21
The Impact of the Affordable Care Act Medicaid Expansions on Applications for Federal Disability Benefits

Tuesday, June 12, 2018
Lullwater Ballroom - Garden Level (Emory Conference Center Hotel)

Presenter: Priyanka Anand

Co-Authors: Jody Schimmel; Margaret Colby; Lauren Hula; Paul O'Leary


We estimate the impact of the ACA Medicaid expansions on applications to Supplemental Security Income (SSI) and DI using data collected on disability applicants by the Social Security Administration. Past work has examined the impact of Medicaid expansions on SSI and DI participation, but little is known about changes in the application rate. We use a difference-in-difference regression estimation approach to compare SSI and DI applications before and after the Medicaid expansion in states that expanded Medicaid to applications in the same period for the comparison states that did not expand Medicaid. To ensure comparability between the expansion and comparison states, we conducted a matching procedure that selects Public Use Microdata Areas (PUMAs) within nonexpanion states that were similar to PUMAs in expansion states in terms of their pre-ACA Medicaid policies, trends in SSI and DI application rates, demographics and socioeconomic characteristics. In addition to estimating an overall effect for all states, we conduct a subgroup analysis to examine whether there is heterogeneity in the impact by state.

Our findings reveal that there is no statistically significant relationship between the ACA Medicaid expansions and applications to either SSDI or SSI. However, we find evidence of heterogeneity in this relationships between states, with some states experiencing a positive impact on disability applications and others experiencing a negative impact. We were not able to identify a consistent pattern for these state-level results. Our findings suggest that the new availability of health insurance coverage does not have a consistent impact on the incentives of individuals with disabilities to apply for SSI and DI benefits but that state-level features that are unobservable in our data may influence the direction of this relationship.