to discontinue the STEP pilot program, conducting a broad evaluation of FEMA’s emergency sheltering programs and the agency’s options for addressing emergency sheltering needs would be useful to ensure that FEMA is prepared to respond effectively to future disasters. Conducting such an evaluation would help FEMA understand its ability to provide sheltering options and to properly plan for the provision of effective emergency sheltering assistance to disaster- affected communities. The USVI and FEMA Established Structures for Overseeing Recovery Efforts, but FEMA Has Not Consolidated Hazard Mitigation Grant Program Monitoring Guidance The USVI and FEMA Have Structures in Place for Overseeing Recovery Activities USVI. As the recipient of federal disaster funding, the USVI is responsible for providing oversight over the Public Assistance program and Hazard Mitigation Grant Program to ensure they are implemented in compliance with applicable laws and regulations, as well as FEMA policies and guidance. Following the 2017 hurricanes, the USVI took steps to address its responsibilities for receiving grant funding through these programs, including by: (1) developing administrative plans, (2) designating two territorial entities to manage the administration of disaster recovery funding, and (3) establishing the new Office of Disaster Recovery. First, as required by FEMA, the USVI developed administrative plans for the Public Assistance program and Hazard Mitigation Grant Program to ensure that subrecipients are in compliance with the conditions of these grant programs. These plans outline programmatic and project monitoring activities as well as the financial and administrative procedures for both programs. The plans require, among other things, the USVI to submit quarterly progress and financial reports on the status of projects to FEMA and describe the USVI’s specific roles and responsibilities for implementing and overseeing these two programs. Second, the Governor of the USVI designated two territorial entities to manage and oversee the implementation of recovery programs—the Virgin Islands Territorial Emergency Management Agency as the programmatic manager and the governor’s authorized representative as the grant administrator for all federal recovery funding in the USVI. Among other responsibilities, these entities are responsible for ensuring that Public Assistance program and Hazard Mitigation Grant Program participants are in compliance with all programmatic and administrative requirements. For example, the Virgin Islands Territorial Emergency Management Agency is responsible for, among other things, preparing and submitting quarterly progress and financial reports to FEMA for certain Public Assistance program projects and for all Hazard Mitigation Grant Program projects. In addition, the governor’s authorized representative is responsible for ensuring the territory’s compliance with all requirements outlined in the FEMA-approved administrative plans. Third, in February 2019, the USVI’s new Office of Disaster Recovery assumed responsibility for overseeing the Public Assistance program and Hazard Mitigation Grant Program, including tracking and reporting on the progress of individual projects and overseeing the submission of reimbursement requests for completed work. The office is also responsible for monitoring and publicly reporting the status of federal recovery funding at http://www.usviodr.com/. In addition, according to the office’s Director, her team also meets with FEMA officials on a bi- weekly basis to discuss recovery activities more generally and raise any potential issues for discussion, such as challenges with submitting quarterly progress and financial reports to FEMA in a timely manner. FEMA. FEMA officials at the USVI, regional, and headquarters level are responsible for overseeing the USVI’s implementation of federal recovery programs. Specifically, once FEMA obligates grant funding for a project, FEMA officials on the ground in the USVI are responsible for the day-to- day monitoring of individual projects using a variety of tools. For example, FEMA officials stated they use the information included in quarterly progress and financial reports to help ensure that subrecipients are in compliance with applicable federal requirements and that potential problems are identified and addressed in a timely manner. Further, FEMA officials are to conduct quarterly meetings with USVI officials to ensure regular communication and coordination and to discuss program implementation, raise potential challenges, and identify solutions. In addition to monitoring ongoing projects, FEMA officials in the USVI are responsible for managing the process for closing out Public Assistance program and Hazard Mitigation Grant Program projects that have been completed. To facilitate this close-out process, the USVI is to compile all required documentation for an individual project and submit this paperwork to FEMA. FEMA is to review the documentation to ensure that all work