Housing First is the dominant approach to homelessness policy in America. Housing First proponents believe that permanent housing, subsidized by the government, is the only legitimate solution to homelessness, and that housing programs should not impose preconditions such as sobriety, work, or participation in social services. The United States spends billions of dollars annually on homeless services, but the federal government tends to support only those programs that operate under Housing First principles.

A recent article in Vox by the journalist Rachel Cohen acknowledged Housing First’s vast influence over homelessness policy but worries that that could change, thanks to recent criticism, including criticism published in City Journal. The article provides a competent, if conventional, restatement of the case for Housing First. But it fails to answer critics’ core objection, which is that Housing First is more dominant than it deserves to be.

Housing First’s empire rests on its reputation as having been vindicated by science. Its rise is a case study in the use and abuse of social science in policymaking and offers several lessons relevant to many other debates. 

Ignore flows. Housing First studies examine the application of programs to a set number of currently homeless people, not how a system must respond to a dynamic situation—people flowing constantly into homelessness, which is, of course, the task in the real world. In other words, the studies consider stocks rather than flows, yet flows are the whole problem. How can a government program keep up with the thousands falling into homelessness every year? Will policymakers simply project how many there will be and then plan on building a unit for each of them every year, forever? Guaranteeing permanent housing to everyone who falls into homelessness is sure to attract more people into homelessness.

De-emphasize fadeout. The Housing First literature shows that, over time, many people who get Housing First–style housing leave it or die, and that many people who don’t get it eventually do find their way out of homelessness. The longer the study frame stretches, the more pronounced this effect becomes. The treatment group’s outcomes look more impressive than the control group’s at one to two years than they do at five to ten years and beyond.

Scale through simplification. In a straightforward sense, Housing First’s inability to operate at scale is its most obvious failing. But in a messaging sense, it has done so very successfully. Social science scales through simplification: “This is a cheap and simple way to end homelessness.” Simplification has a populist thrust, as scientists arrive to affirm the layman’s simple, non-scientific understanding of homelessness. As for Housing First’s opponents, well, why would they prefer complexity to simplicity? They must be bad people.

Root causes don’t matter, except when they’re all that matters. As a form of harm reduction, Housing First denies that root causes can be addressed. Homelessness is homelessness, the argument goes; don’t let anyone tell you it’s mental illness. This stands in contrast to the progressive public-safety view that crime is not crime, but the product of other factors.

The epistemological two-step. Social scientists are typically cautious in their peer-reviewed studies but less so when they enter the public debate. In the Housing First context, they often become unscientifically suggestive that Housing First could achieve something beyond warehousing people when programs are well managed and supported. Recourse to qualifications about management and appropriate support also happens when proponents have to explain why a program fails. But arguing that Housing First can work only under certain conditions is the opposite of arguing that because Housing First works everywhere, it, and only it, must be implemented.

The only principled defense Housing First imperialists have left is inadequate funding. We know this works, we just don’t know how to fund it, they say. But money is not the only relevant resource. There are also human resources: some communities are more likely to have good people who can run sober housing than good people who can run low-barrier housing. To accommodate that reality, the federal government would have to weaken Housing First’s influence. 

The economist and blogger Arnold Kling recently criticized a scholar in an unrelated situation for “using ‘studies’ like a lawyer to make a case rather than like a judge to try to figure out a complex situation.” Few social-policy areas are more complex than homelessness. Participants in the homelessness debate could stand to be more judicious in their use of social science and dispense with the faux-lawyerly attitude that a given study proves everything. Translated into policy, that judiciousness would provide more intellectual respectability, and much more government funding, for work- and sobriety-oriented programs for homeless people.

Photo by Mel Melcon / Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next