Chairman Sessions, Ranking Member Mfume, and members of the Government Operations Subcommittee of the Committee on Oversight and Government Reform, thank you for today’s opportunity to testify. My name is Jennifer Wagner. I am the Director of Medicaid Eligibility and Enrollment at the Center on Budget and Policy Priorities, a nonprofit, nonpartisan policy institute located here in Washington, D.C. CBPP conducts research and analysis on a range of federal and state policy issues affecting low- and moderate-income families.
I am honored to be here today to talk about something I have spent over 20 years working on — improving government systems to get the right benefits to eligible people in a streamlined and accurate way. I started my career as a food stamp caseworker in North Carolina, was a state advocate, oversaw policy and operations for the Illinois Department of Human Services, and now work on eligibility and enrollment issues. I am eager to discuss available solutions to reduce improper payments and how program cuts and barriers to benefits should not be confused with targeted proposals to improve program integrity.
Let’s start with the facts. The vast majority of people enrolled in health and economic security programs are eligible for those benefits and services. Most improper payments are paperwork issues, not fraud.
Further, increasing red tape doesn’t improve accuracy. Nor will cutting programs; reducing improper payments requires more resources for states and the federal agencies overseeing these programs. States need sufficient staff to make accurate eligibility determinations and implement IT solutions, while the federal agencies need staff to provide technical assistance and oversight.
Notably, program integrity measures don’t consider those who are eligible but unable to enroll due to confusing notices, unnecessary requests for documentation, or inaccurate application of eligibility rules. When measuring program integrity and discussing solutions to improper payments, we must also consider eligible people who aren’t able to enroll and those who inaccurately lose coverage because of barriers in the eligibility and enrollment process.
It’s important to draw a clear distinction between improper payments and fraud. The vast majority of people enrolled in health and economic security programs are eligible for those benefits and services. Most improper payments are paperwork issues or honest mistakes by agency staff or beneficiaries navigating complex rules. Fraud, on the other hand, involves getting something of value through willful misrepresentation. Fraud does exist in health and economic security programs like Medicaid, the Supplemental Nutrition Assistance Program (SNAP), the Temporary Assistance for Needy Families program (TANF), and unemployment insurance (UI), but most of it is not beneficiary fraud.
In Medicaid, most fraud is committed by a small number of providers engaging in criminal behavior to scam money from the program. For example, a business owner in North Carolina was sentenced to 200 months in prison for scamming Medicaid out of $11 million by performing large amounts of medically unnecessary tests.[1] Beneficiary fraud has been found to be negligible in reports issued by the Department of Justice, the Department of Health and Human Services’s Office of the Inspector General, and others outlining fraud prosecutions and settlements, while ambulance service providers, durable medical equipment suppliers, and other providers have been found to account for the vast majority of fraud.[2]
In SNAP, the most egregious instances of fraud involve criminals stealing electronic benefit transfer, or EBT, card information and stealing benefits off these cards. This “skimming” leaves participants without funds to buy food for their families, since the program no longer restores benefits for people whose funds were stolen.
The most high-profile case of fraud in the TANF program did not revolve around recipients of assistance from the program, but rather state officials and contractors who were administering it. A 2020 audit of Mississippi’s TANF spending found tens of millions of dollars were egregiously misspent, including on the construction of a volleyball stadium and payments to a pharmaceutical startup.[3] Seven people have pled guilty in schemes related to fraudulent spending, including the state’s Director of Human Services.
Finally, pandemic unemployment insurance programs, which kept about 5 million people from falling into poverty in 2020 and provided vital support to the economy, were vulnerable to fraud due to years of administrative neglect and because new programs had to be established rapidly given the regular UI system’s very limited scope and responsiveness. And of course, any benefit system would struggle administratively when claims rise by roughly 3,000 percent in a very short time, as happened with UI in the spring of 2020. In September 2020, the Justice Department warned: “Fraudsters, some of which are transnational criminal organizations, are using stolen identities of U.S. citizens to open accounts and file fraudulent claims for UI benefits,” highlighting that much of the fraud was due to criminal rings and not UI claimants.[4]
Improper payments, on the other hand, are primarily the result of administrative errors or missing paperwork. When evaluating proposals to make significant changes to programs like SNAP and Medicaid in the name of “waste, fraud, and abuse,” it is important to understand how improper payments are measured in these programs.
The Medicaid Payment Error Rate Measurement, known as PERM, measures payments that do not meet Medicaid’s statutory, regulatory, or administrative requirements. In most cases, the error results from an administrative or paperwork issue. It is not a measure of fraud, and does not indicate the Medicaid enrollee was not eligible for the coverage they received.[5] There are three components of PERM: the fee-for-service rate, the managed care rate, and the eligibility rate. The overall rate in 2024 was 5.09 percent, down from 8.58 percent in 2023. The rate specific to eligibility determinations was 3.31 percent, down from 5.95 percent in 2023.[6]
The eligibility component of PERM measures whether a state or local eligibility worker (and eligibility system) followed state and federal requirements in determining eligibility. The leading cause of these errors is insufficient documentation, resulting in the PERM reviewer not being able to evaluate all of the information the eligibility worker used in making the determination. For example, an applicant may have attested to their income on their application. The eligibility worker then checked electronic income data sources and used that to verify the client statement. However, the worker may not have saved a record of the electronic data that was used in the eligibility system. So when a reviewer looks at the case, they can’t confirm that the worker’s assessment was correct. There is no reason to believe that the applicant was not in fact eligible, but since the paperwork is missing, it’s considered an improper payment. This is a technical systems and training problem, not a fraud problem.
