Estimating eviction prevalence across the United States

Drawing on 99.9 million court records, we construct national estimates of the annual prevalence of eviction filings and households threatened with eviction in the United States. Using Bayesian hierarchical modeling, we reconcile data from multiple sources to create comprehensive estimates permitting comparisons of eviction filing risk within and between states. This method indicates that relying solely on court-issued data undercounts eviction filings by approximately 1 million cases a year due to omission of counties for which these data cannot be obtained. In an average year between 2000 and 2018, landlords filed more than 3.6 million eviction cases, resulting in almost 7% of renting households facing an eviction lawsuit. During this time, the number of eviction filings nationally increased by 21.5%; however, an expanding renter population has outpaced the growth in filings, resulting in declining filing rates in recent years. Nationwide data reveal stark disparities in eviction filing rates between states that are not explained by variation in sociodemographic composition. Rather, regression discontinuity models indicate a robust association between a simple housing policy—requiring landlords to provide notice to tenants prior to filing an eviction case for nonpayment of rent—and the county-level eviction filing rate, demonstrating that larger structural factors, including state-level landlord–tenant law, could play an important role in shaping risk of receiving an eviction filing. We make aggregated data publicly available to serve as a tool for researchers, policymakers, and members of the public to examine the prevalence, causes, and consequences of eviction lawsuits.