has been completed in accordance with the project’s scope of work as well as relevant laws, federal regulations, and program requirements. FEMA Region II officials and FEMA officials in the Grant Programs Directorate in headquarters also provide higher-level oversight of Public Assistance program and Hazard Mitigation Grant Program implementation. For example, FEMA Region II officials stated they analyze the USVI’s quarterly financial reports to identify patterns that may indicate financial challenges, such as irregularities in the amount of funding the USVI has drawn down for projects or an excess or lack of reimbursements to subrecipients for completed work. At FEMA headquarters, officials in the Grant Programs Directorate stated they assess the financial condition of projects annually to identify potential challenges in administering grant funding and may enhance monitoring efforts as needed. These officials stated they also review the timeliness, completeness, and accuracy of information submitted in quarterly reports to help monitor project milestones and identify potential challenges that require FEMA’s attention. FEMA Has a Consolidated Standard Operating Procedures Document for Monitoring Public Assistance Projects, but Not for Hazard Mitigation Grant Program Projects FEMA has issued numerous documents that provide useful information and guidance for implementing and monitoring the Public Assistance program and the Hazard Mitigation Grant Program, such as program and policy documents, fact sheets, job aids, and operational manuals. For example, one document for monitoring the Public Assistance program is the Public Assistance Program Management and Grant Closeout Standard Operating Procedure. This standard operating procedures document provides FEMA officials across all disasters nationwide with a common understanding of the expectations and requirements for managing projects. It also clearly and concisely outlines the roles and responsibilities, requirements, key tasks and milestones, and performance measures associated with monitoring and closing out Public Assistance program projects. However, FEMA has not developed a similar consolidated standard operating procedures document to provide a clear and concise roadmap for monitoring and closing out projects under the Hazard Mitigation Grant Program. In July 2019, FEMA officials in headquarters and FEMA Region II told us that guidance on key policies and procedures for managing FEMA’s varied hazard mitigation efforts can be found across multiple sources. However, these officials stated that FEMA has not consolidated this guidance into a single document for FEMA-wide use that focuses on the oversight of Hazard Mitigation Grant Program projects specifically. Further, FEMA’s existing guidance documents do not provide a concise roadmap that outlines roles and responsibilities, key tasks and milestones, and performance measures for FEMA officials to use when monitoring and closing out individual program projects for any given disaster. For instance, the Hazard Mitigation Assistance Guidance includes information on many topics, such as program eligibility requirements; roles and responsibilities for recipient and subrecipient personnel; and oversight requirements spanning three separate FEMA mitigation programs. Likewise, FEMA’s 250-page Hazard Mitigation Field Operations Guide includes operating procedures, descriptions of major tasks for mitigation positions, and job aids, including samples and templates of tools to help hazard mitigation staff implement defined tasks. While these and other FEMA documentation collectively provide important information regarding the agency’s broader hazard mitigation efforts, they are not focused specifically on the Hazard Mitigation Grant Program, or the detailed processes FEMA officials on the ground should use to conduct the day-to-day monitoring of individual projects. Further, these documents do not detail the use of performance measures in effectively monitoring Hazard Mitigation Grant Program projects— information that is included in FEMA’s standard operating procedures document for the Public Assistance program. When we asked FEMA officials responsible for implementing the program in the USVI why the agency had not developed a consolidated standard operating procedures document specific to the Hazard Mitigation Grant Program, they stated that developing this document had not been a priority for the agency—which could be due to the relatively small size of this program compared to larger and more complex recovery programs, such as the Public Assistance program. Although FEMA obligates more funding through the Public Assistance program, the Hazard Mitigation Grant Program nonetheless plays a critical role in ensuring that disaster- affected communities can undertake mitigation measures specifically designed to enhance the resilience of their infrastructure during future disasters. Standards for Internal Control in the Federal Government states that management should document