Statement of Sigurd R. Nilsen, Ph.D., Director, Education Workforce, and Income Security Issues,
U.S. General Accounting Office

Testimony Before the Subcommittee on Human Resources
of the House Committee on Ways and Means

Hearing on Unemployment Fraud and Abuse

June 11, 2002

Mr. Chairman and Members of the Subcommittee:

I am pleased to be here today to discuss the Department of Labor’s Unemployment Insurance (UI) program, which is a key component in ensuring the financial security of America’s workforce. The UI program is a federal-state partnership designed to partially replace lost earnings of individuals who become unemployed through no fault of their own and to stabilize the economy in times of economic downturn. The UI program paid about $30 billion in benefits in calendar year 2001 to workers who lost their jobs. The health of each state’s UI program depends, in part, on the ability of the state to control its benefit payments by accurately determining individuals’ eligibility for UI benefits in a timely manner. Inaccurate or untimely eligibility information may contribute to overpayments and fraud.

Reports from Labor’s Office of Inspector General (OIG) and others have identified numerous aspects of the UI program that may be vulnerable to overpayments and fraud. Today, I will be providing information from our draft report that we have provided to Labor for its comment on our findings, conclusions, and recommendations. Our report is due to be issued in July 2002. I will discuss (1) the extent and type of overpayments in the UI program, including those that may be attributable to fraud or abuse; (2) the factors that contribute to overpayments in the UI program; and (3) the broader management issues that may affect the states’ ability to effectively control their UI benefit payments.

To address these issues, we reviewed internal Labor guidance and documentation, performance plans and reports, performance data, as well as overpayment data from Labor’s Benefit Accuracy Measurement (BAM) and Benefit Payment Control (BPC) systems. In addition, we conducted in-depth interviews with more than 100 management and line staff in Labor’s headquarters and 6 regional offices, as well as UI officials in 6 states—California, Colorado, Illinois, Maryland, Massachusetts, and New York.^[1] We selected these states based on numerous criteria, including performance data from the Department of Labor, size of their workforce, availability of overpayment detection and recovery tools, and geographic location. Finally, we spoke with other groups that are involved in unemployment insurance, such as employer representatives and the National Association of State Workforce Agencies.

In summary, our work shows that of the $30 billion in UI benefits paid in calendar year 2001, Labor estimates that this includes about $2.4 billion in overpayments, including $560 million attributable to fraud or abuse. Labor’s analysis also suggests that the states could have detected and/or recovered about $1.3 billion of the total overpayments given their current policies and procedures. Labor based these estimates on data from its quality assurance system, which involves an in-depth analysis of individual UI claims in each state. Labor’s quality assurance data document numerous categories of overpayments, including individuals who work while receiving benefits, or misrepresent their identity. Other sources of overpayments include agency errors and inaccurate or untimely information provided by employers. Our work shows that management and operational practices at both the state and federal level contribute to overpayments in the UI program. At the state level, many states place a higher priority on quickly processing and paying UI claims than on taking the necessary steps to adequately verify claimants’ initial and continued eligibility for UI benefits. As a result, we found that many states do not adequately verify information reported by claimants. At the federal level, we found that Labor’s policies and directives emphasize quickly processing and paying claims, with only limited attention given to payment accuracy. While we recognize the importance of paying benefits to individuals in a timely manner, Labor’s performance measurement system does not provide sufficient incentives and sanctions for states to balance the need for payment timeliness with the need for payment accuracy.

Background

The UI program was established by Title III of the Social Security Act in 1935 and is a key component in ensuring the financial security of America’s workforce. This complex program, which is administered jointly by the federal Department of Labor’s Employment and Training Administration and the states, provides temporary cash benefits to workers who lose their jobs through no fault of their own. Labor is responsible for monitoring state operations and procedures, providing technical assistance and training, as well as analyzing UI program data to diagnose potential problems. Although Labor provides oversight and guidance to ensure that each state operates its program in a manner that is consistent with federal guidelines, primary responsibility for administering the program lies with the states.

State claims representatives determine claimants’ eligibility for UI benefits by gathering essential information, such as their identity, employment history, and other sources of income they may have. To enhance the efficiency and cost-effectiveness of their UI systems, many states have established centralized service centers that allow claimants to apply for benefits by telephone, fax, or the Internet, rather than in person at a local office. To be eligible for UI benefits in most states, claimants must (1) have worked for a specified amount of time in a job that is covered by the unemployment insurance program; (2) have left their prior jobs involuntarily (such as by employer layoff) or have quit their jobs for “good cause”; (3) be currently “able and available” for work, and, in most states, actively seeking work; (4) enroll in employment services or job training programs (in some states); and (5) be legally eligible to work—for example, noncitizens must be lawfully admitted to work in the United States, or lawfully present for other reasons. States are generally expected to provide benefits to the claimant within 14 to 35 days of application.

