An Evaluation of the Plan to End Homelessness: The Chicago Housing First Model

Purpose:  The “Housing First” strategy (Kertesz, Crouch, Milby, Cusimano, and Schmacher, 2009), has been a major approach in our efforts to ameliorate homelessness over the past decade. Under this approach, permanent housing is the central support for homeless people and social services are seen as (client) optional supplements.  The emphasis of the system is on getting people housed as soon as possible. Chicago was one of the first cities to adapt this approach, incorporating its ideology into a Plan to End Homelessness.  This oral paper presentation uses data from a multi-year evaluation of the Chicago Plan.  It details the demographic characteristics, paths into homelessness and needs of individuals in the different types of housing supported under the Plan, comparing individuals still using emergency shelter services to those in programs that ideally reflect the Housing First ideology.  It also traces client outcomes a year after entry into the system.

 Methods:  Data for this presentation come from a stratified random sample of clients residing in programs comprising Chicago’s homeless services system. A multi-stage, stratified random sample design was used, where programs were randomly selected and individual clients were randomly sampled within programs.  In total, 554 clients were included from 59 different programs and interviewed three times over the course of a year.  Participants were interviewed using a structured questionnaire with questions about individual background, service receipt, and perceptions of programs. Major dependent variables included whether or not an individual was housed in permanent housing at the conclusion of the study and the number of days the individual reported being homeless at the final interview.  Important independent variables were type of housing program, whether the individual was in a family or a single, and chronic personal problems including psychiatric, health and substance abuse problems. Mediating variables included receipt of key services.  Analysis included both bivariate and multivariate regression.

 Results:  Findings suggest that clients who were in programs reflecting the Housing First philosophy got more supportive services and were more likely to be housed at the conclusion of the study.  They also experienced fewer days of homelessness.  Families had better outcomes than individuals, but there was no clear evidence of variation in outcomes related to things like mental or physical health problems or disabilities.  

 Implications: This is the first completed evaluation of any major city’s Plan to End homelessness.  Also, there have been no systematic evaluations of Housing First approaches on the scale of the present study. The study is important because the results suggest that Housing First models, particularly interim housing programs that provide housing for up to 120 days and also include service rich environments, may be especially effective in ameliorating homelessness.  At the same time, there are lessons to be learned about problems with implementation, fragmentation and access.  Such information can be important to policy makers and advocates as they plan and modify systems of service to the homeless in their own cities.