SNAP has one of the most rigorous quality control systems of any federal program, with extensive upfront and ongoing eligibility checks. Households applying for SNAP report their income and other relevant information. A state eligibility worker interviews a household member and verifies the accuracy of the information through data matches and by collecting documentation from the household. Households must reapply for benefits periodically, usually every six or 12 months, and between reapplications must report income changes that would affect their eligibility or benefit level. Then separate state Quality Control (QC) workers review a statistically valid sample of SNAP cases to determine whether the eligibility and benefit determination was correct. About a third of these cases are re-reviewed by a federal QC team.[7]
The state SNAP error rates that this system produces were modestly higher in 2022 and 2023 (overpayments were about 10 percent), in part due to temporary pandemic-related administrative flexibilities, which were largely still in place throughout 2022 and most of 2023. These flexibilities helped states manage the challenges of the pandemic and successfully prevented a spike in food insecurity but may have contributed to higher error rates.
SNAP’s error rate does not indicate fraud. The overwhelming majority of payment errors result from honest mistakes by participants, eligibility workers, data entry clerks, or computer programmers. In fact, the Department of Agriculture reported that more than half of the dollar amount of errors in fiscal year 2022 were the fault of the state agency rather than recipients.[8] Much of the remaining overpayments resulted from innocent errors by households who had difficulty navigating a program with complex rules. And households typically must repay overpaid benefits, even if the overpayment was the agency’s mistake.
One shortcoming of improper payment measures in programs like Medicaid and SNAP is that they measure improper payments only for those who receive benefits. They don’t consider those who are eligible but were not able to enroll due to confusing notices, unnecessary requests for documentation, or inaccurate application of eligibility rules. Discussions about improper payments must be part of a larger discussion about program integrity that includes making it easier for people who are eligible to receive benefits. Otherwise we create solutions that erect barriers for eligible people seeking to access programs and we decrease accuracy.
We should keep this broader definition of program integrity in mind when evaluating proposed solutions to improper payments, since many of the solutions being discussed in Congress would in fact worsen program integrity. In general, increasing administrative burden and red tape through increasingly complex policy raises the chance for errors.[9] Understaffed state agencies are often unable to keep up with additional requirements and make mistakes, while program participants don’t always understand what they have to do to maintain their benefits. As a result, errors go up and eligible families incorrectly lose benefits.
One idea currently under consideration that would increase errors is requiring that states conduct renewals for Medicaid enrollees every six months, rather than every year as is the case currently for most enrollees. That would cause more eligible people to inappropriately lose coverage, as we saw when states “unwound” the pandemic-era Medicaid continuous coverage provision and millions of enrollees lost coverage for procedural reasons — not because they were determined to no longer qualify for Medicaid, but because they didn’t receive a notice, didn’t understand what they needed to do to keep their coverage, didn’t return documents on time, or the Medicaid agency didn’t process the paperwork in a timely way.[10]
Other proposals that are currently in the news would add red tape and would increase the likelihood of errors and the risk that eligible people would lose coverage. Taking benefits away from people who don’t meet work requirements would require new complex policy related to exemptions, countable hours and activities, and good cause reasons for non-compliance. This would increase improper payments resulting from already overworked eligibility workers not following the correct complex procedures and documentation requirements, and lead to erroneous loss of coverage for enrollees who don’t understand what they need to do to comply with the new requirements. Studies have shown that work requirements don’t increase employment but do cause eligible people to lose coverage.[11]
Another approach that appears to be under consideration is substantially changing the funding structures of Medicaid and SNAP, which would shift costs to the states. This is unlikely to drive greater efficiency, but is likely to drive states to dramatically reduce services offered and/or populations covered, as we saw with the TANF program when funding was block granted. And some of these proposals would reduce funding for staff and IT systems, resulting in fewer resources to accurately determine eligibility and fix issues identified through audits.
The TANF fraud in Mississippi points to a larger problem that may have increased the possibility of large-scale fraud, and that certainly diminishes TANF’s ability to help families living in poverty. In 2022, Mississippi spent about $59 million in federal and state TANF funds, with only 7 percent of these funds going toward basic cash assistance.[12]
This partly explains why, in 2020, only about 1 in 25 families experiencing poverty in Mississippi received TANF cash assistance, down from almost 4 in 10 families in poverty receiving such aid at the time TANF was implemented in 1996.[13] And for those who did receive TANF assistance in 2020, the monthly benefit equaled only $260 for a family of three. Mississippi spends much greater portions of its TANF funds for other purposes, including 19 percent for child welfare services, 41 percent for work and education activities, and 25 percent on other services.