The UI program is funded through federal and state taxes levied on employers. States’ taxes pay the actual unemployment insurance benefits, whereas administrative costs are generally financed through the federal tax. Labor holds these funds in the Unemployment Trust Fund of the U.S. Treasury. To obtain annual funding from Labor to administer their programs, states submit a request via their annual State Quality Service Plan (SQSP). Labor reviews each state’s plan and makes adjustments in funding as necessary. In fiscal year 2001, Labor provided about $2.3 billion to states to administer their programs.

To ensure UI program integrity, Labor funds two principal kinds of activities for detecting and measuring UI overpayments at the state level—Benefit Payment Control and Benefit Accuracy Measurement. Each state is required to operate a benefit payment control division that is responsible for detecting and recovering overpayments. Each state is required to report overpayment data to Labor on a quarterly basis. By contrast, Labor’s benefit accuracy measurement data is an estimate of the total overpayments in the UI program—in each state and the nation as a whole—based on an examination of a sample of paid and denied claims. Benefit accuracy measurement is one of the main quality assurance systems that Labor uses to assess payment accuracy in the program.

More Than $2 Billion in Overpayments Detected in 2001

Labor’s data show that of the $2.4 billion in estimated overpayments about $1.3 billion could have been detected and/or recovered by the states in 2001 given their existing policies and procedures.^[2] In contrast, the states reported that $650 million in overpayments were made in 2001, of which $370 million was actually recovered. The difference in the overpayment figures produced by the two systems can be attributed to the fact that Labor’s quality assurance estimate is based on a more comprehensive examination of individual UI claims than the states’ benefit payment control activities can generally produce. Our analysis suggests that Labor’s quality assurance system estimate is a more complete assessment of the true level of overpayments in the UI program, partly because the system provides a more in-depth review of individual UI cases and causes of payment errors. We are currently in the process of verifying the precision of these estimates.^[3]

Overpayments Have Changed Little During
the Last 10 Years

Over the past 10 years, the annual overpayment rate estimated by Labor’s quality assurance system has remained fairly constant as a percentage of total benefits paid—ranging from a low of 7.9 percent in 2001 to 9.2 percent in 1999, and averaging about 8.4 percent during that period. Overpayments averaged about $1.8 billion per year and reached a high of $2.4 billion in 2001. (See fig. 1.)

Figure 1: Overpayments Estimated by Labor’s Quality Assurance System, 1992 to 2001

Source: Department of Labor quality assurance data.

The overpayments estimated by Labor’s quality assurance data occur for a number of reasons. Some overpayments result from errors in claimants’ reporting or the state agency’s recording of important eligibility information, such as wages or other sources of income that a claimant obtained while receiving UI benefits (“benefit year earnings” or “base period wages”). Overpayments also occur because claimants are not able and/or available to work, fail to register for employment services as required by their state, or fail to look for a new job as required (“eligibility” violations). Claimants may also be overpaid because they become unemployed for reasons not covered by state law—such as being fired (“separation” issues). Finally, overpayments may occur due to erroneous reporting or recording of a claimant’s dependent information (“dependency” issues), or other causes such as reversal of benefits paid due to an appeals decision (“other” causes). (See fig. 2.) The quality assurance data also classifies overpayments as being “fraud” or “nonfraud.” Fraud can occur when claimants intentionally misrepresent eligibility information, employers file fraudulent claims, or state UI program personnel misuse their access to sensitive information. Of the total overpayments estimated by Labor in 2001, about $560 million (24 percent) were attributed to fraud. Of this amount, about $313 million (56 percent) were due to unreported earnings. However, we found that the states differ substantially in how they define fraud. For example, some states may include overpayments resulting from unreported earnings such as fraud, while other states do not. Thus, state-to-state comparisons of the level of fraud in the UI program and the activities that constitute fraud are difficult to make.

Figure 2: Categories of $2.4 Billion in Overpayments Estimated by Labor’s Quality Assurance System (2001)

Note: Numbers in parentheses are in millions of dollars.

Source: Labor’s quality assurance data.

Although some categories of overpayments are more difficult than others to detect or recover, Labor’s analysis suggests that the states could have detected and recovered about $1.3 billion of the $2.4 billion in estimated overpayments in 2001. In particular, Labor’s data show that existing state processes and procedures could have detected more overpayments attributable to unreported recipient income and wages and payments to individuals who are not entitled to UI benefits due to the circumstances under which they became unemployed. Labor’s analysis also suggests that other types of overpayments are likely to be detected by most states given their current policies and procedures. These include income from social security programs, unreported vacation or severance pay, and illegal aliens claiming benefits. Furthermore, Labor’s analysis showed that a substantial proportion of the overpayments detected by the states could be recovered using commonly available procedures, such as offsetting claimants’ current and future benefits, and intercepting other sources of income, such as state tax refunds. Labor determined that the remaining $1.1 billion in estimated overpayments could probably not be detected or recovered by the states due to limitations in their existing policies and procedures. For example, overpayments caused by state agency errors are generally not pursued for recovery.

