A National Analysis of the Spatial Patterns and Correlates of Evictions in the United States

This national study investigates the spatial patterns and correlates of housing evictions in the U.S., a critical crisis of local and national significance. A national data set on county-level eviction rates was obtained from Princeton University’s Eviction Lab database, and the data on risk factors of eviction were acquired from the Centers for Disease Control and Prevention (CDC)’s Social Vulnerability Index (SVI). First, we examined disparate patterns of evictions across U.S. counties, followed by a hot spot analysis to determine clusters of counties with significantly high or low values of eviction rates. The analysis concluded with multiple regression to investigate the associations of eviction rates and 15 CDC SVI indicators. Evictions are most prevalent in populous urban or metropolitan counties; however, eviction rates can be higher in less-populous suburban counties even nonmetropolitan communities. Clusters of counties with significantly high eviction rates (i.e., hot spots) were mainly concentrated in lower Michigan-upper Indiana and the along the east coast that spanned Virginia, North Carolina, and South Carolina. The spatial regression model explained a moderate degree of the variations of eviction rates and social vulnerability indicators including income, minority population, single-parent, unemployment rate, apartment living, and low education status were most helpful predictors of eviction. This national study can inform government resource allocation efforts where it provides new insights into the spatial disparities and potential contributing factors of eviction rates across U.S. counties, thus enhancing our understanding of the eviction crisis nationwide.