The Impact of Emergency Shelter, Transitional Housing, and Permanent Supportive Housing on Unsheltered Homeless in the United States

45 pagesUnsheltered homeless in the United States commonly identify as chronically homeless per the Department of Housing and Urban Development (HUD) definition. This means they suffer from a disabling condition and have either been homeless for at least a year or have experienced at least four episodes of homelessness in the past three years. Understanding their relationship to public and private support systems from an aggregate perspective is imperative in determining proper channels for aid. This population also typically exhibits transient characteristics. Identifying patterns of movement in this population would also aid future research and funding decisions. To investigate the effect of shelter beds on unsheltered homeless rates and potential migration, I analyze panel data provided by HUD from 2007 to 2017. I develop an econometric model to firstly identify a relationship between increases in emergency shelter, transitional housing and permanent supportive housing beds on unsheltered homeless counts. I then add neighboring CoC and region variables in the attempt to measure a substitution effect of unsheltered homeless between CoCs and states. I initially find suggestive evidence that increases in shelter beds effect unsheltered homeless rates at the 0.01 significance level. However, upon adding controls for potential omitted variable bias and reverse causality, I discover that these suggested patterns were incorrectly specified. I find no systematic evidence of a relationship or migration on a regional or national level