Labor’s Quality Assurance System Data Provide a More Complete Representation of UI Overpayments

In contrast to Labor’s quality assurance overpayment estimate, the states’ benefit payment control systems reported about $650 million in overpayments in 2001, of which about $370 million was recovered. Based on our analysis as well as analysis performed by Labor’s Division of Performance Management, we believe that Labor’s quality assurance system data represent a more complete assessment of the true level of UI overpayments than the benefit payment control figure reported by the states. In particular, the quality assurance system is able to estimate all the potential overpayments that have occurred in each state’s UI program because it is based on a statistically valid sample of UI claims from each state. Moreover, quality assurance investigators are able to conduct a more detailed, comprehensive analysis of each case reviewed than is typically possible for most states’ benefit payment control operations. For example, investigators are generally able to spend more time verifying the accuracy of the claims information by personally contacting employers, claimants, and third parties. They also typically commit between 5 and 8 hours examining a single case, allowing for a more in-depth review of a claimant’s eligibility. By contrast, the states’ benefit payment control activities are often affected by factors that limit their ability to detect and/or recover overpayments. These factors include (1) limited staffing and funding and (2) a lack of access to timely data sources. Moreover, benefit payment control personnel are required to quickly examine thousands of cases to identify overpayments, thus potentially limiting their ability to thoroughly review cases for payment accuracy.

Overpayments Caused by Management and Operational Practices at the State and Federal Level

We identified various management and operational practices at both the federal and state level that contribute to UI overpayments. In particular, both Labor and the states tend to place primary emphasis on quickly processing and paying UI claims and may not sufficiently balance the need to make timely payments with ensuring payment accuracy. While we recognize the importance of providing UI benefits in a timely manner to individuals who are unemployed, our work suggests that Labor and the states do not always take the necessary steps to adequately verify claimants’ initial and continuing eligibility for benefits. While some of the states we visited use automated data sources to determine if claimants are working or obtaining other benefits while receiving UI, others rely heavily on self-reported information from claimants to make payment decisions. In addition, we found that Labor’s performance measures generally emphasize payment timeliness at the expense of payment accuracy. Moreover, Labor has been reluctant to link the states’ performance on payment accuracy to the annual administrative funding process as a way of holding states accountable for performance. Despite these problems, we found that Labor is taking some actions to improve UI program integrity, such as working to help states obtain automated data sources essential to making more accurate and timely eligibility decisions.

States Do Not Always Balance Need for Payment Timeliness with Payment Accuracy

The emphasis that an agency places on critical program activities can be measured, in part, by the level of staff and other resources devoted to those activities. Consistent with stated program objectives, most of the states we visited place a primary emphasis on quickly processing and paying UI claims, but do not always balance this focus with adequate attention to program integrity. In particular, we found that program managers commonly moved staff assigned to program integrity activities (such as benefit payment control) to claims processing positions in response to increases in the number of UI claims being filed. For example, one state was using only 4 of the 16 positions (25 percent) it was allotted by Labor for benefit payment control. Only one of the six states we visited was fully staffing its benefit payment control operations. The remaining states had transferred staff into other positions, including claims processing. Another state stopped drawing its quality assurance sample for a period of time and moved staff responsible for these operations into claims processing positions when unemployment claims increased during the third quarter of 2001.^[4] Many federal and state officials we interviewed told us that states move staff into claims processing roles from other positions because they lack funding to properly administer all the necessary activities of their UI programs.

States Vary in Their Use of Automation to Independently Verify Claimants’ Information

While states differed in the level of staff and resources devoted to program integrity activities, we also found variation in the processes and tools they used to verify information that could affect a claimant’s eligibility for UI benefits, such as identity, alien status, wages, employment status, or receipt of other federal or state benefits. All of the states we visited conduct basic computer matches that detect potential UI overpayments due to unreported earnings. For example, each state regularly conducts a “Wage/Benefit Crossmatch” that compares the database of UI claimants with the state’s database of individuals’ wages to identify UI recipients who may have unreported income in the same state in which they are receiving UI benefits. However, because state wage data are only available quarterly, the crossmatch relies on information that may be several months old by the time the match is conducted. This delay allows some overpayments to remain undetected for a long period of time. Officials at Labor and in some states emphasized that overpayments are more likely to be recovered if they can be detected quickly. States generally recover a substantial proportion of the overpayments they detect by offsetting a claimant’s current and future UI benefits. However, UI benefits tend to be paid out over a relatively short period of time—about 14 weeks on average—and overpayment detection and recovery activities may begin long after individuals leave the UI rolls. This inability to obtain timely eligibility information places the program at substantial risk for overpayments that may never be recovered.

