Do SNAP Eligibility Expansions Increase Take-Up Among Households That Were Already Eligible?

Anders and Rafkin (2025) examine whether expanding SNAP income-eligibility limits increases program use among households that were already eligible. Using administrative SNAP Quality Control data from 1996–2016, they track state-level changes in eligibility thresholds and their effect on participation. They find that a 10-percentage-point increase in the income limit raises take-up among always-eligible households by about 1–2 percent. For every newly eligible person who enrolls, about 0.99 previously eligible individuals also begin participating. They argue that improved information—not changes in stigma—best explains these increases, and they show that such eligibility expansions are likely to raise overall social welfare.

Why This Article Was Selected for The Policy Scientist

This article addresses a consequential policy issue: how eligibility rules in major transfer programs shape real participation, administrative burdens, and the effective reach of the social safety net. That question is especially timely as governments reconsider means-testing thresholds, program complexity, and information barriers in the wake of large pandemic-era policy changes. Anders and Rafkin, who have written extensively on program participation and administrative frictions, offer a useful contribution by pairing high-quality administrative SNAP data with a clear empirical strategy. Their identification approach is stronger than standard cross-sectional regressions. The dataset’s national coverage supports cautious generalization to other U.S. jurisdictions with similar eligibility structures.

Full Citation and Link to Article

Anders, J., & Rafkin, C. (Forthcoming 2025). The welfare effects of eligibility expansions: Theory and evidence from SNAP. American Economic Journal: Economic Policy. https://www.aeaweb.org/articles?id=10.1257/pol.20220738 

Central Research Question

Anders and Rafkin (2025) investigate whether expanding SNAP’s income-eligibility threshold alters program participation among households that were already eligible under baseline federal rules. Their central aim is to determine whether loosening eligibility criteria influences take-up behavior among inframarginal households—those whose incomes fall well below all plausible thresholds—and to identify the mechanisms that drive this behavioral response. A related objective is to quantify the welfare implications of such changes by integrating empirical estimates with a theoretical model of program participation that incorporates information frictions and heterogeneous stigma costs. Together, these aims allow the authors to evaluate eligibility design not merely as an administrative constraint but as a behavioral policy instrument that shapes participation patterns across the income distribution. 

Previous Literature

The study builds on an extensive literature analyzing incomplete take-up of social programs, where researchers have long noted the roles of transaction costs, stigma, and limited information in depressing participation among eligible households. Prior work has documented that similar patterns arise in health insurance programs, including Medicaid, where expansions sometimes increase enrollment among individuals who were already eligible. Research in public economics has also emphasized how means-testing rules shape both enrollment incentives and broader perceptions of program legitimacy. The authors contribute by clarifying that eligibility thresholds themselves—not just administrative burdens or benefit levels—can produce behavioral responses among inframarginal households. While earlier studies often relied on treatment effects identified at program cutoffs or within newly eligible populations, Anders and Rafkin extend the literature by showing that eligibility design can influence participation far from the margin, and by developing a welfare framework that distinguishes between information-driven and stigma-driven responses. They also situate their analysis within recent work examining barriers to benefit access, which increasingly recognizes that misperceptions of eligibility and lack of program awareness are often more consequential than previously assumed. 

Data

The authors rely primarily on administrative SNAP Quality Control (QC) data from 1996–2016, which provide highly accurate information on participant income, household characteristics, and benefit receipt. Because these data represent a quality-assured sample of all participants, they avoid many of the measurement errors that plague household surveys, particularly underreporting of program participation. The QC data allow the authors to compute enrollment counts among specific income bands, focusing on households between 50–115 percent of the Federal Poverty Level (FPL), a group that is always eligible regardless of state policy. This design isolates changes in participation among households whose eligibility status should theoretically remain stable.

To construct denominators for take-up rates, the authors use CPS ASEC data, while acknowledging its known income-measurement limitations. They also incorporate detailed state-level policy data, including the timing of eligibility expansions permitted under Broad-Based Categorical Eligibility (BBCE), SNAP outreach spending, unemployment rates, and other contemporaneous program rules. Additional analyses incorporate the Food Stamp Program Access Study (FSPAS), a nationally representative survey capturing information and stigma perceptions among both participants and nonparticipants. This dataset—rarely used in academic research—enables the authors to connect demographic correlates of information gaps and stigma to the take-up patterns observed in administrative data. Overall, the dataset is unusually rich for policy research, with national coverage and high-quality administrative measures, which strengthens both internal validity and generalizability to other U.S. jurisdictions with comparable administrative structures. 

