Medicaid expansions and labor supply among low-income childless adults: Evidence from 2000-2013

In many U.S. states, Medicaid expansions were used to provide low-income childless adults with access to healthcare prior to the passage of the Affordable Care Act (ACA). Accumulated evidence from these pre-ACA state expansions can lend insight into Medicaid’s impact on labor supply. Specifically, an examination of pre-ACA expansions can isolate the effect of Medicaid expansions alone without the confounding effects of other ACA provisions such as the exchange program. This paper studies the labor supply effects of Medicaid expansions to low-income childless adults between the ages of 20 and 64 years using the natural experiment that arises from pre-ACA variation in childless adult eligibility and allows us to compare individuals in states that expanded Medicaid prior to the ACA to those in states that did not implement expansions.

We use two nationally representative surveys, the Current Population Survey (CPS) and the Survey of Income and Program Participation (SIPP), one of which is a large repeated cross-sectional survey while the other is longitudinal. We supplement these data with compiled state level Medicaid eligibility and coverage criteria from 2000 to 2013. Our main outcome variables are employment, full-time employment, and weekly hours worked, conditional on full- and part-time and hourly employment. We estimate a simulated eligibility (SE) variable to mitigate endogeneity between Medicaid eligibility and labor supply by capturing the generosity of a state’s Medicaid eligibility threshold without relying on actual coverage. When estimate labor supply outcomes using OLS including the SE variable along with individual characteristics and state and year fixed effects. We also estimate IV equations with SE eligibility in the first stage and the labor supply outcome of interest in the second stage.

Labor supply reductions among low-income childless adults following Medicaid expansion are likely to be modest, if they occur at all. This finding was consistent across two samples and two different estimation approaches. It was also consistent for women and men. IV estimations show that reductions in women employment hover around 1 to 2 percentage points associated with becoming eligible for Medicaid while reductions in men employment are about 2 to 4 percentage points associated with becoming eligible for Medicaid. OLS estimates suggest that men were about 1.8 percentage point less likely to be employed for every 10 percentage point increase in Medicaid eligibility. Full-time female workers and part-time male workers had a small negative effect on weekly hours worked to Medicaid eligibility. We also observed a small positive effect on weekly hours worked for hourly workers, which could result from more latitude in hours worked without losing Medicaid eligibility.

We find evidence that Medicaid eligibility creates a small incentive to reduce employment. We also find some evidence that Medicaid eligibility may result in a small negative effect on hours worked among full- and part-time employees, but a positive effect for women who work hourly. Taken together, Medicaid eligibility appears to decrease labor supply among low-income childless adults.