in policies and procedures each unit’s responsibility for an operational process’s objectives in the appropriate level of detail to allow management to effectively monitor the control activity. In addition, these standards state that management should define objectives in specific and measurable terms so they are understood at all levels of the organization. This includes clearly defining what is to be achieved, who is to achieve it, how it will be achieved, and the timeframes for achievement. Further, leading practices identified in the Program Management Institute’s Standard for Program Management call for agencies to develop a program roadmap outlining information on the program’s intended direction and providing a set of documented success criteria for each key milestone and decision point. This roadmap can be a valuable tool for managing the execution of the program and for assessing the program’s progress toward achieving its goals. FEMA’s consolidated standard operating procedures document for the Public Assistance program provides such a roadmap, stating that FEMA’s main goal is to provide effective assistance and excellent customer service necessary to assist disaster-affected communities to recover while also ensuring the responsible stewardship of public funds. FEMA does have guidance on key policies and procedures for managing its hazard mitigation efforts, including through the Hazard Mitigation Grant Program. However, FEMA could further strengthen its existing guidance by consolidating this information into a single document for FEMA-wide use, similar to the one FEMA uses for the Public Assistance program. FEMA officials told us that having a detailed standard operating procedures document for the Hazard Mitigation Grant Program that clearly and concisely outlined roles and responsibilities, key objectives and tasks, and milestones for conducting monitoring and close-out activities would be helpful in effectively overseeing program projects. In addition to FEMA officials, the governor’s authorized representative in the USVI also told us that such a document would help to ensure that both FEMA and USVI officials are following the necessary procedures and guidance when conducting program management and close-out activities. Assessing the need for such a consolidated standard operating procedures document for the Hazard Mitigation Grant Program would help FEMA determine whether existing guidance should be strengthened to ensure that agency officials across all disasters are using a consistent approach in carrying out their responsibilities under the program. Conclusions As of June 30, 2019, FEMA had obligated more than $1.9 billion in grant funding through the Public Assistance program and Hazard Mitigation Grant Program to help the USVI recover from the catastrophic 2017 hurricane season. As part of the Public Assistance program, FEMA authorized the STEP pilot program in the USVI—and in other locations— to supplement other FEMA programs and provide necessary additional capacity to help address the emergency sheltering needs of disaster survivors. While FEMA decided to discontinue the STEP pilot program, the agency has not evaluated how this decision will affect its ability to provide emergency sheltering assistance in the future. FEMA has a responsibility to provide assistance, when requested, to address the emergency sheltering needs of disaster survivors. Given that FEMA will no longer use the STEP pilot program, taking steps to evaluate its options for addressing these needs will help FEMA to assess its capacity for providing effective emergency sheltering assistance in the future and to properly plan for when the next disaster inevitably strikes. In addition, FEMA has issued numerous policy documents, guides, and other useful documents to assist FEMA officials to effectively monitor and oversee Hazard Mitigation Grant Program projects. However, FEMA has not developed a consolidated standard operating procedures document specific to the Hazard Mitigation Grant Program that provides a clear and concise roadmap for FEMA officials’ use in monitoring individual projects. Assessing the need for a consolidated roadmap for agency-wide use would help FEMA determine whether existing guidance for effectively monitoring Hazard Mitigation Grant Program projects should be strengthened. Recommendations for Executive Action We are making the following two recommendations to FEMA: The FEMA Administrator should evaluate the agency’s options for providing future emergency sheltering assistance. (Recommendation 1) The FEMA Administrator should assess the need for an agency-wide consolidated standard operating procedures document for the Hazard Mitigation Grant Program that provides detailed information on the roles and responsibilities, requirements, and key tasks and milestones for monitoring and closing out program projects. (Recommendation 2) Agency Comments and Our Evaluation We provided a draft of this product to DHS and the USVI government for review and