Abstract
This thesis addresses the problem that cities, counties and states are experiencing with a rising number of homeless veterans. Specifically, it analyzes the relationship between homeless veterans in Sacramento County, California, and long-term homelessness. The hypothesis is that when all other factors of homelessness are held constant, a veteran of the US. Military is more likely to experience long-term homelessness than a non-veteran.
The data used to determine this relationship is from the 2007 Sacramento County homeless survey. Multivariate logistic regression models analyzed this data in order to determine the relationship between long-term homelessness and different possible causal factors. The regression results indicated that there is no discernable relationship between veteran status and long-term homelessness, thus disproving the hypothesis of the thesis. The model also indicated that the drug abuse, alcohol abuse, former foster care/group home resident, and age variables are positive and statistically significant predictors of long-term homelessness. In addition, the demographic African-American variable reported as a statistically significant negative predictor of long-term homelessness.