The Impact of Affordable Care Act Medicaid Expansions on Applications to Federal Disability Programs
DRC Working Paper Number: 2017-01
Publisher: Washington, DC: Center for Studying Disability Policy, Mathematica Policy Research
In this paper, we estimate the impact of Medicaid expansions via the Patient Protection and Affordable Care Act (ACA) on applications to federal disability programs in 15 states that expanded Medicaid in 2014. We use difference-in-differences regression to compare disability application rates in geographic areas within states that expanded Medicaid to rates in carefully matched areas of non-expansion states. We first selected the non-expansion states as potential comparison states for each expansion state based on pre-ACA Medicaid program features that might have affected the perceived value of disability benefits. We then propensity score matched each Public Use Microdata Area (PUMA) in an expansion state to PUMAs in selected non-expansion states, accounting for pre-period trends in disability applications and a set of PUMA-level characteristics. We find 6 states that experienced a statistically significant impact on SSI application rates and 6 that experienced a statistically significant impact on SSDI-only application rates, with effects varying in size and direction.
We find significant cross-state variation in estimated impacts. Of the 15 expansion states for which we generated estimates, we found 6 that experienced a statistically significant impact on SSI application rates and 6 that experienced a statistically significant impact on SSDI-only application rates. Yet, these estimates varied in size and direction, with some of the estimated effects positive and others negative. Competing factors could drive the net effect in either direction, depending on which is dominant, but we are not able to confirm that any of these competing factors explain the variation in the results. It is also possible that state variation in the direction of the effect is due to methodological issues resulting from limitations in our ability to identify an appropriately matched comparison group for each PUMA in expansion states.
Disability Research Consortium
Social Security Administration