Association for Public Policy Analysis & Management

Background: Under the Affordable Care Act (ACA), premium subsidy is the major provision that enables lower income households with limited access to employer-sponsored insurance and other public insurance such as Medicaid to purchase affordable individual health insurance on the “Marketplace.” Households with incomes between 100 and 400% Federal Poverty Level (FPL) are eligible to receive a substantial subsidy when they purchase private health insurance. In states where Medicaid was not expanded, a coverage gap has emerged where some groups of households are eligible for neither Medicaid nor a premium subsidy (e.g., those with incomes below 100% FPL but above Medicaid eligibility threshold), while households with income above 100% FPL are eligible for a premium subsidy. Therefore, those households around the cutoff face very different insurance premium costs.

Goal: Most previous studies of the effect of the ACA on insurance coverage are either descriptive or focus on the Medicaid expansion. In this study, we assess the causal effect of the premium subsidies on health insurance coverage among lower income households using a representative sample of Americans.

Data and Methods: We use data from the Current Population Survey (CPS) March 2012-2017. The premium subsidy eligibility creates exogenous variation in the insurance premiums that families must pay. We exploit this policy-induced variation using a regression discontinuity (RD) research design, and we measure the effect of the premium subsidies on coverage by comparing the probability of having coverage between households whose income are just above the eligibility threshold (100% FPL) and households whose income are just below the threshold. We further examine the effect for different subgroups based on their household type and likelihood of having health insurance (e.g., single adults with and without children, married couples with and without children).

Results: We find limited evidence that premium subsidies increased the probability of having coverage among the average household around the cutoff income level. However, premium subsidies increased the likelihood of being insured among single childless adults with incomes just above the cutoff relative to those just below the cutoff.

Discussion: Premium subsidies might have differential effects on insurance coverage across different types of lower income households. Health insurance might be more valuable to single childless adults who are more likely to lack social support (either financially or mentally).