The Trump administration is facing roadblocks in its efforts to alter the criteria by which the Department of Housing and Urban Development (HUD) awards homelessness grants. U.S. District Judge Mary McElroy—a Trump nominee—ruled that the administration’s “slapdash imposition of political whims” is illegal. An appeals court upheld the decision the following day.
The controversy was not about policy, but about bureaucratic procedure, which the Administrative Procedure Act regulates. Yet the appellate decision took a detour from procedural considerations to defend the prevailing homelessness policy, known as “Housing First.” This ideological digression by U.S. Circuit Court Judge Julie Rikelman reflects widely held fallacies about Housing First.
Housing First refers to a homelessness strategy that prioritizes immediate housing with no strings attached. By contrast, the Clinton-era strategy—often referred to as “treatment first”—required people to maintain sobriety and participate in programs designed to promote self-sufficiency. Treatment first was considered successful when clients attained independence, while Housing First measures success by the lower standard of “housing stability” with ongoing subsidies.
Housing First became official federal policy in 2013 after the 2009 HEARTH Act took effect. HUD enforces Housing First by withholding funds from organizations that pursue alternative models of care, such as providing sober living facilities for recovering drug addicts. The Trump administration’s new funding criteria promised a return to a more treatment-oriented approach.
Yet Judge Rikelman insisted that Housing First is an “evidence-based” solution for homelessness that “has proven effective.” This language is repeated ad nauseam in pro-Housing First literature. The claim that Housing First is an “evidence-based strategy” has essentially become a religious creed, obediently parroted by the faithful.
The problem is that the Church of Housing First is highly selective about what “evidence” it considers valid. Judge Rikelman references the original Housing First experiment, conducted in the 1990s, which found that 88 percent of Housing First clients remained stably housed after five years (in subsidized housing), compared to 47 percent of the treatment-first group (without subsidies). However, she neglects to mention that the experiment only accepted homeless persons with diagnosed mental disorders, such as schizophrenia—in other words, people who are far less likely to achieve self-sufficiency under any program.
Experimental case studies provide the “evidence-based” claims about Housing First, but it is frankly inappropriate to use them for policy analysis. This is because they suffer from an inherent sample-bias problem, in that they only include people who are accepted into the program. They fail to account for the thousands of desperate souls who, for want of resources, are left to languish on housing waitlists until they are deemed sufficiently vulnerable, according to dystopian “vulnerability assessment tests” administered by outreach workers.
With this in mind, it is easier to recognize why Housing First case studies report dramatic successes, yet aggregated homelessness data tells a story of abject failure. Over the past decade, homelessness has skyrocketed by 40 percent, even as homelessness spending has ballooned.
Housing First, with its emphasis on permanent subsidies, is enormously expensive, and the lion’s share of the budget is consumed by a small minority of beneficiaries who have already been housed. San Francisco—which adopted Housing First in 2009—is illustrative, with 60 percent of its massive homelessness budget being spent on housing subsidies for the formerly homeless.
Although Housing First may boast high rates of housing retention among those who are placed (as reflected in the case studies), the resource demands prevent most of the homeless population from receiving any help at all (as reflected in the national data). Treatment first may have had a lower success rate in percentage terms, but because it helped free clients from dependence on taxpayer support, it could reach far more people, producing better outcomes overall.
While the Trump administration may have violated procedural rules when it altered funding criteria, the decision to move toward a treatment-based model is far more of an “evidence-based” policy than Housing First ever was.
Christopher Calton, Ph.D., is the research fellow in housing and homelessness with the Independent Institute in Oakland, Calif., and a contributor to the book Beyond Homeless: Good Intentions, Bad Outcomes, Transformative Solutions.