More timely sources of data than the “Wage/Benefit Crossmatch” exist to verify a claimant’s employment status. State new hires data can provide information on individuals’ current employment status.^[5] States that use this data source have reported that it is helpful in detecting overpayments more quickly. However, we found that the new hires data are not routinely used in all states. Two of the six states we visited do not currently use their new hires data to verify claimants’ earnings or employment status. ^[6] Yet, one of the states we visited reported that because the new hires data detect overpayments earlier than other detection methods, the size of its average overpayment at the time of detection has been reduced by nearly 75 percent, from about $2,800 to roughly $750. Labor’s OIG has identified the new hire database as a potentially useful tool for detecting overpayments resulting from unreported income, which represents a substantial portion of the total UI overpayments each year.^[7] Although Labor has encouraged each state to use its own new hires database for purposes of administering their UI program, a number of states nationwide still do not use it.

While the states’ directory of new hires data are useful for verifying claimants’ employment status, a main limitation is that they only identify this information for claimants within a given state. To detect unreported or underreported wages in other states, some states also use an “Interstate Crossmatch” that is facilitated by Labor.^[8] However, this match also typically relies on wage data that are about 4 to 6 months old. Another type of match called the “Interstate Inquiry” allows states to check a claimant’s UI and employment status in other states. However, this system can generally only be used to check individual claimants and is not designed to verify the status of large numbers of claimants simultaneously.

To enhance the ability of states to verify the status of claimants who could be working or receiving UI benefits in other states, many officials we spoke with advocated giving states access to the Office of Child Support Enforcement’s National Directory of New Hires (NDNH). The NDNH is a comprehensive source of unemployment insurance, wage, and new hires data for the whole nation. However, current law limits access to the NDNH and does not permit individual states to obtain data from it for purposes of verifying claimants’ eligibility for UI.^[9] One possible alternative to the NDNH suggested by some officials for tracking interstate wages and UI benefit receipt is the Department of Labor’s Wage Record Interchange System (WRIS). This system, which was developed in response to the Workforce Investment Act (WIA) of 1998, is a “data clearinghouse” that makes UI wage records available to states seeking employment and wage information on individuals in other states.^[10] Certain federal officials and others familiar with WRIS told us that with some modification—such as incorporating the more timely new hires data from the states—WRIS could be a logical alternative to the NDNH because the computer network for sharing data among the states already exists. However, WRIS currently lacks important pieces of information (such as states’ new hires data) that would make it most useful as an interstate verification tool. Moreover, in a recent report, we noted that some states have been reluctant to become involved with WRIS, partly because of concerns about the cost of administering the system.^[11] Furthermore, we noted that if not all states participate, the value of WRIS will be diminished—even for participating states—because no data will be available from nonparticipating states’ UI wage records.

Some States May Not Verify Claimants’ Receipt of Other Programs’ Benefits

Claimants’ eligibility for UI benefits may be affected if they are receiving benefits from other state or federal programs. For example, claimants in some states are ineligible for UI benefits, or they may receive reduced benefits if they are receiving workers’ compensation. Overpayments can occur if claimants do not accurately report the existence or amount of such benefits when they apply for UI, or if the state employment security agency fails to verify the information in a timely manner.^[12] Only two of the six states we visited verify claimants’ receipt of workers’ compensation using independent sources of information. Moreover, at least one of these states only checks for receipt of workers’ compensation if the claimant self-reports that they are currently receiving such benefits. Similarly, receipt of some federal benefits such as cash payments from Social Security programs may affect a UI claimant’s eligibility for or amount of benefits.^[13] For example, one state we visited requires claims representatives to ask claimants if they are currently receiving Social Security Disability Insurance (DI), which could reduce or eliminate the UI benefits they are eligible to receive. However, if a claimant states that he or she is not receiving DI benefits, then no further actions are taken to independently verify this information. Labor’s quality assurance data estimate that in 2001, about $30 million in UI overpayments were due to unreported social security benefits, such as DI.

