Did the Affordable Care Act's Medicaid eligibility expansions crowd out private health insurance coverage?

The Affordable Care Act (ACA) provided funding to help states expand Medicaid eligibility to those earning up to 138% of the Federal Poverty Level. Such expansions in Medicaid eligibility, however, could "crowd out" private insurance coverage, including changes in coverage relating to other ACA provisions. To estimate the extent of such crowd out, I use a difference-in-differences empirical approach, examining changes in health insurance coverage sources among low-income Americans in states that expanded eligibility relative to comparable individuals in states that did not. Using American Community Survey data from 2009 to 2019, I find a 43% crowd-out rate, consisting of a 10.7 percentage point relative increase in Medicaid coverage among low-income adults and a 4.6 percentage point relative decline in private health insurance among respondents in states that expanded Medicaid eligibility. Among working adults, my estimates imply a larger 56% rate of crowding out. Event study analyses provide support for a causal interpretation for my findings. I further show that my estimates are robust to different sample restrictions and estimation choices, are not subject to the issues raised by the new difference-in-differences literature, and are similar when I use approaches to identifying crowd out common in the existing literature.