comment. DHS provided written comments, which are reprinted in appendix IV and summarized below. DHS and the USVI government provided technical comments, which we incorporated as appropriate. DHS concurred with both our recommendations and described the actions it plans to take in response. With regard to our first recommendation, DHS stated that it will evaluate FEMA’s options for providing emergency sheltering assistance through its Individual Assistance Division and provide any recommendations for action, as appropriate, to FEMA’s Assistant Administrator for Recovery. DHS anticipates completing this evaluation by February 2020. This action, if fully implemented, should address the intent of the recommendation. With regard to our second recommendation, DHS stated that FEMA will assess the need for an agency-wide consolidated standard operating procedures document for the Hazard Mitigation Grant Program and, if deemed necessary, FEMA will develop this document. DHS anticipates this effort will be completed by August 2020. This action, if fully implemented, should address the intent of the recommendation. In the meantime, DHS noted that FEMA will consider updating its website to include a single portal providing access to all existing guidance documents relevant to monitoring and overseeing Hazard Mitigation Grant Program projects. We will monitor DHS’s and FEMA’s efforts to address these two recommendations. We are sending copies of this report to the appropriate congressional committees, the Secretary of Homeland Security, the FEMA Administrator, the USVI government, and other interested parties. In addition, the report is available at no charge on the GAO website at http://www.gao.gov. If you and your staff have any questions, please contact me at (202) 512- 8777 or curriec@gao.gov. GAO staff who made key contributions to this report are listed in appendix V. Appendix I: Status of Public Assistance Program Funding in the U.S. Virgin Islands The Federal Emergency Management Agency (FEMA) obligated more than $1.8 billion in Public Assistance grant funding for 618 projects across the U.S. Virgin Islands (USVI) as of June 30, 2019. Specifically, FEMA obligated more than $1.1 billion for emergency work projects (categories A and B), about $588.5 million for permanent work projects (categories C through G), and about $141.2 million for management costs (category Z). As of that date, the USVI expended nearly $1.1 billion—about 59 percent of total Public Assistance obligations to the USVI—to reimburse subrecipients for completed work. Of this nearly $1.1 billion, the USVI expended about $857.5 million (78 percent) for emergency work projects, $211.3 million (19 percent) for permanent work projects, and $29.9 million (3 percent) for management costs. The majority of FEMA’s obligations and the funding the USVI expended as of June 30, 2019, are for emergency work because these projects began soon after the disasters struck and focused on debris removal and providing assistance to address immediate threats to life and property. In contrast, permanent work projects take time to identify, develop, and ultimately complete as they represent the longer-term repair and restoration of public infrastructure. While the data below represent the status of Public Assistance funding as of June 30, 2019, the amount of grant funding FEMA obligates and the USVI expends will likely increase over time as additional projects are finalized and approved. Emergency work. Of the more than $1.8 billion FEMA obligated as of June 30, 2019, more than 1.1 billion (61 percent) was obligated for 410 emergency work projects in Public Assistance program categories A and B. Category A: Debris removal. FEMA obligated about $139.9 million for 88 projects focused on debris removal activities across the territory. For example, FEMA obligated $81.8 million to the USVI Water and Power Authority for territory-wide debris removal efforts (see fig. 9). Of the $139.9 million FEMA obligated for debris removal, the USVI expended about $76.9 million (55 percent) as of June 30, 2019. Category B: Emergency protective measures. FEMA obligated about $985.6 million for 322 projects focused on emergency measures. For example, FEMA obligated about $278.2 million for the Sheltering and Temporary Essential Power (STEP) pilot program as of June 30, 2019, to fund certain types of temporary repairs to private homes. In addition, FEMA obligated approximately $111.9 million for the purchase and installation of modular units to be used as temporary classrooms and other facilities while permanent school buildings are repaired or replaced (see fig. 10). Of the $985.6 million FEMA obligated for emergency protective measures, the USVI expended about $780.7 million (79 percent) as of June 30, 2019. Permanent work. Of the more than $1.8 billion in Public Assistance grant funding FEMA obligated as of June 30, 2019, about $588.5 million (32 percent) was obligated for 200 permanent work projects across categories C through G. These permanent