Some States Fail to Adequately Verify Claimants’ Identity and Whether They Are Legal Residents

To ensure that UI benefits are paid only to individuals who are eligible to receive them, it is important that states verify claimants’ identity and whether they are legal residents.^[14] However, states may be vulnerable to fraud and overpayments because they rely heavily on claimants to self-report important identity information such as their social security number (SSN), or are unable to verify such information in a timely manner. Prior investigations by Labor’s OIG demonstrate that the failure or inability of state employment security agencies to verify claimants’ identity have likely contributed to millions of dollars in UI overpayments stemming from fraud. One audit conducted in four states (Florida, Georgia, North Carolina, and Texas) revealed that almost 3,000 UI claims totaling about $3.2 million were paid to individuals using SSNs that did not exist, or belonged to deceased individuals. Furthermore, the OIG concluded that illegal aliens filed a substantial proportion of these claims.^[15]

We found that vulnerabilities remain with regard to verifying claimants’ identity and citizenship status. For example, none of the six states we visited have access to the Social Security Administration’s (SSA) State Online Query (SOLQ) system, which can be used to verify the identity of claimants applying for UI by matching their name, date of birth, and SSN in real time. At the time of our review, only two states had access to this system because they were participating in a pilot project with SSA. The states we visited generally use a batch file method in which large numbers of SSNs are periodically sent to SSA for verification.^[16] This process tends to be less timely than online access for verifying claimants’ initial eligibility for benefits. One state we visited reported that it does not perform any verification of the SSNs that UI claimants submit because a prior system it used for verifying SSNs identified only a small number of potential violations. In addition, all six states we visited rely mainly on claimants to accurately self-report their citizenship status when they first apply for UI benefits. State officials told us that they generally do not verify this information with the Immigration and Naturalization Service (INS) unless the claimant states that he or she is a noncitizen. Labor estimates that about $30 million in overpayments in 2001 were due to illegal alien violations.

Even if individuals do not misrepresent their identity or citizenship status to illegally obtain UI benefits, the potential for fraud and abuse may still exist. For example, one state we visited revealed that they, along with a bordering state, identified nine SSNs that are currently being illegally used by multiple individuals as proof of eligibility for employment. Upon further investigation, we determined that these SSNs are being used by approximately 700 individuals in at least 29 states, and that seven of the SSNs belonged to deceased individuals. Although we did not find any instances in which UI benefits were obtained by those individuals earning wages under these numbers, both state and federal officials agreed that the potential for these individuals to fraudulently apply for and receive UI benefits in the future was possible. At the Subcommittee’s request, our Office of Special Investigations is currently investigating the use of these SSNs. Initial indications are that the individuals involved are illegal aliens.

States May Not Receive Timely Information from Employers

To varying degrees, officials from all of the six states we visited told us that employers or their agents do not always comply in a timely manner with state requests for information needed to determine a claimant’s eligibility for UI benefits. For example, one state UI Director reported that about 75 percent of employers fail to respond to requests for wage information in a timely manner. In addition, a Labor OIG audit conducted between 1996 and 1998 revealed that 22 out of 53 states experienced a nonresponse rate of 25 percent or higher for wage requests sent to employers.^[17] A more in-depth review of seven states in this audit also showed that $17 million in overpayments occurred in four of the states because employers did not respond to the states’ request for wage information. We discussed these issues with an official from a national employer representative organization who told us that some employers may resist requests to fill out paperwork from states because they view the process as cumbersome, time-consuming, and cannot always see how fraud and UI overpayments can affect their tax rate. In particular, because employers are unlikely to experience an immediate increase in the UI taxes they pay to the state as a direct result of overpayments, they do not see the benefit in complying with state requests for wage data in a timely manner. Although Labor has taken some limited actions to address this issue, our work to date shows that failure of employers to respond to requests for information in a timely manner is still a problem.^[18]

States Vary in Their Ability To Recover Overpayments

While most states recover a large proportion of their overpayments by offsetting claimants’ current or future benefits, some of the states we visited have additional overpayment recovery tools for individuals who are no longer receiving UI. These tools include state tax refund offset, wage garnishment, and use of private collection agencies.^[19] Some of these procedures, such as the state tax refund offset, are viewed as particularly effective. For example, one state reported overpayment collections of about $11 million annually between 1998 and 2000 resulting from this process. Other states have increased overpayment collections by allowing more aggressive criminal penalties for individuals who are suspected of UI fraud. For example, one state prosecutes UI fraud cases that exceed a minimum threshold as felonies instead of misdemeanors. Officials in this state told us that the threat of imprisonment often encourages claimants suspected of fraud to make restitution for UI overpayments. According to state officials, this initiative resulted in $37 million in additional overpayment collections in calendar years 2000 and 2001. However, other states we visited lacked many of these tools. For example, one state relied primarily on offsets against current UI claims to recover overpayments because its laws and policies did not permit the use of many of the tools that other states have found to be effective for collecting overpayments from individuals who have left the UI rolls.

