Project:
Does SNAP Discourage Claiming Disability Insurance Benefits?

Year: 2017

Investigator: Drabo, Emmanuel Fulgence

Institution: Johns Hopkins University

Project Contact:
Emmanuel Fulgence Drabo
Johns Hopkins University
Department of Health Policy and Management
Johns Hopkins Bloomberg School of Public Health
624 North Broadway, Suite 402
Baltimore, MD 21205

Recent studies have called for more research to understand the determinants and consequences of participation in the Supplemental Nutrition Assistance Program (SNAP), a critical step for evaluating SNAP’s effectiveness at supplementing both private and public incomes, reducing dependency, and helping individuals with disabilities remain in the labor force. Of particular interest is the way in which SNAP may interfere with the Social Security Disability Insurance (SSDI) program, to encourage or discourage participation in the labor force, given that recent studies found that an increasing share of SNAP households with a non-elderly working-age disabled person (47 percent in 2009 versus 51.4 percent in 2014) have claimed SSDI benefits. Understanding whether SNAP discourages SSDI-seeking behavior (application to SSDI), and why the SSDI application rate among SNAP households is increasing, is therefore critical. It is also of policy relevance, as States can request a temporary waiver of SNAP’s time-limit requirements for the able-bodied adults without dependents (ABAWD) population when unemployment is high or when there are not enough jobs available; determining the effects of such a waiver would inform states regarding the consequences of SNAP policy exemptions.

This study examines whether SNAP’s general and ABAWD work requirements discourage or encourage seeking SSDI benefits, after accounting for other factors that have been found to also affect SSDI- seeking behavior, measured by applications to SSDI. This study first discusses pathways through which SNAP’s work requirements may discourage or encourage SSDI-seeking behavior among SNAP participants. SNAP may encourage work through its general and able-bodied adults without dependents (ABAWD) work requirements, as well as its earned income tax credit provision for qualifying working beneficiaries, and hence encourage beneficiaries to remain in the labor force and only seek SSDI benefits as a last resort. Despite this theoretical encouragement for work, the work requirement for SNAP may encourage SSDI applications because a significant number of people with disabilities are still considered ABAWD by the U. S. Department of Agriculture’s (USDA) definitions and may seek SSDI because they have a tenuous relationship with the labor force. Hence, SNAP participants could be incentivized to seek SSDI more actively given the SNAP work rule and the high risk of this population becoming temporarily unemployed. This SSDI seeking behavior may discourage labor force participation, because any evidence of work would jeopardize SSDI receipt.

This study selected a sample of SNAP eligible individuals (aged 17-57 years) from the U.S. Census Bureau’s SIPP Synthetic Beta v6.02 (SSB), which links data from the SIPP to SSA administrative data. This study characterized this population in terms of their socio-demographic and economic characteristics, as well as their ABAWD status. To minimize potential selection and simultaneity biases of estimates based on observational data (due to the endogenous nature of SNAP participation), this study used previously-identified instrumental variables (IVs) for SNAP participation (e.g. State broad based categorical eligibility [BBCE] policy) to estimate the causal effect of SNAP participation on SSDI- seeking behavior (application for SSDI). To estimate the effects of SNAP’s ABAWD work requirements on SSDI seeking behavior, this study estimates a heterogeneous treatment effect (HTE) model within the IV framework. This study also conducts robust analyses of these HTE estimates, using the causal forest modeling (CFM) approach.

Results from the IV analyses show that SNAP participation is associated with reduced likelihood of SSDI-seeking behavior. However, the HTE and CFM analyses yield consistent results, showing that SNAP’s work requirement policy for the ABAWDs may disincentivize labor force participation and incentivize SSDI-seeking behavior among those participants.

Two main conclusions emerge from this study. First, SNAP participation is associated with reduced likelihood of SSDI-seeking behavior. Second, however, the ABAWD work requirement is associated with more pronounced SSDI-seeking behavior among SNAP participants who can be categorized as ABAWDs. These findings may be explained by the fact that a significant number of people with disabilities are still considered ABAWD by the USDA’s definitions and may seek SSDI because they have a tenuous relationship with the labor force. Hence, SNAP participants may be incentivized to seek SSDI more actively given the SNAP work rule and the high risk of this population becoming temporarily unemployed. These results help shed light on the potential unintended consequences of SNAP participation and work requirements on SSDI-seeking behaviors.