work projects included more than $383.1 million for cost-effective hazard mitigation measures to reduce the future risk of disaster damage to infrastructure. Category C: Roads and bridges. FEMA obligated about $5.9 million for 40 projects focused on repairing roads and bridges in the territory, 9 of which included hazard mitigation measures totaling about $1.6 million. For example, FEMA obligated about $233,000 for one project to repair a road on St. Croix damaged by floodwaters. This project included approximately $61,000 for hazard mitigation measures to reduce the likelihood of erosion during future flooding events. Of the $5.9 million FEMA obligated for category C, the USVI expended about $86,000 (1.5 percent) as of June 30, 2019. Category D: Water control facilities. As of June 30, 2019, FEMA did not have any projects in this category. According to FEMA officials, the USVI does not have water control infrastructure—such as berms or levees—that would fall under category D. Category E: Buildings and equipment. FEMA obligated about $68.3 million for 101 projects focused on repairing damaged structures in the territory, 49 of which included hazard mitigation measures totaling about $3.1 million. For example, FEMA obligated about $59.7 million for one project to replace 5 heavily damaged buildings in a public housing facility in St. Thomas (see fig. 11). While FEMA obligated this project through the standard Public Assistance program, FEMA and the USVI plan to work to develop a fixed-cost estimate with the intention of transitioning this project to the alternative procedures program, according to FEMA documentation. Further, this documentation states that hazard mitigation measures will be incorporated into the new structures by implementing internationally adopted building codes and standards for wall and window replacements. Of the $68.3 million FEMA obligated for Category E, the USVI expended about $533,000 (0.8 percent) as of June 30, 2019. Category F: Utilities. Of the $588.5 million FEMA obligated for permanent work projects, $505.6 million (86 percent) was obligated for 23 projects focused on repairing utilities, 13 of which included hazard mitigation measures totaling about $378.2 million. Specifically, FEMA obligated $481.8 million—or 95 percent of the $505.6 million—through the standard Public Assistance program for projects focused on territory-wide permanent electrical distribution system repairs. This includes replacing damaged wooden utility poles with more resilient composite fiberglass poles that can withstand 200 mile per hour winds as well as power transmission lines and transformers (see fig. 12). Of the $505.6 million FEMA obligated for category F, the USVI expended about $210.4 million (42 percent) as of June 30, 2019. Category G: Parks, recreational, and other facilities. As of June 30, 2019, FEMA obligated about $8.8 million for 36 projects focused on repairing parks, playgrounds, and other facilities, 5 of which included hazard mitigation measures totaling about $214,000. For example, FEMA obligated about $1.5 million in March 2019 for two projects to repair the USVI’s Tsunami Early Warning System, which comprises a network of warning stations that alert residents of a potential tsunami event (see fig. 13). As of September 2019, these were the only projects FEMA had obligated under the Public Assistance alternative procedures program in the territory, according to FEMA officials. These projects included about $185,000 for hazard mitigation measures to replace wooden poles with higher-rated steel poles that are able to withstand high winds and impacts from flying debris during a storm. Of the $8.8 million FEMA obligated for category G, the USVI expended about $246,000 (3 percent) as of June 30, 2019. Appendix II: Status of Hazard Mitigation Grant Program Funding in the U.S. Virgin Islands As of June 30, 2019, the Federal Emergency Management Agency (FEMA) obligated about $60.6 million for 22 Hazard Mitigation Grant Program projects in the U.S. Virgin Islands (USVI) and the USVI expended about $1.7 million (3 percent) across 5 projects. Unlike Public Assistance program projects that, in many cases, are focused on rapidly providing emergency services or repairing critical disaster-damaged infrastructure and systems, Hazard Mitigation Grant Program projects are designed to fund a variety of measures to increase the longer-term resilience of the USVI’s infrastructure during future disasters. Information on selected Hazard Mitigation Grant Program projects in the USVI that received obligations as of June 30, 2019, is detailed below. Virgin Islands Territorial Emergency Management Agency and Bureau of Information Technology Emergency Operations Center and Safe Room Retrofit. FEMA obligated about $22.5 million to fund the retrofit of the USVI Territorial Emergency Management Agency’s Emergency Operations Center. The new facility will serve as the headquarters for both the USVI Territorial Emergency Management Agency and the USVI Bureau of Information