Housing First

What the evidence says

On its main outcome, keeping people housed, the evidence is strong. A 2019 meta-analysis of four randomized trials in Canada and the United States, with about 3,410 adults, found that Housing First participants were about two and a half times as likely to be stably housed after 18 to 24 months (risk ratio 2.46). A 2018 Campbell review of 43 randomized trials of homelessness programs, 37 of them in the United States, found that housing without an abstinence requirement probably adds about 110 days of stable housing compared with usual services, with moderate certainty.

Canada’s At Home/Chez Soi trial ran in several cities. Among people with high support needs, Housing First participants spent 71 percent of their time in stable housing, compared with 29 percent of those receiving usual services. For people with moderate needs, rent supplements plus intensive case management also increased housing stability over two years.

Other outcomes are less clear. The 2019 meta-analysis found no clear differences in mental health, physical health or substance use, though problem substance use fell in both groups. Housing First participants made fewer emergency department visits and had fewer hospital admissions. The At Home/Chez Soi trial for people with moderate needs found no difference in quality of life. All four trials in the meta-analysis were judged at high risk of bias, and follow-up was limited to about two years.

The grade is strong for housing stability. Reviewers flag risks of bias, such as poor reporting and the lack of blinding, but the housing effect is large and consistent across trials. For health and well-being, the evidence is weak or mixed. Charitology has not found a reliable published cost per outcome that would fit the Impact Calculator.

Where it sits on the depth scale

Housing First as tested is repair (depth 1). It restores a home to a person who has lost one, quickly and without preconditions. It doesn’t change the housing supply, prices or other rules that produce homelessness. When a government adopts Housing First as the organizing rule of its whole homelessness system, the policy acts closer to structure, but that is a different claim from what the trials measured.

Open questions

• What happens after two years? Few trials follow people longer.
• Most of the evidence comes from North America. How well does it travel?
• Housing First keeps people housed while it is funded. What would it take to stop people losing homes in the first place? That is the Symptom Trap in a single program, and part of the government question.