State Variation in SNAP Performance: Trade-offs and Drivers Across Accuracy, Access, and Timeliness

• In Event: SNAP Program Design: Implications for Participants and Policy

Abstract

The U.S. Department of Agriculture (USDA) has statutorily designated four metrics as annual measures of state performance in administering SNAP, and each of these has at different times resulted in performance bonuses or sanctions. The degree of access to SNAP is captured by the Program Access Index (PAI); compliance with national guidelines for the timeliness of processing initial applications is measured by Application Processing Timeliness (APT); and there are two metrics of accuracy: Payment Error Rates (PER) and Case and Procedural Error Rates (CAPER). 

Historically, all SNAP benefits were paid for by the federal government, and administrative costs were split 50–50 between states and the federal government. In summer 2025, however, the One Big Beautiful Bill mandated that states with PER above 6% would be required to cover between 5% and 15% of SNAP benefit costs beginning in October 2027. As a result, understanding variation in SNAP performance metrics—particularly PER—is of substantial policy and fiscal importance.

Specifically, we examine the following research questions:
1) How do SNAP performance metrics (PER, CAPER, PAI, and APT) vary over time and by state?
2) What trade-offs do states appear to make in terms of accuracy, access, and timeliness?
3) Which state-level demographic characteristics, economic conditions, and SNAP policies are associated with each performance metric?

We use USDA state-year data on each of the four performance metrics as well as information from the USDA Quality Control Data, the American Community Survey, the USDA SNAP Policy Database, and the University of Kentucky National Welfare Data. We examine trends over time and across states and use cluster analysis to sort states into five different groups based on the trade-offs they make across the four different performance metrics. We then estimate models for each metric to examine the role of state-level demographic characteristics, economic conditions, SNAP policies, and monthly per-case administrative costs in predicting state-year values.

Our findings will deepen our understanding of the extent to which USDA performance metrics reflect trade-offs consistent with public management theory. States may find our results helpful in understanding the forces at play in determining their own performance so that they can make evidence-based decisions to improve it. Finally, policymakers may find this information useful for understanding why states make trade-offs across accuracy, access, and timeliness.

Authors

• Colleen M Heflin, Syracuse University
  Presenting Author

• Michele Ver Ploeg
  Non-Presenting Co-Author

• Hyeryung Jo, Syracuse University
  Non-Presenting Co-Author