Technology and house a 911 Emergency Call Center. According to FEMA documentation, the facility will include a safe room to allow emergency personnel to shelter in place during disasters and will contain sufficient space to house FEMA and other federal personnel, as necessary. Further, the facility will include a hardened communications system to ensure emergency responders are able to effectively communicate during emergency events, among other improvements. The Comprehensive Territorial Hazard Mitigation and Resilience Plan Project. FEMA obligated nearly $5.0 million to fund the development of an in-depth, comprehensive hazard mitigation and resilience plan for territory-wide use. FEMA officials stated that unlike in the contiguous United States, the USVI does not have any entities responsible for formally developing similar plans to guide operations and mitigation activities across various sectors, such as protecting the potable water supply and assessing economically feasible options for development. As a result, FEMA officials told us that this project represents an important effort to develop a holistic, territory-wide hazard mitigation plan that would cover all relevant sectors. The Spring Gut Watershed Green Space Acquisition and Stormwater Management Project. FEMA obligated nearly $1.0 million to fund the first phase of a $2.0 million project to purchase 50 acres of undeveloped land and develop storm water retention measures—such as berms and rock linings—to reduce downstream flooding and the associated damages to roads, homes, and infrastructure (see fig. 14). FEMA officials told us that the first phase of the project included an environmental and historical preservation review of the target locations to confirm program eligibility before actual construction activities can begin. Fortuna/Bordeaux Fire Station Retrofit. FEMA obligated more than $470,000 to fund the first phase of a nearly $5.0 million project to retrofit a fire station in St. Thomas. Specifically, the project will upgrade the facility’s structure to applicable codes and standards to mitigate the risks posed by hurricane-force winds, including the dangers posed by flying debris. Further, the retrofit will include the installation of a steel-reinforced concrete safe room and a back-up emergency power generator to ensure the safety and protection of emergency personnel and the continuity of emergency response activities during a disaster, according to FEMA documentation. Appendix III: The Governor Juan F. Luis Hospital and Medical Center in St. Croix, U.S. Virgin Islands The Governor Juan F. Luis Hospital and Medical Center (JFL hospital) in St. Croix provides an illustrative example of the processes and challenges associated with developing and implementing Public Assistance program projects in the U.S. Virgin Islands (USVI). In September 2017, Hurricane Maria’s strong winds and torrential rains caused severe damage to the facility’s roof; heating, ventilation, and air conditioning system; and electrical, water, and sewage systems, according to Federal Emergency Management Agency (FEMA) documentation. Further, the infusion of water—both during and after the storm—saturated the interior of the hospital, destroyed medical equipment and hospital furnishings, and facilitated the growth of hazardous mold. Figure 15 details selected hurricane damage to the facility. Due to the extensive damage, the JFL hospital has been operating at reduced capacity since the hurricane and certain functions have been relocated to undamaged areas, according to FEMA documentation. This documentation states that while the hospital continues to provide limited medical services to St. Croix residents, it is no longer capable of providing critical care services. Since alternate options were either limited or unavailable on the island following the storms, St. Croix residents in need of life sustaining medical treatments such as chemotherapy infusions or who are experiencing life-threatening health events such as cardiac failure or trauma must be transported out of the territory to receive life- saving care, according to FEMA documentation. Public Assistance Program Emergency Work Projects at the JFL Hospital Following Hurricane Maria, FEMA obligated grant funding for several Public Assistance emergency work projects to help keep the JFL hospital functioning and capable of providing limited medical services to St. Croix residents. For example, in August 2018, FEMA obligated $119,000 in grant funding to reimburse the JFL hospital for the use of an emergency backup generator through Public Assistance program category B, which provides funding for emergency protective measures. According to JFL personnel, this funding covered the cost of using the backup generator until the facility’s primary electrical system could be restored. In another example, FEMA obligated about $2.4 million in August 2018—also through Public Assistance program category B—to fund the rental of mobile dialysis trailers (see fig. 16). According to