Methods

The primary empirical strategy is an event-study design exploiting variation in the timing of state-level eligibility expansions under BBCE. Because only some states chose to extend eligibility beyond the federal 130% FPL threshold, and because expansions occurred at different times from 2001–2016, the authors use this staggered rollout to estimate how changes in eligibility thresholds affect SNAP enrollment among inframarginal households. The specifications include state and year fixed effects, controls for the size of the eligible population, macroeconomic conditions, and other SNAP policies. The event-study structure permits examination of pre-trends, which appear flat once key controls are included, improving confidence in the causal interpretation.

The authors supplement the event-study with instrumental variables estimates that treat the eligibility threshold as an instrument for the share of the population eligible for SNAP, enabling calculation of an elasticity of take-up with respect to eligibility. Though not an RCT, the empirical design approximates causal inference by utilizing plausibly exogenous policy timing and by conducting placebo tests in states that adopted BBCE but did not change eligibility limits. These placebo states exhibit no shift in take-up, strengthening the argument that eligibility expansions—rather than other administrative modifications under BBCE—drive the observed effects.

For mechanism testing, the authors use (1) an online experiment that randomizes beliefs about the prevalence of SNAP eligibility and examines effects on stigma measures and (2) the FSPAS survey to compare demographic correlates of stigma, information gaps, and enrollment responses. Finally, they build a theoretical model that integrates information frictions and stigma costs into welfare analysis, illustrating how the source of behavioral responses matters for social welfare calculations. While the empirical strategy is rigorous, future work using randomized or quasi-experimental variation could further refine causal inference about mechanisms. 

Findings/Size Effects

The central finding is that expanding SNAP eligibility thresholds increases participation among households that were already eligible. A 10–percentage-point increase in the income threshold (for example, from 130% to 140% of FPL) produces a 1–2 percent increase in take-up among households between 50–115% of FPL. When states adopt the modal expansion—from 130% to 200% of FPL—take-up among this group rises by roughly 8.4 percent. Strikingly, for every newly eligible enrollee, approximately 0.99 previously eligible individuals also enroll. These size effects suggest that expansions influence large segments of the low-income population who are not near the eligibility cutoff.

Mechanism tests indicate that information plays a larger role than stigma. The demographic groups showing the strongest enrollment responses are those with the weakest baseline program knowledge, not those most sensitive to stigma shifts in the experiment. Although the randomized information intervention modestly reduces perceived stigma, these reductions do not align with the demographic pattern of enrollment responses. This suggests that eligibility expansions function partly as informational signals about program accessibility.

The welfare analysis indicates that eligibility expansions increase social welfare under reasonable assumptions, especially when new take-up arises because previously uninformed households become aware of their eligibility. In this case, the welfare gains reflect first-order improvements in household utility, since previously uninformed households could not optimize their participation decisions. The authors estimate that making one additional household eligible for SNAP yields a welfare gain of approximately $697, including the mechanical fiscal cost of expanding eligibility. These estimates depend on modeling assumptions, but they illustrate the significance of information as a determinant of both behavioral response and welfare impact. 

Conclusion

Anders and Rafkin demonstrate that eligibility thresholds shape program participation far beyond the margins traditionally examined in policy research. By showing that expansions increase enrollment among already-eligible households, they reveal that eligibility design performs an informational function in addition to a screening function. Their integration of administrative data, policy variation, survey evidence, and a welfare model provides a cohesive assessment of how eligibility rules affect both behavior and social welfare. The study advances the literature by reframing eligibility expansions as a behavioral policy lever rather than a purely administrative boundary. With U.S. debates increasingly focused on access barriers, administrative burdens, and the role of means-tested programs, this research provides timely evidence that program design indirectly alters take-up patterns through information channels. Future work that employs randomized or quasi-experimental designs could further clarify causal pathways, but the present study offers a substantial and methodologically careful contribution.