The Connection Between Family Homelessness and Foreclosures over the Most Recent Housing Cycle
*Names in bold indicate Presenter
The analysis is conducted in two parts. First, we expand upon earlier work on the determinants of area-level homelessness (Honig and Filer, 1993; Boston, 2012; Byrne et al. 2012; and Fargo et al., 2013) and exploit variation in homelessness both over time (2009 to 2014) and across over 200 local areas identified by HUD as “CoC’s. (A CoC is defined as a geographic area over which local or regional organizations provide services to the homeless.) Using the HUD data, we focus on persons living in homeless families in metro areas. (For purposes of comparison, we also look at the chronically homeless.) A key covariate is a measure of MSA-level foreclosures, obtained from the proprietary data source, CoreLogic. Other area-level covariates include area-level unemployment rates, fair market rents, and changes in states and local-area policies affecting lower income persons (e.g. welfare generosity, access to Medicaid and mental health resources, etc.)
In the second part of the analysis, we examine the determinants of families being at risk of homelessness using household-level data from the American Housing Survey. The focus is on the housing status of low-educated single mothers since these families are the most economically vulnerable. These households are regarded as at risk of homelessness if they are living in substandard or overcrowded housing, doubling-up out of economic necessity, or paying more than 50 percent of their income in rent. These measures of being “at-risk” of homelessness are drawn from definitions of homelessness used by HUD based on the HEARTH Act of 2009, the definition of homelessness used by the Department of Health and Human Services, and research-based concepts. In this portion of the analysis, we use the same set of area-level covariates plus controls for household-level characteristics. The advantage of this second approach is two-fold; first, by defining the dependent variable at the household level, we can better pin down the direction of causality. Second, from a policy standpoint, we are interested not only in understanding the determinants of homelessness per se, but also the factors that place families at the margin of becoming homeless. Policies to assist families may be most effective at this juncture.