What states can expect with the new SNAP match: Options to reduce state error rates

Summary:

• The food stamp error rate is far too high. Between 2013 and 2024, the food stamp error rate more than tripled from 3.2 percent to 10.93 percent.
• The One Big Beautiful Bill Act requires states to have skin in the game: if they continue to have high error rates, they will be responsible for a small percentage of their food stamp benefits.
• Based on their 2024 error rate, eight states will not see any match, but most states will begin paying a state match in 2028.
• To reduce their error rates, states should act now to reform their eligibility practices.

---

Without dramatic changes, many states will soon pay millions of dollars out of their state budgets to provide food assistance to their citizens. The One Big Beautiful Bill Act revolutionized the federal-state responsibility for the Supplemental Nutrition Assistance Program (SNAP). Unlike Medicaid, in which the federal government and state governments share spending obligations to cover benefit costs, the federal government has covered 100 percent of SNAP benefits since 1965. This has created a perverse incentive structure for states to be less diligent in addressing waste, fraud, and abuse. Now states will have new skin in the game: if they continue to have high error rates, they will be responsible for a small portion of food stamp benefit costs for their state citizens.

The One Big Beautiful Bill Act (OBBBA) changed the federal-state share in food stamps because the payment error rate has become unreasonably high. The error rate reached a low of 3.2 percent in Fiscal Year 2013 after 15 years of continuous improvement. In 2019, the food stamp error rate more than doubled the low rate, with 7.36 percent of every dollar being sent out the door incorrectly. After two years without measuring the error rate because of COVID, the error rate reached 11.54 percent in 2022 and remains nearly as high at 10.93 percent as of 2024—more than three times higher than its low rate in FY2013. Nearly one in every nine dollars spent by SNAP—over $10.2 billion—does not reach the intended beneficiary. Prior to the pandemic, erroneous SNAP payments never exceeded $4.2 billion. Figure 1 shows SNAP error rates since FY1997.

Why Have Error Rates Increased?

The Food and Nutrition Service (FNS) of the U.S. Department of Agriculture has allowed states flexibility to streamline their eligibility processes, but this “streamlining” often means not fully investigating whether an applicant is eligible or accurately assessing the precise benefits for which they qualify. As states started making use of FNS flexibilities in administering the program, this led to higher error rates as they allowed individuals to self-attest income, simplified reporting, and lengthened certification periods. 

There was also a severe error increase after COVID. In 2020, FNS introduced significantly greater leniencies regarding eligibility and accuracy, including extended certification periods, waiving of periodic reporting, and waiving face-to-face interviews, etc. These policies increased fraud by making it easier to access benefits without proof of eligibility. Ineligible recipients often remained on the benefit for up to 48 months. At the same time, many states responded to COVID shutdown policies by allowing SNAP EBT cards to be used for online food purchasing, which opened up benefits to new avenues for fraud (for instance, as of 2025, the majority of states did not have EBT chip security, which has been widespread in the private industry for years). In addition, FNS didn’t release the error rates in 2020 or 2021, so the extent of the fraud remained hidden.

As benefits have grown more generous, it became more likely for some benefits to be sent in error. Food stamps cost federal taxpayers $60.3 billion in 2019 and reached $100 billion in 2024—mostly due to benefit increases. Congress authorized a temporary 15 percent benefit increase in 2020. After that ended, the Biden administration increased benefits by 23 percent through the Thrifty Food Plan.

FNS did have options to address states with high error rates. States with higher error rates would be required to submit corrective action plans. FNS also had the option to fine states if their error rate was above the national average for two consecutive years. However, if the states opted to reinvest half of the fine to reduce the cause of the error, the other half of the fine would be held “at-risk” if the error rate does not subside. Despite these tools, the error rate has continued to climb.

Figure 1: Error Rate Graph

OBBBA Changes to State Responsibility

The bill makes three major changes to rein in the SNAP error rate:

New State Match. States will have a state match in FY 2028,* anchored to the state’s payment error rate if the state’s error rate equals or exceeds 6 percent. (Sec. 10105)

• Below six percent error rates: zero percent match.
• Six percent or above but below eight percent error rate: five percent match.
• Eight percent or above but below 10 percent error rates: 10 percent match.
• Above 10 percent error rate: 15 percent match.

*However, the state match will be delayed one or two years, to FY2029 or FY2030 if the error is equal to or above 13 1/3 percent in FY2025 or FY2026, respectively.

Lower Administrative Match. The OBBBA also lowers the federal match administrative costs from 50 percent to 25 percent starting in FY2027.* (Sec. 10106)

*For several states, the match may not be implemented until 2029. The statute contains a provision that if the FY2025 payment error rate of a state is multiplied by 1.5 is equal to or above 20 percent, the match will not be implemented until FY2029. The same applies for the FY2026 error rate for a delay until 2030.

