The Rent is Too Damn High and the Coverage is Too Damn Low

In this paper, I use a novel dataset on evictions and a quasi-experimental research design to shed light on the relationship between health insurance and housing stability. I exploit variation in Medicaid expansion implementation to measure the expansion's impact on country-level eviction rates. In my preferred difference-in-differences specification, I estimate that the expansion decreased eviction rates by 12.5%, leading to one fewer eviction per 400 renting households. As the 95% confidence interval on this estimate ranges from -25.7% to 0.7%, I fail to reject both that the expansion had a more substantial negative impact or that it had no impact. The expansion appears to have a heterogeneous impact based on county characteristics. For example, it appears to have a larger impact on rent-burdened counties. Counterintuitively, I find that a county's baseline uninsured rate has little relationship with the impact of Medicaid expansion on the county's eviction rates. My paper adds to a growing literature on the impact of Medicaid expansion on a variety of labor, health, and economic outcomes and is the first to use this national dataset of evictions.