Youth, Crime, and the Potential Cost Offset to Housing First Programs

Abstract

Housing First (HF) is an approach that emphasizes providing housing as a precondition for assisting people experiencing homelessness. To the extent that housing reduces contacts with police, HF may reduce criminal behaviour and so reduce costs borne by the justice system. This may be particularly true for youth whose homelessness often forces them to adopt survival behaviour that exposes them to police and bylaw enforcement officers. Using regression analysis, we employ linked administrative data sets from police and from HF programs to examine how interactions of youth with police change following admission to a HF program. An important contribution of our study is the use of administrative police records rather than self-reported data on the number of criminal incidents and their severity. Unconditional quantile regression is used to observe HF’s effect on changes in both the number and severity of criminal incidents. Controlling for demographic characteristics of youth and for type of housing program and using administrative police records as opposed to self-reported police interactions, we find only weak evidence to suggest that the number of criminal incidents falls following admission to a HF program and only weak evidence of a fall in the seriousness of crimes.