Targeted support services can help cut costly use of homeless shelters

‘Our findings suggest that housing and support services targeting the episodic and long stay users are likely to be the most efficient strategy for reducing the shelter population in Canadian cities.’

By Tim Aubry

Twenty years ago, Randy Kuhn and Dennis Culhane published the results of a groundbreaking study in the American Journal of Community Psychology showing that over 80% of people who use emergency shelters in large American cities experienced homelessness for short periods of time and most frequently as a one-time event in their lives. In contrast, a relatively small minority demonstrated a cyclical use of shelters with multiple episodes for short periods of time or a small number of stays but for long periods of time. Although these latter two groups were relatively small making up 10% each of the shelter population, they were shown to occupy over 60% of the shelter beds.

Recently, I collaborated with a group of colleagues (Susan Farrell, Stephen Hwang, and Melissa Calhoun) from the University of Ottawa and St. Michaels Hospital on a study that evaluated four years of shelter data from three Canadian cities of different sizes, namely Toronto, Ottawa, and Guelph (this study replicated the analyses of the influential Kuhn and Culhane study). In line with the results of the American study, our research identified three distinct and similar patterns of shelter stays in the three Ontario cities.

The largest group of shelter users, characterized as “temporary”, experienced a small number of homeless episodes for relatively short periods of time. This group comprised 88% of shelter users in both Toronto and Ottawa. We also found a smaller group of users that we defined as having “episodic” shelter use, who experienced multiple homeless episodes for short periods of time. This group represented 11% and 9% of shelter users in Toronto and Ottawa respectively. Finally, the smallest group of shelter users, representing 4% of shelter users in Toronto and 2% in Ottawa, had a relatively small number of homeless episodes but for long periods of time.

The two smaller groups of shelter users occupied over one half of the shelter beds in both Toronto and Ottawa. Similar to the American research, our findings suggest that housing and support services targeting the episodic and long stay users are likely to be the most efficient strategy for reducing the shelter population in Canadian cities. A paper on the findings of our research has been published in Housing Studies. The paper is available online here.