Incentive(less)? The Effectiveness of Tax Credits and Cost-Sharing Subsidies in the Affordable Care Act

The Patient Protection and Affordable Care Act of 2010 implemented a complex, broad set of changes in the U.S. health care system. In 2014, several large ACA components went into effect: individual and large employer mandates, health insurance exchanges, and expansion of the Medicaid program. A wave of early, nationally representative polls have suggested that the ACA has reduced the proportion of the uninsured population (Sommers et al. 2015a, Sommers et al. 2014).  The early investigations into the impact of the ACA have to-date used pre-post, difference-in-difference, and interrupted time series methods.  Identification is a large concern in examining the ACA – a feasible control group to which a counterfactual can be drawn is a daunting task given the widespread influence of the ACA policies.  Although preliminary evidence suggests insurance coverage increased substantially, disentangling the mechanisms by which the consumer behaviors are affected is of critical importance.  This analysis focuses on consumer behavior in the private, non-employer-sponsored health insurance market.  I use a regression discontinuity design to identify impacts of the tax credits available to individuals below 400% FPL and cost-sharing subsidies available to individuals below 250% FPL on directly purchased health insurance.  Data are drawn from the most recent Current Population Survey Annual Social and Economic Supplement. Preliminary results confirm a large increase in directly purchased health insurance coverage found in reports from rapid response surveys. Increases in directly purchased insurance are highest between 138% and 250% FPL. I find no discernable differences in coverage attributable to the cost-sharing subsidies. There are suggestive effects of tax credits on private insurance coverage, with stronger effects in states that implemented a state-based health insurance exchange.