The solution to the problem of TANF’s limited reach is certainly not to cut TANF funding, which has already lost 50 percent of its real, inflation-adjusted value since it was established. Instead, reforms should be put in place to ensure that states spend a meaningful portion of their TANF funds on the core purpose of the program — providing basic cash assistance to needy families and helping them find a path to employment.
Reducing improper payments actually requires more resources for states and the federal agencies overseeing these programs, not less. These programs serving millions of people are incredibly complex to administer, and state and local agencies often lack the time and resources to tackle technical programs as they juggle staffing shortages, budget pressures, and increased responsibilities. Federal agencies like the Centers for Medicare & Medicaid Services (CMS) and the Food and Nutrition Service can do more to support states in improving their systems, but they need sufficient staff to do so. This includes key roles such as inspectors general to review eligibility determinations, policy staff to help states sort through complicated rules, and tech expertise to assist with IT system development and maintenance.[14]
The Government Accountability Office identified two common strategies federal agencies have used to successfully reduce improper payment rates: establishing accountability and facilitating internal collaboration; and providing technology, tools, and training targeted to root causes.[15] For instance, federal agencies need sufficient staff with the right expertise to meaningfully engage with states when they submit a corrective action plan to address high rates of improper payments. This is necessary to leverage the audit findings to make meaningful changes and hold states accountable for addressing shortcomings in their processes.
A great example of federal agency engagement in making meaningful improvements to state agency operations happened during the “unwinding” of the Medicaid continuous coverage provision when states had to renew the eligibility of all of their enrollees after a pause in terminations during the pandemic. Federal officials identified states were not using a required process known as ex parte renewals that would keep eligible people enrolled while greatly decreasing agency workload. Ex parte renewals use available data to confirm ongoing eligibility without requiring enrollees to fill out and submit complex paperwork and are often automated in the eligibility system.[16]
CMS published detailed guidance and offered waivers and technical assistance to states to improve their ex parte rates. But they went further and partnered with U.S. Digital Services (USDS) to send teams of experts into states to help them quickly modify their eligibility systems and improve their rates, resulting in more than 5 million people renewed for health coverage with less red tape in 2024.[17] CBPP supported USDS in this effort, and worked with Code for America to do similar work improving ex parte rates in other states, reducing processing time for renewals in one state from 70 minutes to 11 minutes per case.[18]
The federal government also took significant steps to reduce UI fraud during and after the pandemic, including by sending “tiger teams” of experts to help states strengthen identity verification and improve the administration of their UI programs. With these efforts, and the expiration of the pandemic programs, estimated improper unemployment insurance payments fell by over 92 percent from fiscal year 2021 to fiscal year 2024 (from $78 billion to $6 billion), and the improper payment rate has declined from 19 percent to 16 percent.[19]
Further improvements could be made by ensuring states have sufficient and consistent funds to properly administer their UI programs. Additionally, Senators Wyden and Crapo put forward bipartisan legislation last year that includes a number of positive proposals to improve the administration of UI, including detecting and preventing fraud.[20]
Other program areas have similar opportunities to reduce improper payments, improve client experiences, and reduce burden on state agency staff. Income verification is one of the most important, yet most difficult, parts of the eligibility determination process. An accurate and efficient way to verify income is to use state, federal, and commercial data sources, such as state quarterly wage data from the state workforce agency. Yet, the rules around which program can use which data source and in what way are very complicated which, combined with complex IT systems and data-sharing agreements, can make it difficult for states to maximize use of these data sources when verifying income. Instead, workers may default to requiring clients to send in paystubs, which delays the processing of benefits and increases the chance of something going wrong in the process and the denial of benefits to an eligible household. Federal support through detailed guidance and on-the-ground implementation support could lead to more timely and accurate eligibility determinations.
Incredibly complex eligibility systems are the source of many challenges in eligibility determinations for these programs. While IT vendors manage the systems in most states, state staff must design the systems and oversee the vendors. States often lack enough staff with the right skill sets for this task. An innovative approach to improve the systems clients and eligibility workers interact with is the creation of state digital service teams with expertise in digital government policy, agile product management, and user-centered research and design.[21] Expanding the use of teams like these can improve both customer experience and the accuracy of benefit determinations.
Another solution to reduce improper payments in Medicaid is already underway as part of the eligibility and enrollment rule that was finalized in April 2024. It updates outdated and unclear regulations on maintenance of case records. The revised rule specifies the types of information Medicaid agencies must keep in their records, specifies a retention period, and requires states to store this information electronically. This should substantially reduce the frequency of improper payments resulting from inadequate records.
It's not an easy task to address improper payments in large programs serving millions of people administered by federal, state, and local governments. Proposals to dramatically restructure program funding or rules are unlikely to address improper payments, especially since most improper payments are paperwork issues rather than fraud. But carefully tailored approaches can bring us closer to the goal of getting the right benefits to eligible people timely and accurately.
Thank you for the opportunity to testify. I look forward to responding to your questions.