Labor’s Management Places Insufficient Emphasis on Program Integrity

In general, Labor’s approach to managing the UI program has emphasized quickly processing and paying UI claims, with only limited attention to overpayment prevention, detection, and collection. This approach is most evident in the priorities that are emphasized in Labor’s recent annual performance plans, the UI program’s performance measurement system, and the limited use of quality assurance data to correct vulnerabilities in states’ UI operations. For example, Labor’s recent annual performance plans required under the Government Performance and Results Act of 1993 have not included strategies or goals to improve payment accuracy in state UI programs. In addition, we found that Labor’s system for measuring and improving UI program performance is primarily geared to assess the timeliness of various state operations. ^[20] Most of the first 12 performance measures (called “Tier I”) assess whether states meet specified timeframes for certain activities, such as the percentage of first payments made to claimants within 14 to 35 days. However, none of the Tier I measures gauge the accuracy of UI payments. Labor also gives Tier I measures more weight than the remaining measures (called “Tier II”), which assess other aspects of state performance, including overpayment collections. Labor has developed national criteria specifying the minimum acceptable level of performance for most Tier I measures.^[21] States that fail to meet the minimum established criteria are generally required to submit a “Corrective Action Plan” to Labor. Moreover, Labor has indicated that it may withhold the administrative funding of states that continually do not meet Tier I performance goals. By contrast, the Tier II measures do not have national minimum performance criteria and are generally not enforced as strictly by Labor.  Labor could set Tier II criteria on a state-by-state basis and withhold funding in case of subsequent noncompliance.

Officials from most of the states we visited also told us that the Tier I and Tier II measures make the UI program complex to administer and may contribute to an environment in which overpayments are more likely. In particular, these officials told us that because the measures are so numerous and are designed to monitor a wide range of activities, it is difficult to place sufficient emphasis on more fundamental management issues, such as payment accuracy. There are currently more than 70 Tier I and Tier II measures that gauge how states perform in terms of the timeliness, quality, and accuracy of benefit decisions. Faced with competing priorities, some states tend to focus most of their staff and resources on meeting certain measures such as payment timeliness, but may neglect other activities such as those dealing with program integrity.

We believe, however, that Labor can do more to encourage states to balance payment timeliness with the need for payment accuracy in a manner that does not require the complete withholding of administrative funds. For example, under federal regulations covering funds to states, Labor may temporarily withhold cash payments, disallow costs, or terminate part of a state’s administrative funding due to noncompliance with grant agreements or statutes.^[22] Withholding or delaying a portion of these funds is one way Labor can potentially persuade states to implement basic payment control policies and procedures. In addition, while completing the annual budget process, Labor could prioritize additional administrative funding to states to help them achieve or surpass agreed upon payment accuracy performance levels. ^[23] However, we found that Labor is only using such tools to a limited degree to help states enhance their program integrity activities.

Labor Has Not Fully Utilized Its Quality Assurance Data to Improve State Operations

Labor has also been reluctant to use its quality assurance data as a management tool to encourage states to place greater emphasis on program integrity. According to an internal agency performance report and Labor officials, quality assurance data should be used to identify vulnerabilities in state program operations, measure the effectiveness of efforts to address these vulnerabilities, and help states develop mechanisms that prevent overpayments from occurring. ^[24] However, as currently administered, Labor’s quality assurance system does not achieve all of these objectives. In particular, Labor lacks an effective mechanism to link its quality assurance data with specific improvements that are needed in states’ operations. For example, over the last decade, payment errors due to unreported income have consistently represented between 20 and 30 percent of annual UI overpayments. While Labor’s quality assurance system has repeatedly identified income reporting as a vulnerable area, it has not always played an active role in helping states develop specific strategies for improving their performance in this area. Of particular concern to us is that the overpayment rate for the nation has shown little improvement over the last 10 years. This suggests that Labor and some of the states are not adequately using quality assurance data to address program policies and procedures that allow overpayments to occur.

Labor Gives Inadequate Attention to Overpayment Recoveries

Finally, Labor has given limited attention to overpayment collections. Currently, Labor requires states to collect at least 55 percent of all the overpayments they establish annually through their benefit payment control operations. This 55 percent performance target has not been modified since 1979 despite advancements in technology over the last decade, such as automatic state tax refund intercepts, that could make overpayment recovery more efficient. At the time of our review, only 34 out of 53 states met or exceeded the minimum standard of 55 percent. A small number of federal and state officials told us that states tend to devote the minimum possible resources to meet it each year. However, our work shows that Labor has not actively sought to improve overpayment collections by requiring states to incrementally increase the percentage of overpayments they recover each year.

