Panel Paper: Homelessness Among Veterans: Evidence From the Greater Richmond Area

Thursday, November 7, 2013 : 12:10 PM
3015 Madison (Washington Marriott)

*Names in bold indicate Presenter

Shannon A. McKay, Federal Reserve Bank of Richmond and Margot Ackermann, Homeward
With the military combat efforts in Iraq having ended and those in Afghanistan winding down, more veterans will be returning stateside with housing, employment and health needs. Among the issues confronting veterans is homelessness. While homelessness among veterans has been on the decline, more than 62,000 veterans were homeless in 2012. In 2010, the Obama administration set a goal to end homelessness among veterans by 2015.

To effectively target homelessness reduction efforts, it is important to identify the risk factors contributing to homelessness among veterans. Our study investigates the factors that affect the risk of homelessness among the low-income veteran population in the Greater Richmond, Virginia area. We look at the role of demographic, socioeconomic, mental health and behavioral characteristics on veterans’ likelihood of homelessness. Our study is based on data collected from a semi-annual survey of low income and homeless populations between 2008 and 2012. The survey is conducted by Homeward, a planning and coordinating organization for homeless services in the Greater Richmond area.

To address the questions, we model an individual’s likelihood of homelessness as follows:

Let Yi be a binary random variable that assumes a value of 1 if individual i is homeless and 0 otherwise. Then the probability of being homeless, after controlling for individual characteristics is given by:

P[Yi = 1|Xi = x]  =  Φ(β0  + βxi   + γVi  + δVi *xi)                

where Vi assumes a value 1 if an individual is a veteran and 0 otherwise,  xi is a vector of demographic, socioeconomic, behavioral  and other characteristics of the individual and Vi *xi are the interactions of individual characteristics with the veteran variable. Thus, the model enables us to measure the effects of the risk factors on veterans and compare them to those of the overall low-income population. We estimate the model using maximum likelihood techniques with logistic specification for Φ.

Among the variables included in our study are those that have been found to be significant determinants of homelessness in prior research, such as demographic and socioeconomic characteristics (Lazaryan et al, 2013; Early, 1999), history of substance abuse (Early, 2005; Allgood and Warren, 2006), adverse life events (Lazaryan et al, 2013), and access to affordable housing and mental health programs (Elliccot and Krivo, 1991; Honig and Filler 1993; Early 2002; Lazayan et al, 2013). We also test whether access to VA benefits plays a role in reducing the risk of homelessness among low-income veterans.

The contribution of our research is to provide up-to-date evidence that will inform policy directed towards homelessness elimination efforts among veterans. Our findings will help identify the veterans most at risk of homelessness and help policymakers in targeting resources towards the most effective strategies.

Full Paper: