Does Broad-Based Categorical Eligibility Increase SNAP Participation?

Wang et al. (2026) ask whether state adoption of broad-based categorical eligibility (BBCE) increases SNAP participation, and how much of that increase comes from newly eligible versus already eligible households. They analyze state-level SNAP participation from 1996–2016 using administrative SNAP Quality Control data combined with state policy and economic data. Using a difference-in-differences design that allows effects to vary over time and across states, they find that BBCE increased SNAP participation by about 15 percent, more than double earlier estimates. Effects grew over time, reaching roughly 25 percent after seven years, and were driven mainly by higher take-up among households already income-eligible.

Why This Article Was Selected for The Policy Scientist

This article addresses a broad and important policy issue: how the structure and administration of public benefit programs affect whether eligible people actually participate. SNAP is one of the largest social programs in the United States, and small design choices can influence millions of households and billions of dollars in spending. Understanding participation is therefore essential not only for SNAP but also for programs such as Medicaid, housing assistance, and energy assistance, which rely on similar eligibility rules and administrative processes. Wang et al. (2026), who have written extensively on SNAP policy, show that eligibility expansions can raise participation mainly by increasing take-up among people who were already eligible. This insight is timely as policymakers continue to reconsider program complexity, access, and administrative burden across the social safety net.

Full Citation and Link to Article

Wang, X., Valizadeh, P., Nayga, R. M., Bryant, H. L., & Fischer, B. (2025). Broad-Based Categorical Eligibility Policy and SNAP Participation. Journal of Policy Analysis and Management, 45(1). https://doi.org/10.1002/pam.70063 

Central Research Question

This article asks whether state adoption of broad-based categorical eligibility (BBCE) causally increases participation in the Supplemental Nutrition Assistance Program (SNAP), and whether prior research has underestimated this effect due to methodological limitations. The authors further examine how BBCE’s impact evolves over time, whether effects differ across states with different adoption timing, and which subpopulations drive observed changes in participation. A central concern is whether BBCE primarily expands SNAP to higher-income households newly made eligible, or instead increases take-up among households already eligible under federal income rules. By addressing these questions, the study aims to clarify both the magnitude and the mechanisms of BBCE’s role in SNAP growth.

Previous Literature

A substantial literature examines state-level SNAP policy variation as a determinant of program participation. Widely cited studies, including Klerman and Danielson, Ziliak, and Ganong and Liebman, generally find that BBCE increased SNAP participation by roughly 4 to 8 percent. These studies rely almost exclusively on static two-way fixed effects (TWFE) difference-in-differences models exploiting staggered state adoption of BBCE. However, recent econometric research has shown that TWFE estimators can produce biased or misleading estimates when treatment effects vary across units or over time, particularly under staggered adoption. The authors position their contribution squarely within this critique, arguing that earlier BBCE estimates may lack a clear causal interpretation. Building on advances by Callaway and Sant’Anna and others, they re-estimate BBCE’s effects using methods explicitly designed to handle treatment effect heterogeneity, thereby reassessing a well-established empirical result in the SNAP literature.

Data

The analysis uses high-quality administrative data from the SNAP Quality Control (QC) database covering the period from 1996 through 2016. These data provide detailed, household-level information on SNAP participants and are aggregated to the state-year level to construct per-capita participation measures. The authors combine these data with state population figures from the U.S. Census Bureau, state unemployment rates from the Bureau of Labor Statistics, and detailed policy information from the USDA Economic Research Service SNAP Policy Database. This policy database includes adoption timing and specific features of BBCE, such as gross income and asset test thresholds, as well as other SNAP-related administrative policies. The resulting panel includes all 50 states and the District of Columbia over 21 years, yielding more than 1,000 state-year observations. Overall, the dataset is well suited to identifying policy-induced changes in participation over time.

Methods

The authors begin by estimating BBCE’s effect using the standard static TWFE difference-in-differences model to replicate the approach used in earlier studies. They then assess the limitations of this model using a Goodman–Bacon decomposition, which reveals substantial weight placed on problematic comparisons between earlier- and later-treated states. To address these concerns, the core analysis employs the Callaway and Sant’Anna (2021) difference-in-differences estimator, which allows treatment effects to vary across states and over time and avoids negative weighting. The authors estimate group-time average treatment effects based on adoption cohorts and aggregate them using policy-relevant weights. They also implement dynamic event-study versions of both TWFE and Callaway–Sant’Anna estimators to trace the evolution of BBCE’s impact over time. Covariates are incorporated through a doubly robust procedure, strengthening the plausibility of conditional parallel trends. This approach represents a strong application of modern causal inference techniques to an established policy question.

Findings/Size Effects

Using the static TWFE model with covariates, the authors estimate that BBCE increased SNAP participation by about 6 percent, consistent with prior studies. In contrast, the heterogeneity-robust Callaway–Sant’Anna estimator produces an estimated increase of roughly 15 percent, more than double the TWFE estimate. Event-study results show that BBCE’s effect is small in the year of implementation, approximately 2 to 3 percent, but grows steadily over time, reaching about 24 to 25 percent after seven years before leveling off. Importantly, this growth pattern appears similar across adoption cohorts, explaining why static TWFE models understate the average effect. Heterogeneity analyses show that although participation among households just above federal income limits is responsive to BBCE, most of the overall increase in SNAP participation comes from higher take-up among households already income-eligible. Counterfactual simulations suggest that BBCE accounted for roughly 1.6 million additional participants per year on average between 2000 and 2016, while only about 11.5 percent of participation growth and 3.8 percent of added spending came from newly eligible higher-income households.

Conclusion

This study provides a substantial reassessment of BBCE’s role in SNAP participation growth by applying contemporary causal inference methods to a long-standing policy question. The findings suggest that earlier estimates based on TWFE models significantly understated BBCE’s impact due to unaddressed treatment effect dynamics. The results also clarify that BBCE’s primary effect operates through increased take-up among already eligible households rather than through large expansions to higher-income groups. Beyond SNAP, the analysis has broader implications for evaluating administrative and eligibility reforms in means-tested programs, especially those implemented gradually across jurisdictions. By demonstrating how methodological choices shape substantive conclusions, the article makes an important contribution to the applied policy evaluation literature and sets a higher standard for future work in this area.