Association for Public Policy Analysis & Management

The housing cost burden among renters has increased in recent years, placing many households at risk of eviction. While a handful of local studies have identified the frequency and negative consequences of eviction for low-income renter households, we have no comprehensive estimate of the prevalence of eviction in the United States. To address this, we compiled, standardized, and geocoded over 81 million court records of eviction cases that occurred between 2000 and 2016 to construct the first national database of evictions in the United States. We validated eviction case volume at the county-level to ensure consistent coverage across space, a key step in creating comprehensive estimates. We used Bayesian Hierarchical modeling to predict total case volume in counties without a complete accounting of cases in the court records. We then used a set of negative binomial regression models to estimate the number of unique households threatened by eviction, as well as the number of cases that resulted in an eviction judgment against the tenant(s). We find that, on average, 3.5 million eviction cases are filed each year in the U.S. As a result, 2.7 million households are threatened by eviction and 1.4 million households receive a judgment to leave their homes annually. Eviction is most concentrated in suburban and fringe metropolitan areas in the Southeastern U.S. outside the large metropolitan centers that usually dominate discussions of access to affordable housing. We find substantial differences in eviction rates across states, even after accounting for the distributions of population demographics. These new estimates—the first of their kind—provide us with a foundation to examine the scope and concentration of eviction in a national perspective. This is crucial to our ability to study high- and low-displacement areas and evaluate the effectiveness of housing policies and regulatory environments.