Labor is Taking Actions To Improve Program Integrity

At the time of our review, Labor was continuing to implement a series of actions to help states with the administration of their UI programs. For example, Labor is helping states use the Information Technology Support Center (ITSC) as a resource for states to obtain technical information and best practices for administering their UI programs.^[25]

Labor also provides technical assistance and training for state personnel, as well as coordination and support for periodic program integrity conferences. In its annual budget justification, Labor has requested a limited amount of funding for the states for program integrity purposes, such as $35 million in fiscal year 2001 for states to improve benefit overpayment detection and collection, eligibility reviews, and field tax audits. More recently, Labor has been developing a new payment accuracy indicator in its Annual Performance Plan for fiscal year 2003 for the states’ UI programs that will establish a baseline measurement for benefit payment accuracy during 2002. Labor also plans to provide states with additional quality assurance data on the nature and cause of overpayments to help them better target areas of vulnerability and identify more effective means of preventing overpayments.

At the time of our review, Labor was also developing a legislative proposal to give state employment security agencies access to the NDNH to verify UI claimants’ employment and benefit status in other states. Our analysis suggests that use of this data source could potentially help states reduce their exposure to overpayments. For example, if the directory had been used by all states to detect claimants’ unreported or underreported income, it could have helped prevent or detect hundreds of millions of dollars in overpayments in 2001 alone.^[26] In addition, Labor is working to develop an agreement with the Social Security Administration that would grant states access to SSA’s SOLQ system. States that used this system would be able to more quickly validate the accuracy of each claimant’s SSN and identity at the time of application for UI benefits.

Conclusions

Despite the various efforts by Labor and some states to improve the integrity of the UI program, problems still exist. The vulnerabilities that we have identified are partly attributable to a management approach in Labor and many states that emphasizes quickly processing and paying UI claims without a similar focus on controlling program payments. While we recognize the importance of paying unemployed individuals in a timely manner, this approach has likely contributed to the consistently high level of overpayments over time, and as such, may have increased the burden placed on some state UI trust funds. As the number of UI claimants has risen over the last year, many states have felt pressured to quickly process and pay additional claims. The results of our work suggest that, in this environment, overpayments are not likely to abate and could increase.

Labor is taking some steps to improve UI program integrity by helping enhance existing state operations, such as working to obtain access to important data sources. Our prior work suggests that using more front-end automated data sources to verify claimant eligibility before overpayments are made is a more efficient method of protecting program funds than trying to recover overpayments after they have occurred. In the case of the UI program, access to data sources such as the NDNH or WRIS could help states reduce overpayments caused by unreported income, which accounts for more than one-third of the overpayments in 2001. However, absent a change in the current approach to managing the UI program at both the federal and state level, it is unlikely that the deficiencies we identified will be sufficiently addressed. In particular, without more active involvement from Labor in emphasizing the need to balance payment timeliness with payment accuracy, states may be reluctant to implement needed changes in their management philosophy and operations. With increased emphasis on payment accuracy, Labor’s system of performance measures could help encourage states to place a higher priority on program integrity activities. Moreover, an effective strategy to help states control benefit payments will also require use of its quality assurance data to identify areas for improvement and work with the states to implement changes to policies and procedures that allow overpayments to occur. However, Labor must be willing to link state performance in the area of program integrity to tangible incentives and disincentives, such as through the annual administrative funding process. Ultimately, a coordinated effort between Labor and the states is needed to address the weaknesses we have identified and reduce the program’s exposure to improper payments. Without such an effort, Labor risks continuing the policies and procedures that have contributed to consistently high levels of UI overpayments over the last decade.

Mr. Chairman, this concludes my prepared statement. I will be happy to respond to any questions you or other members of the Subcommittee may have.

GAO Contacts and Staff Acknowledgements

For information regarding this testimony, please contact Sigurd R. Nilsen, Director, Education, Workforce, and Income Security Issues, at (202) 512-7215. Individuals who made key contributions to this testimony include Daniel Bertoni, Jeremy Cox, Cheryn Powell, and Salvatore Sorbello.

Related GAO Products

Workforce Investment Act: Improvements Needed in Performance Measures to Provide a More Accurate Picture of WIA’s Effectiveness. GAO-02-275 Washington, D.C.: February1, 2002.

Strategies to Manage Improper Payments: Learning from Public and Private Sector Organizations. GAO-02-69G. Washington, D.C.: October 2001.

Department of Labor: Status of Achieving Key Outcomes and Addressing Major Management Challenges. GAO-01-779. Washington, D.C.: June 15, 2001.

Unemployment Insurance: Role as Safety Net for Low-Wage Workers is Limited. GAO-01-181. Washington, D.C.: December 29, 2000.

Benefit and Loan Programs: Improved Data Sharing Could Enhance Program Integrity. GAO/HEHS-00-119. Washington, D.C.: September 13, 2000.

