Medicaid, Earnings, and Heterogeneous Treatment Effects: Evidence from the Oregon Health Insurance Experiment

Krieg Tidemann
DOI: https://doi.org/10.1515/bejeap-2020-0270
2021-10-01
Abstract:Abstract The Medicaid and labor supply empirical literature offers competing conclusions of zero effects and significant reductions in earnings. However, zero effects are only theoretically consistent with the earnings distribution’s extremes. Medicaid participants with positive pre-treatment labor supply should unequivocally decrease earnings. This paper clarifies the literature’s ambiguity by combining quantile regression with data from the Oregon Health Insurance Experiment. The distributional impacts imply that zero effects are not universally representative of Medicaid households. The annual earnings impact of Medicaid participation ranges between increases of $1400 to deceases of $3120 for single adults. Pre-existing mental illness or health constraints on work account for counterintuitive positive earnings impacts. By demonstrating that sample compositional differences determine whether Medicaid’s labor supply impact is zero or negative, this paper offers a reconciliation to the range of existing estimates in the empirical literature.
What problem does this paper attempt to address?