While eight states (Idaho, Nebraska, Nevada, South Dakota, Utah, Vermont, Wisconsin, and Wyoming) would not face a state match based on their 2024 error rate, most states will begin paying a state match in 2028 if their rates would remain the same. Another 10 states would have delayed cost sharing in FY2029 or FY2030, if their rates stay the same. These annual costs could hit as high as $1.8 billion for California, but would be significantly less for smaller states, especially if matching costs are five percent, such as $6 million for North Dakota. (See below for a 50-state projection of new state matches based on current error rates).

While 42 states and the District of Columbia currently have error rates above the six percent threshold where a cost sharing would occur, just four states and D.C. had error rates above that threshold in FY2013, indicating that lowering payment error rates should be feasible for all states.

State Options to lower error rates

Because the federal government has traditionally covered 100 percent of benefit costs, but only 50 percent of administrative costs, states had an incentive to reduce case-working hours and data checks that cost state dollars. This led many states to use options that can lead to higher error rates. With the advent of steep financial penalties for ongoing error rates, however, states face a financial incentive to end those flexibilities and ensure accurate benefit distribution. To reform their SNAP program, states should:

1. Cross-Check Applicant Income and Deduction Information with In-State or Third-Party Wage Data. FNS does require verification of gross income but does not require complete verification for net income. Current guidance tells states, “best practice allows the state agency an opportunity to resolve potential inconsistencies without requiring additional contact or additional verification.” Despite this, states retain the option to ensure net income accuracy by using in-state wage data or third-party payroll data to verify all applications.
   
2. Reduce Recertification Periods. For non-elderly or disabled individuals, states can set a reasonable six-month recertification period. This ensures that states are closely monitoring eligibility, but it also gives the state the option to offer assistance like employment and training programs to non-disabled households who continue to report zero or extremely low income.
   
3. Repeal Broad-Based Categorical Eligibility. The SNAP statute allows states to automatically enroll individuals who receive benefits from Supplemental Security Income (SSI) and Temporary Assistance for Needy Families (TANF), programs with similar eligibility standards. However, in 2009, FNS issued guidance to create an administrative option called “broad-based categorical eligibility” (BBCE), where “benefit” is defined so broadly that it includes the online distribution of pamphlets and 1-800 numbers. This definition allows states to eliminate the SNAP asset limit and raise to gross income limit up to 200 percent of the Federal Poverty Level (FPL). Forty-five states now opt into BBCE, though some simply adjust the statutory asset limit.
   
4. End Simplified Reporting. Simplified reporting allows states to maintain a recipient’s benefits at the same level, without checking income or deduction changes, until recertification. Recipients only need to report increases in income if they exceed 130 percent FPL. FNS first created simplified reporting as a state option via regulation in 2000 for recipients with earnings, and it was then expanded as an option for all recipients in the 2002 Farm Bill. States generally found that shorter certification periods increased benefit accuracy, but likewise increased administrative effort and, thus, state costs. Now that they face steep penalties for continued inaccuracy, this math likely changes for many of them.

States Need More Tools in The TOolbox

While state offices and local administrators are the front line in reducing error rates, their federal partners can open additional tools to assist them, including:

1. State Access to Federal Data Sources. FNS can enter Memorandums of Understanding for state SNAP eligibility systems to utilize data system maintained by federal partners, particularly the Department of Health and Human Services and the Department of Labor. These can include the Public Assistance Reporting Information System (PARIS) and National Directory of New Hires (NDNH), and key federal data hubs, like social security data through the Social Security Administration.
   
2. Amend SNAP “Right to File.” FNS requires states to process any application that contains a name, address, and signature, regardless of whether the application is incomplete. States must also issue benefits within 30 days of receipt of even an incomplete application. With the rise of artificial intelligence, states have witnessed a dramatic rise in bots submitting applications. “Right to File” drives up fraudulent payments because state must spend time processing incomplete applications and issuing benefits when they cannot complete processing in the allotted timeframe. USDA can issue guidance for state options to validate application humanity—for instance, proper identification before the submission of an application.
   
3. Revise Merit Guidance. Unlike many federal programs, SNAP statute does require that merit system employees (generally understood as state agency employees) must certify applications. FNS has issued guidance to clarify that states may request FNS approval to pursue options like call center support. FNS can update this guidance to clarify that state employees may use technology to authorize or disapprove cases. 

Conclusion

The recent federal legislation reset the federal-state food stamp relationship. With additional skin-in-the-game, states will see increased incentive to drive down errors rates and ensure that every taxpayer dollar reaches those in need. This new relationship also means that states can drive reform in this essential safety net program to ensure that all recipients find self-sufficiency.