ON THE DETERMINANTS OF HOMELESSNESS IN U.S. METROPOLITAN AREAS

Sociological research on homelessness has long considered its structural determinants, mainly those factors related to local housing and labor markets. The purpose of this dissertation is to expand and update this understanding by examining previously untested drivers that have important implications for community rates of homelessness: income segregation, eviction, and social capital. In chapter one, I investigate the relationship between income segregation and local homelessness rates. I find a positive association between the spatial segregation of low-income households and the prevalence of the condition, which confirms qualitative research suggesting that homelessness thrives when poverty is highly concentrated and homelessness can be obscured from advantaged social actors and prime economic spaces. In chapter two, I examine associations between local eviction rates and rates of homelessness. I arrive at a series of null results across a range of homelessness measures and model specifications. I argue that these null findings are likely due to the high prevalence of informal forced moves which contribute to overall housing instability but are not captured by official eviction records. In the final chapter, I explore the relationship between average local temperatures and rates of unsheltered homelessness. While this bivariate relationship is strong and positive, it is largely explained by local stocks of homeless shelter beds relative to the size of local homeless populations. Stated differently, I show that warmer communities tend to provide fewer shelter beds per homeless person. Upon further exploration, I find communities that undersupply shelter beds tend to be located in more exclusive rental markets and score lower on measures of social capital, broadly implying that beds are less likely to be supplied at a sufficient level when the average renter is farther from poverty and civic engagement and feelings of public spiritedness are low