JFL personnel, these trailers were acquired to replace all 14 of the facility’s dialysis units that were destroyed during the storm. In addition to Public Assistance projects focused on maintaining the existing facility, FEMA obligated $43.2 million in January 2018 to fund the purchase and installation of modular units to serve as a temporary medical facility through category B emergency work. According to FEMA documentation, this temporary facility is intended to provide critical medical services to St. Croix residents—including an emergency room, pediatric care, a labor and delivery ward, and an intensive care unit, among other services—until a permanent facility is completed (see fig. 17). JFL personnel told us that completing this project is a key priority as it will enable them to transition all medical services from the main facility, which continues to deteriorate over time. However, they stated that implementing this project has been challenging. For example, JFL personnel told us that when former JFL administrators were developing the project, they incorporated the cost of procuring and installing the modular units, but omitted the costs associated with acquiring, installing, and certifying new medical equipment for use in the interim facility. These personnel stated that these costs should have been incorporated into the original project paperwork and clarified that the acquisition and certification of new equipment was critical in ensuring the hospital’s provision of medical services would not be disrupted. Specifically, they explained that relocating the facility’s existing medical equipment was not a feasible option as it would result in an unacceptable lapse in medical services during the time-consuming process of deconstructing, transferring, reinstalling, and recertifying this equipment in the modular facility. JFL personnel stated they worked closely with FEMA officials and contractor personnel to update the project’s paperwork and request additional program funding for this new medical equipment. In August 2019, FEMA approved the updated paperwork and obligated additional funding for this project, according to FEMA documentation. JFL personnel also stated they have limited capacity to effectively manage and oversee the construction of the temporary facility due to competing responsibilities. In addition to managing this project, they stated they were occupied with the continuous maintenance challenges associated with keeping the deteriorating main facility functioning while also working to develop options for a permanent facility through the Public Assistance alternative procedures program, as discussed below. They told us FEMA officials and contractor personnel had been helpful in providing assistance, but stated they would benefit from a larger hospital management team that could focus specifically on planning and implementing the facility’s numerous recovery efforts. As of July 2019, JFL personnel stated their aim is to officially open the temporary facility in the spring of 2020. Public Assistance Alternative Procedures Permanent Work at the JFL Hospital In conjunction with the temporary facility’s construction, FEMA officials and JFL personnel are working to develop a permanent work project under the Public Assistance alternative procedures program to replace the damaged hospital. Further, in providing a critical service, the JFL hospital is eligible to use the new flexibilities provided by the Bipartisan Budget Act of 2018. This Act allows FEMA—when using the alternative procedures—to fund the repair or full replacement of the hospital to accepted industry standards regardless of any pre-disaster damage or wear and tear the facility may have sustained prior to the 2017 hurricanes. FEMA officials and JFL personnel stated the Act therefore provides a valuable opportunity to restore the facility to a better condition than it was in prior to the storms. However, JFL personnel told us that pursuing this permanent work project included challenges. For example, JFL personnel told us that maintaining the damaged facility while FEMA determined whether the hospital was eligible for repairs to the existing structure or a complete replacement under the Bipartisan Budget Act was challenging. Specifically, they explained that while FEMA worked to finalize this determination, management was in the difficult position of deciding where and how to invest its finite resources to keep the constantly deteriorating facility functioning. For example, these personnel explained that if FEMA determined that the facility is ineligible for replacement under the Act, they would immediately invest money into the existing facility to address critical components that require urgent attention, such as the water and wastewater systems. In contrast, if FEMA determined that the facility is indeed eligible for replacement, management would strategically invest the minimum amount of resources required to keep the facility functioning with the full knowledge that it would