Supplemental Security Income: Action Needed on Long-Standing Problems Affecting Program Integrity. GAO/HEHS-98-158. Washington, D.C.: September14, 1998.

Supplemental Security Income: Opportunities Exist for Improving Payment Accuracy. GAO/HEHS-98-75. Washington, D.C.: March 27, 1998.

Supplemental Security Income: Administrative and Program Savings Possible by Directly Accessing State Data. GAO/HEHS-96-163. Washington, D.C.: August 29, 1996.

^[1] We also interviewed the Utah UI Director by telephone because this state has been utilizing some practices that other states could use to verify claimants’ eligibility for UI benefits, such as on-line access to the Social Security Administration’s State Online Query system to verify the validity of individuals’ social security numbers.

^[2] These estimates are based on preliminary data from Labor available at the time of our review.

^[3] We have not yet been able to obtain data on confidence intervals, so we are unsure of the precision of these estimates.

^[4] Several state officials told us that the number of UI claims have increased since the terrorist attacks of September 11, 2001, and have forced them to move staff resources from benefit payment control or benefit accuracy measurement activities into claims taking positions.

^[5] Each state is required to maintain a database of individuals who were recently hired to help state child support enforcement agencies locate non-custodial parents who owe child support payments.

^[6] All states were required to create a state directory of newly hired employees as part of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. Each state’s directory periodically reports state unemployment insurance, wage and new hires data to the National Directory of New Hires for purposes of locating noncustodial parents in other states who owe child support payments.

^[7] See the U.S. Department of Labor, Office of Inspector General, Unemployment Insurance Integrity: Fraud and Vulnerabilities in the System (1P-03-315-0001-PE) March 31, 1999.

^[8] This match is conducted using Labor’s Interstate Connection Network.

^[9] See 42 U.S.C. 653 (l).

^[10] WRIS helps participating states track the employment status of individuals who have participated in WIA job training programs in other states.

^[11] Labor agreed to fund WRIS for the first year of its operation, but has not committed to funding future years. The estimated annual cost of administering the system is $2 million. See Workforce Investment Act: Improvements Needed in Performance Measures to Provide a More Accurate Picture of WIA’s Effectiveness, GAO‑02‑275, (Washington, D.C.: Feb. 1, 2002).

^[12] State laws differ from one another in terms of how benefits that are received from other federal or state programs affect claimants’ eligibility for UI benefits.

^[13] The Social Security Administration is responsible for administering programs including the Old Age and Survivors Insurance, Supplemental Security Income, and Disability Insurance .

^[14] Although some categories of noncitizens may be eligible for UI benefits, such as those authorized to work in the United States at the time they apply for benefits, others, including illegal aliens, are not. See Federal Unemployment Tax Act 3304 Section (a)(14)(A).

^[15] See Department of Labor Office of Inspector General, Verification of Social Security Numbers Could Prevent Unemployment Insurance Payments to Illegal Aliens, 04-98-001-03-315, March 2, 1998.

^[16] States report sending SSNs to SSA for verification in intervals ranging from daily to once per quarter (every 3 months).

^[17] See U.S. Department of Labor, Office of Inspector General, Examination of UI Benefit/Wage Crossmatch and Analysis of Employers Who Fail to Respond to the States’ Requests for Weekly Wage Data (05-99-005-03-315) March 1999.

^[18] Labor recently funded a grant to one state to facilitate more effective coordination and cooperation between the state and its employers. As a result of its actions, this state reported that about 80 percent of the state’s employers comply with state requests for information in a timely manner.

^[19] For UI claimants who have outstanding overpayments, the state tax refund offset allows a state to intercept the individual’s state tax refund to recover an overpayment; wage garnishment allows the state to recover UI overpayments from an individual’s paycheck when they return to work; and private collection agencies can pursue overpayments when the state has been unsuccessful in recovering using its existing collection procedures.

^[20] This system, called “UI Performs,” was developed with input and coordination from the states. The system incorporates more than 70 performance measures to gauge states’ performance, including the timeliness, quality, and accuracy of benefit decisions.

^[21] The national minimum performance criteria are performance measures that are applied uniformly to all states.

^[22] See 29 C.F.R. 97.43.

^[23] See 20 C.F.R. 601.6.

^[24] See Department of Labor, Employment and Training Administration, UI Performs 2000 Annual Report, p.9.

^[25] ITSC is a collaborative effort involving the Department of Labor, state employment security agencies, private sector organizations, and the state of Maryland. It was created in 1994 to help states adopt more efficient, timely, and cost-effective service for their unemployment service claimants.

^[26] This assumes that the top two categories of overpayments (“benefit year earnings” and “base period wages”) were substantially reduced or eliminated by use of the NDNH.