Abstract

Supplemental Nutrition Assistance Program (SNAP) caseloads declined by 10% nationally from June 2025 to April 2026, even as the unemployment rate ticked up. Recent commentary has attributed the caseload decline to reforms to SNAP enacted in July 2025, especially a provision that ties required state funding contributions to their payment error rates. I first show how this reform shapes state incentives regarding enrollment and benefit amount decisions: States have an incentive to enroll households if their expected payment error as a share of their benefit payment is less than the state’s current payment error rate. Additionally, overpaying SNAP benefits to a household does less to increase the payment error rate than underpaying by the same amount, and modest overpayments reduce the payment error rate more than paying households their accurate benefit. Next, I exploit provisions in the new law that in its initial years create different incentives for payment error rate reductions across states. Using a difference-in-differences style event study framework, I do not find an association between states’ varying incentives to reduce payment error rates and their caseload declines. Though states may have responded to the cost-sharing reform, the lack of a differential response suggests caution in drawing conclusions about the drivers of the SNAP caseload decline.

Read full paper here