eventually be demolished. Figure 18 details selected examples of temporary fixes JFL personnel implemented to keep the facility functioning. In May 2019, FEMA officially determined that the JFL hospital was eligible under the Bipartisan Budget Act of 2018 for a complete replacement through the Public Assistance alternative procedures program. JFL personnel told us they are working with FEMA officials and medical industry experts to ensure that they take advantage of the flexibilities provided by the Bipartisan Budget Act when developing the project. As of July 2019, these personnel explained they are in the early stages of working with FEMA officials and territorial stakeholders to assess options for the replacement facility and are designing a strategy to ensure the future hospital is able to sufficiently address the healthcare needs of USVI residents. Appendix IV: Comments from the Department of Homeland Security Appendix V: GAO Contact and Staff Acknowledgments GAO Contact Chris Currie, (202) 512-8777 or curriec@gao.gov. Staff Acknowledgments In addition to the contact named above, Joel Aldape (Assistant Director), Bryan Bourgault (Analyst in Charge), Aaron Gluck, Eric Hauswirth, Brian Lipman, Amanda Miller, Heidi Nielson, and Kevin Reeves made key contributions to this report.

Why GAO Did This Study In September 2017, two major hurricanes—Irma and Maria—struck the USVI, causing billions of dollars in damage. FEMA is the lead federal agency responsible for assisting the USVI to recover from natural disasters. FEMA administers the Public Assistance program and Hazard Mitigation Grant Program in partnership with the USVI government, providing grant funding for response and recovery activities, including life-saving emergency protective measures, the repair or replacement of public infrastructure, and measures to increase the territory's resilience during future disasters. GAO was asked to review the federal government's response and recovery efforts in the USVI. This report examines (1) the status of Public Assistance program and Hazard Mitigation Grant Program funding and challenges, if any, with implementation, (2) the STEP pilot program, and (3) the oversight of these programs. GAO reviewed documentation and data on the Public Assistance program and Hazard Mitigation Grant Program in the USVI as of June 30, 2019. GAO interviewed FEMA and USVI officials regarding the status of recovery efforts and associated challenges, and conducted site visits to the USVI islands of St. Croix, St. Thomas, and St. John. What GAO Found As of June 30, 2019, FEMA obligated more than $1.9 billion in grant funding for 640 projects in the U.S. Virgin Islands (USVI) through the Public Assistance program and Hazard Mitigation Grant Program in response to the 2017 hurricanes. However, the limited availability of local USVI personnel to staff key recovery positions and the territory's difficult fiscal situation presented challenges in implementing these programs. Further, FEMA and USVI officials stated they faced challenges with implementing the Public Assistance alternative procedures program, which provides the USVI with flexibility in determining when and how to fund projects. Specifically, these officials stated that developing accurate fixed-cost estimates and using new flexibilities authorized by law delayed longer-term recovery projects. USVI officials told GAO they plan to take a cautious approach when deciding whether to pursue projects using the alternative procedures. FEMA expanded its Sheltering and Temporary Essential Power (STEP) pilot program in the USVI to address the lack of other sheltering options for survivors, such as hotels. The program aimed to provide minimal, temporary repairs to damaged homes to quickly make them habitable. In May 2019, FEMA decided it would not use the STEP pilot program in the future since it did not provide assistance as rapidly as intended. Historically, the program was used to address survivors' emergency sheltering needs. However, since ending it, FEMA has not evaluated options for providing future emergency sheltering assistance. Doing so could help FEMA plan for when the next disaster inevitably strikes. The USVI and FEMA established structures for overseeing recovery efforts. For example, the USVI established a new office to oversee federal recovery programs and FEMA has processes in place to oversee recovery projects at the local, regional, and headquarters levels. However, GAO found that FEMA does not have a consolidated standard operating procedures document for monitoring Hazard Mitigation Grant Program projects. Assessing the need for a consolidated document would help FEMA determine whether its existing guidance should be strengthened. What GAO Recommends GAO recommends that FEMA (1) evaluate its options for providing emergency sheltering and (2) assess the need for a consolidated standard operating procedures document for the Hazard Mitigation Grant Program. The Department of Homeland Security concurred with these recommendations.